diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,86158 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12304,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.273040591428128,
+      "learning_rate": 5.405405405405406e-08,
+      "loss": 0.8615,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.14660204198603,
+      "learning_rate": 1.0810810810810812e-07,
+      "loss": 0.8351,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5136363685645717,
+      "learning_rate": 1.6216216216216218e-07,
+      "loss": 0.7338,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.84314659649236,
+      "learning_rate": 2.1621621621621625e-07,
+      "loss": 0.3097,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.3116383284825615,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.9433,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.151887895391798,
+      "learning_rate": 3.2432432432432436e-07,
+      "loss": 0.8352,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.359720301871594,
+      "learning_rate": 3.7837837837837843e-07,
+      "loss": 0.8869,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.069900304039918,
+      "learning_rate": 4.324324324324325e-07,
+      "loss": 0.8074,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.221531701629939,
+      "learning_rate": 4.864864864864865e-07,
+      "loss": 0.8024,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.223144226951507,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.793,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.3223820324853826,
+      "learning_rate": 5.945945945945947e-07,
+      "loss": 0.9239,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8854492475804197,
+      "learning_rate": 6.486486486486487e-07,
+      "loss": 0.7622,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5468659118607886,
+      "learning_rate": 7.027027027027028e-07,
+      "loss": 0.8072,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.239336624444215,
+      "learning_rate": 7.567567567567569e-07,
+      "loss": 0.8146,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.476905400596688,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.7157,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8515112912848704,
+      "learning_rate": 8.64864864864865e-07,
+      "loss": 0.6896,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.363378808068876,
+      "learning_rate": 9.189189189189191e-07,
+      "loss": 0.7681,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.646198456699034,
+      "learning_rate": 9.72972972972973e-07,
+      "loss": 0.8277,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.958244139103593,
+      "learning_rate": 1.027027027027027e-06,
+      "loss": 0.707,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4628987030897243,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.8058,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0814254976956748,
+      "learning_rate": 1.1351351351351352e-06,
+      "loss": 0.796,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.134834005763383,
+      "learning_rate": 1.1891891891891893e-06,
+      "loss": 0.7277,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9432058635310248,
+      "learning_rate": 1.2432432432432434e-06,
+      "loss": 0.6734,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2219674117937425,
+      "learning_rate": 1.2972972972972974e-06,
+      "loss": 0.7524,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6618761393680126,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.6931,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7996160383043525,
+      "learning_rate": 1.4054054054054056e-06,
+      "loss": 0.7208,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7728768994451718,
+      "learning_rate": 1.4594594594594596e-06,
+      "loss": 0.6902,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8967993526053564,
+      "learning_rate": 1.5135135135135137e-06,
+      "loss": 0.7391,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7443476331916143,
+      "learning_rate": 1.5675675675675678e-06,
+      "loss": 0.7357,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.789052299834985,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.6854,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.584750261961582,
+      "learning_rate": 1.675675675675676e-06,
+      "loss": 0.6375,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.789236566080917,
+      "learning_rate": 1.72972972972973e-06,
+      "loss": 0.7515,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5893620129109587,
+      "learning_rate": 1.783783783783784e-06,
+      "loss": 0.6829,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7943146599443252,
+      "learning_rate": 1.8378378378378381e-06,
+      "loss": 0.7003,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.57691567184717,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.6645,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5392384888398358,
+      "learning_rate": 1.945945945945946e-06,
+      "loss": 0.7122,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8831190046317889,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.3036,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3157264544284732,
+      "learning_rate": 2.054054054054054e-06,
+      "loss": 0.6175,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.432915059081968,
+      "learning_rate": 2.1081081081081085e-06,
+      "loss": 0.7022,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4613657485470233,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.6661,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5547873171988644,
+      "learning_rate": 2.2162162162162166e-06,
+      "loss": 0.6233,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6877086684917586,
+      "learning_rate": 2.2702702702702705e-06,
+      "loss": 0.6675,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7184407982106682,
+      "learning_rate": 2.3243243243243247e-06,
+      "loss": 0.7379,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7731563332171092,
+      "learning_rate": 2.3783783783783786e-06,
+      "loss": 0.6577,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.801920902456086,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.7479,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6573379957461731,
+      "learning_rate": 2.4864864864864867e-06,
+      "loss": 0.6584,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4084165272724114,
+      "learning_rate": 2.540540540540541e-06,
+      "loss": 0.6129,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6525233813389781,
+      "learning_rate": 2.594594594594595e-06,
+      "loss": 0.704,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3782860124442546,
+      "learning_rate": 2.648648648648649e-06,
+      "loss": 0.6154,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4337365267953683,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.6052,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2702372133594575,
+      "learning_rate": 2.7567567567567573e-06,
+      "loss": 0.5269,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2788406696387828,
+      "learning_rate": 2.810810810810811e-06,
+      "loss": 0.559,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5336812095086203,
+      "learning_rate": 2.8648648648648654e-06,
+      "loss": 0.688,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3528711974952585,
+      "learning_rate": 2.9189189189189193e-06,
+      "loss": 0.6331,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.333332096964607,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.6096,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4784282183701558,
+      "learning_rate": 3.0270270270270274e-06,
+      "loss": 0.6252,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4658057231898225,
+      "learning_rate": 3.0810810810810817e-06,
+      "loss": 0.6609,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4561817591061057,
+      "learning_rate": 3.1351351351351356e-06,
+      "loss": 0.6048,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.355095612044604,
+      "learning_rate": 3.1891891891891894e-06,
+      "loss": 0.5902,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3067364183441053,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.5688,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3751486378946132,
+      "learning_rate": 3.2972972972972976e-06,
+      "loss": 0.6159,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2862193884453008,
+      "learning_rate": 3.351351351351352e-06,
+      "loss": 0.593,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4789049778412198,
+      "learning_rate": 3.4054054054054057e-06,
+      "loss": 0.6427,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4546804071157595,
+      "learning_rate": 3.45945945945946e-06,
+      "loss": 0.632,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4395735150129738,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.6129,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4346828370676308,
+      "learning_rate": 3.567567567567568e-06,
+      "loss": 0.6517,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4377476761384835,
+      "learning_rate": 3.621621621621622e-06,
+      "loss": 0.6353,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8803533563697833,
+      "learning_rate": 3.6756756756756763e-06,
+      "loss": 0.3199,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3957121329522453,
+      "learning_rate": 3.72972972972973e-06,
+      "loss": 0.6205,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4125486289001528,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.6586,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.431189021990164,
+      "learning_rate": 3.837837837837838e-06,
+      "loss": 0.6508,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2515213521059998,
+      "learning_rate": 3.891891891891892e-06,
+      "loss": 0.5511,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4889319759544897,
+      "learning_rate": 3.945945945945947e-06,
+      "loss": 0.6802,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3850309722543301,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5624,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5352306040196768,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.6158,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.306638038154721,
+      "learning_rate": 4.108108108108108e-06,
+      "loss": 0.5853,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4703554455302463,
+      "learning_rate": 4.162162162162163e-06,
+      "loss": 0.6308,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3157754223842182,
+      "learning_rate": 4.216216216216217e-06,
+      "loss": 0.6677,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2784647086503598,
+      "learning_rate": 4.270270270270271e-06,
+      "loss": 0.5917,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4109447882750359,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.635,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.793412137191073,
+      "learning_rate": 4.378378378378379e-06,
+      "loss": 0.6547,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4467943130716234,
+      "learning_rate": 4.432432432432433e-06,
+      "loss": 0.6441,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5009537855152142,
+      "learning_rate": 4.486486486486487e-06,
+      "loss": 0.6814,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.298628336993122,
+      "learning_rate": 4.540540540540541e-06,
+      "loss": 0.6049,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.599083383122989,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.7032,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3744837239712366,
+      "learning_rate": 4.6486486486486495e-06,
+      "loss": 0.5911,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4225318479566327,
+      "learning_rate": 4.702702702702703e-06,
+      "loss": 0.5956,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3958885743685878,
+      "learning_rate": 4.756756756756757e-06,
+      "loss": 0.5795,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3249125605774177,
+      "learning_rate": 4.810810810810811e-06,
+      "loss": 0.5479,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3298464814569155,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.5754,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4683698015671058,
+      "learning_rate": 4.91891891891892e-06,
+      "loss": 0.5928,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3388444498030598,
+      "learning_rate": 4.9729729729729735e-06,
+      "loss": 0.6335,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3594399027949904,
+      "learning_rate": 5.027027027027027e-06,
+      "loss": 0.602,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8636527147025339,
+      "learning_rate": 5.081081081081082e-06,
+      "loss": 0.3149,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2168309642609405,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.5471,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2351839782392957,
+      "learning_rate": 5.18918918918919e-06,
+      "loss": 0.5424,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.292604070171166,
+      "learning_rate": 5.243243243243244e-06,
+      "loss": 0.5823,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.311816649798154,
+      "learning_rate": 5.297297297297298e-06,
+      "loss": 0.6077,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8143613433965735,
+      "learning_rate": 5.351351351351351e-06,
+      "loss": 0.3432,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3116387212090366,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.6164,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3535109082047654,
+      "learning_rate": 5.45945945945946e-06,
+      "loss": 0.6392,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3443165530744305,
+      "learning_rate": 5.513513513513515e-06,
+      "loss": 0.568,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.414863407793237,
+      "learning_rate": 5.567567567567568e-06,
+      "loss": 0.6686,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3009721885803063,
+      "learning_rate": 5.621621621621622e-06,
+      "loss": 0.5804,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2774240010244946,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.6211,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8344894256527007,
+      "learning_rate": 5.729729729729731e-06,
+      "loss": 0.3314,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3879213164427744,
+      "learning_rate": 5.783783783783784e-06,
+      "loss": 0.656,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.375044697622862,
+      "learning_rate": 5.837837837837839e-06,
+      "loss": 0.5651,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4877329295170625,
+      "learning_rate": 5.8918918918918924e-06,
+      "loss": 0.6075,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2983195899347304,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.5463,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2218725662034848,
+      "learning_rate": 6e-06,
+      "loss": 0.6036,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3436025991798024,
+      "learning_rate": 6.054054054054055e-06,
+      "loss": 0.6227,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2977100495834533,
+      "learning_rate": 6.108108108108109e-06,
+      "loss": 0.5586,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3507646576725383,
+      "learning_rate": 6.162162162162163e-06,
+      "loss": 0.5956,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.399931221889171,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.6347,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4071118149538326,
+      "learning_rate": 6.270270270270271e-06,
+      "loss": 0.6294,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5751588266470815,
+      "learning_rate": 6.324324324324325e-06,
+      "loss": 0.5842,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5903449377258876,
+      "learning_rate": 6.378378378378379e-06,
+      "loss": 0.6512,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6710624509177674,
+      "learning_rate": 6.432432432432433e-06,
+      "loss": 0.5931,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1494354077899656,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.5491,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3044888172319544,
+      "learning_rate": 6.540540540540541e-06,
+      "loss": 0.615,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.268346144656469,
+      "learning_rate": 6.594594594594595e-06,
+      "loss": 0.5698,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3160275141042568,
+      "learning_rate": 6.648648648648649e-06,
+      "loss": 0.5061,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.246078708770176,
+      "learning_rate": 6.702702702702704e-06,
+      "loss": 0.5999,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.427653128474058,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.6213,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.303477151542448,
+      "learning_rate": 6.810810810810811e-06,
+      "loss": 0.6192,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3710171885376965,
+      "learning_rate": 6.864864864864865e-06,
+      "loss": 0.5732,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3631429826418746,
+      "learning_rate": 6.91891891891892e-06,
+      "loss": 0.5732,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4099354465501006,
+      "learning_rate": 6.972972972972973e-06,
+      "loss": 0.6389,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4281747107854714,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.6006,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4595708181113594,
+      "learning_rate": 7.0810810810810815e-06,
+      "loss": 0.6472,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.489743891332324,
+      "learning_rate": 7.135135135135136e-06,
+      "loss": 0.6437,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2738442793830842,
+      "learning_rate": 7.189189189189189e-06,
+      "loss": 0.5813,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6295009674146617,
+      "learning_rate": 7.243243243243244e-06,
+      "loss": 0.7095,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5620836204569906,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.5914,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5060790781694255,
+      "learning_rate": 7.3513513513513525e-06,
+      "loss": 0.6339,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.388913535992189,
+      "learning_rate": 7.4054054054054055e-06,
+      "loss": 0.6277,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4863516194848927,
+      "learning_rate": 7.45945945945946e-06,
+      "loss": 0.5646,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3445803127161098,
+      "learning_rate": 7.513513513513514e-06,
+      "loss": 0.6292,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3384584779109399,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.5435,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4926838153450348,
+      "learning_rate": 7.621621621621622e-06,
+      "loss": 0.6428,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3869777102360936,
+      "learning_rate": 7.675675675675676e-06,
+      "loss": 0.5966,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3409434328050722,
+      "learning_rate": 7.72972972972973e-06,
+      "loss": 0.6088,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2762128491212468,
+      "learning_rate": 7.783783783783784e-06,
+      "loss": 0.6029,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.314550003433995,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.5568,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4580077217392413,
+      "learning_rate": 7.891891891891894e-06,
+      "loss": 0.6383,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.460486772621834,
+      "learning_rate": 7.945945945945946e-06,
+      "loss": 0.6107,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3790848419120394,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.6367,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3986888028757734,
+      "learning_rate": 8.054054054054055e-06,
+      "loss": 0.5709,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5066361422827574,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.5565,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.231260969143522,
+      "learning_rate": 8.162162162162163e-06,
+      "loss": 0.5974,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3662141363352507,
+      "learning_rate": 8.216216216216217e-06,
+      "loss": 0.5617,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.379757692211869,
+      "learning_rate": 8.27027027027027e-06,
+      "loss": 0.6216,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4032029868039735,
+      "learning_rate": 8.324324324324326e-06,
+      "loss": 0.5659,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.246102895886861,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.6069,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3129486277734963,
+      "learning_rate": 8.432432432432434e-06,
+      "loss": 0.5649,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5949807239779763,
+      "learning_rate": 8.486486486486488e-06,
+      "loss": 0.6402,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2367723511070507,
+      "learning_rate": 8.540540540540542e-06,
+      "loss": 0.5507,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3732449159415614,
+      "learning_rate": 8.594594594594595e-06,
+      "loss": 0.6039,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3522671789714542,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.583,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3857986224694943,
+      "learning_rate": 8.702702702702703e-06,
+      "loss": 0.6238,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6041881805697953,
+      "learning_rate": 8.756756756756759e-06,
+      "loss": 0.5787,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3844553479323647,
+      "learning_rate": 8.810810810810811e-06,
+      "loss": 0.6455,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.372775499931938,
+      "learning_rate": 8.864864864864866e-06,
+      "loss": 0.5469,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3096092196718552,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.6066,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3873598772944997,
+      "learning_rate": 8.972972972972974e-06,
+      "loss": 0.6354,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4261461525288062,
+      "learning_rate": 9.027027027027028e-06,
+      "loss": 0.5119,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3805251800039318,
+      "learning_rate": 9.081081081081082e-06,
+      "loss": 0.5846,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8240477026882622,
+      "learning_rate": 9.135135135135136e-06,
+      "loss": 0.3422,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3817497767181797,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.5916,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4599858495694384,
+      "learning_rate": 9.243243243243243e-06,
+      "loss": 0.5474,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3796151016083495,
+      "learning_rate": 9.297297297297299e-06,
+      "loss": 0.5706,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.409777572587561,
+      "learning_rate": 9.351351351351353e-06,
+      "loss": 0.6052,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4002947829604957,
+      "learning_rate": 9.405405405405407e-06,
+      "loss": 0.5698,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.405204434110899,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.596,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.462514585076768,
+      "learning_rate": 9.513513513513514e-06,
+      "loss": 0.62,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7499683079184325,
+      "learning_rate": 9.567567567567568e-06,
+      "loss": 0.5936,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4505838663306836,
+      "learning_rate": 9.621621621621622e-06,
+      "loss": 0.6323,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2193208841128076,
+      "learning_rate": 9.675675675675676e-06,
+      "loss": 0.5516,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5461709008460212,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.6565,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3783256914341504,
+      "learning_rate": 9.783783783783785e-06,
+      "loss": 0.6273,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3009129274113025,
+      "learning_rate": 9.83783783783784e-06,
+      "loss": 0.5709,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4331042802705694,
+      "learning_rate": 9.891891891891893e-06,
+      "loss": 0.6054,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6637836990959596,
+      "learning_rate": 9.945945945945947e-06,
+      "loss": 0.5736,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4268744416818444,
+      "learning_rate": 1e-05,
+      "loss": 0.5935,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.325262107647321,
+      "learning_rate": 1.0054054054054055e-05,
+      "loss": 0.5971,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.477582803403984,
+      "learning_rate": 1.0108108108108109e-05,
+      "loss": 0.6255,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3006442007786023,
+      "learning_rate": 1.0162162162162164e-05,
+      "loss": 0.5756,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3884979600616545,
+      "learning_rate": 1.0216216216216216e-05,
+      "loss": 0.6581,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2998097970689109,
+      "learning_rate": 1.027027027027027e-05,
+      "loss": 0.5781,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3950495411925943,
+      "learning_rate": 1.0324324324324324e-05,
+      "loss": 0.5968,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.36884172166217,
+      "learning_rate": 1.037837837837838e-05,
+      "loss": 0.5746,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3242805937768347,
+      "learning_rate": 1.0432432432432433e-05,
+      "loss": 0.6482,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.387103904758459,
+      "learning_rate": 1.0486486486486487e-05,
+      "loss": 0.5908,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3226413426613004,
+      "learning_rate": 1.0540540540540541e-05,
+      "loss": 0.5772,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4041121660539917,
+      "learning_rate": 1.0594594594594597e-05,
+      "loss": 0.5884,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4683717432308583,
+      "learning_rate": 1.0648648648648649e-05,
+      "loss": 0.6196,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3287374075081768,
+      "learning_rate": 1.0702702702702703e-05,
+      "loss": 0.5308,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.31673514609701,
+      "learning_rate": 1.0756756756756757e-05,
+      "loss": 0.5923,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2248739732041203,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 0.5463,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4582091435807332,
+      "learning_rate": 1.0864864864864866e-05,
+      "loss": 0.5624,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3974633462736783,
+      "learning_rate": 1.091891891891892e-05,
+      "loss": 0.5518,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3098255750598122,
+      "learning_rate": 1.0972972972972974e-05,
+      "loss": 0.5978,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4020937292739093,
+      "learning_rate": 1.102702702702703e-05,
+      "loss": 0.6133,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3006764141808833,
+      "learning_rate": 1.1081081081081081e-05,
+      "loss": 0.5437,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3121807130049612,
+      "learning_rate": 1.1135135135135135e-05,
+      "loss": 0.5422,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1807616133678631,
+      "learning_rate": 1.1189189189189189e-05,
+      "loss": 0.551,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.411361256781282,
+      "learning_rate": 1.1243243243243245e-05,
+      "loss": 0.594,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4437358882210654,
+      "learning_rate": 1.1297297297297298e-05,
+      "loss": 0.6544,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.359156252589416,
+      "learning_rate": 1.1351351351351352e-05,
+      "loss": 0.6272,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.29652118501433,
+      "learning_rate": 1.1405405405405404e-05,
+      "loss": 0.6323,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3097621304296205,
+      "learning_rate": 1.1459459459459462e-05,
+      "loss": 0.6287,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2169668382664947,
+      "learning_rate": 1.1513513513513514e-05,
+      "loss": 0.5423,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4669873874010895,
+      "learning_rate": 1.1567567567567568e-05,
+      "loss": 0.6193,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2954066617571018,
+      "learning_rate": 1.1621621621621622e-05,
+      "loss": 0.5883,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.160631713695103,
+      "learning_rate": 1.1675675675675677e-05,
+      "loss": 0.5929,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.34874266847612,
+      "learning_rate": 1.1729729729729731e-05,
+      "loss": 0.6387,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.439317708700643,
+      "learning_rate": 1.1783783783783785e-05,
+      "loss": 0.5649,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.502437313426451,
+      "learning_rate": 1.1837837837837837e-05,
+      "loss": 0.6514,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3337609956026828,
+      "learning_rate": 1.1891891891891894e-05,
+      "loss": 0.5162,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2481859007998901,
+      "learning_rate": 1.1945945945945946e-05,
+      "loss": 0.6038,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3067393315538185,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5858,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5757684424914125,
+      "learning_rate": 1.2054054054054054e-05,
+      "loss": 0.6355,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5277737917360357,
+      "learning_rate": 1.210810810810811e-05,
+      "loss": 0.6498,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3038105939520328,
+      "learning_rate": 1.2162162162162164e-05,
+      "loss": 0.5666,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3439900623746683,
+      "learning_rate": 1.2216216216216217e-05,
+      "loss": 0.6461,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4499492648511714,
+      "learning_rate": 1.227027027027027e-05,
+      "loss": 0.5793,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1895262665587139,
+      "learning_rate": 1.2324324324324327e-05,
+      "loss": 0.615,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4385484255224155,
+      "learning_rate": 1.2378378378378379e-05,
+      "loss": 0.5863,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2568226128615418,
+      "learning_rate": 1.2432432432432433e-05,
+      "loss": 0.5336,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4055308076396922,
+      "learning_rate": 1.2486486486486487e-05,
+      "loss": 0.5927,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3604519277780607,
+      "learning_rate": 1.2540540540540542e-05,
+      "loss": 0.6365,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3976912250958944,
+      "learning_rate": 1.2594594594594596e-05,
+      "loss": 0.6016,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3870555370772695,
+      "learning_rate": 1.264864864864865e-05,
+      "loss": 0.6367,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1506464171651247,
+      "learning_rate": 1.2702702702702702e-05,
+      "loss": 0.3983,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.105610354688429,
+      "learning_rate": 1.2756756756756758e-05,
+      "loss": 0.4861,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2922528747586577,
+      "learning_rate": 1.2810810810810812e-05,
+      "loss": 0.5429,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4117421331128992,
+      "learning_rate": 1.2864864864864865e-05,
+      "loss": 0.7149,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1561171907937822,
+      "learning_rate": 1.291891891891892e-05,
+      "loss": 0.5846,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2823597420925605,
+      "learning_rate": 1.2972972972972975e-05,
+      "loss": 0.6354,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.331199921173987,
+      "learning_rate": 1.3027027027027029e-05,
+      "loss": 0.6929,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3632982831874763,
+      "learning_rate": 1.3081081081081083e-05,
+      "loss": 0.6142,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.30032739139169,
+      "learning_rate": 1.3135135135135135e-05,
+      "loss": 0.6053,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2669107246418034,
+      "learning_rate": 1.318918918918919e-05,
+      "loss": 0.6165,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4024543498429707,
+      "learning_rate": 1.3243243243243244e-05,
+      "loss": 0.5811,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2282938927207665,
+      "learning_rate": 1.3297297297297298e-05,
+      "loss": 0.5824,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2230011147853819,
+      "learning_rate": 1.3351351351351352e-05,
+      "loss": 0.5773,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4405143504114788,
+      "learning_rate": 1.3405405405405407e-05,
+      "loss": 0.5856,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.407496997500501,
+      "learning_rate": 1.3459459459459461e-05,
+      "loss": 0.6758,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3481381140295823,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 0.6091,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3772848449020698,
+      "learning_rate": 1.3567567567567567e-05,
+      "loss": 0.6378,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3399199799458354,
+      "learning_rate": 1.3621621621621623e-05,
+      "loss": 0.6966,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.418137767899028,
+      "learning_rate": 1.3675675675675677e-05,
+      "loss": 0.6543,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3872559456216722,
+      "learning_rate": 1.372972972972973e-05,
+      "loss": 0.6297,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4057293135350568,
+      "learning_rate": 1.3783783783783784e-05,
+      "loss": 0.6444,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2832126600113147,
+      "learning_rate": 1.383783783783784e-05,
+      "loss": 0.5735,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1665636697135973,
+      "learning_rate": 1.3891891891891894e-05,
+      "loss": 0.606,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2594265086154244,
+      "learning_rate": 1.3945945945945946e-05,
+      "loss": 0.5554,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2758336481579695,
+      "learning_rate": 1.4e-05,
+      "loss": 0.5898,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.300264913132061,
+      "learning_rate": 1.4054054054054055e-05,
+      "loss": 0.6402,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3315419378572004,
+      "learning_rate": 1.410810810810811e-05,
+      "loss": 0.548,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.219427993604917,
+      "learning_rate": 1.4162162162162163e-05,
+      "loss": 0.5622,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1897819364789977,
+      "learning_rate": 1.4216216216216217e-05,
+      "loss": 0.6132,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.333457147167597,
+      "learning_rate": 1.4270270270270272e-05,
+      "loss": 0.6269,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1467620301405113,
+      "learning_rate": 1.4324324324324326e-05,
+      "loss": 0.5014,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2858858581704378,
+      "learning_rate": 1.4378378378378378e-05,
+      "loss": 0.6116,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.425258049531029,
+      "learning_rate": 1.4432432432432432e-05,
+      "loss": 0.5601,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1845124721882538,
+      "learning_rate": 1.4486486486486488e-05,
+      "loss": 0.5994,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2324371046729103,
+      "learning_rate": 1.4540540540540542e-05,
+      "loss": 0.5734,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2291795518358182,
+      "learning_rate": 1.4594594594594596e-05,
+      "loss": 0.5702,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2848681461436653,
+      "learning_rate": 1.464864864864865e-05,
+      "loss": 0.6426,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3483627165006886,
+      "learning_rate": 1.4702702702702705e-05,
+      "loss": 0.6642,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4277089296969299,
+      "learning_rate": 1.4756756756756759e-05,
+      "loss": 0.5671,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3093511771695634,
+      "learning_rate": 1.4810810810810811e-05,
+      "loss": 0.5995,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2312759429949,
+      "learning_rate": 1.4864864864864865e-05,
+      "loss": 0.6446,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.398767031842918,
+      "learning_rate": 1.491891891891892e-05,
+      "loss": 0.6307,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2914932022710714,
+      "learning_rate": 1.4972972972972974e-05,
+      "loss": 0.5712,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2827552834296592,
+      "learning_rate": 1.5027027027027028e-05,
+      "loss": 0.6311,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.17125498981903,
+      "learning_rate": 1.5081081081081082e-05,
+      "loss": 0.5277,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2532330404418561,
+      "learning_rate": 1.5135135135135138e-05,
+      "loss": 0.6122,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1771467390533292,
+      "learning_rate": 1.5189189189189191e-05,
+      "loss": 0.5551,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3178622763909202,
+      "learning_rate": 1.5243243243243244e-05,
+      "loss": 0.6077,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2923685907272977,
+      "learning_rate": 1.5297297297297297e-05,
+      "loss": 0.5549,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2778488850748242,
+      "learning_rate": 1.5351351351351353e-05,
+      "loss": 0.6552,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8710331713669297,
+      "learning_rate": 1.540540540540541e-05,
+      "loss": 0.6568,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.288199139989406,
+      "learning_rate": 1.545945945945946e-05,
+      "loss": 0.5902,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.452547686281634,
+      "learning_rate": 1.5513513513513513e-05,
+      "loss": 0.591,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2416649407225506,
+      "learning_rate": 1.556756756756757e-05,
+      "loss": 0.6087,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.251564791733786,
+      "learning_rate": 1.5621621621621624e-05,
+      "loss": 0.6281,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2801045493742762,
+      "learning_rate": 1.5675675675675676e-05,
+      "loss": 0.5922,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2655122609805658,
+      "learning_rate": 1.572972972972973e-05,
+      "loss": 0.604,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1773901975659047,
+      "learning_rate": 1.5783783783783787e-05,
+      "loss": 0.5544,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1546196570757032,
+      "learning_rate": 1.583783783783784e-05,
+      "loss": 0.5362,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3076851584325124,
+      "learning_rate": 1.589189189189189e-05,
+      "loss": 0.5621,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2826279531334102,
+      "learning_rate": 1.5945945945945947e-05,
+      "loss": 0.5728,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.282623769972093,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.5627,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3408958446169106,
+      "learning_rate": 1.6054054054054055e-05,
+      "loss": 0.611,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2348334073578877,
+      "learning_rate": 1.610810810810811e-05,
+      "loss": 0.6068,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2975389815390468,
+      "learning_rate": 1.6162162162162163e-05,
+      "loss": 0.6597,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2096404489865318,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 0.5774,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1813743688009573,
+      "learning_rate": 1.6270270270270274e-05,
+      "loss": 0.5426,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1681240565989068,
+      "learning_rate": 1.6324324324324326e-05,
+      "loss": 0.5872,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3077141773660117,
+      "learning_rate": 1.6378378378378378e-05,
+      "loss": 0.5684,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.302528145958617,
+      "learning_rate": 1.6432432432432434e-05,
+      "loss": 0.6124,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0758705697342368,
+      "learning_rate": 1.648648648648649e-05,
+      "loss": 0.6116,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4131860721941674,
+      "learning_rate": 1.654054054054054e-05,
+      "loss": 0.609,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2851791287758079,
+      "learning_rate": 1.6594594594594597e-05,
+      "loss": 0.6528,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.366338400597577,
+      "learning_rate": 1.6648648648648652e-05,
+      "loss": 0.6228,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.497776907617428,
+      "learning_rate": 1.6702702702702704e-05,
+      "loss": 0.6945,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1602743781779306,
+      "learning_rate": 1.6756756756756757e-05,
+      "loss": 0.5984,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4018561667642373,
+      "learning_rate": 1.6810810810810812e-05,
+      "loss": 0.6094,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4040754275179608,
+      "learning_rate": 1.6864864864864868e-05,
+      "loss": 0.6365,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2110215218082543,
+      "learning_rate": 1.691891891891892e-05,
+      "loss": 0.5932,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2478904633212702,
+      "learning_rate": 1.6972972972972975e-05,
+      "loss": 0.5906,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.219674052199782,
+      "learning_rate": 1.7027027027027028e-05,
+      "loss": 0.5978,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1334469527342586,
+      "learning_rate": 1.7081081081081083e-05,
+      "loss": 0.3721,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2986498696320505,
+      "learning_rate": 1.7135135135135135e-05,
+      "loss": 0.6508,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.432617075086687,
+      "learning_rate": 1.718918918918919e-05,
+      "loss": 0.6134,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3291322605886884,
+      "learning_rate": 1.7243243243243243e-05,
+      "loss": 0.5531,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.250068899737447,
+      "learning_rate": 1.72972972972973e-05,
+      "loss": 0.6449,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2273279396721362,
+      "learning_rate": 1.7351351351351354e-05,
+      "loss": 0.63,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.288258537341939,
+      "learning_rate": 1.7405405405405406e-05,
+      "loss": 0.614,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.207319610961464,
+      "learning_rate": 1.745945945945946e-05,
+      "loss": 0.6219,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1602660996710994,
+      "learning_rate": 1.7513513513513517e-05,
+      "loss": 0.5795,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3070055487401167,
+      "learning_rate": 1.756756756756757e-05,
+      "loss": 0.7466,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1048493267116712,
+      "learning_rate": 1.7621621621621622e-05,
+      "loss": 0.5529,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2579563285152189,
+      "learning_rate": 1.7675675675675677e-05,
+      "loss": 0.6631,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3235404147237428,
+      "learning_rate": 1.7729729729729733e-05,
+      "loss": 0.6216,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1827495846841425,
+      "learning_rate": 1.7783783783783785e-05,
+      "loss": 0.6543,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2601381738043989,
+      "learning_rate": 1.783783783783784e-05,
+      "loss": 0.6269,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2430301530941215,
+      "learning_rate": 1.7891891891891893e-05,
+      "loss": 0.5677,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2692760414700273,
+      "learning_rate": 1.7945945945945948e-05,
+      "loss": 0.6412,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2001150074501838,
+      "learning_rate": 1.8e-05,
+      "loss": 0.5676,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3075487794158436,
+      "learning_rate": 1.8054054054054056e-05,
+      "loss": 0.6472,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.456133317177465,
+      "learning_rate": 1.8108108108108108e-05,
+      "loss": 0.6338,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1939888671400902,
+      "learning_rate": 1.8162162162162164e-05,
+      "loss": 0.6184,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1981881909785101,
+      "learning_rate": 1.821621621621622e-05,
+      "loss": 0.5323,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.20091981296182,
+      "learning_rate": 1.827027027027027e-05,
+      "loss": 0.618,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.41989949023168,
+      "learning_rate": 1.8324324324324324e-05,
+      "loss": 0.6336,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3139662867417505,
+      "learning_rate": 1.8378378378378383e-05,
+      "loss": 0.5965,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.333158999654752,
+      "learning_rate": 1.8432432432432435e-05,
+      "loss": 0.6264,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4531220569740995,
+      "learning_rate": 1.8486486486486487e-05,
+      "loss": 0.5439,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3131403063055023,
+      "learning_rate": 1.8540540540540542e-05,
+      "loss": 0.6107,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3701905301886614,
+      "learning_rate": 1.8594594594594598e-05,
+      "loss": 0.6123,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3653791912501723,
+      "learning_rate": 1.864864864864865e-05,
+      "loss": 0.6613,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1142972006714638,
+      "learning_rate": 1.8702702702702706e-05,
+      "loss": 0.5526,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4299690140309247,
+      "learning_rate": 1.8756756756756758e-05,
+      "loss": 0.6966,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2593444176121449,
+      "learning_rate": 1.8810810810810813e-05,
+      "loss": 0.5774,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1765124465821895,
+      "learning_rate": 1.8864864864864866e-05,
+      "loss": 0.6137,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3065008361798192,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 0.5976,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4087498976721118,
+      "learning_rate": 1.8972972972972973e-05,
+      "loss": 0.6289,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2341841476607793,
+      "learning_rate": 1.902702702702703e-05,
+      "loss": 0.6139,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.328079281513009,
+      "learning_rate": 1.9081081081081084e-05,
+      "loss": 0.6233,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2477408998174193,
+      "learning_rate": 1.9135135135135137e-05,
+      "loss": 0.6108,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2816918475775614,
+      "learning_rate": 1.918918918918919e-05,
+      "loss": 0.6245,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2016024948487902,
+      "learning_rate": 1.9243243243243244e-05,
+      "loss": 0.5771,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.217828445501427,
+      "learning_rate": 1.92972972972973e-05,
+      "loss": 0.578,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9731305095764395,
+      "learning_rate": 1.9351351351351352e-05,
+      "loss": 0.4656,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3567485383775508,
+      "learning_rate": 1.9405405405405408e-05,
+      "loss": 0.6423,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5361091360932055,
+      "learning_rate": 1.9459459459459463e-05,
+      "loss": 0.6508,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1645320502210965,
+      "learning_rate": 1.9513513513513515e-05,
+      "loss": 0.5934,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0958112327792273,
+      "learning_rate": 1.956756756756757e-05,
+      "loss": 0.575,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3740750676813598,
+      "learning_rate": 1.9621621621621623e-05,
+      "loss": 0.6114,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.262910997447116,
+      "learning_rate": 1.967567567567568e-05,
+      "loss": 0.6542,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3570416080679524,
+      "learning_rate": 1.972972972972973e-05,
+      "loss": 0.6802,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1127799363489526,
+      "learning_rate": 1.9783783783783786e-05,
+      "loss": 0.5859,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2463121588632178,
+      "learning_rate": 1.983783783783784e-05,
+      "loss": 0.6215,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2379252903825178,
+      "learning_rate": 1.9891891891891894e-05,
+      "loss": 0.6024,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1412956122460987,
+      "learning_rate": 1.994594594594595e-05,
+      "loss": 0.5442,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1359940245417317,
+      "learning_rate": 2e-05,
+      "loss": 0.6442,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1800301173417773,
+      "learning_rate": 1.9999999653504437e-05,
+      "loss": 0.6515,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2781092218781143,
+      "learning_rate": 1.9999998614017768e-05,
+      "loss": 0.6085,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.172788213355845,
+      "learning_rate": 1.999999688154006e-05,
+      "loss": 0.5846,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2016265117946634,
+      "learning_rate": 1.999999445607144e-05,
+      "loss": 0.6501,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1201465902325103,
+      "learning_rate": 1.9999991337612076e-05,
+      "loss": 0.6075,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.27326710987819,
+      "learning_rate": 1.9999987526162182e-05,
+      "loss": 0.7154,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2675834152794216,
+      "learning_rate": 1.9999983021722023e-05,
+      "loss": 0.6373,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3527705851821266,
+      "learning_rate": 1.999997782429191e-05,
+      "loss": 0.6465,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.258748188700293,
+      "learning_rate": 1.99999719338722e-05,
+      "loss": 0.6348,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.213322715499661,
+      "learning_rate": 1.999996535046331e-05,
+      "loss": 0.6418,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3321652569073361,
+      "learning_rate": 1.999995807406569e-05,
+      "loss": 0.722,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.245916973083892,
+      "learning_rate": 1.9999950104679847e-05,
+      "loss": 0.6258,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1799491737624703,
+      "learning_rate": 1.9999941442306328e-05,
+      "loss": 0.6077,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.187391126708189,
+      "learning_rate": 1.9999932086945735e-05,
+      "loss": 0.641,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2541712864332293,
+      "learning_rate": 1.9999922038598724e-05,
+      "loss": 0.5921,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.316135226338434,
+      "learning_rate": 1.9999911297265987e-05,
+      "loss": 0.5714,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2810307718824059,
+      "learning_rate": 1.999989986294826e-05,
+      "loss": 0.612,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2079749802503408,
+      "learning_rate": 1.999988773564635e-05,
+      "loss": 0.6461,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.221100438003557,
+      "learning_rate": 1.9999874915361083e-05,
+      "loss": 0.5395,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2321815633471245,
+      "learning_rate": 1.999986140209336e-05,
+      "loss": 0.626,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2602679527635106,
+      "learning_rate": 1.9999847195844104e-05,
+      "loss": 0.6293,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1795010152484604,
+      "learning_rate": 1.999983229661431e-05,
+      "loss": 0.5701,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2089494729842096,
+      "learning_rate": 1.999981670440501e-05,
+      "loss": 0.6006,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3261960086679885,
+      "learning_rate": 1.9999800419217285e-05,
+      "loss": 0.5857,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.258321316496881,
+      "learning_rate": 1.999978344105226e-05,
+      "loss": 0.6662,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1705754710200273,
+      "learning_rate": 1.9999765769911108e-05,
+      "loss": 0.647,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1891364177409702,
+      "learning_rate": 1.9999747405795057e-05,
+      "loss": 0.5929,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2102667886877647,
+      "learning_rate": 1.9999728348705386e-05,
+      "loss": 0.6013,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2641137026960239,
+      "learning_rate": 1.9999708598643405e-05,
+      "loss": 0.6163,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2937634579523318,
+      "learning_rate": 1.999968815561049e-05,
+      "loss": 0.6061,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3553502126852537,
+      "learning_rate": 1.9999667019608058e-05,
+      "loss": 0.6649,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1871908937053395,
+      "learning_rate": 1.999964519063757e-05,
+      "loss": 0.5435,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2660823944720954,
+      "learning_rate": 1.999962266870054e-05,
+      "loss": 0.6907,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2083772436641733,
+      "learning_rate": 1.9999599453798523e-05,
+      "loss": 0.6537,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1642945571775338,
+      "learning_rate": 1.999957554593314e-05,
+      "loss": 0.6076,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2599482071313974,
+      "learning_rate": 1.9999550945106038e-05,
+      "loss": 0.6471,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3562495151348162,
+      "learning_rate": 1.999952565131893e-05,
+      "loss": 0.7162,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0991869227445243,
+      "learning_rate": 1.999949966457356e-05,
+      "loss": 0.6373,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4761791840477894,
+      "learning_rate": 1.9999472984871734e-05,
+      "loss": 0.681,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2778912175636392,
+      "learning_rate": 1.99994456122153e-05,
+      "loss": 0.6335,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3259903218416293,
+      "learning_rate": 1.9999417546606153e-05,
+      "loss": 0.6059,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.30223892918819,
+      "learning_rate": 1.9999388788046238e-05,
+      "loss": 0.6614,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2694327949012507,
+      "learning_rate": 1.9999359336537552e-05,
+      "loss": 0.5949,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2617795135910619,
+      "learning_rate": 1.9999329192082132e-05,
+      "loss": 0.6396,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2451258095266837,
+      "learning_rate": 1.999929835468207e-05,
+      "loss": 0.6391,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.276674414464419,
+      "learning_rate": 1.9999266824339502e-05,
+      "loss": 0.7109,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3589387839394185,
+      "learning_rate": 1.999923460105661e-05,
+      "loss": 0.6363,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2702756200783802,
+      "learning_rate": 1.999920168483563e-05,
+      "loss": 0.6659,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2708618223956263,
+      "learning_rate": 1.9999168075678842e-05,
+      "loss": 0.6351,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2391455379672414,
+      "learning_rate": 1.999913377358858e-05,
+      "loss": 0.6541,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1599886274827143,
+      "learning_rate": 1.999909877856721e-05,
+      "loss": 0.665,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.288184833339696,
+      "learning_rate": 1.999906309061717e-05,
+      "loss": 0.6596,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2045342383158713,
+      "learning_rate": 1.999902670974092e-05,
+      "loss": 0.6208,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1790516976577494,
+      "learning_rate": 1.9998989635940996e-05,
+      "loss": 0.6316,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1881008635767092,
+      "learning_rate": 1.9998951869219954e-05,
+      "loss": 0.6216,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2987033612753498,
+      "learning_rate": 1.9998913409580418e-05,
+      "loss": 0.641,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1570461964852865,
+      "learning_rate": 1.9998874257025055e-05,
+      "loss": 0.6407,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1249351566142112,
+      "learning_rate": 1.999883441155657e-05,
+      "loss": 0.6047,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1065769182994274,
+      "learning_rate": 1.999879387317773e-05,
+      "loss": 0.6333,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2437237331960347,
+      "learning_rate": 1.9998752641891347e-05,
+      "loss": 0.6552,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3090764868276303,
+      "learning_rate": 1.9998710717700272e-05,
+      "loss": 0.6934,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4015060478983412,
+      "learning_rate": 1.9998668100607414e-05,
+      "loss": 0.6688,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1350594119836854,
+      "learning_rate": 1.9998624790615722e-05,
+      "loss": 0.5912,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1184799612791112,
+      "learning_rate": 1.9998580787728207e-05,
+      "loss": 0.6357,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2121276795334905,
+      "learning_rate": 1.9998536091947907e-05,
+      "loss": 0.6295,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.243908237449138,
+      "learning_rate": 1.999849070327793e-05,
+      "loss": 0.6512,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1423620436885662,
+      "learning_rate": 1.9998444621721413e-05,
+      "loss": 0.5638,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.195628182158961,
+      "learning_rate": 1.9998397847281548e-05,
+      "loss": 0.6488,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.245661950095741,
+      "learning_rate": 1.9998350379961583e-05,
+      "loss": 0.6223,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2069261086286653,
+      "learning_rate": 1.9998302219764806e-05,
+      "loss": 0.5926,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1260911895558972,
+      "learning_rate": 1.9998253366694555e-05,
+      "loss": 0.6861,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1454069464798424,
+      "learning_rate": 1.9998203820754213e-05,
+      "loss": 0.6685,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1441632874889192,
+      "learning_rate": 1.9998153581947216e-05,
+      "loss": 0.5732,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2796985229602802,
+      "learning_rate": 1.9998102650277046e-05,
+      "loss": 0.6827,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.259238092044345,
+      "learning_rate": 1.9998051025747223e-05,
+      "loss": 0.6325,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1999964484464798,
+      "learning_rate": 1.999799870836134e-05,
+      "loss": 0.6663,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2378447884686448,
+      "learning_rate": 1.999794569812301e-05,
+      "loss": 0.6847,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.180576325266745,
+      "learning_rate": 1.9997891995035914e-05,
+      "loss": 0.6204,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1858681540882434,
+      "learning_rate": 1.9997837599103772e-05,
+      "loss": 0.6072,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1677804559635043,
+      "learning_rate": 1.9997782510330352e-05,
+      "loss": 0.6739,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1158809705378172,
+      "learning_rate": 1.9997726728719468e-05,
+      "loss": 0.5778,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2044311500249878,
+      "learning_rate": 1.9997670254274992e-05,
+      "loss": 0.6658,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1467332394365353,
+      "learning_rate": 1.9997613087000833e-05,
+      "loss": 0.6158,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1366647161281065,
+      "learning_rate": 1.9997555226900957e-05,
+      "loss": 0.6127,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3568063589428707,
+      "learning_rate": 1.9997496673979375e-05,
+      "loss": 0.6049,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2692474254544384,
+      "learning_rate": 1.9997437428240136e-05,
+      "loss": 0.6066,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.359777571955867,
+      "learning_rate": 1.999737748968735e-05,
+      "loss": 0.7016,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.260913107073566,
+      "learning_rate": 1.9997316858325177e-05,
+      "loss": 0.6022,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2523330977441332,
+      "learning_rate": 1.9997255534157814e-05,
+      "loss": 0.6213,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1390397288959015,
+      "learning_rate": 1.9997193517189505e-05,
+      "loss": 0.6717,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3578840331768134,
+      "learning_rate": 1.9997130807424556e-05,
+      "loss": 0.6383,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.441823833716857,
+      "learning_rate": 1.999706740486731e-05,
+      "loss": 0.5689,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2290512975605155,
+      "learning_rate": 1.999700330952216e-05,
+      "loss": 0.5764,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1972836304503411,
+      "learning_rate": 1.9996938521393542e-05,
+      "loss": 0.6016,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.190460722585969,
+      "learning_rate": 1.9996873040485957e-05,
+      "loss": 0.6477,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1123058896187992,
+      "learning_rate": 1.9996806866803937e-05,
+      "loss": 0.615,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0935102246732809,
+      "learning_rate": 1.9996740000352068e-05,
+      "loss": 0.5393,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2682103650008913,
+      "learning_rate": 1.9996672441134987e-05,
+      "loss": 0.7129,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1267973119694383,
+      "learning_rate": 1.999660418915737e-05,
+      "loss": 0.5966,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2916713629818193,
+      "learning_rate": 1.9996535244423947e-05,
+      "loss": 0.5527,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1237915877887323,
+      "learning_rate": 1.99964656069395e-05,
+      "loss": 0.5915,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2113708220129957,
+      "learning_rate": 1.9996395276708856e-05,
+      "loss": 0.6178,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2780784931519782,
+      "learning_rate": 1.9996324253736884e-05,
+      "loss": 0.6967,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2129306566189202,
+      "learning_rate": 1.999625253802851e-05,
+      "loss": 0.6319,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2176291595097406,
+      "learning_rate": 1.99961801295887e-05,
+      "loss": 0.602,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1928293808417556,
+      "learning_rate": 1.9996107028422474e-05,
+      "loss": 0.6077,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4041430903030365,
+      "learning_rate": 1.9996033234534895e-05,
+      "loss": 0.6844,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1082709009194491,
+      "learning_rate": 1.9995958747931083e-05,
+      "loss": 0.6288,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2591220583252387,
+      "learning_rate": 1.9995883568616195e-05,
+      "loss": 0.624,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1472894687965085,
+      "learning_rate": 1.9995807696595442e-05,
+      "loss": 0.6267,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.282837857275939,
+      "learning_rate": 1.9995731131874082e-05,
+      "loss": 0.6904,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2078256390579016,
+      "learning_rate": 1.9995653874457418e-05,
+      "loss": 0.6489,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6494002270659436,
+      "learning_rate": 1.9995575924350813e-05,
+      "loss": 0.6999,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2895378872657652,
+      "learning_rate": 1.9995497281559658e-05,
+      "loss": 0.689,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.350853543232127,
+      "learning_rate": 1.9995417946089407e-05,
+      "loss": 0.6829,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1943843703640955,
+      "learning_rate": 1.999533791794556e-05,
+      "loss": 0.6556,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3439418508960037,
+      "learning_rate": 1.999525719713366e-05,
+      "loss": 0.4694,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5174981490776334,
+      "learning_rate": 1.9995175783659304e-05,
+      "loss": 0.5545,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1823217060352558,
+      "learning_rate": 1.999509367752813e-05,
+      "loss": 0.58,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3439366064000626,
+      "learning_rate": 1.999501087874583e-05,
+      "loss": 0.6784,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1052186698208823,
+      "learning_rate": 1.9994927387318142e-05,
+      "loss": 0.6275,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1759748514990198,
+      "learning_rate": 1.9994843203250853e-05,
+      "loss": 0.6861,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3358621470858025,
+      "learning_rate": 1.9994758326549794e-05,
+      "loss": 0.6388,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2111472045537026,
+      "learning_rate": 1.9994672757220845e-05,
+      "loss": 0.5908,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3290851723488153,
+      "learning_rate": 1.9994586495269944e-05,
+      "loss": 0.6672,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1800279999404348,
+      "learning_rate": 1.9994499540703062e-05,
+      "loss": 0.6758,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.187891209226948,
+      "learning_rate": 1.9994411893526226e-05,
+      "loss": 0.699,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4327628101310725,
+      "learning_rate": 1.9994323553745515e-05,
+      "loss": 0.4593,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2353310670768305,
+      "learning_rate": 1.9994234521367043e-05,
+      "loss": 0.6498,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.217220834744078,
+      "learning_rate": 1.9994144796396985e-05,
+      "loss": 0.637,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1012119323943392,
+      "learning_rate": 1.9994054378841557e-05,
+      "loss": 0.5268,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.109010150374482,
+      "learning_rate": 1.999396326870702e-05,
+      "loss": 0.6481,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3104432477385424,
+      "learning_rate": 1.99938714659997e-05,
+      "loss": 0.6435,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0225947862526552,
+      "learning_rate": 1.9993778970725953e-05,
+      "loss": 0.5475,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1807483240933856,
+      "learning_rate": 1.9993685782892184e-05,
+      "loss": 0.6164,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1036461749929372,
+      "learning_rate": 1.9993591902504854e-05,
+      "loss": 0.6136,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2879988587362776,
+      "learning_rate": 1.9993497329570473e-05,
+      "loss": 0.6551,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1963291092976445,
+      "learning_rate": 1.999340206409559e-05,
+      "loss": 0.6794,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1047181103321142,
+      "learning_rate": 1.9993306106086808e-05,
+      "loss": 0.5978,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1150390655450892,
+      "learning_rate": 1.9993209455550773e-05,
+      "loss": 0.5817,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.233783965559147,
+      "learning_rate": 1.999311211249419e-05,
+      "loss": 0.6413,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2297400715670368,
+      "learning_rate": 1.9993014076923803e-05,
+      "loss": 0.6877,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2022050419955803,
+      "learning_rate": 1.9992915348846403e-05,
+      "loss": 0.6793,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3398503987350763,
+      "learning_rate": 1.9992815928268832e-05,
+      "loss": 0.7163,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0694563998559343,
+      "learning_rate": 1.999271581519798e-05,
+      "loss": 0.6376,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0402449566393575,
+      "learning_rate": 1.999261500964079e-05,
+      "loss": 0.6029,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.227880870571574,
+      "learning_rate": 1.999251351160424e-05,
+      "loss": 0.5998,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1310312614228673,
+      "learning_rate": 1.9992411321095366e-05,
+      "loss": 0.614,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0086715856563895,
+      "learning_rate": 1.9992308438121253e-05,
+      "loss": 0.5827,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.240841306459145,
+      "learning_rate": 1.999220486268903e-05,
+      "loss": 0.5864,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1996928015356496,
+      "learning_rate": 1.9992100594805866e-05,
+      "loss": 0.6422,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4282724911086893,
+      "learning_rate": 1.9991995634479e-05,
+      "loss": 0.6399,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1202283620976718,
+      "learning_rate": 1.9991889981715696e-05,
+      "loss": 0.5842,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.206185140830073,
+      "learning_rate": 1.9991783636523282e-05,
+      "loss": 0.6642,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.090229502573641,
+      "learning_rate": 1.9991676598909124e-05,
+      "loss": 0.6654,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3103010963837258,
+      "learning_rate": 1.999156886888064e-05,
+      "loss": 0.6399,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.243129736124134,
+      "learning_rate": 1.9991460446445297e-05,
+      "loss": 0.5981,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2535060220321637,
+      "learning_rate": 1.9991351331610606e-05,
+      "loss": 0.6845,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0397220665110378,
+      "learning_rate": 1.999124152438413e-05,
+      "loss": 0.6322,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.119272269521874,
+      "learning_rate": 1.9991131024773478e-05,
+      "loss": 0.6278,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.176775643440753,
+      "learning_rate": 1.9991019832786308e-05,
+      "loss": 0.625,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.161217955544425,
+      "learning_rate": 1.9990907948430327e-05,
+      "loss": 0.6539,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1368426586875748,
+      "learning_rate": 1.999079537171329e-05,
+      "loss": 0.6275,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.338855766945942,
+      "learning_rate": 1.9990682102642987e-05,
+      "loss": 0.6615,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1168425819506358,
+      "learning_rate": 1.9990568141227284e-05,
+      "loss": 0.6824,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1617301433270104,
+      "learning_rate": 1.9990453487474067e-05,
+      "loss": 0.6133,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1564343856241175,
+      "learning_rate": 1.9990338141391284e-05,
+      "loss": 0.6341,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2885256022165874,
+      "learning_rate": 1.9990222102986935e-05,
+      "loss": 0.6475,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2400580863381165,
+      "learning_rate": 1.999010537226905e-05,
+      "loss": 0.5882,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2885791998719807,
+      "learning_rate": 1.9989987949245725e-05,
+      "loss": 0.6527,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.177901799429232,
+      "learning_rate": 1.9989869833925094e-05,
+      "loss": 0.6551,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7818280325226765,
+      "learning_rate": 1.9989751026315347e-05,
+      "loss": 0.723,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3058762683856173,
+      "learning_rate": 1.9989631526424716e-05,
+      "loss": 0.6873,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.150498168071998,
+      "learning_rate": 1.998951133426148e-05,
+      "loss": 0.7088,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1648183483500902,
+      "learning_rate": 1.9989390449833968e-05,
+      "loss": 0.634,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.242716598701499,
+      "learning_rate": 1.998926887315056e-05,
+      "loss": 0.6164,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1423813392922217,
+      "learning_rate": 1.998914660421968e-05,
+      "loss": 0.6478,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3749975258731517,
+      "learning_rate": 1.99890236430498e-05,
+      "loss": 0.618,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1455085795035111,
+      "learning_rate": 1.9988899989649438e-05,
+      "loss": 0.5982,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2543874159946773,
+      "learning_rate": 1.9988775644027172e-05,
+      "loss": 0.633,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.174490222564184,
+      "learning_rate": 1.9988650606191614e-05,
+      "loss": 0.6266,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1832096627499953,
+      "learning_rate": 1.9988524876151425e-05,
+      "loss": 0.6144,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.21550652390854,
+      "learning_rate": 1.9988398453915322e-05,
+      "loss": 0.659,
+      "step": 553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1543957372759528,
+      "learning_rate": 1.998827133949207e-05,
+      "loss": 0.6467,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0967112499183245,
+      "learning_rate": 1.998814353289047e-05,
+      "loss": 0.6141,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0267622312276812,
+      "learning_rate": 1.9988015034119385e-05,
+      "loss": 0.6064,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1774243726685896,
+      "learning_rate": 1.9987885843187717e-05,
+      "loss": 0.6174,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0625061356459073,
+      "learning_rate": 1.9987755960104418e-05,
+      "loss": 0.6006,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.084560871180758,
+      "learning_rate": 1.9987625384878493e-05,
+      "loss": 0.6063,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1889027689490759,
+      "learning_rate": 1.9987494117518986e-05,
+      "loss": 0.6729,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1335356972818682,
+      "learning_rate": 1.9987362158034996e-05,
+      "loss": 0.6896,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4050654834507526,
+      "learning_rate": 1.9987229506435666e-05,
+      "loss": 0.5667,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1973334017795196,
+      "learning_rate": 1.998709616273019e-05,
+      "loss": 0.6441,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1948117517144539,
+      "learning_rate": 1.998696212692781e-05,
+      "loss": 0.7147,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1921590585442832,
+      "learning_rate": 1.998682739903781e-05,
+      "loss": 0.6434,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.243769075857278,
+      "learning_rate": 1.9986691979069532e-05,
+      "loss": 0.6608,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.268764077048501,
+      "learning_rate": 1.9986555867032357e-05,
+      "loss": 0.6149,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0901318091600034,
+      "learning_rate": 1.998641906293572e-05,
+      "loss": 0.6869,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1808960933093477,
+      "learning_rate": 1.99862815667891e-05,
+      "loss": 0.6034,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1064884659907313,
+      "learning_rate": 1.9986143378602026e-05,
+      "loss": 0.6145,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2629968487327867,
+      "learning_rate": 1.998600449838407e-05,
+      "loss": 0.6354,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1098239795799647,
+      "learning_rate": 1.998586492614486e-05,
+      "loss": 0.6446,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1172317251983954,
+      "learning_rate": 1.998572466189407e-05,
+      "loss": 0.6059,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1776210955533486,
+      "learning_rate": 1.9985583705641418e-05,
+      "loss": 0.6194,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1421240125909013,
+      "learning_rate": 1.9985442057396675e-05,
+      "loss": 0.5688,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3033078081535983,
+      "learning_rate": 1.9985299717169654e-05,
+      "loss": 0.6404,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1946419016655512,
+      "learning_rate": 1.9985156684970214e-05,
+      "loss": 0.577,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2188327655634794,
+      "learning_rate": 1.9985012960808275e-05,
+      "loss": 0.6336,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2471832078397793,
+      "learning_rate": 1.9984868544693795e-05,
+      "loss": 0.6152,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2827823919534669,
+      "learning_rate": 1.9984723436636785e-05,
+      "loss": 0.6819,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.154753356121878,
+      "learning_rate": 1.9984577636647292e-05,
+      "loss": 0.6452,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2501698234663974,
+      "learning_rate": 1.9984431144735426e-05,
+      "loss": 0.6386,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3338990753512137,
+      "learning_rate": 1.998428396091134e-05,
+      "loss": 0.6606,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1873704742882054,
+      "learning_rate": 1.9984136085185232e-05,
+      "loss": 0.591,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.166414930342365,
+      "learning_rate": 1.9983987517567348e-05,
+      "loss": 0.6735,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1362047762422045,
+      "learning_rate": 1.998383825806799e-05,
+      "loss": 0.5989,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1785883612942007,
+      "learning_rate": 1.9983688306697488e-05,
+      "loss": 0.6682,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1575253422468532,
+      "learning_rate": 1.9983537663466244e-05,
+      "loss": 0.5574,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2171303979561474,
+      "learning_rate": 1.9983386328384696e-05,
+      "loss": 0.6653,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1779477719716205,
+      "learning_rate": 1.998323430146333e-05,
+      "loss": 0.6468,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0520298733113895,
+      "learning_rate": 1.9983081582712684e-05,
+      "loss": 0.5669,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2413891317205863,
+      "learning_rate": 1.9982928172143337e-05,
+      "loss": 0.6552,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0805023983063284,
+      "learning_rate": 1.9982774069765923e-05,
+      "loss": 0.5755,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2462092352478134,
+      "learning_rate": 1.9982619275591124e-05,
+      "loss": 0.6539,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2311934123337847,
+      "learning_rate": 1.998246378962966e-05,
+      "loss": 0.6769,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2862807486957082,
+      "learning_rate": 1.9982307611892314e-05,
+      "loss": 0.7153,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1501963851498185,
+      "learning_rate": 1.9982150742389897e-05,
+      "loss": 0.586,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0185449240780169,
+      "learning_rate": 1.9981993181133297e-05,
+      "loss": 0.6049,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.390711835950105,
+      "learning_rate": 1.998183492813342e-05,
+      "loss": 0.6899,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4417654309382693,
+      "learning_rate": 1.9981675983401234e-05,
+      "loss": 0.6118,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0392600350000416,
+      "learning_rate": 1.9981516346947757e-05,
+      "loss": 0.5494,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.124642591279403,
+      "learning_rate": 1.998135601878405e-05,
+      "loss": 0.5698,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1218889954717481,
+      "learning_rate": 1.9981194998921226e-05,
+      "loss": 0.6356,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.189137467170877,
+      "learning_rate": 1.9981033287370443e-05,
+      "loss": 0.6391,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1882587920699188,
+      "learning_rate": 1.9980870884142906e-05,
+      "loss": 0.6681,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.255753276002731,
+      "learning_rate": 1.9980707789249866e-05,
+      "loss": 0.6781,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2023510062253613,
+      "learning_rate": 1.9980544002702635e-05,
+      "loss": 0.7198,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.054688112742724,
+      "learning_rate": 1.998037952451255e-05,
+      "loss": 0.6138,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.107499007509298,
+      "learning_rate": 1.9980214354691022e-05,
+      "loss": 0.5985,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0589520103441592,
+      "learning_rate": 1.998004849324949e-05,
+      "loss": 0.6205,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1975815219590236,
+      "learning_rate": 1.997988194019945e-05,
+      "loss": 0.5926,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2336481921456244,
+      "learning_rate": 1.9979714695552444e-05,
+      "loss": 0.6479,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0394534716312003,
+      "learning_rate": 1.997954675932006e-05,
+      "loss": 0.5713,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1298571096569328,
+      "learning_rate": 1.997937813151394e-05,
+      "loss": 0.606,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1594385853901574,
+      "learning_rate": 1.9979208812145766e-05,
+      "loss": 0.6047,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0480989117397763,
+      "learning_rate": 1.9979038801227273e-05,
+      "loss": 0.6042,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.11743684858994,
+      "learning_rate": 1.9978868098770244e-05,
+      "loss": 0.629,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0789247877309214,
+      "learning_rate": 1.9978696704786505e-05,
+      "loss": 0.7054,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.214704874693797,
+      "learning_rate": 1.9978524619287937e-05,
+      "loss": 0.6837,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2762010798985541,
+      "learning_rate": 1.997835184228646e-05,
+      "loss": 0.6524,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.036839257758854,
+      "learning_rate": 1.9978178373794055e-05,
+      "loss": 0.6761,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0783286513719492,
+      "learning_rate": 1.9978004213822736e-05,
+      "loss": 0.6654,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2133745178374113,
+      "learning_rate": 1.997782936238458e-05,
+      "loss": 0.6946,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2098030928382038,
+      "learning_rate": 1.9977653819491696e-05,
+      "loss": 0.5995,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1151817187068271,
+      "learning_rate": 1.9977477585156252e-05,
+      "loss": 0.586,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3347118405295793,
+      "learning_rate": 1.9977300659390463e-05,
+      "loss": 0.6752,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1815468771856832,
+      "learning_rate": 1.997712304220659e-05,
+      "loss": 0.5802,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1188855784750988,
+      "learning_rate": 1.9976944733616935e-05,
+      "loss": 0.63,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0985965917225944,
+      "learning_rate": 1.9976765733633866e-05,
+      "loss": 0.5863,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2046569696689249,
+      "learning_rate": 1.9976586042269776e-05,
+      "loss": 0.4522,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1672364801364987,
+      "learning_rate": 1.9976405659537123e-05,
+      "loss": 0.6388,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1689539536673572,
+      "learning_rate": 1.9976224585448407e-05,
+      "loss": 0.6218,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2637252555860425,
+      "learning_rate": 1.9976042820016176e-05,
+      "loss": 0.6143,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1814014645815236,
+      "learning_rate": 1.997586036325303e-05,
+      "loss": 0.6382,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.222151385844291,
+      "learning_rate": 1.9975677215171606e-05,
+      "loss": 0.6453,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0914309372643383,
+      "learning_rate": 1.9975493375784598e-05,
+      "loss": 0.6333,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.201970236345254,
+      "learning_rate": 1.997530884510475e-05,
+      "loss": 0.6586,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.054938735797658,
+      "learning_rate": 1.9975123623144847e-05,
+      "loss": 0.6288,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3866748760527043,
+      "learning_rate": 1.9974937709917722e-05,
+      "loss": 0.6907,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1652988956147512,
+      "learning_rate": 1.9974751105436266e-05,
+      "loss": 0.659,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2370818741683758,
+      "learning_rate": 1.9974563809713406e-05,
+      "loss": 0.4051,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2982928662518747,
+      "learning_rate": 1.9974375822762117e-05,
+      "loss": 0.6157,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0479843603076193,
+      "learning_rate": 1.9974187144595433e-05,
+      "loss": 0.4918,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2429258136089865,
+      "learning_rate": 1.9973997775226424e-05,
+      "loss": 0.653,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2192984610891346,
+      "learning_rate": 1.9973807714668224e-05,
+      "loss": 0.6706,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1251401683601023,
+      "learning_rate": 1.997361696293399e-05,
+      "loss": 0.6229,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0708970840194627,
+      "learning_rate": 1.9973425520036948e-05,
+      "loss": 0.5686,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1553407807232856,
+      "learning_rate": 1.9973233385990364e-05,
+      "loss": 0.6734,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1466860232715055,
+      "learning_rate": 1.997304056080755e-05,
+      "loss": 0.5551,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.350493965217335,
+      "learning_rate": 1.9972847044501876e-05,
+      "loss": 0.5949,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2454267382525472,
+      "learning_rate": 1.9972652837086746e-05,
+      "loss": 0.6507,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.029352302353856,
+      "learning_rate": 1.997245793857562e-05,
+      "loss": 0.5958,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1497968726198924,
+      "learning_rate": 1.9972262348982e-05,
+      "loss": 0.67,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.088989161085879,
+      "learning_rate": 1.997206606831945e-05,
+      "loss": 0.5854,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1790833944364503,
+      "learning_rate": 1.997186909660157e-05,
+      "loss": 0.6825,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0384747944909554,
+      "learning_rate": 1.9971671433842e-05,
+      "loss": 0.6763,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1213533648567533,
+      "learning_rate": 1.9971473080054445e-05,
+      "loss": 0.6894,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1609933428042643,
+      "learning_rate": 1.9971274035252653e-05,
+      "loss": 0.5707,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2081990423276776,
+      "learning_rate": 1.9971074299450414e-05,
+      "loss": 0.6806,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0724208806076496,
+      "learning_rate": 1.9970873872661567e-05,
+      "loss": 0.5806,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0610754332454024,
+      "learning_rate": 1.997067275490001e-05,
+      "loss": 0.6022,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1629971401024213,
+      "learning_rate": 1.997047094617967e-05,
+      "loss": 0.6024,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2436199105527355,
+      "learning_rate": 1.9970268446514543e-05,
+      "loss": 0.6317,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1442220153552487,
+      "learning_rate": 1.997006525591865e-05,
+      "loss": 0.6532,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3021218257030307,
+      "learning_rate": 1.9969861374406086e-05,
+      "loss": 0.6593,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1111386090671853,
+      "learning_rate": 1.9969656801990967e-05,
+      "loss": 0.6495,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1031034022345894,
+      "learning_rate": 1.9969451538687474e-05,
+      "loss": 0.6161,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0933845844784758,
+      "learning_rate": 1.9969245584509832e-05,
+      "loss": 0.5477,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1712771182304644,
+      "learning_rate": 1.9969038939472315e-05,
+      "loss": 0.6338,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2155964784926518,
+      "learning_rate": 1.9968831603589243e-05,
+      "loss": 0.6716,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1244215186538657,
+      "learning_rate": 1.9968623576874984e-05,
+      "loss": 0.6159,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.11184462359373,
+      "learning_rate": 1.996841485934395e-05,
+      "loss": 0.5756,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.136388764225708,
+      "learning_rate": 1.996820545101061e-05,
+      "loss": 0.6362,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2594325771585668,
+      "learning_rate": 1.9967995351889476e-05,
+      "loss": 0.6782,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.077894031554686,
+      "learning_rate": 1.9967784561995103e-05,
+      "loss": 0.5933,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.085391720228807,
+      "learning_rate": 1.9967573081342103e-05,
+      "loss": 0.6247,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.092634209209323,
+      "learning_rate": 1.996736090994513e-05,
+      "loss": 0.5373,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.198269526162391,
+      "learning_rate": 1.9967148047818884e-05,
+      "loss": 0.5686,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2183730301752702,
+      "learning_rate": 1.996693449497812e-05,
+      "loss": 0.6678,
+      "step": 679
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1399175734048983,
+      "learning_rate": 1.9966720251437635e-05,
+      "loss": 0.6017,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0273161496250076,
+      "learning_rate": 1.996650531721228e-05,
+      "loss": 0.5534,
+      "step": 681
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.047089911204804,
+      "learning_rate": 1.9966289692316944e-05,
+      "loss": 0.5956,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.217569905753707,
+      "learning_rate": 1.9966073376766575e-05,
+      "loss": 0.6095,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1145304786124874,
+      "learning_rate": 1.9965856370576163e-05,
+      "loss": 0.613,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1977722179345607,
+      "learning_rate": 1.9965638673760738e-05,
+      "loss": 0.6071,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2160224513488096,
+      "learning_rate": 1.9965420286335397e-05,
+      "loss": 0.6376,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3450712803411724,
+      "learning_rate": 1.996520120831527e-05,
+      "loss": 0.6046,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1398839262388,
+      "learning_rate": 1.9964981439715532e-05,
+      "loss": 0.5862,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1615547015849625,
+      "learning_rate": 1.9964760980551428e-05,
+      "loss": 0.5632,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1095424326717258,
+      "learning_rate": 1.996453983083822e-05,
+      "loss": 0.6577,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1245069632159537,
+      "learning_rate": 1.9964317990591243e-05,
+      "loss": 0.6358,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0739036008737486,
+      "learning_rate": 1.9964095459825866e-05,
+      "loss": 0.6257,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0889267792227462,
+      "learning_rate": 1.9963872238557516e-05,
+      "loss": 0.6612,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.101405073966227,
+      "learning_rate": 1.9963648326801653e-05,
+      "loss": 0.6636,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.061467009295685,
+      "learning_rate": 1.99634237245738e-05,
+      "loss": 0.4699,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0916752453273684,
+      "learning_rate": 1.9963198431889523e-05,
+      "loss": 0.5637,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0034377408449018,
+      "learning_rate": 1.996297244876443e-05,
+      "loss": 0.5266,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1009331610859694,
+      "learning_rate": 1.9962745775214187e-05,
+      "loss": 0.6557,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.006040563282083,
+      "learning_rate": 1.9962518411254493e-05,
+      "loss": 0.571,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2442846642824592,
+      "learning_rate": 1.996229035690111e-05,
+      "loss": 0.6507,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0822018700447882,
+      "learning_rate": 1.9962061612169844e-05,
+      "loss": 0.648,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2097081161212462,
+      "learning_rate": 1.9961832177076544e-05,
+      "loss": 0.6196,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2030315390368393,
+      "learning_rate": 1.996160205163711e-05,
+      "loss": 0.6099,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1297382460923315,
+      "learning_rate": 1.9961371235867494e-05,
+      "loss": 0.6596,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1674304329946574,
+      "learning_rate": 1.9961139729783683e-05,
+      "loss": 0.6333,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2304187056475613,
+      "learning_rate": 1.9960907533401722e-05,
+      "loss": 0.6619,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1425239340299234,
+      "learning_rate": 1.996067464673771e-05,
+      "loss": 0.5948,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2514604903606674,
+      "learning_rate": 1.9960441069807778e-05,
+      "loss": 0.6098,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1006930003492308,
+      "learning_rate": 1.996020680262811e-05,
+      "loss": 0.6225,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0897684407070278,
+      "learning_rate": 1.9959971845214953e-05,
+      "loss": 0.6381,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0351947993478245,
+      "learning_rate": 1.9959736197584577e-05,
+      "loss": 0.5927,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2403363792215993,
+      "learning_rate": 1.9959499859753317e-05,
+      "loss": 0.647,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1981011259553083,
+      "learning_rate": 1.995926283173755e-05,
+      "loss": 0.6881,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9841149549694405,
+      "learning_rate": 1.9959025113553706e-05,
+      "loss": 0.6494,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0600192658138001,
+      "learning_rate": 1.9958786705218254e-05,
+      "loss": 0.6265,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2811772952743912,
+      "learning_rate": 1.9958547606747715e-05,
+      "loss": 0.5239,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1521958975518616,
+      "learning_rate": 1.9958307818158662e-05,
+      "loss": 0.6341,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1179962043127805,
+      "learning_rate": 1.995806733946771e-05,
+      "loss": 0.6329,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0146153054249425,
+      "learning_rate": 1.995782617069152e-05,
+      "loss": 0.5976,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1645750270250357,
+      "learning_rate": 1.9957584311846814e-05,
+      "loss": 0.576,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2491529995294561,
+      "learning_rate": 1.9957341762950346e-05,
+      "loss": 0.7102,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1764586868629787,
+      "learning_rate": 1.9957098524018925e-05,
+      "loss": 0.6875,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1389846638956975,
+      "learning_rate": 1.995685459506941e-05,
+      "loss": 0.6148,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.057344796664354,
+      "learning_rate": 1.9956609976118704e-05,
+      "loss": 0.5285,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.246786646040238,
+      "learning_rate": 1.9956364667183755e-05,
+      "loss": 0.5909,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2410890769714888,
+      "learning_rate": 1.9956118668281568e-05,
+      "loss": 0.6763,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1142586997360597,
+      "learning_rate": 1.9955871979429188e-05,
+      "loss": 0.6469,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1276135732050507,
+      "learning_rate": 1.9955624600643712e-05,
+      "loss": 0.6592,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.066854522714197,
+      "learning_rate": 1.9955376531942278e-05,
+      "loss": 0.5611,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.08373527229625,
+      "learning_rate": 1.9955127773342086e-05,
+      "loss": 0.5927,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2196226278884847,
+      "learning_rate": 1.9954878324860365e-05,
+      "loss": 0.6232,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1409979528374246,
+      "learning_rate": 1.995462818651441e-05,
+      "loss": 0.6322,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.074608252737145,
+      "learning_rate": 1.9954377358321547e-05,
+      "loss": 0.5813,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0325700029995777,
+      "learning_rate": 1.9954125840299165e-05,
+      "loss": 0.4502,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1297723614367152,
+      "learning_rate": 1.995387363246469e-05,
+      "loss": 0.6199,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0916886554647816,
+      "learning_rate": 1.9953620734835603e-05,
+      "loss": 0.6186,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1978421868431721,
+      "learning_rate": 1.995336714742943e-05,
+      "loss": 0.6573,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.146234644774168,
+      "learning_rate": 1.9953112870263737e-05,
+      "loss": 0.6158,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0833865452731306,
+      "learning_rate": 1.9952857903356155e-05,
+      "loss": 0.6236,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0277210942729749,
+      "learning_rate": 1.9952602246724348e-05,
+      "loss": 0.6171,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1046435282049112,
+      "learning_rate": 1.995234590038603e-05,
+      "loss": 0.6509,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.187550009575104,
+      "learning_rate": 1.995208886435897e-05,
+      "loss": 0.6832,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1013173491692863,
+      "learning_rate": 1.995183113866098e-05,
+      "loss": 0.6326,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0349313608491895,
+      "learning_rate": 1.9951572723309918e-05,
+      "loss": 0.5911,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.13623690688545,
+      "learning_rate": 1.9951313618323696e-05,
+      "loss": 0.6131,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.05623891586562,
+      "learning_rate": 1.9951053823720267e-05,
+      "loss": 0.6311,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.098775832907385,
+      "learning_rate": 1.9950793339517632e-05,
+      "loss": 0.6321,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0964712573130224,
+      "learning_rate": 1.9950532165733847e-05,
+      "loss": 0.6329,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1245652087511098,
+      "learning_rate": 1.995027030238701e-05,
+      "loss": 0.6432,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2375929509587618,
+      "learning_rate": 1.9950007749495263e-05,
+      "loss": 0.611,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.080590540348563,
+      "learning_rate": 1.9949744507076806e-05,
+      "loss": 0.6639,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.106532099604561,
+      "learning_rate": 1.994948057514988e-05,
+      "loss": 0.6232,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0356971574762561,
+      "learning_rate": 1.994921595373278e-05,
+      "loss": 0.5809,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2054321702962159,
+      "learning_rate": 1.9948950642843836e-05,
+      "loss": 0.6959,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1866427603796048,
+      "learning_rate": 1.9948684642501433e-05,
+      "loss": 0.5915,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1568127923932405,
+      "learning_rate": 1.9948417952724014e-05,
+      "loss": 0.6201,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0679478432065364,
+      "learning_rate": 1.9948150573530054e-05,
+      "loss": 0.6118,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2389782363966395,
+      "learning_rate": 1.994788250493808e-05,
+      "loss": 0.6178,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1320990625121634,
+      "learning_rate": 1.9947613746966678e-05,
+      "loss": 0.5965,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.055659570175554,
+      "learning_rate": 1.9947344299634464e-05,
+      "loss": 0.6197,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0158590766537203,
+      "learning_rate": 1.9947074162960113e-05,
+      "loss": 0.6052,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.213723272660111,
+      "learning_rate": 1.9946803336962346e-05,
+      "loss": 0.5871,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1870814088915298,
+      "learning_rate": 1.994653182165993e-05,
+      "loss": 0.6191,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2498400851275093,
+      "learning_rate": 1.994625961707168e-05,
+      "loss": 0.7248,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1683672596182015,
+      "learning_rate": 1.9945986723216463e-05,
+      "loss": 0.5657,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1686459028088645,
+      "learning_rate": 1.9945713140113188e-05,
+      "loss": 0.6563,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.174814353567215,
+      "learning_rate": 1.9945438867780814e-05,
+      "loss": 0.5865,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1195506080858777,
+      "learning_rate": 1.9945163906238347e-05,
+      "loss": 0.6169,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2211466830279876,
+      "learning_rate": 1.9944888255504846e-05,
+      "loss": 0.6599,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.098094186565308,
+      "learning_rate": 1.994461191559941e-05,
+      "loss": 0.5711,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0342907512334356,
+      "learning_rate": 1.9944334886541184e-05,
+      "loss": 0.6,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1118274230511065,
+      "learning_rate": 1.9944057168349374e-05,
+      "loss": 0.5939,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0468608125121714,
+      "learning_rate": 1.9943778761043223e-05,
+      "loss": 0.5908,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0957138400050503,
+      "learning_rate": 1.994349966464202e-05,
+      "loss": 0.6157,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1635606740555189,
+      "learning_rate": 1.9943219879165113e-05,
+      "loss": 0.6674,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0570904827261942,
+      "learning_rate": 1.9942939404631893e-05,
+      "loss": 0.5812,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1047314581338863,
+      "learning_rate": 1.9942658241061785e-05,
+      "loss": 0.6725,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1209323619076677,
+      "learning_rate": 1.9942376388474282e-05,
+      "loss": 0.6099,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0531072607859158,
+      "learning_rate": 1.9942093846888912e-05,
+      "loss": 0.654,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1245284813161793,
+      "learning_rate": 1.9941810616325262e-05,
+      "loss": 0.6879,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2806531327448978,
+      "learning_rate": 1.994152669680295e-05,
+      "loss": 0.6672,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1367446672914951,
+      "learning_rate": 1.994124208834166e-05,
+      "loss": 0.6449,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.123488436414863,
+      "learning_rate": 1.9940956790961108e-05,
+      "loss": 0.6283,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1621898434373996,
+      "learning_rate": 1.9940670804681068e-05,
+      "loss": 0.6227,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1273964782620118,
+      "learning_rate": 1.994038412952136e-05,
+      "loss": 0.6902,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1176649212729894,
+      "learning_rate": 1.994009676550185e-05,
+      "loss": 0.6174,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1072199332026769,
+      "learning_rate": 1.993980871264245e-05,
+      "loss": 0.6046,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0486363905651077,
+      "learning_rate": 1.993951997096312e-05,
+      "loss": 0.5663,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.033014159421138,
+      "learning_rate": 1.9939230540483873e-05,
+      "loss": 0.5875,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1065308629200603,
+      "learning_rate": 1.9938940421224768e-05,
+      "loss": 0.6169,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0786912963350155,
+      "learning_rate": 1.9938649613205907e-05,
+      "loss": 0.6306,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.129046654293702,
+      "learning_rate": 1.9938358116447444e-05,
+      "loss": 0.631,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9863221302179781,
+      "learning_rate": 1.9938065930969578e-05,
+      "loss": 0.5559,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.089501085757703,
+      "learning_rate": 1.993777305679256e-05,
+      "loss": 0.6775,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2022727340828454,
+      "learning_rate": 1.993747949393668e-05,
+      "loss": 0.6469,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1368424852076264,
+      "learning_rate": 1.993718524242229e-05,
+      "loss": 0.6475,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0820497403961937,
+      "learning_rate": 1.9936890302269773e-05,
+      "loss": 0.5831,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0898352110514833,
+      "learning_rate": 1.9936594673499578e-05,
+      "loss": 0.6381,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2294389908535341,
+      "learning_rate": 1.993629835613218e-05,
+      "loss": 0.4407,
+      "step": 799
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0918414115098216,
+      "learning_rate": 1.993600135018812e-05,
+      "loss": 0.6081,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9537633730917028,
+      "learning_rate": 1.9935703655687982e-05,
+      "loss": 0.5166,
+      "step": 801
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0660382716002714,
+      "learning_rate": 1.993540527265239e-05,
+      "loss": 0.6609,
+      "step": 802
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1379226647047838,
+      "learning_rate": 1.9935106201102032e-05,
+      "loss": 0.6515,
+      "step": 803
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0589300124561416,
+      "learning_rate": 1.993480644105762e-05,
+      "loss": 0.5814,
+      "step": 804
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0175478881295723,
+      "learning_rate": 1.9934505992539934e-05,
+      "loss": 0.5841,
+      "step": 805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1396476742438806,
+      "learning_rate": 1.99342048555698e-05,
+      "loss": 0.6405,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1621349884764827,
+      "learning_rate": 1.9933903030168075e-05,
+      "loss": 0.7015,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0632458504231408,
+      "learning_rate": 1.9933600516355684e-05,
+      "loss": 0.6147,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.06885761464636,
+      "learning_rate": 1.9933297314153593e-05,
+      "loss": 0.6383,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0962532335479325,
+      "learning_rate": 1.99329934235828e-05,
+      "loss": 0.5765,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1464346102920349,
+      "learning_rate": 1.993268884466438e-05,
+      "loss": 0.5723,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2509669034212627,
+      "learning_rate": 1.9932383577419432e-05,
+      "loss": 0.5771,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0414712734658653,
+      "learning_rate": 1.9932077621869112e-05,
+      "loss": 0.6552,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2035876318057512,
+      "learning_rate": 1.993177097803462e-05,
+      "loss": 0.6744,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.118292884144844,
+      "learning_rate": 1.993146364593721e-05,
+      "loss": 0.6673,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2133348072449912,
+      "learning_rate": 1.993115562559818e-05,
+      "loss": 0.6326,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.054058689703446,
+      "learning_rate": 1.9930846917038873e-05,
+      "loss": 0.6147,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2012351073159693,
+      "learning_rate": 1.9930537520280684e-05,
+      "loss": 0.6592,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.164198742439426,
+      "learning_rate": 1.9930227435345053e-05,
+      "loss": 0.6111,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0662066688946203,
+      "learning_rate": 1.992991666225347e-05,
+      "loss": 0.5349,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.999433802047172,
+      "learning_rate": 1.9929605201027468e-05,
+      "loss": 0.6664,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1356669429058914,
+      "learning_rate": 1.9929293051688634e-05,
+      "loss": 0.5944,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0915185271313848,
+      "learning_rate": 1.9928980214258597e-05,
+      "loss": 0.6321,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1512780011586605,
+      "learning_rate": 1.992866668875904e-05,
+      "loss": 0.6711,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.211306666748613,
+      "learning_rate": 1.992835247521169e-05,
+      "loss": 0.6296,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1157941995912357,
+      "learning_rate": 1.9928037573638316e-05,
+      "loss": 0.6819,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1529125372179114,
+      "learning_rate": 1.9927721984060747e-05,
+      "loss": 0.6221,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.172281554294927,
+      "learning_rate": 1.992740570650085e-05,
+      "loss": 0.662,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.150850811381565,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.683,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0835533671026325,
+      "learning_rate": 1.992677108752179e-05,
+      "loss": 0.6373,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1311275755980512,
+      "learning_rate": 1.9926452746146605e-05,
+      "loss": 0.6024,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0663614825308445,
+      "learning_rate": 1.992613371687705e-05,
+      "loss": 0.4717,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.122966022638328,
+      "learning_rate": 1.9925813999735238e-05,
+      "loss": 0.6189,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2502857431361363,
+      "learning_rate": 1.992549359474332e-05,
+      "loss": 0.6694,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9883676453300628,
+      "learning_rate": 1.992517250192349e-05,
+      "loss": 0.6629,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0843568044317529,
+      "learning_rate": 1.9924850721298017e-05,
+      "loss": 0.5797,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1302426860122095,
+      "learning_rate": 1.992452825288919e-05,
+      "loss": 0.6683,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1279702636357174,
+      "learning_rate": 1.992420509671936e-05,
+      "loss": 0.6624,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1548522444197566,
+      "learning_rate": 1.9923881252810917e-05,
+      "loss": 0.6606,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2189793417281167,
+      "learning_rate": 1.9923556721186308e-05,
+      "loss": 0.6906,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0057515507994603,
+      "learning_rate": 1.9923231501868018e-05,
+      "loss": 0.6407,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.03480201906667,
+      "learning_rate": 1.992290559487859e-05,
+      "loss": 0.6212,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2312157692437504,
+      "learning_rate": 1.9922579000240602e-05,
+      "loss": 0.6539,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1510651239803475,
+      "learning_rate": 1.9922251717976697e-05,
+      "loss": 0.6423,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.228234587897638,
+      "learning_rate": 1.992192374810954e-05,
+      "loss": 0.6523,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3035456156567933,
+      "learning_rate": 1.9921595090661872e-05,
+      "loss": 0.6438,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.125010510695676,
+      "learning_rate": 1.9921265745656466e-05,
+      "loss": 0.6128,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0702335652144521,
+      "learning_rate": 1.9920935713116144e-05,
+      "loss": 0.6311,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1455341862560653,
+      "learning_rate": 1.9920604993063777e-05,
+      "loss": 0.6406,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.203676850522986,
+      "learning_rate": 1.992027358552228e-05,
+      "loss": 0.7314,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0271676262087595,
+      "learning_rate": 1.991994149051463e-05,
+      "loss": 0.5971,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.205644875995286,
+      "learning_rate": 1.9919608708063826e-05,
+      "loss": 0.6107,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2046978136025066,
+      "learning_rate": 1.991927523819294e-05,
+      "loss": 0.6406,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0945845324462102,
+      "learning_rate": 1.991894108092508e-05,
+      "loss": 0.5819,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1130735560807379,
+      "learning_rate": 1.99186062362834e-05,
+      "loss": 0.6191,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0389898930015686,
+      "learning_rate": 1.9918270704291104e-05,
+      "loss": 0.6716,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2716474232262376,
+      "learning_rate": 1.991793448497145e-05,
+      "loss": 0.6004,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0389131172541453,
+      "learning_rate": 1.991759757834773e-05,
+      "loss": 0.5947,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0950374146231157,
+      "learning_rate": 1.9917259984443295e-05,
+      "loss": 0.6049,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0364117726822706,
+      "learning_rate": 1.991692170328154e-05,
+      "loss": 0.6233,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1011938895926205,
+      "learning_rate": 1.9916582734885906e-05,
+      "loss": 0.7002,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.061748573780894,
+      "learning_rate": 1.991624307927989e-05,
+      "loss": 0.5697,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0460051979027172,
+      "learning_rate": 1.991590273648702e-05,
+      "loss": 0.607,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0497730613084617,
+      "learning_rate": 1.9915561706530882e-05,
+      "loss": 0.5858,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0650382321114669,
+      "learning_rate": 1.9915219989435117e-05,
+      "loss": 0.6258,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0583106709875194,
+      "learning_rate": 1.9914877585223403e-05,
+      "loss": 0.6135,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.073826426367363,
+      "learning_rate": 1.9914534493919464e-05,
+      "loss": 0.6618,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0532675508825666,
+      "learning_rate": 1.991419071554708e-05,
+      "loss": 0.6091,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.126468696862115,
+      "learning_rate": 1.9913846250130074e-05,
+      "loss": 0.6568,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2196257672975,
+      "learning_rate": 1.9913501097692312e-05,
+      "loss": 0.6105,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1475489520819309,
+      "learning_rate": 1.9913155258257724e-05,
+      "loss": 0.5912,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1808552294384747,
+      "learning_rate": 1.9912808731850265e-05,
+      "loss": 0.6991,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1584108242615976,
+      "learning_rate": 1.991246151849396e-05,
+      "loss": 0.6977,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0289436915103616,
+      "learning_rate": 1.9912113618212852e-05,
+      "loss": 0.6501,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0439752354711855,
+      "learning_rate": 1.991176503103107e-05,
+      "loss": 0.6131,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1188632994545695,
+      "learning_rate": 1.9911415756972764e-05,
+      "loss": 0.6592,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2134771338604364,
+      "learning_rate": 1.9911065796062137e-05,
+      "loss": 0.6363,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1070321810757595,
+      "learning_rate": 1.9910715148323438e-05,
+      "loss": 0.6432,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1159201260679672,
+      "learning_rate": 1.9910363813780975e-05,
+      "loss": 0.5572,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.082698575626692,
+      "learning_rate": 1.9910011792459086e-05,
+      "loss": 0.5655,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0551268128565945,
+      "learning_rate": 1.9909659084382172e-05,
+      "loss": 0.5494,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1463091432748589,
+      "learning_rate": 1.9909305689574672e-05,
+      "loss": 0.6349,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0274960646007008,
+      "learning_rate": 1.9908951608061078e-05,
+      "loss": 0.6146,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0541607414609717,
+      "learning_rate": 1.9908596839865927e-05,
+      "loss": 0.5856,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.416168177554952,
+      "learning_rate": 1.9908241385013804e-05,
+      "loss": 0.4781,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.068463683699775,
+      "learning_rate": 1.990788524352934e-05,
+      "loss": 0.5868,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1117140695290186,
+      "learning_rate": 1.990752841543722e-05,
+      "loss": 0.681,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.108831597329571,
+      "learning_rate": 1.9907170900762164e-05,
+      "loss": 0.6391,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1082692126241613,
+      "learning_rate": 1.9906812699528956e-05,
+      "loss": 0.6302,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.084315446777715,
+      "learning_rate": 1.9906453811762415e-05,
+      "loss": 0.5957,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0997233557871868,
+      "learning_rate": 1.990609423748741e-05,
+      "loss": 0.6264,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1273421993059851,
+      "learning_rate": 1.9905733976728862e-05,
+      "loss": 0.6095,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1057839655029467,
+      "learning_rate": 1.990537302951174e-05,
+      "loss": 0.571,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0514643831580244,
+      "learning_rate": 1.9905011395861048e-05,
+      "loss": 0.5988,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0361450310583142,
+      "learning_rate": 1.9904649075801852e-05,
+      "loss": 0.5558,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.150559012228703,
+      "learning_rate": 1.9904286069359263e-05,
+      "loss": 0.6765,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.005249808253615,
+      "learning_rate": 1.9903922376558432e-05,
+      "loss": 0.5878,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9205219013292413,
+      "learning_rate": 1.9903557997424565e-05,
+      "loss": 0.5287,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0380631233474868,
+      "learning_rate": 1.9903192931982916e-05,
+      "loss": 0.635,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1151631244071423,
+      "learning_rate": 1.9902827180258778e-05,
+      "loss": 0.6371,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1385580867661342,
+      "learning_rate": 1.99024607422775e-05,
+      "loss": 0.6381,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1749410505707398,
+      "learning_rate": 1.9902093618064483e-05,
+      "loss": 0.676,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0927232181593396,
+      "learning_rate": 1.9901725807645154e-05,
+      "loss": 0.6469,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0974191506180526,
+      "learning_rate": 1.990135731104501e-05,
+      "loss": 0.5542,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.124203680324317,
+      "learning_rate": 1.9900988128289593e-05,
+      "loss": 0.6122,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0088657889858017,
+      "learning_rate": 1.990061825940447e-05,
+      "loss": 0.6634,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0044706318253844,
+      "learning_rate": 1.990024770441529e-05,
+      "loss": 0.5614,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1291401742975522,
+      "learning_rate": 1.9899876463347727e-05,
+      "loss": 0.5792,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1340186318427607,
+      "learning_rate": 1.9899504536227505e-05,
+      "loss": 0.6871,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1230863530970432,
+      "learning_rate": 1.98991319230804e-05,
+      "loss": 0.6134,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1644758801324757,
+      "learning_rate": 1.989875862393223e-05,
+      "loss": 0.5701,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.094310361279521,
+      "learning_rate": 1.989838463880887e-05,
+      "loss": 0.6342,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0246166924294686,
+      "learning_rate": 1.9898009967736236e-05,
+      "loss": 0.5564,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.029422191382958,
+      "learning_rate": 1.989763461074029e-05,
+      "loss": 0.5958,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.107913272959802,
+      "learning_rate": 1.989725856784704e-05,
+      "loss": 0.5847,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.096318807524696,
+      "learning_rate": 1.9896881839082554e-05,
+      "loss": 0.6037,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2018583652830437,
+      "learning_rate": 1.9896504424472936e-05,
+      "loss": 0.6224,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1961833371889385,
+      "learning_rate": 1.9896126324044338e-05,
+      "loss": 0.6426,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2817682573246538,
+      "learning_rate": 1.9895747537822965e-05,
+      "loss": 0.6955,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.14955929893788,
+      "learning_rate": 1.989536806583506e-05,
+      "loss": 0.6328,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1540859308447506,
+      "learning_rate": 1.9894987908106933e-05,
+      "loss": 0.6459,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0375748859320604,
+      "learning_rate": 1.9894607064664914e-05,
+      "loss": 0.6201,
+      "step": 922
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1298354214589668,
+      "learning_rate": 1.9894225535535407e-05,
+      "loss": 0.6913,
+      "step": 923
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0021066744486649,
+      "learning_rate": 1.9893843320744845e-05,
+      "loss": 0.5619,
+      "step": 924
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0168328583368194,
+      "learning_rate": 1.9893460420319716e-05,
+      "loss": 0.6607,
+      "step": 925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9896091925300838,
+      "learning_rate": 1.9893076834286557e-05,
+      "loss": 0.5867,
+      "step": 926
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2502531797998122,
+      "learning_rate": 1.9892692562671944e-05,
+      "loss": 0.718,
+      "step": 927
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0353809825535139,
+      "learning_rate": 1.9892307605502514e-05,
+      "loss": 0.5753,
+      "step": 928
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0514084405024644,
+      "learning_rate": 1.9891921962804942e-05,
+      "loss": 0.5604,
+      "step": 929
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0772823283544717,
+      "learning_rate": 1.9891535634605954e-05,
+      "loss": 0.5712,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0807441080391822,
+      "learning_rate": 1.989114862093232e-05,
+      "loss": 0.6179,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0458888781155307,
+      "learning_rate": 1.9890760921810856e-05,
+      "loss": 0.6274,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0511812832675538,
+      "learning_rate": 1.9890372537268433e-05,
+      "loss": 0.6443,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.078728574604796,
+      "learning_rate": 1.988998346733197e-05,
+      "loss": 0.6147,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1166802869475498,
+      "learning_rate": 1.9889593712028422e-05,
+      "loss": 0.6403,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0624234559060666,
+      "learning_rate": 1.9889203271384803e-05,
+      "loss": 0.5563,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0577128119712451,
+      "learning_rate": 1.9888812145428172e-05,
+      "loss": 0.673,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1039049361248896,
+      "learning_rate": 1.9888420334185627e-05,
+      "loss": 0.6499,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2272573165212521,
+      "learning_rate": 1.9888027837684326e-05,
+      "loss": 0.6061,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1160443934529196,
+      "learning_rate": 1.9887634655951464e-05,
+      "loss": 0.6287,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.104634072856478,
+      "learning_rate": 1.988724078901429e-05,
+      "loss": 0.6851,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.105786043699008,
+      "learning_rate": 1.9886846236900102e-05,
+      "loss": 0.7122,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0171741574616147,
+      "learning_rate": 1.9886450999636243e-05,
+      "loss": 0.5775,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2819694827608143,
+      "learning_rate": 1.9886055077250092e-05,
+      "loss": 0.7037,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1271597552993238,
+      "learning_rate": 1.9885658469769094e-05,
+      "loss": 0.5979,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1455028853874523,
+      "learning_rate": 1.9885261177220737e-05,
+      "loss": 0.5962,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0639290716254242,
+      "learning_rate": 1.9884863199632546e-05,
+      "loss": 0.6347,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0144440835622914,
+      "learning_rate": 1.9884464537032103e-05,
+      "loss": 0.6397,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.036460767711225,
+      "learning_rate": 1.9884065189447036e-05,
+      "loss": 0.6134,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0006656100539892,
+      "learning_rate": 1.9883665156905015e-05,
+      "loss": 0.6241,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.058228820011543,
+      "learning_rate": 1.988326443943377e-05,
+      "loss": 0.6417,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1734905485423708,
+      "learning_rate": 1.988286303706106e-05,
+      "loss": 0.626,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1270162617667199,
+      "learning_rate": 1.9882460949814716e-05,
+      "loss": 0.5621,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1408949836430282,
+      "learning_rate": 1.988205817772259e-05,
+      "loss": 0.6351,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1396736021976535,
+      "learning_rate": 1.9881654720812594e-05,
+      "loss": 0.6667,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0169098488256985,
+      "learning_rate": 1.9881250579112694e-05,
+      "loss": 0.6072,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0907079731731821,
+      "learning_rate": 1.9880845752650896e-05,
+      "loss": 0.5655,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.169059792113472,
+      "learning_rate": 1.988044024145525e-05,
+      "loss": 0.6416,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1156233349996894,
+      "learning_rate": 1.9880034045553858e-05,
+      "loss": 0.6269,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1084136434129432,
+      "learning_rate": 1.9879627164974868e-05,
+      "loss": 0.6429,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1176885510223709,
+      "learning_rate": 1.9879219599746486e-05,
+      "loss": 0.6685,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8867395514211373,
+      "learning_rate": 1.987881134989694e-05,
+      "loss": 0.4933,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.990682541637075,
+      "learning_rate": 1.9878402415454534e-05,
+      "loss": 0.6724,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0056387200473818,
+      "learning_rate": 1.9877992796447604e-05,
+      "loss": 0.5682,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1467756136126264,
+      "learning_rate": 1.9877582492904533e-05,
+      "loss": 0.695,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1088054224503603,
+      "learning_rate": 1.987717150485376e-05,
+      "loss": 0.6608,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0436562408251457,
+      "learning_rate": 1.9876759832323756e-05,
+      "loss": 0.6355,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.024013032934522,
+      "learning_rate": 1.9876347475343062e-05,
+      "loss": 0.6029,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.123425037307103,
+      "learning_rate": 1.9875934433940248e-05,
+      "loss": 0.6084,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0661739928360128,
+      "learning_rate": 1.9875520708143933e-05,
+      "loss": 0.6411,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0218598836399087,
+      "learning_rate": 1.9875106297982798e-05,
+      "loss": 0.6291,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1134477148440822,
+      "learning_rate": 1.987469120348555e-05,
+      "loss": 0.6534,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1517210895453271,
+      "learning_rate": 1.9874275424680966e-05,
+      "loss": 0.6704,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1140417226147281,
+      "learning_rate": 1.987385896159785e-05,
+      "loss": 0.621,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.015703441507511,
+      "learning_rate": 1.987344181426507e-05,
+      "loss": 0.6387,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.031557022872771,
+      "learning_rate": 1.987302398271153e-05,
+      "loss": 0.619,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.179749176771803,
+      "learning_rate": 1.987260546696618e-05,
+      "loss": 0.62,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1224651935196714,
+      "learning_rate": 1.987218626705803e-05,
+      "loss": 0.677,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0949746162974316,
+      "learning_rate": 1.9871766383016127e-05,
+      "loss": 0.6365,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0828147418781917,
+      "learning_rate": 1.9871345814869575e-05,
+      "loss": 0.6582,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0417686332976923,
+      "learning_rate": 1.9870924562647512e-05,
+      "loss": 0.6457,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2736955707692121,
+      "learning_rate": 1.9870502626379127e-05,
+      "loss": 0.6502,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0004273125024994,
+      "learning_rate": 1.9870080006093674e-05,
+      "loss": 0.578,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1096665328492759,
+      "learning_rate": 1.9869656701820424e-05,
+      "loss": 0.5603,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2125446185211894,
+      "learning_rate": 1.9869232713588724e-05,
+      "loss": 0.464,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1221372062657058,
+      "learning_rate": 1.9868808041427948e-05,
+      "loss": 0.5887,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.996001225280882,
+      "learning_rate": 1.9868382685367533e-05,
+      "loss": 0.6238,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1367037471800612,
+      "learning_rate": 1.9867956645436944e-05,
+      "loss": 0.6152,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1129390935514154,
+      "learning_rate": 1.9867529921665713e-05,
+      "loss": 0.6287,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2269397679391247,
+      "learning_rate": 1.9867102514083415e-05,
+      "loss": 0.6655,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.26827046275962,
+      "learning_rate": 1.9866674422719666e-05,
+      "loss": 0.6604,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0594102839215263,
+      "learning_rate": 1.9866245647604128e-05,
+      "loss": 0.6356,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0773242595576222,
+      "learning_rate": 1.9865816188766516e-05,
+      "loss": 0.6223,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1461188807936942,
+      "learning_rate": 1.9865386046236597e-05,
+      "loss": 0.6325,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2375769131659358,
+      "learning_rate": 1.9864955220044175e-05,
+      "loss": 0.6041,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1529673106809384,
+      "learning_rate": 1.9864523710219107e-05,
+      "loss": 0.6545,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0715029734939707,
+      "learning_rate": 1.986409151679129e-05,
+      "loss": 0.567,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9652259721065576,
+      "learning_rate": 1.9863658639790686e-05,
+      "loss": 0.622,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1205944623699782,
+      "learning_rate": 1.9863225079247286e-05,
+      "loss": 0.6425,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.251314945813738,
+      "learning_rate": 1.9862790835191137e-05,
+      "loss": 0.6876,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9323839167997477,
+      "learning_rate": 1.9862355907652332e-05,
+      "loss": 0.5943,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9863978580378259,
+      "learning_rate": 1.986192029666101e-05,
+      "loss": 0.5605,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0486628796809216,
+      "learning_rate": 1.9861484002247357e-05,
+      "loss": 0.5606,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1196085211902809,
+      "learning_rate": 1.9861047024441614e-05,
+      "loss": 0.6215,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0658424922162064,
+      "learning_rate": 1.9860609363274056e-05,
+      "loss": 0.6589,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0050420253517978,
+      "learning_rate": 1.9860171018775018e-05,
+      "loss": 0.5239,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0584283466966222,
+      "learning_rate": 1.9859731990974867e-05,
+      "loss": 0.6244,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0585855584799557,
+      "learning_rate": 1.9859292279904043e-05,
+      "loss": 0.5561,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9990564994313156,
+      "learning_rate": 1.9858851885593004e-05,
+      "loss": 0.4815,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0178981238404532,
+      "learning_rate": 1.9858410808072278e-05,
+      "loss": 0.58,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9590816481827416,
+      "learning_rate": 1.9857969047372422e-05,
+      "loss": 0.5601,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0622275660970883,
+      "learning_rate": 1.985752660352406e-05,
+      "loss": 0.5345,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1863129467779967,
+      "learning_rate": 1.9857083476557846e-05,
+      "loss": 0.6909,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.052063234705519,
+      "learning_rate": 1.9856639666504492e-05,
+      "loss": 0.5748,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.002701005360324,
+      "learning_rate": 1.9856195173394754e-05,
+      "loss": 0.589,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.938763483372456,
+      "learning_rate": 1.985574999725943e-05,
+      "loss": 0.5357,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0966699932094541,
+      "learning_rate": 1.985530413812937e-05,
+      "loss": 0.6467,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9847762667527534,
+      "learning_rate": 1.9854857596035476e-05,
+      "loss": 0.636,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1229985693999278,
+      "learning_rate": 1.9854410371008693e-05,
+      "loss": 0.6408,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0914436618280507,
+      "learning_rate": 1.9853962463080013e-05,
+      "loss": 0.6649,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.063621268903449,
+      "learning_rate": 1.9853513872280476e-05,
+      "loss": 0.6554,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0312095056964494,
+      "learning_rate": 1.985306459864117e-05,
+      "loss": 0.6262,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.117013859147517,
+      "learning_rate": 1.985261464219322e-05,
+      "loss": 0.6764,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.036906745742879,
+      "learning_rate": 1.9852164002967818e-05,
+      "loss": 0.6228,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0208147244129753,
+      "learning_rate": 1.9851712680996188e-05,
+      "loss": 0.6193,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9589392180387819,
+      "learning_rate": 1.985126067630961e-05,
+      "loss": 0.6018,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1675221592473077,
+      "learning_rate": 1.9850807988939405e-05,
+      "loss": 0.6075,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0544466663271506,
+      "learning_rate": 1.9850354618916942e-05,
+      "loss": 0.6635,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0481367773090586,
+      "learning_rate": 1.9849900566273642e-05,
+      "loss": 0.5868,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.969942783915277,
+      "learning_rate": 1.984944583104097e-05,
+      "loss": 0.5484,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9991758991063624,
+      "learning_rate": 1.9848990413250436e-05,
+      "loss": 0.5921,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9426679105129601,
+      "learning_rate": 1.9848534312933606e-05,
+      "loss": 0.5454,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.108668062134207,
+      "learning_rate": 1.9848077530122083e-05,
+      "loss": 0.6393,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.163748991275181,
+      "learning_rate": 1.9847620064847522e-05,
+      "loss": 0.6919,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1011142742465692,
+      "learning_rate": 1.9847161917141626e-05,
+      "loss": 0.6098,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1603642028495122,
+      "learning_rate": 1.984670308703614e-05,
+      "loss": 0.6514,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9787388127791858,
+      "learning_rate": 1.9846243574562866e-05,
+      "loss": 0.5918,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1162496201649623,
+      "learning_rate": 1.9845783379753648e-05,
+      "loss": 0.627,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.104476587435251,
+      "learning_rate": 1.9845322502640374e-05,
+      "loss": 0.5546,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1535168566665752,
+      "learning_rate": 1.9844860943254983e-05,
+      "loss": 0.6147,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1166245608533367,
+      "learning_rate": 1.984439870162946e-05,
+      "loss": 0.6482,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0427963654809782,
+      "learning_rate": 1.984393577779584e-05,
+      "loss": 0.5811,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0767589535184312,
+      "learning_rate": 1.9843472171786204e-05,
+      "loss": 0.7003,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9479657664193446,
+      "learning_rate": 1.9843007883632674e-05,
+      "loss": 0.5707,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0463700626631678,
+      "learning_rate": 1.984254291336743e-05,
+      "loss": 0.6024,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0773084022527417,
+      "learning_rate": 1.984207726102269e-05,
+      "loss": 0.6155,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0902382782675282,
+      "learning_rate": 1.984161092663073e-05,
+      "loss": 0.5474,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9644531785730229,
+      "learning_rate": 1.984114391022386e-05,
+      "loss": 0.529,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.127131720367706,
+      "learning_rate": 1.984067621183445e-05,
+      "loss": 0.5964,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.094880066095062,
+      "learning_rate": 1.9840207831494903e-05,
+      "loss": 0.6238,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1393198512420912,
+      "learning_rate": 1.983973876923768e-05,
+      "loss": 0.7522,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0059937311188745,
+      "learning_rate": 1.9839269025095293e-05,
+      "loss": 0.5853,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4109933801849883,
+      "learning_rate": 1.9838798599100286e-05,
+      "loss": 0.6001,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1132444057463866,
+      "learning_rate": 1.9838327491285266e-05,
+      "loss": 0.5767,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9701633973894861,
+      "learning_rate": 1.9837855701682875e-05,
+      "loss": 0.5808,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1247541865700246,
+      "learning_rate": 1.983738323032581e-05,
+      "loss": 0.6039,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0092591983705888,
+      "learning_rate": 1.9836910077246813e-05,
+      "loss": 0.6334,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9705654534952596,
+      "learning_rate": 1.9836436242478676e-05,
+      "loss": 0.5847,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9904458846833634,
+      "learning_rate": 1.9835961726054228e-05,
+      "loss": 0.6517,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1197036023727012,
+      "learning_rate": 1.983548652800636e-05,
+      "loss": 0.6023,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1771133490975412,
+      "learning_rate": 1.9835010648368e-05,
+      "loss": 0.6963,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0699175937730134,
+      "learning_rate": 1.9834534087172126e-05,
+      "loss": 0.6147,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1551718492888678,
+      "learning_rate": 1.983405684445176e-05,
+      "loss": 0.6549,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0603070951538438,
+      "learning_rate": 1.983357892023998e-05,
+      "loss": 0.6094,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0402062222116446,
+      "learning_rate": 1.98331003145699e-05,
+      "loss": 0.6219,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9939060911914949,
+      "learning_rate": 1.983262102747469e-05,
+      "loss": 0.6104,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9774031117289627,
+      "learning_rate": 1.983214105898757e-05,
+      "loss": 0.5856,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0914284577741291,
+      "learning_rate": 1.983166040914179e-05,
+      "loss": 0.5925,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0680432076381006,
+      "learning_rate": 1.983117907797067e-05,
+      "loss": 0.682,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0942550327326452,
+      "learning_rate": 1.9830697065507554e-05,
+      "loss": 0.5849,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0321035757944577,
+      "learning_rate": 1.9830214371785858e-05,
+      "loss": 0.5567,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0419076031734429,
+      "learning_rate": 1.982973099683902e-05,
+      "loss": 0.6708,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.023472160735429,
+      "learning_rate": 1.9829246940700543e-05,
+      "loss": 0.6217,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9986553016654867,
+      "learning_rate": 1.9828762203403973e-05,
+      "loss": 0.5932,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.078259151860152,
+      "learning_rate": 1.98282767849829e-05,
+      "loss": 0.6501,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9435008967700403,
+      "learning_rate": 1.9827790685470963e-05,
+      "loss": 0.5278,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.058301399489676,
+      "learning_rate": 1.9827303904901853e-05,
+      "loss": 0.6602,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1175064502389007,
+      "learning_rate": 1.9826816443309294e-05,
+      "loss": 0.6301,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.99965556015855,
+      "learning_rate": 1.9826328300727074e-05,
+      "loss": 0.6088,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1386141728126398,
+      "learning_rate": 1.9825839477189017e-05,
+      "loss": 0.676,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.082303668746878,
+      "learning_rate": 1.9825349972729003e-05,
+      "loss": 0.6496,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9824070847379709,
+      "learning_rate": 1.982485978738095e-05,
+      "loss": 0.6093,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0204342899951846,
+      "learning_rate": 1.9824368921178825e-05,
+      "loss": 0.6787,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.004722031213302,
+      "learning_rate": 1.9823877374156647e-05,
+      "loss": 0.6069,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.036218483815217,
+      "learning_rate": 1.9823385146348485e-05,
+      "loss": 0.6353,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0611860755386595,
+      "learning_rate": 1.9822892237788448e-05,
+      "loss": 0.5991,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1588271079524333,
+      "learning_rate": 1.9822398648510684e-05,
+      "loss": 0.7502,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.03100217892194,
+      "learning_rate": 1.9821904378549414e-05,
+      "loss": 0.5934,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.052827521858649,
+      "learning_rate": 1.9821409427938878e-05,
+      "loss": 0.6325,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9408025909515155,
+      "learning_rate": 1.982091379671338e-05,
+      "loss": 0.5921,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1543920923976936,
+      "learning_rate": 1.982041748490727e-05,
+      "loss": 0.6921,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0745995961987254,
+      "learning_rate": 1.9819920492554935e-05,
+      "loss": 0.651,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.101414274214897,
+      "learning_rate": 1.9819422819690824e-05,
+      "loss": 0.6983,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0657813614519533,
+      "learning_rate": 1.9818924466349422e-05,
+      "loss": 0.6255,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0039151439925784,
+      "learning_rate": 1.981842543256526e-05,
+      "loss": 0.613,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9727132720114751,
+      "learning_rate": 1.981792571837293e-05,
+      "loss": 0.6061,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.089952183620595,
+      "learning_rate": 1.981742532380705e-05,
+      "loss": 0.6576,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1378152254765532,
+      "learning_rate": 1.9816924248902304e-05,
+      "loss": 0.6077,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0000243008860976,
+      "learning_rate": 1.9816422493693417e-05,
+      "loss": 0.6,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0262601717430377,
+      "learning_rate": 1.9815920058215157e-05,
+      "loss": 0.5675,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0424400127184634,
+      "learning_rate": 1.9815416942502346e-05,
+      "loss": 0.6893,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1959462863993415,
+      "learning_rate": 1.9814913146589847e-05,
+      "loss": 0.621,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0532552468562892,
+      "learning_rate": 1.9814408670512572e-05,
+      "loss": 0.6032,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0106259099736372,
+      "learning_rate": 1.981390351430548e-05,
+      "loss": 0.5914,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1403825524665405,
+      "learning_rate": 1.981339767800358e-05,
+      "loss": 0.6111,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.043511667204388,
+      "learning_rate": 1.9812891161641927e-05,
+      "loss": 0.6132,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1476441506456099,
+      "learning_rate": 1.981238396525562e-05,
+      "loss": 0.6988,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0601807753560835,
+      "learning_rate": 1.9811876088879808e-05,
+      "loss": 0.6176,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1079517298243802,
+      "learning_rate": 1.9811367532549686e-05,
+      "loss": 0.5924,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0216589393619098,
+      "learning_rate": 1.9810858296300496e-05,
+      "loss": 0.5809,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1727244067131004,
+      "learning_rate": 1.9810348380167527e-05,
+      "loss": 0.666,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0590507767162933,
+      "learning_rate": 1.9809837784186117e-05,
+      "loss": 0.5926,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0800020752477553,
+      "learning_rate": 1.9809326508391653e-05,
+      "loss": 0.6388,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.113372749895697,
+      "learning_rate": 1.980881455281956e-05,
+      "loss": 0.6648,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1485772029405572,
+      "learning_rate": 1.980830191750532e-05,
+      "loss": 0.6499,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8997588895538492,
+      "learning_rate": 1.980778860248446e-05,
+      "loss": 0.533,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.027167278167877,
+      "learning_rate": 1.9807274607792545e-05,
+      "loss": 0.6515,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0088579416146564,
+      "learning_rate": 1.98067599334652e-05,
+      "loss": 0.5844,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1311769529954938,
+      "learning_rate": 1.980624457953809e-05,
+      "loss": 0.6711,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0829367163761348,
+      "learning_rate": 1.980572854604693e-05,
+      "loss": 0.5857,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1391803149878377,
+      "learning_rate": 1.980521183302748e-05,
+      "loss": 0.6258,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.125268403164557,
+      "learning_rate": 1.980469444051554e-05,
+      "loss": 0.6142,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0778526344981947,
+      "learning_rate": 1.980417636854698e-05,
+      "loss": 0.5555,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.082392494479735,
+      "learning_rate": 1.9803657617157693e-05,
+      "loss": 0.603,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0569189959865126,
+      "learning_rate": 1.9803138186383628e-05,
+      "loss": 0.6449,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0852474738625402,
+      "learning_rate": 1.9802618076260784e-05,
+      "loss": 0.6242,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1175750495002248,
+      "learning_rate": 1.9802097286825197e-05,
+      "loss": 0.6315,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0773717623972658,
+      "learning_rate": 1.9801575818112964e-05,
+      "loss": 0.664,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.162007196070372,
+      "learning_rate": 1.980105367016022e-05,
+      "loss": 0.6763,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.136601052174248,
+      "learning_rate": 1.9800530843003157e-05,
+      "loss": 0.6117,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.143985763323429,
+      "learning_rate": 1.9800007336677994e-05,
+      "loss": 0.6176,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0404379302337705,
+      "learning_rate": 1.979948315122102e-05,
+      "loss": 0.6334,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9958232338449807,
+      "learning_rate": 1.979895828666855e-05,
+      "loss": 0.6101,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0200571531476168,
+      "learning_rate": 1.9798432743056964e-05,
+      "loss": 0.5884,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0324428907512637,
+      "learning_rate": 1.979790652042268e-05,
+      "loss": 0.6089,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1115915371843572,
+      "learning_rate": 1.9797379618802163e-05,
+      "loss": 0.6658,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1328477500696639,
+      "learning_rate": 1.9796852038231932e-05,
+      "loss": 0.6374,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2353642392222939,
+      "learning_rate": 1.9796323778748544e-05,
+      "loss": 0.6538,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.053963539963818,
+      "learning_rate": 1.9795794840388605e-05,
+      "loss": 0.568,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1496538850517588,
+      "learning_rate": 1.9795265223188775e-05,
+      "loss": 0.5619,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1157661180851675,
+      "learning_rate": 1.9794734927185756e-05,
+      "loss": 0.6437,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1599396320995672,
+      "learning_rate": 1.979420395241629e-05,
+      "loss": 0.6507,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.112262206229225,
+      "learning_rate": 1.9793672298917178e-05,
+      "loss": 0.6211,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0729221065115775,
+      "learning_rate": 1.9793139966725264e-05,
+      "loss": 0.5873,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1470954186434454,
+      "learning_rate": 1.9792606955877437e-05,
+      "loss": 0.6533,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0966433183148414,
+      "learning_rate": 1.979207326641063e-05,
+      "loss": 0.6198,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0887800674211725,
+      "learning_rate": 1.979153889836184e-05,
+      "loss": 0.6142,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0505608118632632,
+      "learning_rate": 1.979100385176808e-05,
+      "loss": 0.6201,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0637276276013699,
+      "learning_rate": 1.979046812666644e-05,
+      "loss": 0.6334,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0959557081058233,
+      "learning_rate": 1.9789931723094046e-05,
+      "loss": 0.5997,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1145558485853704,
+      "learning_rate": 1.9789394641088068e-05,
+      "loss": 0.6324,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1470672684736019,
+      "learning_rate": 1.978885688068572e-05,
+      "loss": 0.6474,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.059988951594,
+      "learning_rate": 1.9788318441924276e-05,
+      "loss": 0.6878,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0449670410814962,
+      "learning_rate": 1.9787779324841045e-05,
+      "loss": 0.6136,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1036223223053379,
+      "learning_rate": 1.978723952947339e-05,
+      "loss": 0.4832,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9439899179136536,
+      "learning_rate": 1.9786699055858715e-05,
+      "loss": 0.5875,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1271381575003854,
+      "learning_rate": 1.9786157904034476e-05,
+      "loss": 0.6729,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1559246313644351,
+      "learning_rate": 1.9785616074038177e-05,
+      "loss": 0.6624,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1229234851699839,
+      "learning_rate": 1.978507356590736e-05,
+      "loss": 0.5696,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1417411033525822,
+      "learning_rate": 1.978453037967963e-05,
+      "loss": 0.6108,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9829968484078301,
+      "learning_rate": 1.978398651539262e-05,
+      "loss": 0.6475,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0605076321999238,
+      "learning_rate": 1.9783441973084023e-05,
+      "loss": 0.61,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9719439884646509,
+      "learning_rate": 1.9782896752791576e-05,
+      "loss": 0.574,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0185093398635798,
+      "learning_rate": 1.978235085455306e-05,
+      "loss": 0.6139,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9513968903641321,
+      "learning_rate": 1.9781804278406308e-05,
+      "loss": 0.6326,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0634305191214797,
+      "learning_rate": 1.9781257024389194e-05,
+      "loss": 0.6322,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9513957696631159,
+      "learning_rate": 1.9780709092539647e-05,
+      "loss": 0.6513,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.899895870947211,
+      "learning_rate": 1.9780160482895633e-05,
+      "loss": 0.5656,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0053864478482981,
+      "learning_rate": 1.9779611195495177e-05,
+      "loss": 0.6136,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0556166213683182,
+      "learning_rate": 1.9779061230376334e-05,
+      "loss": 0.5876,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0620148974297947,
+      "learning_rate": 1.9778510587577226e-05,
+      "loss": 0.6286,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0944898098419875,
+      "learning_rate": 1.9777959267136005e-05,
+      "loss": 0.6408,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1262428474198718,
+      "learning_rate": 1.977740726909088e-05,
+      "loss": 0.5868,
+      "step": 1173
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1348189568565432,
+      "learning_rate": 1.9776854593480107e-05,
+      "loss": 0.6325,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0265007969572906,
+      "learning_rate": 1.977630124034198e-05,
+      "loss": 0.6002,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0045863640942072,
+      "learning_rate": 1.9775747209714847e-05,
+      "loss": 0.4731,
+      "step": 1176
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1767076857181433,
+      "learning_rate": 1.9775192501637104e-05,
+      "loss": 0.5284,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0981349566632057,
+      "learning_rate": 1.9774637116147194e-05,
+      "loss": 0.6136,
+      "step": 1178
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1161235671651721,
+      "learning_rate": 1.97740810532836e-05,
+      "loss": 0.6547,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0884324364235982,
+      "learning_rate": 1.9773524313084857e-05,
+      "loss": 0.6237,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1129008869770969,
+      "learning_rate": 1.977296689558955e-05,
+      "loss": 0.6,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.096193804673158,
+      "learning_rate": 1.9772408800836308e-05,
+      "loss": 0.6041,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9666265897405518,
+      "learning_rate": 1.9771850028863802e-05,
+      "loss": 0.5926,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2019818907270623,
+      "learning_rate": 1.977129057971076e-05,
+      "loss": 0.6472,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0893335099998267,
+      "learning_rate": 1.977073045341594e-05,
+      "loss": 0.6338,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0960683097540216,
+      "learning_rate": 1.977016965001817e-05,
+      "loss": 0.6508,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2142205404788615,
+      "learning_rate": 1.9769608169556314e-05,
+      "loss": 0.6288,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0610027884391453,
+      "learning_rate": 1.9769046012069273e-05,
+      "loss": 0.611,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0705797801046106,
+      "learning_rate": 1.9768483177596008e-05,
+      "loss": 0.6007,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9870195777939947,
+      "learning_rate": 1.9767919666175526e-05,
+      "loss": 0.579,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0850603051142014,
+      "learning_rate": 1.976735547784687e-05,
+      "loss": 0.5486,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.976803281526567,
+      "learning_rate": 1.976679061264915e-05,
+      "loss": 0.5985,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0568799468567787,
+      "learning_rate": 1.97662250706215e-05,
+      "loss": 0.5601,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0791769089869665,
+      "learning_rate": 1.9765658851803116e-05,
+      "loss": 0.593,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0301803244650218,
+      "learning_rate": 1.9765091956233235e-05,
+      "loss": 0.5781,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1232153943591479,
+      "learning_rate": 1.9764524383951147e-05,
+      "loss": 0.681,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.978549446797339,
+      "learning_rate": 1.9763956134996176e-05,
+      "loss": 0.6041,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0988170899210346,
+      "learning_rate": 1.9763387209407706e-05,
+      "loss": 0.6253,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9992464023597344,
+      "learning_rate": 1.9762817607225163e-05,
+      "loss": 0.5952,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0648084957537778,
+      "learning_rate": 1.976224732848802e-05,
+      "loss": 0.6382,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0094155962654219,
+      "learning_rate": 1.9761676373235797e-05,
+      "loss": 0.5395,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9813734130095451,
+      "learning_rate": 1.976110474150806e-05,
+      "loss": 0.6242,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9800418269518998,
+      "learning_rate": 1.976053243334442e-05,
+      "loss": 0.5816,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0461807540210366,
+      "learning_rate": 1.975995944878454e-05,
+      "loss": 0.6228,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1634591956582694,
+      "learning_rate": 1.9759385787868128e-05,
+      "loss": 0.642,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0139084190327168,
+      "learning_rate": 1.9758811450634936e-05,
+      "loss": 0.5179,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1264320315929353,
+      "learning_rate": 1.9758236437124768e-05,
+      "loss": 0.6243,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0325778222003517,
+      "learning_rate": 1.975766074737747e-05,
+      "loss": 0.5712,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0593457101808865,
+      "learning_rate": 1.975708438143294e-05,
+      "loss": 0.6367,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1923816644294367,
+      "learning_rate": 1.9756507339331115e-05,
+      "loss": 0.6276,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9729967459849977,
+      "learning_rate": 1.9755929621111985e-05,
+      "loss": 0.6313,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0460689918701482,
+      "learning_rate": 1.9755351226815586e-05,
+      "loss": 0.6645,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9693482523850563,
+      "learning_rate": 1.9754772156482e-05,
+      "loss": 0.5295,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.124750196318656,
+      "learning_rate": 1.9754192410151357e-05,
+      "loss": 0.5686,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.073152082986142,
+      "learning_rate": 1.975361198786383e-05,
+      "loss": 0.6052,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.023364933183271,
+      "learning_rate": 1.9753030889659644e-05,
+      "loss": 0.5647,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.081775326478639,
+      "learning_rate": 1.975244911557907e-05,
+      "loss": 0.632,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1735100931930877,
+      "learning_rate": 1.9751866665662424e-05,
+      "loss": 0.5785,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0741437321863894,
+      "learning_rate": 1.9751283539950065e-05,
+      "loss": 0.6068,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0562432053843551,
+      "learning_rate": 1.9750699738482403e-05,
+      "loss": 0.636,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0989013126993985,
+      "learning_rate": 1.9750115261299903e-05,
+      "loss": 0.6054,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0119053434260734,
+      "learning_rate": 1.9749530108443063e-05,
+      "loss": 0.6065,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.079234448458603,
+      "learning_rate": 1.9748944279952433e-05,
+      "loss": 0.6835,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0915946403546992,
+      "learning_rate": 1.9748357775868615e-05,
+      "loss": 0.6293,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0920504111740896,
+      "learning_rate": 1.9747770596232247e-05,
+      "loss": 0.7179,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1510367657359035,
+      "learning_rate": 1.974718274108402e-05,
+      "loss": 0.6247,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9528179845476008,
+      "learning_rate": 1.974659421046468e-05,
+      "loss": 0.5738,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0980600804946157,
+      "learning_rate": 1.9746005004415004e-05,
+      "loss": 0.603,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.112778681258694,
+      "learning_rate": 1.9745415122975825e-05,
+      "loss": 0.6356,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0704814670074079,
+      "learning_rate": 1.9744824566188027e-05,
+      "loss": 0.6157,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0033594940133337,
+      "learning_rate": 1.9744233334092525e-05,
+      "loss": 0.6019,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2024696465537366,
+      "learning_rate": 1.9743641426730297e-05,
+      "loss": 0.6621,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9820245364504157,
+      "learning_rate": 1.9743048844142364e-05,
+      "loss": 0.6052,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.143158745985614,
+      "learning_rate": 1.9742455586369786e-05,
+      "loss": 0.6646,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2942234573798506,
+      "learning_rate": 1.9741861653453672e-05,
+      "loss": 0.6221,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2151223441956862,
+      "learning_rate": 1.9741267045435193e-05,
+      "loss": 0.6063,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0487514310045238,
+      "learning_rate": 1.9740671762355548e-05,
+      "loss": 0.5818,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0123848172684526,
+      "learning_rate": 1.9740075804255987e-05,
+      "loss": 0.6388,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9570590630526808,
+      "learning_rate": 1.9739479171177816e-05,
+      "loss": 0.6237,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.156696361880616,
+      "learning_rate": 1.9738881863162372e-05,
+      "loss": 0.7091,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0897508472632145,
+      "learning_rate": 1.973828388025106e-05,
+      "loss": 0.6423,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1737063919986686,
+      "learning_rate": 1.9737685222485307e-05,
+      "loss": 0.6774,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0938566735390922,
+      "learning_rate": 1.9737085889906608e-05,
+      "loss": 0.6219,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1437393726188527,
+      "learning_rate": 1.9736485882556495e-05,
+      "loss": 0.6295,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0791457714353598,
+      "learning_rate": 1.9735885200476545e-05,
+      "loss": 0.6326,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.105337478068499,
+      "learning_rate": 1.9735283843708384e-05,
+      "loss": 0.5601,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9900255732554479,
+      "learning_rate": 1.973468181229369e-05,
+      "loss": 0.612,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9638114049500947,
+      "learning_rate": 1.9734079106274185e-05,
+      "loss": 0.5639,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.155607623256461,
+      "learning_rate": 1.9733475725691627e-05,
+      "loss": 0.656,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8552961484865126,
+      "learning_rate": 1.9732871670587835e-05,
+      "loss": 0.5105,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.010264435160565,
+      "learning_rate": 1.973226694100467e-05,
+      "loss": 0.6195,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7056228909411109,
+      "learning_rate": 1.9731661536984038e-05,
+      "loss": 0.4986,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9955221005221241,
+      "learning_rate": 1.9731055458567895e-05,
+      "loss": 0.6386,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3420947012150832,
+      "learning_rate": 1.973044870579824e-05,
+      "loss": 0.5083,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0090464596512863,
+      "learning_rate": 1.972984127871712e-05,
+      "loss": 0.6675,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0806608477579371,
+      "learning_rate": 1.972923317736663e-05,
+      "loss": 0.6017,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.12514539614613,
+      "learning_rate": 1.9728624401788908e-05,
+      "loss": 0.625,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0903568183651684,
+      "learning_rate": 1.972801495202615e-05,
+      "loss": 0.6665,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9965450584629888,
+      "learning_rate": 1.972740482812058e-05,
+      "loss": 0.589,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0261512414679927,
+      "learning_rate": 1.9726794030114484e-05,
+      "loss": 0.4879,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2427966939144413,
+      "learning_rate": 1.972618255805019e-05,
+      "loss": 0.6667,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.101568478078574,
+      "learning_rate": 1.9725570411970074e-05,
+      "loss": 0.6222,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0828180407176138,
+      "learning_rate": 1.972495759191655e-05,
+      "loss": 0.6233,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0487641243984773,
+      "learning_rate": 1.9724344097932097e-05,
+      "loss": 0.6602,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1526684403152978,
+      "learning_rate": 1.972372993005922e-05,
+      "loss": 0.6448,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.029942884679848,
+      "learning_rate": 1.9723115088340483e-05,
+      "loss": 0.6205,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0887794032526703,
+      "learning_rate": 1.9722499572818496e-05,
+      "loss": 0.5804,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.315334282162586,
+      "learning_rate": 1.972188338353591e-05,
+      "loss": 0.6214,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1271196813059783,
+      "learning_rate": 1.9721266520535435e-05,
+      "loss": 0.6682,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.163495523822124,
+      "learning_rate": 1.972064898385981e-05,
+      "loss": 0.6164,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0811227308269724,
+      "learning_rate": 1.972003077355183e-05,
+      "loss": 0.6423,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0236819263499997,
+      "learning_rate": 1.971941188965434e-05,
+      "loss": 0.6196,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0352641260022146,
+      "learning_rate": 1.971879233221023e-05,
+      "loss": 0.6194,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0848684317413644,
+      "learning_rate": 1.971817210126243e-05,
+      "loss": 0.6133,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0566291958722285,
+      "learning_rate": 1.9717551196853925e-05,
+      "loss": 0.5973,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0355223511760434,
+      "learning_rate": 1.9716929619027734e-05,
+      "loss": 0.5948,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1732139139098121,
+      "learning_rate": 1.971630736782695e-05,
+      "loss": 0.7294,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9992577370321353,
+      "learning_rate": 1.9715684443294677e-05,
+      "loss": 0.6432,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0221808293116248,
+      "learning_rate": 1.971506084547409e-05,
+      "loss": 0.5805,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0768858195353592,
+      "learning_rate": 1.9714436574408408e-05,
+      "loss": 0.6353,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9226845697587781,
+      "learning_rate": 1.9713811630140885e-05,
+      "loss": 0.592,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0268897107440507,
+      "learning_rate": 1.971318601271483e-05,
+      "loss": 0.6324,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0428076161671564,
+      "learning_rate": 1.9712559722173602e-05,
+      "loss": 0.6377,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0859639624879796,
+      "learning_rate": 1.9711932758560604e-05,
+      "loss": 0.6022,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0210349956207485,
+      "learning_rate": 1.971130512191928e-05,
+      "loss": 0.6474,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0599033695812297,
+      "learning_rate": 1.971067681229312e-05,
+      "loss": 0.6067,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.120979412441857,
+      "learning_rate": 1.971004782972567e-05,
+      "loss": 0.6269,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2802519696010277,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.547,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0022250775277506,
+      "learning_rate": 1.9708787845941306e-05,
+      "loss": 0.6117,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0787715179894646,
+      "learning_rate": 1.97081568448117e-05,
+      "loss": 0.6176,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0037129554011988,
+      "learning_rate": 1.970752517091544e-05,
+      "loss": 0.6323,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0334591262363808,
+      "learning_rate": 1.9706892824296297e-05,
+      "loss": 0.6304,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0374628218351345,
+      "learning_rate": 1.9706259804998093e-05,
+      "loss": 0.4893,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9865355112514083,
+      "learning_rate": 1.970562611306469e-05,
+      "loss": 0.5467,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.116948900462893,
+      "learning_rate": 1.9704991748540004e-05,
+      "loss": 0.611,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.084751087051627,
+      "learning_rate": 1.9704356711468e-05,
+      "loss": 0.644,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1208899298642612,
+      "learning_rate": 1.9703721001892685e-05,
+      "loss": 0.6399,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0538062893584283,
+      "learning_rate": 1.9703084619858112e-05,
+      "loss": 0.6364,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9944067068990796,
+      "learning_rate": 1.9702447565408382e-05,
+      "loss": 0.6172,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.022208238295205,
+      "learning_rate": 1.970180983858764e-05,
+      "loss": 0.6118,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0199055126077872,
+      "learning_rate": 1.970117143944008e-05,
+      "loss": 0.6156,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.982700829054324,
+      "learning_rate": 1.9700532368009947e-05,
+      "loss": 0.5647,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0121386373965509,
+      "learning_rate": 1.9699892624341527e-05,
+      "loss": 0.604,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9843522129415369,
+      "learning_rate": 1.9699252208479147e-05,
+      "loss": 0.4893,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0805374392292688,
+      "learning_rate": 1.9698611120467196e-05,
+      "loss": 0.6397,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0497480677687236,
+      "learning_rate": 1.9697969360350098e-05,
+      "loss": 0.5588,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0958726599081376,
+      "learning_rate": 1.9697326928172323e-05,
+      "loss": 0.6535,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9666448228023902,
+      "learning_rate": 1.9696683823978392e-05,
+      "loss": 0.576,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9946083522677421,
+      "learning_rate": 1.9696040047812874e-05,
+      "loss": 0.6022,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0237237273450226,
+      "learning_rate": 1.9695395599720385e-05,
+      "loss": 0.6396,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0853962147117462,
+      "learning_rate": 1.9694750479745573e-05,
+      "loss": 0.6019,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.002745208425993,
+      "learning_rate": 1.969410468793316e-05,
+      "loss": 0.576,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.033603915650126,
+      "learning_rate": 1.9693458224327886e-05,
+      "loss": 0.5707,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0369646792701839,
+      "learning_rate": 1.9692811088974556e-05,
+      "loss": 0.5698,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0388865698456857,
+      "learning_rate": 1.9692163281918016e-05,
+      "loss": 0.623,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9886888943099889,
+      "learning_rate": 1.9691514803203157e-05,
+      "loss": 0.6026,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0760978377547694,
+      "learning_rate": 1.969086565287492e-05,
+      "loss": 0.6156,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0024953962331788,
+      "learning_rate": 1.9690215830978286e-05,
+      "loss": 0.6297,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.092907645473543,
+      "learning_rate": 1.968956533755829e-05,
+      "loss": 0.6497,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0441774948342657,
+      "learning_rate": 1.968891417266001e-05,
+      "loss": 0.6091,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9898586379948117,
+      "learning_rate": 1.9688262336328576e-05,
+      "loss": 0.5755,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9397873413080308,
+      "learning_rate": 1.9687609828609156e-05,
+      "loss": 0.6226,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0412928068520495,
+      "learning_rate": 1.9686956649546964e-05,
+      "loss": 0.5881,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9954215004050639,
+      "learning_rate": 1.9686302799187272e-05,
+      "loss": 0.6102,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1638641323105938,
+      "learning_rate": 1.9685648277575385e-05,
+      "loss": 0.6323,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9463279359139796,
+      "learning_rate": 1.9684993084756664e-05,
+      "loss": 0.5582,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.071442865518808,
+      "learning_rate": 1.9684337220776514e-05,
+      "loss": 0.5199,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0170811281607908,
+      "learning_rate": 1.9683680685680382e-05,
+      "loss": 0.5709,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9828891043357619,
+      "learning_rate": 1.9683023479513768e-05,
+      "loss": 0.6285,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9806341501252827,
+      "learning_rate": 1.968236560232222e-05,
+      "loss": 0.6161,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0613457807345021,
+      "learning_rate": 1.968170705415132e-05,
+      "loss": 0.6188,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0685738980115753,
+      "learning_rate": 1.9681047835046708e-05,
+      "loss": 0.6087,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9840438544734612,
+      "learning_rate": 1.9680387945054073e-05,
+      "loss": 0.5443,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.140764435462584,
+      "learning_rate": 1.9679727384219137e-05,
+      "loss": 0.6209,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0182327607866783,
+      "learning_rate": 1.967906615258768e-05,
+      "loss": 0.6131,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0021196334036369,
+      "learning_rate": 1.9678404250205522e-05,
+      "loss": 0.6491,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9975411361548076,
+      "learning_rate": 1.9677741677118536e-05,
+      "loss": 0.5964,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.924058375002181,
+      "learning_rate": 1.9677078433372635e-05,
+      "loss": 0.6155,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1148488713326163,
+      "learning_rate": 1.9676414519013782e-05,
+      "loss": 0.6719,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0508767522012055,
+      "learning_rate": 1.9675749934087988e-05,
+      "loss": 0.6367,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0723392595965906,
+      "learning_rate": 1.9675084678641303e-05,
+      "loss": 0.615,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0466812905500391,
+      "learning_rate": 1.9674418752719835e-05,
+      "loss": 0.6187,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0192894391011744,
+      "learning_rate": 1.9673752156369726e-05,
+      "loss": 0.6268,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9594859291872142,
+      "learning_rate": 1.9673084889637172e-05,
+      "loss": 0.5834,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0881934969983387,
+      "learning_rate": 1.9672416952568416e-05,
+      "loss": 0.6063,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1647734316504663,
+      "learning_rate": 1.9671748345209746e-05,
+      "loss": 0.6656,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1321357566390124,
+      "learning_rate": 1.9671079067607495e-05,
+      "loss": 0.683,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0163862126726073,
+      "learning_rate": 1.9670409119808042e-05,
+      "loss": 0.6353,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.148257709769176,
+      "learning_rate": 1.9669738501857812e-05,
+      "loss": 0.6746,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0982853234518442,
+      "learning_rate": 1.9669067213803287e-05,
+      "loss": 0.6424,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.126143133493401,
+      "learning_rate": 1.9668395255690975e-05,
+      "loss": 0.686,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9792936415990227,
+      "learning_rate": 1.966772262756745e-05,
+      "loss": 0.6292,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1548354716465838,
+      "learning_rate": 1.966704932947932e-05,
+      "loss": 0.6186,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.000836433337724,
+      "learning_rate": 1.966637536147325e-05,
+      "loss": 0.5942,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0117949869053557,
+      "learning_rate": 1.966570072359594e-05,
+      "loss": 0.5791,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.054083198075337,
+      "learning_rate": 1.966502541589414e-05,
+      "loss": 0.5827,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1281975366177772,
+      "learning_rate": 1.9664349438414656e-05,
+      "loss": 0.6375,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0217925796235703,
+      "learning_rate": 1.9663672791204328e-05,
+      "loss": 0.6147,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1560400317141977,
+      "learning_rate": 1.9662995474310042e-05,
+      "loss": 0.6497,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0824588534286652,
+      "learning_rate": 1.9662317487778745e-05,
+      "loss": 0.6342,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.066511214990864,
+      "learning_rate": 1.9661638831657414e-05,
+      "loss": 0.6056,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1489912524465191,
+      "learning_rate": 1.9660959505993086e-05,
+      "loss": 0.6158,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0074515766780978,
+      "learning_rate": 1.966027951083283e-05,
+      "loss": 0.6348,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0830137168473368,
+      "learning_rate": 1.9659598846223775e-05,
+      "loss": 0.6652,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9800063722365412,
+      "learning_rate": 1.9658917512213084e-05,
+      "loss": 0.5851,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9938348549484941,
+      "learning_rate": 1.9658235508847982e-05,
+      "loss": 0.6333,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9375863809230456,
+      "learning_rate": 1.9657552836175725e-05,
+      "loss": 0.6282,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2414875465032795,
+      "learning_rate": 1.965686949424362e-05,
+      "loss": 0.6691,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9846915989441325,
+      "learning_rate": 1.9656185483099027e-05,
+      "loss": 0.6228,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9855716559708518,
+      "learning_rate": 1.9655500802789342e-05,
+      "loss": 0.5688,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9690623160434885,
+      "learning_rate": 1.9654815453362016e-05,
+      "loss": 0.5783,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0343685845795267,
+      "learning_rate": 1.9654129434864545e-05,
+      "loss": 0.6056,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0140780390900874,
+      "learning_rate": 1.965344274734447e-05,
+      "loss": 0.5836,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0009332862499594,
+      "learning_rate": 1.965275539084937e-05,
+      "loss": 0.6123,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.016702561351821,
+      "learning_rate": 1.9652067365426887e-05,
+      "loss": 0.5728,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0162934957404974,
+      "learning_rate": 1.96513786711247e-05,
+      "loss": 0.6103,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0463167905694821,
+      "learning_rate": 1.9650689307990522e-05,
+      "loss": 0.5792,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0308461139411522,
+      "learning_rate": 1.964999927607214e-05,
+      "loss": 0.5638,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0748914011352861,
+      "learning_rate": 1.9649308575417372e-05,
+      "loss": 0.6312,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0594179627576494,
+      "learning_rate": 1.9648617206074073e-05,
+      "loss": 0.624,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9715447357154672,
+      "learning_rate": 1.9647925168090162e-05,
+      "loss": 0.5731,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0014903173207736,
+      "learning_rate": 1.9647232461513597e-05,
+      "loss": 0.6465,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0461118034809385,
+      "learning_rate": 1.9646539086392376e-05,
+      "loss": 0.6042,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1278315412347864,
+      "learning_rate": 1.9645845042774555e-05,
+      "loss": 0.6311,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9916310533107556,
+      "learning_rate": 1.9645150330708225e-05,
+      "loss": 0.5808,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1730449228013495,
+      "learning_rate": 1.9644454950241532e-05,
+      "loss": 0.6499,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0680519469964425,
+      "learning_rate": 1.9643758901422673e-05,
+      "loss": 0.6125,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0495618373555042,
+      "learning_rate": 1.964306218429987e-05,
+      "loss": 0.6534,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.991210791391256,
+      "learning_rate": 1.964236479892141e-05,
+      "loss": 0.5843,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1017210183429609,
+      "learning_rate": 1.9641666745335626e-05,
+      "loss": 0.6282,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1241495120151075,
+      "learning_rate": 1.9640968023590887e-05,
+      "loss": 0.6856,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.02748779670214,
+      "learning_rate": 1.9640268633735616e-05,
+      "loss": 0.587,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1048813470548953,
+      "learning_rate": 1.963956857581828e-05,
+      "loss": 0.6463,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9678613319611744,
+      "learning_rate": 1.963886784988739e-05,
+      "loss": 0.6009,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9820811076689967,
+      "learning_rate": 1.9638166455991508e-05,
+      "loss": 0.6384,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0235020554571306,
+      "learning_rate": 1.963746439417924e-05,
+      "loss": 0.6096,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0644057126732112,
+      "learning_rate": 1.963676166449924e-05,
+      "loss": 0.6501,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0743452283499784,
+      "learning_rate": 1.9636058267000203e-05,
+      "loss": 0.6937,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0666403475309791,
+      "learning_rate": 1.9635354201730874e-05,
+      "loss": 0.6369,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0795702053351024,
+      "learning_rate": 1.9634649468740048e-05,
+      "loss": 0.5609,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0842891736824818,
+      "learning_rate": 1.963394406807656e-05,
+      "loss": 0.6054,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.232357196608936,
+      "learning_rate": 1.963323799978929e-05,
+      "loss": 0.7096,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0627038461122897,
+      "learning_rate": 1.9632531263927173e-05,
+      "loss": 0.5423,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.158864262721317,
+      "learning_rate": 1.963182386053918e-05,
+      "loss": 0.6784,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1047999928694838,
+      "learning_rate": 1.9631115789674343e-05,
+      "loss": 0.5742,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0348252416708916,
+      "learning_rate": 1.963040705138172e-05,
+      "loss": 0.6467,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.100756898992012,
+      "learning_rate": 1.9629697645710432e-05,
+      "loss": 0.6564,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0134805396263211,
+      "learning_rate": 1.962898757270964e-05,
+      "loss": 0.6013,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1551860789638277,
+      "learning_rate": 1.9628276832428548e-05,
+      "loss": 0.6122,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0523592965335977,
+      "learning_rate": 1.962756542491641e-05,
+      "loss": 0.6456,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0072420388751542,
+      "learning_rate": 1.9626853350222535e-05,
+      "loss": 0.6206,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9389508294024854,
+      "learning_rate": 1.962614060839626e-05,
+      "loss": 0.5318,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9785971976060436,
+      "learning_rate": 1.9625427199486973e-05,
+      "loss": 0.5988,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0565134094275008,
+      "learning_rate": 1.962471312354412e-05,
+      "loss": 0.6017,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0769950453386896,
+      "learning_rate": 1.9623998380617187e-05,
+      "loss": 0.5855,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0424660062990199,
+      "learning_rate": 1.9623282970755702e-05,
+      "loss": 0.6113,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9809188594461099,
+      "learning_rate": 1.9622566894009247e-05,
+      "loss": 0.6357,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0366035114434953,
+      "learning_rate": 1.962185015042744e-05,
+      "loss": 0.5987,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0288494452139783,
+      "learning_rate": 1.962113274005995e-05,
+      "loss": 0.5874,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0099375316753343,
+      "learning_rate": 1.9620414662956494e-05,
+      "loss": 0.5621,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9666975626620397,
+      "learning_rate": 1.9619695919166836e-05,
+      "loss": 0.5965,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9862166627261453,
+      "learning_rate": 1.9618976508740782e-05,
+      "loss": 0.5448,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8706924255202901,
+      "learning_rate": 1.961825643172819e-05,
+      "loss": 0.4272,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9861330858786174,
+      "learning_rate": 1.961753568817896e-05,
+      "loss": 0.6266,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9746193528706869,
+      "learning_rate": 1.9616814278143038e-05,
+      "loss": 0.6087,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0273042877502343,
+      "learning_rate": 1.9616092201670415e-05,
+      "loss": 0.5207,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9975075185558374,
+      "learning_rate": 1.961536945881113e-05,
+      "loss": 0.5777,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.074244733642685,
+      "learning_rate": 1.9614646049615273e-05,
+      "loss": 0.6868,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.019900271457737,
+      "learning_rate": 1.961392197413297e-05,
+      "loss": 0.5854,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0169010906420888,
+      "learning_rate": 1.9613197232414405e-05,
+      "loss": 0.6737,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0616525197387021,
+      "learning_rate": 1.96124718245098e-05,
+      "loss": 0.6442,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0010044272938432,
+      "learning_rate": 1.961174575046942e-05,
+      "loss": 0.5957,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0575345130965779,
+      "learning_rate": 1.9611019010343585e-05,
+      "loss": 0.5818,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1147214676236437,
+      "learning_rate": 1.9610291604182658e-05,
+      "loss": 0.6263,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1387158998687177,
+      "learning_rate": 1.960956353203705e-05,
+      "loss": 0.6204,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2318473099005043,
+      "learning_rate": 1.960883479395721e-05,
+      "loss": 0.5864,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0377594896122029,
+      "learning_rate": 1.9608105389993644e-05,
+      "loss": 0.5611,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9970888793455268,
+      "learning_rate": 1.9607375320196892e-05,
+      "loss": 0.6519,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0175219908144566,
+      "learning_rate": 1.960664458461756e-05,
+      "loss": 0.6332,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0990453693654425,
+      "learning_rate": 1.9605913183306272e-05,
+      "loss": 0.4677,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1027567360059916,
+      "learning_rate": 1.9605181116313725e-05,
+      "loss": 0.641,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9380109470025033,
+      "learning_rate": 1.9604448383690644e-05,
+      "loss": 0.6192,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9869363044160382,
+      "learning_rate": 1.9603714985487813e-05,
+      "loss": 0.6524,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.013485737719734,
+      "learning_rate": 1.9602980921756046e-05,
+      "loss": 0.5573,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9702032466898539,
+      "learning_rate": 1.9602246192546224e-05,
+      "loss": 0.6087,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1239818572944578,
+      "learning_rate": 1.9601510797909257e-05,
+      "loss": 0.6198,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0349287078717646,
+      "learning_rate": 1.9600774737896106e-05,
+      "loss": 0.6169,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0023446170801873,
+      "learning_rate": 1.960003801255778e-05,
+      "loss": 0.5426,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0838020795025674,
+      "learning_rate": 1.959930062194534e-05,
+      "loss": 0.6661,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1445880774155552,
+      "learning_rate": 1.959856256610988e-05,
+      "loss": 0.6241,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0756990406078983,
+      "learning_rate": 1.959782384510255e-05,
+      "loss": 0.596,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0238726201838086,
+      "learning_rate": 1.959708445897454e-05,
+      "loss": 0.6427,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1307731133424648,
+      "learning_rate": 1.9596344407777085e-05,
+      "loss": 0.643,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0285910561385603,
+      "learning_rate": 1.9595603691561477e-05,
+      "loss": 0.6756,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9949420844408791,
+      "learning_rate": 1.9594862310379046e-05,
+      "loss": 0.6079,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.040586161086886,
+      "learning_rate": 1.959412026428117e-05,
+      "loss": 0.5854,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9126335616675043,
+      "learning_rate": 1.959337755331926e-05,
+      "loss": 0.5551,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0176160660649483,
+      "learning_rate": 1.9592634177544803e-05,
+      "loss": 0.5906,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.086794686195696,
+      "learning_rate": 1.9591890137009308e-05,
+      "loss": 0.6345,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0365211797145388,
+      "learning_rate": 1.9591145431764327e-05,
+      "loss": 0.6051,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1849633403345523,
+      "learning_rate": 1.959040006186148e-05,
+      "loss": 0.6648,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0349945536963618,
+      "learning_rate": 1.9589654027352412e-05,
+      "loss": 0.6387,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9929411170861682,
+      "learning_rate": 1.958890732828883e-05,
+      "loss": 0.5947,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9923671261831932,
+      "learning_rate": 1.9588159964722474e-05,
+      "loss": 0.5977,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0723403570973749,
+      "learning_rate": 1.9587411936705135e-05,
+      "loss": 0.6061,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1309137191054779,
+      "learning_rate": 1.9586663244288655e-05,
+      "loss": 0.6172,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0228119546392276,
+      "learning_rate": 1.9585913887524914e-05,
+      "loss": 0.5978,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.055575387496972,
+      "learning_rate": 1.9585163866465847e-05,
+      "loss": 0.6854,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0321171852377182,
+      "learning_rate": 1.958441318116342e-05,
+      "loss": 0.6108,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9265607906492737,
+      "learning_rate": 1.9583661831669664e-05,
+      "loss": 0.5967,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1081882645274406,
+      "learning_rate": 1.9582909818036648e-05,
+      "loss": 0.5702,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.03719923918571,
+      "learning_rate": 1.9582157140316472e-05,
+      "loss": 0.5178,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.119059131475791,
+      "learning_rate": 1.9581403798561314e-05,
+      "loss": 0.5767,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0311596675673023,
+      "learning_rate": 1.9580649792823368e-05,
+      "loss": 0.61,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0583962918596255,
+      "learning_rate": 1.957989512315489e-05,
+      "loss": 0.579,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.084334762026287,
+      "learning_rate": 1.957913978960818e-05,
+      "loss": 0.6214,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0887905371592597,
+      "learning_rate": 1.9578383792235573e-05,
+      "loss": 0.5538,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.057229505151521,
+      "learning_rate": 1.957762713108947e-05,
+      "loss": 0.5807,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.081309748615656,
+      "learning_rate": 1.95768698062223e-05,
+      "loss": 0.7256,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0835482753095111,
+      "learning_rate": 1.957611181768655e-05,
+      "loss": 0.6008,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9472949645396779,
+      "learning_rate": 1.957535316553474e-05,
+      "loss": 0.6263,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0463220262272768,
+      "learning_rate": 1.9574593849819453e-05,
+      "loss": 0.5526,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0668226136034622,
+      "learning_rate": 1.9573833870593307e-05,
+      "loss": 0.6201,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2624562734003215,
+      "learning_rate": 1.957307322790896e-05,
+      "loss": 0.6532,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0860664742619095,
+      "learning_rate": 1.9572311921819135e-05,
+      "loss": 0.5966,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0627301848870327,
+      "learning_rate": 1.957154995237658e-05,
+      "loss": 0.656,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0151482279119024,
+      "learning_rate": 1.9570787319634107e-05,
+      "loss": 0.6252,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0381758566758736,
+      "learning_rate": 1.957002402364456e-05,
+      "loss": 0.5607,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.318043710299476,
+      "learning_rate": 1.9569260064460837e-05,
+      "loss": 0.7168,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0484950881712423,
+      "learning_rate": 1.9568495442135878e-05,
+      "loss": 0.5741,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.257735140002447,
+      "learning_rate": 1.9567730156722672e-05,
+      "loss": 0.6263,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1015010819265607,
+      "learning_rate": 1.9566964208274254e-05,
+      "loss": 0.5963,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9913609664622401,
+      "learning_rate": 1.9566197596843702e-05,
+      "loss": 0.5649,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0272315395873561,
+      "learning_rate": 1.956543032248414e-05,
+      "loss": 0.5699,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0427619537229365,
+      "learning_rate": 1.9564662385248743e-05,
+      "loss": 0.5767,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0631391429171742,
+      "learning_rate": 1.9563893785190728e-05,
+      "loss": 0.6465,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9870702083748742,
+      "learning_rate": 1.9563124522363357e-05,
+      "loss": 0.6024,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2315661829412479,
+      "learning_rate": 1.9562354596819938e-05,
+      "loss": 0.6791,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1144355699428936,
+      "learning_rate": 1.9561584008613826e-05,
+      "loss": 0.6348,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1130939812444343,
+      "learning_rate": 1.9560812757798423e-05,
+      "loss": 0.6291,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2017247690800552,
+      "learning_rate": 1.956004084442718e-05,
+      "loss": 0.6411,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.126123159735174,
+      "learning_rate": 1.955926826855358e-05,
+      "loss": 0.6034,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.003296902988338,
+      "learning_rate": 1.9558495030231174e-05,
+      "loss": 0.6248,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9657714590592112,
+      "learning_rate": 1.9557721129513538e-05,
+      "loss": 0.5877,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9224512012672073,
+      "learning_rate": 1.9556946566454308e-05,
+      "loss": 0.5968,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.032329308432359,
+      "learning_rate": 1.9556171341107152e-05,
+      "loss": 0.5938,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0343115532513394,
+      "learning_rate": 1.9555395453525806e-05,
+      "loss": 0.5759,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1010501401367896,
+      "learning_rate": 1.9554618903764026e-05,
+      "loss": 0.6259,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0039742973115253,
+      "learning_rate": 1.9553841691875632e-05,
+      "loss": 0.655,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.100021199901766,
+      "learning_rate": 1.9553063817914482e-05,
+      "loss": 0.4912,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0913235913602368,
+      "learning_rate": 1.9552285281934484e-05,
+      "loss": 0.5687,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1085906438760087,
+      "learning_rate": 1.9551506083989592e-05,
+      "loss": 0.6476,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0239746332752369,
+      "learning_rate": 1.9550726224133795e-05,
+      "loss": 0.6609,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0811194986543498,
+      "learning_rate": 1.9549945702421144e-05,
+      "loss": 0.5572,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1327547585506779,
+      "learning_rate": 1.9549164518905727e-05,
+      "loss": 0.6541,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1449440442891867,
+      "learning_rate": 1.954838267364168e-05,
+      "loss": 0.6387,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0080649164608988,
+      "learning_rate": 1.9547600166683184e-05,
+      "loss": 0.6361,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0634607963289726,
+      "learning_rate": 1.954681699808446e-05,
+      "loss": 0.58,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0450288464842377,
+      "learning_rate": 1.9546033167899788e-05,
+      "loss": 0.6228,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9794147398904083,
+      "learning_rate": 1.9545248676183486e-05,
+      "loss": 0.646,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9439257970714194,
+      "learning_rate": 1.9544463522989917e-05,
+      "loss": 0.6007,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9593080154837881,
+      "learning_rate": 1.9543677708373496e-05,
+      "loss": 0.5715,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0550240084527236,
+      "learning_rate": 1.954289123238867e-05,
+      "loss": 0.698,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.010980510854376,
+      "learning_rate": 1.9542104095089946e-05,
+      "loss": 0.5256,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0780155128709732,
+      "learning_rate": 1.9541316296531875e-05,
+      "loss": 0.6156,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0783401810137738,
+      "learning_rate": 1.9540527836769047e-05,
+      "loss": 0.6011,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0220845234322393,
+      "learning_rate": 1.95397387158561e-05,
+      "loss": 0.6436,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.006269151060652,
+      "learning_rate": 1.9538948933847727e-05,
+      "loss": 0.5765,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0502520128791262,
+      "learning_rate": 1.953815849079865e-05,
+      "loss": 0.5935,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0099151207552917,
+      "learning_rate": 1.953736738676365e-05,
+      "loss": 0.5783,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0115633381309157,
+      "learning_rate": 1.9536575621797546e-05,
+      "loss": 0.7009,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0314261395753932,
+      "learning_rate": 1.9535783195955215e-05,
+      "loss": 0.625,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0305413122838198,
+      "learning_rate": 1.9534990109291568e-05,
+      "loss": 0.6119,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0589889413813205,
+      "learning_rate": 1.953419636186156e-05,
+      "loss": 0.6151,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0537572089897898,
+      "learning_rate": 1.9533401953720204e-05,
+      "loss": 0.732,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1561489357870232,
+      "learning_rate": 1.9532606884922547e-05,
+      "loss": 0.6193,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.036129485903809,
+      "learning_rate": 1.953181115552369e-05,
+      "loss": 0.6357,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9621713367722898,
+      "learning_rate": 1.9531014765578774e-05,
+      "loss": 0.5871,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.094823343969492,
+      "learning_rate": 1.9530217715142987e-05,
+      "loss": 0.6485,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9785272655837435,
+      "learning_rate": 1.9529420004271568e-05,
+      "loss": 0.5039,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.083586772678874,
+      "learning_rate": 1.9528621633019792e-05,
+      "loss": 0.6447,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0002925976491284,
+      "learning_rate": 1.952782260144299e-05,
+      "loss": 0.536,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9262307923258066,
+      "learning_rate": 1.9527022909596537e-05,
+      "loss": 0.54,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9800418846378972,
+      "learning_rate": 1.9526222557535842e-05,
+      "loss": 0.6442,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9628400126389941,
+      "learning_rate": 1.9525421545316378e-05,
+      "loss": 0.5566,
+      "step": 1545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9428595296385236,
+      "learning_rate": 1.9524619872993648e-05,
+      "loss": 0.5485,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1165341782891196,
+      "learning_rate": 1.9523817540623208e-05,
+      "loss": 0.5906,
+      "step": 1547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0112460723323362,
+      "learning_rate": 1.9523014548260657e-05,
+      "loss": 0.635,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0185326086531783,
+      "learning_rate": 1.9522210895961648e-05,
+      "loss": 0.5875,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.00351955353462,
+      "learning_rate": 1.9521406583781872e-05,
+      "loss": 0.6074,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0794945145826804,
+      "learning_rate": 1.9520601611777065e-05,
+      "loss": 0.6069,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9679784995016987,
+      "learning_rate": 1.9519795980003007e-05,
+      "loss": 0.614,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1043952426403454,
+      "learning_rate": 1.9518989688515533e-05,
+      "loss": 0.5914,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1527097920161586,
+      "learning_rate": 1.9518182737370515e-05,
+      "loss": 0.7452,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1059153106910997,
+      "learning_rate": 1.9517375126623882e-05,
+      "loss": 0.6206,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9977261438193731,
+      "learning_rate": 1.9516566856331593e-05,
+      "loss": 0.6321,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0201871612144144,
+      "learning_rate": 1.951575792654966e-05,
+      "loss": 0.5564,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9601699304044047,
+      "learning_rate": 1.9514948337334144e-05,
+      "loss": 0.4204,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9655527593749033,
+      "learning_rate": 1.9514138088741146e-05,
+      "loss": 0.5994,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.021938738185002,
+      "learning_rate": 1.951332718082682e-05,
+      "loss": 0.6092,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1164618252129554,
+      "learning_rate": 1.9512515613647358e-05,
+      "loss": 0.6469,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9661864554484725,
+      "learning_rate": 1.9511703387259e-05,
+      "loss": 0.543,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9647944075014523,
+      "learning_rate": 1.9510890501718037e-05,
+      "loss": 0.5907,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0424734503862871,
+      "learning_rate": 1.95100769570808e-05,
+      "loss": 0.6481,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0588923703277024,
+      "learning_rate": 1.9509262753403656e-05,
+      "loss": 0.65,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8810729326303413,
+      "learning_rate": 1.950844789074305e-05,
+      "loss": 0.5341,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0265294325695518,
+      "learning_rate": 1.950763236915543e-05,
+      "loss": 0.6191,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.059898351367038,
+      "learning_rate": 1.9506816188697322e-05,
+      "loss": 0.6289,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0009322743834865,
+      "learning_rate": 1.950599934942529e-05,
+      "loss": 0.5897,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.061526486833266,
+      "learning_rate": 1.9505181851395928e-05,
+      "loss": 0.6192,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.081852926219257,
+      "learning_rate": 1.9504363694665897e-05,
+      "loss": 0.6464,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1260267561080988,
+      "learning_rate": 1.9503544879291893e-05,
+      "loss": 0.6535,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1258767221569566,
+      "learning_rate": 1.950272540533066e-05,
+      "loss": 0.6928,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0346108077510237,
+      "learning_rate": 1.9501905272838983e-05,
+      "loss": 0.5869,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.149899484431447,
+      "learning_rate": 1.95010844818737e-05,
+      "loss": 0.664,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1020327223722928,
+      "learning_rate": 1.9500263032491688e-05,
+      "loss": 0.6623,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0297737960298794,
+      "learning_rate": 1.9499440924749878e-05,
+      "loss": 0.6437,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9343334803708131,
+      "learning_rate": 1.9498618158705235e-05,
+      "loss": 0.6093,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0287446925236747,
+      "learning_rate": 1.9497794734414782e-05,
+      "loss": 0.5599,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0489118928687258,
+      "learning_rate": 1.9496970651935575e-05,
+      "loss": 0.6262,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1295782786398594,
+      "learning_rate": 1.9496145911324724e-05,
+      "loss": 0.6088,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.042843824083153,
+      "learning_rate": 1.949532051263939e-05,
+      "loss": 0.5938,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1030489445427343,
+      "learning_rate": 1.9494494455936763e-05,
+      "loss": 0.6502,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0155020298359885,
+      "learning_rate": 1.9493667741274093e-05,
+      "loss": 0.5839,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.065572552200458,
+      "learning_rate": 1.9492840368708668e-05,
+      "loss": 0.5917,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.311197032426637,
+      "learning_rate": 1.949201233829783e-05,
+      "loss": 0.5324,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3655083579169545,
+      "learning_rate": 1.9491183650098953e-05,
+      "loss": 0.5474,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.179113938400328,
+      "learning_rate": 1.9490354304169467e-05,
+      "loss": 0.6661,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0957183409700058,
+      "learning_rate": 1.9489524300566845e-05,
+      "loss": 0.6467,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.051786424467458,
+      "learning_rate": 1.948869363934861e-05,
+      "loss": 0.6013,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1924075929286637,
+      "learning_rate": 1.948786232057232e-05,
+      "loss": 0.6942,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0584652636189806,
+      "learning_rate": 1.9487030344295586e-05,
+      "loss": 0.6294,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9447168266155662,
+      "learning_rate": 1.9486197710576063e-05,
+      "loss": 0.5499,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.972937373205438,
+      "learning_rate": 1.9485364419471454e-05,
+      "loss": 0.664,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9792901817040632,
+      "learning_rate": 1.948453047103951e-05,
+      "loss": 0.5629,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0196538962335693,
+      "learning_rate": 1.948369586533801e-05,
+      "loss": 0.61,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0479317330475613,
+      "learning_rate": 1.94828606024248e-05,
+      "loss": 0.6615,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0404865485674981,
+      "learning_rate": 1.948202468235776e-05,
+      "loss": 0.6868,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9845855698699403,
+      "learning_rate": 1.9481188105194827e-05,
+      "loss": 0.6699,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0453875361089522,
+      "learning_rate": 1.948035087099396e-05,
+      "loss": 0.6004,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9423906166078881,
+      "learning_rate": 1.9479512979813193e-05,
+      "loss": 0.5634,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0283084189763398,
+      "learning_rate": 1.947867443171058e-05,
+      "loss": 0.5786,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.141933735094667,
+      "learning_rate": 1.9477835226744243e-05,
+      "loss": 0.5785,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1207636909607657,
+      "learning_rate": 1.9476995364972327e-05,
+      "loss": 0.6612,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1185052250128316,
+      "learning_rate": 1.9476154846453037e-05,
+      "loss": 0.5691,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0395024380287932,
+      "learning_rate": 1.9475313671244624e-05,
+      "loss": 0.6653,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1227322427576973,
+      "learning_rate": 1.9474471839405377e-05,
+      "loss": 0.6205,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0231221460248499,
+      "learning_rate": 1.9473629350993633e-05,
+      "loss": 0.5498,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9523212008119725,
+      "learning_rate": 1.947278620606778e-05,
+      "loss": 0.5849,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.094776684086471,
+      "learning_rate": 1.9471942404686247e-05,
+      "loss": 0.6354,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0211375159615892,
+      "learning_rate": 1.9471097946907506e-05,
+      "loss": 0.6031,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2412258044393085,
+      "learning_rate": 1.947025283279008e-05,
+      "loss": 0.6913,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9912415664046444,
+      "learning_rate": 1.9469407062392528e-05,
+      "loss": 0.5713,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9588341015952317,
+      "learning_rate": 1.946856063577347e-05,
+      "loss": 0.5535,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0301367696445638,
+      "learning_rate": 1.9467713552991557e-05,
+      "loss": 0.6127,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9800192103314544,
+      "learning_rate": 1.9466865814105493e-05,
+      "loss": 0.6247,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9889924807240417,
+      "learning_rate": 1.9466017419174027e-05,
+      "loss": 0.5599,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0075065416062952,
+      "learning_rate": 1.9465168368255946e-05,
+      "loss": 0.6166,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0671090763975322,
+      "learning_rate": 1.9464318661410097e-05,
+      "loss": 0.6176,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.053639753510594,
+      "learning_rate": 1.9463468298695357e-05,
+      "loss": 0.6008,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.043584000165112,
+      "learning_rate": 1.9462617280170657e-05,
+      "loss": 0.5603,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9855988604094579,
+      "learning_rate": 1.9461765605894974e-05,
+      "loss": 0.6136,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9785494268307853,
+      "learning_rate": 1.9460913275927326e-05,
+      "loss": 0.5665,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9185300086052195,
+      "learning_rate": 1.9460060290326784e-05,
+      "loss": 0.526,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9419639343149278,
+      "learning_rate": 1.9459206649152452e-05,
+      "loss": 0.5603,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0604969546819052,
+      "learning_rate": 1.945835235246349e-05,
+      "loss": 0.6027,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9976023956722005,
+      "learning_rate": 1.9457497400319097e-05,
+      "loss": 0.6019,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0103924418233958,
+      "learning_rate": 1.9456641792778527e-05,
+      "loss": 0.565,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0289771525148863,
+      "learning_rate": 1.9455785529901064e-05,
+      "loss": 0.6336,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.958196944777394,
+      "learning_rate": 1.945492861174606e-05,
+      "loss": 0.4424,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.036359306845461,
+      "learning_rate": 1.945407103837288e-05,
+      "loss": 0.6501,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0229971933809134,
+      "learning_rate": 1.9453212809840965e-05,
+      "loss": 0.6873,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.997114764638409,
+      "learning_rate": 1.945235392620979e-05,
+      "loss": 0.4905,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9633455772418739,
+      "learning_rate": 1.9451494387538873e-05,
+      "loss": 0.5667,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0761532151031492,
+      "learning_rate": 1.9450634193887776e-05,
+      "loss": 0.6529,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1922896878104507,
+      "learning_rate": 1.9449773345316113e-05,
+      "loss": 0.6218,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.000698930279899,
+      "learning_rate": 1.944891184188354e-05,
+      "loss": 0.5953,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0527551660365246,
+      "learning_rate": 1.9448049683649753e-05,
+      "loss": 0.6841,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1807838377158926,
+      "learning_rate": 1.9447186870674505e-05,
+      "loss": 0.6784,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2843482015093421,
+      "learning_rate": 1.944632340301759e-05,
+      "loss": 0.5106,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9922564540621734,
+      "learning_rate": 1.9445459280738838e-05,
+      "loss": 0.5037,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.952673106460145,
+      "learning_rate": 1.944459450389814e-05,
+      "loss": 0.5759,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0789735287269921,
+      "learning_rate": 1.9443729072555417e-05,
+      "loss": 0.5833,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9929128562914025,
+      "learning_rate": 1.9442862986770645e-05,
+      "loss": 0.6195,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9939666202490641,
+      "learning_rate": 1.9441996246603848e-05,
+      "loss": 0.6348,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9146036701079484,
+      "learning_rate": 1.9441128852115083e-05,
+      "loss": 0.611,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9358084165527047,
+      "learning_rate": 1.9440260803364463e-05,
+      "loss": 0.603,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9996578346711003,
+      "learning_rate": 1.9439392100412145e-05,
+      "loss": 0.5301,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1106725835561553,
+      "learning_rate": 1.9438522743318327e-05,
+      "loss": 0.581,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0223341153440686,
+      "learning_rate": 1.9437652732143252e-05,
+      "loss": 0.594,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.061253275018964,
+      "learning_rate": 1.9436782066947215e-05,
+      "loss": 0.5873,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0961476565306238,
+      "learning_rate": 1.943591074779055e-05,
+      "loss": 0.6504,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1131513809579108,
+      "learning_rate": 1.9435038774733644e-05,
+      "loss": 0.6113,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9959591458602137,
+      "learning_rate": 1.9434166147836917e-05,
+      "loss": 0.6193,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0898356458340683,
+      "learning_rate": 1.9433292867160843e-05,
+      "loss": 0.6632,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.012741168027966,
+      "learning_rate": 1.9432418932765942e-05,
+      "loss": 0.6281,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9008564402084792,
+      "learning_rate": 1.9431544344712776e-05,
+      "loss": 0.5148,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9579588034766762,
+      "learning_rate": 1.9430669103061953e-05,
+      "loss": 0.5577,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9758486715067173,
+      "learning_rate": 1.9429793207874126e-05,
+      "loss": 0.5811,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9899522905676923,
+      "learning_rate": 1.9428916659209995e-05,
+      "loss": 0.6214,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9120782629931777,
+      "learning_rate": 1.94280394571303e-05,
+      "loss": 0.5678,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9752034280998241,
+      "learning_rate": 1.9427161601695833e-05,
+      "loss": 0.6131,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0109589990611225,
+      "learning_rate": 1.942628309296743e-05,
+      "loss": 0.5384,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1014070080522274,
+      "learning_rate": 1.9425403931005968e-05,
+      "loss": 0.677,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.009005362126033,
+      "learning_rate": 1.9424524115872375e-05,
+      "loss": 0.6068,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.008399993710228,
+      "learning_rate": 1.9423643647627625e-05,
+      "loss": 0.6021,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9525007427813187,
+      "learning_rate": 1.9422762526332723e-05,
+      "loss": 0.5691,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0926859628696797,
+      "learning_rate": 1.942188075204874e-05,
+      "loss": 0.6686,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0722394254713838,
+      "learning_rate": 1.9420998324836777e-05,
+      "loss": 0.5931,
+      "step": 1669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9650163359668696,
+      "learning_rate": 1.9420115244757985e-05,
+      "loss": 0.6437,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.017397111933101,
+      "learning_rate": 1.941923151187356e-05,
+      "loss": 0.5942,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0942653548901369,
+      "learning_rate": 1.9418347126244754e-05,
+      "loss": 0.6195,
+      "step": 1672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0256751446156467,
+      "learning_rate": 1.941746208793284e-05,
+      "loss": 0.5689,
+      "step": 1673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0420200004454279,
+      "learning_rate": 1.9416576396999156e-05,
+      "loss": 0.5454,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1282627502666678,
+      "learning_rate": 1.941569005350508e-05,
+      "loss": 0.6112,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1223142256728877,
+      "learning_rate": 1.941480305751204e-05,
+      "loss": 0.5911,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9574998557547988,
+      "learning_rate": 1.9413915409081496e-05,
+      "loss": 0.5814,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9789856639171853,
+      "learning_rate": 1.9413027108274964e-05,
+      "loss": 0.6239,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9079658859319869,
+      "learning_rate": 1.9412138155154e-05,
+      "loss": 0.5424,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.997675552438307,
+      "learning_rate": 1.941124854978022e-05,
+      "loss": 0.5783,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0240956770351866,
+      "learning_rate": 1.9410358292215252e-05,
+      "loss": 0.68,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0531747697737566,
+      "learning_rate": 1.9409467382520805e-05,
+      "loss": 0.5563,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0106078332022375,
+      "learning_rate": 1.9408575820758616e-05,
+      "loss": 0.6239,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1336686212019484,
+      "learning_rate": 1.940768360699047e-05,
+      "loss": 0.5812,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9427888254007594,
+      "learning_rate": 1.9406790741278188e-05,
+      "loss": 0.6002,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0708900224334923,
+      "learning_rate": 1.940589722368366e-05,
+      "loss": 0.6759,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0544731535821026,
+      "learning_rate": 1.940500305426879e-05,
+      "loss": 0.5946,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.000572268213837,
+      "learning_rate": 1.9404108233095557e-05,
+      "loss": 0.5565,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9469861618458745,
+      "learning_rate": 1.940321276022596e-05,
+      "loss": 0.6116,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.984815886481496,
+      "learning_rate": 1.9402316635722062e-05,
+      "loss": 0.6322,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0360745957580155,
+      "learning_rate": 1.9401419859645958e-05,
+      "loss": 0.5955,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0390331354965414,
+      "learning_rate": 1.9400522432059802e-05,
+      "loss": 0.4928,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.095133833358097,
+      "learning_rate": 1.9399624353025774e-05,
+      "loss": 0.6053,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0481745310455743,
+      "learning_rate": 1.939872562260612e-05,
+      "loss": 0.5901,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9635325259605018,
+      "learning_rate": 1.9397826240863113e-05,
+      "loss": 0.5477,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.056255123418143,
+      "learning_rate": 1.9396926207859085e-05,
+      "loss": 0.5865,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8758001421262852,
+      "learning_rate": 1.939602552365641e-05,
+      "loss": 0.5967,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0623347770055493,
+      "learning_rate": 1.9395124188317493e-05,
+      "loss": 0.5864,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.046069421727186,
+      "learning_rate": 1.9394222201904806e-05,
+      "loss": 0.6322,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0635388569825572,
+      "learning_rate": 1.9393319564480854e-05,
+      "loss": 0.5859,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0396994085777351,
+      "learning_rate": 1.9392416276108192e-05,
+      "loss": 0.6369,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.069197810429597,
+      "learning_rate": 1.9391512336849406e-05,
+      "loss": 0.6187,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9948691983607679,
+      "learning_rate": 1.939060774676715e-05,
+      "loss": 0.6283,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9369154927339403,
+      "learning_rate": 1.9389702505924106e-05,
+      "loss": 0.5582,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9110705801015708,
+      "learning_rate": 1.9388796614383008e-05,
+      "loss": 0.566,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9312148329941633,
+      "learning_rate": 1.938789007220663e-05,
+      "loss": 0.5327,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9391005657013416,
+      "learning_rate": 1.9386982879457795e-05,
+      "loss": 0.5459,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0106676478778127,
+      "learning_rate": 1.9386075036199378e-05,
+      "loss": 0.5732,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1020962008149404,
+      "learning_rate": 1.938516654249428e-05,
+      "loss": 0.6255,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.161045643502798,
+      "learning_rate": 1.9384257398405473e-05,
+      "loss": 0.6387,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0899649970282215,
+      "learning_rate": 1.938334760399595e-05,
+      "loss": 0.6095,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9554662845585481,
+      "learning_rate": 1.9382437159328758e-05,
+      "loss": 0.5472,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0773572457764136,
+      "learning_rate": 1.9381526064466995e-05,
+      "loss": 0.4818,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9980357980012017,
+      "learning_rate": 1.9380614319473798e-05,
+      "loss": 0.5929,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0965656273453313,
+      "learning_rate": 1.9379701924412344e-05,
+      "loss": 0.6584,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9565851250803735,
+      "learning_rate": 1.937878887934587e-05,
+      "loss": 0.6129,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9284045836171315,
+      "learning_rate": 1.9377875184337647e-05,
+      "loss": 0.5907,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1134948767026893,
+      "learning_rate": 1.9376960839450988e-05,
+      "loss": 0.5701,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9792430599995569,
+      "learning_rate": 1.9376045844749267e-05,
+      "loss": 0.6148,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9237201545998925,
+      "learning_rate": 1.937513020029588e-05,
+      "loss": 0.5524,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9695000739742969,
+      "learning_rate": 1.9374213906154286e-05,
+      "loss": 0.6064,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0509066073915099,
+      "learning_rate": 1.9373296962387988e-05,
+      "loss": 0.5842,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0137941184215924,
+      "learning_rate": 1.937237936906052e-05,
+      "loss": 0.5337,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0478769449551844,
+      "learning_rate": 1.9371461126235474e-05,
+      "loss": 0.6209,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0231670900137617,
+      "learning_rate": 1.937054223397649e-05,
+      "loss": 0.6478,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0087171579611938,
+      "learning_rate": 1.9369622692347233e-05,
+      "loss": 0.5918,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0541550223771612,
+      "learning_rate": 1.936870250141144e-05,
+      "loss": 0.7013,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0331248553066827,
+      "learning_rate": 1.936778166123287e-05,
+      "loss": 0.5751,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0075349881213924,
+      "learning_rate": 1.9366860171875345e-05,
+      "loss": 0.5509,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.975520975539256,
+      "learning_rate": 1.9365938033402715e-05,
+      "loss": 0.5705,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0499924966908039,
+      "learning_rate": 1.9365015245878892e-05,
+      "loss": 0.5684,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9711685551889372,
+      "learning_rate": 1.936409180936781e-05,
+      "loss": 0.6005,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.052822654931391,
+      "learning_rate": 1.9363167723933477e-05,
+      "loss": 0.6444,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9489608663159473,
+      "learning_rate": 1.9362242989639926e-05,
+      "loss": 0.518,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0523021470731204,
+      "learning_rate": 1.936131760655124e-05,
+      "loss": 0.506,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9665533177577489,
+      "learning_rate": 1.9360391574731547e-05,
+      "loss": 0.5292,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.005596907178615,
+      "learning_rate": 1.935946489424502e-05,
+      "loss": 0.6312,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0262942511052289,
+      "learning_rate": 1.935853756515588e-05,
+      "loss": 0.6343,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0438665597916685,
+      "learning_rate": 1.9357609587528385e-05,
+      "loss": 0.6709,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0675839441468322,
+      "learning_rate": 1.9356680961426847e-05,
+      "loss": 0.6619,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0499414830717126,
+      "learning_rate": 1.9355751686915617e-05,
+      "loss": 0.6456,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9769748381304898,
+      "learning_rate": 1.9354821764059094e-05,
+      "loss": 0.6537,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9619271758171696,
+      "learning_rate": 1.935389119292172e-05,
+      "loss": 0.6137,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0273752317519276,
+      "learning_rate": 1.9352959973567984e-05,
+      "loss": 0.6247,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0540513963093687,
+      "learning_rate": 1.9352028106062417e-05,
+      "loss": 0.6241,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.019786109985117,
+      "learning_rate": 1.9351095590469596e-05,
+      "loss": 0.5784,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1593224796693433,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.6446,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.003402218113443,
+      "learning_rate": 1.9349228615280736e-05,
+      "loss": 0.6559,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9323779466330664,
+      "learning_rate": 1.9348294155814078e-05,
+      "loss": 0.5848,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9649266600056559,
+      "learning_rate": 1.934735904851892e-05,
+      "loss": 0.5913,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0189695057359012,
+      "learning_rate": 1.9346423293460078e-05,
+      "loss": 0.6156,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9508957159234627,
+      "learning_rate": 1.9345486890702386e-05,
+      "loss": 0.5786,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8771419924453374,
+      "learning_rate": 1.9344549840310743e-05,
+      "loss": 0.5926,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0763350268212386,
+      "learning_rate": 1.9343612142350085e-05,
+      "loss": 0.5966,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9946198344432027,
+      "learning_rate": 1.9342673796885395e-05,
+      "loss": 0.6017,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9539611035425803,
+      "learning_rate": 1.93417348039817e-05,
+      "loss": 0.5984,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9344252040983774,
+      "learning_rate": 1.934079516370406e-05,
+      "loss": 0.6351,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0915301423903436,
+      "learning_rate": 1.933985487611761e-05,
+      "loss": 0.6586,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8947065911393781,
+      "learning_rate": 1.93389139412875e-05,
+      "loss": 0.5607,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1073120365984666,
+      "learning_rate": 1.9337972359278935e-05,
+      "loss": 0.5979,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9693167721107211,
+      "learning_rate": 1.9337030130157166e-05,
+      "loss": 0.628,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9625722731315666,
+      "learning_rate": 1.9336087253987495e-05,
+      "loss": 0.5942,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9396578385873692,
+      "learning_rate": 1.9335143730835258e-05,
+      "loss": 0.5917,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9613382312549996,
+      "learning_rate": 1.933419956076584e-05,
+      "loss": 0.5517,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9495156501733618,
+      "learning_rate": 1.933325474384467e-05,
+      "loss": 0.5303,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0290084286898358,
+      "learning_rate": 1.9332309280137227e-05,
+      "loss": 0.6439,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9716484435289793,
+      "learning_rate": 1.933136316970903e-05,
+      "loss": 0.5967,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.968535746654822,
+      "learning_rate": 1.933041641262564e-05,
+      "loss": 0.603,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.045616747126607,
+      "learning_rate": 1.9329469008952668e-05,
+      "loss": 0.6284,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.069043035466113,
+      "learning_rate": 1.932852095875577e-05,
+      "loss": 0.6672,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9707470847011743,
+      "learning_rate": 1.9327572262100642e-05,
+      "loss": 0.5747,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8752792859825459,
+      "learning_rate": 1.9326622919053034e-05,
+      "loss": 0.5589,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9614197721357719,
+      "learning_rate": 1.9325672929678728e-05,
+      "loss": 0.5549,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0074353729006507,
+      "learning_rate": 1.932472229404356e-05,
+      "loss": 0.5593,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9120956947478422,
+      "learning_rate": 1.932377101221341e-05,
+      "loss": 0.5694,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9156991949273016,
+      "learning_rate": 1.9322819084254197e-05,
+      "loss": 0.5573,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9199254380244064,
+      "learning_rate": 1.9321866510231887e-05,
+      "loss": 0.6179,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0861265552584947,
+      "learning_rate": 1.93209132902125e-05,
+      "loss": 0.6715,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1236405448947937,
+      "learning_rate": 1.9319959424262092e-05,
+      "loss": 0.6493,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0467816961232732,
+      "learning_rate": 1.931900491244676e-05,
+      "loss": 0.621,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0688593278210219,
+      "learning_rate": 1.9318049754832656e-05,
+      "loss": 0.6196,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9995772369930963,
+      "learning_rate": 1.9317093951485963e-05,
+      "loss": 0.5881,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0178102088596666,
+      "learning_rate": 1.931613750247293e-05,
+      "loss": 0.6829,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0075909234904905,
+      "learning_rate": 1.9315180407859828e-05,
+      "loss": 0.6581,
+      "step": 1784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0256750596460658,
+      "learning_rate": 1.9314222667712988e-05,
+      "loss": 0.6293,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0177433162472909,
+      "learning_rate": 1.931326428209878e-05,
+      "loss": 0.6093,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0862864909273302,
+      "learning_rate": 1.9312305251083613e-05,
+      "loss": 0.5152,
+      "step": 1787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9256285006173005,
+      "learning_rate": 1.9311345574733958e-05,
+      "loss": 0.5556,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0230687624511863,
+      "learning_rate": 1.9310385253116307e-05,
+      "loss": 0.6125,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.020000318712854,
+      "learning_rate": 1.930942428629722e-05,
+      "loss": 0.593,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.060391790298629,
+      "learning_rate": 1.9308462674343288e-05,
+      "loss": 0.6621,
+      "step": 1791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.96873478565776,
+      "learning_rate": 1.9307500417321154e-05,
+      "loss": 0.6247,
+      "step": 1792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9259799452753706,
+      "learning_rate": 1.930653751529749e-05,
+      "loss": 0.6099,
+      "step": 1793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1069947124278503,
+      "learning_rate": 1.9305573968339032e-05,
+      "loss": 0.6069,
+      "step": 1794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.058441133773948,
+      "learning_rate": 1.930460977651255e-05,
+      "loss": 0.6325,
+      "step": 1795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.977173212643966,
+      "learning_rate": 1.930364493988487e-05,
+      "loss": 0.6179,
+      "step": 1796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0266606354391452,
+      "learning_rate": 1.9302679458522844e-05,
+      "loss": 0.5753,
+      "step": 1797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9743030871356761,
+      "learning_rate": 1.9301713332493386e-05,
+      "loss": 0.5387,
+      "step": 1798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9153132297294244,
+      "learning_rate": 1.930074656186344e-05,
+      "loss": 0.5802,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0851112564975114,
+      "learning_rate": 1.929977914670001e-05,
+      "loss": 0.6124,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0126714996986839,
+      "learning_rate": 1.9298811087070134e-05,
+      "loss": 0.6734,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1123818538124828,
+      "learning_rate": 1.9297842383040898e-05,
+      "loss": 0.7079,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0228876548907846,
+      "learning_rate": 1.9296873034679427e-05,
+      "loss": 0.6336,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9516639256057758,
+      "learning_rate": 1.9295903042052907e-05,
+      "loss": 0.5531,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9704534136451547,
+      "learning_rate": 1.929493240522855e-05,
+      "loss": 0.6174,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9801162091417275,
+      "learning_rate": 1.9293961124273623e-05,
+      "loss": 0.6326,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0425771118361655,
+      "learning_rate": 1.929298919925543e-05,
+      "loss": 0.6636,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9599469190171349,
+      "learning_rate": 1.9292016630241334e-05,
+      "loss": 0.6003,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8707841858381885,
+      "learning_rate": 1.9291043417298723e-05,
+      "loss": 0.5516,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0020013467161477,
+      "learning_rate": 1.9290069560495042e-05,
+      "loss": 0.5347,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.095080749599255,
+      "learning_rate": 1.9289095059897787e-05,
+      "loss": 0.6277,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.090773935898295,
+      "learning_rate": 1.9288119915574485e-05,
+      "loss": 0.6111,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.978967808007286,
+      "learning_rate": 1.9287144127592704e-05,
+      "loss": 0.5623,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9302789832381554,
+      "learning_rate": 1.9286167696020076e-05,
+      "loss": 0.549,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.974690334183085,
+      "learning_rate": 1.9285190620924267e-05,
+      "loss": 0.6169,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9767003754849144,
+      "learning_rate": 1.9284212902372978e-05,
+      "loss": 0.6241,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9907255241813573,
+      "learning_rate": 1.928323454043397e-05,
+      "loss": 0.6229,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1142070623511469,
+      "learning_rate": 1.9282255535175047e-05,
+      "loss": 0.6103,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0248511092891297,
+      "learning_rate": 1.928127588666405e-05,
+      "loss": 0.5623,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.994991335298417,
+      "learning_rate": 1.9280295594968863e-05,
+      "loss": 0.5865,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9812727318761492,
+      "learning_rate": 1.9279314660157423e-05,
+      "loss": 0.6418,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9630743577652199,
+      "learning_rate": 1.927833308229771e-05,
+      "loss": 0.5103,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9996298075093891,
+      "learning_rate": 1.927735086145774e-05,
+      "loss": 0.6354,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9805994014100998,
+      "learning_rate": 1.9276367997705584e-05,
+      "loss": 0.5739,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9010497876657586,
+      "learning_rate": 1.927538449110936e-05,
+      "loss": 0.5258,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9976777723502999,
+      "learning_rate": 1.9274400341737214e-05,
+      "loss": 0.5761,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0130011150060951,
+      "learning_rate": 1.927341554965735e-05,
+      "loss": 0.5522,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0349919103598368,
+      "learning_rate": 1.9272430114938018e-05,
+      "loss": 0.5707,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0269075489815958,
+      "learning_rate": 1.92714440376475e-05,
+      "loss": 0.6424,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9245815228929377,
+      "learning_rate": 1.9270457317854135e-05,
+      "loss": 0.5445,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.042268266511188,
+      "learning_rate": 1.92694699556263e-05,
+      "loss": 0.5685,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0755570703561435,
+      "learning_rate": 1.926848195103242e-05,
+      "loss": 0.7127,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8802799775002621,
+      "learning_rate": 1.926749330414096e-05,
+      "loss": 0.5497,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9749207683721478,
+      "learning_rate": 1.926650401502044e-05,
+      "loss": 0.6509,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9936837660861811,
+      "learning_rate": 1.9265514083739404e-05,
+      "loss": 0.6173,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8941460948798987,
+      "learning_rate": 1.9264523510366463e-05,
+      "loss": 0.5846,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9556447858034413,
+      "learning_rate": 1.9263532294970263e-05,
+      "loss": 0.5685,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9581469049694387,
+      "learning_rate": 1.9262540437619488e-05,
+      "loss": 0.608,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0159830950902855,
+      "learning_rate": 1.926154793838288e-05,
+      "loss": 0.6803,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.006614746978762,
+      "learning_rate": 1.926055479732921e-05,
+      "loss": 0.5681,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9604027364713731,
+      "learning_rate": 1.925956101452731e-05,
+      "loss": 0.606,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0826894078754479,
+      "learning_rate": 1.9258566590046047e-05,
+      "loss": 0.5971,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0223462945685502,
+      "learning_rate": 1.9257571523954328e-05,
+      "loss": 0.5968,
+      "step": 1843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9614839255093992,
+      "learning_rate": 1.9256575816321114e-05,
+      "loss": 0.5565,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.924425557439055,
+      "learning_rate": 1.925557946721541e-05,
+      "loss": 0.5685,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.056866996923417,
+      "learning_rate": 1.9254582476706254e-05,
+      "loss": 0.6002,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9334277386485584,
+      "learning_rate": 1.9253584844862745e-05,
+      "loss": 0.6184,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8767870072335794,
+      "learning_rate": 1.9252586571754013e-05,
+      "loss": 0.5253,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9113482795540835,
+      "learning_rate": 1.925158765744924e-05,
+      "loss": 0.5904,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.964622683320886,
+      "learning_rate": 1.9250588102017643e-05,
+      "loss": 0.6002,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9209021189317579,
+      "learning_rate": 1.92495879055285e-05,
+      "loss": 0.5529,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0517115998700501,
+      "learning_rate": 1.924858706805112e-05,
+      "loss": 0.6084,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0121355327781139,
+      "learning_rate": 1.924758558965486e-05,
+      "loss": 0.5884,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.035522236552528,
+      "learning_rate": 1.924658347040912e-05,
+      "loss": 0.6331,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9098667167283655,
+      "learning_rate": 1.9245580710383344e-05,
+      "loss": 0.5066,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9882838782668055,
+      "learning_rate": 1.924457730964703e-05,
+      "loss": 0.6573,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8170470779204368,
+      "learning_rate": 1.9243573268269706e-05,
+      "loss": 0.5192,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1153812484779042,
+      "learning_rate": 1.9242568586320956e-05,
+      "loss": 0.6242,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.034925508619141,
+      "learning_rate": 1.92415632638704e-05,
+      "loss": 0.6048,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.943648489456906,
+      "learning_rate": 1.9240557300987705e-05,
+      "loss": 0.5746,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0863408480572274,
+      "learning_rate": 1.923955069774259e-05,
+      "loss": 0.7041,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0642094533973365,
+      "learning_rate": 1.9238543454204802e-05,
+      "loss": 0.6411,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0069358328656293,
+      "learning_rate": 1.923753557044415e-05,
+      "loss": 0.6549,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9637795959365351,
+      "learning_rate": 1.9236527046530476e-05,
+      "loss": 0.6205,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9083232656443532,
+      "learning_rate": 1.923551788253367e-05,
+      "loss": 0.5804,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0283659794931066,
+      "learning_rate": 1.923450807852367e-05,
+      "loss": 0.581,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0104856417456372,
+      "learning_rate": 1.9233497634570446e-05,
+      "loss": 0.6503,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.105383745788613,
+      "learning_rate": 1.923248655074403e-05,
+      "loss": 0.7004,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0774461890083162,
+      "learning_rate": 1.923147482711448e-05,
+      "loss": 0.6321,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0318851189457057,
+      "learning_rate": 1.923046246375192e-05,
+      "loss": 0.6791,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9327788418837317,
+      "learning_rate": 1.9229449460726495e-05,
+      "loss": 0.5738,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9803706876908969,
+      "learning_rate": 1.9228435818108408e-05,
+      "loss": 0.6639,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9657285086391221,
+      "learning_rate": 1.9227421535967906e-05,
+      "loss": 0.6011,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9644259081561086,
+      "learning_rate": 1.9226406614375276e-05,
+      "loss": 0.4687,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0937221730243076,
+      "learning_rate": 1.922539105340085e-05,
+      "loss": 0.5926,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0355689113799338,
+      "learning_rate": 1.922437485311501e-05,
+      "loss": 0.6252,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9493040784838319,
+      "learning_rate": 1.9223358013588172e-05,
+      "loss": 0.5499,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9474996681844716,
+      "learning_rate": 1.9222340534890803e-05,
+      "loss": 0.5851,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0338117204194743,
+      "learning_rate": 1.922132241709342e-05,
+      "loss": 0.6586,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9455740270706581,
+      "learning_rate": 1.9220303660266568e-05,
+      "loss": 0.5658,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9766084713327681,
+      "learning_rate": 1.9219284264480854e-05,
+      "loss": 0.5583,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1407050220765134,
+      "learning_rate": 1.9218264229806917e-05,
+      "loss": 0.6909,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0189913748914254,
+      "learning_rate": 1.9217243556315445e-05,
+      "loss": 0.5941,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.005943329132577,
+      "learning_rate": 1.9216222244077173e-05,
+      "loss": 0.6116,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0085992680953428,
+      "learning_rate": 1.921520029316287e-05,
+      "loss": 0.5959,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9530355673098095,
+      "learning_rate": 1.9214177703643365e-05,
+      "loss": 0.5689,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9927108368990538,
+      "learning_rate": 1.9213154475589513e-05,
+      "loss": 0.4917,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.145381087906433,
+      "learning_rate": 1.921213060907223e-05,
+      "loss": 0.6237,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9237290032126951,
+      "learning_rate": 1.921110610416247e-05,
+      "loss": 0.5401,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1816105930140846,
+      "learning_rate": 1.9210080960931224e-05,
+      "loss": 0.6396,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.045015765819876,
+      "learning_rate": 1.920905517944954e-05,
+      "loss": 0.5987,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0438281050104827,
+      "learning_rate": 1.9208028759788496e-05,
+      "loss": 0.6267,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1258432834265213,
+      "learning_rate": 1.920700170201923e-05,
+      "loss": 0.6134,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0806923013990861,
+      "learning_rate": 1.920597400621291e-05,
+      "loss": 0.7003,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0215362002067059,
+      "learning_rate": 1.9204945672440757e-05,
+      "loss": 0.5316,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1174940694287256,
+      "learning_rate": 1.9203916700774035e-05,
+      "loss": 0.688,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0585002502030911,
+      "learning_rate": 1.920288709128405e-05,
+      "loss": 0.5828,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.10077227527449,
+      "learning_rate": 1.920185684404215e-05,
+      "loss": 0.6515,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.047385362560591,
+      "learning_rate": 1.9200825959119736e-05,
+      "loss": 0.6398,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0580131941502218,
+      "learning_rate": 1.9199794436588244e-05,
+      "loss": 0.6211,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0182865825702039,
+      "learning_rate": 1.9198762276519156e-05,
+      "loss": 0.6557,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9618402593349183,
+      "learning_rate": 1.9197729478984003e-05,
+      "loss": 0.5659,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9881111937134388,
+      "learning_rate": 1.9196696044054354e-05,
+      "loss": 0.6162,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9798061834622409,
+      "learning_rate": 1.9195661971801825e-05,
+      "loss": 0.5748,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0083357679462275,
+      "learning_rate": 1.9194627262298082e-05,
+      "loss": 0.5276,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.979514797051948,
+      "learning_rate": 1.9193591915614824e-05,
+      "loss": 0.5788,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.086160667609808,
+      "learning_rate": 1.9192555931823798e-05,
+      "loss": 0.6366,
+      "step": 1907
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9793396668250293,
+      "learning_rate": 1.9191519310996806e-05,
+      "loss": 0.5931,
+      "step": 1908
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9361137077940025,
+      "learning_rate": 1.9190482053205673e-05,
+      "loss": 0.5084,
+      "step": 1909
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9618308068511757,
+      "learning_rate": 1.9189444158522287e-05,
+      "loss": 0.6516,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9732343315606669,
+      "learning_rate": 1.9188405627018573e-05,
+      "loss": 0.5654,
+      "step": 1911
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.132773160557045,
+      "learning_rate": 1.9187366458766497e-05,
+      "loss": 0.6753,
+      "step": 1912
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0402651050839569,
+      "learning_rate": 1.9186326653838075e-05,
+      "loss": 0.6445,
+      "step": 1913
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.996823757390506,
+      "learning_rate": 1.918528621230537e-05,
+      "loss": 0.6372,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.073328738789437,
+      "learning_rate": 1.918424513424047e-05,
+      "loss": 0.6432,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9053958985827631,
+      "learning_rate": 1.918320341971553e-05,
+      "loss": 0.5817,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9971311966062738,
+      "learning_rate": 1.9182161068802742e-05,
+      "loss": 0.548,
+      "step": 1917
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9791716632717283,
+      "learning_rate": 1.9181118081574336e-05,
+      "loss": 0.5843,
+      "step": 1918
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9691858920466895,
+      "learning_rate": 1.918007445810259e-05,
+      "loss": 0.5861,
+      "step": 1919
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0086233222980516,
+      "learning_rate": 1.9179030198459822e-05,
+      "loss": 0.6408,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9223756495219673,
+      "learning_rate": 1.917798530271841e-05,
+      "loss": 0.5596,
+      "step": 1921
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0383476005597099,
+      "learning_rate": 1.9176939770950753e-05,
+      "loss": 0.5413,
+      "step": 1922
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9672186642469593,
+      "learning_rate": 1.917589360322931e-05,
+      "loss": 0.5994,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0489479306416298,
+      "learning_rate": 1.9174846799626584e-05,
+      "loss": 0.5887,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0890574446451857,
+      "learning_rate": 1.9173799360215106e-05,
+      "loss": 0.5113,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8828386718344657,
+      "learning_rate": 1.917275128506747e-05,
+      "loss": 0.5213,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.050030722223755,
+      "learning_rate": 1.9171702574256314e-05,
+      "loss": 0.6337,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9529536751310449,
+      "learning_rate": 1.91706532278543e-05,
+      "loss": 0.5547,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1090583641257683,
+      "learning_rate": 1.916960324593415e-05,
+      "loss": 0.697,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8848474899547881,
+      "learning_rate": 1.9168552628568632e-05,
+      "loss": 0.5998,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9502043834849424,
+      "learning_rate": 1.9167501375830543e-05,
+      "loss": 0.5589,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0050861666142,
+      "learning_rate": 1.9166449487792746e-05,
+      "loss": 0.584,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0436702439254533,
+      "learning_rate": 1.916539696452813e-05,
+      "loss": 0.634,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9459373406054806,
+      "learning_rate": 1.916434380610963e-05,
+      "loss": 0.5895,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.054908650188775,
+      "learning_rate": 1.916329001261024e-05,
+      "loss": 0.568,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9330365256475838,
+      "learning_rate": 1.9162235584102973e-05,
+      "loss": 0.5557,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0055011485384573,
+      "learning_rate": 1.916118052066091e-05,
+      "loss": 0.6304,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.02211232342174,
+      "learning_rate": 1.9160124822357162e-05,
+      "loss": 0.6059,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9944380434012641,
+      "learning_rate": 1.915906848926489e-05,
+      "loss": 0.6317,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9493365884245534,
+      "learning_rate": 1.9158011521457296e-05,
+      "loss": 0.6037,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0219440943167644,
+      "learning_rate": 1.9156953919007625e-05,
+      "loss": 0.6231,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.135401366313444,
+      "learning_rate": 1.915589568198917e-05,
+      "loss": 0.5845,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9750445054134775,
+      "learning_rate": 1.9154836810475266e-05,
+      "loss": 0.5453,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9264585897691986,
+      "learning_rate": 1.9153777304539295e-05,
+      "loss": 0.5591,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0085873685393283,
+      "learning_rate": 1.9152717164254668e-05,
+      "loss": 0.664,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0278958093912496,
+      "learning_rate": 1.915165638969487e-05,
+      "loss": 0.5891,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.064403603956213,
+      "learning_rate": 1.9150594980933392e-05,
+      "loss": 0.5634,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9931787664117413,
+      "learning_rate": 1.9149532938043803e-05,
+      "loss": 0.6159,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0329314277757113,
+      "learning_rate": 1.9148470261099698e-05,
+      "loss": 0.6072,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1039277621455925,
+      "learning_rate": 1.9147406950174715e-05,
+      "loss": 0.5811,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.125424655598283,
+      "learning_rate": 1.9146343005342546e-05,
+      "loss": 0.6605,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0717462179634367,
+      "learning_rate": 1.914527842667692e-05,
+      "loss": 0.6044,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0645384893136742,
+      "learning_rate": 1.914421321425161e-05,
+      "loss": 0.6595,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9880743315660508,
+      "learning_rate": 1.914314736814044e-05,
+      "loss": 0.516,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.990695340025184,
+      "learning_rate": 1.914208088841726e-05,
+      "loss": 0.5579,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0861237703124709,
+      "learning_rate": 1.914101377515599e-05,
+      "loss": 0.5526,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9343367407195712,
+      "learning_rate": 1.9139946028430568e-05,
+      "loss": 0.5441,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0557213241662768,
+      "learning_rate": 1.9138877648314994e-05,
+      "loss": 0.6169,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1145568689622156,
+      "learning_rate": 1.9137808634883307e-05,
+      "loss": 0.5841,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9547072242413849,
+      "learning_rate": 1.9136738988209585e-05,
+      "loss": 0.6044,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0031154617029014,
+      "learning_rate": 1.913566870836796e-05,
+      "loss": 0.6126,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9713050505949087,
+      "learning_rate": 1.913459779543259e-05,
+      "loss": 0.6042,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0593675597978944,
+      "learning_rate": 1.91335262494777e-05,
+      "loss": 0.579,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1283732724343774,
+      "learning_rate": 1.913245407057754e-05,
+      "loss": 0.6801,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8975735142461613,
+      "learning_rate": 1.9131381258806417e-05,
+      "loss": 0.5387,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9882741415747127,
+      "learning_rate": 1.9130307814238672e-05,
+      "loss": 0.5941,
+      "step": 1966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9637366097998488,
+      "learning_rate": 1.912923373694869e-05,
+      "loss": 0.6374,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9766823600750104,
+      "learning_rate": 1.912815902701091e-05,
+      "loss": 0.5561,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9260778389900388,
+      "learning_rate": 1.9127083684499805e-05,
+      "loss": 0.5718,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.091377532249139,
+      "learning_rate": 1.9126007709489896e-05,
+      "loss": 0.5721,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0418857534347066,
+      "learning_rate": 1.912493110205575e-05,
+      "loss": 0.6359,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9657653994400671,
+      "learning_rate": 1.912385386227197e-05,
+      "loss": 0.5676,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9697092686869903,
+      "learning_rate": 1.9122775990213212e-05,
+      "loss": 0.6102,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0133425995852425,
+      "learning_rate": 1.9121697485954168e-05,
+      "loss": 0.6055,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9679797921794574,
+      "learning_rate": 1.912061834956958e-05,
+      "loss": 0.581,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9154155156284548,
+      "learning_rate": 1.911953858113423e-05,
+      "loss": 0.546,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.933799306983249,
+      "learning_rate": 1.9118458180722945e-05,
+      "loss": 0.6079,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.053758507399398,
+      "learning_rate": 1.91173771484106e-05,
+      "loss": 0.5769,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.138191845494953,
+      "learning_rate": 1.9116295484272102e-05,
+      "loss": 0.6154,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9134720259284926,
+      "learning_rate": 1.9115213188382413e-05,
+      "loss": 0.5892,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.059907173398205,
+      "learning_rate": 1.9114130260816534e-05,
+      "loss": 0.6303,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9855166008746266,
+      "learning_rate": 1.9113046701649517e-05,
+      "loss": 0.5789,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0597633161782598,
+      "learning_rate": 1.9111962510956442e-05,
+      "loss": 0.5726,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8120227538838763,
+      "learning_rate": 1.9110877688812452e-05,
+      "loss": 0.5493,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.022270679868322,
+      "learning_rate": 1.9109792235292715e-05,
+      "loss": 0.6067,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0319294759236322,
+      "learning_rate": 1.9108706150472457e-05,
+      "loss": 0.5586,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9043951748653178,
+      "learning_rate": 1.9107619434426944e-05,
+      "loss": 0.6206,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8981673775030506,
+      "learning_rate": 1.9106532087231483e-05,
+      "loss": 0.5786,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9107278405121401,
+      "learning_rate": 1.9105444108961423e-05,
+      "loss": 0.6036,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.044523349649377,
+      "learning_rate": 1.9104355499692166e-05,
+      "loss": 0.633,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0117957050741717,
+      "learning_rate": 1.9103266259499146e-05,
+      "loss": 0.5685,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.965600636959254,
+      "learning_rate": 1.910217638845785e-05,
+      "loss": 0.6139,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0093507104367392,
+      "learning_rate": 1.9101085886643804e-05,
+      "loss": 0.6291,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0634999781981085,
+      "learning_rate": 1.909999475413258e-05,
+      "loss": 0.6767,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9359622179269197,
+      "learning_rate": 1.909890299099979e-05,
+      "loss": 0.5753,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9781846611171376,
+      "learning_rate": 1.9097810597321095e-05,
+      "loss": 0.5734,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9658528853661609,
+      "learning_rate": 1.9096717573172192e-05,
+      "loss": 0.6279,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0353063421004602,
+      "learning_rate": 1.909562391862883e-05,
+      "loss": 0.6234,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8673949782339934,
+      "learning_rate": 1.90945296337668e-05,
+      "loss": 0.5474,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0514599651032068,
+      "learning_rate": 1.909343471866193e-05,
+      "loss": 0.6215,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0759403952207536,
+      "learning_rate": 1.9092339173390108e-05,
+      "loss": 0.6361,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1525565183994508,
+      "learning_rate": 1.909124299802724e-05,
+      "loss": 0.6463,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0991145737955368,
+      "learning_rate": 1.9090146192649293e-05,
+      "loss": 0.6291,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9114300270012013,
+      "learning_rate": 1.9089048757332285e-05,
+      "loss": 0.5504,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1271425098911323,
+      "learning_rate": 1.908795069215226e-05,
+      "loss": 0.6001,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9186969798037613,
+      "learning_rate": 1.9086851997185307e-05,
+      "loss": 0.5815,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9598657717472785,
+      "learning_rate": 1.908575267250757e-05,
+      "loss": 0.6459,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0136639599893646,
+      "learning_rate": 1.9084652718195237e-05,
+      "loss": 0.5951,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9345431131491013,
+      "learning_rate": 1.908355213432453e-05,
+      "loss": 0.5771,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8956818361321761,
+      "learning_rate": 1.9082450920971712e-05,
+      "loss": 0.593,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9994855629040246,
+      "learning_rate": 1.9081349078213105e-05,
+      "loss": 0.6248,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2337518096291753,
+      "learning_rate": 1.908024660612506e-05,
+      "loss": 0.5699,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.938035174519471,
+      "learning_rate": 1.9079143504783982e-05,
+      "loss": 0.6318,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9892323617279899,
+      "learning_rate": 1.9078039774266308e-05,
+      "loss": 0.5445,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.922117987267186,
+      "learning_rate": 1.9076935414648533e-05,
+      "loss": 0.5127,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9604587542539131,
+      "learning_rate": 1.9075830426007184e-05,
+      "loss": 0.5714,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9977282997590355,
+      "learning_rate": 1.9074724808418837e-05,
+      "loss": 0.537,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0716207428072122,
+      "learning_rate": 1.907361856196011e-05,
+      "loss": 0.6096,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0384083043576222,
+      "learning_rate": 1.9072511686707663e-05,
+      "loss": 0.5372,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9594529053049637,
+      "learning_rate": 1.9071404182738206e-05,
+      "loss": 0.5936,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9420925394401806,
+      "learning_rate": 1.9070296050128486e-05,
+      "loss": 0.532,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0147759141041532,
+      "learning_rate": 1.9069187288955296e-05,
+      "loss": 0.5981,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9816266033078141,
+      "learning_rate": 1.9068077899295468e-05,
+      "loss": 0.6499,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9879841579914928,
+      "learning_rate": 1.9066967881225887e-05,
+      "loss": 0.5568,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.058523530099737,
+      "learning_rate": 1.906585723482347e-05,
+      "loss": 0.6412,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9878797709934289,
+      "learning_rate": 1.9064745960165196e-05,
+      "loss": 0.6247,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8936292649973512,
+      "learning_rate": 1.906363405732806e-05,
+      "loss": 0.5601,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.011701229053685,
+      "learning_rate": 1.9062521526389126e-05,
+      "loss": 0.5808,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1187731519116415,
+      "learning_rate": 1.906140836742549e-05,
+      "loss": 0.6699,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0718544915843096,
+      "learning_rate": 1.9060294580514293e-05,
+      "loss": 0.6348,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9462997682760822,
+      "learning_rate": 1.905918016573271e-05,
+      "loss": 0.6116,
+      "step": 2031
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9663357125634172,
+      "learning_rate": 1.9058065123157985e-05,
+      "loss": 0.6202,
+      "step": 2032
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9246530545030645,
+      "learning_rate": 1.905694945286738e-05,
+      "loss": 0.5894,
+      "step": 2033
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9493007285658821,
+      "learning_rate": 1.9055833154938208e-05,
+      "loss": 0.5899,
+      "step": 2034
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9030284864206434,
+      "learning_rate": 1.9054716229447835e-05,
+      "loss": 0.5221,
+      "step": 2035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9601465057030015,
+      "learning_rate": 1.9053598676473656e-05,
+      "loss": 0.6482,
+      "step": 2036
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0732783069225416,
+      "learning_rate": 1.905248049609312e-05,
+      "loss": 0.6566,
+      "step": 2037
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1640059200373662,
+      "learning_rate": 1.9051361688383715e-05,
+      "loss": 0.6609,
+      "step": 2038
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.922087630723464,
+      "learning_rate": 1.9050242253422975e-05,
+      "loss": 0.5481,
+      "step": 2039
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0115892884144417,
+      "learning_rate": 1.9049122191288473e-05,
+      "loss": 0.5152,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9810015503795328,
+      "learning_rate": 1.9048001502057828e-05,
+      "loss": 0.5688,
+      "step": 2041
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9424808502813058,
+      "learning_rate": 1.9046880185808706e-05,
+      "loss": 0.629,
+      "step": 2042
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0318164669425995,
+      "learning_rate": 1.9045758242618813e-05,
+      "loss": 0.6051,
+      "step": 2043
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0273118739479803,
+      "learning_rate": 1.9044635672565898e-05,
+      "loss": 0.5613,
+      "step": 2044
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8830868072185081,
+      "learning_rate": 1.904351247572775e-05,
+      "loss": 0.5698,
+      "step": 2045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1183889457477896,
+      "learning_rate": 1.904238865218221e-05,
+      "loss": 0.4665,
+      "step": 2046
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0225856305965528,
+      "learning_rate": 1.9041264202007158e-05,
+      "loss": 0.5748,
+      "step": 2047
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0694762111194591,
+      "learning_rate": 1.9040139125280517e-05,
+      "loss": 0.5748,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0199752555319277,
+      "learning_rate": 1.9039013422080255e-05,
+      "loss": 0.6081,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9842016088499462,
+      "learning_rate": 1.9037887092484377e-05,
+      "loss": 0.5637,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.023273970692155,
+      "learning_rate": 1.903676013657094e-05,
+      "loss": 0.6317,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.030403375603567,
+      "learning_rate": 1.9035632554418045e-05,
+      "loss": 0.6595,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9432884806250941,
+      "learning_rate": 1.9034504346103825e-05,
+      "loss": 0.6023,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0073480752386745,
+      "learning_rate": 1.9033375511706466e-05,
+      "loss": 0.6112,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.068931655724002,
+      "learning_rate": 1.90322460513042e-05,
+      "loss": 0.6017,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0386163177149474,
+      "learning_rate": 1.9031115964975295e-05,
+      "loss": 0.6221,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0051235548300301,
+      "learning_rate": 1.9029985252798062e-05,
+      "loss": 0.6493,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0869761990452371,
+      "learning_rate": 1.902885391485086e-05,
+      "loss": 0.5659,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9535848607413672,
+      "learning_rate": 1.9027721951212092e-05,
+      "loss": 0.6128,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0412077321045292,
+      "learning_rate": 1.90265893619602e-05,
+      "loss": 0.6478,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0429919366838214,
+      "learning_rate": 1.9025456147173668e-05,
+      "loss": 0.6557,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9722372011141724,
+      "learning_rate": 1.9024322306931035e-05,
+      "loss": 0.585,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.026797061134865,
+      "learning_rate": 1.902318784131087e-05,
+      "loss": 0.661,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9701379700065667,
+      "learning_rate": 1.902205275039179e-05,
+      "loss": 0.6493,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9532383610104006,
+      "learning_rate": 1.902091703425246e-05,
+      "loss": 0.5722,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0321123189440147,
+      "learning_rate": 1.901978069297158e-05,
+      "loss": 0.5949,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9594333455983534,
+      "learning_rate": 1.9018643726627894e-05,
+      "loss": 0.4708,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9797937635347074,
+      "learning_rate": 1.90175061353002e-05,
+      "loss": 0.4845,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.029429760101533,
+      "learning_rate": 1.9016367919067332e-05,
+      "loss": 0.6088,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0400521070299615,
+      "learning_rate": 1.9015229078008163e-05,
+      "loss": 0.6772,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.984440854452079,
+      "learning_rate": 1.9014089612201612e-05,
+      "loss": 0.5406,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0205451452855927,
+      "learning_rate": 1.901294952172665e-05,
+      "loss": 0.5767,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9840334858354977,
+      "learning_rate": 1.901180880666228e-05,
+      "loss": 0.6672,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9427674229758277,
+      "learning_rate": 1.9010667467087554e-05,
+      "loss": 0.5223,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1013136774216972,
+      "learning_rate": 1.9009525503081565e-05,
+      "loss": 0.5849,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0761601500099016,
+      "learning_rate": 1.900838291472345e-05,
+      "loss": 0.5946,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0176861590592503,
+      "learning_rate": 1.900723970209239e-05,
+      "loss": 0.5184,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0460873298295472,
+      "learning_rate": 1.9006095865267605e-05,
+      "loss": 0.5795,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9584819948779129,
+      "learning_rate": 1.9004951404328363e-05,
+      "loss": 0.6414,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.087426505655538,
+      "learning_rate": 1.9003806319353985e-05,
+      "loss": 0.5963,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9346911893859755,
+      "learning_rate": 1.9002660610423808e-05,
+      "loss": 0.5711,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0277683323783546,
+      "learning_rate": 1.9001514277617236e-05,
+      "loss": 0.6062,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9992807356109709,
+      "learning_rate": 1.900036732101371e-05,
+      "loss": 0.6467,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8751132896048749,
+      "learning_rate": 1.8999219740692716e-05,
+      "loss": 0.5005,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9918946212928599,
+      "learning_rate": 1.899807153673377e-05,
+      "loss": 0.6396,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8375400677513595,
+      "learning_rate": 1.8996922709216456e-05,
+      "loss": 0.5189,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0006365110478812,
+      "learning_rate": 1.8995773258220374e-05,
+      "loss": 0.6274,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9937520191210955,
+      "learning_rate": 1.8994623183825183e-05,
+      "loss": 0.6189,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1116281752553856,
+      "learning_rate": 1.8993472486110586e-05,
+      "loss": 0.694,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0213234557429571,
+      "learning_rate": 1.899232116515632e-05,
+      "loss": 0.6026,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0023366265849354,
+      "learning_rate": 1.8991169221042173e-05,
+      "loss": 0.5995,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9910945945231302,
+      "learning_rate": 1.8990016653847978e-05,
+      "loss": 0.6022,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1200184181031805,
+      "learning_rate": 1.8988863463653603e-05,
+      "loss": 0.6562,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0155184591151794,
+      "learning_rate": 1.8987709650538958e-05,
+      "loss": 0.6389,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9650011456012725,
+      "learning_rate": 1.898655521458401e-05,
+      "loss": 0.5576,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0579642872444046,
+      "learning_rate": 1.8985400155868756e-05,
+      "loss": 0.5898,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.968661935774855,
+      "learning_rate": 1.898424447447324e-05,
+      "loss": 0.5597,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0804058104288323,
+      "learning_rate": 1.8983088170477556e-05,
+      "loss": 0.6229,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1630361930612874,
+      "learning_rate": 1.8981931243961823e-05,
+      "loss": 0.6395,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9472132177706004,
+      "learning_rate": 1.8980773695006226e-05,
+      "loss": 0.5772,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9232106027800606,
+      "learning_rate": 1.897961552369098e-05,
+      "loss": 0.5748,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9375393267389712,
+      "learning_rate": 1.8978456730096336e-05,
+      "loss": 0.5782,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0526881739286673,
+      "learning_rate": 1.897729731430261e-05,
+      "loss": 0.5435,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1440929482692097,
+      "learning_rate": 1.8976137276390145e-05,
+      "loss": 0.6177,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9239252211596448,
+      "learning_rate": 1.897497661643932e-05,
+      "loss": 0.6119,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9664453721766127,
+      "learning_rate": 1.8973815334530583e-05,
+      "loss": 0.523,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9504949286899973,
+      "learning_rate": 1.8972653430744403e-05,
+      "loss": 0.6081,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0457199660560634,
+      "learning_rate": 1.8971490905161297e-05,
+      "loss": 0.6772,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0378069524788978,
+      "learning_rate": 1.897032775786183e-05,
+      "loss": 0.626,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9729500857706217,
+      "learning_rate": 1.8969163988926606e-05,
+      "loss": 0.5945,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.533917801839161,
+      "learning_rate": 1.896799959843627e-05,
+      "loss": 0.6287,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0979853305051799,
+      "learning_rate": 1.8966834586471517e-05,
+      "loss": 0.656,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9436270666351112,
+      "learning_rate": 1.8965668953113083e-05,
+      "loss": 0.5879,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9454702686057739,
+      "learning_rate": 1.8964502698441745e-05,
+      "loss": 0.5991,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8873844998178761,
+      "learning_rate": 1.8963335822538317e-05,
+      "loss": 0.5835,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.147471428553273,
+      "learning_rate": 1.896216832548367e-05,
+      "loss": 0.6603,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1057922519103787,
+      "learning_rate": 1.8961000207358707e-05,
+      "loss": 0.6675,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9412420554305279,
+      "learning_rate": 1.895983146824438e-05,
+      "loss": 0.6009,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9299496677178599,
+      "learning_rate": 1.8958662108221677e-05,
+      "loss": 0.5487,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0024697235170035,
+      "learning_rate": 1.8957492127371635e-05,
+      "loss": 0.644,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9525929588641228,
+      "learning_rate": 1.8956321525775337e-05,
+      "loss": 0.628,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0418909039513666,
+      "learning_rate": 1.8955150303513902e-05,
+      "loss": 0.5646,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9514053674277251,
+      "learning_rate": 1.895397846066849e-05,
+      "loss": 0.5671,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.040858447434405,
+      "learning_rate": 1.8952805997320315e-05,
+      "loss": 0.6042,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0380436753980238,
+      "learning_rate": 1.8951632913550625e-05,
+      "loss": 0.6072,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0234302731992189,
+      "learning_rate": 1.8950459209440716e-05,
+      "loss": 0.6109,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0945225898261712,
+      "learning_rate": 1.8949284885071917e-05,
+      "loss": 0.6402,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9642259329102109,
+      "learning_rate": 1.8948109940525622e-05,
+      "loss": 0.5854,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9262640462813323,
+      "learning_rate": 1.894693437588324e-05,
+      "loss": 0.5699,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0546036035380781,
+      "learning_rate": 1.8945758191226242e-05,
+      "loss": 0.6399,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.892279726784369,
+      "learning_rate": 1.8944581386636137e-05,
+      "loss": 0.5991,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9601886962094532,
+      "learning_rate": 1.8943403962194477e-05,
+      "loss": 0.5391,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9820769851632366,
+      "learning_rate": 1.8942225917982854e-05,
+      "loss": 0.6186,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0907957831446005,
+      "learning_rate": 1.8941047254082903e-05,
+      "loss": 0.5005,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8872934358107167,
+      "learning_rate": 1.8939867970576315e-05,
+      "loss": 0.5791,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.012616664948051,
+      "learning_rate": 1.8938688067544802e-05,
+      "loss": 0.5794,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.023609395455567,
+      "learning_rate": 1.893750754507014e-05,
+      "loss": 0.5418,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0245431610313795,
+      "learning_rate": 1.8936326403234125e-05,
+      "loss": 0.5804,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0381393955766731,
+      "learning_rate": 1.893514464211862e-05,
+      "loss": 0.617,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9191246443163674,
+      "learning_rate": 1.8933962261805515e-05,
+      "loss": 0.6222,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0053856875035811,
+      "learning_rate": 1.893277926237675e-05,
+      "loss": 0.5887,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9937571856876615,
+      "learning_rate": 1.8931595643914307e-05,
+      "loss": 0.6551,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0103025400896615,
+      "learning_rate": 1.893041140650021e-05,
+      "loss": 0.4889,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1326283733191278,
+      "learning_rate": 1.8929226550216522e-05,
+      "loss": 0.6579,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0345882359058862,
+      "learning_rate": 1.8928041075145352e-05,
+      "loss": 0.6248,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9638961572425462,
+      "learning_rate": 1.892685498136886e-05,
+      "loss": 0.5776,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9503035496039794,
+      "learning_rate": 1.892566826896923e-05,
+      "loss": 0.5758,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0265755834645953,
+      "learning_rate": 1.8924480938028708e-05,
+      "loss": 0.5906,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9772350507987435,
+      "learning_rate": 1.8923292988629575e-05,
+      "loss": 0.5762,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9678168956550066,
+      "learning_rate": 1.892210442085415e-05,
+      "loss": 0.5854,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9550725284032997,
+      "learning_rate": 1.8920915234784805e-05,
+      "loss": 0.6434,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0418780540372863,
+      "learning_rate": 1.8919725430503946e-05,
+      "loss": 0.6381,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.951117776669499,
+      "learning_rate": 1.8918535008094028e-05,
+      "loss": 0.5835,
+      "step": 2153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8789893915475889,
+      "learning_rate": 1.891734396763754e-05,
+      "loss": 0.5071,
+      "step": 2154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9637330021078095,
+      "learning_rate": 1.891615230921703e-05,
+      "loss": 0.602,
+      "step": 2155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9730380922707319,
+      "learning_rate": 1.8914960032915072e-05,
+      "loss": 0.5574,
+      "step": 2156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9745286624861224,
+      "learning_rate": 1.891376713881429e-05,
+      "loss": 0.6078,
+      "step": 2157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9153110919261715,
+      "learning_rate": 1.8912573626997354e-05,
+      "loss": 0.5904,
+      "step": 2158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.011667820668252,
+      "learning_rate": 1.891137949754697e-05,
+      "loss": 0.6148,
+      "step": 2159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0902155476345923,
+      "learning_rate": 1.891018475054589e-05,
+      "loss": 0.6056,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9472518475231407,
+      "learning_rate": 1.890898938607691e-05,
+      "loss": 0.544,
+      "step": 2161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0552234696767666,
+      "learning_rate": 1.890779340422287e-05,
+      "loss": 0.6252,
+      "step": 2162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0569263154227653,
+      "learning_rate": 1.8906596805066648e-05,
+      "loss": 0.5732,
+      "step": 2163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0492075165471888,
+      "learning_rate": 1.8905399588691165e-05,
+      "loss": 0.621,
+      "step": 2164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0089167595328195,
+      "learning_rate": 1.890420175517939e-05,
+      "loss": 0.6021,
+      "step": 2165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0353057215864783,
+      "learning_rate": 1.8903003304614332e-05,
+      "loss": 0.5981,
+      "step": 2166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9646737678180385,
+      "learning_rate": 1.8901804237079043e-05,
+      "loss": 0.6035,
+      "step": 2167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0961600407946666,
+      "learning_rate": 1.8900604552656615e-05,
+      "loss": 0.6611,
+      "step": 2168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9352215688772914,
+      "learning_rate": 1.889940425143019e-05,
+      "loss": 0.5734,
+      "step": 2169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9609782533708203,
+      "learning_rate": 1.889820333348294e-05,
+      "loss": 0.569,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8452540176519757,
+      "learning_rate": 1.8897001798898093e-05,
+      "loss": 0.5334,
+      "step": 2171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9851803544469124,
+      "learning_rate": 1.8895799647758912e-05,
+      "loss": 0.6187,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0293889942861165,
+      "learning_rate": 1.889459688014871e-05,
+      "loss": 0.6058,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0018336586296221,
+      "learning_rate": 1.8893393496150828e-05,
+      "loss": 0.6061,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.008102851968431,
+      "learning_rate": 1.889218949584867e-05,
+      "loss": 0.6237,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0072450005393212,
+      "learning_rate": 1.8890984879325664e-05,
+      "loss": 0.6138,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.187526031541943,
+      "learning_rate": 1.888977964666529e-05,
+      "loss": 0.6545,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9867953653591311,
+      "learning_rate": 1.8888573797951078e-05,
+      "loss": 0.6127,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9387087958810846,
+      "learning_rate": 1.888736733326658e-05,
+      "loss": 0.6117,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0722541801802747,
+      "learning_rate": 1.8886160252695413e-05,
+      "loss": 0.6302,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9779850206491939,
+      "learning_rate": 1.8884952556321223e-05,
+      "loss": 0.6469,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0439762284857736,
+      "learning_rate": 1.8883744244227697e-05,
+      "loss": 0.6531,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0267706789612425,
+      "learning_rate": 1.8882535316498577e-05,
+      "loss": 0.6373,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9898580400329446,
+      "learning_rate": 1.888132577321764e-05,
+      "loss": 0.6467,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8775000630268309,
+      "learning_rate": 1.8880115614468705e-05,
+      "loss": 0.5477,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9595056001380599,
+      "learning_rate": 1.8878904840335635e-05,
+      "loss": 0.669,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9219339134739535,
+      "learning_rate": 1.887769345090233e-05,
+      "loss": 0.5872,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.034499138231239,
+      "learning_rate": 1.887648144625275e-05,
+      "loss": 0.6174,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9426548133498441,
+      "learning_rate": 1.8875268826470875e-05,
+      "loss": 0.5727,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0290765921044192,
+      "learning_rate": 1.8874055591640746e-05,
+      "loss": 0.5592,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8568541726447252,
+      "learning_rate": 1.887284174184643e-05,
+      "loss": 0.499,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.009808118159287,
+      "learning_rate": 1.8871627277172058e-05,
+      "loss": 0.5931,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0151562984941591,
+      "learning_rate": 1.887041219770178e-05,
+      "loss": 0.6362,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9196707775669238,
+      "learning_rate": 1.8869196503519807e-05,
+      "loss": 0.5952,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9065759621006254,
+      "learning_rate": 1.8867980194710382e-05,
+      "loss": 0.6028,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.027461981521237,
+      "learning_rate": 1.88667632713578e-05,
+      "loss": 0.5995,
+      "step": 2196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9316606805491132,
+      "learning_rate": 1.886554573354638e-05,
+      "loss": 0.561,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0300501829703554,
+      "learning_rate": 1.886432758136051e-05,
+      "loss": 0.5255,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0239573905462445,
+      "learning_rate": 1.8863108814884602e-05,
+      "loss": 0.6205,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0534000129161385,
+      "learning_rate": 1.8861889434203112e-05,
+      "loss": 0.599,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0750868919378282,
+      "learning_rate": 1.8860669439400543e-05,
+      "loss": 0.5538,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9883194803662289,
+      "learning_rate": 1.8859448830561445e-05,
+      "loss": 0.6598,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9895234269568443,
+      "learning_rate": 1.8858227607770398e-05,
+      "loss": 0.668,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9954174791349467,
+      "learning_rate": 1.885700577111204e-05,
+      "loss": 0.632,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1474192952722595,
+      "learning_rate": 1.8855783320671034e-05,
+      "loss": 0.5958,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9192615979698529,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.5911,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.013275376886477,
+      "learning_rate": 1.8853336578779994e-05,
+      "loss": 0.5722,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9358020044247859,
+      "learning_rate": 1.8852112287499518e-05,
+      "loss": 0.5594,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.983445447661659,
+      "learning_rate": 1.8850887382775507e-05,
+      "loss": 0.5761,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.026955046322064,
+      "learning_rate": 1.884966186469286e-05,
+      "loss": 0.5661,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.159699459442207,
+      "learning_rate": 1.8848435733336487e-05,
+      "loss": 0.6511,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.001777768746846,
+      "learning_rate": 1.884720898879137e-05,
+      "loss": 0.5746,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8785765469744956,
+      "learning_rate": 1.8845981631142518e-05,
+      "loss": 0.5533,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0559320675134651,
+      "learning_rate": 1.8844753660474985e-05,
+      "loss": 0.6484,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.060208265156859,
+      "learning_rate": 1.8843525076873866e-05,
+      "loss": 0.6731,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9241440849844474,
+      "learning_rate": 1.8842295880424305e-05,
+      "loss": 0.5179,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.960927482449649,
+      "learning_rate": 1.8841066071211485e-05,
+      "loss": 0.5848,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0117277009292722,
+      "learning_rate": 1.8839835649320622e-05,
+      "loss": 0.5871,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1051299188661905,
+      "learning_rate": 1.8838604614837e-05,
+      "loss": 0.6599,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1126022281978067,
+      "learning_rate": 1.8837372967845907e-05,
+      "loss": 0.6209,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0437512390828085,
+      "learning_rate": 1.883614070843271e-05,
+      "loss": 0.6129,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9347343174247541,
+      "learning_rate": 1.88349078366828e-05,
+      "loss": 0.5732,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.978072948910801,
+      "learning_rate": 1.8833674352681613e-05,
+      "loss": 0.5862,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9035635392386544,
+      "learning_rate": 1.8832440256514633e-05,
+      "loss": 0.5817,
+      "step": 2224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.946716318845138,
+      "learning_rate": 1.8831205548267375e-05,
+      "loss": 0.6018,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9114345496424868,
+      "learning_rate": 1.8829970228025405e-05,
+      "loss": 0.5422,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.947695890004665,
+      "learning_rate": 1.882873429587433e-05,
+      "loss": 0.6219,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9607336606853346,
+      "learning_rate": 1.8827497751899798e-05,
+      "loss": 0.5899,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.056301862698836,
+      "learning_rate": 1.8826260596187505e-05,
+      "loss": 0.6012,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9839236383385103,
+      "learning_rate": 1.882502282882318e-05,
+      "loss": 0.5476,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9453761983415658,
+      "learning_rate": 1.88237844498926e-05,
+      "loss": 0.5613,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0769594253515593,
+      "learning_rate": 1.8822545459481585e-05,
+      "loss": 0.6289,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9504981321207882,
+      "learning_rate": 1.8821305857675997e-05,
+      "loss": 0.5588,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9703487793124386,
+      "learning_rate": 1.8820065644561736e-05,
+      "loss": 0.5812,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9863394180870134,
+      "learning_rate": 1.8818824820224747e-05,
+      "loss": 0.6304,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9529494000210472,
+      "learning_rate": 1.8817583384751023e-05,
+      "loss": 0.5819,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9677249493395993,
+      "learning_rate": 1.881634133822659e-05,
+      "loss": 0.6066,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9420220803818873,
+      "learning_rate": 1.8815098680737523e-05,
+      "loss": 0.5745,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.016139132803753,
+      "learning_rate": 1.881385541236994e-05,
+      "loss": 0.6346,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1089714899910492,
+      "learning_rate": 1.881261153320999e-05,
+      "loss": 0.6067,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9853924873312063,
+      "learning_rate": 1.881136704334388e-05,
+      "loss": 0.6338,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8155046887597646,
+      "learning_rate": 1.8810121942857848e-05,
+      "loss": 0.6043,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9104861985868933,
+      "learning_rate": 1.880887623183818e-05,
+      "loss": 0.5181,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9228901883726971,
+      "learning_rate": 1.8807629910371203e-05,
+      "loss": 0.5156,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9801612127255177,
+      "learning_rate": 1.8806382978543283e-05,
+      "loss": 0.6033,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.056989468750248,
+      "learning_rate": 1.8805135436440837e-05,
+      "loss": 0.6078,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1457163395568721,
+      "learning_rate": 1.8803887284150317e-05,
+      "loss": 0.6381,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0805771612792152,
+      "learning_rate": 1.8802638521758214e-05,
+      "loss": 0.6548,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3202820723326176,
+      "learning_rate": 1.880138914935107e-05,
+      "loss": 0.636,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9148686090902376,
+      "learning_rate": 1.8800139167015466e-05,
+      "loss": 0.5745,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0389049536285395,
+      "learning_rate": 1.8798888574838023e-05,
+      "loss": 0.6133,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.030661980018216,
+      "learning_rate": 1.8797637372905407e-05,
+      "loss": 0.548,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0276577285974946,
+      "learning_rate": 1.8796385561304323e-05,
+      "loss": 0.6184,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0196070836335733,
+      "learning_rate": 1.8795133140121522e-05,
+      "loss": 0.5818,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0889379693225758,
+      "learning_rate": 1.8793880109443797e-05,
+      "loss": 0.6733,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9353604600412142,
+      "learning_rate": 1.8792626469357983e-05,
+      "loss": 0.5649,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9000891999282137,
+      "learning_rate": 1.879137221995095e-05,
+      "loss": 0.558,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9242703330932535,
+      "learning_rate": 1.879011736130962e-05,
+      "loss": 0.6322,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1666243279578408,
+      "learning_rate": 1.8788861893520954e-05,
+      "loss": 0.6027,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9173273086450375,
+      "learning_rate": 1.8787605816671956e-05,
+      "loss": 0.582,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9437307573005449,
+      "learning_rate": 1.8786349130849667e-05,
+      "loss": 0.5623,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8570905229082192,
+      "learning_rate": 1.8785091836141177e-05,
+      "loss": 0.522,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5564688988373234,
+      "learning_rate": 1.8783833932633617e-05,
+      "loss": 0.6621,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0150904173397395,
+      "learning_rate": 1.8782575420414155e-05,
+      "loss": 0.652,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9376520417247656,
+      "learning_rate": 1.8781316299570007e-05,
+      "loss": 0.5854,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9330936778529615,
+      "learning_rate": 1.878005657018843e-05,
+      "loss": 0.5916,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0036469337468876,
+      "learning_rate": 1.877879623235672e-05,
+      "loss": 0.6498,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0443264323120567,
+      "learning_rate": 1.8777535286162217e-05,
+      "loss": 0.678,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9581630196481067,
+      "learning_rate": 1.8776273731692306e-05,
+      "loss": 0.6016,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.083876675418555,
+      "learning_rate": 1.8775011569034405e-05,
+      "loss": 0.6118,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9873027524445195,
+      "learning_rate": 1.877374879827599e-05,
+      "loss": 0.629,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9486312587682523,
+      "learning_rate": 1.8772485419504566e-05,
+      "loss": 0.5635,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.011383681981482,
+      "learning_rate": 1.877122143280768e-05,
+      "loss": 0.642,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.054871890426257,
+      "learning_rate": 1.8769956838272937e-05,
+      "loss": 0.624,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9243443195958089,
+      "learning_rate": 1.8768691635987957e-05,
+      "loss": 0.6012,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9137583522941478,
+      "learning_rate": 1.8767425826040426e-05,
+      "loss": 0.6101,
+      "step": 2276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8732029787589334,
+      "learning_rate": 1.8766159408518062e-05,
+      "loss": 0.5436,
+      "step": 2277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0490382062184178,
+      "learning_rate": 1.8764892383508626e-05,
+      "loss": 0.5305,
+      "step": 2278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0350121153065304,
+      "learning_rate": 1.8763624751099924e-05,
+      "loss": 0.6107,
+      "step": 2279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8036004247191376,
+      "learning_rate": 1.8762356511379796e-05,
+      "loss": 0.5169,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9817024833816537,
+      "learning_rate": 1.8761087664436137e-05,
+      "loss": 0.5483,
+      "step": 2281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0869713946968007,
+      "learning_rate": 1.8759818210356874e-05,
+      "loss": 0.6059,
+      "step": 2282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9333008852297026,
+      "learning_rate": 1.8758548149229978e-05,
+      "loss": 0.5559,
+      "step": 2283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9797631089608451,
+      "learning_rate": 1.8757277481143467e-05,
+      "loss": 0.5846,
+      "step": 2284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8580503736403456,
+      "learning_rate": 1.8756006206185388e-05,
+      "loss": 0.5484,
+      "step": 2285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9459834566720147,
+      "learning_rate": 1.8754734324443853e-05,
+      "loss": 0.5814,
+      "step": 2286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.129380323888552,
+      "learning_rate": 1.875346183600699e-05,
+      "loss": 0.6183,
+      "step": 2287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9513144228125862,
+      "learning_rate": 1.8752188740962986e-05,
+      "loss": 0.607,
+      "step": 2288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0304282864628564,
+      "learning_rate": 1.8750915039400068e-05,
+      "loss": 0.6071,
+      "step": 2289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9340505697364657,
+      "learning_rate": 1.87496407314065e-05,
+      "loss": 0.5504,
+      "step": 2290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.023533370754278,
+      "learning_rate": 1.8748365817070586e-05,
+      "loss": 0.6182,
+      "step": 2291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9882283676462045,
+      "learning_rate": 1.8747090296480683e-05,
+      "loss": 0.5844,
+      "step": 2292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9320546167398589,
+      "learning_rate": 1.8745814169725183e-05,
+      "loss": 0.5435,
+      "step": 2293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9867593700841457,
+      "learning_rate": 1.8744537436892517e-05,
+      "loss": 0.5975,
+      "step": 2294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0258212231425314,
+      "learning_rate": 1.8743260098071163e-05,
+      "loss": 0.5049,
+      "step": 2295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0907723321403282,
+      "learning_rate": 1.8741982153349642e-05,
+      "loss": 0.6586,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.950702288691101,
+      "learning_rate": 1.8740703602816506e-05,
+      "loss": 0.6177,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0254255423493217,
+      "learning_rate": 1.8739424446560365e-05,
+      "loss": 0.6089,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0918084848642433,
+      "learning_rate": 1.8738144684669867e-05,
+      "loss": 0.6807,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0255381404566133,
+      "learning_rate": 1.8736864317233688e-05,
+      "loss": 0.6707,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9287783578796915,
+      "learning_rate": 1.873558334434056e-05,
+      "loss": 0.6023,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9522040066938621,
+      "learning_rate": 1.873430176607926e-05,
+      "loss": 0.5756,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9664634365215445,
+      "learning_rate": 1.8733019582538595e-05,
+      "loss": 0.5544,
+      "step": 2303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9438264204012479,
+      "learning_rate": 1.8731736793807417e-05,
+      "loss": 0.5802,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9450277499788748,
+      "learning_rate": 1.873045339997462e-05,
+      "loss": 0.5775,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9226491997621605,
+      "learning_rate": 1.872916940112915e-05,
+      "loss": 0.5655,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9316094898328231,
+      "learning_rate": 1.8727884797359984e-05,
+      "loss": 0.571,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9161774303898466,
+      "learning_rate": 1.8726599588756144e-05,
+      "loss": 0.5742,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0212729130725422,
+      "learning_rate": 1.8725313775406693e-05,
+      "loss": 0.6725,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0371994533029285,
+      "learning_rate": 1.8724027357400737e-05,
+      "loss": 0.6275,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9387275440330368,
+      "learning_rate": 1.872274033482742e-05,
+      "loss": 0.609,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.018429708491077,
+      "learning_rate": 1.8721452707775935e-05,
+      "loss": 0.5579,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0025377413107226,
+      "learning_rate": 1.8720164476335516e-05,
+      "loss": 0.6228,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1300305737441967,
+      "learning_rate": 1.8718875640595432e-05,
+      "loss": 0.661,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0204356367555911,
+      "learning_rate": 1.8717586200645002e-05,
+      "loss": 0.5863,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9527375433742457,
+      "learning_rate": 1.8716296156573578e-05,
+      "loss": 0.6602,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9901798497245836,
+      "learning_rate": 1.8715005508470565e-05,
+      "loss": 0.5693,
+      "step": 2317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.053822223301016,
+      "learning_rate": 1.8713714256425396e-05,
+      "loss": 0.5913,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8833294760007623,
+      "learning_rate": 1.8712422400527556e-05,
+      "loss": 0.5931,
+      "step": 2319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9639514016390054,
+      "learning_rate": 1.8711129940866577e-05,
+      "loss": 0.5981,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0823666757152655,
+      "learning_rate": 1.870983687753202e-05,
+      "loss": 0.6608,
+      "step": 2321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8931959289690588,
+      "learning_rate": 1.8708543210613492e-05,
+      "loss": 0.5971,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9858619839716677,
+      "learning_rate": 1.8707248940200643e-05,
+      "loss": 0.5847,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.931992339599603,
+      "learning_rate": 1.8705954066383166e-05,
+      "loss": 0.5913,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9600650161274669,
+      "learning_rate": 1.8704658589250795e-05,
+      "loss": 0.5694,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9139115608171103,
+      "learning_rate": 1.8703362508893302e-05,
+      "loss": 0.599,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9079993112354389,
+      "learning_rate": 1.870206582540051e-05,
+      "loss": 0.6328,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.930373994652327,
+      "learning_rate": 1.8700768538862274e-05,
+      "loss": 0.6067,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.935081675704654,
+      "learning_rate": 1.8699470649368496e-05,
+      "loss": 0.5887,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.974106870385751,
+      "learning_rate": 1.8698172157009124e-05,
+      "loss": 0.5886,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9335074917711323,
+      "learning_rate": 1.8696873061874127e-05,
+      "loss": 0.556,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9289390886527041,
+      "learning_rate": 1.8695573364053548e-05,
+      "loss": 0.6101,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9494119915399611,
+      "learning_rate": 1.8694273063637444e-05,
+      "loss": 0.6225,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.868203726273747,
+      "learning_rate": 1.869297216071593e-05,
+      "loss": 0.5634,
+      "step": 2334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9645235066076571,
+      "learning_rate": 1.8691670655379157e-05,
+      "loss": 0.5783,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9291663000652597,
+      "learning_rate": 1.8690368547717313e-05,
+      "loss": 0.5383,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.030978162010167,
+      "learning_rate": 1.8689065837820642e-05,
+      "loss": 0.6168,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.01006563723956,
+      "learning_rate": 1.8687762525779412e-05,
+      "loss": 0.5955,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9572350447649162,
+      "learning_rate": 1.8686458611683948e-05,
+      "loss": 0.617,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0413589226973223,
+      "learning_rate": 1.8685154095624605e-05,
+      "loss": 0.7026,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.006611036261011,
+      "learning_rate": 1.8683848977691784e-05,
+      "loss": 0.6215,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.902856036151998,
+      "learning_rate": 1.868254325797594e-05,
+      "loss": 0.5864,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0432428298350762,
+      "learning_rate": 1.868123693656754e-05,
+      "loss": 0.5814,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0310284547691086,
+      "learning_rate": 1.8679930013557127e-05,
+      "loss": 0.5941,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9342025335462472,
+      "learning_rate": 1.867862248903526e-05,
+      "loss": 0.6058,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0046093700308836,
+      "learning_rate": 1.8677314363092555e-05,
+      "loss": 0.5298,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9711270887222404,
+      "learning_rate": 1.867600563581966e-05,
+      "loss": 0.5702,
+      "step": 2347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.077095212889763,
+      "learning_rate": 1.867469630730727e-05,
+      "loss": 0.5995,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8931131934883211,
+      "learning_rate": 1.867338637764612e-05,
+      "loss": 0.4927,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0369626372647212,
+      "learning_rate": 1.867207584692699e-05,
+      "loss": 0.577,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9567874811250195,
+      "learning_rate": 1.867076471524069e-05,
+      "loss": 0.4779,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9766997211054103,
+      "learning_rate": 1.866945298267809e-05,
+      "loss": 0.6276,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0546535855796144,
+      "learning_rate": 1.866814064933009e-05,
+      "loss": 0.6283,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0842469473548701,
+      "learning_rate": 1.8666827715287627e-05,
+      "loss": 0.66,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9133052767488332,
+      "learning_rate": 1.8665514180641697e-05,
+      "loss": 0.5781,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9991322639797208,
+      "learning_rate": 1.8664200045483314e-05,
+      "loss": 0.5336,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9699785492721534,
+      "learning_rate": 1.8662885309903558e-05,
+      "loss": 0.5715,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0024836346982164,
+      "learning_rate": 1.8661569973993533e-05,
+      "loss": 0.6319,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0298945721796273,
+      "learning_rate": 1.866025403784439e-05,
+      "loss": 0.6255,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8974998871243925,
+      "learning_rate": 1.865893750154732e-05,
+      "loss": 0.5178,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.057894104135965,
+      "learning_rate": 1.8657620365193566e-05,
+      "loss": 0.6258,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9877141542623917,
+      "learning_rate": 1.8656302628874402e-05,
+      "loss": 0.6648,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9542372663196041,
+      "learning_rate": 1.8654984292681142e-05,
+      "loss": 0.5872,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8546229947901854,
+      "learning_rate": 1.8653665356705146e-05,
+      "loss": 0.5864,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9260957967141475,
+      "learning_rate": 1.865234582103782e-05,
+      "loss": 0.5882,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.038963914135987,
+      "learning_rate": 1.86510256857706e-05,
+      "loss": 0.6273,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9630029696755222,
+      "learning_rate": 1.8649704950994976e-05,
+      "loss": 0.6044,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0408025228163054,
+      "learning_rate": 1.864838361680247e-05,
+      "loss": 0.5721,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9901450030781432,
+      "learning_rate": 1.864706168328465e-05,
+      "loss": 0.6231,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0128039127196022,
+      "learning_rate": 1.8645739150533123e-05,
+      "loss": 0.6232,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.035782290703731,
+      "learning_rate": 1.8644416018639547e-05,
+      "loss": 0.6019,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0569343629525514,
+      "learning_rate": 1.8643092287695604e-05,
+      "loss": 0.6118,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9739875131987521,
+      "learning_rate": 1.8641767957793037e-05,
+      "loss": 0.6424,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9480212105192773,
+      "learning_rate": 1.864044302902361e-05,
+      "loss": 0.5794,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9633841127891988,
+      "learning_rate": 1.8639117501479143e-05,
+      "loss": 0.5918,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8875048063352914,
+      "learning_rate": 1.8637791375251505e-05,
+      "loss": 0.5661,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9553926429206897,
+      "learning_rate": 1.863646465043258e-05,
+      "loss": 0.6066,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9406156212474246,
+      "learning_rate": 1.8635137327114317e-05,
+      "loss": 0.5694,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0082468868263996,
+      "learning_rate": 1.8633809405388697e-05,
+      "loss": 0.584,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9697465631492217,
+      "learning_rate": 1.8632480885347744e-05,
+      "loss": 0.636,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9103226982286848,
+      "learning_rate": 1.863115176708352e-05,
+      "loss": 0.5912,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.903505883017083,
+      "learning_rate": 1.8629822050688138e-05,
+      "loss": 0.5725,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9399914412587055,
+      "learning_rate": 1.862849173625374e-05,
+      "loss": 0.5703,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.865161610202597,
+      "learning_rate": 1.862716082387252e-05,
+      "loss": 0.5792,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9859677288051738,
+      "learning_rate": 1.8625829313636707e-05,
+      "loss": 0.6196,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.994656223718024,
+      "learning_rate": 1.862449720563857e-05,
+      "loss": 0.6377,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0886207567360782,
+      "learning_rate": 1.862316449997043e-05,
+      "loss": 0.5995,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9986334346192072,
+      "learning_rate": 1.862183119672464e-05,
+      "loss": 0.6196,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0365462981424516,
+      "learning_rate": 1.862049729599359e-05,
+      "loss": 0.5983,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0433892804940201,
+      "learning_rate": 1.8619162797869728e-05,
+      "loss": 0.5875,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.956003796235037,
+      "learning_rate": 1.861782770244553e-05,
+      "loss": 0.5334,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8948968349280684,
+      "learning_rate": 1.8616492009813516e-05,
+      "loss": 0.5317,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.967022027051535,
+      "learning_rate": 1.8615155720066247e-05,
+      "loss": 0.6288,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9790421801306229,
+      "learning_rate": 1.861381883329633e-05,
+      "loss": 0.6143,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1236596585523422,
+      "learning_rate": 1.8612481349596406e-05,
+      "loss": 0.6214,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.946020311289503,
+      "learning_rate": 1.8611143269059165e-05,
+      "loss": 0.5774,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0809689030226681,
+      "learning_rate": 1.8609804591777333e-05,
+      "loss": 0.5975,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0135643705288075,
+      "learning_rate": 1.860846531784368e-05,
+      "loss": 0.6491,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9544748502585023,
+      "learning_rate": 1.8607125447351017e-05,
+      "loss": 0.6125,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9163164402222126,
+      "learning_rate": 1.8605784980392193e-05,
+      "loss": 0.5258,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.14714289345333,
+      "learning_rate": 1.86044439170601e-05,
+      "loss": 0.5343,
+      "step": 2401
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9996593655533791,
+      "learning_rate": 1.8603102257447686e-05,
+      "loss": 0.6477,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0396410331618158,
+      "learning_rate": 1.860176000164791e-05,
+      "loss": 0.5779,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9694426587894854,
+      "learning_rate": 1.8600417149753794e-05,
+      "loss": 0.5503,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9399594621363393,
+      "learning_rate": 1.85990737018584e-05,
+      "loss": 0.5529,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9611673318625494,
+      "learning_rate": 1.8597729658054827e-05,
+      "loss": 0.577,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.063842825267057,
+      "learning_rate": 1.8596385018436214e-05,
+      "loss": 0.6513,
+      "step": 2407
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9440600590318581,
+      "learning_rate": 1.8595039783095747e-05,
+      "loss": 0.522,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9333040800805044,
+      "learning_rate": 1.859369395212664e-05,
+      "loss": 0.6214,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0208532063205993,
+      "learning_rate": 1.859234752562217e-05,
+      "loss": 0.6271,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.081775506940939,
+      "learning_rate": 1.8591000503675635e-05,
+      "loss": 0.5961,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9756559941848068,
+      "learning_rate": 1.8589652886380387e-05,
+      "loss": 0.6023,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.061469168824822,
+      "learning_rate": 1.8588304673829814e-05,
+      "loss": 0.64,
+      "step": 2413
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9643883392364556,
+      "learning_rate": 1.8586955866117345e-05,
+      "loss": 0.6009,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9622948386380665,
+      "learning_rate": 1.8585606463336448e-05,
+      "loss": 0.5491,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9804391313646905,
+      "learning_rate": 1.8584256465580642e-05,
+      "loss": 0.6679,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1901230134706908,
+      "learning_rate": 1.8582905872943477e-05,
+      "loss": 0.5767,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0298855448133197,
+      "learning_rate": 1.8581554685518543e-05,
+      "loss": 0.5731,
+      "step": 2418
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9403946452528907,
+      "learning_rate": 1.8580202903399484e-05,
+      "loss": 0.5958,
+      "step": 2419
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0789858376109014,
+      "learning_rate": 1.8578850526679976e-05,
+      "loss": 0.6047,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9846049002431896,
+      "learning_rate": 1.8577497555453735e-05,
+      "loss": 0.5542,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9920273561796581,
+      "learning_rate": 1.8576143989814524e-05,
+      "loss": 0.5957,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.012926254158362,
+      "learning_rate": 1.857478982985614e-05,
+      "loss": 0.5816,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1060199873713032,
+      "learning_rate": 1.8573435075672422e-05,
+      "loss": 0.6219,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9665465330730769,
+      "learning_rate": 1.8572079727357265e-05,
+      "loss": 0.5141,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.010619313409472,
+      "learning_rate": 1.8570723785004583e-05,
+      "loss": 0.6223,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0701981612275158,
+      "learning_rate": 1.8569367248708343e-05,
+      "loss": 0.6207,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.968653341581941,
+      "learning_rate": 1.8568010118562556e-05,
+      "loss": 0.6153,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0285147099330336,
+      "learning_rate": 1.8566652394661268e-05,
+      "loss": 0.5845,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0401009522193112,
+      "learning_rate": 1.856529407709857e-05,
+      "loss": 0.5614,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0312486719248375,
+      "learning_rate": 1.8563935165968584e-05,
+      "loss": 0.5535,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8939351482825008,
+      "learning_rate": 1.8562575661365493e-05,
+      "loss": 0.5404,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0252632002781987,
+      "learning_rate": 1.8561215563383496e-05,
+      "loss": 0.5902,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9425198617957646,
+      "learning_rate": 1.855985487211686e-05,
+      "loss": 0.579,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9890648746831118,
+      "learning_rate": 1.8558493587659874e-05,
+      "loss": 0.621,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9366563149047062,
+      "learning_rate": 1.8557131710106873e-05,
+      "loss": 0.5563,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9984953205723499,
+      "learning_rate": 1.8555769239552232e-05,
+      "loss": 0.6493,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.980140534084269,
+      "learning_rate": 1.8554406176090377e-05,
+      "loss": 0.6212,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.00668465629207,
+      "learning_rate": 1.8553042519815756e-05,
+      "loss": 0.6299,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9874709621095749,
+      "learning_rate": 1.8551678270822878e-05,
+      "loss": 0.6188,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9247042108253141,
+      "learning_rate": 1.8550313429206282e-05,
+      "loss": 0.557,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9952811181588312,
+      "learning_rate": 1.8548947995060547e-05,
+      "loss": 0.6042,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0020526733733572,
+      "learning_rate": 1.85475819684803e-05,
+      "loss": 0.5833,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.955455293951105,
+      "learning_rate": 1.8546215349560204e-05,
+      "loss": 0.6151,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0341207506844388,
+      "learning_rate": 1.8544848138394965e-05,
+      "loss": 0.5803,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0264964629842621,
+      "learning_rate": 1.854348033507933e-05,
+      "loss": 0.6286,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9302590580265883,
+      "learning_rate": 1.8542111939708086e-05,
+      "loss": 0.5781,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8796676362843857,
+      "learning_rate": 1.854074295237606e-05,
+      "loss": 0.5421,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.004870683626179,
+      "learning_rate": 1.8539373373178126e-05,
+      "loss": 0.5445,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0665094713061753,
+      "learning_rate": 1.8538003202209186e-05,
+      "loss": 0.6193,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0457962229818203,
+      "learning_rate": 1.8536632439564203e-05,
+      "loss": 0.5028,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9640419849222112,
+      "learning_rate": 1.853526108533816e-05,
+      "loss": 0.6291,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0268636636017952,
+      "learning_rate": 1.8533889139626096e-05,
+      "loss": 0.5408,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0358135887820035,
+      "learning_rate": 1.8532516602523087e-05,
+      "loss": 0.5789,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.046182089719115,
+      "learning_rate": 1.853114347412424e-05,
+      "loss": 0.5674,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.126374433619317,
+      "learning_rate": 1.8529769754524724e-05,
+      "loss": 0.6586,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9045955142000208,
+      "learning_rate": 1.8528395443819725e-05,
+      "loss": 0.5948,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8992249493330345,
+      "learning_rate": 1.8527020542104487e-05,
+      "loss": 0.5861,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0028727195091176,
+      "learning_rate": 1.852564504947429e-05,
+      "loss": 0.6158,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1317843702334225,
+      "learning_rate": 1.852426896602445e-05,
+      "loss": 0.632,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0718645139741845,
+      "learning_rate": 1.8522892291850335e-05,
+      "loss": 0.6324,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1507465013420966,
+      "learning_rate": 1.8521515027047344e-05,
+      "loss": 0.5917,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0390501368556815,
+      "learning_rate": 1.8520137171710923e-05,
+      "loss": 0.623,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0199145522160606,
+      "learning_rate": 1.851875872593655e-05,
+      "loss": 0.6655,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9765738987075874,
+      "learning_rate": 1.8517379689819752e-05,
+      "loss": 0.5973,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0897129702395667,
+      "learning_rate": 1.85160000634561e-05,
+      "loss": 0.5654,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9737301064012777,
+      "learning_rate": 1.8514619846941192e-05,
+      "loss": 0.589,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9419077307422697,
+      "learning_rate": 1.851323904037069e-05,
+      "loss": 0.5888,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0853730530182204,
+      "learning_rate": 1.8511857643840264e-05,
+      "loss": 0.6198,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.992302096237278,
+      "learning_rate": 1.8510475657445656e-05,
+      "loss": 0.5962,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0140166287761607,
+      "learning_rate": 1.8509093081282636e-05,
+      "loss": 0.6112,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8910914912445789,
+      "learning_rate": 1.8507709915447013e-05,
+      "loss": 0.5705,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9660251973192866,
+      "learning_rate": 1.8506326160034638e-05,
+      "loss": 0.556,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9516083077316758,
+      "learning_rate": 1.8504941815141406e-05,
+      "loss": 0.622,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0276900810420124,
+      "learning_rate": 1.850355688086325e-05,
+      "loss": 0.6428,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0234009192262177,
+      "learning_rate": 1.8502171357296144e-05,
+      "loss": 0.521,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9300947641179861,
+      "learning_rate": 1.8500785244536104e-05,
+      "loss": 0.5487,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9803225475297762,
+      "learning_rate": 1.849939854267919e-05,
+      "loss": 0.5837,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9718620665209154,
+      "learning_rate": 1.849801125182149e-05,
+      "loss": 0.5954,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8568227724968096,
+      "learning_rate": 1.8496623372059152e-05,
+      "loss": 0.5528,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9975366955008645,
+      "learning_rate": 1.849523490348835e-05,
+      "loss": 0.6434,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9704820831780121,
+      "learning_rate": 1.8493845846205303e-05,
+      "loss": 0.6076,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0470017044262758,
+      "learning_rate": 1.8492456200306276e-05,
+      "loss": 0.6127,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0002178090901197,
+      "learning_rate": 1.8491065965887568e-05,
+      "loss": 0.5508,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9119469336380572,
+      "learning_rate": 1.8489675143045516e-05,
+      "loss": 0.5624,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0505339179736148,
+      "learning_rate": 1.8488283731876508e-05,
+      "loss": 0.6058,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8875674907841911,
+      "learning_rate": 1.848689173247697e-05,
+      "loss": 0.5268,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9865949243859709,
+      "learning_rate": 1.8485499144943358e-05,
+      "loss": 0.5906,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9625178791840348,
+      "learning_rate": 1.8484105969372184e-05,
+      "loss": 0.5549,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8948390139961174,
+      "learning_rate": 1.8482712205859992e-05,
+      "loss": 0.6188,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.047076142564619,
+      "learning_rate": 1.848131785450337e-05,
+      "loss": 0.6543,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9444702882438062,
+      "learning_rate": 1.8479922915398937e-05,
+      "loss": 0.579,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0852998525665014,
+      "learning_rate": 1.8478527388643375e-05,
+      "loss": 0.6392,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.951577608183831,
+      "learning_rate": 1.8477131274333383e-05,
+      "loss": 0.5678,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9536234809843884,
+      "learning_rate": 1.847573457256571e-05,
+      "loss": 0.5663,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0085519410123744,
+      "learning_rate": 1.8474337283437155e-05,
+      "loss": 0.6148,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9733488355019365,
+      "learning_rate": 1.8472939407044536e-05,
+      "loss": 0.6036,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9837751916587119,
+      "learning_rate": 1.847154094348474e-05,
+      "loss": 0.5707,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0040076270633222,
+      "learning_rate": 1.847014189285466e-05,
+      "loss": 0.5996,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8408621469655635,
+      "learning_rate": 1.8468742255251268e-05,
+      "loss": 0.5432,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0111470886873506,
+      "learning_rate": 1.846734203077155e-05,
+      "loss": 0.6538,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.965137606513294,
+      "learning_rate": 1.8465941219512533e-05,
+      "loss": 0.581,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8809664551569435,
+      "learning_rate": 1.8464539821571302e-05,
+      "loss": 0.5743,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0831610501744091,
+      "learning_rate": 1.8463137837044973e-05,
+      "loss": 0.5693,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9377551417262783,
+      "learning_rate": 1.8461735266030696e-05,
+      "loss": 0.5784,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0527879433579985,
+      "learning_rate": 1.8460332108625668e-05,
+      "loss": 0.629,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.992612826274164,
+      "learning_rate": 1.8458928364927137e-05,
+      "loss": 0.5982,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9554499534353595,
+      "learning_rate": 1.8457524035032364e-05,
+      "loss": 0.5667,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0322061914645393,
+      "learning_rate": 1.8456119119038683e-05,
+      "loss": 0.6673,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.012298631254966,
+      "learning_rate": 1.8454713617043448e-05,
+      "loss": 0.6267,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9751534686903088,
+      "learning_rate": 1.8453307529144055e-05,
+      "loss": 0.641,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0246962915110827,
+      "learning_rate": 1.845190085543795e-05,
+      "loss": 0.6305,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9089309094696517,
+      "learning_rate": 1.845049359602261e-05,
+      "loss": 0.5929,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8998356265099821,
+      "learning_rate": 1.8449085750995564e-05,
+      "loss": 0.5446,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.008345723946353,
+      "learning_rate": 1.8447677320454367e-05,
+      "loss": 0.596,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9908774036336532,
+      "learning_rate": 1.8446268304496624e-05,
+      "loss": 0.6383,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0072737095100066,
+      "learning_rate": 1.8444858703219982e-05,
+      "loss": 0.6442,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8652406980423714,
+      "learning_rate": 1.844344851672212e-05,
+      "loss": 0.5321,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9981622954590514,
+      "learning_rate": 1.844203774510077e-05,
+      "loss": 0.6602,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8713331836320304,
+      "learning_rate": 1.8440626388453686e-05,
+      "loss": 0.5823,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9236692999869184,
+      "learning_rate": 1.8439214446878685e-05,
+      "loss": 0.7023,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9117951641284442,
+      "learning_rate": 1.8437801920473605e-05,
+      "loss": 0.5547,
+      "step": 2522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9483077352466959,
+      "learning_rate": 1.8436388809336338e-05,
+      "loss": 0.561,
+      "step": 2523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9665947626085446,
+      "learning_rate": 1.8434975113564804e-05,
+      "loss": 0.6055,
+      "step": 2524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9385491506378937,
+      "learning_rate": 1.8433560833256986e-05,
+      "loss": 0.6063,
+      "step": 2525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9527023510611712,
+      "learning_rate": 1.8432145968510878e-05,
+      "loss": 0.5277,
+      "step": 2526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9662826327001135,
+      "learning_rate": 1.8430730519424532e-05,
+      "loss": 0.5898,
+      "step": 2527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.034736667191442,
+      "learning_rate": 1.8429314486096042e-05,
+      "loss": 0.5866,
+      "step": 2528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0689261718103882,
+      "learning_rate": 1.8427897868623535e-05,
+      "loss": 0.6371,
+      "step": 2529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8653291692283432,
+      "learning_rate": 1.8426480667105178e-05,
+      "loss": 0.568,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9098599831792978,
+      "learning_rate": 1.842506288163919e-05,
+      "loss": 0.5838,
+      "step": 2531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9191263895725162,
+      "learning_rate": 1.8423644512323814e-05,
+      "loss": 0.6342,
+      "step": 2532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9569768590266543,
+      "learning_rate": 1.8422225559257345e-05,
+      "loss": 0.6281,
+      "step": 2533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8285310527806168,
+      "learning_rate": 1.8420806022538115e-05,
+      "loss": 0.5172,
+      "step": 2534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8930960260286901,
+      "learning_rate": 1.8419385902264497e-05,
+      "loss": 0.6172,
+      "step": 2535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9269553677227064,
+      "learning_rate": 1.8417965198534907e-05,
+      "loss": 0.6097,
+      "step": 2536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0740954878523123,
+      "learning_rate": 1.841654391144779e-05,
+      "loss": 0.6397,
+      "step": 2537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9446598393823056,
+      "learning_rate": 1.841512204110165e-05,
+      "loss": 0.5599,
+      "step": 2538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9103665763797646,
+      "learning_rate": 1.8413699587595016e-05,
+      "loss": 0.5865,
+      "step": 2539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9054825790388551,
+      "learning_rate": 1.841227655102646e-05,
+      "loss": 0.5674,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8263852032716875,
+      "learning_rate": 1.8410852931494606e-05,
+      "loss": 0.6027,
+      "step": 2541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0279982369799479,
+      "learning_rate": 1.84094287290981e-05,
+      "loss": 0.6128,
+      "step": 2542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8627449323576147,
+      "learning_rate": 1.8408003943935643e-05,
+      "loss": 0.5885,
+      "step": 2543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9312463352872257,
+      "learning_rate": 1.8406578576105973e-05,
+      "loss": 0.5843,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8876688206289071,
+      "learning_rate": 1.8405152625707863e-05,
+      "loss": 0.5971,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8334420024858418,
+      "learning_rate": 1.840372609284013e-05,
+      "loss": 0.5071,
+      "step": 2546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0547438702977643,
+      "learning_rate": 1.8402298977601636e-05,
+      "loss": 0.6659,
+      "step": 2547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9384283089727256,
+      "learning_rate": 1.8400871280091274e-05,
+      "loss": 0.62,
+      "step": 2548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8741990456683056,
+      "learning_rate": 1.839944300040798e-05,
+      "loss": 0.5506,
+      "step": 2549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9156548395431,
+      "learning_rate": 1.8398014138650742e-05,
+      "loss": 0.575,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9491831775479801,
+      "learning_rate": 1.839658469491857e-05,
+      "loss": 0.5716,
+      "step": 2551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.970076589441962,
+      "learning_rate": 1.839515466931053e-05,
+      "loss": 0.6099,
+      "step": 2552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.004362696843191,
+      "learning_rate": 1.8393724061925714e-05,
+      "loss": 0.6277,
+      "step": 2553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9440776440870434,
+      "learning_rate": 1.839229287286327e-05,
+      "loss": 0.5533,
+      "step": 2554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9078265991640199,
+      "learning_rate": 1.839086110222237e-05,
+      "loss": 0.5973,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9495101497226277,
+      "learning_rate": 1.8389428750102238e-05,
+      "loss": 0.5827,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9223923326893704,
+      "learning_rate": 1.8387995816602137e-05,
+      "loss": 0.5098,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8775065736224732,
+      "learning_rate": 1.8386562301821363e-05,
+      "loss": 0.5311,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9730830661108568,
+      "learning_rate": 1.8385128205859267e-05,
+      "loss": 0.5869,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9741033272488008,
+      "learning_rate": 1.8383693528815218e-05,
+      "loss": 0.5783,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9586616755962616,
+      "learning_rate": 1.8382258270788648e-05,
+      "loss": 0.5866,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9620923146243326,
+      "learning_rate": 1.8380822431879012e-05,
+      "loss": 0.5727,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9470891414477499,
+      "learning_rate": 1.8379386012185813e-05,
+      "loss": 0.5911,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1488547373729767,
+      "learning_rate": 1.83779490118086e-05,
+      "loss": 0.626,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0020985253821808,
+      "learning_rate": 1.837651143084695e-05,
+      "loss": 0.5889,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9519069969151185,
+      "learning_rate": 1.8375073269400488e-05,
+      "loss": 0.6057,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.040029345268677,
+      "learning_rate": 1.8373634527568877e-05,
+      "loss": 0.4678,
+      "step": 2567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0185322739046456,
+      "learning_rate": 1.8372195205451822e-05,
+      "loss": 0.6452,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9846593231833829,
+      "learning_rate": 1.8370755303149064e-05,
+      "loss": 0.5689,
+      "step": 2569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0532067696810197,
+      "learning_rate": 1.8369314820760386e-05,
+      "loss": 0.6109,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.062830205061241,
+      "learning_rate": 1.836787375838562e-05,
+      "loss": 0.6192,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8932031597560607,
+      "learning_rate": 1.836643211612462e-05,
+      "loss": 0.5726,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.938840490483992,
+      "learning_rate": 1.8364989894077297e-05,
+      "loss": 0.6245,
+      "step": 2573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9301251286808755,
+      "learning_rate": 1.8363547092343593e-05,
+      "loss": 0.5844,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9830204831340076,
+      "learning_rate": 1.8362103711023498e-05,
+      "loss": 0.5838,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8233563546163213,
+      "learning_rate": 1.836065975021703e-05,
+      "loss": 0.4613,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8804645354736946,
+      "learning_rate": 1.835921521002426e-05,
+      "loss": 0.6636,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9507404367699335,
+      "learning_rate": 1.8357770090545285e-05,
+      "loss": 0.5366,
+      "step": 2578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.932122035312336,
+      "learning_rate": 1.835632439188026e-05,
+      "loss": 0.6617,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9172867112989296,
+      "learning_rate": 1.8354878114129368e-05,
+      "loss": 0.5361,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9350228745493447,
+      "learning_rate": 1.835343125739283e-05,
+      "loss": 0.5781,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0337044836734066,
+      "learning_rate": 1.8351983821770915e-05,
+      "loss": 0.5823,
+      "step": 2582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9051354658937525,
+      "learning_rate": 1.835053580736393e-05,
+      "loss": 0.5522,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.081713649725828,
+      "learning_rate": 1.8349087214272222e-05,
+      "loss": 0.6442,
+      "step": 2584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.986286911076109,
+      "learning_rate": 1.8347638042596177e-05,
+      "loss": 0.5924,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9145313347195219,
+      "learning_rate": 1.834618829243622e-05,
+      "loss": 0.5414,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.033848476051229,
+      "learning_rate": 1.8344737963892813e-05,
+      "loss": 0.6362,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9634793666586047,
+      "learning_rate": 1.834328705706647e-05,
+      "loss": 0.5892,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8135673361799275,
+      "learning_rate": 1.8341835572057735e-05,
+      "loss": 0.5019,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9544914367776864,
+      "learning_rate": 1.834038350896719e-05,
+      "loss": 0.5692,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9695356063078714,
+      "learning_rate": 1.833893086789547e-05,
+      "loss": 0.6074,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8833866555407887,
+      "learning_rate": 1.8337477648943236e-05,
+      "loss": 0.5801,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0225031754148959,
+      "learning_rate": 1.8336023852211197e-05,
+      "loss": 0.5934,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9512387406439328,
+      "learning_rate": 1.83345694778001e-05,
+      "loss": 0.6627,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9865220706641923,
+      "learning_rate": 1.8333114525810726e-05,
+      "loss": 0.5865,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0745395462324694,
+      "learning_rate": 1.833165899634391e-05,
+      "loss": 0.6437,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9406556691737249,
+      "learning_rate": 1.8330202889500518e-05,
+      "loss": 0.6085,
+      "step": 2597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9666420944891609,
+      "learning_rate": 1.8328746205381453e-05,
+      "loss": 0.5894,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9022430872716137,
+      "learning_rate": 1.8327288944087663e-05,
+      "loss": 0.5847,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9809080782353568,
+      "learning_rate": 1.8325831105720135e-05,
+      "loss": 0.6222,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9673317294393174,
+      "learning_rate": 1.8324372690379896e-05,
+      "loss": 0.5582,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9761900184362405,
+      "learning_rate": 1.8322913698168014e-05,
+      "loss": 0.6216,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9661851831216987,
+      "learning_rate": 1.8321454129185597e-05,
+      "loss": 0.5772,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9906529066084958,
+      "learning_rate": 1.831999398353379e-05,
+      "loss": 0.5657,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0410741717992296,
+      "learning_rate": 1.831853326131378e-05,
+      "loss": 0.5796,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.105172965112674,
+      "learning_rate": 1.831707196262679e-05,
+      "loss": 0.5994,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0392581992735035,
+      "learning_rate": 1.8315610087574088e-05,
+      "loss": 0.6008,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.053043125787598,
+      "learning_rate": 1.831414763625699e-05,
+      "loss": 0.6217,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.085606036923765,
+      "learning_rate": 1.831268460877683e-05,
+      "loss": 0.5928,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9618254873284892,
+      "learning_rate": 1.8311221005235e-05,
+      "loss": 0.566,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9879925417829869,
+      "learning_rate": 1.830975682573293e-05,
+      "loss": 0.6077,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9013566267896668,
+      "learning_rate": 1.8308292070372084e-05,
+      "loss": 0.6134,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9471939550715492,
+      "learning_rate": 1.8306826739253965e-05,
+      "loss": 0.554,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0117778301036309,
+      "learning_rate": 1.8305360832480118e-05,
+      "loss": 0.601,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8994234802143227,
+      "learning_rate": 1.8303894350152138e-05,
+      "loss": 0.4759,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.009902132319506,
+      "learning_rate": 1.830242729237164e-05,
+      "loss": 0.6027,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9314631626290966,
+      "learning_rate": 1.8300959659240292e-05,
+      "loss": 0.5654,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9663152894675163,
+      "learning_rate": 1.829949145085981e-05,
+      "loss": 0.6001,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0527759807116104,
+      "learning_rate": 1.829802266733193e-05,
+      "loss": 0.554,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9441814027278527,
+      "learning_rate": 1.829655330875844e-05,
+      "loss": 0.5764,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9669943416679583,
+      "learning_rate": 1.829508337524116e-05,
+      "loss": 0.5185,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9532324404339418,
+      "learning_rate": 1.8293612866881965e-05,
+      "loss": 0.6193,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9122909261664486,
+      "learning_rate": 1.8292141783782754e-05,
+      "loss": 0.5344,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9184525412283985,
+      "learning_rate": 1.829067012604547e-05,
+      "loss": 0.6079,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0400039581612621,
+      "learning_rate": 1.8289197893772103e-05,
+      "loss": 0.6262,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0703975227084082,
+      "learning_rate": 1.8287725087064673e-05,
+      "loss": 0.5594,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9508531842550776,
+      "learning_rate": 1.8286251706025245e-05,
+      "loss": 0.5872,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0260098313678139,
+      "learning_rate": 1.828477775075592e-05,
+      "loss": 0.5755,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9910777135295307,
+      "learning_rate": 1.8283303221358854e-05,
+      "loss": 0.6457,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9249482186538088,
+      "learning_rate": 1.8281828117936217e-05,
+      "loss": 0.5589,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9602553560831182,
+      "learning_rate": 1.8280352440590236e-05,
+      "loss": 0.6608,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0260206878932465,
+      "learning_rate": 1.827887618942318e-05,
+      "loss": 0.5475,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0142523857260723,
+      "learning_rate": 1.8277399364537345e-05,
+      "loss": 0.6027,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9068347442951924,
+      "learning_rate": 1.8275921966035076e-05,
+      "loss": 0.5952,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9662920231000809,
+      "learning_rate": 1.8274443994018754e-05,
+      "loss": 0.608,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9831243383495013,
+      "learning_rate": 1.8272965448590807e-05,
+      "loss": 0.5982,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0072628424109062,
+      "learning_rate": 1.827148632985369e-05,
+      "loss": 0.5552,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9561275626706207,
+      "learning_rate": 1.8270006637909907e-05,
+      "loss": 0.5363,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9382299009178503,
+      "learning_rate": 1.8268526372862e-05,
+      "loss": 0.61,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.036330966633798,
+      "learning_rate": 1.8267045534812547e-05,
+      "loss": 0.5729,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0591982746515018,
+      "learning_rate": 1.8265564123864174e-05,
+      "loss": 0.5942,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0280607493711271,
+      "learning_rate": 1.826408214011954e-05,
+      "loss": 0.6189,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9694580927695198,
+      "learning_rate": 1.826259958368134e-05,
+      "loss": 0.5646,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9148632149342607,
+      "learning_rate": 1.826111645465232e-05,
+      "loss": 0.5231,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0528796357090346,
+      "learning_rate": 1.8259632753135257e-05,
+      "loss": 0.6494,
+      "step": 2645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8927493697457675,
+      "learning_rate": 1.825814847923297e-05,
+      "loss": 0.605,
+      "step": 2646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9746220924690185,
+      "learning_rate": 1.825666363304832e-05,
+      "loss": 0.5596,
+      "step": 2647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.104502002989997,
+      "learning_rate": 1.82551782146842e-05,
+      "loss": 0.5217,
+      "step": 2648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9522011695494644,
+      "learning_rate": 1.825369222424356e-05,
+      "loss": 0.5457,
+      "step": 2649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9974276810353772,
+      "learning_rate": 1.8252205661829364e-05,
+      "loss": 0.6035,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8536988050512456,
+      "learning_rate": 1.8250718527544636e-05,
+      "loss": 0.561,
+      "step": 2651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.014452009587213,
+      "learning_rate": 1.824923082149243e-05,
+      "loss": 0.5939,
+      "step": 2652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0286657701497812,
+      "learning_rate": 1.824774254377585e-05,
+      "loss": 0.6005,
+      "step": 2653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8738102658423137,
+      "learning_rate": 1.8246253694498024e-05,
+      "loss": 0.5788,
+      "step": 2654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9880182661920288,
+      "learning_rate": 1.8244764273762133e-05,
+      "loss": 0.632,
+      "step": 2655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8950089128352056,
+      "learning_rate": 1.8243274281671392e-05,
+      "loss": 0.5739,
+      "step": 2656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.904422421895005,
+      "learning_rate": 1.824178371832905e-05,
+      "loss": 0.603,
+      "step": 2657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9873253816907256,
+      "learning_rate": 1.824029258383841e-05,
+      "loss": 0.5945,
+      "step": 2658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1132475701850797,
+      "learning_rate": 1.8238800878302804e-05,
+      "loss": 0.6235,
+      "step": 2659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1216405221969532,
+      "learning_rate": 1.8237308601825604e-05,
+      "loss": 0.5969,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0266161431706164,
+      "learning_rate": 1.8235815754510227e-05,
+      "loss": 0.519,
+      "step": 2661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9705185677800771,
+      "learning_rate": 1.823432233646012e-05,
+      "loss": 0.5804,
+      "step": 2662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9579417158184198,
+      "learning_rate": 1.8232828347778778e-05,
+      "loss": 0.675,
+      "step": 2663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.999957902123899,
+      "learning_rate": 1.8231333788569737e-05,
+      "loss": 0.5755,
+      "step": 2664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.996708421437918,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.6197,
+      "step": 2665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8644195494317504,
+      "learning_rate": 1.8228342958982874e-05,
+      "loss": 0.5294,
+      "step": 2666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9366598979644085,
+      "learning_rate": 1.8226846688812314e-05,
+      "loss": 0.6038,
+      "step": 2667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8966307476883493,
+      "learning_rate": 1.8225349848528574e-05,
+      "loss": 0.5667,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0839760777914489,
+      "learning_rate": 1.822385243823539e-05,
+      "loss": 0.6408,
+      "step": 2669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8627457075836641,
+      "learning_rate": 1.8222354458036523e-05,
+      "loss": 0.5858,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9506321931490626,
+      "learning_rate": 1.8220855908035783e-05,
+      "loss": 0.5744,
+      "step": 2671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8980323219691284,
+      "learning_rate": 1.8219356788337027e-05,
+      "loss": 0.588,
+      "step": 2672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9614101469490036,
+      "learning_rate": 1.8217857099044128e-05,
+      "loss": 0.5708,
+      "step": 2673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9732045139725454,
+      "learning_rate": 1.8216356840261028e-05,
+      "loss": 0.6515,
+      "step": 2674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9802795442791357,
+      "learning_rate": 1.8214856012091684e-05,
+      "loss": 0.634,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9608737963408859,
+      "learning_rate": 1.8213354614640105e-05,
+      "loss": 0.6326,
+      "step": 2676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9140455271956307,
+      "learning_rate": 1.8211852648010338e-05,
+      "loss": 0.6295,
+      "step": 2677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9696423375138519,
+      "learning_rate": 1.8210350112306466e-05,
+      "loss": 0.5804,
+      "step": 2678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8442936328053501,
+      "learning_rate": 1.8208847007632613e-05,
+      "loss": 0.5968,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9691958315893788,
+      "learning_rate": 1.8207343334092944e-05,
+      "loss": 0.6059,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.882560444021095,
+      "learning_rate": 1.820583909179166e-05,
+      "loss": 0.5787,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.036455415023574,
+      "learning_rate": 1.8204334280833005e-05,
+      "loss": 0.5966,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0057190615511393,
+      "learning_rate": 1.8202828901321265e-05,
+      "loss": 0.6027,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.020603493240722,
+      "learning_rate": 1.8201322953360758e-05,
+      "loss": 0.5888,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9371289052205584,
+      "learning_rate": 1.8199816437055843e-05,
+      "loss": 0.5487,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0491755776905296,
+      "learning_rate": 1.8198309352510924e-05,
+      "loss": 0.6072,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8660714216318303,
+      "learning_rate": 1.8196801699830437e-05,
+      "loss": 0.6076,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9862596290015903,
+      "learning_rate": 1.8195293479118863e-05,
+      "loss": 0.6277,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0437377771063885,
+      "learning_rate": 1.819378469048072e-05,
+      "loss": 0.5915,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9439824865798379,
+      "learning_rate": 1.8192275334020565e-05,
+      "loss": 0.5828,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0225402759808864,
+      "learning_rate": 1.8190765409842997e-05,
+      "loss": 0.6008,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9958776066891274,
+      "learning_rate": 1.818925491805265e-05,
+      "loss": 0.5392,
+      "step": 2692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9329456271958139,
+      "learning_rate": 1.8187743858754206e-05,
+      "loss": 0.6347,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9906642944077385,
+      "learning_rate": 1.818623223205237e-05,
+      "loss": 0.6293,
+      "step": 2694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.940956884211602,
+      "learning_rate": 1.8184720038051905e-05,
+      "loss": 0.5864,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9430103458543017,
+      "learning_rate": 1.8183207276857596e-05,
+      "loss": 0.5725,
+      "step": 2696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9750056603913353,
+      "learning_rate": 1.8181693948574285e-05,
+      "loss": 0.6003,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.936158816403639,
+      "learning_rate": 1.818018005330684e-05,
+      "loss": 0.5746,
+      "step": 2698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0049390990235507,
+      "learning_rate": 1.817866559116017e-05,
+      "loss": 0.6166,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9979185376293706,
+      "learning_rate": 1.8177150562239236e-05,
+      "loss": 0.5793,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0203902925197001,
+      "learning_rate": 1.8175634966649015e-05,
+      "loss": 0.5755,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9142012279899165,
+      "learning_rate": 1.8174118804494548e-05,
+      "loss": 0.5524,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0338022613929347,
+      "learning_rate": 1.8172602075880893e-05,
+      "loss": 0.6003,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0650332972018386,
+      "learning_rate": 1.8171084780913165e-05,
+      "loss": 0.5795,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9413835902741449,
+      "learning_rate": 1.8169566919696512e-05,
+      "loss": 0.6581,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9301794167686281,
+      "learning_rate": 1.8168048492336116e-05,
+      "loss": 0.4732,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9971512309322477,
+      "learning_rate": 1.81665294989372e-05,
+      "loss": 0.5338,
+      "step": 2707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9626054977861396,
+      "learning_rate": 1.8165009939605037e-05,
+      "loss": 0.575,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9686254044597287,
+      "learning_rate": 1.816348981444493e-05,
+      "loss": 0.6684,
+      "step": 2709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.817821965462608,
+      "learning_rate": 1.816196912356222e-05,
+      "loss": 0.5142,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9906203998916217,
+      "learning_rate": 1.8160447867062286e-05,
+      "loss": 0.6254,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0145536544226452,
+      "learning_rate": 1.8158926045050553e-05,
+      "loss": 0.6246,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8877817517453693,
+      "learning_rate": 1.8157403657632485e-05,
+      "loss": 0.5587,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9169924356992728,
+      "learning_rate": 1.8155880704913577e-05,
+      "loss": 0.5368,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7941288099049386,
+      "learning_rate": 1.8154357186999368e-05,
+      "loss": 0.5503,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9409239063898954,
+      "learning_rate": 1.8152833103995443e-05,
+      "loss": 0.6428,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9536368923356101,
+      "learning_rate": 1.8151308456007416e-05,
+      "loss": 0.6105,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9242821682058803,
+      "learning_rate": 1.814978324314094e-05,
+      "loss": 0.5262,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8923447734732718,
+      "learning_rate": 1.8148257465501718e-05,
+      "loss": 0.5968,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0346232613906077,
+      "learning_rate": 1.814673112319548e-05,
+      "loss": 0.6651,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8967208895801116,
+      "learning_rate": 1.8145204216327998e-05,
+      "loss": 0.5616,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9789527728091927,
+      "learning_rate": 1.8143676745005093e-05,
+      "loss": 0.5564,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9301175229831795,
+      "learning_rate": 1.814214870933261e-05,
+      "loss": 0.5414,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9279945540168463,
+      "learning_rate": 1.8140620109416445e-05,
+      "loss": 0.5891,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9583822050583957,
+      "learning_rate": 1.8139090945362525e-05,
+      "loss": 0.6199,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.963837056971241,
+      "learning_rate": 1.8137561217276823e-05,
+      "loss": 0.5775,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9193008965088793,
+      "learning_rate": 1.8136030925265347e-05,
+      "loss": 0.6201,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9595001172487877,
+      "learning_rate": 1.8134500069434144e-05,
+      "loss": 0.5929,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8662619104759794,
+      "learning_rate": 1.81329686498893e-05,
+      "loss": 0.5501,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.972240880760536,
+      "learning_rate": 1.8131436666736945e-05,
+      "loss": 0.5691,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9615177720700632,
+      "learning_rate": 1.8129904120083243e-05,
+      "loss": 0.5681,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9229624236657873,
+      "learning_rate": 1.8128371010034394e-05,
+      "loss": 0.555,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8890895179572698,
+      "learning_rate": 1.8126837336696645e-05,
+      "loss": 0.5756,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9067723189735549,
+      "learning_rate": 1.8125303100176275e-05,
+      "loss": 0.5928,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9007052662584437,
+      "learning_rate": 1.812376830057961e-05,
+      "loss": 0.6127,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9437405590348689,
+      "learning_rate": 1.812223293801301e-05,
+      "loss": 0.6169,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9646217361783188,
+      "learning_rate": 1.8120697012582863e-05,
+      "loss": 0.5928,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9682755714496285,
+      "learning_rate": 1.8119160524395622e-05,
+      "loss": 0.6008,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.999055226351776,
+      "learning_rate": 1.8117623473557758e-05,
+      "loss": 0.6105,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9724026619999333,
+      "learning_rate": 1.8116085860175788e-05,
+      "loss": 0.5347,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0081974793755035,
+      "learning_rate": 1.8114547684356264e-05,
+      "loss": 0.5553,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0293986423368018,
+      "learning_rate": 1.8113008946205787e-05,
+      "loss": 0.6537,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9711812270768454,
+      "learning_rate": 1.8111469645830983e-05,
+      "loss": 0.5677,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8855189900145181,
+      "learning_rate": 1.810992978333853e-05,
+      "loss": 0.5616,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9973233513608902,
+      "learning_rate": 1.8108389358835135e-05,
+      "loss": 0.5849,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.919178257973744,
+      "learning_rate": 1.810684837242755e-05,
+      "loss": 0.6009,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2479235413643202,
+      "learning_rate": 1.810530682422256e-05,
+      "loss": 0.612,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0580936700356294,
+      "learning_rate": 1.8103764714327004e-05,
+      "loss": 0.5988,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.016274277429732,
+      "learning_rate": 1.8102222042847735e-05,
+      "loss": 0.5768,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0392968760654928,
+      "learning_rate": 1.8100678809891668e-05,
+      "loss": 0.6248,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9281711927515615,
+      "learning_rate": 1.8099135015565745e-05,
+      "loss": 0.58,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.893762521635862,
+      "learning_rate": 1.8097590659976946e-05,
+      "loss": 0.5739,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8673575882784395,
+      "learning_rate": 1.8096045743232303e-05,
+      "loss": 0.5668,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9189763319220341,
+      "learning_rate": 1.8094500265438866e-05,
+      "loss": 0.5817,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8740099428626087,
+      "learning_rate": 1.8092954226703742e-05,
+      "loss": 0.5745,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0177735915260575,
+      "learning_rate": 1.8091407627134067e-05,
+      "loss": 0.6324,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8755765204054051,
+      "learning_rate": 1.8089860466837023e-05,
+      "loss": 0.591,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9844508074811542,
+      "learning_rate": 1.8088312745919823e-05,
+      "loss": 0.6863,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9178781550779036,
+      "learning_rate": 1.8086764464489723e-05,
+      "loss": 0.5362,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8438837000342646,
+      "learning_rate": 1.8085215622654023e-05,
+      "loss": 0.5502,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.027995036155377,
+      "learning_rate": 1.8083666220520045e-05,
+      "loss": 0.6827,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0072627378618417,
+      "learning_rate": 1.8082116258195173e-05,
+      "loss": 0.609,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0404432889174489,
+      "learning_rate": 1.8080565735786813e-05,
+      "loss": 0.5405,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.960224909807464,
+      "learning_rate": 1.8079014653402414e-05,
+      "loss": 0.5984,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8957444706981957,
+      "learning_rate": 1.8077463011149464e-05,
+      "loss": 0.5948,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8909499775341096,
+      "learning_rate": 1.80759108091355e-05,
+      "loss": 0.5498,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9552071296438046,
+      "learning_rate": 1.807435804746807e-05,
+      "loss": 0.5756,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0301741647540563,
+      "learning_rate": 1.8072804726254792e-05,
+      "loss": 0.6114,
+      "step": 2768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9687543183786282,
+      "learning_rate": 1.807125084560331e-05,
+      "loss": 0.5714,
+      "step": 2769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8673956813094961,
+      "learning_rate": 1.80696964056213e-05,
+      "loss": 0.572,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8849403306028029,
+      "learning_rate": 1.8068141406416487e-05,
+      "loss": 0.5341,
+      "step": 2771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9106492006599914,
+      "learning_rate": 1.8066585848096637e-05,
+      "loss": 0.6088,
+      "step": 2772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9742053568515937,
+      "learning_rate": 1.8065029730769534e-05,
+      "loss": 0.6063,
+      "step": 2773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9903274449807491,
+      "learning_rate": 1.806347305454303e-05,
+      "loss": 0.5429,
+      "step": 2774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9181460398525245,
+      "learning_rate": 1.8061915819524995e-05,
+      "loss": 0.5536,
+      "step": 2775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0098508357388716,
+      "learning_rate": 1.8060358025823344e-05,
+      "loss": 0.6372,
+      "step": 2776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9213271798021149,
+      "learning_rate": 1.8058799673546032e-05,
+      "loss": 0.5686,
+      "step": 2777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9557490562689361,
+      "learning_rate": 1.805724076280105e-05,
+      "loss": 0.5554,
+      "step": 2778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0084940234973645,
+      "learning_rate": 1.805568129369643e-05,
+      "loss": 0.665,
+      "step": 2779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9126637077556393,
+      "learning_rate": 1.805412126634024e-05,
+      "loss": 0.5613,
+      "step": 2780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9577382335710649,
+      "learning_rate": 1.8052560680840595e-05,
+      "loss": 0.6643,
+      "step": 2781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0449965708611912,
+      "learning_rate": 1.8050999537305634e-05,
+      "loss": 0.5253,
+      "step": 2782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9938981486681303,
+      "learning_rate": 1.8049437835843545e-05,
+      "loss": 0.5189,
+      "step": 2783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9753113641821504,
+      "learning_rate": 1.8047875576562556e-05,
+      "loss": 0.6217,
+      "step": 2784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9553770251097057,
+      "learning_rate": 1.8046312759570924e-05,
+      "loss": 0.5544,
+      "step": 2785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9919726546590415,
+      "learning_rate": 1.804474938497696e-05,
+      "loss": 0.6089,
+      "step": 2786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9808570279080948,
+      "learning_rate": 1.8043185452888997e-05,
+      "loss": 0.5294,
+      "step": 2787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0518621525037852,
+      "learning_rate": 1.8041620963415418e-05,
+      "loss": 0.5776,
+      "step": 2788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1022635177347593,
+      "learning_rate": 1.804005591666464e-05,
+      "loss": 0.6217,
+      "step": 2789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9429626058610409,
+      "learning_rate": 1.8038490312745116e-05,
+      "loss": 0.6275,
+      "step": 2790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.930398967716542,
+      "learning_rate": 1.8036924151765345e-05,
+      "loss": 0.5554,
+      "step": 2791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.887025266457902,
+      "learning_rate": 1.803535743383386e-05,
+      "loss": 0.5701,
+      "step": 2792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9402776215475784,
+      "learning_rate": 1.8033790159059224e-05,
+      "loss": 0.6054,
+      "step": 2793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0326325235514582,
+      "learning_rate": 1.8032222327550063e-05,
+      "loss": 0.6075,
+      "step": 2794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9775021711373921,
+      "learning_rate": 1.803065393941502e-05,
+      "loss": 0.5606,
+      "step": 2795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9157898288063075,
+      "learning_rate": 1.802908499476278e-05,
+      "loss": 0.5374,
+      "step": 2796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8959173642295047,
+      "learning_rate": 1.8027515493702075e-05,
+      "loss": 0.487,
+      "step": 2797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9516732846513793,
+      "learning_rate": 1.8025945436341663e-05,
+      "loss": 0.5696,
+      "step": 2798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9225055633764292,
+      "learning_rate": 1.8024374822790355e-05,
+      "loss": 0.5725,
+      "step": 2799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9139056227761806,
+      "learning_rate": 1.8022803653156983e-05,
+      "loss": 0.5483,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9438666489422416,
+      "learning_rate": 1.802123192755044e-05,
+      "loss": 0.5711,
+      "step": 2801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.994929728112408,
+      "learning_rate": 1.8019659646079636e-05,
+      "loss": 0.5899,
+      "step": 2802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9894663256980869,
+      "learning_rate": 1.8018086808853535e-05,
+      "loss": 0.6254,
+      "step": 2803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0052732190279563,
+      "learning_rate": 1.8016513415981128e-05,
+      "loss": 0.5144,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0144845417320454,
+      "learning_rate": 1.801493946757145e-05,
+      "loss": 0.5835,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0252420292101474,
+      "learning_rate": 1.801336496373358e-05,
+      "loss": 0.5536,
+      "step": 2806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8971116968992333,
+      "learning_rate": 1.8011789904576624e-05,
+      "loss": 0.5506,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8970316281501216,
+      "learning_rate": 1.8010214290209735e-05,
+      "loss": 0.6684,
+      "step": 2808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9915460509347839,
+      "learning_rate": 1.80086381207421e-05,
+      "loss": 0.6473,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9679950558017996,
+      "learning_rate": 1.8007061396282944e-05,
+      "loss": 0.5946,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9297039333204747,
+      "learning_rate": 1.800548411694154e-05,
+      "loss": 0.5475,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8258902225497208,
+      "learning_rate": 1.8003906282827186e-05,
+      "loss": 0.5406,
+      "step": 2812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7445951245744602,
+      "learning_rate": 1.8002327894049225e-05,
+      "loss": 0.5067,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8985526515890513,
+      "learning_rate": 1.800074895071704e-05,
+      "loss": 0.579,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0104156103033566,
+      "learning_rate": 1.799916945294005e-05,
+      "loss": 0.6262,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0213484101551884,
+      "learning_rate": 1.7997589400827712e-05,
+      "loss": 0.6357,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0685631612624997,
+      "learning_rate": 1.799600879448952e-05,
+      "loss": 0.5975,
+      "step": 2817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9967993589986436,
+      "learning_rate": 1.7994427634035016e-05,
+      "loss": 0.5412,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9180677224628448,
+      "learning_rate": 1.799284591957376e-05,
+      "loss": 0.5347,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9747429803370978,
+      "learning_rate": 1.799126365121538e-05,
+      "loss": 0.6033,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9163259787114536,
+      "learning_rate": 1.798968082906951e-05,
+      "loss": 0.5792,
+      "step": 2821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9211695025727105,
+      "learning_rate": 1.798809745324585e-05,
+      "loss": 0.5119,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8676048846449058,
+      "learning_rate": 1.798651352385412e-05,
+      "loss": 0.5284,
+      "step": 2823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9015393452040311,
+      "learning_rate": 1.798492904100409e-05,
+      "loss": 0.4627,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9387971158929248,
+      "learning_rate": 1.7983344004805555e-05,
+      "loss": 0.6055,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8471959162297319,
+      "learning_rate": 1.7981758415368365e-05,
+      "loss": 0.5055,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.988588879294854,
+      "learning_rate": 1.7980172272802398e-05,
+      "loss": 0.5528,
+      "step": 2827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9326776260735447,
+      "learning_rate": 1.7978585577217568e-05,
+      "loss": 0.5196,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8966730811382917,
+      "learning_rate": 1.7976998328723833e-05,
+      "loss": 0.5468,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8490311284377158,
+      "learning_rate": 1.7975410527431195e-05,
+      "loss": 0.5447,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8813316008855758,
+      "learning_rate": 1.797382217344968e-05,
+      "loss": 0.4973,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.089434215108231,
+      "learning_rate": 1.7972233266889356e-05,
+      "loss": 0.6135,
+      "step": 2832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9659587366357253,
+      "learning_rate": 1.797064380786034e-05,
+      "loss": 0.5099,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9218666717431676,
+      "learning_rate": 1.7969053796472783e-05,
+      "loss": 0.5723,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9277864344934234,
+      "learning_rate": 1.796746323283686e-05,
+      "loss": 0.5605,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9776799469352457,
+      "learning_rate": 1.7965872117062806e-05,
+      "loss": 0.5077,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9318298969039761,
+      "learning_rate": 1.796428044926088e-05,
+      "loss": 0.6138,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9075948043353205,
+      "learning_rate": 1.7962688229541382e-05,
+      "loss": 0.5704,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9306145438415679,
+      "learning_rate": 1.7961095458014655e-05,
+      "loss": 0.6038,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8884000710243785,
+      "learning_rate": 1.795950213479107e-05,
+      "loss": 0.5054,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8569607622945014,
+      "learning_rate": 1.795790825998105e-05,
+      "loss": 0.5184,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9791079837049262,
+      "learning_rate": 1.7956313833695046e-05,
+      "loss": 0.5739,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9588531159827303,
+      "learning_rate": 1.795471885604355e-05,
+      "loss": 0.5694,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0297830187508645,
+      "learning_rate": 1.7953123327137093e-05,
+      "loss": 0.6476,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9288421640089412,
+      "learning_rate": 1.7951527247086243e-05,
+      "loss": 0.5351,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9602300038067393,
+      "learning_rate": 1.794993061600161e-05,
+      "loss": 0.596,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9075377212248501,
+      "learning_rate": 1.7948333433993833e-05,
+      "loss": 0.5389,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.995373021003706,
+      "learning_rate": 1.7946735701173604e-05,
+      "loss": 0.6393,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9164049272640322,
+      "learning_rate": 1.7945137417651638e-05,
+      "loss": 0.6477,
+      "step": 2849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9477115032759619,
+      "learning_rate": 1.7943538583538696e-05,
+      "loss": 0.6198,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.90846432248897,
+      "learning_rate": 1.7941939198945574e-05,
+      "loss": 0.52,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9174204792362805,
+      "learning_rate": 1.7940339263983112e-05,
+      "loss": 0.5913,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9792924779362212,
+      "learning_rate": 1.7938738778762182e-05,
+      "loss": 0.5644,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9005528079322241,
+      "learning_rate": 1.7937137743393695e-05,
+      "loss": 0.5532,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9674891956317448,
+      "learning_rate": 1.7935536157988605e-05,
+      "loss": 0.5931,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0392442539723934,
+      "learning_rate": 1.79339340226579e-05,
+      "loss": 0.5858,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9010473425300569,
+      "learning_rate": 1.79323313375126e-05,
+      "loss": 0.5254,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.908624006675908,
+      "learning_rate": 1.7930728102663775e-05,
+      "loss": 0.5139,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0041814177937274,
+      "learning_rate": 1.792912431822253e-05,
+      "loss": 0.591,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8424943230616456,
+      "learning_rate": 1.79275199843e-05,
+      "loss": 0.5597,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9827303241823689,
+      "learning_rate": 1.7925915101007366e-05,
+      "loss": 0.5951,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9747069232699177,
+      "learning_rate": 1.792430966845585e-05,
+      "loss": 0.5873,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0293522334126957,
+      "learning_rate": 1.7922703686756697e-05,
+      "loss": 0.6401,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.011927611023172,
+      "learning_rate": 1.792109715602121e-05,
+      "loss": 0.5973,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9273786221962907,
+      "learning_rate": 1.7919490076360714e-05,
+      "loss": 0.6032,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.093940323247069,
+      "learning_rate": 1.7917882447886585e-05,
+      "loss": 0.6462,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.890019833154301,
+      "learning_rate": 1.7916274270710218e-05,
+      "loss": 0.6138,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9769921244269482,
+      "learning_rate": 1.7914665544943072e-05,
+      "loss": 0.6104,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9485632509377442,
+      "learning_rate": 1.791305627069662e-05,
+      "loss": 0.6246,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0018934058623412,
+      "learning_rate": 1.791144644808239e-05,
+      "loss": 0.6317,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9875398922579854,
+      "learning_rate": 1.7909836077211936e-05,
+      "loss": 0.6921,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8942446600706555,
+      "learning_rate": 1.790822515819686e-05,
+      "loss": 0.5872,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9604170179855209,
+      "learning_rate": 1.7906613691148796e-05,
+      "loss": 0.5829,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8942090505558472,
+      "learning_rate": 1.7905001676179414e-05,
+      "loss": 0.5426,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0487502187628917,
+      "learning_rate": 1.7903389113400427e-05,
+      "loss": 0.5667,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.959663943230379,
+      "learning_rate": 1.790177600292359e-05,
+      "loss": 0.572,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9917149402947344,
+      "learning_rate": 1.790016234486068e-05,
+      "loss": 0.5906,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.937418302425181,
+      "learning_rate": 1.789854813932353e-05,
+      "loss": 0.5202,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7861228868319374,
+      "learning_rate": 1.7896933386423998e-05,
+      "loss": 0.4862,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0141607091776144,
+      "learning_rate": 1.7895318086273986e-05,
+      "loss": 0.5141,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0273508303059105,
+      "learning_rate": 1.7893702238985433e-05,
+      "loss": 0.5389,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9399823550683523,
+      "learning_rate": 1.7892085844670318e-05,
+      "loss": 0.5422,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9049524508006479,
+      "learning_rate": 1.7890468903440656e-05,
+      "loss": 0.5719,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9956943417222794,
+      "learning_rate": 1.7888851415408495e-05,
+      "loss": 0.5924,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.883186802287362,
+      "learning_rate": 1.788723338068593e-05,
+      "loss": 0.5544,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9971492225911742,
+      "learning_rate": 1.7885614799385086e-05,
+      "loss": 0.6009,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8355205506640969,
+      "learning_rate": 1.7883995671618133e-05,
+      "loss": 0.5485,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9074735160025519,
+      "learning_rate": 1.7882375997497273e-05,
+      "loss": 0.574,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0213868782570381,
+      "learning_rate": 1.788075577713475e-05,
+      "loss": 0.5807,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.944087464291218,
+      "learning_rate": 1.7879135010642836e-05,
+      "loss": 0.6015,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9115513698895477,
+      "learning_rate": 1.787751369813386e-05,
+      "loss": 0.573,
+      "step": 2891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.937984094940414,
+      "learning_rate": 1.787589183972017e-05,
+      "loss": 0.6324,
+      "step": 2892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1049632286846252,
+      "learning_rate": 1.787426943551416e-05,
+      "loss": 0.6556,
+      "step": 2893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8892815904572252,
+      "learning_rate": 1.7872646485628266e-05,
+      "loss": 0.6171,
+      "step": 2894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.890347163387591,
+      "learning_rate": 1.7871022990174948e-05,
+      "loss": 0.5293,
+      "step": 2895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8997722667227449,
+      "learning_rate": 1.7869398949266724e-05,
+      "loss": 0.565,
+      "step": 2896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9329374331298593,
+      "learning_rate": 1.786777436301613e-05,
+      "loss": 0.5565,
+      "step": 2897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9746227868782539,
+      "learning_rate": 1.7866149231535754e-05,
+      "loss": 0.5568,
+      "step": 2898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8773883378129755,
+      "learning_rate": 1.786452355493821e-05,
+      "loss": 0.5434,
+      "step": 2899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9631742937919332,
+      "learning_rate": 1.7862897333336162e-05,
+      "loss": 0.6124,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9779033162507079,
+      "learning_rate": 1.78612705668423e-05,
+      "loss": 0.6597,
+      "step": 2901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9886719845976905,
+      "learning_rate": 1.7859643255569364e-05,
+      "loss": 0.5886,
+      "step": 2902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9661761439948027,
+      "learning_rate": 1.785801539963012e-05,
+      "loss": 0.5725,
+      "step": 2903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9641311236560048,
+      "learning_rate": 1.785638699913738e-05,
+      "loss": 0.5953,
+      "step": 2904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9824821479543113,
+      "learning_rate": 1.785475805420399e-05,
+      "loss": 0.6451,
+      "step": 2905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9574330850350337,
+      "learning_rate": 1.7853128564942834e-05,
+      "loss": 0.591,
+      "step": 2906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.002233316642695,
+      "learning_rate": 1.7851498531466833e-05,
+      "loss": 0.6056,
+      "step": 2907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0215148904505245,
+      "learning_rate": 1.784986795388895e-05,
+      "loss": 0.6765,
+      "step": 2908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.976072703279664,
+      "learning_rate": 1.7848236832322175e-05,
+      "loss": 0.5366,
+      "step": 2909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0272588509160951,
+      "learning_rate": 1.7846605166879555e-05,
+      "loss": 0.667,
+      "step": 2910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9826121519395441,
+      "learning_rate": 1.7844972957674156e-05,
+      "loss": 0.5788,
+      "step": 2911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.978849061408577,
+      "learning_rate": 1.7843340204819087e-05,
+      "loss": 0.6177,
+      "step": 2912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0479357254213726,
+      "learning_rate": 1.78417069084275e-05,
+      "loss": 0.5245,
+      "step": 2913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9221183355541194,
+      "learning_rate": 1.784007306861258e-05,
+      "loss": 0.623,
+      "step": 2914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9128945043954562,
+      "learning_rate": 1.783843868548755e-05,
+      "loss": 0.5235,
+      "step": 2915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8916286639889864,
+      "learning_rate": 1.7836803759165673e-05,
+      "loss": 0.5629,
+      "step": 2916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0138599166669626,
+      "learning_rate": 1.7835168289760248e-05,
+      "loss": 0.6256,
+      "step": 2917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9382854752966578,
+      "learning_rate": 1.7833532277384607e-05,
+      "loss": 0.561,
+      "step": 2918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8932203928499848,
+      "learning_rate": 1.783189572215213e-05,
+      "loss": 0.5669,
+      "step": 2919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9635695529223409,
+      "learning_rate": 1.7830258624176224e-05,
+      "loss": 0.6151,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8869233883388015,
+      "learning_rate": 1.782862098357034e-05,
+      "loss": 0.5696,
+      "step": 2921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0316281872406312,
+      "learning_rate": 1.782698280044797e-05,
+      "loss": 0.6051,
+      "step": 2922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0455886074412064,
+      "learning_rate": 1.7825344074922633e-05,
+      "loss": 0.6081,
+      "step": 2923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9910576483300347,
+      "learning_rate": 1.782370480710789e-05,
+      "loss": 0.6417,
+      "step": 2924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9651470637576416,
+      "learning_rate": 1.7822064997117348e-05,
+      "loss": 0.5922,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9345754499780848,
+      "learning_rate": 1.7820424645064635e-05,
+      "loss": 0.513,
+      "step": 2926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9903327113291599,
+      "learning_rate": 1.7818783751063433e-05,
+      "loss": 0.6042,
+      "step": 2927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9698448430291365,
+      "learning_rate": 1.7817142315227452e-05,
+      "loss": 0.6555,
+      "step": 2928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9449001506430976,
+      "learning_rate": 1.7815500337670442e-05,
+      "loss": 0.5328,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9499642731886148,
+      "learning_rate": 1.7813857818506194e-05,
+      "loss": 0.5382,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9114911296436924,
+      "learning_rate": 1.7812214757848523e-05,
+      "loss": 0.5718,
+      "step": 2931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0555112155422826,
+      "learning_rate": 1.7810571155811307e-05,
+      "loss": 0.6417,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0012259749389651,
+      "learning_rate": 1.7808927012508436e-05,
+      "loss": 0.5963,
+      "step": 2933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8588757511049373,
+      "learning_rate": 1.7807282328053847e-05,
+      "loss": 0.5725,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8962264917282546,
+      "learning_rate": 1.7805637102561516e-05,
+      "loss": 0.6398,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0126953458804542,
+      "learning_rate": 1.7803991336145462e-05,
+      "loss": 0.5541,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9485077698206339,
+      "learning_rate": 1.7802345028919728e-05,
+      "loss": 0.585,
+      "step": 2937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9528169390915887,
+      "learning_rate": 1.7800698180998406e-05,
+      "loss": 0.6129,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9279286190560277,
+      "learning_rate": 1.7799050792495617e-05,
+      "loss": 0.5792,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.907201226645678,
+      "learning_rate": 1.7797402863525528e-05,
+      "loss": 0.5317,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9794055147062374,
+      "learning_rate": 1.7795754394202334e-05,
+      "loss": 0.5912,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0109889092334905,
+      "learning_rate": 1.7794105384640277e-05,
+      "loss": 0.6199,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0500004928260396,
+      "learning_rate": 1.779245583495363e-05,
+      "loss": 0.5977,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9722339821250309,
+      "learning_rate": 1.7790805745256703e-05,
+      "loss": 0.5688,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9618036445187186,
+      "learning_rate": 1.7789155115663853e-05,
+      "loss": 0.5664,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8426791654221744,
+      "learning_rate": 1.778750394628946e-05,
+      "loss": 0.5497,
+      "step": 2946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9612700706138941,
+      "learning_rate": 1.7785852237247952e-05,
+      "loss": 0.5915,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9261565886562699,
+      "learning_rate": 1.778419998865379e-05,
+      "loss": 0.5658,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9728962618141286,
+      "learning_rate": 1.7782547200621475e-05,
+      "loss": 0.5725,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.045860367370037,
+      "learning_rate": 1.7780893873265536e-05,
+      "loss": 0.6454,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9788455225264904,
+      "learning_rate": 1.777924000670056e-05,
+      "loss": 0.6038,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0945387612010085,
+      "learning_rate": 1.777758560104115e-05,
+      "loss": 0.6038,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8957453512244588,
+      "learning_rate": 1.777593065640195e-05,
+      "loss": 0.5951,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9979474133002743,
+      "learning_rate": 1.777427517289766e-05,
+      "loss": 0.589,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9429662984163877,
+      "learning_rate": 1.7772619150642996e-05,
+      "loss": 0.5365,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9358474973512778,
+      "learning_rate": 1.777096258975272e-05,
+      "loss": 0.6543,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9096719871801495,
+      "learning_rate": 1.7769305490341623e-05,
+      "loss": 0.592,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9537438553085594,
+      "learning_rate": 1.776764785252455e-05,
+      "loss": 0.5968,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9563472444119449,
+      "learning_rate": 1.7765989676416374e-05,
+      "loss": 0.5488,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9289889239231146,
+      "learning_rate": 1.7764330962132e-05,
+      "loss": 0.6065,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9496563907877692,
+      "learning_rate": 1.7762671709786375e-05,
+      "loss": 0.5092,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9243170842255678,
+      "learning_rate": 1.776101191949449e-05,
+      "loss": 0.591,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0710754976090913,
+      "learning_rate": 1.775935159137136e-05,
+      "loss": 0.635,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9780586967963613,
+      "learning_rate": 1.7757690725532048e-05,
+      "loss": 0.5785,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9906049593399756,
+      "learning_rate": 1.7756029322091647e-05,
+      "loss": 0.5637,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.91575458677578,
+      "learning_rate": 1.7754367381165298e-05,
+      "loss": 0.5285,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.977131199479803,
+      "learning_rate": 1.7752704902868164e-05,
+      "loss": 0.6075,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9857150899066561,
+      "learning_rate": 1.775104188731546e-05,
+      "loss": 0.6261,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9576292313368313,
+      "learning_rate": 1.774937833462243e-05,
+      "loss": 0.5617,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8445171511314254,
+      "learning_rate": 1.7747714244904348e-05,
+      "loss": 0.5282,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9229718629347127,
+      "learning_rate": 1.7746049618276545e-05,
+      "loss": 0.5479,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8690650163482897,
+      "learning_rate": 1.7744384454854377e-05,
+      "loss": 0.5541,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9648710618900104,
+      "learning_rate": 1.7742718754753232e-05,
+      "loss": 0.6267,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9128519455250741,
+      "learning_rate": 1.774105251808855e-05,
+      "loss": 0.5086,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.903480230508501,
+      "learning_rate": 1.7739385744975788e-05,
+      "loss": 0.5827,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9501921693907465,
+      "learning_rate": 1.773771843553046e-05,
+      "loss": 0.5899,
+      "step": 2976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9375234127105827,
+      "learning_rate": 1.773605058986811e-05,
+      "loss": 0.5068,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0022123910340575,
+      "learning_rate": 1.7734382208104314e-05,
+      "loss": 0.6204,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9271276387355356,
+      "learning_rate": 1.7732713290354694e-05,
+      "loss": 0.5993,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8412067356706011,
+      "learning_rate": 1.77310438367349e-05,
+      "loss": 0.5356,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8963614258194879,
+      "learning_rate": 1.772937384736063e-05,
+      "loss": 0.5543,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.941153800669109,
+      "learning_rate": 1.77277033223476e-05,
+      "loss": 0.5908,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8840202297399563,
+      "learning_rate": 1.772603226181159e-05,
+      "loss": 0.6028,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9729775167227305,
+      "learning_rate": 1.7724360665868395e-05,
+      "loss": 0.6125,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8596207995883371,
+      "learning_rate": 1.772268853463386e-05,
+      "loss": 0.5613,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0696073709578997,
+      "learning_rate": 1.7721015868223858e-05,
+      "loss": 0.596,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8680112938013765,
+      "learning_rate": 1.7719342666754307e-05,
+      "loss": 0.5516,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9696281293778938,
+      "learning_rate": 1.7717668930341152e-05,
+      "loss": 0.6295,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9755273934994584,
+      "learning_rate": 1.771599465910039e-05,
+      "loss": 0.6291,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9345644481453044,
+      "learning_rate": 1.771431985314804e-05,
+      "loss": 0.5621,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9037245373504185,
+      "learning_rate": 1.7712644512600163e-05,
+      "loss": 0.5962,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0638475033986836,
+      "learning_rate": 1.7710968637572866e-05,
+      "loss": 0.6574,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8344440667470702,
+      "learning_rate": 1.770929222818228e-05,
+      "loss": 0.4977,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8894745605787729,
+      "learning_rate": 1.7707615284544585e-05,
+      "loss": 0.5996,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9326281101476411,
+      "learning_rate": 1.7705937806775986e-05,
+      "loss": 0.5731,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8695247751218795,
+      "learning_rate": 1.7704259794992734e-05,
+      "loss": 0.5895,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0317406351652776,
+      "learning_rate": 1.7702581249311107e-05,
+      "loss": 0.6543,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8077967565584185,
+      "learning_rate": 1.7700902169847434e-05,
+      "loss": 0.518,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.003810213369707,
+      "learning_rate": 1.769922255671807e-05,
+      "loss": 0.5289,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9748601712071312,
+      "learning_rate": 1.7697542410039413e-05,
+      "loss": 0.5974,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9220205485487152,
+      "learning_rate": 1.7695861729927896e-05,
+      "loss": 0.5275,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9939522843169762,
+      "learning_rate": 1.7694180516499986e-05,
+      "loss": 0.6003,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9431839294240236,
+      "learning_rate": 1.769249876987219e-05,
+      "loss": 0.529,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9163038128226331,
+      "learning_rate": 1.7690816490161054e-05,
+      "loss": 0.5529,
+      "step": 3004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9612779979723355,
+      "learning_rate": 1.768913367748316e-05,
+      "loss": 0.5566,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9193225717062538,
+      "learning_rate": 1.7687450331955115e-05,
+      "loss": 0.5653,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8985009069505546,
+      "learning_rate": 1.7685766453693584e-05,
+      "loss": 0.5281,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9995819562166177,
+      "learning_rate": 1.7684082042815255e-05,
+      "loss": 0.5755,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0488660971736616,
+      "learning_rate": 1.768239709943686e-05,
+      "loss": 0.6405,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0060682914657895,
+      "learning_rate": 1.7680711623675155e-05,
+      "loss": 0.6043,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9685403613457841,
+      "learning_rate": 1.767902561564695e-05,
+      "loss": 0.5787,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8615980271016578,
+      "learning_rate": 1.767733907546908e-05,
+      "loss": 0.5722,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9853089580676793,
+      "learning_rate": 1.7675652003258427e-05,
+      "loss": 0.6401,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0187436271337882,
+      "learning_rate": 1.7673964399131895e-05,
+      "loss": 0.5397,
+      "step": 3014
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.034056628820441,
+      "learning_rate": 1.7672276263206433e-05,
+      "loss": 0.5924,
+      "step": 3015
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8746264193882949,
+      "learning_rate": 1.7670587595599034e-05,
+      "loss": 0.4671,
+      "step": 3016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0132891832260866,
+      "learning_rate": 1.7668898396426717e-05,
+      "loss": 0.5868,
+      "step": 3017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8973560235274429,
+      "learning_rate": 1.766720866580655e-05,
+      "loss": 0.5662,
+      "step": 3018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9830770810376366,
+      "learning_rate": 1.7665518403855614e-05,
+      "loss": 0.6166,
+      "step": 3019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9692899867659497,
+      "learning_rate": 1.766382761069106e-05,
+      "loss": 0.6223,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9607598785885566,
+      "learning_rate": 1.7662136286430046e-05,
+      "loss": 0.6089,
+      "step": 3021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9226827698337999,
+      "learning_rate": 1.766044443118978e-05,
+      "loss": 0.5055,
+      "step": 3022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.020928123282797,
+      "learning_rate": 1.7658752045087516e-05,
+      "loss": 0.6092,
+      "step": 3023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.996536739973853,
+      "learning_rate": 1.7657059128240526e-05,
+      "loss": 0.5707,
+      "step": 3024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0197047633927752,
+      "learning_rate": 1.765536568076613e-05,
+      "loss": 0.6489,
+      "step": 3025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9591301041949271,
+      "learning_rate": 1.7653671702781685e-05,
+      "loss": 0.5411,
+      "step": 3026
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8719742546833583,
+      "learning_rate": 1.7651977194404578e-05,
+      "loss": 0.5787,
+      "step": 3027
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.046589072837998,
+      "learning_rate": 1.765028215575224e-05,
+      "loss": 0.5943,
+      "step": 3028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.93558659226997,
+      "learning_rate": 1.7648586586942134e-05,
+      "loss": 0.5325,
+      "step": 3029
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0084710284795635,
+      "learning_rate": 1.764689048809176e-05,
+      "loss": 0.5347,
+      "step": 3030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9808260064257344,
+      "learning_rate": 1.7645193859318658e-05,
+      "loss": 0.614,
+      "step": 3031
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9556969370434116,
+      "learning_rate": 1.7643496700740407e-05,
+      "loss": 0.6143,
+      "step": 3032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9066448474711181,
+      "learning_rate": 1.7641799012474608e-05,
+      "loss": 0.5614,
+      "step": 3033
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.936506703901719,
+      "learning_rate": 1.764010079463892e-05,
+      "loss": 0.5505,
+      "step": 3034
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9555289651987778,
+      "learning_rate": 1.7638402047351025e-05,
+      "loss": 0.6526,
+      "step": 3035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9611056933445485,
+      "learning_rate": 1.7636702770728637e-05,
+      "loss": 0.5742,
+      "step": 3036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9059821946830916,
+      "learning_rate": 1.7635002964889527e-05,
+      "loss": 0.5577,
+      "step": 3037
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9573450830444244,
+      "learning_rate": 1.763330262995148e-05,
+      "loss": 0.6275,
+      "step": 3038
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9601163319268974,
+      "learning_rate": 1.7631601766032337e-05,
+      "loss": 0.5637,
+      "step": 3039
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.002866210234754,
+      "learning_rate": 1.7629900373249956e-05,
+      "loss": 0.5704,
+      "step": 3040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8760623142199797,
+      "learning_rate": 1.7628198451722247e-05,
+      "loss": 0.5525,
+      "step": 3041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9523505020216828,
+      "learning_rate": 1.7626496001567154e-05,
+      "loss": 0.5804,
+      "step": 3042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.965206293258815,
+      "learning_rate": 1.7624793022902648e-05,
+      "loss": 0.58,
+      "step": 3043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9022195843672679,
+      "learning_rate": 1.7623089515846752e-05,
+      "loss": 0.6024,
+      "step": 3044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9534846451535837,
+      "learning_rate": 1.7621385480517514e-05,
+      "loss": 0.5893,
+      "step": 3045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9509893948267736,
+      "learning_rate": 1.7619680917033023e-05,
+      "loss": 0.5674,
+      "step": 3046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9667981591904946,
+      "learning_rate": 1.7617975825511403e-05,
+      "loss": 0.6137,
+      "step": 3047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9457533097952052,
+      "learning_rate": 1.7616270206070814e-05,
+      "loss": 0.5797,
+      "step": 3048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8800567240746717,
+      "learning_rate": 1.7614564058829454e-05,
+      "loss": 0.5534,
+      "step": 3049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9306818061440629,
+      "learning_rate": 1.7612857383905565e-05,
+      "loss": 0.5664,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0394494750240004,
+      "learning_rate": 1.7611150181417406e-05,
+      "loss": 0.6443,
+      "step": 3051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9143008362220608,
+      "learning_rate": 1.7609442451483292e-05,
+      "loss": 0.5357,
+      "step": 3052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8459253561004924,
+      "learning_rate": 1.7607734194221565e-05,
+      "loss": 0.5381,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8342798033382928,
+      "learning_rate": 1.7606025409750608e-05,
+      "loss": 0.5503,
+      "step": 3054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9497336635018657,
+      "learning_rate": 1.760431609818884e-05,
+      "loss": 0.5588,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9574074351131134,
+      "learning_rate": 1.7602606259654704e-05,
+      "loss": 0.5993,
+      "step": 3056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2266619564048968,
+      "learning_rate": 1.7600895894266702e-05,
+      "loss": 0.5867,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9543613597242573,
+      "learning_rate": 1.7599185002143357e-05,
+      "loss": 0.6301,
+      "step": 3058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9433524901855496,
+      "learning_rate": 1.759747358340323e-05,
+      "loss": 0.5699,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9989762059557455,
+      "learning_rate": 1.7595761638164924e-05,
+      "loss": 0.5463,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9912262872442057,
+      "learning_rate": 1.7594049166547073e-05,
+      "loss": 0.5611,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.955654711601771,
+      "learning_rate": 1.7592336168668352e-05,
+      "loss": 0.6187,
+      "step": 3062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9407381255668952,
+      "learning_rate": 1.7590622644647466e-05,
+      "loss": 0.571,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8974550887945183,
+      "learning_rate": 1.7588908594603165e-05,
+      "loss": 0.4889,
+      "step": 3064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8719072751485014,
+      "learning_rate": 1.758719401865423e-05,
+      "loss": 0.5625,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8891693890171455,
+      "learning_rate": 1.758547891691948e-05,
+      "loss": 0.5671,
+      "step": 3066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9430776803567622,
+      "learning_rate": 1.7583763289517767e-05,
+      "loss": 0.6137,
+      "step": 3067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9764665774810648,
+      "learning_rate": 1.7582047136567987e-05,
+      "loss": 0.6012,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0107046063269252,
+      "learning_rate": 1.7580330458189066e-05,
+      "loss": 0.644,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9512873214249711,
+      "learning_rate": 1.757861325449997e-05,
+      "loss": 0.6462,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.003399707071991,
+      "learning_rate": 1.7576895525619693e-05,
+      "loss": 0.6177,
+      "step": 3071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9317498927261336,
+      "learning_rate": 1.757517727166728e-05,
+      "loss": 0.5163,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.884093708449396,
+      "learning_rate": 1.7573458492761802e-05,
+      "loss": 0.473,
+      "step": 3073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9546080085115011,
+      "learning_rate": 1.7571739189022365e-05,
+      "loss": 0.5957,
+      "step": 3074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0330919558923133,
+      "learning_rate": 1.7570019360568117e-05,
+      "loss": 0.5716,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9623272014338214,
+      "learning_rate": 1.7568299007518247e-05,
+      "loss": 0.5663,
+      "step": 3076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9888262495455644,
+      "learning_rate": 1.7566578129991966e-05,
+      "loss": 0.6225,
+      "step": 3077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.892592052293103,
+      "learning_rate": 1.756485672810853e-05,
+      "loss": 0.5885,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9551702736887312,
+      "learning_rate": 1.7563134801987235e-05,
+      "loss": 0.6147,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9954287831755133,
+      "learning_rate": 1.7561412351747406e-05,
+      "loss": 0.5876,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9158818809267135,
+      "learning_rate": 1.7559689377508413e-05,
+      "loss": 0.5339,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0174523252448435,
+      "learning_rate": 1.7557965879389644e-05,
+      "loss": 0.5493,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8947025622357997,
+      "learning_rate": 1.7556241857510547e-05,
+      "loss": 0.5544,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2270341670509768,
+      "learning_rate": 1.7554517311990592e-05,
+      "loss": 0.6078,
+      "step": 3084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0170400035892802,
+      "learning_rate": 1.7552792242949287e-05,
+      "loss": 0.6123,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9139996199480314,
+      "learning_rate": 1.755106665050618e-05,
+      "loss": 0.5805,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9622769309264265,
+      "learning_rate": 1.7549340534780852e-05,
+      "loss": 0.6121,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9809266179096972,
+      "learning_rate": 1.754761389589292e-05,
+      "loss": 0.6003,
+      "step": 3088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9361453357606391,
+      "learning_rate": 1.7545886733962044e-05,
+      "loss": 0.5877,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9521538902460409,
+      "learning_rate": 1.7544159049107902e-05,
+      "loss": 0.5804,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8638250061554852,
+      "learning_rate": 1.7542430841450236e-05,
+      "loss": 0.5138,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8724382049228834,
+      "learning_rate": 1.7540702111108803e-05,
+      "loss": 0.5124,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0254553888380893,
+      "learning_rate": 1.7538972858203397e-05,
+      "loss": 0.5292,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9330508544120452,
+      "learning_rate": 1.7537243082853866e-05,
+      "loss": 0.5537,
+      "step": 3094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9432771599677009,
+      "learning_rate": 1.753551278518007e-05,
+      "loss": 0.579,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8919899667127843,
+      "learning_rate": 1.7533781965301924e-05,
+      "loss": 0.4923,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.961077027530501,
+      "learning_rate": 1.753205062333937e-05,
+      "loss": 0.597,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9069964785933305,
+      "learning_rate": 1.753031875941239e-05,
+      "loss": 0.5856,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9659625727638037,
+      "learning_rate": 1.7528586373640997e-05,
+      "loss": 0.5876,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0598241751866497,
+      "learning_rate": 1.7526853466145248e-05,
+      "loss": 0.632,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0778620313469889,
+      "learning_rate": 1.7525120037045227e-05,
+      "loss": 0.5799,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8954798615992657,
+      "learning_rate": 1.7523386086461065e-05,
+      "loss": 0.6049,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.935250732427432,
+      "learning_rate": 1.7521651614512918e-05,
+      "loss": 0.6179,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0023476270439846,
+      "learning_rate": 1.751991662132099e-05,
+      "loss": 0.5978,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9002821778685863,
+      "learning_rate": 1.751818110700551e-05,
+      "loss": 0.5656,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8800925304899958,
+      "learning_rate": 1.751644507168674e-05,
+      "loss": 0.5569,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9330859593957906,
+      "learning_rate": 1.7514708515485002e-05,
+      "loss": 0.6456,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9532235450165991,
+      "learning_rate": 1.7512971438520626e-05,
+      "loss": 0.5808,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8998905570101068,
+      "learning_rate": 1.7511233840913994e-05,
+      "loss": 0.5211,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9128687245400994,
+      "learning_rate": 1.7509495722785518e-05,
+      "loss": 0.5159,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.075402357728703,
+      "learning_rate": 1.7507757084255652e-05,
+      "loss": 0.6144,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9429480058993237,
+      "learning_rate": 1.750601792544488e-05,
+      "loss": 0.5949,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9129715759556368,
+      "learning_rate": 1.750427824647372e-05,
+      "loss": 0.5553,
+      "step": 3113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9127366900187914,
+      "learning_rate": 1.7502538047462737e-05,
+      "loss": 0.5545,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9397256695327718,
+      "learning_rate": 1.750079732853252e-05,
+      "loss": 0.5552,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0024073480566253,
+      "learning_rate": 1.74990560898037e-05,
+      "loss": 0.6514,
+      "step": 3116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9620678706748186,
+      "learning_rate": 1.7497314331396946e-05,
+      "loss": 0.616,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.980921040306892,
+      "learning_rate": 1.7495572053432962e-05,
+      "loss": 0.6389,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.93016075302683,
+      "learning_rate": 1.749382925603248e-05,
+      "loss": 0.5954,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9513139305328434,
+      "learning_rate": 1.749208593931628e-05,
+      "loss": 0.5899,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9098228343772,
+      "learning_rate": 1.7490342103405168e-05,
+      "loss": 0.6347,
+      "step": 3121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9982505813028052,
+      "learning_rate": 1.748859774841999e-05,
+      "loss": 0.6127,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9229136534442328,
+      "learning_rate": 1.748685287448163e-05,
+      "loss": 0.6267,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8754740252477136,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.6227,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9122049883525993,
+      "learning_rate": 1.748336157022908e-05,
+      "loss": 0.5726,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8813794257915903,
+      "learning_rate": 1.7481615140156837e-05,
+      "loss": 0.5837,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0392588455372822,
+      "learning_rate": 1.747986819161529e-05,
+      "loss": 0.5378,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.063738607824746,
+      "learning_rate": 1.747812072472552e-05,
+      "loss": 0.6057,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0799665881792668,
+      "learning_rate": 1.7476372739608615e-05,
+      "loss": 0.6201,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.922842839007906,
+      "learning_rate": 1.7474624236385706e-05,
+      "loss": 0.5879,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9427134960414176,
+      "learning_rate": 1.747287521517797e-05,
+      "loss": 0.5887,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9182269254202275,
+      "learning_rate": 1.7471125676106613e-05,
+      "loss": 0.6023,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.992985791336242,
+      "learning_rate": 1.7469375619292873e-05,
+      "loss": 0.5516,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0000627224938892,
+      "learning_rate": 1.7467625044858025e-05,
+      "loss": 0.5806,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0464342928375443,
+      "learning_rate": 1.7465873952923386e-05,
+      "loss": 0.6352,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8489039252817476,
+      "learning_rate": 1.7464122343610307e-05,
+      "loss": 0.5423,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8344056548884692,
+      "learning_rate": 1.7462370217040167e-05,
+      "loss": 0.6058,
+      "step": 3137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9413618313258731,
+      "learning_rate": 1.7460617573334393e-05,
+      "loss": 0.588,
+      "step": 3138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0008486902893026,
+      "learning_rate": 1.7458864412614436e-05,
+      "loss": 0.6089,
+      "step": 3139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9376106409264786,
+      "learning_rate": 1.745711073500179e-05,
+      "loss": 0.571,
+      "step": 3140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.887158908415457,
+      "learning_rate": 1.7455356540617988e-05,
+      "loss": 0.5797,
+      "step": 3141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9221618609988214,
+      "learning_rate": 1.745360182958459e-05,
+      "loss": 0.5654,
+      "step": 3142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9051428050775634,
+      "learning_rate": 1.7451846602023196e-05,
+      "loss": 0.5332,
+      "step": 3143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.972614082330035,
+      "learning_rate": 1.745009085805544e-05,
+      "loss": 0.5517,
+      "step": 3144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0267964133100653,
+      "learning_rate": 1.7448334597803e-05,
+      "loss": 0.5996,
+      "step": 3145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8669897880071976,
+      "learning_rate": 1.7446577821387575e-05,
+      "loss": 0.5437,
+      "step": 3146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9444554670732166,
+      "learning_rate": 1.7444820528930914e-05,
+      "loss": 0.5424,
+      "step": 3147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1496699362879719,
+      "learning_rate": 1.7443062720554796e-05,
+      "loss": 0.5784,
+      "step": 3148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0157973115543528,
+      "learning_rate": 1.744130439638103e-05,
+      "loss": 0.634,
+      "step": 3149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8366603399421715,
+      "learning_rate": 1.7439545556531473e-05,
+      "loss": 0.5062,
+      "step": 3150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.97088446679413,
+      "learning_rate": 1.7437786201128003e-05,
+      "loss": 0.5379,
+      "step": 3151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9939309659717451,
+      "learning_rate": 1.743602633029255e-05,
+      "loss": 0.5846,
+      "step": 3152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8990135518929009,
+      "learning_rate": 1.7434265944147068e-05,
+      "loss": 0.5164,
+      "step": 3153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.019817941557951,
+      "learning_rate": 1.743250504281355e-05,
+      "loss": 0.6309,
+      "step": 3154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8928122311880572,
+      "learning_rate": 1.7430743626414024e-05,
+      "loss": 0.4856,
+      "step": 3155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9782714945133913,
+      "learning_rate": 1.7428981695070558e-05,
+      "loss": 0.6257,
+      "step": 3156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8879693848442041,
+      "learning_rate": 1.7427219248905246e-05,
+      "loss": 0.5216,
+      "step": 3157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.089049966696691,
+      "learning_rate": 1.7425456288040236e-05,
+      "loss": 0.6037,
+      "step": 3158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9056987893721266,
+      "learning_rate": 1.7423692812597682e-05,
+      "loss": 0.5546,
+      "step": 3159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0574500523825905,
+      "learning_rate": 1.7421928822699805e-05,
+      "loss": 0.5797,
+      "step": 3160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0050064235457727,
+      "learning_rate": 1.7420164318468845e-05,
+      "loss": 0.6173,
+      "step": 3161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9148162925439718,
+      "learning_rate": 1.741839930002708e-05,
+      "loss": 0.5662,
+      "step": 3162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8762180869341742,
+      "learning_rate": 1.741663376749682e-05,
+      "loss": 0.5885,
+      "step": 3163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8594642521189727,
+      "learning_rate": 1.7414867721000423e-05,
+      "loss": 0.5508,
+      "step": 3164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9794014670832364,
+      "learning_rate": 1.7413101160660267e-05,
+      "loss": 0.5757,
+      "step": 3165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9907022513665806,
+      "learning_rate": 1.7411334086598775e-05,
+      "loss": 0.5629,
+      "step": 3166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8991676755470959,
+      "learning_rate": 1.7409566498938405e-05,
+      "loss": 0.5953,
+      "step": 3167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.091552960265873,
+      "learning_rate": 1.740779839780165e-05,
+      "loss": 0.5151,
+      "step": 3168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0550607969869947,
+      "learning_rate": 1.7406029783311036e-05,
+      "loss": 0.6001,
+      "step": 3169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.023053375591623,
+      "learning_rate": 1.7404260655589128e-05,
+      "loss": 0.6253,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0158274844468704,
+      "learning_rate": 1.7402491014758526e-05,
+      "loss": 0.565,
+      "step": 3171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8856276226898813,
+      "learning_rate": 1.740072086094186e-05,
+      "loss": 0.5784,
+      "step": 3172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9297964339077929,
+      "learning_rate": 1.7398950194261803e-05,
+      "loss": 0.4888,
+      "step": 3173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0394334239483127,
+      "learning_rate": 1.739717901484106e-05,
+      "loss": 0.6169,
+      "step": 3174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.000153477231492,
+      "learning_rate": 1.7395407322802374e-05,
+      "loss": 0.581,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9310772778163637,
+      "learning_rate": 1.739363511826852e-05,
+      "loss": 0.5245,
+      "step": 3176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9003978982215697,
+      "learning_rate": 1.739186240136231e-05,
+      "loss": 0.5652,
+      "step": 3177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0088358143669862,
+      "learning_rate": 1.7390089172206594e-05,
+      "loss": 0.6085,
+      "step": 3178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9211698202843335,
+      "learning_rate": 1.7388315430924253e-05,
+      "loss": 0.5584,
+      "step": 3179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8748172476650534,
+      "learning_rate": 1.73865411776382e-05,
+      "loss": 0.5751,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8417660121346013,
+      "learning_rate": 1.7384766412471405e-05,
+      "loss": 0.5499,
+      "step": 3181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9408868417098869,
+      "learning_rate": 1.7382991135546842e-05,
+      "loss": 0.6425,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.940025872410451,
+      "learning_rate": 1.7381215346987538e-05,
+      "loss": 0.5604,
+      "step": 3183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8701094523091236,
+      "learning_rate": 1.7379439046916564e-05,
+      "loss": 0.5617,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9589378666465185,
+      "learning_rate": 1.737766223545701e-05,
+      "loss": 0.5457,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8717363773584366,
+      "learning_rate": 1.7375884912732004e-05,
+      "loss": 0.5277,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9283899269806596,
+      "learning_rate": 1.7374107078864716e-05,
+      "loss": 0.6118,
+      "step": 3187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9544424219775307,
+      "learning_rate": 1.7372328733978348e-05,
+      "loss": 0.5704,
+      "step": 3188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9255212308304944,
+      "learning_rate": 1.737054987819614e-05,
+      "loss": 0.6063,
+      "step": 3189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9358163965712596,
+      "learning_rate": 1.7368770511641365e-05,
+      "loss": 0.5423,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9010266984747565,
+      "learning_rate": 1.7366990634437328e-05,
+      "loss": 0.6014,
+      "step": 3191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9571620756129031,
+      "learning_rate": 1.736521024670737e-05,
+      "loss": 0.5905,
+      "step": 3192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9217502531680845,
+      "learning_rate": 1.736342934857488e-05,
+      "loss": 0.6337,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8466450609798067,
+      "learning_rate": 1.7361647940163266e-05,
+      "loss": 0.5326,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9985771457839538,
+      "learning_rate": 1.735986602159598e-05,
+      "loss": 0.6586,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9246511819018729,
+      "learning_rate": 1.7358083592996507e-05,
+      "loss": 0.5962,
+      "step": 3196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9695534053036614,
+      "learning_rate": 1.7356300654488367e-05,
+      "loss": 0.6388,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8667598153295203,
+      "learning_rate": 1.7354517206195115e-05,
+      "loss": 0.561,
+      "step": 3198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8842929810545559,
+      "learning_rate": 1.7352733248240347e-05,
+      "loss": 0.543,
+      "step": 3199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.893497524895569,
+      "learning_rate": 1.7350948780747684e-05,
+      "loss": 0.5941,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8583532551872464,
+      "learning_rate": 1.734916380384079e-05,
+      "loss": 0.512,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0429301416865169,
+      "learning_rate": 1.7347378317643368e-05,
+      "loss": 0.6715,
+      "step": 3202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9002956701588745,
+      "learning_rate": 1.7345592322279143e-05,
+      "loss": 0.5906,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0766543533538397,
+      "learning_rate": 1.7343805817871885e-05,
+      "loss": 0.614,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9758413415717146,
+      "learning_rate": 1.73420188045454e-05,
+      "loss": 0.6033,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0511316691987547,
+      "learning_rate": 1.734023128242352e-05,
+      "loss": 0.5785,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8933442945857915,
+      "learning_rate": 1.7338443251630125e-05,
+      "loss": 0.5391,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8938768216765625,
+      "learning_rate": 1.7336654712289125e-05,
+      "loss": 0.5729,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9178495968780668,
+      "learning_rate": 1.733486566452446e-05,
+      "loss": 0.6117,
+      "step": 3209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0081580578635072,
+      "learning_rate": 1.733307610846011e-05,
+      "loss": 0.549,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9144248732928212,
+      "learning_rate": 1.7331286044220086e-05,
+      "loss": 0.5454,
+      "step": 3211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0079526594157142,
+      "learning_rate": 1.7329495471928446e-05,
+      "loss": 0.5632,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9673535210182386,
+      "learning_rate": 1.732770439170927e-05,
+      "loss": 0.5216,
+      "step": 3213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8366252919513592,
+      "learning_rate": 1.732591280368668e-05,
+      "loss": 0.5358,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8997963514421738,
+      "learning_rate": 1.732412070798483e-05,
+      "loss": 0.544,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0674135355568224,
+      "learning_rate": 1.732232810472791e-05,
+      "loss": 0.6194,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9130503380839831,
+      "learning_rate": 1.7320534994040148e-05,
+      "loss": 0.5702,
+      "step": 3217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9125939393730079,
+      "learning_rate": 1.7318741376045806e-05,
+      "loss": 0.5232,
+      "step": 3218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9152888605297167,
+      "learning_rate": 1.731694725086918e-05,
+      "loss": 0.5589,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.918053811287748,
+      "learning_rate": 1.7315152618634594e-05,
+      "loss": 0.5515,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9517138283797776,
+      "learning_rate": 1.731335747946642e-05,
+      "loss": 0.491,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9695937631844701,
+      "learning_rate": 1.7311561833489065e-05,
+      "loss": 0.6387,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9344378625169966,
+      "learning_rate": 1.730976568082696e-05,
+      "loss": 0.5961,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9918672455564193,
+      "learning_rate": 1.7307969021604574e-05,
+      "loss": 0.5898,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9510075197303961,
+      "learning_rate": 1.730617185594642e-05,
+      "loss": 0.6031,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9366114096543621,
+      "learning_rate": 1.7304374183977032e-05,
+      "loss": 0.6217,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0382255870102077,
+      "learning_rate": 1.7302576005820997e-05,
+      "loss": 0.6121,
+      "step": 3227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8940002944032726,
+      "learning_rate": 1.730077732160292e-05,
+      "loss": 0.5346,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9389001006992637,
+      "learning_rate": 1.729897813144745e-05,
+      "loss": 0.6312,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.027925431969123,
+      "learning_rate": 1.729717843547927e-05,
+      "loss": 0.6487,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.973280328221155,
+      "learning_rate": 1.7295378233823096e-05,
+      "loss": 0.5505,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8950161903045436,
+      "learning_rate": 1.7293577526603684e-05,
+      "loss": 0.5313,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0185304119690635,
+      "learning_rate": 1.7291776313945817e-05,
+      "loss": 0.5965,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9786576159121729,
+      "learning_rate": 1.728997459597432e-05,
+      "loss": 0.5804,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0669721345581529,
+      "learning_rate": 1.728817237281405e-05,
+      "loss": 0.5765,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9704256322697749,
+      "learning_rate": 1.7286369644589897e-05,
+      "loss": 0.631,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9319412904056394,
+      "learning_rate": 1.728456641142679e-05,
+      "loss": 0.5408,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0092120792907255,
+      "learning_rate": 1.7282762673449695e-05,
+      "loss": 0.6245,
+      "step": 3238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9394081643734624,
+      "learning_rate": 1.7280958430783608e-05,
+      "loss": 0.6109,
+      "step": 3239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9187040965972737,
+      "learning_rate": 1.7279153683553556e-05,
+      "loss": 0.574,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.890777039494145,
+      "learning_rate": 1.7277348431884613e-05,
+      "loss": 0.5282,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9816871876859573,
+      "learning_rate": 1.7275542675901876e-05,
+      "loss": 0.5562,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0160416140142166,
+      "learning_rate": 1.7273736415730488e-05,
+      "loss": 0.6288,
+      "step": 3243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.976610433292306,
+      "learning_rate": 1.7271929651495617e-05,
+      "loss": 0.5985,
+      "step": 3244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9581355875399314,
+      "learning_rate": 1.7270122383322473e-05,
+      "loss": 0.6334,
+      "step": 3245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1556542224283102,
+      "learning_rate": 1.7268314611336296e-05,
+      "loss": 0.5319,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0106194182617843,
+      "learning_rate": 1.726650633566236e-05,
+      "loss": 0.6098,
+      "step": 3247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9072489025762995,
+      "learning_rate": 1.726469755642598e-05,
+      "loss": 0.5464,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8495528879265195,
+      "learning_rate": 1.7262888273752505e-05,
+      "loss": 0.5391,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9946732239643433,
+      "learning_rate": 1.7261078487767317e-05,
+      "loss": 0.5863,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9445316998047976,
+      "learning_rate": 1.7259268198595828e-05,
+      "loss": 0.5619,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9169313514431681,
+      "learning_rate": 1.7257457406363495e-05,
+      "loss": 0.5631,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9302146760899365,
+      "learning_rate": 1.72556461111958e-05,
+      "loss": 0.5708,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9320385142628118,
+      "learning_rate": 1.725383431321826e-05,
+      "loss": 0.6121,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9053121367203344,
+      "learning_rate": 1.725202201255644e-05,
+      "loss": 0.5386,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9140031954486985,
+      "learning_rate": 1.725020920933593e-05,
+      "loss": 0.6397,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0485591789305269,
+      "learning_rate": 1.7248395903682347e-05,
+      "loss": 0.6717,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9622445974231173,
+      "learning_rate": 1.724658209572136e-05,
+      "loss": 0.5676,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9008820906676879,
+      "learning_rate": 1.724476778557866e-05,
+      "loss": 0.5995,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8493129216065038,
+      "learning_rate": 1.7242952973379983e-05,
+      "loss": 0.536,
+      "step": 3260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0225594519850092,
+      "learning_rate": 1.7241137659251087e-05,
+      "loss": 0.5215,
+      "step": 3261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9949755006324897,
+      "learning_rate": 1.723932184331777e-05,
+      "loss": 0.5143,
+      "step": 3262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0648617924766293,
+      "learning_rate": 1.7237505525705875e-05,
+      "loss": 0.6225,
+      "step": 3263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9205169832526819,
+      "learning_rate": 1.7235688706541266e-05,
+      "loss": 0.5073,
+      "step": 3264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0317464947048756,
+      "learning_rate": 1.723387138594985e-05,
+      "loss": 0.586,
+      "step": 3265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9378484870577685,
+      "learning_rate": 1.723205356405756e-05,
+      "loss": 0.5489,
+      "step": 3266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9899383610437812,
+      "learning_rate": 1.7230235240990373e-05,
+      "loss": 0.565,
+      "step": 3267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0498755184573376,
+      "learning_rate": 1.72284164168743e-05,
+      "loss": 0.5998,
+      "step": 3268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9073391433511635,
+      "learning_rate": 1.7226597091835377e-05,
+      "loss": 0.5493,
+      "step": 3269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8451787420231008,
+      "learning_rate": 1.7224777265999688e-05,
+      "loss": 0.5525,
+      "step": 3270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9488853496030325,
+      "learning_rate": 1.722295693949334e-05,
+      "loss": 0.5827,
+      "step": 3271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0495687319964833,
+      "learning_rate": 1.7221136112442487e-05,
+      "loss": 0.6238,
+      "step": 3272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.94853336956771,
+      "learning_rate": 1.7219314784973304e-05,
+      "loss": 0.523,
+      "step": 3273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9628064262893254,
+      "learning_rate": 1.721749295721201e-05,
+      "loss": 0.5904,
+      "step": 3274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9345929252013101,
+      "learning_rate": 1.7215670629284856e-05,
+      "loss": 0.5494,
+      "step": 3275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8830597865479197,
+      "learning_rate": 1.7213847801318128e-05,
+      "loss": 0.5605,
+      "step": 3276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9572622089997017,
+      "learning_rate": 1.7212024473438145e-05,
+      "loss": 0.6316,
+      "step": 3277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9562194162392834,
+      "learning_rate": 1.7210200645771268e-05,
+      "loss": 0.5326,
+      "step": 3278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0553393174265637,
+      "learning_rate": 1.7208376318443877e-05,
+      "loss": 0.634,
+      "step": 3279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9970413860899692,
+      "learning_rate": 1.72065514915824e-05,
+      "loss": 0.6207,
+      "step": 3280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9318499998745525,
+      "learning_rate": 1.72047261653133e-05,
+      "loss": 0.5454,
+      "step": 3281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9223364403973837,
+      "learning_rate": 1.7202900339763066e-05,
+      "loss": 0.6575,
+      "step": 3282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.929898055654162,
+      "learning_rate": 1.7201074015058226e-05,
+      "loss": 0.5548,
+      "step": 3283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9324984663522081,
+      "learning_rate": 1.7199247191325347e-05,
+      "loss": 0.6458,
+      "step": 3284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9534813809660702,
+      "learning_rate": 1.7197419868691022e-05,
+      "loss": 0.5562,
+      "step": 3285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.953733613695882,
+      "learning_rate": 1.719559204728188e-05,
+      "loss": 0.6061,
+      "step": 3286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9526278381395721,
+      "learning_rate": 1.7193763727224596e-05,
+      "loss": 0.6133,
+      "step": 3287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9974347688291391,
+      "learning_rate": 1.719193490864587e-05,
+      "loss": 0.611,
+      "step": 3288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9452782503892461,
+      "learning_rate": 1.719010559167243e-05,
+      "loss": 0.5824,
+      "step": 3289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8358480245034036,
+      "learning_rate": 1.7188275776431048e-05,
+      "loss": 0.5364,
+      "step": 3290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.918407324135863,
+      "learning_rate": 1.7186445463048533e-05,
+      "loss": 0.5966,
+      "step": 3291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9047942682262397,
+      "learning_rate": 1.7184614651651723e-05,
+      "loss": 0.5637,
+      "step": 3292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9728046891966552,
+      "learning_rate": 1.718278334236749e-05,
+      "loss": 0.6377,
+      "step": 3293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9405103120350403,
+      "learning_rate": 1.7180951535322742e-05,
+      "loss": 0.5708,
+      "step": 3294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8591298849891674,
+      "learning_rate": 1.717911923064442e-05,
+      "loss": 0.5249,
+      "step": 3295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8752323427602121,
+      "learning_rate": 1.7177286428459505e-05,
+      "loss": 0.5671,
+      "step": 3296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9414564811464446,
+      "learning_rate": 1.717545312889501e-05,
+      "loss": 0.5671,
+      "step": 3297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.02907718626476,
+      "learning_rate": 1.7173619332077972e-05,
+      "loss": 0.6245,
+      "step": 3298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.883595641347089,
+      "learning_rate": 1.717178503813548e-05,
+      "loss": 0.58,
+      "step": 3299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9260784685712767,
+      "learning_rate": 1.7169950247194646e-05,
+      "loss": 0.5683,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9018370832687636,
+      "learning_rate": 1.716811495938262e-05,
+      "loss": 0.6049,
+      "step": 3301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9249549260329984,
+      "learning_rate": 1.716627917482658e-05,
+      "loss": 0.5928,
+      "step": 3302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.950760435439383,
+      "learning_rate": 1.716444289365376e-05,
+      "loss": 0.592,
+      "step": 3303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0057267824279712,
+      "learning_rate": 1.7162606115991395e-05,
+      "loss": 0.6063,
+      "step": 3304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9109873091350366,
+      "learning_rate": 1.7160768841966785e-05,
+      "loss": 0.5169,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9393355629622872,
+      "learning_rate": 1.7158931071707242e-05,
+      "loss": 0.6203,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9332032789248998,
+      "learning_rate": 1.7157092805340126e-05,
+      "loss": 0.5535,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9745364105146304,
+      "learning_rate": 1.7155254042992827e-05,
+      "loss": 0.554,
+      "step": 3308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8993080723884361,
+      "learning_rate": 1.715341478479277e-05,
+      "loss": 0.5725,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9404230807613553,
+      "learning_rate": 1.715157503086741e-05,
+      "loss": 0.5964,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0544615187533903,
+      "learning_rate": 1.7149734781344247e-05,
+      "loss": 0.6241,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9740583185354336,
+      "learning_rate": 1.7147894036350804e-05,
+      "loss": 0.559,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.906313386144417,
+      "learning_rate": 1.7146052796014646e-05,
+      "loss": 0.5712,
+      "step": 3313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.922204166116946,
+      "learning_rate": 1.7144211060463368e-05,
+      "loss": 0.5948,
+      "step": 3314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9019297900829841,
+      "learning_rate": 1.7142368829824602e-05,
+      "loss": 0.5747,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8758645949867254,
+      "learning_rate": 1.714052610422601e-05,
+      "loss": 0.575,
+      "step": 3316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9339216019045296,
+      "learning_rate": 1.7138682883795292e-05,
+      "loss": 0.5893,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.892853688298582,
+      "learning_rate": 1.713683916866018e-05,
+      "loss": 0.5385,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8869662452289221,
+      "learning_rate": 1.7134994958948444e-05,
+      "loss": 0.5819,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9363151281743876,
+      "learning_rate": 1.713315025478789e-05,
+      "loss": 0.5378,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9454884558940029,
+      "learning_rate": 1.713130505630635e-05,
+      "loss": 0.5796,
+      "step": 3321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9600187437555181,
+      "learning_rate": 1.7129459363631692e-05,
+      "loss": 0.5741,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9129181561719604,
+      "learning_rate": 1.7127613176891824e-05,
+      "loss": 0.5531,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8436607312653893,
+      "learning_rate": 1.7125766496214687e-05,
+      "loss": 0.5581,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9634247002821723,
+      "learning_rate": 1.712391932172825e-05,
+      "loss": 0.5728,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9110565883221894,
+      "learning_rate": 1.712207165356053e-05,
+      "loss": 0.6041,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8991361572067563,
+      "learning_rate": 1.7120223491839553e-05,
+      "loss": 0.5818,
+      "step": 3327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9050015040122568,
+      "learning_rate": 1.7118374836693407e-05,
+      "loss": 0.6224,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8981409776838637,
+      "learning_rate": 1.71165256882502e-05,
+      "loss": 0.616,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9038761710482833,
+      "learning_rate": 1.7114676046638076e-05,
+      "loss": 0.577,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9331379208091041,
+      "learning_rate": 1.7112825911985207e-05,
+      "loss": 0.5678,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9866541584168005,
+      "learning_rate": 1.7110975284419814e-05,
+      "loss": 0.6266,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.938458776562073,
+      "learning_rate": 1.7109124164070144e-05,
+      "loss": 0.5473,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9282242071375044,
+      "learning_rate": 1.710727255106447e-05,
+      "loss": 0.5782,
+      "step": 3334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9217922793401815,
+      "learning_rate": 1.710542044553112e-05,
+      "loss": 0.5536,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8435229885105086,
+      "learning_rate": 1.710356784759843e-05,
+      "loss": 0.5292,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8888554609533063,
+      "learning_rate": 1.7101714757394792e-05,
+      "loss": 0.6026,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9747928015600499,
+      "learning_rate": 1.7099861175048617e-05,
+      "loss": 0.6023,
+      "step": 3338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0846321842053752,
+      "learning_rate": 1.7098007100688362e-05,
+      "loss": 0.5917,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9185007365976529,
+      "learning_rate": 1.7096152534442515e-05,
+      "loss": 0.5267,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8376651511974664,
+      "learning_rate": 1.7094297476439585e-05,
+      "loss": 0.5167,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8460925887281229,
+      "learning_rate": 1.7092441926808138e-05,
+      "loss": 0.5659,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9936106466309771,
+      "learning_rate": 1.7090585885676753e-05,
+      "loss": 0.5279,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.064099435320851,
+      "learning_rate": 1.7088729353174054e-05,
+      "loss": 0.5633,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9457408843760768,
+      "learning_rate": 1.7086872329428702e-05,
+      "loss": 0.6306,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8350392579479472,
+      "learning_rate": 1.708501481456938e-05,
+      "loss": 0.5424,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9524852783334381,
+      "learning_rate": 1.7083156808724817e-05,
+      "loss": 0.6225,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9605946312330363,
+      "learning_rate": 1.7081298312023773e-05,
+      "loss": 0.5867,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0630205296787159,
+      "learning_rate": 1.7079439324595038e-05,
+      "loss": 0.5759,
+      "step": 3349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8757051239915971,
+      "learning_rate": 1.7077579846567435e-05,
+      "loss": 0.5896,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.911951669277883,
+      "learning_rate": 1.7075719878069822e-05,
+      "loss": 0.5148,
+      "step": 3351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9017122704150898,
+      "learning_rate": 1.7073859419231104e-05,
+      "loss": 0.5533,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.130210862193503,
+      "learning_rate": 1.70719984701802e-05,
+      "loss": 0.6358,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.015208620424021,
+      "learning_rate": 1.7070137031046074e-05,
+      "loss": 0.6288,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.98786674028703,
+      "learning_rate": 1.7068275101957724e-05,
+      "loss": 0.5977,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9258362716954789,
+      "learning_rate": 1.7066412683044176e-05,
+      "loss": 0.5598,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9371702688369251,
+      "learning_rate": 1.7064549774434502e-05,
+      "loss": 0.5474,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9476739249381467,
+      "learning_rate": 1.7062686376257792e-05,
+      "loss": 0.6134,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8500421461497814,
+      "learning_rate": 1.706082248864318e-05,
+      "loss": 0.538,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8320273702343286,
+      "learning_rate": 1.7058958111719836e-05,
+      "loss": 0.5137,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9597941408747482,
+      "learning_rate": 1.7057093245616953e-05,
+      "loss": 0.6073,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9336154644932648,
+      "learning_rate": 1.705522789046377e-05,
+      "loss": 0.5306,
+      "step": 3362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9007035249997226,
+      "learning_rate": 1.7053362046389553e-05,
+      "loss": 0.5615,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8606210782971025,
+      "learning_rate": 1.7051495713523598e-05,
+      "loss": 0.5333,
+      "step": 3364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.037431116748183,
+      "learning_rate": 1.7049628891995245e-05,
+      "loss": 0.6167,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8689014841123899,
+      "learning_rate": 1.7047761581933867e-05,
+      "loss": 0.5588,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8812212259999989,
+      "learning_rate": 1.704589378346886e-05,
+      "loss": 0.5603,
+      "step": 3367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9829983554137566,
+      "learning_rate": 1.7044025496729665e-05,
+      "loss": 0.5547,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9032978495659435,
+      "learning_rate": 1.7042156721845754e-05,
+      "loss": 0.497,
+      "step": 3369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8688514896023216,
+      "learning_rate": 1.7040287458946623e-05,
+      "loss": 0.5186,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8515062816030871,
+      "learning_rate": 1.7038417708161817e-05,
+      "loss": 0.58,
+      "step": 3371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.879654908280097,
+      "learning_rate": 1.7036547469620908e-05,
+      "loss": 0.5517,
+      "step": 3372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9111200905537216,
+      "learning_rate": 1.70346767434535e-05,
+      "loss": 0.5802,
+      "step": 3373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9598633626437251,
+      "learning_rate": 1.7032805529789233e-05,
+      "loss": 0.5801,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9596570514252659,
+      "learning_rate": 1.7030933828757785e-05,
+      "loss": 0.5795,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.961433589710115,
+      "learning_rate": 1.7029061640488855e-05,
+      "loss": 0.5757,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9407600851727718,
+      "learning_rate": 1.702718896511219e-05,
+      "loss": 0.5584,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8088236897988393,
+      "learning_rate": 1.7025315802757558e-05,
+      "loss": 0.5237,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9058206622747288,
+      "learning_rate": 1.7023442153554776e-05,
+      "loss": 0.554,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9935709406567976,
+      "learning_rate": 1.7021568017633683e-05,
+      "loss": 0.6106,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9189552590843816,
+      "learning_rate": 1.7019693395124153e-05,
+      "loss": 0.5729,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9358733733184447,
+      "learning_rate": 1.70178182861561e-05,
+      "loss": 0.5921,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8672815001532652,
+      "learning_rate": 1.701594269085946e-05,
+      "loss": 0.6447,
+      "step": 3383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9180923664460106,
+      "learning_rate": 1.701406660936422e-05,
+      "loss": 0.6046,
+      "step": 3384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0033780298482418,
+      "learning_rate": 1.7012190041800384e-05,
+      "loss": 0.5508,
+      "step": 3385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9146946932023774,
+      "learning_rate": 1.7010312988297993e-05,
+      "loss": 0.589,
+      "step": 3386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9725045335715133,
+      "learning_rate": 1.7008435448987134e-05,
+      "loss": 0.6529,
+      "step": 3387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9226018976731254,
+      "learning_rate": 1.7006557423997917e-05,
+      "loss": 0.5691,
+      "step": 3388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9765084315182954,
+      "learning_rate": 1.7004678913460483e-05,
+      "loss": 0.6344,
+      "step": 3389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.072975921776912,
+      "learning_rate": 1.7002799917505014e-05,
+      "loss": 0.5671,
+      "step": 3390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9540991904709908,
+      "learning_rate": 1.700092043626172e-05,
+      "loss": 0.5818,
+      "step": 3391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9795979551718766,
+      "learning_rate": 1.6999040469860852e-05,
+      "loss": 0.5807,
+      "step": 3392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8628947844434518,
+      "learning_rate": 1.6997160018432688e-05,
+      "loss": 0.5442,
+      "step": 3393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8811809802727653,
+      "learning_rate": 1.6995279082107537e-05,
+      "loss": 0.5267,
+      "step": 3394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8814467993180549,
+      "learning_rate": 1.6993397661015754e-05,
+      "loss": 0.5939,
+      "step": 3395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.892270166357497,
+      "learning_rate": 1.6991515755287715e-05,
+      "loss": 0.5586,
+      "step": 3396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9363609939411836,
+      "learning_rate": 1.6989633365053837e-05,
+      "loss": 0.5826,
+      "step": 3397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9460263488365313,
+      "learning_rate": 1.6987750490444565e-05,
+      "loss": 0.5738,
+      "step": 3398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8830829644992519,
+      "learning_rate": 1.6985867131590383e-05,
+      "loss": 0.6229,
+      "step": 3399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0074862329859187,
+      "learning_rate": 1.6983983288621807e-05,
+      "loss": 0.6148,
+      "step": 3400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8830992932928085,
+      "learning_rate": 1.6982098961669383e-05,
+      "loss": 0.5447,
+      "step": 3401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9669231327429048,
+      "learning_rate": 1.6980214150863692e-05,
+      "loss": 0.5977,
+      "step": 3402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9375796919681045,
+      "learning_rate": 1.6978328856335354e-05,
+      "loss": 0.548,
+      "step": 3403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8929206405304951,
+      "learning_rate": 1.6976443078215015e-05,
+      "loss": 0.5252,
+      "step": 3404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8971001962743499,
+      "learning_rate": 1.697455681663336e-05,
+      "loss": 0.5643,
+      "step": 3405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8973473047538455,
+      "learning_rate": 1.69726700717211e-05,
+      "loss": 0.5758,
+      "step": 3406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0046297232651533,
+      "learning_rate": 1.6970782843608994e-05,
+      "loss": 0.5857,
+      "step": 3407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.991860327566595,
+      "learning_rate": 1.6968895132427817e-05,
+      "loss": 0.5205,
+      "step": 3408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9344982001418214,
+      "learning_rate": 1.696700693830839e-05,
+      "loss": 0.5794,
+      "step": 3409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8598062778766361,
+      "learning_rate": 1.6965118261381557e-05,
+      "loss": 0.5485,
+      "step": 3410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9926075241949736,
+      "learning_rate": 1.6963229101778215e-05,
+      "loss": 0.5576,
+      "step": 3411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8706724751709342,
+      "learning_rate": 1.696133945962927e-05,
+      "loss": 0.5495,
+      "step": 3412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.950973621277761,
+      "learning_rate": 1.695944933506567e-05,
+      "loss": 0.5053,
+      "step": 3413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0268360922092827,
+      "learning_rate": 1.695755872821841e-05,
+      "loss": 0.666,
+      "step": 3414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9053834277128515,
+      "learning_rate": 1.6955667639218497e-05,
+      "loss": 0.5685,
+      "step": 3415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.082944483643275,
+      "learning_rate": 1.695377606819699e-05,
+      "loss": 0.5695,
+      "step": 3416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8542060396820743,
+      "learning_rate": 1.6951884015284966e-05,
+      "loss": 0.5983,
+      "step": 3417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9432598922230787,
+      "learning_rate": 1.694999148061355e-05,
+      "loss": 0.5653,
+      "step": 3418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9270649827866421,
+      "learning_rate": 1.6948098464313886e-05,
+      "loss": 0.5379,
+      "step": 3419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.075767072500345,
+      "learning_rate": 1.6946204966517165e-05,
+      "loss": 0.5268,
+      "step": 3420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8705734453427241,
+      "learning_rate": 1.6944310987354597e-05,
+      "loss": 0.5886,
+      "step": 3421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9925730390794844,
+      "learning_rate": 1.6942416526957438e-05,
+      "loss": 0.6483,
+      "step": 3422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.907248303985114,
+      "learning_rate": 1.694052158545697e-05,
+      "loss": 0.5032,
+      "step": 3423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9043231411925035,
+      "learning_rate": 1.6938626162984516e-05,
+      "loss": 0.5614,
+      "step": 3424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9461336445647923,
+      "learning_rate": 1.6936730259671423e-05,
+      "loss": 0.5063,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9913856016305469,
+      "learning_rate": 1.6934833875649074e-05,
+      "loss": 0.5747,
+      "step": 3426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9253569554243637,
+      "learning_rate": 1.693293701104889e-05,
+      "loss": 0.5865,
+      "step": 3427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8644437828373465,
+      "learning_rate": 1.693103966600232e-05,
+      "loss": 0.5782,
+      "step": 3428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.925353380941438,
+      "learning_rate": 1.692914184064085e-05,
+      "loss": 0.5667,
+      "step": 3429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9749291183376588,
+      "learning_rate": 1.6927243535095995e-05,
+      "loss": 0.5387,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8819179929608888,
+      "learning_rate": 1.6925344749499308e-05,
+      "loss": 0.5265,
+      "step": 3431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8728789653879663,
+      "learning_rate": 1.6923445483982376e-05,
+      "loss": 0.5412,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9042520287309848,
+      "learning_rate": 1.6921545738676807e-05,
+      "loss": 0.5677,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9097494183903341,
+      "learning_rate": 1.691964551371426e-05,
+      "loss": 0.5969,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9853158768766778,
+      "learning_rate": 1.691774480922642e-05,
+      "loss": 0.5842,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8166520864997642,
+      "learning_rate": 1.6915843625344997e-05,
+      "loss": 0.5148,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8609842442951521,
+      "learning_rate": 1.6913941962201747e-05,
+      "loss": 0.5372,
+      "step": 3437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8727021376495987,
+      "learning_rate": 1.691203981992845e-05,
+      "loss": 0.6025,
+      "step": 3438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9009716445148898,
+      "learning_rate": 1.6910137198656925e-05,
+      "loss": 0.5617,
+      "step": 3439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9791987469560515,
+      "learning_rate": 1.6908234098519024e-05,
+      "loss": 0.5147,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9245555800454982,
+      "learning_rate": 1.6906330519646622e-05,
+      "loss": 0.6286,
+      "step": 3441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9655064426884022,
+      "learning_rate": 1.6904426462171647e-05,
+      "loss": 0.6019,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8731797030694629,
+      "learning_rate": 1.690252192622604e-05,
+      "loss": 0.5452,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.930713438796544,
+      "learning_rate": 1.6900616911941783e-05,
+      "loss": 0.5488,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9807267518738538,
+      "learning_rate": 1.6898711419450897e-05,
+      "loss": 0.5992,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9244327054086132,
+      "learning_rate": 1.689680544888543e-05,
+      "loss": 0.5108,
+      "step": 3446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8835917131112488,
+      "learning_rate": 1.6894899000377462e-05,
+      "loss": 0.536,
+      "step": 3447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9100881276299515,
+      "learning_rate": 1.689299207405911e-05,
+      "loss": 0.6109,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9843100087872552,
+      "learning_rate": 1.6891084670062517e-05,
+      "loss": 0.5497,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9185017074722286,
+      "learning_rate": 1.6889176788519876e-05,
+      "loss": 0.5741,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8697110048282695,
+      "learning_rate": 1.6887268429563387e-05,
+      "loss": 0.5425,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0280757547635149,
+      "learning_rate": 1.688535959332531e-05,
+      "loss": 0.6179,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0022941901330085,
+      "learning_rate": 1.688345027993792e-05,
+      "loss": 0.6289,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0396302382621723,
+      "learning_rate": 1.6881540489533527e-05,
+      "loss": 0.6032,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9853194918842026,
+      "learning_rate": 1.6879630222244487e-05,
+      "loss": 0.6601,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9947368061129188,
+      "learning_rate": 1.6877719478203172e-05,
+      "loss": 0.4907,
+      "step": 3456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8532268987285451,
+      "learning_rate": 1.6875808257541998e-05,
+      "loss": 0.5378,
+      "step": 3457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9071315028978325,
+      "learning_rate": 1.6873896560393413e-05,
+      "loss": 0.5627,
+      "step": 3458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9136588626603805,
+      "learning_rate": 1.687198438688989e-05,
+      "loss": 0.5592,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9719371861308499,
+      "learning_rate": 1.6870071737163948e-05,
+      "loss": 0.5675,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9505609502388157,
+      "learning_rate": 1.6868158611348124e-05,
+      "loss": 0.6057,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8039359178601575,
+      "learning_rate": 1.6866245009575e-05,
+      "loss": 0.4808,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9390886461339968,
+      "learning_rate": 1.686433093197719e-05,
+      "loss": 0.6191,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9434824431556162,
+      "learning_rate": 1.686241637868734e-05,
+      "loss": 0.5357,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9336655413258358,
+      "learning_rate": 1.6860501349838114e-05,
+      "loss": 0.545,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9940741569523487,
+      "learning_rate": 1.685858584556223e-05,
+      "loss": 0.5828,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9147782411036046,
+      "learning_rate": 1.6856669865992437e-05,
+      "loss": 0.5497,
+      "step": 3467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.846982347796428,
+      "learning_rate": 1.68547534112615e-05,
+      "loss": 0.5163,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9498171544263488,
+      "learning_rate": 1.685283648150223e-05,
+      "loss": 0.6197,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9199470650867926,
+      "learning_rate": 1.6850919076847474e-05,
+      "loss": 0.6108,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0394969040093853,
+      "learning_rate": 1.68490011974301e-05,
+      "loss": 0.5443,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8709749528807426,
+      "learning_rate": 1.684708284338302e-05,
+      "loss": 0.5751,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.855026890162834,
+      "learning_rate": 1.684516401483917e-05,
+      "loss": 0.5543,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9374281824638894,
+      "learning_rate": 1.6843244711931526e-05,
+      "loss": 0.6044,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9238721362463175,
+      "learning_rate": 1.6841324934793096e-05,
+      "loss": 0.558,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9072585266971033,
+      "learning_rate": 1.6839404683556914e-05,
+      "loss": 0.4736,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9143915587444463,
+      "learning_rate": 1.6837483958356054e-05,
+      "loss": 0.6134,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.983794665294586,
+      "learning_rate": 1.6835562759323622e-05,
+      "loss": 0.5457,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9357786768311311,
+      "learning_rate": 1.683364108659275e-05,
+      "loss": 0.6289,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8728876774010064,
+      "learning_rate": 1.6831718940296617e-05,
+      "loss": 0.5035,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8528829004574374,
+      "learning_rate": 1.6829796320568416e-05,
+      "loss": 0.6151,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.954694895551522,
+      "learning_rate": 1.6827873227541393e-05,
+      "loss": 0.6269,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9438303988784065,
+      "learning_rate": 1.6825949661348812e-05,
+      "loss": 0.5902,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9450631524045531,
+      "learning_rate": 1.682402562212397e-05,
+      "loss": 0.5956,
+      "step": 3484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0221615000328568,
+      "learning_rate": 1.6822101110000207e-05,
+      "loss": 0.6432,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9340166290868366,
+      "learning_rate": 1.6820176125110886e-05,
+      "loss": 0.6274,
+      "step": 3486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9159751842545095,
+      "learning_rate": 1.681825066758941e-05,
+      "loss": 0.5367,
+      "step": 3487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8341778683720619,
+      "learning_rate": 1.6816324737569215e-05,
+      "loss": 0.5904,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9721777267639525,
+      "learning_rate": 1.681439833518376e-05,
+      "loss": 0.5366,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8214085641113914,
+      "learning_rate": 1.681247146056654e-05,
+      "loss": 0.6098,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8811256460597657,
+      "learning_rate": 1.6810544113851096e-05,
+      "loss": 0.5111,
+      "step": 3491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9484985396662438,
+      "learning_rate": 1.6808616295170983e-05,
+      "loss": 0.5921,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8813089051357941,
+      "learning_rate": 1.6806688004659803e-05,
+      "loss": 0.4919,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9816279349466661,
+      "learning_rate": 1.6804759242451177e-05,
+      "loss": 0.5854,
+      "step": 3494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.862825517856782,
+      "learning_rate": 1.6802830008678777e-05,
+      "loss": 0.5217,
+      "step": 3495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9204080613604179,
+      "learning_rate": 1.6800900303476286e-05,
+      "loss": 0.5527,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9825728162668876,
+      "learning_rate": 1.679897012697744e-05,
+      "loss": 0.5336,
+      "step": 3497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9982188434933679,
+      "learning_rate": 1.6797039479315994e-05,
+      "loss": 0.6479,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.913708551786946,
+      "learning_rate": 1.679510836062574e-05,
+      "loss": 0.5342,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9274094365407106,
+      "learning_rate": 1.6793176771040504e-05,
+      "loss": 0.5949,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9273663856313644,
+      "learning_rate": 1.6791244710694144e-05,
+      "loss": 0.5546,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.98301986004304,
+      "learning_rate": 1.678931217972055e-05,
+      "loss": 0.4835,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.84103745533568,
+      "learning_rate": 1.6787379178253642e-05,
+      "loss": 0.4985,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9698588777438701,
+      "learning_rate": 1.678544570642738e-05,
+      "loss": 0.5881,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.951776115340784,
+      "learning_rate": 1.6783511764375745e-05,
+      "loss": 0.6007,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9741451637827968,
+      "learning_rate": 1.678157735223277e-05,
+      "loss": 0.5965,
+      "step": 3506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9065524129003147,
+      "learning_rate": 1.6779642470132487e-05,
+      "loss": 0.5405,
+      "step": 3507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9250130238196568,
+      "learning_rate": 1.6777707118209004e-05,
+      "loss": 0.6035,
+      "step": 3508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9393499737202737,
+      "learning_rate": 1.6775771296596427e-05,
+      "loss": 0.5909,
+      "step": 3509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8589539160299299,
+      "learning_rate": 1.677383500542891e-05,
+      "loss": 0.5071,
+      "step": 3510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8916436165229539,
+      "learning_rate": 1.6771898244840636e-05,
+      "loss": 0.5841,
+      "step": 3511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8795522853651708,
+      "learning_rate": 1.676996101496582e-05,
+      "loss": 0.5698,
+      "step": 3512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9140986536382631,
+      "learning_rate": 1.6768023315938708e-05,
+      "loss": 0.6178,
+      "step": 3513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0834371963750848,
+      "learning_rate": 1.6766085147893583e-05,
+      "loss": 0.6471,
+      "step": 3514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.935683335427111,
+      "learning_rate": 1.6764146510964762e-05,
+      "loss": 0.6731,
+      "step": 3515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8334877532242587,
+      "learning_rate": 1.676220740528659e-05,
+      "loss": 0.4929,
+      "step": 3516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8479848617399672,
+      "learning_rate": 1.676026783099344e-05,
+      "loss": 0.5383,
+      "step": 3517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9399644753508193,
+      "learning_rate": 1.6758327788219722e-05,
+      "loss": 0.5711,
+      "step": 3518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8976978401914122,
+      "learning_rate": 1.6756387277099885e-05,
+      "loss": 0.5118,
+      "step": 3519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9096358011034394,
+      "learning_rate": 1.6754446297768404e-05,
+      "loss": 0.5343,
+      "step": 3520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8756936077029496,
+      "learning_rate": 1.6752504850359785e-05,
+      "loss": 0.5871,
+      "step": 3521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8481418927380738,
+      "learning_rate": 1.6750562935008572e-05,
+      "loss": 0.5126,
+      "step": 3522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.875692294080384,
+      "learning_rate": 1.6748620551849333e-05,
+      "loss": 0.6195,
+      "step": 3523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9289899034178508,
+      "learning_rate": 1.6746677701016675e-05,
+      "loss": 0.6014,
+      "step": 3524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.798198379662321,
+      "learning_rate": 1.674473438264524e-05,
+      "loss": 0.5155,
+      "step": 3525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8364506278435189,
+      "learning_rate": 1.674279059686969e-05,
+      "loss": 0.5638,
+      "step": 3526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9674610184024626,
+      "learning_rate": 1.6740846343824734e-05,
+      "loss": 0.5908,
+      "step": 3527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8897752692194084,
+      "learning_rate": 1.6738901623645107e-05,
+      "loss": 0.5301,
+      "step": 3528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0070755438073646,
+      "learning_rate": 1.6736956436465573e-05,
+      "loss": 0.6214,
+      "step": 3529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.853310413688308,
+      "learning_rate": 1.6735010782420934e-05,
+      "loss": 0.5819,
+      "step": 3530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.002604659974855,
+      "learning_rate": 1.6733064661646023e-05,
+      "loss": 0.5757,
+      "step": 3531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8825996930380876,
+      "learning_rate": 1.67311180742757e-05,
+      "loss": 0.5913,
+      "step": 3532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9185858297292431,
+      "learning_rate": 1.672917102044487e-05,
+      "loss": 0.6152,
+      "step": 3533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.898173400038949,
+      "learning_rate": 1.6727223500288458e-05,
+      "loss": 0.6118,
+      "step": 3534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0034091029734553,
+      "learning_rate": 1.672527551394142e-05,
+      "loss": 0.5808,
+      "step": 3535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.882816022561754,
+      "learning_rate": 1.6723327061538753e-05,
+      "loss": 0.5979,
+      "step": 3536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8911535124949506,
+      "learning_rate": 1.672137814321549e-05,
+      "loss": 0.5396,
+      "step": 3537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8796186206450183,
+      "learning_rate": 1.6719428759106676e-05,
+      "loss": 0.5201,
+      "step": 3538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9169457655302597,
+      "learning_rate": 1.6717478909347417e-05,
+      "loss": 0.5862,
+      "step": 3539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9814266566832116,
+      "learning_rate": 1.671552859407282e-05,
+      "loss": 0.6225,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8622223722404974,
+      "learning_rate": 1.6713577813418058e-05,
+      "loss": 0.5983,
+      "step": 3541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9993934642157488,
+      "learning_rate": 1.67116265675183e-05,
+      "loss": 0.5768,
+      "step": 3542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9092726740215555,
+      "learning_rate": 1.6709674856508775e-05,
+      "loss": 0.565,
+      "step": 3543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9232881931366909,
+      "learning_rate": 1.6707722680524735e-05,
+      "loss": 0.5864,
+      "step": 3544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9530394049860118,
+      "learning_rate": 1.6705770039701464e-05,
+      "loss": 0.5408,
+      "step": 3545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9506612173293659,
+      "learning_rate": 1.670381693417428e-05,
+      "loss": 0.6284,
+      "step": 3546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8434394278434862,
+      "learning_rate": 1.6701863364078524e-05,
+      "loss": 0.5427,
+      "step": 3547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9724131503049097,
+      "learning_rate": 1.6699909329549583e-05,
+      "loss": 0.4966,
+      "step": 3548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9893191331651435,
+      "learning_rate": 1.669795483072287e-05,
+      "loss": 0.5897,
+      "step": 3549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8701298054577694,
+      "learning_rate": 1.6695999867733824e-05,
+      "loss": 0.5424,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9389970593482257,
+      "learning_rate": 1.669404444071793e-05,
+      "loss": 0.6622,
+      "step": 3551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9024202866939837,
+      "learning_rate": 1.6692088549810695e-05,
+      "loss": 0.4986,
+      "step": 3552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9196638626707861,
+      "learning_rate": 1.6690132195147655e-05,
+      "loss": 0.5415,
+      "step": 3553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9315837796528372,
+      "learning_rate": 1.668817537686439e-05,
+      "loss": 0.5866,
+      "step": 3554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9883947924463582,
+      "learning_rate": 1.6686218095096506e-05,
+      "loss": 0.5928,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9627628774229461,
+      "learning_rate": 1.6684260349979637e-05,
+      "loss": 0.5831,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9476768527246094,
+      "learning_rate": 1.6682302141649452e-05,
+      "loss": 0.5994,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9485565562128317,
+      "learning_rate": 1.668034347024166e-05,
+      "loss": 0.5723,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9760330608078868,
+      "learning_rate": 1.667838433589199e-05,
+      "loss": 0.5349,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8864957985018198,
+      "learning_rate": 1.6676424738736208e-05,
+      "loss": 0.5838,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9307026791978568,
+      "learning_rate": 1.6674464678910117e-05,
+      "loss": 0.5658,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9044244733439744,
+      "learning_rate": 1.667250415654954e-05,
+      "loss": 0.5329,
+      "step": 3562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9358310455208178,
+      "learning_rate": 1.6670543171790347e-05,
+      "loss": 0.6028,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9945180142864156,
+      "learning_rate": 1.6668581724768423e-05,
+      "loss": 0.5394,
+      "step": 3564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8602807961739487,
+      "learning_rate": 1.6666619815619703e-05,
+      "loss": 0.539,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8948693345748759,
+      "learning_rate": 1.6664657444480145e-05,
+      "loss": 0.6245,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8565794868024554,
+      "learning_rate": 1.666269461148574e-05,
+      "loss": 0.4878,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.866446067109164,
+      "learning_rate": 1.6660731316772503e-05,
+      "loss": 0.5559,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9300978205223425,
+      "learning_rate": 1.6658767560476494e-05,
+      "loss": 0.5421,
+      "step": 3569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8743919086457418,
+      "learning_rate": 1.6656803342733804e-05,
+      "loss": 0.6016,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9887159020971205,
+      "learning_rate": 1.6654838663680542e-05,
+      "loss": 0.5941,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9935905572056851,
+      "learning_rate": 1.6652873523452867e-05,
+      "loss": 0.5825,
+      "step": 3572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9917578015678681,
+      "learning_rate": 1.6650907922186958e-05,
+      "loss": 0.5464,
+      "step": 3573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8836985763212661,
+      "learning_rate": 1.6648941860019028e-05,
+      "loss": 0.5667,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8837790954049046,
+      "learning_rate": 1.6646975337085323e-05,
+      "loss": 0.5275,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0543057172385015,
+      "learning_rate": 1.6645008353522122e-05,
+      "loss": 0.6152,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9161591991044842,
+      "learning_rate": 1.6643040909465743e-05,
+      "loss": 0.5601,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9601951864694586,
+      "learning_rate": 1.6641073005052516e-05,
+      "loss": 0.5857,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9765714977543812,
+      "learning_rate": 1.663910464041882e-05,
+      "loss": 0.5903,
+      "step": 3579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8749919554487148,
+      "learning_rate": 1.6637135815701066e-05,
+      "loss": 0.5547,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.943868588563792,
+      "learning_rate": 1.663516653103568e-05,
+      "loss": 0.5624,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9406057711354047,
+      "learning_rate": 1.6633196786559143e-05,
+      "loss": 0.5454,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9602637556293583,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.6411,
+      "step": 3583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8818966099422427,
+      "learning_rate": 1.662925591871864e-05,
+      "loss": 0.5759,
+      "step": 3584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0002142171935597,
+      "learning_rate": 1.6627284795627777e-05,
+      "loss": 0.5652,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8726073504452184,
+      "learning_rate": 1.6625313213271953e-05,
+      "loss": 0.5773,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9105240364055169,
+      "learning_rate": 1.66233411717878e-05,
+      "loss": 0.4951,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8467521769576934,
+      "learning_rate": 1.6621368671311973e-05,
+      "loss": 0.5547,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9618991888429478,
+      "learning_rate": 1.6619395711981183e-05,
+      "loss": 0.64,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9131903387125504,
+      "learning_rate": 1.661742229393213e-05,
+      "loss": 0.6045,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9734975634683273,
+      "learning_rate": 1.6615448417301588e-05,
+      "loss": 0.6041,
+      "step": 3591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9554196630977829,
+      "learning_rate": 1.6613474082226338e-05,
+      "loss": 0.6134,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0722452324867415,
+      "learning_rate": 1.6611499288843194e-05,
+      "loss": 0.6594,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8659926589103515,
+      "learning_rate": 1.660952403728902e-05,
+      "loss": 0.5241,
+      "step": 3594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0027663664973614,
+      "learning_rate": 1.6607548327700694e-05,
+      "loss": 0.671,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8707474663912872,
+      "learning_rate": 1.6605572160215126e-05,
+      "loss": 0.5783,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0087074986539963,
+      "learning_rate": 1.6603595534969266e-05,
+      "loss": 0.638,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9712820924532131,
+      "learning_rate": 1.6601618452100098e-05,
+      "loss": 0.6031,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9282173136440852,
+      "learning_rate": 1.659964091174462e-05,
+      "loss": 0.5563,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9858593134307094,
+      "learning_rate": 1.6597662914039885e-05,
+      "loss": 0.5695,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9790081161566879,
+      "learning_rate": 1.659568445912296e-05,
+      "loss": 0.6717,
+      "step": 3601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.92396838455597,
+      "learning_rate": 1.6593705547130955e-05,
+      "loss": 0.5798,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8781891791299089,
+      "learning_rate": 1.6591726178201e-05,
+      "loss": 0.5568,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9532094188863465,
+      "learning_rate": 1.6589746352470275e-05,
+      "loss": 0.5711,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9605308425776762,
+      "learning_rate": 1.6587766070075965e-05,
+      "loss": 0.5674,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9075877004257134,
+      "learning_rate": 1.6585785331155312e-05,
+      "loss": 0.5982,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9778240081915002,
+      "learning_rate": 1.6583804135845582e-05,
+      "loss": 0.6316,
+      "step": 3607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8063905073185061,
+      "learning_rate": 1.658182248428406e-05,
+      "loss": 0.5393,
+      "step": 3608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9095242303970597,
+      "learning_rate": 1.6579840376608076e-05,
+      "loss": 0.5845,
+      "step": 3609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9260816695897505,
+      "learning_rate": 1.6577857812954994e-05,
+      "loss": 0.5652,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.998827823223787,
+      "learning_rate": 1.65758747934622e-05,
+      "loss": 0.528,
+      "step": 3611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9114557389191057,
+      "learning_rate": 1.6573891318267113e-05,
+      "loss": 0.629,
+      "step": 3612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9823821174946192,
+      "learning_rate": 1.6571907387507194e-05,
+      "loss": 0.6058,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0511732864639134,
+      "learning_rate": 1.6569923001319916e-05,
+      "loss": 0.5967,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9686847202243349,
+      "learning_rate": 1.6567938159842807e-05,
+      "loss": 0.6398,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8933691431087123,
+      "learning_rate": 1.6565952863213407e-05,
+      "loss": 0.5597,
+      "step": 3616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8945158188311984,
+      "learning_rate": 1.65639671115693e-05,
+      "loss": 0.5664,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8154853016245658,
+      "learning_rate": 1.6561980905048087e-05,
+      "loss": 0.5101,
+      "step": 3618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8273776997836692,
+      "learning_rate": 1.6559994243787427e-05,
+      "loss": 0.4892,
+      "step": 3619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0302635716889421,
+      "learning_rate": 1.655800712792498e-05,
+      "loss": 0.5976,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0043279125071218,
+      "learning_rate": 1.6556019557598453e-05,
+      "loss": 0.5788,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9753388668576959,
+      "learning_rate": 1.6554031532945588e-05,
+      "loss": 0.6504,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9597580412921606,
+      "learning_rate": 1.6552043054104153e-05,
+      "loss": 0.5883,
+      "step": 3623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0242495713982198,
+      "learning_rate": 1.6550054121211946e-05,
+      "loss": 0.5753,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7932872760131129,
+      "learning_rate": 1.6548064734406798e-05,
+      "loss": 0.4609,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9463995135467075,
+      "learning_rate": 1.654607489382657e-05,
+      "loss": 0.588,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9566650907230213,
+      "learning_rate": 1.654408459960916e-05,
+      "loss": 0.6569,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9198226238212484,
+      "learning_rate": 1.6542093851892493e-05,
+      "loss": 0.5615,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9137048670369079,
+      "learning_rate": 1.654010265081452e-05,
+      "loss": 0.5889,
+      "step": 3629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0686054851272107,
+      "learning_rate": 1.653811099651324e-05,
+      "loss": 0.5917,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8316623722805385,
+      "learning_rate": 1.6536118889126665e-05,
+      "loss": 0.5438,
+      "step": 3631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.900597140469948,
+      "learning_rate": 1.6534126328792846e-05,
+      "loss": 0.5737,
+      "step": 3632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.983236693684268,
+      "learning_rate": 1.653213331564987e-05,
+      "loss": 0.5542,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8923703352786861,
+      "learning_rate": 1.653013984983585e-05,
+      "loss": 0.5854,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.860331550092371,
+      "learning_rate": 1.6528145931488934e-05,
+      "loss": 0.4834,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.934322417674138,
+      "learning_rate": 1.6526151560747294e-05,
+      "loss": 0.5552,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0653721164248258,
+      "learning_rate": 1.6524156737749132e-05,
+      "loss": 0.6018,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8718089486691982,
+      "learning_rate": 1.6522161462632705e-05,
+      "loss": 0.5718,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9025461147431527,
+      "learning_rate": 1.6520165735536268e-05,
+      "loss": 0.4984,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0448045557872034,
+      "learning_rate": 1.651816955659813e-05,
+      "loss": 0.6191,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9643104817920352,
+      "learning_rate": 1.6516172925956624e-05,
+      "loss": 0.5801,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8573755137208006,
+      "learning_rate": 1.6514175843750112e-05,
+      "loss": 0.5469,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9168281111480824,
+      "learning_rate": 1.6512178310116994e-05,
+      "loss": 0.498,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9980497424146102,
+      "learning_rate": 1.6510180325195696e-05,
+      "loss": 0.6049,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8352927441471474,
+      "learning_rate": 1.6508181889124678e-05,
+      "loss": 0.562,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9098235155117534,
+      "learning_rate": 1.650618300204242e-05,
+      "loss": 0.5712,
+      "step": 3646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9318970570266037,
+      "learning_rate": 1.6504183664087458e-05,
+      "loss": 0.5857,
+      "step": 3647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9322556307904534,
+      "learning_rate": 1.6502183875398335e-05,
+      "loss": 0.4988,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8689832946583307,
+      "learning_rate": 1.6500183636113637e-05,
+      "loss": 0.5214,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9994019497128986,
+      "learning_rate": 1.649818294637198e-05,
+      "loss": 0.589,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9892900604091126,
+      "learning_rate": 1.6496181806312005e-05,
+      "loss": 0.5811,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8117548953306063,
+      "learning_rate": 1.6494180216072397e-05,
+      "loss": 0.5841,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9213653596995112,
+      "learning_rate": 1.649217817579186e-05,
+      "loss": 0.5614,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.988994948049542,
+      "learning_rate": 1.6490175685609133e-05,
+      "loss": 0.5912,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9457260692079452,
+      "learning_rate": 1.6488172745662984e-05,
+      "loss": 0.5493,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8973934121192513,
+      "learning_rate": 1.6486169356092224e-05,
+      "loss": 0.6416,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0003701542148136,
+      "learning_rate": 1.648416551703568e-05,
+      "loss": 0.601,
+      "step": 3657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9375819350879385,
+      "learning_rate": 1.6482161228632217e-05,
+      "loss": 0.5633,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8715810769204608,
+      "learning_rate": 1.648015649102073e-05,
+      "loss": 0.5143,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.956343085905747,
+      "learning_rate": 1.6478151304340144e-05,
+      "loss": 0.598,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8466436290465038,
+      "learning_rate": 1.647614566872942e-05,
+      "loss": 0.5448,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0018889533958015,
+      "learning_rate": 1.6474139584327548e-05,
+      "loss": 0.6558,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9768304608041964,
+      "learning_rate": 1.647213305127354e-05,
+      "loss": 0.4928,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9476325871815199,
+      "learning_rate": 1.6470126069706456e-05,
+      "loss": 0.6763,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9116612300802762,
+      "learning_rate": 1.6468118639765376e-05,
+      "loss": 0.5577,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9705610218173512,
+      "learning_rate": 1.646611076158941e-05,
+      "loss": 0.6202,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9044189603954035,
+      "learning_rate": 1.6464102435317702e-05,
+      "loss": 0.5508,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8682864121536775,
+      "learning_rate": 1.6462093661089432e-05,
+      "loss": 0.643,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9042250301625764,
+      "learning_rate": 1.64600844390438e-05,
+      "loss": 0.614,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.915033666481628,
+      "learning_rate": 1.6458074769320046e-05,
+      "loss": 0.5764,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8218929072923022,
+      "learning_rate": 1.6456064652057443e-05,
+      "loss": 0.5159,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8778388677902503,
+      "learning_rate": 1.6454054087395284e-05,
+      "loss": 0.5133,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8389863239156715,
+      "learning_rate": 1.6452043075472898e-05,
+      "loss": 0.5176,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9435807760681507,
+      "learning_rate": 1.6450031616429655e-05,
+      "loss": 0.6082,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9169409518546056,
+      "learning_rate": 1.6448019710404938e-05,
+      "loss": 0.554,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9801572359431846,
+      "learning_rate": 1.6446007357538178e-05,
+      "loss": 0.5588,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9262934877911382,
+      "learning_rate": 1.6443994557968826e-05,
+      "loss": 0.5349,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9778690714497104,
+      "learning_rate": 1.6441981311836363e-05,
+      "loss": 0.5439,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9550533350922927,
+      "learning_rate": 1.643996761928031e-05,
+      "loss": 0.603,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9514029700405338,
+      "learning_rate": 1.6437953480440217e-05,
+      "loss": 0.6207,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9208468742014804,
+      "learning_rate": 1.6435938895455653e-05,
+      "loss": 0.5565,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9614563981238538,
+      "learning_rate": 1.6433923864466235e-05,
+      "loss": 0.5365,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0040116203536171,
+      "learning_rate": 1.6431908387611604e-05,
+      "loss": 0.5948,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8657093495585827,
+      "learning_rate": 1.642989246503142e-05,
+      "loss": 0.5876,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9683313092980244,
+      "learning_rate": 1.6427876096865394e-05,
+      "loss": 0.5575,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9735112189429884,
+      "learning_rate": 1.6425859283253255e-05,
+      "loss": 0.6037,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8483864927774368,
+      "learning_rate": 1.642384202433477e-05,
+      "loss": 0.5197,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0289501655311173,
+      "learning_rate": 1.6421824320249732e-05,
+      "loss": 0.5883,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8065027388887831,
+      "learning_rate": 1.641980617113796e-05,
+      "loss": 0.5542,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9721774427241994,
+      "learning_rate": 1.6417787577139317e-05,
+      "loss": 0.5892,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.903531584027664,
+      "learning_rate": 1.641576853839369e-05,
+      "loss": 0.5987,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9247698353933245,
+      "learning_rate": 1.641374905504099e-05,
+      "loss": 0.5728,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9114264057369905,
+      "learning_rate": 1.641172912722117e-05,
+      "loss": 0.5938,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9351345133128272,
+      "learning_rate": 1.640970875507421e-05,
+      "loss": 0.5918,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.847331838532848,
+      "learning_rate": 1.640768793874012e-05,
+      "loss": 0.5788,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9834974607562615,
+      "learning_rate": 1.6405666678358934e-05,
+      "loss": 0.553,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9635080252497199,
+      "learning_rate": 1.6403644974070732e-05,
+      "loss": 0.5437,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9081995856472028,
+      "learning_rate": 1.6401622826015616e-05,
+      "loss": 0.5518,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8448121229095162,
+      "learning_rate": 1.6399600234333716e-05,
+      "loss": 0.5327,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9396638616959802,
+      "learning_rate": 1.6397577199165192e-05,
+      "loss": 0.6707,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8932577772825789,
+      "learning_rate": 1.639555372065025e-05,
+      "loss": 0.5025,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9081954188369703,
+      "learning_rate": 1.6393529798929103e-05,
+      "loss": 0.5871,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8853290866239732,
+      "learning_rate": 1.639150543414201e-05,
+      "loss": 0.5313,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9233280166174495,
+      "learning_rate": 1.6389480626429262e-05,
+      "loss": 0.5942,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9463063765768046,
+      "learning_rate": 1.6387455375931174e-05,
+      "loss": 0.5392,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8730763067649788,
+      "learning_rate": 1.6385429682788095e-05,
+      "loss": 0.544,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9316457082221453,
+      "learning_rate": 1.63834035471404e-05,
+      "loss": 0.5758,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9542199542196371,
+      "learning_rate": 1.6381376969128508e-05,
+      "loss": 0.6123,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9416443659815187,
+      "learning_rate": 1.6379349948892845e-05,
+      "loss": 0.4821,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9333823224097688,
+      "learning_rate": 1.6377322486573892e-05,
+      "loss": 0.602,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8409452002423059,
+      "learning_rate": 1.637529458231215e-05,
+      "loss": 0.5059,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9280610886053118,
+      "learning_rate": 1.637326623624814e-05,
+      "loss": 0.6287,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9096302289339779,
+      "learning_rate": 1.637123744852244e-05,
+      "loss": 0.5049,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0049544036616702,
+      "learning_rate": 1.6369208219275635e-05,
+      "loss": 0.6629,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9013101190306692,
+      "learning_rate": 1.6367178548648347e-05,
+      "loss": 0.5969,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0162996039363488,
+      "learning_rate": 1.6365148436781235e-05,
+      "loss": 0.5464,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0263918742129827,
+      "learning_rate": 1.6363117883814986e-05,
+      "loss": 0.6456,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.860333941053944,
+      "learning_rate": 1.6361086889890307e-05,
+      "loss": 0.5763,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8528743046913408,
+      "learning_rate": 1.635905545514795e-05,
+      "loss": 0.5548,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9660427612544795,
+      "learning_rate": 1.635702357972869e-05,
+      "loss": 0.6462,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9134307436287689,
+      "learning_rate": 1.6354991263773338e-05,
+      "loss": 0.5735,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.902943882911617,
+      "learning_rate": 1.6352958507422727e-05,
+      "loss": 0.53,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9348713243516186,
+      "learning_rate": 1.635092531081772e-05,
+      "loss": 0.6143,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8468682556085794,
+      "learning_rate": 1.634889167409923e-05,
+      "loss": 0.4976,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8311154744443643,
+      "learning_rate": 1.6346857597408174e-05,
+      "loss": 0.455,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8780250093580937,
+      "learning_rate": 1.634482308088552e-05,
+      "loss": 0.579,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8297776086320948,
+      "learning_rate": 1.6342788124672255e-05,
+      "loss": 0.5381,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9539167630401393,
+      "learning_rate": 1.634075272890939e-05,
+      "loss": 0.5416,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0019019974902386,
+      "learning_rate": 1.6338716893737995e-05,
+      "loss": 0.6171,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9788484051442959,
+      "learning_rate": 1.6336680619299138e-05,
+      "loss": 0.5867,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9222090771321823,
+      "learning_rate": 1.6334643905733932e-05,
+      "loss": 0.5908,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.876664958229384,
+      "learning_rate": 1.633260675318353e-05,
+      "loss": 0.5143,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8933905718102726,
+      "learning_rate": 1.633056916178909e-05,
+      "loss": 0.5128,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8924163368771747,
+      "learning_rate": 1.6328531131691823e-05,
+      "loss": 0.5776,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8375540811736663,
+      "learning_rate": 1.6326492663032964e-05,
+      "loss": 0.5809,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9864446004257903,
+      "learning_rate": 1.6324453755953772e-05,
+      "loss": 0.6347,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9735173503758445,
+      "learning_rate": 1.6322414410595548e-05,
+      "loss": 0.5306,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9369967945864461,
+      "learning_rate": 1.6320374627099612e-05,
+      "loss": 0.5496,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8931827555344554,
+      "learning_rate": 1.631833440560732e-05,
+      "loss": 0.5092,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0429319179996712,
+      "learning_rate": 1.631629374626006e-05,
+      "loss": 0.5573,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0092123786527019,
+      "learning_rate": 1.6314252649199244e-05,
+      "loss": 0.5645,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9221786043515009,
+      "learning_rate": 1.6312211114566322e-05,
+      "loss": 0.5584,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8863761142956672,
+      "learning_rate": 1.6310169142502767e-05,
+      "loss": 0.525,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9347174132990639,
+      "learning_rate": 1.6308126733150088e-05,
+      "loss": 0.6126,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0632686311470156,
+      "learning_rate": 1.6306083886649823e-05,
+      "loss": 0.6301,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0074442161292392,
+      "learning_rate": 1.6304040603143537e-05,
+      "loss": 0.5749,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9485310866229572,
+      "learning_rate": 1.6301996882772828e-05,
+      "loss": 0.6036,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9259055405982872,
+      "learning_rate": 1.6299952725679325e-05,
+      "loss": 0.6138,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9216543019525134,
+      "learning_rate": 1.6297908132004688e-05,
+      "loss": 0.5322,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8867586367561826,
+      "learning_rate": 1.6295863101890603e-05,
+      "loss": 0.5266,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.930621927448145,
+      "learning_rate": 1.6293817635478787e-05,
+      "loss": 0.5562,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.834710674438063,
+      "learning_rate": 1.629177173291099e-05,
+      "loss": 0.5912,
+      "step": 3752
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.945039577387616,
+      "learning_rate": 1.6289725394328998e-05,
+      "loss": 0.5675,
+      "step": 3753
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0046881856885677,
+      "learning_rate": 1.6287678619874614e-05,
+      "loss": 0.5822,
+      "step": 3754
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9444310869381672,
+      "learning_rate": 1.628563140968968e-05,
+      "loss": 0.6171,
+      "step": 3755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8936053185510102,
+      "learning_rate": 1.6283583763916062e-05,
+      "loss": 0.5566,
+      "step": 3756
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.851601255245641,
+      "learning_rate": 1.6281535682695663e-05,
+      "loss": 0.5427,
+      "step": 3757
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9061461756195821,
+      "learning_rate": 1.6279487166170412e-05,
+      "loss": 0.4949,
+      "step": 3758
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9271870381103675,
+      "learning_rate": 1.627743821448227e-05,
+      "loss": 0.5722,
+      "step": 3759
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0091095183053302,
+      "learning_rate": 1.6275388827773235e-05,
+      "loss": 0.5942,
+      "step": 3760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8860061586258776,
+      "learning_rate": 1.627333900618531e-05,
+      "loss": 0.5878,
+      "step": 3761
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9550515256655273,
+      "learning_rate": 1.627128874986056e-05,
+      "loss": 0.534,
+      "step": 3762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9466939591667332,
+      "learning_rate": 1.626923805894107e-05,
+      "loss": 0.5818,
+      "step": 3763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8593634170820906,
+      "learning_rate": 1.6267186933568934e-05,
+      "loss": 0.623,
+      "step": 3764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9055110244963029,
+      "learning_rate": 1.6265135373886303e-05,
+      "loss": 0.5561,
+      "step": 3765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9522349882853491,
+      "learning_rate": 1.6263083380035352e-05,
+      "loss": 0.6138,
+      "step": 3766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9804021740033947,
+      "learning_rate": 1.6261030952158275e-05,
+      "loss": 0.6136,
+      "step": 3767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9771039368379923,
+      "learning_rate": 1.625897809039731e-05,
+      "loss": 0.5797,
+      "step": 3768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.993286275378227,
+      "learning_rate": 1.625692479489471e-05,
+      "loss": 0.6008,
+      "step": 3769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9382597208111868,
+      "learning_rate": 1.6254871065792776e-05,
+      "loss": 0.6317,
+      "step": 3770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9130802421175084,
+      "learning_rate": 1.6252816903233822e-05,
+      "loss": 0.6397,
+      "step": 3771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8870309436502534,
+      "learning_rate": 1.6250762307360206e-05,
+      "loss": 0.5487,
+      "step": 3772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8630507613282848,
+      "learning_rate": 1.6248707278314303e-05,
+      "loss": 0.6063,
+      "step": 3773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8204465669044464,
+      "learning_rate": 1.6246651816238533e-05,
+      "loss": 0.5359,
+      "step": 3774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8791486581655621,
+      "learning_rate": 1.6244595921275327e-05,
+      "loss": 0.5696,
+      "step": 3775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8884869209713185,
+      "learning_rate": 1.624253959356717e-05,
+      "loss": 0.5646,
+      "step": 3776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8825282203386668,
+      "learning_rate": 1.6240482833256548e-05,
+      "loss": 0.5236,
+      "step": 3777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9366188651664406,
+      "learning_rate": 1.6238425640486005e-05,
+      "loss": 0.5926,
+      "step": 3778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9339163870662345,
+      "learning_rate": 1.62363680153981e-05,
+      "loss": 0.5729,
+      "step": 3779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9361957987894465,
+      "learning_rate": 1.623430995813542e-05,
+      "loss": 0.6206,
+      "step": 3780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9142015976316188,
+      "learning_rate": 1.6232251468840593e-05,
+      "loss": 0.6301,
+      "step": 3781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8585683173377849,
+      "learning_rate": 1.6230192547656264e-05,
+      "loss": 0.5282,
+      "step": 3782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8833783575926998,
+      "learning_rate": 1.622813319472512e-05,
+      "loss": 0.5974,
+      "step": 3783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8608490199464518,
+      "learning_rate": 1.622607341018987e-05,
+      "loss": 0.5524,
+      "step": 3784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8848172429593518,
+      "learning_rate": 1.622401319419325e-05,
+      "loss": 0.5606,
+      "step": 3785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8042328269827678,
+      "learning_rate": 1.6221952546878044e-05,
+      "loss": 0.5154,
+      "step": 3786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8958394235761792,
+      "learning_rate": 1.621989146838704e-05,
+      "loss": 0.6145,
+      "step": 3787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8922850703740579,
+      "learning_rate": 1.6217829958863077e-05,
+      "loss": 0.5611,
+      "step": 3788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9013966060877183,
+      "learning_rate": 1.6215768018449015e-05,
+      "loss": 0.5549,
+      "step": 3789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9854311867726121,
+      "learning_rate": 1.621370564728774e-05,
+      "loss": 0.5854,
+      "step": 3790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9686953885478913,
+      "learning_rate": 1.6211642845522173e-05,
+      "loss": 0.5728,
+      "step": 3791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.848322047870251,
+      "learning_rate": 1.620957961329527e-05,
+      "loss": 0.573,
+      "step": 3792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9168811248245282,
+      "learning_rate": 1.620751595075001e-05,
+      "loss": 0.5687,
+      "step": 3793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8563048407779456,
+      "learning_rate": 1.6205451858029392e-05,
+      "loss": 0.5997,
+      "step": 3794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9474923997403427,
+      "learning_rate": 1.620338733527647e-05,
+      "loss": 0.5352,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9299568257108631,
+      "learning_rate": 1.6201322382634307e-05,
+      "loss": 0.5198,
+      "step": 3796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9148308383657838,
+      "learning_rate": 1.6199257000246004e-05,
+      "loss": 0.5742,
+      "step": 3797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9363226570594413,
+      "learning_rate": 1.6197191188254692e-05,
+      "loss": 0.5975,
+      "step": 3798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0230584582962952,
+      "learning_rate": 1.6195124946803527e-05,
+      "loss": 0.6454,
+      "step": 3799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8742058995001382,
+      "learning_rate": 1.6193058276035696e-05,
+      "loss": 0.545,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8711637158069439,
+      "learning_rate": 1.6190991176094416e-05,
+      "loss": 0.6298,
+      "step": 3801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9418798072740513,
+      "learning_rate": 1.6188923647122946e-05,
+      "loss": 0.5593,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0387832078071524,
+      "learning_rate": 1.6186855689264556e-05,
+      "loss": 0.6113,
+      "step": 3803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.926162906855895,
+      "learning_rate": 1.618478730266255e-05,
+      "loss": 0.572,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.885613716903195,
+      "learning_rate": 1.6182718487460274e-05,
+      "loss": 0.5771,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.943864432630467,
+      "learning_rate": 1.618064924380109e-05,
+      "loss": 0.5278,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9306896144782546,
+      "learning_rate": 1.6178579571828392e-05,
+      "loss": 0.5818,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9218870472956118,
+      "learning_rate": 1.6176509471685616e-05,
+      "loss": 0.5568,
+      "step": 3808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8969001171613205,
+      "learning_rate": 1.6174438943516206e-05,
+      "loss": 0.5828,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9669693971910587,
+      "learning_rate": 1.617236798746366e-05,
+      "loss": 0.58,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9136607722559579,
+      "learning_rate": 1.6170296603671483e-05,
+      "loss": 0.5508,
+      "step": 3811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8605010059912521,
+      "learning_rate": 1.6168224792283226e-05,
+      "loss": 0.5363,
+      "step": 3812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9501176650973622,
+      "learning_rate": 1.616615255344246e-05,
+      "loss": 0.5826,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0011978284197116,
+      "learning_rate": 1.6164079887292795e-05,
+      "loss": 0.6175,
+      "step": 3814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8646956502967698,
+      "learning_rate": 1.6162006793977858e-05,
+      "loss": 0.5554,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8299241484026281,
+      "learning_rate": 1.615993327364132e-05,
+      "loss": 0.57,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9461613392861088,
+      "learning_rate": 1.6157859326426865e-05,
+      "loss": 0.6507,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9065221382423563,
+      "learning_rate": 1.6155784952478227e-05,
+      "loss": 0.544,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9547433861128523,
+      "learning_rate": 1.6153710151939145e-05,
+      "loss": 0.5617,
+      "step": 3819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9843060940520709,
+      "learning_rate": 1.615163492495341e-05,
+      "loss": 0.6266,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9216513794765518,
+      "learning_rate": 1.6149559271664835e-05,
+      "loss": 0.5406,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8997307743019719,
+      "learning_rate": 1.6147483192217252e-05,
+      "loss": 0.5437,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8841110559575263,
+      "learning_rate": 1.614540668675454e-05,
+      "loss": 0.5412,
+      "step": 3823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8190597944819852,
+      "learning_rate": 1.6143329755420592e-05,
+      "loss": 0.5199,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8811575545591398,
+      "learning_rate": 1.6141252398359347e-05,
+      "loss": 0.5517,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9495478226528136,
+      "learning_rate": 1.6139174615714753e-05,
+      "loss": 0.6203,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8797998768085065,
+      "learning_rate": 1.6137096407630805e-05,
+      "loss": 0.5824,
+      "step": 3827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9478838526700144,
+      "learning_rate": 1.613501777425152e-05,
+      "loss": 0.5624,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8509726001493284,
+      "learning_rate": 1.6132938715720946e-05,
+      "loss": 0.5516,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9303984729844625,
+      "learning_rate": 1.6130859232183155e-05,
+      "loss": 0.5794,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8570119246962986,
+      "learning_rate": 1.612877932378226e-05,
+      "loss": 0.5966,
+      "step": 3831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.903863068393604,
+      "learning_rate": 1.6126698990662393e-05,
+      "loss": 0.6146,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.910826964802132,
+      "learning_rate": 1.6124618232967722e-05,
+      "loss": 0.5759,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8503801686254258,
+      "learning_rate": 1.6122537050842443e-05,
+      "loss": 0.5785,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8646395305483849,
+      "learning_rate": 1.612045544443077e-05,
+      "loss": 0.5956,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8979625542240427,
+      "learning_rate": 1.611837341387697e-05,
+      "loss": 0.5682,
+      "step": 3836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.877739368460374,
+      "learning_rate": 1.6116290959325318e-05,
+      "loss": 0.567,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8941178925109462,
+      "learning_rate": 1.6114208080920125e-05,
+      "loss": 0.4819,
+      "step": 3838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9798634018310648,
+      "learning_rate": 1.6112124778805734e-05,
+      "loss": 0.6096,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9032849513071829,
+      "learning_rate": 1.611004105312652e-05,
+      "loss": 0.4802,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9156165559397718,
+      "learning_rate": 1.610795690402688e-05,
+      "loss": 0.5825,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9257049329284944,
+      "learning_rate": 1.6105872331651245e-05,
+      "loss": 0.5467,
+      "step": 3842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9240778610764073,
+      "learning_rate": 1.610378733614407e-05,
+      "loss": 0.6246,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9901370283969468,
+      "learning_rate": 1.6101701917649852e-05,
+      "loss": 0.6262,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0349073072890402,
+      "learning_rate": 1.60996160763131e-05,
+      "loss": 0.5461,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8941247919258477,
+      "learning_rate": 1.6097529812278364e-05,
+      "loss": 0.5748,
+      "step": 3846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8401322694110076,
+      "learning_rate": 1.6095443125690222e-05,
+      "loss": 0.4761,
+      "step": 3847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9457057504482168,
+      "learning_rate": 1.609335601669328e-05,
+      "loss": 0.6018,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.898059756360942,
+      "learning_rate": 1.6091268485432165e-05,
+      "loss": 0.6018,
+      "step": 3849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9574558404478161,
+      "learning_rate": 1.6089180532051552e-05,
+      "loss": 0.5048,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9188939124978395,
+      "learning_rate": 1.6087092156696127e-05,
+      "loss": 0.5522,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9691554158650906,
+      "learning_rate": 1.6085003359510616e-05,
+      "loss": 0.5651,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9953360714177077,
+      "learning_rate": 1.6082914140639768e-05,
+      "loss": 0.5312,
+      "step": 3853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9516548297556112,
+      "learning_rate": 1.6080824500228367e-05,
+      "loss": 0.5831,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9744591097344041,
+      "learning_rate": 1.607873443842122e-05,
+      "loss": 0.5357,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9916786491593069,
+      "learning_rate": 1.607664395536317e-05,
+      "loss": 0.5184,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0783830257128157,
+      "learning_rate": 1.6074553051199084e-05,
+      "loss": 0.589,
+      "step": 3857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9327083156351761,
+      "learning_rate": 1.6072461726073856e-05,
+      "loss": 0.6027,
+      "step": 3858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.072289447924501,
+      "learning_rate": 1.6070369980132425e-05,
+      "loss": 0.6064,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8299534240728982,
+      "learning_rate": 1.6068277813519733e-05,
+      "loss": 0.5044,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9522751323222246,
+      "learning_rate": 1.606618522638077e-05,
+      "loss": 0.537,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9331646723727852,
+      "learning_rate": 1.6064092218860553e-05,
+      "loss": 0.5846,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9322832535527018,
+      "learning_rate": 1.6061998791104125e-05,
+      "loss": 0.5319,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8936974228375246,
+      "learning_rate": 1.6059904943256557e-05,
+      "loss": 0.5557,
+      "step": 3864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8527992025507102,
+      "learning_rate": 1.605781067546295e-05,
+      "loss": 0.5803,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.858642960030146,
+      "learning_rate": 1.605571598786844e-05,
+      "loss": 0.5272,
+      "step": 3866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7971785297359876,
+      "learning_rate": 1.605362088061818e-05,
+      "loss": 0.5018,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8986389122506443,
+      "learning_rate": 1.6051525353857364e-05,
+      "loss": 0.5706,
+      "step": 3868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9365303829334433,
+      "learning_rate": 1.604942940773121e-05,
+      "loss": 0.6165,
+      "step": 3869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0017125459415919,
+      "learning_rate": 1.604733304238496e-05,
+      "loss": 0.5739,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9031465462671923,
+      "learning_rate": 1.60452362579639e-05,
+      "loss": 0.6438,
+      "step": 3871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0046701734849914,
+      "learning_rate": 1.6043139054613326e-05,
+      "loss": 0.5714,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9879187604061295,
+      "learning_rate": 1.6041041432478573e-05,
+      "loss": 0.5833,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9316863553413524,
+      "learning_rate": 1.603894339170501e-05,
+      "loss": 0.6159,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.946948808147859,
+      "learning_rate": 1.6036844932438028e-05,
+      "loss": 0.5334,
+      "step": 3875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.997715143378706,
+      "learning_rate": 1.603474605482305e-05,
+      "loss": 0.6068,
+      "step": 3876
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0769994222963795,
+      "learning_rate": 1.6032646759005515e-05,
+      "loss": 0.6482,
+      "step": 3877
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9817520043293302,
+      "learning_rate": 1.6030547045130912e-05,
+      "loss": 0.5853,
+      "step": 3878
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9648721660602143,
+      "learning_rate": 1.6028446913344754e-05,
+      "loss": 0.5616,
+      "step": 3879
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9278261013834327,
+      "learning_rate": 1.6026346363792565e-05,
+      "loss": 0.5881,
+      "step": 3880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.976882951080119,
+      "learning_rate": 1.6024245396619923e-05,
+      "loss": 0.5802,
+      "step": 3881
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.982550166661968,
+      "learning_rate": 1.6022144011972415e-05,
+      "loss": 0.588,
+      "step": 3882
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8778654274821055,
+      "learning_rate": 1.6020042209995674e-05,
+      "loss": 0.5717,
+      "step": 3883
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9761246649548841,
+      "learning_rate": 1.601793999083534e-05,
+      "loss": 0.5979,
+      "step": 3884
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9127386804169086,
+      "learning_rate": 1.601583735463711e-05,
+      "loss": 0.5232,
+      "step": 3885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.943428272432976,
+      "learning_rate": 1.6013734301546682e-05,
+      "loss": 0.596,
+      "step": 3886
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8873528540453057,
+      "learning_rate": 1.6011630831709802e-05,
+      "loss": 0.4994,
+      "step": 3887
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0020297517768617,
+      "learning_rate": 1.6009526945272243e-05,
+      "loss": 0.6594,
+      "step": 3888
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.969974182360959,
+      "learning_rate": 1.600742264237979e-05,
+      "loss": 0.4831,
+      "step": 3889
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8565424953400176,
+      "learning_rate": 1.600531792317828e-05,
+      "loss": 0.5725,
+      "step": 3890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.943954280881943,
+      "learning_rate": 1.6003212787813566e-05,
+      "loss": 0.5465,
+      "step": 3891
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8430972074442779,
+      "learning_rate": 1.6001107236431525e-05,
+      "loss": 0.5275,
+      "step": 3892
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8757053789419513,
+      "learning_rate": 1.5999001269178082e-05,
+      "loss": 0.5166,
+      "step": 3893
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0103937818297317,
+      "learning_rate": 1.5996894886199167e-05,
+      "loss": 0.578,
+      "step": 3894
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9004178001640264,
+      "learning_rate": 1.599478808764076e-05,
+      "loss": 0.6012,
+      "step": 3895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9064418582095709,
+      "learning_rate": 1.5992680873648852e-05,
+      "loss": 0.5688,
+      "step": 3896
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.913659760784911,
+      "learning_rate": 1.5990573244369478e-05,
+      "loss": 0.5659,
+      "step": 3897
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0124956917343046,
+      "learning_rate": 1.5988465199948692e-05,
+      "loss": 0.6485,
+      "step": 3898
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8901834527319934,
+      "learning_rate": 1.5986356740532577e-05,
+      "loss": 0.5982,
+      "step": 3899
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8764696457969312,
+      "learning_rate": 1.5984247866267253e-05,
+      "loss": 0.5469,
+      "step": 3900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9406358564079779,
+      "learning_rate": 1.5982138577298857e-05,
+      "loss": 0.5012,
+      "step": 3901
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8906184319109551,
+      "learning_rate": 1.5980028873773563e-05,
+      "loss": 0.5248,
+      "step": 3902
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9257365052996489,
+      "learning_rate": 1.5977918755837576e-05,
+      "loss": 0.5771,
+      "step": 3903
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8697721075412118,
+      "learning_rate": 1.5975808223637117e-05,
+      "loss": 0.6153,
+      "step": 3904
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9582866753626065,
+      "learning_rate": 1.5973697277318452e-05,
+      "loss": 0.6421,
+      "step": 3905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9999765624853517,
+      "learning_rate": 1.5971585917027864e-05,
+      "loss": 0.6416,
+      "step": 3906
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9084148944426773,
+      "learning_rate": 1.596947414291167e-05,
+      "loss": 0.5444,
+      "step": 3907
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9706443582125923,
+      "learning_rate": 1.5967361955116207e-05,
+      "loss": 0.5201,
+      "step": 3908
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9972933163735651,
+      "learning_rate": 1.596524935378786e-05,
+      "loss": 0.5934,
+      "step": 3909
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8746715492924646,
+      "learning_rate": 1.5963136339073023e-05,
+      "loss": 0.5472,
+      "step": 3910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8610502255050615,
+      "learning_rate": 1.5961022911118124e-05,
+      "loss": 0.6129,
+      "step": 3911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.908968711863703,
+      "learning_rate": 1.5958909070069627e-05,
+      "loss": 0.5596,
+      "step": 3912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8760887786049828,
+      "learning_rate": 1.5956794816074015e-05,
+      "loss": 0.5631,
+      "step": 3913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9159331172803,
+      "learning_rate": 1.5954680149277807e-05,
+      "loss": 0.5787,
+      "step": 3914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8659525657404382,
+      "learning_rate": 1.5952565069827544e-05,
+      "loss": 0.47,
+      "step": 3915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8250608098886822,
+      "learning_rate": 1.5950449577869807e-05,
+      "loss": 0.5444,
+      "step": 3916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8310286356585561,
+      "learning_rate": 1.594833367355119e-05,
+      "loss": 0.4955,
+      "step": 3917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.940905437018659,
+      "learning_rate": 1.5946217357018322e-05,
+      "loss": 0.61,
+      "step": 3918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9313010060308121,
+      "learning_rate": 1.594410062841787e-05,
+      "loss": 0.5724,
+      "step": 3919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9309610664870516,
+      "learning_rate": 1.5941983487896515e-05,
+      "loss": 0.5331,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8901671550634485,
+      "learning_rate": 1.5939865935600976e-05,
+      "loss": 0.5906,
+      "step": 3921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9543630353086977,
+      "learning_rate": 1.5937747971677996e-05,
+      "loss": 0.5637,
+      "step": 3922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8939676646410317,
+      "learning_rate": 1.5935629596274345e-05,
+      "loss": 0.5744,
+      "step": 3923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8888549200288502,
+      "learning_rate": 1.593351080953683e-05,
+      "loss": 0.5727,
+      "step": 3924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8968565334675708,
+      "learning_rate": 1.5931391611612283e-05,
+      "loss": 0.5727,
+      "step": 3925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9970235070283097,
+      "learning_rate": 1.5929272002647554e-05,
+      "loss": 0.6307,
+      "step": 3926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8795770579019754,
+      "learning_rate": 1.5927151982789535e-05,
+      "loss": 0.5426,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8724285002585467,
+      "learning_rate": 1.592503155218514e-05,
+      "loss": 0.5284,
+      "step": 3928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.909354024769601,
+      "learning_rate": 1.592291071098132e-05,
+      "loss": 0.5683,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.004209989415013,
+      "learning_rate": 1.5920789459325034e-05,
+      "loss": 0.563,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9043095354424483,
+      "learning_rate": 1.5918667797363295e-05,
+      "loss": 0.5364,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9581734823349979,
+      "learning_rate": 1.5916545725243124e-05,
+      "loss": 0.5751,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9672287773164598,
+      "learning_rate": 1.5914423243111582e-05,
+      "loss": 0.5979,
+      "step": 3933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9256755293346672,
+      "learning_rate": 1.591230035111576e-05,
+      "loss": 0.5826,
+      "step": 3934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9089934348766084,
+      "learning_rate": 1.5910177049402762e-05,
+      "loss": 0.579,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9235722253896989,
+      "learning_rate": 1.5908053338119743e-05,
+      "loss": 0.5909,
+      "step": 3936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9773177163752066,
+      "learning_rate": 1.590592921741386e-05,
+      "loss": 0.6234,
+      "step": 3937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9177701352719017,
+      "learning_rate": 1.5903804687432325e-05,
+      "loss": 0.5361,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8916835330713223,
+      "learning_rate": 1.5901679748322367e-05,
+      "loss": 0.5138,
+      "step": 3939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0161600179707424,
+      "learning_rate": 1.5899554400231233e-05,
+      "loss": 0.545,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9160973920565795,
+      "learning_rate": 1.5897428643306207e-05,
+      "loss": 0.5696,
+      "step": 3941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9168375963244876,
+      "learning_rate": 1.5895302477694614e-05,
+      "loss": 0.5171,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9596387856101649,
+      "learning_rate": 1.5893175903543788e-05,
+      "loss": 0.557,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8806056535691328,
+      "learning_rate": 1.5891048921001094e-05,
+      "loss": 0.5787,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9184353595815069,
+      "learning_rate": 1.5888921530213938e-05,
+      "loss": 0.5489,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9008051585907707,
+      "learning_rate": 1.5886793731329743e-05,
+      "loss": 0.5742,
+      "step": 3946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9214844580354372,
+      "learning_rate": 1.5884665524495965e-05,
+      "loss": 0.633,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0020293010722414,
+      "learning_rate": 1.5882536909860086e-05,
+      "loss": 0.5728,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8578081842433017,
+      "learning_rate": 1.5880407887569617e-05,
+      "loss": 0.5171,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8587555597349382,
+      "learning_rate": 1.5878278457772095e-05,
+      "loss": 0.5575,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0361111128796203,
+      "learning_rate": 1.5876148620615094e-05,
+      "loss": 0.5906,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8893338118813109,
+      "learning_rate": 1.5874018376246204e-05,
+      "loss": 0.5418,
+      "step": 3952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9951994215390055,
+      "learning_rate": 1.587188772481305e-05,
+      "loss": 0.6434,
+      "step": 3953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8760987314596554,
+      "learning_rate": 1.586975666646328e-05,
+      "loss": 0.5719,
+      "step": 3954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9741708146173176,
+      "learning_rate": 1.586762520134459e-05,
+      "loss": 0.5114,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9053015926217912,
+      "learning_rate": 1.586549332960467e-05,
+      "loss": 0.5896,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0322490493095118,
+      "learning_rate": 1.586336105139127e-05,
+      "loss": 0.6103,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.912987608954962,
+      "learning_rate": 1.5861228366852148e-05,
+      "loss": 0.5358,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.945625272219544,
+      "learning_rate": 1.58590952761351e-05,
+      "loss": 0.547,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0679542684563517,
+      "learning_rate": 1.5856961779387945e-05,
+      "loss": 0.5971,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8642824221120665,
+      "learning_rate": 1.5854827876758535e-05,
+      "loss": 0.5196,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.990179303335577,
+      "learning_rate": 1.5852693568394743e-05,
+      "loss": 0.5959,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8666422528701844,
+      "learning_rate": 1.585055885444448e-05,
+      "loss": 0.4666,
+      "step": 3963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9884825573976187,
+      "learning_rate": 1.584842373505568e-05,
+      "loss": 0.5883,
+      "step": 3964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8598450286175446,
+      "learning_rate": 1.58462882103763e-05,
+      "loss": 0.5026,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8665646932315791,
+      "learning_rate": 1.5844152280554333e-05,
+      "loss": 0.4779,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9650599543426119,
+      "learning_rate": 1.5842015945737798e-05,
+      "loss": 0.6127,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8378359622706862,
+      "learning_rate": 1.583987920607474e-05,
+      "loss": 0.5244,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9662255575979611,
+      "learning_rate": 1.583774206171323e-05,
+      "loss": 0.6374,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8429884326613644,
+      "learning_rate": 1.5835604512801375e-05,
+      "loss": 0.553,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9376564970901118,
+      "learning_rate": 1.5833466559487305e-05,
+      "loss": 0.5976,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9324971037844922,
+      "learning_rate": 1.5831328201919175e-05,
+      "loss": 0.6169,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0128425702126547,
+      "learning_rate": 1.5829189440245175e-05,
+      "loss": 0.6142,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0372645855687352,
+      "learning_rate": 1.5827050274613512e-05,
+      "loss": 0.6401,
+      "step": 3974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9380399027238666,
+      "learning_rate": 1.5824910705172437e-05,
+      "loss": 0.5905,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9415163524530699,
+      "learning_rate": 1.5822770732070222e-05,
+      "loss": 0.5514,
+      "step": 3976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.866600935516856,
+      "learning_rate": 1.5820630355455155e-05,
+      "loss": 0.5763,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9152366905386166,
+      "learning_rate": 1.5818489575475564e-05,
+      "loss": 0.5363,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9422393689041517,
+      "learning_rate": 1.5816348392279814e-05,
+      "loss": 0.6887,
+      "step": 3979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9066783559458047,
+      "learning_rate": 1.5814206806016273e-05,
+      "loss": 0.6047,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9600494430271688,
+      "learning_rate": 1.581206481683336e-05,
+      "loss": 0.6092,
+      "step": 3981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8981869604826622,
+      "learning_rate": 1.580992242487951e-05,
+      "loss": 0.5736,
+      "step": 3982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8496609966141786,
+      "learning_rate": 1.580777963030319e-05,
+      "loss": 0.5108,
+      "step": 3983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8995294057437336,
+      "learning_rate": 1.5805636433252892e-05,
+      "loss": 0.5595,
+      "step": 3984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8649395089955623,
+      "learning_rate": 1.5803492833877143e-05,
+      "loss": 0.5506,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9194681120880075,
+      "learning_rate": 1.5801348832324483e-05,
+      "loss": 0.5092,
+      "step": 3986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9022113242229857,
+      "learning_rate": 1.5799204428743497e-05,
+      "loss": 0.5777,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9521786178423841,
+      "learning_rate": 1.5797059623282787e-05,
+      "loss": 0.5979,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.975208854083481,
+      "learning_rate": 1.5794914416090988e-05,
+      "loss": 0.5444,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0034484704900803,
+      "learning_rate": 1.579276880731676e-05,
+      "loss": 0.6227,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0597278667233871,
+      "learning_rate": 1.579062279710879e-05,
+      "loss": 0.6034,
+      "step": 3991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8784580768653654,
+      "learning_rate": 1.57884763856158e-05,
+      "loss": 0.5405,
+      "step": 3992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9531035237183869,
+      "learning_rate": 1.5786329572986527e-05,
+      "loss": 0.6006,
+      "step": 3993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9197563734914485,
+      "learning_rate": 1.578418235936975e-05,
+      "loss": 0.5324,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9454280511184845,
+      "learning_rate": 1.5782034744914264e-05,
+      "loss": 0.5458,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8474779750455188,
+      "learning_rate": 1.57798867297689e-05,
+      "loss": 0.5221,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9766094107559433,
+      "learning_rate": 1.5777738314082514e-05,
+      "loss": 0.6148,
+      "step": 3997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9217892458525768,
+      "learning_rate": 1.5775589498003984e-05,
+      "loss": 0.5567,
+      "step": 3998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9409235290685977,
+      "learning_rate": 1.5773440281682226e-05,
+      "loss": 0.5966,
+      "step": 3999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.904426711240461,
+      "learning_rate": 1.577129066526618e-05,
+      "loss": 0.553,
+      "step": 4000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9370662463188123,
+      "learning_rate": 1.5769140648904806e-05,
+      "loss": 0.5454,
+      "step": 4001
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8846935856523952,
+      "learning_rate": 1.5766990232747106e-05,
+      "loss": 0.5365,
+      "step": 4002
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9434664117402615,
+      "learning_rate": 1.5764839416942097e-05,
+      "loss": 0.5992,
+      "step": 4003
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3096888551499557,
+      "learning_rate": 1.576268820163883e-05,
+      "loss": 0.521,
+      "step": 4004
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9698155276466377,
+      "learning_rate": 1.576053658698638e-05,
+      "loss": 0.6299,
+      "step": 4005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9371335851184125,
+      "learning_rate": 1.5758384573133857e-05,
+      "loss": 0.5855,
+      "step": 4006
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0166311336574105,
+      "learning_rate": 1.5756232160230388e-05,
+      "loss": 0.6351,
+      "step": 4007
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9017351451582668,
+      "learning_rate": 1.5754079348425137e-05,
+      "loss": 0.5309,
+      "step": 4008
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0872595226025266,
+      "learning_rate": 1.575192613786729e-05,
+      "loss": 0.5589,
+      "step": 4009
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9429036522959239,
+      "learning_rate": 1.574977252870607e-05,
+      "loss": 0.5597,
+      "step": 4010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.006473815437657,
+      "learning_rate": 1.5747618521090706e-05,
+      "loss": 0.6188,
+      "step": 4011
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9831280828790899,
+      "learning_rate": 1.574546411517048e-05,
+      "loss": 0.5073,
+      "step": 4012
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9614882582761742,
+      "learning_rate": 1.5743309311094687e-05,
+      "loss": 0.5916,
+      "step": 4013
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9834440770393784,
+      "learning_rate": 1.574115410901265e-05,
+      "loss": 0.5796,
+      "step": 4014
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.862865878132095,
+      "learning_rate": 1.573899850907373e-05,
+      "loss": 0.5184,
+      "step": 4015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9369647382877944,
+      "learning_rate": 1.5736842511427302e-05,
+      "loss": 0.5914,
+      "step": 4016
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8918627312384875,
+      "learning_rate": 1.5734686116222775e-05,
+      "loss": 0.502,
+      "step": 4017
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.976668862121526,
+      "learning_rate": 1.573252932360959e-05,
+      "loss": 0.5887,
+      "step": 4018
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8604324291777963,
+      "learning_rate": 1.5730372133737206e-05,
+      "loss": 0.5676,
+      "step": 4019
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9355398542457567,
+      "learning_rate": 1.5728214546755117e-05,
+      "loss": 0.5784,
+      "step": 4020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0053396355524888,
+      "learning_rate": 1.572605656281284e-05,
+      "loss": 0.5561,
+      "step": 4021
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8865428977414068,
+      "learning_rate": 1.572389818205992e-05,
+      "loss": 0.5344,
+      "step": 4022
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.917616140853694,
+      "learning_rate": 1.5721739404645937e-05,
+      "loss": 0.5448,
+      "step": 4023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0039270053413536,
+      "learning_rate": 1.5719580230720485e-05,
+      "loss": 0.5648,
+      "step": 4024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8715238705686509,
+      "learning_rate": 1.57174206604332e-05,
+      "loss": 0.6044,
+      "step": 4025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9609013329719558,
+      "learning_rate": 1.571526069393373e-05,
+      "loss": 0.5953,
+      "step": 4026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9475604172383607,
+      "learning_rate": 1.5713100331371768e-05,
+      "loss": 0.5605,
+      "step": 4027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8078213940067273,
+      "learning_rate": 1.5710939572897018e-05,
+      "loss": 0.5468,
+      "step": 4028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9691697196000543,
+      "learning_rate": 1.570877841865922e-05,
+      "loss": 0.5711,
+      "step": 4029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.005924058700349,
+      "learning_rate": 1.5706616868808142e-05,
+      "loss": 0.5627,
+      "step": 4030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8603428183105295,
+      "learning_rate": 1.5704454923493577e-05,
+      "loss": 0.5344,
+      "step": 4031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0494686065684906,
+      "learning_rate": 1.570229258286535e-05,
+      "loss": 0.5895,
+      "step": 4032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8563575513133033,
+      "learning_rate": 1.5700129847073298e-05,
+      "loss": 0.5631,
+      "step": 4033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0982265801325355,
+      "learning_rate": 1.5697966716267308e-05,
+      "loss": 0.6157,
+      "step": 4034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8426234383113499,
+      "learning_rate": 1.5695803190597275e-05,
+      "loss": 0.5765,
+      "step": 4035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9077511704024396,
+      "learning_rate": 1.5693639270213138e-05,
+      "loss": 0.5276,
+      "step": 4036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8654428007933888,
+      "learning_rate": 1.5691474955264848e-05,
+      "loss": 0.4938,
+      "step": 4037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9792980677325999,
+      "learning_rate": 1.568931024590239e-05,
+      "loss": 0.5882,
+      "step": 4038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0353408848047134,
+      "learning_rate": 1.568714514227578e-05,
+      "loss": 0.6413,
+      "step": 4039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8556489487154524,
+      "learning_rate": 1.5684979644535053e-05,
+      "loss": 0.5847,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9468726518340129,
+      "learning_rate": 1.5682813752830284e-05,
+      "loss": 0.5498,
+      "step": 4041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9710687607873638,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.5876,
+      "step": 4042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8165640547630525,
+      "learning_rate": 1.5678480788129003e-05,
+      "loss": 0.5817,
+      "step": 4043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8358262932296227,
+      "learning_rate": 1.567631371543277e-05,
+      "loss": 0.5138,
+      "step": 4044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8043492548198713,
+      "learning_rate": 1.5674146249373027e-05,
+      "loss": 0.5326,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8762669905406324,
+      "learning_rate": 1.5671978390099985e-05,
+      "loss": 0.5032,
+      "step": 4046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.962549549006086,
+      "learning_rate": 1.566981013776387e-05,
+      "loss": 0.5345,
+      "step": 4047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8584925617245548,
+      "learning_rate": 1.5667641492514942e-05,
+      "loss": 0.4845,
+      "step": 4048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9768208555080496,
+      "learning_rate": 1.5665472454503484e-05,
+      "loss": 0.6085,
+      "step": 4049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8910712523051837,
+      "learning_rate": 1.566330302387981e-05,
+      "loss": 0.5695,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8993957926471542,
+      "learning_rate": 1.566113320079426e-05,
+      "loss": 0.5671,
+      "step": 4051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9509042340172726,
+      "learning_rate": 1.5658962985397202e-05,
+      "loss": 0.5668,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8274785334369242,
+      "learning_rate": 1.565679237783903e-05,
+      "loss": 0.5048,
+      "step": 4053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8602365961998485,
+      "learning_rate": 1.565462137827016e-05,
+      "loss": 0.5589,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9312739718943414,
+      "learning_rate": 1.5652449986841048e-05,
+      "loss": 0.574,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8972888462688227,
+      "learning_rate": 1.5650278203702162e-05,
+      "loss": 0.5738,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9849922783424293,
+      "learning_rate": 1.564810602900401e-05,
+      "loss": 0.6366,
+      "step": 4057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9266568698973598,
+      "learning_rate": 1.564593346289712e-05,
+      "loss": 0.5807,
+      "step": 4058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9689728059389319,
+      "learning_rate": 1.564376050553205e-05,
+      "loss": 0.6002,
+      "step": 4059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9657283266055987,
+      "learning_rate": 1.564158715705938e-05,
+      "loss": 0.5452,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9252710654263362,
+      "learning_rate": 1.563941341762973e-05,
+      "loss": 0.6325,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9067245498427078,
+      "learning_rate": 1.5637239287393725e-05,
+      "loss": 0.601,
+      "step": 4062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9155849496185144,
+      "learning_rate": 1.5635064766502042e-05,
+      "loss": 0.5615,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8936140240745216,
+      "learning_rate": 1.563288985510537e-05,
+      "loss": 0.5471,
+      "step": 4064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.942245882106811,
+      "learning_rate": 1.5630714553354425e-05,
+      "loss": 0.5919,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9363613460777342,
+      "learning_rate": 1.5628538861399956e-05,
+      "loss": 0.5929,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0024484438838643,
+      "learning_rate": 1.5626362779392738e-05,
+      "loss": 0.6062,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8589020510416536,
+      "learning_rate": 1.562418630748357e-05,
+      "loss": 0.5201,
+      "step": 4068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.859399789929067,
+      "learning_rate": 1.5622009445823274e-05,
+      "loss": 0.5157,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9655223138192177,
+      "learning_rate": 1.5619832194562716e-05,
+      "loss": 0.5716,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9628342084902699,
+      "learning_rate": 1.561765455385277e-05,
+      "loss": 0.5767,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9572679965465991,
+      "learning_rate": 1.5615476523844346e-05,
+      "loss": 0.5719,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8314228385114546,
+      "learning_rate": 1.5613298104688383e-05,
+      "loss": 0.4915,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8856555588832423,
+      "learning_rate": 1.5611119296535836e-05,
+      "loss": 0.4953,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8825697950973398,
+      "learning_rate": 1.56089400995377e-05,
+      "loss": 0.5186,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0123348773261922,
+      "learning_rate": 1.560676051384499e-05,
+      "loss": 0.6382,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9949471643486758,
+      "learning_rate": 1.560458053960875e-05,
+      "loss": 0.5926,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9208384971441007,
+      "learning_rate": 1.560240017698005e-05,
+      "loss": 0.5815,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0898988852025966,
+      "learning_rate": 1.5600219426109986e-05,
+      "loss": 0.6395,
+      "step": 4079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9118273390877075,
+      "learning_rate": 1.5598038287149684e-05,
+      "loss": 0.6092,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8525027875549489,
+      "learning_rate": 1.5595856760250296e-05,
+      "loss": 0.5162,
+      "step": 4081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9278775945269585,
+      "learning_rate": 1.5593674845562994e-05,
+      "loss": 0.5954,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.330704959252747,
+      "learning_rate": 1.559149254323899e-05,
+      "loss": 0.6494,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9244176454658858,
+      "learning_rate": 1.558930985342951e-05,
+      "loss": 0.5549,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9455312227753272,
+      "learning_rate": 1.5587126776285818e-05,
+      "loss": 0.6124,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9137367270037585,
+      "learning_rate": 1.5584943311959197e-05,
+      "loss": 0.5225,
+      "step": 4086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9787600920662352,
+      "learning_rate": 1.5582759460600952e-05,
+      "loss": 0.5767,
+      "step": 4087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8621461831215322,
+      "learning_rate": 1.5580575222362435e-05,
+      "loss": 0.5406,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8804137554755774,
+      "learning_rate": 1.5578390597395e-05,
+      "loss": 0.5362,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8723709928624246,
+      "learning_rate": 1.5576205585850052e-05,
+      "loss": 0.4771,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9936210472826803,
+      "learning_rate": 1.5574020187878994e-05,
+      "loss": 0.603,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0090338812951476,
+      "learning_rate": 1.557183440363329e-05,
+      "loss": 0.6171,
+      "step": 4092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8932855373353723,
+      "learning_rate": 1.5569648233264395e-05,
+      "loss": 0.5652,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8737280526971734,
+      "learning_rate": 1.556746167692382e-05,
+      "loss": 0.5322,
+      "step": 4094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9708220775517294,
+      "learning_rate": 1.5565274734763094e-05,
+      "loss": 0.5867,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9697497442178686,
+      "learning_rate": 1.5563087406933762e-05,
+      "loss": 0.5511,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9903770214845385,
+      "learning_rate": 1.5560899693587405e-05,
+      "loss": 0.6059,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9203756944219564,
+      "learning_rate": 1.5558711594875634e-05,
+      "loss": 0.4967,
+      "step": 4098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.942205169285279,
+      "learning_rate": 1.555652311095008e-05,
+      "loss": 0.5684,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9600421552335939,
+      "learning_rate": 1.5554334241962403e-05,
+      "loss": 0.5922,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9174980936365247,
+      "learning_rate": 1.5552144988064292e-05,
+      "loss": 0.5991,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8450449715915833,
+      "learning_rate": 1.5549955349407456e-05,
+      "loss": 0.5217,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8690691429197966,
+      "learning_rate": 1.5547765326143634e-05,
+      "loss": 0.5026,
+      "step": 4103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0294790378362721,
+      "learning_rate": 1.5545574918424602e-05,
+      "loss": 0.6469,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8553897917778106,
+      "learning_rate": 1.5543384126402144e-05,
+      "loss": 0.5178,
+      "step": 4105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.963476645302054,
+      "learning_rate": 1.554119295022808e-05,
+      "loss": 0.5804,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8640671675579383,
+      "learning_rate": 1.5539001390054265e-05,
+      "loss": 0.5261,
+      "step": 4107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9682608000185056,
+      "learning_rate": 1.5536809446032562e-05,
+      "loss": 0.6721,
+      "step": 4108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8860882499843543,
+      "learning_rate": 1.5534617118314882e-05,
+      "loss": 0.5743,
+      "step": 4109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9069502189222868,
+      "learning_rate": 1.553242440705314e-05,
+      "loss": 0.5516,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8936217039676828,
+      "learning_rate": 1.5530231312399294e-05,
+      "loss": 0.5435,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9706173889045864,
+      "learning_rate": 1.5528037834505322e-05,
+      "loss": 0.5701,
+      "step": 4112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9178913160396679,
+      "learning_rate": 1.5525843973523237e-05,
+      "loss": 0.5524,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8145499993607632,
+      "learning_rate": 1.552364972960506e-05,
+      "loss": 0.5285,
+      "step": 4114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.016269762503253,
+      "learning_rate": 1.552145510290286e-05,
+      "loss": 0.5321,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.908990601168826,
+      "learning_rate": 1.5519260093568717e-05,
+      "loss": 0.5547,
+      "step": 4116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9328424886518241,
+      "learning_rate": 1.5517064701754744e-05,
+      "loss": 0.6041,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9016063684659464,
+      "learning_rate": 1.5514868927613084e-05,
+      "loss": 0.5758,
+      "step": 4118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8420443749720986,
+      "learning_rate": 1.5512672771295898e-05,
+      "loss": 0.5029,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9588642379929366,
+      "learning_rate": 1.5510476232955376e-05,
+      "loss": 0.588,
+      "step": 4120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9003302000099962,
+      "learning_rate": 1.5508279312743742e-05,
+      "loss": 0.5722,
+      "step": 4121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.877570962504461,
+      "learning_rate": 1.5506082010813237e-05,
+      "loss": 0.4978,
+      "step": 4122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9989799140340428,
+      "learning_rate": 1.550388432731613e-05,
+      "loss": 0.5816,
+      "step": 4123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8644779147210184,
+      "learning_rate": 1.550168626240472e-05,
+      "loss": 0.5078,
+      "step": 4124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9713792974934381,
+      "learning_rate": 1.549948781623134e-05,
+      "loss": 0.6444,
+      "step": 4125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9157006453630713,
+      "learning_rate": 1.5497288988948326e-05,
+      "loss": 0.5919,
+      "step": 4126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9740797882912834,
+      "learning_rate": 1.5495089780708062e-05,
+      "loss": 0.6244,
+      "step": 4127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9011107822348161,
+      "learning_rate": 1.5492890191662954e-05,
+      "loss": 0.6247,
+      "step": 4128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0004936320232625,
+      "learning_rate": 1.5490690221965424e-05,
+      "loss": 0.5899,
+      "step": 4129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9769522662893291,
+      "learning_rate": 1.5488489871767928e-05,
+      "loss": 0.5603,
+      "step": 4130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9531619489136708,
+      "learning_rate": 1.5486289141222955e-05,
+      "loss": 0.546,
+      "step": 4131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8588746366213443,
+      "learning_rate": 1.5484088030483015e-05,
+      "loss": 0.548,
+      "step": 4132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9155465841984228,
+      "learning_rate": 1.5481886539700636e-05,
+      "loss": 0.4894,
+      "step": 4133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.924147240967755,
+      "learning_rate": 1.5479684669028384e-05,
+      "loss": 0.5814,
+      "step": 4134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8825588133700512,
+      "learning_rate": 1.5477482418618844e-05,
+      "loss": 0.625,
+      "step": 4135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0442236829760536,
+      "learning_rate": 1.547527978862463e-05,
+      "loss": 0.6079,
+      "step": 4136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9227304530173703,
+      "learning_rate": 1.5473076779198385e-05,
+      "loss": 0.5721,
+      "step": 4137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0268902487716691,
+      "learning_rate": 1.547087339049277e-05,
+      "loss": 0.6532,
+      "step": 4138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9282329564893024,
+      "learning_rate": 1.5468669622660487e-05,
+      "loss": 0.5388,
+      "step": 4139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9878857633469456,
+      "learning_rate": 1.5466465475854246e-05,
+      "loss": 0.5986,
+      "step": 4140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0095349326177143,
+      "learning_rate": 1.54642609502268e-05,
+      "loss": 0.5704,
+      "step": 4141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8335990919695064,
+      "learning_rate": 1.5462056045930912e-05,
+      "loss": 0.5412,
+      "step": 4142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9715117775274271,
+      "learning_rate": 1.5459850763119386e-05,
+      "loss": 0.5278,
+      "step": 4143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9114149090614846,
+      "learning_rate": 1.5457645101945046e-05,
+      "loss": 0.5223,
+      "step": 4144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.858592720845518,
+      "learning_rate": 1.5455439062560743e-05,
+      "loss": 0.5791,
+      "step": 4145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8942255917626747,
+      "learning_rate": 1.5453232645119348e-05,
+      "loss": 0.5552,
+      "step": 4146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8826034287798566,
+      "learning_rate": 1.5451025849773773e-05,
+      "loss": 0.5712,
+      "step": 4147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9978211602085963,
+      "learning_rate": 1.544881867667694e-05,
+      "loss": 0.587,
+      "step": 4148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9534289695842948,
+      "learning_rate": 1.5446611125981804e-05,
+      "loss": 0.5829,
+      "step": 4149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9171045438073293,
+      "learning_rate": 1.5444403197841345e-05,
+      "loss": 0.6322,
+      "step": 4150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9103267468648931,
+      "learning_rate": 1.5442194892408583e-05,
+      "loss": 0.5276,
+      "step": 4151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9318835958199996,
+      "learning_rate": 1.5439986209836532e-05,
+      "loss": 0.5889,
+      "step": 4152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9935823989772233,
+      "learning_rate": 1.5437777150278268e-05,
+      "loss": 0.5639,
+      "step": 4153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0297272006608487,
+      "learning_rate": 1.543556771388687e-05,
+      "loss": 0.6397,
+      "step": 4154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9775512865777397,
+      "learning_rate": 1.543335790081545e-05,
+      "loss": 0.5728,
+      "step": 4155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8264711311001698,
+      "learning_rate": 1.5431147711217147e-05,
+      "loss": 0.5647,
+      "step": 4156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7969675893425409,
+      "learning_rate": 1.5428937145245126e-05,
+      "loss": 0.5264,
+      "step": 4157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9149958151946387,
+      "learning_rate": 1.542672620305257e-05,
+      "loss": 0.5749,
+      "step": 4158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8931961491044399,
+      "learning_rate": 1.542451488479271e-05,
+      "loss": 0.5328,
+      "step": 4159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8109905646646205,
+      "learning_rate": 1.5422303190618776e-05,
+      "loss": 0.5047,
+      "step": 4160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9536185768596526,
+      "learning_rate": 1.5420091120684042e-05,
+      "loss": 0.5677,
+      "step": 4161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9340327406436755,
+      "learning_rate": 1.54178786751418e-05,
+      "loss": 0.5517,
+      "step": 4162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8801478055544735,
+      "learning_rate": 1.541566585414537e-05,
+      "loss": 0.5112,
+      "step": 4163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9203855454083592,
+      "learning_rate": 1.5413452657848104e-05,
+      "loss": 0.5836,
+      "step": 4164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9008530027903892,
+      "learning_rate": 1.5411239086403367e-05,
+      "loss": 0.5972,
+      "step": 4165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0356456643420995,
+      "learning_rate": 1.540902513996456e-05,
+      "loss": 0.597,
+      "step": 4166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8433055793215191,
+      "learning_rate": 1.5406810818685113e-05,
+      "loss": 0.5037,
+      "step": 4167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0015286153463043,
+      "learning_rate": 1.5404596122718473e-05,
+      "loss": 0.581,
+      "step": 4168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9712135995909356,
+      "learning_rate": 1.540238105221811e-05,
+      "loss": 0.5599,
+      "step": 4169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9707839063979106,
+      "learning_rate": 1.5400165607337534e-05,
+      "loss": 0.5304,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9938465618035094,
+      "learning_rate": 1.5397949788230275e-05,
+      "loss": 0.575,
+      "step": 4171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1246282226906894,
+      "learning_rate": 1.539573359504988e-05,
+      "loss": 0.5929,
+      "step": 4172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.952586657296879,
+      "learning_rate": 1.539351702794993e-05,
+      "loss": 0.5333,
+      "step": 4173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.976592921038152,
+      "learning_rate": 1.539130008708404e-05,
+      "loss": 0.6206,
+      "step": 4174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8593233700589394,
+      "learning_rate": 1.538908277260583e-05,
+      "loss": 0.5927,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9321633186484158,
+      "learning_rate": 1.538686508466897e-05,
+      "loss": 0.6076,
+      "step": 4176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9736653930607452,
+      "learning_rate": 1.5384647023427136e-05,
+      "loss": 0.6351,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0073840250099277,
+      "learning_rate": 1.538242858903404e-05,
+      "loss": 0.5281,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9670824565967566,
+      "learning_rate": 1.538020978164341e-05,
+      "loss": 0.5519,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9003903583858006,
+      "learning_rate": 1.5377990601409022e-05,
+      "loss": 0.5621,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.07375811358448,
+      "learning_rate": 1.5375771048484657e-05,
+      "loss": 0.5801,
+      "step": 4181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9829663735472544,
+      "learning_rate": 1.5373551123024123e-05,
+      "loss": 0.6094,
+      "step": 4182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8686452093382149,
+      "learning_rate": 1.537133082518126e-05,
+      "loss": 0.5506,
+      "step": 4183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9804770241218727,
+      "learning_rate": 1.536911015510994e-05,
+      "loss": 0.5503,
+      "step": 4184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8689620504507469,
+      "learning_rate": 1.5366889112964044e-05,
+      "loss": 0.5606,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.880493881682588,
+      "learning_rate": 1.5364667698897498e-05,
+      "loss": 0.5309,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9549420150814182,
+      "learning_rate": 1.5362445913064238e-05,
+      "loss": 0.5554,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9255714148261335,
+      "learning_rate": 1.536022375561823e-05,
+      "loss": 0.5805,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9160529514074018,
+      "learning_rate": 1.535800122671347e-05,
+      "loss": 0.5463,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9220363895496618,
+      "learning_rate": 1.535577832650398e-05,
+      "loss": 0.4973,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8529340143381954,
+      "learning_rate": 1.53535550551438e-05,
+      "loss": 0.5442,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9997574057268795,
+      "learning_rate": 1.5351331412787004e-05,
+      "loss": 0.6065,
+      "step": 4192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9291109892458387,
+      "learning_rate": 1.534910739958769e-05,
+      "loss": 0.5558,
+      "step": 4193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9131668497103859,
+      "learning_rate": 1.5346883015699976e-05,
+      "loss": 0.5957,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9013737322445415,
+      "learning_rate": 1.5344658261278013e-05,
+      "loss": 0.6395,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8451928896140083,
+      "learning_rate": 1.5342433136475972e-05,
+      "loss": 0.5138,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9204352581964403,
+      "learning_rate": 1.5340207641448054e-05,
+      "loss": 0.5535,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9285526995779052,
+      "learning_rate": 1.5337981776348484e-05,
+      "loss": 0.52,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.96092597621691,
+      "learning_rate": 1.533575554133151e-05,
+      "loss": 0.5847,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8841256267260578,
+      "learning_rate": 1.533352893655141e-05,
+      "loss": 0.565,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9820667038604786,
+      "learning_rate": 1.5331301962162485e-05,
+      "loss": 0.6118,
+      "step": 4201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.90455636810397,
+      "learning_rate": 1.5329074618319063e-05,
+      "loss": 0.5453,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0119288125177617,
+      "learning_rate": 1.5326846905175497e-05,
+      "loss": 0.5862,
+      "step": 4203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9461879744824403,
+      "learning_rate": 1.5324618822886167e-05,
+      "loss": 0.6043,
+      "step": 4204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.939134668597495,
+      "learning_rate": 1.5322390371605473e-05,
+      "loss": 0.577,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9843968714772094,
+      "learning_rate": 1.532016155148785e-05,
+      "loss": 0.5525,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9416378125162804,
+      "learning_rate": 1.531793236268775e-05,
+      "loss": 0.561,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9043162067384275,
+      "learning_rate": 1.531570280535965e-05,
+      "loss": 0.4951,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.953135119133098,
+      "learning_rate": 1.5313472879658066e-05,
+      "loss": 0.5997,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9904782885873062,
+      "learning_rate": 1.531124258573752e-05,
+      "loss": 0.5398,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9141887774274177,
+      "learning_rate": 1.530901192375258e-05,
+      "loss": 0.5488,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9307386669857576,
+      "learning_rate": 1.530678089385782e-05,
+      "loss": 0.5841,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9567495658547867,
+      "learning_rate": 1.5304549496207848e-05,
+      "loss": 0.5479,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9676728012807723,
+      "learning_rate": 1.5302317730957305e-05,
+      "loss": 0.5915,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9185599117650312,
+      "learning_rate": 1.5300085598260843e-05,
+      "loss": 0.5809,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8203698746978695,
+      "learning_rate": 1.5297853098273148e-05,
+      "loss": 0.5176,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9929516274827571,
+      "learning_rate": 1.529562023114894e-05,
+      "loss": 0.5683,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9217222912479962,
+      "learning_rate": 1.5293386997042943e-05,
+      "loss": 0.5301,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8681487591067317,
+      "learning_rate": 1.5291153396109925e-05,
+      "loss": 0.5411,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9882312511442278,
+      "learning_rate": 1.5288919428504668e-05,
+      "loss": 0.6358,
+      "step": 4220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9066244268708632,
+      "learning_rate": 1.5286685094381984e-05,
+      "loss": 0.5119,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9038436185292014,
+      "learning_rate": 1.5284450393896713e-05,
+      "loss": 0.5951,
+      "step": 4222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9642323326452018,
+      "learning_rate": 1.528221532720372e-05,
+      "loss": 0.546,
+      "step": 4223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9956160521476128,
+      "learning_rate": 1.5279979894457887e-05,
+      "loss": 0.6047,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8354483533906525,
+      "learning_rate": 1.5277744095814132e-05,
+      "loss": 0.5119,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9604418200300491,
+      "learning_rate": 1.5275507931427392e-05,
+      "loss": 0.5881,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9005288726215351,
+      "learning_rate": 1.5273271401452633e-05,
+      "loss": 0.5507,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9390955908421621,
+      "learning_rate": 1.5271034506044838e-05,
+      "loss": 0.5362,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.004379660532751,
+      "learning_rate": 1.5268797245359035e-05,
+      "loss": 0.5816,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.988809361388353,
+      "learning_rate": 1.5266559619550254e-05,
+      "loss": 0.5677,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8965461566573829,
+      "learning_rate": 1.526432162877356e-05,
+      "loss": 0.5972,
+      "step": 4231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8899439036923005,
+      "learning_rate": 1.526208327318405e-05,
+      "loss": 0.5549,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8831555105843113,
+      "learning_rate": 1.5259844552936833e-05,
+      "loss": 0.5199,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9053002337578933,
+      "learning_rate": 1.5257605468187056e-05,
+      "loss": 0.5575,
+      "step": 4234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.076438001507276,
+      "learning_rate": 1.5255366019089883e-05,
+      "loss": 0.4907,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9969810516684846,
+      "learning_rate": 1.5253126205800506e-05,
+      "loss": 0.6066,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9872274573817507,
+      "learning_rate": 1.525088602847414e-05,
+      "loss": 0.5654,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8572405337777793,
+      "learning_rate": 1.5248645487266036e-05,
+      "loss": 0.5443,
+      "step": 4238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8459905204744106,
+      "learning_rate": 1.5246404582331451e-05,
+      "loss": 0.5223,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8089099757270066,
+      "learning_rate": 1.5244163313825684e-05,
+      "loss": 0.5629,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8868275652220692,
+      "learning_rate": 1.524192168190405e-05,
+      "loss": 0.4984,
+      "step": 4241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9299705286995178,
+      "learning_rate": 1.5239679686721892e-05,
+      "loss": 0.5544,
+      "step": 4242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.032961198809958,
+      "learning_rate": 1.5237437328434581e-05,
+      "loss": 0.6112,
+      "step": 4243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8999846365877913,
+      "learning_rate": 1.5235194607197508e-05,
+      "loss": 0.545,
+      "step": 4244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9422703636135591,
+      "learning_rate": 1.5232951523166092e-05,
+      "loss": 0.5203,
+      "step": 4245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8820713677107922,
+      "learning_rate": 1.5230708076495777e-05,
+      "loss": 0.5232,
+      "step": 4246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9241143950002292,
+      "learning_rate": 1.5228464267342036e-05,
+      "loss": 0.5773,
+      "step": 4247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9115901855811518,
+      "learning_rate": 1.5226220095860353e-05,
+      "loss": 0.5443,
+      "step": 4248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9692553383520067,
+      "learning_rate": 1.5223975562206255e-05,
+      "loss": 0.5583,
+      "step": 4249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0601455956061403,
+      "learning_rate": 1.5221730666535285e-05,
+      "loss": 0.6298,
+      "step": 4250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7742548362033347,
+      "learning_rate": 1.5219485409003013e-05,
+      "loss": 0.5086,
+      "step": 4251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0254340546298504,
+      "learning_rate": 1.5217239789765028e-05,
+      "loss": 0.5832,
+      "step": 4252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8661718359978385,
+      "learning_rate": 1.5214993808976956e-05,
+      "loss": 0.5454,
+      "step": 4253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9429372489488378,
+      "learning_rate": 1.5212747466794437e-05,
+      "loss": 0.5834,
+      "step": 4254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9280200691809544,
+      "learning_rate": 1.5210500763373142e-05,
+      "loss": 0.5302,
+      "step": 4255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9484388180375084,
+      "learning_rate": 1.5208253698868766e-05,
+      "loss": 0.6313,
+      "step": 4256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9170638988041517,
+      "learning_rate": 1.5206006273437031e-05,
+      "loss": 0.6159,
+      "step": 4257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0063274313357442,
+      "learning_rate": 1.5203758487233677e-05,
+      "loss": 0.5816,
+      "step": 4258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8113862998367966,
+      "learning_rate": 1.5201510340414473e-05,
+      "loss": 0.5125,
+      "step": 4259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.958709181326403,
+      "learning_rate": 1.5199261833135222e-05,
+      "loss": 0.5564,
+      "step": 4260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8211454039458304,
+      "learning_rate": 1.5197012965551733e-05,
+      "loss": 0.5896,
+      "step": 4261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8304508772583704,
+      "learning_rate": 1.5194763737819856e-05,
+      "loss": 0.5553,
+      "step": 4262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8899457437890497,
+      "learning_rate": 1.519251415009546e-05,
+      "loss": 0.5524,
+      "step": 4263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9641008384200902,
+      "learning_rate": 1.5190264202534442e-05,
+      "loss": 0.5872,
+      "step": 4264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0481063603542773,
+      "learning_rate": 1.5188013895292715e-05,
+      "loss": 0.6124,
+      "step": 4265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0437724166588191,
+      "learning_rate": 1.5185763228526226e-05,
+      "loss": 0.6339,
+      "step": 4266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8858211683890729,
+      "learning_rate": 1.5183512202390951e-05,
+      "loss": 0.5898,
+      "step": 4267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8905028036724305,
+      "learning_rate": 1.518126081704287e-05,
+      "loss": 0.5172,
+      "step": 4268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8816742044266888,
+      "learning_rate": 1.5179009072638016e-05,
+      "loss": 0.5368,
+      "step": 4269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8944448914022917,
+      "learning_rate": 1.5176756969332428e-05,
+      "loss": 0.554,
+      "step": 4270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8512773616826556,
+      "learning_rate": 1.5174504507282168e-05,
+      "loss": 0.485,
+      "step": 4271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9369207232240409,
+      "learning_rate": 1.517225168664334e-05,
+      "loss": 0.5814,
+      "step": 4272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8926504243414912,
+      "learning_rate": 1.5169998507572057e-05,
+      "loss": 0.5155,
+      "step": 4273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9501844007981542,
+      "learning_rate": 1.5167744970224463e-05,
+      "loss": 0.5562,
+      "step": 4274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8999480245495015,
+      "learning_rate": 1.5165491074756723e-05,
+      "loss": 0.5411,
+      "step": 4275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8527653264746171,
+      "learning_rate": 1.5163236821325037e-05,
+      "loss": 0.5883,
+      "step": 4276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8691186223661405,
+      "learning_rate": 1.5160982210085621e-05,
+      "loss": 0.5269,
+      "step": 4277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.933540473664518,
+      "learning_rate": 1.515872724119471e-05,
+      "loss": 0.5769,
+      "step": 4278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9157217785033557,
+      "learning_rate": 1.5156471914808582e-05,
+      "loss": 0.542,
+      "step": 4279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9699683221638261,
+      "learning_rate": 1.5154216231083522e-05,
+      "loss": 0.5454,
+      "step": 4280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8620056534267265,
+      "learning_rate": 1.515196019017585e-05,
+      "loss": 0.4944,
+      "step": 4281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8463023180790658,
+      "learning_rate": 1.5149703792241903e-05,
+      "loss": 0.509,
+      "step": 4282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7887007205212211,
+      "learning_rate": 1.5147447037438055e-05,
+      "loss": 0.4653,
+      "step": 4283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.896276367645115,
+      "learning_rate": 1.5145189925920694e-05,
+      "loss": 0.5098,
+      "step": 4284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9270094024131633,
+      "learning_rate": 1.514293245784623e-05,
+      "loss": 0.5909,
+      "step": 4285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9105702606018491,
+      "learning_rate": 1.514067463337111e-05,
+      "loss": 0.5257,
+      "step": 4286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9046292918014648,
+      "learning_rate": 1.5138416452651803e-05,
+      "loss": 0.5459,
+      "step": 4287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8185620723801201,
+      "learning_rate": 1.5136157915844787e-05,
+      "loss": 0.4844,
+      "step": 4288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9212215102695174,
+      "learning_rate": 1.5133899023106584e-05,
+      "loss": 0.551,
+      "step": 4289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8689911637720217,
+      "learning_rate": 1.5131639774593737e-05,
+      "loss": 0.535,
+      "step": 4290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9015688037502202,
+      "learning_rate": 1.5129380170462802e-05,
+      "loss": 0.5815,
+      "step": 4291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8854452628551268,
+      "learning_rate": 1.512712021087037e-05,
+      "loss": 0.5552,
+      "step": 4292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8375236940962522,
+      "learning_rate": 1.5124859895973058e-05,
+      "loss": 0.5811,
+      "step": 4293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0396346487059804,
+      "learning_rate": 1.51225992259275e-05,
+      "loss": 0.548,
+      "step": 4294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9574274447907284,
+      "learning_rate": 1.5120338200890356e-05,
+      "loss": 0.5502,
+      "step": 4295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.86965100414926,
+      "learning_rate": 1.5118076821018322e-05,
+      "loss": 0.5442,
+      "step": 4296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8883625419684474,
+      "learning_rate": 1.5115815086468103e-05,
+      "loss": 0.5568,
+      "step": 4297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8950422831418536,
+      "learning_rate": 1.511355299739643e-05,
+      "loss": 0.5023,
+      "step": 4298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0147768322049326,
+      "learning_rate": 1.511129055396008e-05,
+      "loss": 0.5961,
+      "step": 4299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.905224107471028,
+      "learning_rate": 1.510902775631582e-05,
+      "loss": 0.542,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9008359298426322,
+      "learning_rate": 1.510676460462047e-05,
+      "loss": 0.5406,
+      "step": 4301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0409019164334707,
+      "learning_rate": 1.5104501099030864e-05,
+      "loss": 0.6519,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.92608839006507,
+      "learning_rate": 1.5102237239703858e-05,
+      "loss": 0.4852,
+      "step": 4303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8558223846325838,
+      "learning_rate": 1.509997302679634e-05,
+      "loss": 0.545,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9071792434994643,
+      "learning_rate": 1.5097708460465214e-05,
+      "loss": 0.5647,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8243596727202004,
+      "learning_rate": 1.5095443540867412e-05,
+      "loss": 0.4917,
+      "step": 4306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8528546266985911,
+      "learning_rate": 1.5093178268159892e-05,
+      "loss": 0.5034,
+      "step": 4307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8606279197401389,
+      "learning_rate": 1.5090912642499635e-05,
+      "loss": 0.5733,
+      "step": 4308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9107650353024779,
+      "learning_rate": 1.5088646664043652e-05,
+      "loss": 0.5562,
+      "step": 4309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8475845684717342,
+      "learning_rate": 1.5086380332948962e-05,
+      "loss": 0.5049,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.900875716155614,
+      "learning_rate": 1.5084113649372634e-05,
+      "loss": 0.5144,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9098977051510746,
+      "learning_rate": 1.5081846613471736e-05,
+      "loss": 0.5301,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9173494547586318,
+      "learning_rate": 1.5079579225403373e-05,
+      "loss": 0.5734,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8997686442804093,
+      "learning_rate": 1.507731148532468e-05,
+      "loss": 0.6031,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0035543282799402,
+      "learning_rate": 1.5075043393392799e-05,
+      "loss": 0.594,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9416183936432486,
+      "learning_rate": 1.5072774949764916e-05,
+      "loss": 0.563,
+      "step": 4316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.111424103703773,
+      "learning_rate": 1.5070506154598228e-05,
+      "loss": 0.6619,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9708532135401428,
+      "learning_rate": 1.5068237008049963e-05,
+      "loss": 0.5821,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.914665472381504,
+      "learning_rate": 1.5065967510277366e-05,
+      "loss": 0.5453,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8798308983590678,
+      "learning_rate": 1.5063697661437713e-05,
+      "loss": 0.5403,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8131661194438965,
+      "learning_rate": 1.5061427461688306e-05,
+      "loss": 0.4882,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9820694718040851,
+      "learning_rate": 1.5059156911186465e-05,
+      "loss": 0.5487,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0046413759018036,
+      "learning_rate": 1.5056886010089536e-05,
+      "loss": 0.6133,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8878118116815175,
+      "learning_rate": 1.5054614758554896e-05,
+      "loss": 0.653,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2206966797071999,
+      "learning_rate": 1.5052343156739933e-05,
+      "loss": 0.6036,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9180573056939928,
+      "learning_rate": 1.5050071204802073e-05,
+      "loss": 0.5868,
+      "step": 4326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0217753319934397,
+      "learning_rate": 1.5047798902898756e-05,
+      "loss": 0.622,
+      "step": 4327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8974867336494516,
+      "learning_rate": 1.5045526251187452e-05,
+      "loss": 0.5798,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9152499171753111,
+      "learning_rate": 1.5043253249825656e-05,
+      "loss": 0.6113,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9093691066885794,
+      "learning_rate": 1.5040979898970883e-05,
+      "loss": 0.588,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1133396282471304,
+      "learning_rate": 1.5038706198780673e-05,
+      "loss": 0.5973,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9198753057184388,
+      "learning_rate": 1.5036432149412592e-05,
+      "loss": 0.5804,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0969343843678707,
+      "learning_rate": 1.5034157751024232e-05,
+      "loss": 0.6763,
+      "step": 4333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9186493768039867,
+      "learning_rate": 1.5031883003773206e-05,
+      "loss": 0.5646,
+      "step": 4334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9582047355056563,
+      "learning_rate": 1.502960790781715e-05,
+      "loss": 0.5473,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8762162500326096,
+      "learning_rate": 1.5027332463313729e-05,
+      "loss": 0.622,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8232637771036523,
+      "learning_rate": 1.5025056670420624e-05,
+      "loss": 0.5395,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.90982048874615,
+      "learning_rate": 1.502278052929555e-05,
+      "loss": 0.5762,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9655525218679801,
+      "learning_rate": 1.5020504040096241e-05,
+      "loss": 0.5324,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7943921727751669,
+      "learning_rate": 1.5018227202980455e-05,
+      "loss": 0.4902,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0548661431756876,
+      "learning_rate": 1.5015950018105976e-05,
+      "loss": 0.4857,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9307571470045382,
+      "learning_rate": 1.5013672485630611e-05,
+      "loss": 0.6182,
+      "step": 4342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.929896560472335,
+      "learning_rate": 1.5011394605712188e-05,
+      "loss": 0.4729,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8867515976963221,
+      "learning_rate": 1.5009116378508564e-05,
+      "loss": 0.5139,
+      "step": 4344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.02226268665367,
+      "learning_rate": 1.5006837804177618e-05,
+      "loss": 0.556,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9191762225421793,
+      "learning_rate": 1.5004558882877254e-05,
+      "loss": 0.6007,
+      "step": 4346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9514399880147593,
+      "learning_rate": 1.5002279614765396e-05,
+      "loss": 0.5309,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9494607283066798,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4908,
+      "step": 4348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1928182550584672,
+      "learning_rate": 1.499772003873904e-05,
+      "loss": 0.5622,
+      "step": 4349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9683394095413308,
+      "learning_rate": 1.4995439731140512e-05,
+      "loss": 0.548,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0536520054444802,
+      "learning_rate": 1.4993159077362445e-05,
+      "loss": 0.6328,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9355912661637906,
+      "learning_rate": 1.499087807756288e-05,
+      "loss": 0.5965,
+      "step": 4352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9159385022666104,
+      "learning_rate": 1.4988596731899889e-05,
+      "loss": 0.6053,
+      "step": 4353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9298009190003068,
+      "learning_rate": 1.4986315040531574e-05,
+      "loss": 0.5606,
+      "step": 4354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9000952798543325,
+      "learning_rate": 1.4984033003616047e-05,
+      "loss": 0.603,
+      "step": 4355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.98380076433322,
+      "learning_rate": 1.4981750621311453e-05,
+      "loss": 0.5949,
+      "step": 4356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9435040964843473,
+      "learning_rate": 1.4979467893775963e-05,
+      "loss": 0.534,
+      "step": 4357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9903874881684613,
+      "learning_rate": 1.4977184821167764e-05,
+      "loss": 0.545,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9306842224405125,
+      "learning_rate": 1.4974901403645068e-05,
+      "loss": 0.4705,
+      "step": 4359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.879005834633366,
+      "learning_rate": 1.497261764136612e-05,
+      "loss": 0.5459,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8701678877283088,
+      "learning_rate": 1.4970333534489179e-05,
+      "loss": 0.6014,
+      "step": 4361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7909788159248554,
+      "learning_rate": 1.4968049083172534e-05,
+      "loss": 0.5239,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8666354990733092,
+      "learning_rate": 1.4965764287574494e-05,
+      "loss": 0.5672,
+      "step": 4363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9548946836991843,
+      "learning_rate": 1.4963479147853393e-05,
+      "loss": 0.5818,
+      "step": 4364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0469311919480322,
+      "learning_rate": 1.496119366416759e-05,
+      "loss": 0.6307,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9131536614559036,
+      "learning_rate": 1.4958907836675467e-05,
+      "loss": 0.5407,
+      "step": 4366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9119903308980676,
+      "learning_rate": 1.495662166553543e-05,
+      "loss": 0.5048,
+      "step": 4367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.139918394904237,
+      "learning_rate": 1.4954335150905905e-05,
+      "loss": 0.5712,
+      "step": 4368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.982964987239106,
+      "learning_rate": 1.4952048292945352e-05,
+      "loss": 0.5777,
+      "step": 4369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9613085225608329,
+      "learning_rate": 1.4949761091812243e-05,
+      "loss": 0.5306,
+      "step": 4370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0194065477967498,
+      "learning_rate": 1.4947473547665081e-05,
+      "loss": 0.6194,
+      "step": 4371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8825081634776647,
+      "learning_rate": 1.4945185660662391e-05,
+      "loss": 0.5097,
+      "step": 4372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9650839534634333,
+      "learning_rate": 1.4942897430962722e-05,
+      "loss": 0.4978,
+      "step": 4373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0318679026761992,
+      "learning_rate": 1.494060885872464e-05,
+      "loss": 0.5238,
+      "step": 4374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8864586334831703,
+      "learning_rate": 1.4938319944106754e-05,
+      "loss": 0.5226,
+      "step": 4375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0151435374926832,
+      "learning_rate": 1.4936030687267672e-05,
+      "loss": 0.5641,
+      "step": 4376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9680320885205316,
+      "learning_rate": 1.4933741088366043e-05,
+      "loss": 0.6272,
+      "step": 4377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0063656243311683,
+      "learning_rate": 1.4931451147560534e-05,
+      "loss": 0.5885,
+      "step": 4378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8969813806528744,
+      "learning_rate": 1.492916086500983e-05,
+      "loss": 0.5366,
+      "step": 4379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.89570113690597,
+      "learning_rate": 1.4926870240872652e-05,
+      "loss": 0.5287,
+      "step": 4380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6806229089415208,
+      "learning_rate": 1.4924579275307738e-05,
+      "loss": 0.5136,
+      "step": 4381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8772278715634582,
+      "learning_rate": 1.492228796847385e-05,
+      "loss": 0.5823,
+      "step": 4382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0148013614453049,
+      "learning_rate": 1.4919996320529768e-05,
+      "loss": 0.6213,
+      "step": 4383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9820137070849825,
+      "learning_rate": 1.4917704331634306e-05,
+      "loss": 0.5012,
+      "step": 4384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8983543389621446,
+      "learning_rate": 1.49154120019463e-05,
+      "loss": 0.5577,
+      "step": 4385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.839493694424055,
+      "learning_rate": 1.4913119331624597e-05,
+      "loss": 0.5441,
+      "step": 4386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9193981006433436,
+      "learning_rate": 1.4910826320828085e-05,
+      "loss": 0.5411,
+      "step": 4387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0090728948785002,
+      "learning_rate": 1.4908532969715663e-05,
+      "loss": 0.5255,
+      "step": 4388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9836631247147275,
+      "learning_rate": 1.490623927844626e-05,
+      "loss": 0.5497,
+      "step": 4389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8860894415981879,
+      "learning_rate": 1.4903945247178828e-05,
+      "loss": 0.5318,
+      "step": 4390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8670465037749358,
+      "learning_rate": 1.4901650876072342e-05,
+      "loss": 0.5369,
+      "step": 4391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9953631170549834,
+      "learning_rate": 1.4899356165285794e-05,
+      "loss": 0.5545,
+      "step": 4392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9063420930863816,
+      "learning_rate": 1.489706111497821e-05,
+      "loss": 0.5357,
+      "step": 4393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.881755479709398,
+      "learning_rate": 1.489476572530864e-05,
+      "loss": 0.5343,
+      "step": 4394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8193129452548926,
+      "learning_rate": 1.489246999643614e-05,
+      "loss": 0.4722,
+      "step": 4395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8783440703065661,
+      "learning_rate": 1.489017392851981e-05,
+      "loss": 0.5433,
+      "step": 4396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8547465151818406,
+      "learning_rate": 1.488787752171877e-05,
+      "loss": 0.5481,
+      "step": 4397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9133780786690582,
+      "learning_rate": 1.4885580776192149e-05,
+      "loss": 0.5353,
+      "step": 4398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8782420674843047,
+      "learning_rate": 1.4883283692099114e-05,
+      "loss": 0.5447,
+      "step": 4399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9650825987663942,
+      "learning_rate": 1.488098626959885e-05,
+      "loss": 0.6096,
+      "step": 4400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9457461408325257,
+      "learning_rate": 1.4878688508850567e-05,
+      "loss": 0.5685,
+      "step": 4401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8396109468560583,
+      "learning_rate": 1.4876390410013498e-05,
+      "loss": 0.5542,
+      "step": 4402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.10276618772219,
+      "learning_rate": 1.48740919732469e-05,
+      "loss": 0.6523,
+      "step": 4403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8469604534554713,
+      "learning_rate": 1.487179319871005e-05,
+      "loss": 0.5672,
+      "step": 4404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9119790850830539,
+      "learning_rate": 1.4869494086562253e-05,
+      "loss": 0.6194,
+      "step": 4405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.887038406000676,
+      "learning_rate": 1.4867194636962836e-05,
+      "loss": 0.5182,
+      "step": 4406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0079039365291105,
+      "learning_rate": 1.4864894850071147e-05,
+      "loss": 0.5612,
+      "step": 4407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8390473208179717,
+      "learning_rate": 1.486259472604656e-05,
+      "loss": 0.5347,
+      "step": 4408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9686207645890165,
+      "learning_rate": 1.4860294265048474e-05,
+      "loss": 0.5913,
+      "step": 4409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0015965610679087,
+      "learning_rate": 1.4857993467236303e-05,
+      "loss": 0.6163,
+      "step": 4410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9101404073900543,
+      "learning_rate": 1.4855692332769494e-05,
+      "loss": 0.537,
+      "step": 4411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9229731168365843,
+      "learning_rate": 1.4853390861807518e-05,
+      "loss": 0.5235,
+      "step": 4412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8095758622279768,
+      "learning_rate": 1.4851089054509852e-05,
+      "loss": 0.5163,
+      "step": 4413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9142132578654567,
+      "learning_rate": 1.4848786911036023e-05,
+      "loss": 0.5656,
+      "step": 4414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0076384074827356,
+      "learning_rate": 1.4846484431545562e-05,
+      "loss": 0.6286,
+      "step": 4415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0666728554711093,
+      "learning_rate": 1.4844181616198028e-05,
+      "loss": 0.598,
+      "step": 4416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9561431209856557,
+      "learning_rate": 1.4841878465153006e-05,
+      "loss": 0.6087,
+      "step": 4417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8399907289741158,
+      "learning_rate": 1.4839574978570098e-05,
+      "loss": 0.5377,
+      "step": 4418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9088906220770147,
+      "learning_rate": 1.4837271156608938e-05,
+      "loss": 0.5189,
+      "step": 4419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.907192125612477,
+      "learning_rate": 1.4834966999429179e-05,
+      "loss": 0.5094,
+      "step": 4420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9549375328800763,
+      "learning_rate": 1.4832662507190493e-05,
+      "loss": 0.5861,
+      "step": 4421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8195915273699955,
+      "learning_rate": 1.4830357680052586e-05,
+      "loss": 0.448,
+      "step": 4422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8457199060127676,
+      "learning_rate": 1.4828052518175172e-05,
+      "loss": 0.5053,
+      "step": 4423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9655078403174225,
+      "learning_rate": 1.4825747021718002e-05,
+      "loss": 0.5754,
+      "step": 4424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8993455107136055,
+      "learning_rate": 1.4823441190840844e-05,
+      "loss": 0.5951,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9695646262062783,
+      "learning_rate": 1.4821135025703491e-05,
+      "loss": 0.5539,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9340908557152009,
+      "learning_rate": 1.4818828526465755e-05,
+      "loss": 0.5463,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9484184703205448,
+      "learning_rate": 1.4816521693287477e-05,
+      "loss": 0.5511,
+      "step": 4428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8090461904484498,
+      "learning_rate": 1.481421452632852e-05,
+      "loss": 0.5391,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9068706093760909,
+      "learning_rate": 1.4811907025748764e-05,
+      "loss": 0.531,
+      "step": 4430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8404838158009921,
+      "learning_rate": 1.480959919170812e-05,
+      "loss": 0.5492,
+      "step": 4431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8289162524796344,
+      "learning_rate": 1.4807291024366519e-05,
+      "loss": 0.5513,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8394841803055377,
+      "learning_rate": 1.4804982523883915e-05,
+      "loss": 0.5078,
+      "step": 4433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9250383568107076,
+      "learning_rate": 1.4802673690420281e-05,
+      "loss": 0.5885,
+      "step": 4434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9495083386718011,
+      "learning_rate": 1.4800364524135622e-05,
+      "loss": 0.5778,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9536242637144737,
+      "learning_rate": 1.4798055025189962e-05,
+      "loss": 0.6067,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8677917692119125,
+      "learning_rate": 1.4795745193743341e-05,
+      "loss": 0.5587,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8866934244445475,
+      "learning_rate": 1.4793435029955832e-05,
+      "loss": 0.5103,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0042081691754168,
+      "learning_rate": 1.4791124533987529e-05,
+      "loss": 0.6197,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.869082561205368,
+      "learning_rate": 1.4788813705998543e-05,
+      "loss": 0.6103,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9407422204488816,
+      "learning_rate": 1.4786502546149015e-05,
+      "loss": 0.5694,
+      "step": 4441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9278704826539301,
+      "learning_rate": 1.4784191054599109e-05,
+      "loss": 0.5968,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9588874501973558,
+      "learning_rate": 1.4781879231509005e-05,
+      "loss": 0.5519,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8591868818411956,
+      "learning_rate": 1.4779567077038912e-05,
+      "loss": 0.5607,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8701437695360021,
+      "learning_rate": 1.477725459134906e-05,
+      "loss": 0.5065,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9023479514656566,
+      "learning_rate": 1.4774941774599703e-05,
+      "loss": 0.5855,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8904185731123726,
+      "learning_rate": 1.4772628626951114e-05,
+      "loss": 0.5713,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8535558813189821,
+      "learning_rate": 1.4770315148563596e-05,
+      "loss": 0.5409,
+      "step": 4448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.962305319537762,
+      "learning_rate": 1.4768001339597471e-05,
+      "loss": 0.5573,
+      "step": 4449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9126805318858925,
+      "learning_rate": 1.4765687200213079e-05,
+      "loss": 0.5475,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9893873429771054,
+      "learning_rate": 1.4763372730570793e-05,
+      "loss": 0.6411,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.935005237685736,
+      "learning_rate": 1.4761057930831002e-05,
+      "loss": 0.5588,
+      "step": 4452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9159587474165702,
+      "learning_rate": 1.475874280115412e-05,
+      "loss": 0.5548,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9045849178952995,
+      "learning_rate": 1.475642734170058e-05,
+      "loss": 0.5059,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9272647834317185,
+      "learning_rate": 1.475411155263085e-05,
+      "loss": 0.5701,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8679504825088333,
+      "learning_rate": 1.47517954341054e-05,
+      "loss": 0.5504,
+      "step": 4456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8784123628733355,
+      "learning_rate": 1.4749478986284743e-05,
+      "loss": 0.5651,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7613363838766739,
+      "learning_rate": 1.4747162209329408e-05,
+      "loss": 0.5165,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.014684584701171,
+      "learning_rate": 1.474484510339994e-05,
+      "loss": 0.6104,
+      "step": 4459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8678082577950834,
+      "learning_rate": 1.4742527668656915e-05,
+      "loss": 0.5358,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9801279027102768,
+      "learning_rate": 1.474020990526093e-05,
+      "loss": 0.5773,
+      "step": 4461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.829818263982647,
+      "learning_rate": 1.4737891813372605e-05,
+      "loss": 0.5223,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8525757596323112,
+      "learning_rate": 1.4735573393152576e-05,
+      "loss": 0.5602,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9103375069043098,
+      "learning_rate": 1.4733254644761514e-05,
+      "loss": 0.5933,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9087295841386293,
+      "learning_rate": 1.4730935568360103e-05,
+      "loss": 0.6359,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9403445729528365,
+      "learning_rate": 1.4728616164109051e-05,
+      "loss": 0.5755,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.879590804392101,
+      "learning_rate": 1.4726296432169095e-05,
+      "loss": 0.5774,
+      "step": 4467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9251101238642484,
+      "learning_rate": 1.472397637270099e-05,
+      "loss": 0.5488,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9440786095944539,
+      "learning_rate": 1.472165598586551e-05,
+      "loss": 0.5501,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.892311824110475,
+      "learning_rate": 1.4719335271823461e-05,
+      "loss": 0.5545,
+      "step": 4470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9222735252882099,
+      "learning_rate": 1.4717014230735661e-05,
+      "loss": 0.5671,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8774550769864838,
+      "learning_rate": 1.471469286276296e-05,
+      "loss": 0.5856,
+      "step": 4472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8945744492384955,
+      "learning_rate": 1.4712371168066227e-05,
+      "loss": 0.5554,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8632486488540376,
+      "learning_rate": 1.4710049146806348e-05,
+      "loss": 0.5571,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9075970561265309,
+      "learning_rate": 1.4707726799144245e-05,
+      "loss": 0.5557,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8623183406655154,
+      "learning_rate": 1.4705404125240849e-05,
+      "loss": 0.562,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8993540931201318,
+      "learning_rate": 1.470308112525712e-05,
+      "loss": 0.5011,
+      "step": 4477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9070876790888006,
+      "learning_rate": 1.470075779935404e-05,
+      "loss": 0.5386,
+      "step": 4478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.894769666418227,
+      "learning_rate": 1.4698434147692618e-05,
+      "loss": 0.6098,
+      "step": 4479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.89583961043636,
+      "learning_rate": 1.4696110170433873e-05,
+      "loss": 0.5813,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8861955251651111,
+      "learning_rate": 1.469378586773886e-05,
+      "loss": 0.565,
+      "step": 4481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9275619282158948,
+      "learning_rate": 1.4691461239768649e-05,
+      "loss": 0.5474,
+      "step": 4482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9046505136392399,
+      "learning_rate": 1.4689136286684335e-05,
+      "loss": 0.5759,
+      "step": 4483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9305391975015446,
+      "learning_rate": 1.4686811008647037e-05,
+      "loss": 0.5531,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9903532887324388,
+      "learning_rate": 1.4684485405817897e-05,
+      "loss": 0.578,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.905053324048193,
+      "learning_rate": 1.4682159478358067e-05,
+      "loss": 0.5344,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9706423919522855,
+      "learning_rate": 1.467983322642874e-05,
+      "loss": 0.5948,
+      "step": 4487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9127425851510094,
+      "learning_rate": 1.4677506650191124e-05,
+      "loss": 0.5651,
+      "step": 4488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9640169198932111,
+      "learning_rate": 1.4675179749806444e-05,
+      "loss": 0.5849,
+      "step": 4489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8409429176257024,
+      "learning_rate": 1.4672852525435954e-05,
+      "loss": 0.5098,
+      "step": 4490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9086225149968713,
+      "learning_rate": 1.4670524977240929e-05,
+      "loss": 0.5602,
+      "step": 4491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9895660646217311,
+      "learning_rate": 1.4668197105382667e-05,
+      "loss": 0.6056,
+      "step": 4492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9110056402144131,
+      "learning_rate": 1.4665868910022485e-05,
+      "loss": 0.569,
+      "step": 4493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.964291622484286,
+      "learning_rate": 1.4663540391321726e-05,
+      "loss": 0.5913,
+      "step": 4494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0321238359078417,
+      "learning_rate": 1.4661211549441756e-05,
+      "loss": 0.6189,
+      "step": 4495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8883996584082586,
+      "learning_rate": 1.465888238454396e-05,
+      "loss": 0.5703,
+      "step": 4496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9221703556073723,
+      "learning_rate": 1.4656552896789746e-05,
+      "loss": 0.5558,
+      "step": 4497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8746129298054462,
+      "learning_rate": 1.4654223086340547e-05,
+      "loss": 0.488,
+      "step": 4498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9292851665842938,
+      "learning_rate": 1.4651892953357816e-05,
+      "loss": 0.5466,
+      "step": 4499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9489222789168605,
+      "learning_rate": 1.4649562498003032e-05,
+      "loss": 0.6406,
+      "step": 4500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9472256638031281,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.5931,
+      "step": 4501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8927187601605701,
+      "learning_rate": 1.4644900620823308e-05,
+      "loss": 0.5251,
+      "step": 4502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9092235192734804,
+      "learning_rate": 1.4642569199321436e-05,
+      "loss": 0.5596,
+      "step": 4503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9911171371540515,
+      "learning_rate": 1.4640237456093636e-05,
+      "loss": 0.6254,
+      "step": 4504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8902682685340019,
+      "learning_rate": 1.4637905391301496e-05,
+      "loss": 0.5597,
+      "step": 4505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9779813845697346,
+      "learning_rate": 1.4635573005106627e-05,
+      "loss": 0.5155,
+      "step": 4506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8593153982172275,
+      "learning_rate": 1.4633240297670661e-05,
+      "loss": 0.5848,
+      "step": 4507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9417037319181273,
+      "learning_rate": 1.463090726915525e-05,
+      "loss": 0.5036,
+      "step": 4508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.826959279039463,
+      "learning_rate": 1.4628573919722073e-05,
+      "loss": 0.5068,
+      "step": 4509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9122443267097082,
+      "learning_rate": 1.4626240249532833e-05,
+      "loss": 0.5697,
+      "step": 4510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.911386237979581,
+      "learning_rate": 1.4623906258749243e-05,
+      "loss": 0.5511,
+      "step": 4511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9836468674217567,
+      "learning_rate": 1.462157194753305e-05,
+      "loss": 0.6439,
+      "step": 4512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9245030210290315,
+      "learning_rate": 1.4619237316046024e-05,
+      "loss": 0.5938,
+      "step": 4513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8823041615405315,
+      "learning_rate": 1.4616902364449947e-05,
+      "loss": 0.5385,
+      "step": 4514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9282146018943377,
+      "learning_rate": 1.4614567092906631e-05,
+      "loss": 0.5289,
+      "step": 4515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8513409879459296,
+      "learning_rate": 1.4612231501577912e-05,
+      "loss": 0.4712,
+      "step": 4516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8970738325052732,
+      "learning_rate": 1.4609895590625635e-05,
+      "loss": 0.5656,
+      "step": 4517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8977821484073438,
+      "learning_rate": 1.4607559360211688e-05,
+      "loss": 0.5297,
+      "step": 4518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8660669647290293,
+      "learning_rate": 1.460522281049796e-05,
+      "loss": 0.5524,
+      "step": 4519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8910709052637866,
+      "learning_rate": 1.460288594164638e-05,
+      "loss": 0.5052,
+      "step": 4520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8727890717084368,
+      "learning_rate": 1.4600548753818884e-05,
+      "loss": 0.4904,
+      "step": 4521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9238031222868925,
+      "learning_rate": 1.4598211247177443e-05,
+      "loss": 0.5652,
+      "step": 4522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9175522190299712,
+      "learning_rate": 1.459587342188404e-05,
+      "loss": 0.5368,
+      "step": 4523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9529440524959538,
+      "learning_rate": 1.4593535278100684e-05,
+      "loss": 0.5888,
+      "step": 4524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8746329567698374,
+      "learning_rate": 1.4591196815989407e-05,
+      "loss": 0.5749,
+      "step": 4525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9699853208903029,
+      "learning_rate": 1.4588858035712266e-05,
+      "loss": 0.5508,
+      "step": 4526
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9555329839997029,
+      "learning_rate": 1.4586518937431332e-05,
+      "loss": 0.5743,
+      "step": 4527
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9458608579519454,
+      "learning_rate": 1.4584179521308703e-05,
+      "loss": 0.608,
+      "step": 4528
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8527041686893861,
+      "learning_rate": 1.45818397875065e-05,
+      "loss": 0.473,
+      "step": 4529
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8928812741342821,
+      "learning_rate": 1.4579499736186864e-05,
+      "loss": 0.5538,
+      "step": 4530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9120715539821165,
+      "learning_rate": 1.4577159367511959e-05,
+      "loss": 0.4966,
+      "step": 4531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0404997402846254,
+      "learning_rate": 1.457481868164397e-05,
+      "loss": 0.5259,
+      "step": 4532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9213843354801686,
+      "learning_rate": 1.45724776787451e-05,
+      "loss": 0.5097,
+      "step": 4533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1301394046908915,
+      "learning_rate": 1.4570136358977589e-05,
+      "loss": 0.5866,
+      "step": 4534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8993396951953636,
+      "learning_rate": 1.456779472250368e-05,
+      "loss": 0.6042,
+      "step": 4535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9597726253126289,
+      "learning_rate": 1.4565452769485644e-05,
+      "loss": 0.6088,
+      "step": 4536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9236890364028195,
+      "learning_rate": 1.4563110500085786e-05,
+      "loss": 0.5624,
+      "step": 4537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8956434628330934,
+      "learning_rate": 1.4560767914466417e-05,
+      "loss": 0.5511,
+      "step": 4538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9639572643026926,
+      "learning_rate": 1.4558425012789873e-05,
+      "loss": 0.5579,
+      "step": 4539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9794697866313931,
+      "learning_rate": 1.4556081795218525e-05,
+      "loss": 0.5269,
+      "step": 4540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8887634982521235,
+      "learning_rate": 1.4553738261914742e-05,
+      "loss": 0.5336,
+      "step": 4541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9542512904237881,
+      "learning_rate": 1.4551394413040942e-05,
+      "loss": 0.5485,
+      "step": 4542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.994243489296905,
+      "learning_rate": 1.4549050248759546e-05,
+      "loss": 0.5648,
+      "step": 4543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.973290658485551,
+      "learning_rate": 1.4546705769233003e-05,
+      "loss": 0.5819,
+      "step": 4544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9707765308575302,
+      "learning_rate": 1.4544360974623781e-05,
+      "loss": 0.5097,
+      "step": 4545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9736719646068369,
+      "learning_rate": 1.4542015865094377e-05,
+      "loss": 0.5682,
+      "step": 4546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9026646459372545,
+      "learning_rate": 1.4539670440807298e-05,
+      "loss": 0.5721,
+      "step": 4547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9847250280289744,
+      "learning_rate": 1.4537324701925088e-05,
+      "loss": 0.632,
+      "step": 4548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9816038435971146,
+      "learning_rate": 1.4534978648610301e-05,
+      "loss": 0.5988,
+      "step": 4549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0238723489848807,
+      "learning_rate": 1.4532632281025514e-05,
+      "loss": 0.6407,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8774092216491454,
+      "learning_rate": 1.4530285599333332e-05,
+      "loss": 0.5169,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8977191065810496,
+      "learning_rate": 1.4527938603696376e-05,
+      "loss": 0.5758,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8679359206799987,
+      "learning_rate": 1.452559129427729e-05,
+      "loss": 0.5466,
+      "step": 4553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8862085480729563,
+      "learning_rate": 1.4523243671238741e-05,
+      "loss": 0.5008,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8560466744924718,
+      "learning_rate": 1.4520895734743419e-05,
+      "loss": 0.5498,
+      "step": 4555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.870445799003857,
+      "learning_rate": 1.4518547484954033e-05,
+      "loss": 0.5446,
+      "step": 4556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9915550536321615,
+      "learning_rate": 1.4516198922033313e-05,
+      "loss": 0.5891,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9766747506463757,
+      "learning_rate": 1.4513850046144015e-05,
+      "loss": 0.5856,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1299468463134879,
+      "learning_rate": 1.451150085744891e-05,
+      "loss": 0.5327,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.876075346471893,
+      "learning_rate": 1.45091513561108e-05,
+      "loss": 0.5676,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9026216340846099,
+      "learning_rate": 1.4506801542292501e-05,
+      "loss": 0.4733,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9271284281161375,
+      "learning_rate": 1.450445141615685e-05,
+      "loss": 0.6201,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9643881947320083,
+      "learning_rate": 1.4502100977866713e-05,
+      "loss": 0.63,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9016319613024337,
+      "learning_rate": 1.449975022758497e-05,
+      "loss": 0.5579,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8301182624895908,
+      "learning_rate": 1.449739916547453e-05,
+      "loss": 0.5391,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0885222691950676,
+      "learning_rate": 1.4495047791698316e-05,
+      "loss": 0.5193,
+      "step": 4566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8532968406074554,
+      "learning_rate": 1.4492696106419275e-05,
+      "loss": 0.5216,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9434837561969038,
+      "learning_rate": 1.4490344109800382e-05,
+      "loss": 0.5765,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8803884583946759,
+      "learning_rate": 1.4487991802004625e-05,
+      "loss": 0.486,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8000039279991835,
+      "learning_rate": 1.4485639183195014e-05,
+      "loss": 0.4766,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9594388992920403,
+      "learning_rate": 1.448328625353459e-05,
+      "loss": 0.5695,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8484072973197494,
+      "learning_rate": 1.4480933013186403e-05,
+      "loss": 0.5692,
+      "step": 4572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1050447422521597,
+      "learning_rate": 1.4478579462313533e-05,
+      "loss": 0.5268,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8681889570895173,
+      "learning_rate": 1.4476225601079078e-05,
+      "loss": 0.5614,
+      "step": 4574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9827528961085784,
+      "learning_rate": 1.447387142964616e-05,
+      "loss": 0.6487,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9192293189477148,
+      "learning_rate": 1.4471516948177921e-05,
+      "loss": 0.5596,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8507728172540402,
+      "learning_rate": 1.4469162156837521e-05,
+      "loss": 0.458,
+      "step": 4577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9116255169315678,
+      "learning_rate": 1.4466807055788152e-05,
+      "loss": 0.543,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8972062853518864,
+      "learning_rate": 1.4464451645193013e-05,
+      "loss": 0.5706,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8844699891420403,
+      "learning_rate": 1.4462095925215336e-05,
+      "loss": 0.5486,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8469427622135454,
+      "learning_rate": 1.445973989601837e-05,
+      "loss": 0.5295,
+      "step": 4581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0531060450779732,
+      "learning_rate": 1.4457383557765385e-05,
+      "loss": 0.6393,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8996525214835874,
+      "learning_rate": 1.4455026910619672e-05,
+      "loss": 0.5354,
+      "step": 4583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8842873124765084,
+      "learning_rate": 1.4452669954744545e-05,
+      "loss": 0.5528,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8787924380403497,
+      "learning_rate": 1.4450312690303342e-05,
+      "loss": 0.629,
+      "step": 4585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8125932722275822,
+      "learning_rate": 1.4447955117459414e-05,
+      "loss": 0.4836,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8355003410967605,
+      "learning_rate": 1.4445597236376143e-05,
+      "loss": 0.5161,
+      "step": 4587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.890935498183146,
+      "learning_rate": 1.4443239047216928e-05,
+      "loss": 0.5175,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9152569502319765,
+      "learning_rate": 1.4440880550145187e-05,
+      "loss": 0.5548,
+      "step": 4589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9338050629782046,
+      "learning_rate": 1.4438521745324363e-05,
+      "loss": 0.605,
+      "step": 4590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8610689574026248,
+      "learning_rate": 1.4436162632917918e-05,
+      "loss": 0.5304,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8372614133811489,
+      "learning_rate": 1.4433803213089341e-05,
+      "loss": 0.5228,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8240098780319006,
+      "learning_rate": 1.443144348600213e-05,
+      "loss": 0.5509,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9114321836325133,
+      "learning_rate": 1.442908345181982e-05,
+      "loss": 0.5135,
+      "step": 4594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8340969162814056,
+      "learning_rate": 1.442672311070595e-05,
+      "loss": 0.5039,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.886391196202956,
+      "learning_rate": 1.44243624628241e-05,
+      "loss": 0.5364,
+      "step": 4596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9536427603728262,
+      "learning_rate": 1.4422001508337853e-05,
+      "loss": 0.5255,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9276882420801554,
+      "learning_rate": 1.4419640247410827e-05,
+      "loss": 0.5701,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8960716066915557,
+      "learning_rate": 1.4417278680206647e-05,
+      "loss": 0.594,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9832722742030179,
+      "learning_rate": 1.4414916806888976e-05,
+      "loss": 0.6219,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9796662164059784,
+      "learning_rate": 1.4412554627621487e-05,
+      "loss": 0.5509,
+      "step": 4601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9263311552853005,
+      "learning_rate": 1.4410192142567874e-05,
+      "loss": 0.4821,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9106880002039524,
+      "learning_rate": 1.4407829351891858e-05,
+      "loss": 0.5703,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8117373702202885,
+      "learning_rate": 1.4405466255757178e-05,
+      "loss": 0.5324,
+      "step": 4604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9051324540084974,
+      "learning_rate": 1.4403102854327595e-05,
+      "loss": 0.5474,
+      "step": 4605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8988441607793508,
+      "learning_rate": 1.4400739147766887e-05,
+      "loss": 0.5718,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9418428015077235,
+      "learning_rate": 1.4398375136238864e-05,
+      "loss": 0.5627,
+      "step": 4607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8659909136750779,
+      "learning_rate": 1.439601081990734e-05,
+      "loss": 0.5206,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9068140592385949,
+      "learning_rate": 1.4393646198936169e-05,
+      "loss": 0.5892,
+      "step": 4609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9967771784960222,
+      "learning_rate": 1.4391281273489216e-05,
+      "loss": 0.6452,
+      "step": 4610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8349913229312583,
+      "learning_rate": 1.438891604373036e-05,
+      "loss": 0.5053,
+      "step": 4611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8885153877992091,
+      "learning_rate": 1.4386550509823515e-05,
+      "loss": 0.5386,
+      "step": 4612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8717848449955791,
+      "learning_rate": 1.4384184671932616e-05,
+      "loss": 0.4851,
+      "step": 4613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.962862508614908,
+      "learning_rate": 1.4381818530221604e-05,
+      "loss": 0.6124,
+      "step": 4614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8910489915347034,
+      "learning_rate": 1.4379452084854455e-05,
+      "loss": 0.538,
+      "step": 4615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.90704637704644,
+      "learning_rate": 1.4377085335995165e-05,
+      "loss": 0.5425,
+      "step": 4616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8258031697206105,
+      "learning_rate": 1.4374718283807738e-05,
+      "loss": 0.5118,
+      "step": 4617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8838282368452399,
+      "learning_rate": 1.4372350928456218e-05,
+      "loss": 0.5603,
+      "step": 4618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8717929142089527,
+      "learning_rate": 1.4369983270104654e-05,
+      "loss": 0.5141,
+      "step": 4619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9120544658922455,
+      "learning_rate": 1.436761530891713e-05,
+      "loss": 0.6108,
+      "step": 4620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9513750558154765,
+      "learning_rate": 1.4365247045057732e-05,
+      "loss": 0.5394,
+      "step": 4621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9271504423453096,
+      "learning_rate": 1.4362878478690595e-05,
+      "loss": 0.6059,
+      "step": 4622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.935788851587579,
+      "learning_rate": 1.4360509609979842e-05,
+      "loss": 0.5753,
+      "step": 4623
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9640035939735568,
+      "learning_rate": 1.4358140439089644e-05,
+      "loss": 0.5396,
+      "step": 4624
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0144514931115078,
+      "learning_rate": 1.435577096618418e-05,
+      "loss": 0.5673,
+      "step": 4625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0271339489351563,
+      "learning_rate": 1.435340119142765e-05,
+      "loss": 0.5648,
+      "step": 4626
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8403509294262873,
+      "learning_rate": 1.4351031114984277e-05,
+      "loss": 0.4805,
+      "step": 4627
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9565545155982719,
+      "learning_rate": 1.434866073701831e-05,
+      "loss": 0.553,
+      "step": 4628
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9695005048984443,
+      "learning_rate": 1.434629005769401e-05,
+      "loss": 0.5715,
+      "step": 4629
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.977899291178307,
+      "learning_rate": 1.4343919077175662e-05,
+      "loss": 0.5849,
+      "step": 4630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.884291550415101,
+      "learning_rate": 1.434154779562758e-05,
+      "loss": 0.5444,
+      "step": 4631
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8531258142757846,
+      "learning_rate": 1.4339176213214084e-05,
+      "loss": 0.5753,
+      "step": 4632
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8577345206288213,
+      "learning_rate": 1.4336804330099525e-05,
+      "loss": 0.5059,
+      "step": 4633
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7924570235333922,
+      "learning_rate": 1.4334432146448272e-05,
+      "loss": 0.5033,
+      "step": 4634
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.014824293404601,
+      "learning_rate": 1.433205966242472e-05,
+      "loss": 0.6218,
+      "step": 4635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9407577361679119,
+      "learning_rate": 1.4329686878193271e-05,
+      "loss": 0.5733,
+      "step": 4636
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8702144654770304,
+      "learning_rate": 1.4327313793918362e-05,
+      "loss": 0.602,
+      "step": 4637
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8761690276292466,
+      "learning_rate": 1.432494040976445e-05,
+      "loss": 0.5269,
+      "step": 4638
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.929781433069658,
+      "learning_rate": 1.4322566725895998e-05,
+      "loss": 0.5445,
+      "step": 4639
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7992557146672608,
+      "learning_rate": 1.432019274247751e-05,
+      "loss": 0.5139,
+      "step": 4640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.773596192547856,
+      "learning_rate": 1.4317818459673496e-05,
+      "loss": 0.4513,
+      "step": 4641
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9862725969212202,
+      "learning_rate": 1.4315443877648494e-05,
+      "loss": 0.6312,
+      "step": 4642
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9346377617043576,
+      "learning_rate": 1.4313068996567054e-05,
+      "loss": 0.5526,
+      "step": 4643
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9096616854418217,
+      "learning_rate": 1.4310693816593766e-05,
+      "loss": 0.5035,
+      "step": 4644
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8896567937140818,
+      "learning_rate": 1.4308318337893214e-05,
+      "loss": 0.5757,
+      "step": 4645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8989146692626034,
+      "learning_rate": 1.4305942560630025e-05,
+      "loss": 0.5482,
+      "step": 4646
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9001044151809969,
+      "learning_rate": 1.4303566484968836e-05,
+      "loss": 0.5518,
+      "step": 4647
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8284432728019371,
+      "learning_rate": 1.4301190111074306e-05,
+      "loss": 0.5221,
+      "step": 4648
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8823025857272053,
+      "learning_rate": 1.4298813439111116e-05,
+      "loss": 0.5092,
+      "step": 4649
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9638802531589524,
+      "learning_rate": 1.4296436469243968e-05,
+      "loss": 0.5593,
+      "step": 4650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8957597475760447,
+      "learning_rate": 1.4294059201637584e-05,
+      "loss": 0.6216,
+      "step": 4651
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9922406505863793,
+      "learning_rate": 1.4291681636456706e-05,
+      "loss": 0.585,
+      "step": 4652
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9067841463591434,
+      "learning_rate": 1.42893037738661e-05,
+      "loss": 0.5457,
+      "step": 4653
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8603834578523885,
+      "learning_rate": 1.4286925614030542e-05,
+      "loss": 0.5728,
+      "step": 4654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9463071681212694,
+      "learning_rate": 1.4284547157114846e-05,
+      "loss": 0.5982,
+      "step": 4655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8547064534994039,
+      "learning_rate": 1.4282168403283829e-05,
+      "loss": 0.5284,
+      "step": 4656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.883673974873926,
+      "learning_rate": 1.4279789352702342e-05,
+      "loss": 0.5759,
+      "step": 4657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9929381579362356,
+      "learning_rate": 1.4277410005535249e-05,
+      "loss": 0.5804,
+      "step": 4658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8075944223992195,
+      "learning_rate": 1.4275030361947438e-05,
+      "loss": 0.5404,
+      "step": 4659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9575072735956454,
+      "learning_rate": 1.427265042210381e-05,
+      "loss": 0.598,
+      "step": 4660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9730461139752252,
+      "learning_rate": 1.4270270186169301e-05,
+      "loss": 0.5552,
+      "step": 4661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8627026160167858,
+      "learning_rate": 1.4267889654308858e-05,
+      "loss": 0.5816,
+      "step": 4662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8134697842892554,
+      "learning_rate": 1.4265508826687442e-05,
+      "loss": 0.5051,
+      "step": 4663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8604152441105043,
+      "learning_rate": 1.4263127703470053e-05,
+      "loss": 0.5473,
+      "step": 4664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8854386728519589,
+      "learning_rate": 1.4260746284821694e-05,
+      "loss": 0.5699,
+      "step": 4665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8920693498069167,
+      "learning_rate": 1.4258364570907395e-05,
+      "loss": 0.5398,
+      "step": 4666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9962386036036195,
+      "learning_rate": 1.4255982561892207e-05,
+      "loss": 0.6245,
+      "step": 4667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8323716865619504,
+      "learning_rate": 1.4253600257941208e-05,
+      "loss": 0.5058,
+      "step": 4668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8699456667642135,
+      "learning_rate": 1.4251217659219481e-05,
+      "loss": 0.6105,
+      "step": 4669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9889583687179783,
+      "learning_rate": 1.4248834765892139e-05,
+      "loss": 0.5819,
+      "step": 4670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9126491520550216,
+      "learning_rate": 1.4246451578124321e-05,
+      "loss": 0.56,
+      "step": 4671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9134335213125362,
+      "learning_rate": 1.4244068096081172e-05,
+      "loss": 0.5301,
+      "step": 4672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9273990704818516,
+      "learning_rate": 1.4241684319927869e-05,
+      "loss": 0.4937,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9268357832113618,
+      "learning_rate": 1.4239300249829606e-05,
+      "loss": 0.5696,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8527064171200383,
+      "learning_rate": 1.4236915885951592e-05,
+      "loss": 0.553,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9562426729410712,
+      "learning_rate": 1.4234531228459069e-05,
+      "loss": 0.5627,
+      "step": 4676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9183753162569681,
+      "learning_rate": 1.4232146277517289e-05,
+      "loss": 0.5514,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9248669788204029,
+      "learning_rate": 1.4229761033291523e-05,
+      "loss": 0.5248,
+      "step": 4678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9321721977272228,
+      "learning_rate": 1.422737549594707e-05,
+      "loss": 0.537,
+      "step": 4679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8532668316951978,
+      "learning_rate": 1.4224989665649248e-05,
+      "loss": 0.5675,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8601920380266586,
+      "learning_rate": 1.4222603542563385e-05,
+      "loss": 0.4955,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.869794347959696,
+      "learning_rate": 1.4220217126854842e-05,
+      "loss": 0.5749,
+      "step": 4682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.972823849878353,
+      "learning_rate": 1.4217830418689e-05,
+      "loss": 0.558,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8801348435436541,
+      "learning_rate": 1.4215443418231248e-05,
+      "loss": 0.5038,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8921011812497122,
+      "learning_rate": 1.4213056125647005e-05,
+      "loss": 0.5297,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9004852636949913,
+      "learning_rate": 1.4210668541101713e-05,
+      "loss": 0.529,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1573939656494636,
+      "learning_rate": 1.4208280664760823e-05,
+      "loss": 0.5374,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8283523390582402,
+      "learning_rate": 1.4205892496789816e-05,
+      "loss": 0.5117,
+      "step": 4688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8123582346827237,
+      "learning_rate": 1.4203504037354192e-05,
+      "loss": 0.5004,
+      "step": 4689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0285817266191963,
+      "learning_rate": 1.4201115286619464e-05,
+      "loss": 0.6204,
+      "step": 4690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8752059281230673,
+      "learning_rate": 1.4198726244751173e-05,
+      "loss": 0.5893,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9099369702348551,
+      "learning_rate": 1.4196336911914878e-05,
+      "loss": 0.5702,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8990956246038558,
+      "learning_rate": 1.419394728827616e-05,
+      "loss": 0.5603,
+      "step": 4693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9622419853620431,
+      "learning_rate": 1.419155737400061e-05,
+      "loss": 0.5678,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8860431919599383,
+      "learning_rate": 1.4189167169253855e-05,
+      "loss": 0.4666,
+      "step": 4695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8895897423278988,
+      "learning_rate": 1.4186776674201533e-05,
+      "loss": 0.5501,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.920359064516808,
+      "learning_rate": 1.4184385889009298e-05,
+      "loss": 0.5521,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8188258847505276,
+      "learning_rate": 1.4181994813842831e-05,
+      "loss": 0.54,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9810026176463054,
+      "learning_rate": 1.4179603448867836e-05,
+      "loss": 0.5882,
+      "step": 4699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9442240534749564,
+      "learning_rate": 1.4177211794250027e-05,
+      "loss": 0.5584,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9776295396882877,
+      "learning_rate": 1.4174819850155148e-05,
+      "loss": 0.5444,
+      "step": 4701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8815069781560079,
+      "learning_rate": 1.4172427616748955e-05,
+      "loss": 0.521,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9598488602858573,
+      "learning_rate": 1.417003509419723e-05,
+      "loss": 0.5725,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.88162059281092,
+      "learning_rate": 1.416764228266577e-05,
+      "loss": 0.5713,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9011421926797095,
+      "learning_rate": 1.4165249182320401e-05,
+      "loss": 0.5621,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8956175379198454,
+      "learning_rate": 1.4162855793326955e-05,
+      "loss": 0.5187,
+      "step": 4706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8036258385169617,
+      "learning_rate": 1.4160462115851292e-05,
+      "loss": 0.5419,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9202754450019062,
+      "learning_rate": 1.4158068150059302e-05,
+      "loss": 0.5333,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.896061684455917,
+      "learning_rate": 1.415567389611687e-05,
+      "loss": 0.5587,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9454485617565561,
+      "learning_rate": 1.4153279354189927e-05,
+      "loss": 0.5858,
+      "step": 4710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9237005211467425,
+      "learning_rate": 1.415088452444441e-05,
+      "loss": 0.527,
+      "step": 4711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9193006748691501,
+      "learning_rate": 1.4148489407046274e-05,
+      "loss": 0.5338,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8406689889011487,
+      "learning_rate": 1.4146094002161501e-05,
+      "loss": 0.5876,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8588860468374754,
+      "learning_rate": 1.4143698309956096e-05,
+      "loss": 0.5727,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9456318467829123,
+      "learning_rate": 1.414130233059607e-05,
+      "loss": 0.5408,
+      "step": 4715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8965301895023075,
+      "learning_rate": 1.4138906064247467e-05,
+      "loss": 0.5793,
+      "step": 4716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9239643757250254,
+      "learning_rate": 1.4136509511076347e-05,
+      "loss": 0.6473,
+      "step": 4717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9022228686323259,
+      "learning_rate": 1.4134112671248783e-05,
+      "loss": 0.4977,
+      "step": 4718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7949190841588801,
+      "learning_rate": 1.4131715544930878e-05,
+      "loss": 0.5347,
+      "step": 4719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9740355679714173,
+      "learning_rate": 1.4129318132288752e-05,
+      "loss": 0.5923,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7708498171790932,
+      "learning_rate": 1.4126920433488542e-05,
+      "loss": 0.5214,
+      "step": 4721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7865502977843916,
+      "learning_rate": 1.4124522448696407e-05,
+      "loss": 0.5338,
+      "step": 4722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8597659959517485,
+      "learning_rate": 1.4122124178078522e-05,
+      "loss": 0.56,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8978989392823783,
+      "learning_rate": 1.4119725621801093e-05,
+      "loss": 0.5677,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9157028799721416,
+      "learning_rate": 1.411732678003033e-05,
+      "loss": 0.6368,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9925066776480296,
+      "learning_rate": 1.411492765293247e-05,
+      "loss": 0.5487,
+      "step": 4726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9113747897282498,
+      "learning_rate": 1.4112528240673779e-05,
+      "loss": 0.5318,
+      "step": 4727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8589278502208879,
+      "learning_rate": 1.4110128543420527e-05,
+      "loss": 0.555,
+      "step": 4728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.950191960101158,
+      "learning_rate": 1.410772856133901e-05,
+      "loss": 0.5285,
+      "step": 4729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9338525505422465,
+      "learning_rate": 1.4105328294595549e-05,
+      "loss": 0.4941,
+      "step": 4730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9615369635942083,
+      "learning_rate": 1.4102927743356481e-05,
+      "loss": 0.6274,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.953732288802404,
+      "learning_rate": 1.4100526907788157e-05,
+      "loss": 0.5737,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0032491436975954,
+      "learning_rate": 1.4098125788056955e-05,
+      "loss": 0.5614,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.889558640015762,
+      "learning_rate": 1.4095724384329272e-05,
+      "loss": 0.5449,
+      "step": 4734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8956467014404496,
+      "learning_rate": 1.4093322696771523e-05,
+      "loss": 0.561,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8551706931065691,
+      "learning_rate": 1.409092072555014e-05,
+      "loss": 0.598,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8548839800087088,
+      "learning_rate": 1.408851847083158e-05,
+      "loss": 0.5593,
+      "step": 4737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8901266047668741,
+      "learning_rate": 1.4086115932782316e-05,
+      "loss": 0.5516,
+      "step": 4738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8561960107371271,
+      "learning_rate": 1.4083713111568841e-05,
+      "loss": 0.5536,
+      "step": 4739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9031494970498228,
+      "learning_rate": 1.4081310007357673e-05,
+      "loss": 0.5372,
+      "step": 4740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9312334703767707,
+      "learning_rate": 1.4078906620315343e-05,
+      "loss": 0.5603,
+      "step": 4741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8599402331636063,
+      "learning_rate": 1.4076502950608397e-05,
+      "loss": 0.544,
+      "step": 4742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8292929383194493,
+      "learning_rate": 1.4074098998403414e-05,
+      "loss": 0.4634,
+      "step": 4743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.815490200853265,
+      "learning_rate": 1.4071694763866988e-05,
+      "loss": 0.538,
+      "step": 4744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.894925411565493,
+      "learning_rate": 1.406929024716572e-05,
+      "loss": 0.6092,
+      "step": 4745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.963522106511423,
+      "learning_rate": 1.4066885448466252e-05,
+      "loss": 0.5206,
+      "step": 4746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8876060513326824,
+      "learning_rate": 1.406448036793523e-05,
+      "loss": 0.5827,
+      "step": 4747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9248326157942703,
+      "learning_rate": 1.406207500573932e-05,
+      "loss": 0.5604,
+      "step": 4748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8549469819980957,
+      "learning_rate": 1.4059669362045216e-05,
+      "loss": 0.5983,
+      "step": 4749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9122826272105434,
+      "learning_rate": 1.4057263437019631e-05,
+      "loss": 0.5094,
+      "step": 4750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.841912287988438,
+      "learning_rate": 1.4054857230829284e-05,
+      "loss": 0.526,
+      "step": 4751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8528346965856147,
+      "learning_rate": 1.4052450743640926e-05,
+      "loss": 0.5588,
+      "step": 4752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9629513938454854,
+      "learning_rate": 1.4050043975621328e-05,
+      "loss": 0.5635,
+      "step": 4753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9684907374115393,
+      "learning_rate": 1.4047636926937278e-05,
+      "loss": 0.5735,
+      "step": 4754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9225498384798198,
+      "learning_rate": 1.4045229597755574e-05,
+      "loss": 0.5369,
+      "step": 4755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8387551385225303,
+      "learning_rate": 1.404282198824305e-05,
+      "loss": 0.5656,
+      "step": 4756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9722381067509114,
+      "learning_rate": 1.4040414098566548e-05,
+      "loss": 0.6086,
+      "step": 4757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8585789480518193,
+      "learning_rate": 1.4038005928892932e-05,
+      "loss": 0.5008,
+      "step": 4758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9809877521334563,
+      "learning_rate": 1.4035597479389088e-05,
+      "loss": 0.5988,
+      "step": 4759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9476898779708328,
+      "learning_rate": 1.4033188750221918e-05,
+      "loss": 0.595,
+      "step": 4760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9190154556053546,
+      "learning_rate": 1.4030779741558345e-05,
+      "loss": 0.5272,
+      "step": 4761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8986826250041086,
+      "learning_rate": 1.402837045356531e-05,
+      "loss": 0.6074,
+      "step": 4762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9669448931268747,
+      "learning_rate": 1.4025960886409777e-05,
+      "loss": 0.5329,
+      "step": 4763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8597820710103147,
+      "learning_rate": 1.4023551040258726e-05,
+      "loss": 0.5572,
+      "step": 4764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9632057604242157,
+      "learning_rate": 1.4021140915279157e-05,
+      "loss": 0.5526,
+      "step": 4765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8111326793433378,
+      "learning_rate": 1.4018730511638087e-05,
+      "loss": 0.4572,
+      "step": 4766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8599006225258523,
+      "learning_rate": 1.4016319829502559e-05,
+      "loss": 0.5486,
+      "step": 4767
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9054438326651731,
+      "learning_rate": 1.4013908869039627e-05,
+      "loss": 0.585,
+      "step": 4768
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.815284889579051,
+      "learning_rate": 1.4011497630416375e-05,
+      "loss": 0.5256,
+      "step": 4769
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0681117562803168,
+      "learning_rate": 1.4009086113799892e-05,
+      "loss": 0.5563,
+      "step": 4770
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9004269746499685,
+      "learning_rate": 1.4006674319357298e-05,
+      "loss": 0.5582,
+      "step": 4771
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9434517262368302,
+      "learning_rate": 1.400426224725573e-05,
+      "loss": 0.5648,
+      "step": 4772
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8505463208839589,
+      "learning_rate": 1.4001849897662337e-05,
+      "loss": 0.5259,
+      "step": 4773
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8630515762899552,
+      "learning_rate": 1.3999437270744296e-05,
+      "loss": 0.5718,
+      "step": 4774
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9791660948704696,
+      "learning_rate": 1.3997024366668802e-05,
+      "loss": 0.6139,
+      "step": 4775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9069769270170313,
+      "learning_rate": 1.3994611185603062e-05,
+      "loss": 0.5569,
+      "step": 4776
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9980723952054046,
+      "learning_rate": 1.399219772771431e-05,
+      "loss": 0.5669,
+      "step": 4777
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8644689973208419,
+      "learning_rate": 1.3989783993169798e-05,
+      "loss": 0.536,
+      "step": 4778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.879292682310879,
+      "learning_rate": 1.3987369982136794e-05,
+      "loss": 0.5866,
+      "step": 4779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9060036514914707,
+      "learning_rate": 1.3984955694782584e-05,
+      "loss": 0.5524,
+      "step": 4780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8978917858742707,
+      "learning_rate": 1.3982541131274485e-05,
+      "loss": 0.5106,
+      "step": 4781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8626340658742684,
+      "learning_rate": 1.3980126291779814e-05,
+      "loss": 0.5919,
+      "step": 4782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8853907390891486,
+      "learning_rate": 1.3977711176465923e-05,
+      "loss": 0.5066,
+      "step": 4783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9719894539559228,
+      "learning_rate": 1.3975295785500176e-05,
+      "loss": 0.6066,
+      "step": 4784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8420668650366185,
+      "learning_rate": 1.3972880119049954e-05,
+      "loss": 0.5569,
+      "step": 4785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0819288026025051,
+      "learning_rate": 1.3970464177282665e-05,
+      "loss": 0.6061,
+      "step": 4786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8661264684608802,
+      "learning_rate": 1.3968047960365733e-05,
+      "loss": 0.529,
+      "step": 4787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9256001465669449,
+      "learning_rate": 1.3965631468466593e-05,
+      "loss": 0.5234,
+      "step": 4788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9031084374098787,
+      "learning_rate": 1.3963214701752714e-05,
+      "loss": 0.5153,
+      "step": 4789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9398534523407612,
+      "learning_rate": 1.396079766039157e-05,
+      "loss": 0.5374,
+      "step": 4790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9908903392902311,
+      "learning_rate": 1.3958380344550659e-05,
+      "loss": 0.5367,
+      "step": 4791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.953551175148032,
+      "learning_rate": 1.3955962754397505e-05,
+      "loss": 0.4969,
+      "step": 4792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.908420529771096,
+      "learning_rate": 1.395354489009964e-05,
+      "loss": 0.5562,
+      "step": 4793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8407068398016209,
+      "learning_rate": 1.3951126751824618e-05,
+      "loss": 0.4925,
+      "step": 4794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0293422320380614,
+      "learning_rate": 1.3948708339740019e-05,
+      "loss": 0.5454,
+      "step": 4795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9554664617714606,
+      "learning_rate": 1.3946289654013435e-05,
+      "loss": 0.6564,
+      "step": 4796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0877753621417259,
+      "learning_rate": 1.3943870694812475e-05,
+      "loss": 0.5433,
+      "step": 4797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8856524373788587,
+      "learning_rate": 1.3941451462304778e-05,
+      "loss": 0.553,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9286230008806331,
+      "learning_rate": 1.393903195665799e-05,
+      "loss": 0.5859,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9510367799250085,
+      "learning_rate": 1.393661217803978e-05,
+      "loss": 0.5666,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9448143716639735,
+      "learning_rate": 1.3934192126617838e-05,
+      "loss": 0.5982,
+      "step": 4801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9379935538051243,
+      "learning_rate": 1.3931771802559875e-05,
+      "loss": 0.5737,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7776618529565054,
+      "learning_rate": 1.3929351206033607e-05,
+      "loss": 0.4976,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8568590847033114,
+      "learning_rate": 1.392693033720679e-05,
+      "loss": 0.4904,
+      "step": 4804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8834314638024736,
+      "learning_rate": 1.3924509196247185e-05,
+      "loss": 0.579,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8940184195213572,
+      "learning_rate": 1.392208778332257e-05,
+      "loss": 0.6216,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9112868877658168,
+      "learning_rate": 1.3919666098600753e-05,
+      "loss": 0.5378,
+      "step": 4807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9216500150943909,
+      "learning_rate": 1.3917244142249551e-05,
+      "loss": 0.5763,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.893654843861322,
+      "learning_rate": 1.3914821914436805e-05,
+      "loss": 0.5533,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8614911309355482,
+      "learning_rate": 1.3912399415330371e-05,
+      "loss": 0.5161,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9230499012788027,
+      "learning_rate": 1.3909976645098131e-05,
+      "loss": 0.5516,
+      "step": 4811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9496706748160366,
+      "learning_rate": 1.3907553603907974e-05,
+      "loss": 0.5344,
+      "step": 4812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8650379902912981,
+      "learning_rate": 1.3905130291927822e-05,
+      "loss": 0.5882,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.998972087470307,
+      "learning_rate": 1.3902706709325603e-05,
+      "loss": 0.5186,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8508144263902477,
+      "learning_rate": 1.3900282856269271e-05,
+      "loss": 0.5418,
+      "step": 4815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0025157638908624,
+      "learning_rate": 1.3897858732926794e-05,
+      "loss": 0.5818,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0516850697205595,
+      "learning_rate": 1.3895434339466167e-05,
+      "loss": 0.6107,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9658348227769601,
+      "learning_rate": 1.3893009676055395e-05,
+      "loss": 0.6161,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0289714416989588,
+      "learning_rate": 1.3890584742862508e-05,
+      "loss": 0.5631,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9460224670710317,
+      "learning_rate": 1.3888159540055544e-05,
+      "loss": 0.5635,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9800985935542305,
+      "learning_rate": 1.3885734067802576e-05,
+      "loss": 0.5972,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9354319412514424,
+      "learning_rate": 1.3883308326271682e-05,
+      "loss": 0.546,
+      "step": 4822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9247901332190117,
+      "learning_rate": 1.3880882315630968e-05,
+      "loss": 0.6057,
+      "step": 4823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8687178724760034,
+      "learning_rate": 1.387845603604855e-05,
+      "loss": 0.5477,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9721356384986916,
+      "learning_rate": 1.387602948769257e-05,
+      "loss": 0.5753,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9264169462238878,
+      "learning_rate": 1.3873602670731184e-05,
+      "loss": 0.5709,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8798247430588303,
+      "learning_rate": 1.387117558533257e-05,
+      "loss": 0.5449,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.027001023657605,
+      "learning_rate": 1.3868748231664918e-05,
+      "loss": 0.5988,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8553348754174525,
+      "learning_rate": 1.3866320609896449e-05,
+      "loss": 0.5224,
+      "step": 4829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9102274530812277,
+      "learning_rate": 1.3863892720195389e-05,
+      "loss": 0.6143,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8638981385026545,
+      "learning_rate": 1.3861464562729992e-05,
+      "loss": 0.4728,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8269196063569016,
+      "learning_rate": 1.3859036137668525e-05,
+      "loss": 0.5418,
+      "step": 4832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0036783596030001,
+      "learning_rate": 1.3856607445179278e-05,
+      "loss": 0.6036,
+      "step": 4833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9065362519579458,
+      "learning_rate": 1.3854178485430554e-05,
+      "loss": 0.5078,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.048824179044158,
+      "learning_rate": 1.3851749258590679e-05,
+      "loss": 0.6005,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8302691187789433,
+      "learning_rate": 1.3849319764828e-05,
+      "loss": 0.4975,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9031582569220298,
+      "learning_rate": 1.3846890004310873e-05,
+      "loss": 0.5631,
+      "step": 4837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9177195645049511,
+      "learning_rate": 1.3844459977207683e-05,
+      "loss": 0.5383,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.949269225994463,
+      "learning_rate": 1.3842029683686826e-05,
+      "loss": 0.5748,
+      "step": 4839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.937659934137604,
+      "learning_rate": 1.3839599123916718e-05,
+      "loss": 0.5762,
+      "step": 4840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8996926326469921,
+      "learning_rate": 1.3837168298065798e-05,
+      "loss": 0.61,
+      "step": 4841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.94884621874138,
+      "learning_rate": 1.3834737206302519e-05,
+      "loss": 0.6057,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.923036476443756,
+      "learning_rate": 1.3832305848795352e-05,
+      "loss": 0.5301,
+      "step": 4843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8541275293194943,
+      "learning_rate": 1.382987422571279e-05,
+      "loss": 0.5602,
+      "step": 4844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9501069778376648,
+      "learning_rate": 1.382744233722334e-05,
+      "loss": 0.5255,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9630480388424695,
+      "learning_rate": 1.382501018349553e-05,
+      "loss": 0.5897,
+      "step": 4846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8762902360405572,
+      "learning_rate": 1.3822577764697908e-05,
+      "loss": 0.5017,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.043217051840068,
+      "learning_rate": 1.3820145080999038e-05,
+      "loss": 0.575,
+      "step": 4848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9439300127928049,
+      "learning_rate": 1.3817712132567503e-05,
+      "loss": 0.5904,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8881934762338176,
+      "learning_rate": 1.3815278919571901e-05,
+      "loss": 0.5236,
+      "step": 4850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9482363818913019,
+      "learning_rate": 1.3812845442180857e-05,
+      "loss": 0.6188,
+      "step": 4851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0714832359599773,
+      "learning_rate": 1.3810411700563005e-05,
+      "loss": 0.5725,
+      "step": 4852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9972613620841716,
+      "learning_rate": 1.3807977694887003e-05,
+      "loss": 0.6182,
+      "step": 4853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9624229426930314,
+      "learning_rate": 1.3805543425321524e-05,
+      "loss": 0.5063,
+      "step": 4854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8981685605321771,
+      "learning_rate": 1.3803108892035259e-05,
+      "loss": 0.522,
+      "step": 4855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8918230664294677,
+      "learning_rate": 1.3800674095196922e-05,
+      "loss": 0.5192,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.021530752522361,
+      "learning_rate": 1.3798239034975243e-05,
+      "loss": 0.5427,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8679200195770674,
+      "learning_rate": 1.3795803711538966e-05,
+      "loss": 0.5067,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0076351564977803,
+      "learning_rate": 1.3793368125056859e-05,
+      "loss": 0.6147,
+      "step": 4859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9477492129643318,
+      "learning_rate": 1.3790932275697708e-05,
+      "loss": 0.5777,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9912531451536267,
+      "learning_rate": 1.378849616363031e-05,
+      "loss": 0.5575,
+      "step": 4861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.008150805432256,
+      "learning_rate": 1.3786059789023487e-05,
+      "loss": 0.5991,
+      "step": 4862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.967448144120769,
+      "learning_rate": 1.3783623152046084e-05,
+      "loss": 0.5237,
+      "step": 4863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0518336967776385,
+      "learning_rate": 1.3781186252866948e-05,
+      "loss": 0.5971,
+      "step": 4864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9082645655876715,
+      "learning_rate": 1.377874909165496e-05,
+      "loss": 0.5647,
+      "step": 4865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9465929099003216,
+      "learning_rate": 1.3776311668579012e-05,
+      "loss": 0.5546,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8694075318857712,
+      "learning_rate": 1.3773873983808014e-05,
+      "loss": 0.5255,
+      "step": 4867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9269426482677092,
+      "learning_rate": 1.3771436037510897e-05,
+      "loss": 0.5775,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9833980689058247,
+      "learning_rate": 1.3768997829856608e-05,
+      "loss": 0.5893,
+      "step": 4869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9336236085410238,
+      "learning_rate": 1.3766559361014113e-05,
+      "loss": 0.5903,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8591684244273873,
+      "learning_rate": 1.3764120631152395e-05,
+      "loss": 0.5636,
+      "step": 4871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.896987510351119,
+      "learning_rate": 1.3761681640440455e-05,
+      "loss": 0.5898,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9260583363210217,
+      "learning_rate": 1.3759242389047315e-05,
+      "loss": 0.5162,
+      "step": 4873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9281921574947163,
+      "learning_rate": 1.375680287714201e-05,
+      "loss": 0.6638,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9548686164779019,
+      "learning_rate": 1.37543631048936e-05,
+      "loss": 0.5717,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.939319463753251,
+      "learning_rate": 1.3751923072471159e-05,
+      "loss": 0.6109,
+      "step": 4876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9370066407879878,
+      "learning_rate": 1.3749482780043773e-05,
+      "loss": 0.4715,
+      "step": 4877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9853199068327506,
+      "learning_rate": 1.3747042227780557e-05,
+      "loss": 0.5326,
+      "step": 4878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9239784176479834,
+      "learning_rate": 1.3744601415850637e-05,
+      "loss": 0.5714,
+      "step": 4879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.910935042749798,
+      "learning_rate": 1.3742160344423164e-05,
+      "loss": 0.5666,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9811225149068671,
+      "learning_rate": 1.3739719013667297e-05,
+      "loss": 0.6076,
+      "step": 4881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9149100233787881,
+      "learning_rate": 1.3737277423752218e-05,
+      "loss": 0.5563,
+      "step": 4882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9053166838456309,
+      "learning_rate": 1.373483557484713e-05,
+      "loss": 0.4929,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7872734502680776,
+      "learning_rate": 1.3732393467121247e-05,
+      "loss": 0.4891,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8742286431117344,
+      "learning_rate": 1.372995110074381e-05,
+      "loss": 0.5329,
+      "step": 4885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8672043564902082,
+      "learning_rate": 1.3727508475884071e-05,
+      "loss": 0.5895,
+      "step": 4886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8414643393353751,
+      "learning_rate": 1.3725065592711299e-05,
+      "loss": 0.4682,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8875840210510944,
+      "learning_rate": 1.3722622451394784e-05,
+      "loss": 0.6033,
+      "step": 4888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8779128007665025,
+      "learning_rate": 1.3720179052103836e-05,
+      "loss": 0.5105,
+      "step": 4889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9046433815894597,
+      "learning_rate": 1.3717735395007786e-05,
+      "loss": 0.4746,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9050851117821455,
+      "learning_rate": 1.3715291480275963e-05,
+      "loss": 0.5518,
+      "step": 4891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9681695345357252,
+      "learning_rate": 1.3712847308077737e-05,
+      "loss": 0.5641,
+      "step": 4892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9480998717859649,
+      "learning_rate": 1.3710402878582487e-05,
+      "loss": 0.5638,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.010665403165591,
+      "learning_rate": 1.3707958191959609e-05,
+      "loss": 0.5386,
+      "step": 4894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9721875042908399,
+      "learning_rate": 1.3705513248378517e-05,
+      "loss": 0.605,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8790064569896765,
+      "learning_rate": 1.3703068048008645e-05,
+      "loss": 0.4673,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8314633413004586,
+      "learning_rate": 1.3700622591019439e-05,
+      "loss": 0.5181,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8243806242658588,
+      "learning_rate": 1.3698176877580372e-05,
+      "loss": 0.5087,
+      "step": 4898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.889784596263977,
+      "learning_rate": 1.3695730907860925e-05,
+      "loss": 0.5393,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9543436360355445,
+      "learning_rate": 1.3693284682030608e-05,
+      "loss": 0.5735,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8393976604033532,
+      "learning_rate": 1.3690838200258936e-05,
+      "loss": 0.5944,
+      "step": 4901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0040665844972214,
+      "learning_rate": 1.368839146271545e-05,
+      "loss": 0.6325,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8661069998555928,
+      "learning_rate": 1.368594446956971e-05,
+      "loss": 0.549,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.980762907582011,
+      "learning_rate": 1.3683497220991286e-05,
+      "loss": 0.5668,
+      "step": 4904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8303767586905748,
+      "learning_rate": 1.3681049717149773e-05,
+      "loss": 0.5261,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0368179040107803,
+      "learning_rate": 1.3678601958214779e-05,
+      "loss": 0.527,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9249813464997794,
+      "learning_rate": 1.367615394435593e-05,
+      "loss": 0.6134,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8358321020913861,
+      "learning_rate": 1.3673705675742875e-05,
+      "loss": 0.5125,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0496005259301522,
+      "learning_rate": 1.3671257152545277e-05,
+      "loss": 0.6041,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.946126940263957,
+      "learning_rate": 1.3668808374932812e-05,
+      "loss": 0.572,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.848440255461486,
+      "learning_rate": 1.3666359343075182e-05,
+      "loss": 0.5319,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8872846916857009,
+      "learning_rate": 1.3663910057142102e-05,
+      "loss": 0.5519,
+      "step": 4912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0799104995159567,
+      "learning_rate": 1.3661460517303304e-05,
+      "loss": 0.6399,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8940584447880352,
+      "learning_rate": 1.3659010723728542e-05,
+      "loss": 0.5479,
+      "step": 4914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9056599511334902,
+      "learning_rate": 1.3656560676587583e-05,
+      "loss": 0.623,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8597632452723915,
+      "learning_rate": 1.3654110376050209e-05,
+      "loss": 0.5268,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9576599634253601,
+      "learning_rate": 1.3651659822286227e-05,
+      "loss": 0.5411,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9237904664858961,
+      "learning_rate": 1.364920901546546e-05,
+      "loss": 0.5251,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8414244460916807,
+      "learning_rate": 1.3646757955757746e-05,
+      "loss": 0.533,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9113629315058073,
+      "learning_rate": 1.3644306643332939e-05,
+      "loss": 0.5311,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.903990767127771,
+      "learning_rate": 1.3641855078360914e-05,
+      "loss": 0.5323,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9785840692406184,
+      "learning_rate": 1.3639403261011563e-05,
+      "loss": 0.5502,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9387542745252239,
+      "learning_rate": 1.3636951191454792e-05,
+      "loss": 0.5561,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.899595616734336,
+      "learning_rate": 1.3634498869860533e-05,
+      "loss": 0.4931,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8787075359867942,
+      "learning_rate": 1.3632046296398724e-05,
+      "loss": 0.5138,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8436660881527684,
+      "learning_rate": 1.3629593471239328e-05,
+      "loss": 0.5069,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9229983052966406,
+      "learning_rate": 1.3627140394552326e-05,
+      "loss": 0.5417,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9687480936383686,
+      "learning_rate": 1.3624687066507709e-05,
+      "loss": 0.6037,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.862557622655597,
+      "learning_rate": 1.3622233487275493e-05,
+      "loss": 0.5621,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8746701383679512,
+      "learning_rate": 1.3619779657025714e-05,
+      "loss": 0.574,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9638263263332979,
+      "learning_rate": 1.3617325575928414e-05,
+      "loss": 0.5793,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8753934118629156,
+      "learning_rate": 1.3614871244153655e-05,
+      "loss": 0.506,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9135589434822865,
+      "learning_rate": 1.3612416661871532e-05,
+      "loss": 0.5127,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9604594234952577,
+      "learning_rate": 1.3609961829252133e-05,
+      "loss": 0.5363,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.028473474567092,
+      "learning_rate": 1.3607506746465584e-05,
+      "loss": 0.5779,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.902538042974246,
+      "learning_rate": 1.360505141368202e-05,
+      "loss": 0.5267,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8992938909760883,
+      "learning_rate": 1.3602595831071586e-05,
+      "loss": 0.5842,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8780589847799097,
+      "learning_rate": 1.3600139998804459e-05,
+      "loss": 0.5513,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9142510765861755,
+      "learning_rate": 1.359768391705082e-05,
+      "loss": 0.5406,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9567959566085759,
+      "learning_rate": 1.3595227585980881e-05,
+      "loss": 0.5539,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8807388652451061,
+      "learning_rate": 1.3592771005764857e-05,
+      "loss": 0.5587,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8651113785797417,
+      "learning_rate": 1.3590314176572989e-05,
+      "loss": 0.5546,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8171458610105073,
+      "learning_rate": 1.3587857098575534e-05,
+      "loss": 0.5267,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0893608287859775,
+      "learning_rate": 1.3585399771942764e-05,
+      "loss": 0.5818,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9605917712592748,
+      "learning_rate": 1.358294219684497e-05,
+      "loss": 0.5289,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7972262798819203,
+      "learning_rate": 1.3580484373452462e-05,
+      "loss": 0.5292,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9419043795441053,
+      "learning_rate": 1.357802630193556e-05,
+      "loss": 0.5731,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8941534978079186,
+      "learning_rate": 1.357556798246461e-05,
+      "loss": 0.51,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9127635202406887,
+      "learning_rate": 1.357310941520997e-05,
+      "loss": 0.5553,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9002276388779009,
+      "learning_rate": 1.3570650600342017e-05,
+      "loss": 0.5921,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8717763667596604,
+      "learning_rate": 1.3568191538031146e-05,
+      "loss": 0.5518,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8191098322900104,
+      "learning_rate": 1.3565732228447766e-05,
+      "loss": 0.4965,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9373761418635452,
+      "learning_rate": 1.3563272671762304e-05,
+      "loss": 0.5527,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9116380184176788,
+      "learning_rate": 1.3560812868145206e-05,
+      "loss": 0.5502,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9782949235050161,
+      "learning_rate": 1.3558352817766935e-05,
+      "loss": 0.5888,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8729097488069214,
+      "learning_rate": 1.355589252079797e-05,
+      "loss": 0.585,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.92953206689005,
+      "learning_rate": 1.3553431977408809e-05,
+      "loss": 0.6116,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9161281687717558,
+      "learning_rate": 1.3550971187769964e-05,
+      "loss": 0.582,
+      "step": 4958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9544317740681652,
+      "learning_rate": 1.3548510152051963e-05,
+      "loss": 0.5308,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9300834006161792,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.5908,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8682564450672268,
+      "learning_rate": 1.3543587343060712e-05,
+      "loss": 0.5659,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8507907896472288,
+      "learning_rate": 1.3541125570128603e-05,
+      "loss": 0.5334,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8841622079095046,
+      "learning_rate": 1.3538663551799636e-05,
+      "loss": 0.6271,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8433659287109949,
+      "learning_rate": 1.3536201288244425e-05,
+      "loss": 0.5682,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9244820516318176,
+      "learning_rate": 1.3533738779633597e-05,
+      "loss": 0.5897,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7981635507567023,
+      "learning_rate": 1.3531276026137807e-05,
+      "loss": 0.4975,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.234573810119623,
+      "learning_rate": 1.3528813027927724e-05,
+      "loss": 0.56,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9720942100727602,
+      "learning_rate": 1.3526349785174025e-05,
+      "loss": 0.5927,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9249900333619757,
+      "learning_rate": 1.3523886298047412e-05,
+      "loss": 0.5049,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8817644529264543,
+      "learning_rate": 1.3521422566718609e-05,
+      "loss": 0.5317,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8806141034443261,
+      "learning_rate": 1.3518958591358345e-05,
+      "loss": 0.4908,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9637460271329971,
+      "learning_rate": 1.3516494372137368e-05,
+      "loss": 0.5397,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9061409021519361,
+      "learning_rate": 1.3514029909226454e-05,
+      "loss": 0.5826,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9483299745746715,
+      "learning_rate": 1.3511565202796381e-05,
+      "loss": 0.5759,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9451773146461331,
+      "learning_rate": 1.3509100253017958e-05,
+      "loss": 0.5814,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8567772541508121,
+      "learning_rate": 1.3506635060062e-05,
+      "loss": 0.5104,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8383479350018068,
+      "learning_rate": 1.350416962409934e-05,
+      "loss": 0.5064,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8912113016113629,
+      "learning_rate": 1.3501703945300832e-05,
+      "loss": 0.5262,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9870956538767125,
+      "learning_rate": 1.349923802383735e-05,
+      "loss": 0.6047,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.977475323915746,
+      "learning_rate": 1.3496771859879774e-05,
+      "loss": 0.5306,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9128506647381485,
+      "learning_rate": 1.349430545359901e-05,
+      "loss": 0.5356,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9323166542632595,
+      "learning_rate": 1.349183880516598e-05,
+      "loss": 0.5505,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9276052946168218,
+      "learning_rate": 1.3489371914751616e-05,
+      "loss": 0.5412,
+      "step": 4983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9888331579532674,
+      "learning_rate": 1.3486904782526876e-05,
+      "loss": 0.5327,
+      "step": 4984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8900191506817302,
+      "learning_rate": 1.3484437408662725e-05,
+      "loss": 0.5338,
+      "step": 4985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8606918204173645,
+      "learning_rate": 1.3481969793330151e-05,
+      "loss": 0.5094,
+      "step": 4986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9423971372459556,
+      "learning_rate": 1.3479501936700161e-05,
+      "loss": 0.5849,
+      "step": 4987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9037119495482759,
+      "learning_rate": 1.3477033838943774e-05,
+      "loss": 0.5204,
+      "step": 4988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0464546656121712,
+      "learning_rate": 1.3474565500232025e-05,
+      "loss": 0.6279,
+      "step": 4989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0047842505728581,
+      "learning_rate": 1.3472096920735966e-05,
+      "loss": 0.6017,
+      "step": 4990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9207006697256435,
+      "learning_rate": 1.3469628100626678e-05,
+      "loss": 0.5511,
+      "step": 4991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9042762572590564,
+      "learning_rate": 1.3467159040075233e-05,
+      "loss": 0.5697,
+      "step": 4992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9258256902752167,
+      "learning_rate": 1.3464689739252741e-05,
+      "loss": 0.5919,
+      "step": 4993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8510889422046063,
+      "learning_rate": 1.346222019833033e-05,
+      "loss": 0.5575,
+      "step": 4994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.923790834710621,
+      "learning_rate": 1.3459750417479125e-05,
+      "loss": 0.5767,
+      "step": 4995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9655024203049409,
+      "learning_rate": 1.3457280396870285e-05,
+      "loss": 0.5502,
+      "step": 4996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8978053659559004,
+      "learning_rate": 1.3454810136674983e-05,
+      "loss": 0.5429,
+      "step": 4997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9173149006811787,
+      "learning_rate": 1.34523396370644e-05,
+      "loss": 0.5388,
+      "step": 4998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8803052233081673,
+      "learning_rate": 1.3449868898209743e-05,
+      "loss": 0.5308,
+      "step": 4999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9213276267952032,
+      "learning_rate": 1.3447397920282232e-05,
+      "loss": 0.5615,
+      "step": 5000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0679262404551646,
+      "learning_rate": 1.3444926703453102e-05,
+      "loss": 0.5619,
+      "step": 5001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8507200211182417,
+      "learning_rate": 1.3442455247893608e-05,
+      "loss": 0.5086,
+      "step": 5002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9397594122523698,
+      "learning_rate": 1.3439983553775018e-05,
+      "loss": 0.5929,
+      "step": 5003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8575371004987608,
+      "learning_rate": 1.3437511621268622e-05,
+      "loss": 0.4855,
+      "step": 5004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8893823279593794,
+      "learning_rate": 1.3435039450545718e-05,
+      "loss": 0.5046,
+      "step": 5005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9879627215267409,
+      "learning_rate": 1.3432567041777624e-05,
+      "loss": 0.5308,
+      "step": 5006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8223979075210116,
+      "learning_rate": 1.3430094395135682e-05,
+      "loss": 0.4693,
+      "step": 5007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0213441059322579,
+      "learning_rate": 1.342762151079124e-05,
+      "loss": 0.5125,
+      "step": 5008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8893830497058866,
+      "learning_rate": 1.3425148388915668e-05,
+      "loss": 0.529,
+      "step": 5009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8453904170689711,
+      "learning_rate": 1.3422675029680352e-05,
+      "loss": 0.5243,
+      "step": 5010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9235327456897732,
+      "learning_rate": 1.342020143325669e-05,
+      "loss": 0.5161,
+      "step": 5011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.934246128185342,
+      "learning_rate": 1.3417727599816101e-05,
+      "loss": 0.5572,
+      "step": 5012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.843513392770215,
+      "learning_rate": 1.3415253529530026e-05,
+      "loss": 0.5877,
+      "step": 5013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8979152627392312,
+      "learning_rate": 1.3412779222569907e-05,
+      "loss": 0.5576,
+      "step": 5014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9027225679116367,
+      "learning_rate": 1.3410304679107214e-05,
+      "loss": 0.5798,
+      "step": 5015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9785640983126853,
+      "learning_rate": 1.3407829899313435e-05,
+      "loss": 0.4948,
+      "step": 5016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9026415340866438,
+      "learning_rate": 1.3405354883360064e-05,
+      "loss": 0.6247,
+      "step": 5017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9659391475419029,
+      "learning_rate": 1.3402879631418621e-05,
+      "loss": 0.5754,
+      "step": 5018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.909795574321967,
+      "learning_rate": 1.3400404143660639e-05,
+      "loss": 0.5855,
+      "step": 5019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9365576954027569,
+      "learning_rate": 1.3397928420257664e-05,
+      "loss": 0.572,
+      "step": 5020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9481969842800917,
+      "learning_rate": 1.3395452461381265e-05,
+      "loss": 0.5493,
+      "step": 5021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9159633447540382,
+      "learning_rate": 1.3392976267203024e-05,
+      "loss": 0.6018,
+      "step": 5022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.929746076040423,
+      "learning_rate": 1.3390499837894533e-05,
+      "loss": 0.5801,
+      "step": 5023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9262633274121936,
+      "learning_rate": 1.3388023173627413e-05,
+      "loss": 0.5262,
+      "step": 5024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9079531784297293,
+      "learning_rate": 1.3385546274573294e-05,
+      "loss": 0.5131,
+      "step": 5025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.161142792927715,
+      "learning_rate": 1.3383069140903816e-05,
+      "loss": 0.5189,
+      "step": 5026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9383572230897598,
+      "learning_rate": 1.3380591772790652e-05,
+      "loss": 0.5646,
+      "step": 5027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8725547375928622,
+      "learning_rate": 1.3378114170405473e-05,
+      "loss": 0.536,
+      "step": 5028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9961487700807296,
+      "learning_rate": 1.3375636333919981e-05,
+      "loss": 0.5026,
+      "step": 5029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9560730427420159,
+      "learning_rate": 1.3373158263505886e-05,
+      "loss": 0.5708,
+      "step": 5030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.889332135064333,
+      "learning_rate": 1.3370679959334911e-05,
+      "loss": 0.4974,
+      "step": 5031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9112265983457682,
+      "learning_rate": 1.336820142157881e-05,
+      "loss": 0.5586,
+      "step": 5032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0140920595535092,
+      "learning_rate": 1.3365722650409336e-05,
+      "loss": 0.6051,
+      "step": 5033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8265136838840981,
+      "learning_rate": 1.3363243645998265e-05,
+      "loss": 0.5112,
+      "step": 5034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8664416233893417,
+      "learning_rate": 1.3360764408517398e-05,
+      "loss": 0.5771,
+      "step": 5035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8928211824327998,
+      "learning_rate": 1.3358284938138532e-05,
+      "loss": 0.552,
+      "step": 5036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8708038249180292,
+      "learning_rate": 1.3355805235033503e-05,
+      "loss": 0.5487,
+      "step": 5037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8648901980307919,
+      "learning_rate": 1.3353325299374147e-05,
+      "loss": 0.5156,
+      "step": 5038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9253251558176558,
+      "learning_rate": 1.3350845131332322e-05,
+      "loss": 0.5946,
+      "step": 5039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9087741111085298,
+      "learning_rate": 1.33483647310799e-05,
+      "loss": 0.5366,
+      "step": 5040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9031770607268709,
+      "learning_rate": 1.3345884098788775e-05,
+      "loss": 0.5449,
+      "step": 5041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9048945632273513,
+      "learning_rate": 1.334340323463085e-05,
+      "loss": 0.5902,
+      "step": 5042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.928909059359663,
+      "learning_rate": 1.3340922138778042e-05,
+      "loss": 0.5314,
+      "step": 5043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9285281521455355,
+      "learning_rate": 1.3338440811402298e-05,
+      "loss": 0.5399,
+      "step": 5044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0519660440082719,
+      "learning_rate": 1.3335959252675566e-05,
+      "loss": 0.6227,
+      "step": 5045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8580949517974473,
+      "learning_rate": 1.3333477462769814e-05,
+      "loss": 0.5642,
+      "step": 5046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9121537302597602,
+      "learning_rate": 1.333099544185703e-05,
+      "loss": 0.5169,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8995134871690074,
+      "learning_rate": 1.332851319010922e-05,
+      "loss": 0.5968,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9135632096680434,
+      "learning_rate": 1.3326030707698399e-05,
+      "loss": 0.5906,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9046364651883315,
+      "learning_rate": 1.3323547994796597e-05,
+      "loss": 0.6055,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9203931898005878,
+      "learning_rate": 1.3321065051575868e-05,
+      "loss": 0.5704,
+      "step": 5051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8673245017777986,
+      "learning_rate": 1.3318581878208279e-05,
+      "loss": 0.4961,
+      "step": 5052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9076849454851069,
+      "learning_rate": 1.3316098474865905e-05,
+      "loss": 0.4916,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9212244712246338,
+      "learning_rate": 1.331361484172085e-05,
+      "loss": 0.5526,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.046665168224382,
+      "learning_rate": 1.3311130978945228e-05,
+      "loss": 0.5865,
+      "step": 5055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8503823922686983,
+      "learning_rate": 1.3308646886711163e-05,
+      "loss": 0.449,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9493204815760115,
+      "learning_rate": 1.3306162565190805e-05,
+      "loss": 0.6048,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9107820615188169,
+      "learning_rate": 1.3303678014556316e-05,
+      "loss": 0.5662,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9714108115018177,
+      "learning_rate": 1.3301193234979865e-05,
+      "loss": 0.5288,
+      "step": 5059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9419033164293819,
+      "learning_rate": 1.3298708226633657e-05,
+      "loss": 0.6086,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.907506779452863,
+      "learning_rate": 1.3296222989689892e-05,
+      "loss": 0.549,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9435728748431338,
+      "learning_rate": 1.3293737524320798e-05,
+      "loss": 0.6021,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9155414972518737,
+      "learning_rate": 1.3291251830698615e-05,
+      "loss": 0.5394,
+      "step": 5063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.016370420589111,
+      "learning_rate": 1.3288765908995598e-05,
+      "loss": 0.5631,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9096054349595017,
+      "learning_rate": 1.3286279759384022e-05,
+      "loss": 0.543,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8356030266153411,
+      "learning_rate": 1.3283793382036175e-05,
+      "loss": 0.5312,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9494496729847453,
+      "learning_rate": 1.3281306777124356e-05,
+      "loss": 0.5967,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0309264218563654,
+      "learning_rate": 1.3278819944820893e-05,
+      "loss": 0.5769,
+      "step": 5068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8414918365636198,
+      "learning_rate": 1.327633288529811e-05,
+      "loss": 0.5492,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8900427211041783,
+      "learning_rate": 1.3273845598728367e-05,
+      "loss": 0.5579,
+      "step": 5070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9043752020683701,
+      "learning_rate": 1.3271358085284029e-05,
+      "loss": 0.53,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9876593193228603,
+      "learning_rate": 1.3268870345137476e-05,
+      "loss": 0.6066,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9081892965471735,
+      "learning_rate": 1.3266382378461109e-05,
+      "loss": 0.5406,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8612314037801095,
+      "learning_rate": 1.3263894185427339e-05,
+      "loss": 0.5397,
+      "step": 5074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9374695390411759,
+      "learning_rate": 1.3261405766208598e-05,
+      "loss": 0.4895,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9984321444647333,
+      "learning_rate": 1.3258917120977327e-05,
+      "loss": 0.6127,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9179346377204336,
+      "learning_rate": 1.3256428249905998e-05,
+      "loss": 0.5219,
+      "step": 5077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9518656711900425,
+      "learning_rate": 1.3253939153167072e-05,
+      "loss": 0.6394,
+      "step": 5078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8958542762619648,
+      "learning_rate": 1.3251449830933052e-05,
+      "loss": 0.5355,
+      "step": 5079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9534508359693109,
+      "learning_rate": 1.3248960283376441e-05,
+      "loss": 0.579,
+      "step": 5080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9344960582822963,
+      "learning_rate": 1.3246470510669766e-05,
+      "loss": 0.5403,
+      "step": 5081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9374250380548944,
+      "learning_rate": 1.3243980512985563e-05,
+      "loss": 0.5178,
+      "step": 5082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8228686351928819,
+      "learning_rate": 1.3241490290496391e-05,
+      "loss": 0.5492,
+      "step": 5083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8978557739748596,
+      "learning_rate": 1.3238999843374814e-05,
+      "loss": 0.5082,
+      "step": 5084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9340313155095844,
+      "learning_rate": 1.323650917179342e-05,
+      "loss": 0.5216,
+      "step": 5085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8916719591757329,
+      "learning_rate": 1.3234018275924814e-05,
+      "loss": 0.5377,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0150853993724218,
+      "learning_rate": 1.3231527155941607e-05,
+      "loss": 0.6204,
+      "step": 5087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9273769401626929,
+      "learning_rate": 1.3229035812016438e-05,
+      "loss": 0.5609,
+      "step": 5088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9169044311151517,
+      "learning_rate": 1.322654424432195e-05,
+      "loss": 0.5639,
+      "step": 5089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9112610005312264,
+      "learning_rate": 1.3224052453030806e-05,
+      "loss": 0.5845,
+      "step": 5090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9805041567073909,
+      "learning_rate": 1.3221560438315689e-05,
+      "loss": 0.6059,
+      "step": 5091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9047437596079848,
+      "learning_rate": 1.3219068200349292e-05,
+      "loss": 0.5612,
+      "step": 5092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9327788511751683,
+      "learning_rate": 1.321657573930432e-05,
+      "loss": 0.5585,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.906266310194228,
+      "learning_rate": 1.3214083055353504e-05,
+      "loss": 0.5256,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9780421675103058,
+      "learning_rate": 1.3211590148669586e-05,
+      "loss": 0.5435,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8590529244903038,
+      "learning_rate": 1.3209097019425317e-05,
+      "loss": 0.5427,
+      "step": 5096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.871310828995196,
+      "learning_rate": 1.3206603667793472e-05,
+      "loss": 0.5254,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.889387767827606,
+      "learning_rate": 1.3204110093946835e-05,
+      "loss": 0.4817,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8515472409039916,
+      "learning_rate": 1.3201616298058214e-05,
+      "loss": 0.5029,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9682710945847225,
+      "learning_rate": 1.3199122280300418e-05,
+      "loss": 0.6037,
+      "step": 5100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9251328046898974,
+      "learning_rate": 1.319662804084629e-05,
+      "loss": 0.5449,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8627728010761364,
+      "learning_rate": 1.3194133579868672e-05,
+      "loss": 0.537,
+      "step": 5102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8675483717514128,
+      "learning_rate": 1.319163889754043e-05,
+      "loss": 0.5234,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8891470238901393,
+      "learning_rate": 1.3189143994034448e-05,
+      "loss": 0.5632,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.878464155700256,
+      "learning_rate": 1.318664886952361e-05,
+      "loss": 0.483,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9519110168495648,
+      "learning_rate": 1.3184153524180837e-05,
+      "loss": 0.5674,
+      "step": 5106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8828085357873369,
+      "learning_rate": 1.3181657958179046e-05,
+      "loss": 0.5326,
+      "step": 5107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0078717687703735,
+      "learning_rate": 1.317916217169118e-05,
+      "loss": 0.6089,
+      "step": 5108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8520385880997858,
+      "learning_rate": 1.3176666164890195e-05,
+      "loss": 0.5646,
+      "step": 5109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.871313113477867,
+      "learning_rate": 1.3174169937949066e-05,
+      "loss": 0.4969,
+      "step": 5110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8740222761172795,
+      "learning_rate": 1.3171673491040772e-05,
+      "loss": 0.5236,
+      "step": 5111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9805578927608394,
+      "learning_rate": 1.3169176824338321e-05,
+      "loss": 0.5786,
+      "step": 5112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8763258975337161,
+      "learning_rate": 1.3166679938014728e-05,
+      "loss": 0.5381,
+      "step": 5113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8597644001374837,
+      "learning_rate": 1.316418283224302e-05,
+      "loss": 0.5418,
+      "step": 5114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.97008825068724,
+      "learning_rate": 1.3161685507196251e-05,
+      "loss": 0.5815,
+      "step": 5115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9828993498197132,
+      "learning_rate": 1.3159187963047481e-05,
+      "loss": 0.571,
+      "step": 5116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8992665461979235,
+      "learning_rate": 1.3156690199969786e-05,
+      "loss": 0.562,
+      "step": 5117
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9991378355301007,
+      "learning_rate": 1.3154192218136261e-05,
+      "loss": 0.6009,
+      "step": 5118
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8981685770096789,
+      "learning_rate": 1.3151694017720016e-05,
+      "loss": 0.5287,
+      "step": 5119
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9229520921845562,
+      "learning_rate": 1.3149195598894167e-05,
+      "loss": 0.5432,
+      "step": 5120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.31688240689969,
+      "learning_rate": 1.3146696961831858e-05,
+      "loss": 0.526,
+      "step": 5121
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8872179410218345,
+      "learning_rate": 1.314419810670624e-05,
+      "loss": 0.5431,
+      "step": 5122
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7965128643545243,
+      "learning_rate": 1.314169903369048e-05,
+      "loss": 0.465,
+      "step": 5123
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9228116133064066,
+      "learning_rate": 1.3139199742957767e-05,
+      "loss": 0.583,
+      "step": 5124
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9708210120494751,
+      "learning_rate": 1.3136700234681294e-05,
+      "loss": 0.6015,
+      "step": 5125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9379346186917816,
+      "learning_rate": 1.313420050903428e-05,
+      "loss": 0.5597,
+      "step": 5126
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8583394317632376,
+      "learning_rate": 1.313170056618995e-05,
+      "loss": 0.511,
+      "step": 5127
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9664397262334126,
+      "learning_rate": 1.3129200406321545e-05,
+      "loss": 0.5993,
+      "step": 5128
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.937342255366917,
+      "learning_rate": 1.312670002960233e-05,
+      "loss": 0.5576,
+      "step": 5129
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8938685765950013,
+      "learning_rate": 1.3124199436205575e-05,
+      "loss": 0.6095,
+      "step": 5130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8795874008243144,
+      "learning_rate": 1.3121698626304574e-05,
+      "loss": 0.5268,
+      "step": 5131
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9572529534621174,
+      "learning_rate": 1.3119197600072624e-05,
+      "loss": 0.5539,
+      "step": 5132
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7892642422145666,
+      "learning_rate": 1.3116696357683047e-05,
+      "loss": 0.5267,
+      "step": 5133
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9124039323834103,
+      "learning_rate": 1.3114194899309175e-05,
+      "loss": 0.5872,
+      "step": 5134
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9034469389534122,
+      "learning_rate": 1.3111693225124365e-05,
+      "loss": 0.5655,
+      "step": 5135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.980671478125758,
+      "learning_rate": 1.310919133530197e-05,
+      "loss": 0.5409,
+      "step": 5136
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7970800799824627,
+      "learning_rate": 1.3106689230015372e-05,
+      "loss": 0.4885,
+      "step": 5137
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8885303582931813,
+      "learning_rate": 1.310418690943797e-05,
+      "loss": 0.5902,
+      "step": 5138
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8250161927876943,
+      "learning_rate": 1.3101684373743166e-05,
+      "loss": 0.5207,
+      "step": 5139
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.809632259881539,
+      "learning_rate": 1.3099181623104386e-05,
+      "loss": 0.5252,
+      "step": 5140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9036939639535884,
+      "learning_rate": 1.3096678657695072e-05,
+      "loss": 0.5677,
+      "step": 5141
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8522063526840421,
+      "learning_rate": 1.3094175477688671e-05,
+      "loss": 0.4812,
+      "step": 5142
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9591812262443219,
+      "learning_rate": 1.3091672083258653e-05,
+      "loss": 0.5718,
+      "step": 5143
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8796134146527177,
+      "learning_rate": 1.3089168474578504e-05,
+      "loss": 0.5509,
+      "step": 5144
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9637346590878184,
+      "learning_rate": 1.3086664651821719e-05,
+      "loss": 0.5226,
+      "step": 5145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8654354312013556,
+      "learning_rate": 1.308416061516181e-05,
+      "loss": 0.5672,
+      "step": 5146
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0063143032868171,
+      "learning_rate": 1.3081656364772308e-05,
+      "loss": 0.5448,
+      "step": 5147
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9170705999289325,
+      "learning_rate": 1.3079151900826752e-05,
+      "loss": 0.5176,
+      "step": 5148
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9637334144792751,
+      "learning_rate": 1.3076647223498703e-05,
+      "loss": 0.553,
+      "step": 5149
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9671932172736195,
+      "learning_rate": 1.3074142332961729e-05,
+      "loss": 0.5069,
+      "step": 5150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9363712624459859,
+      "learning_rate": 1.3071637229389416e-05,
+      "loss": 0.5368,
+      "step": 5151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9080314781366527,
+      "learning_rate": 1.3069131912955368e-05,
+      "loss": 0.5389,
+      "step": 5152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.848854023646483,
+      "learning_rate": 1.3066626383833203e-05,
+      "loss": 0.5374,
+      "step": 5153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.991351160584258,
+      "learning_rate": 1.3064120642196549e-05,
+      "loss": 0.5621,
+      "step": 5154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9563550446244192,
+      "learning_rate": 1.306161468821905e-05,
+      "loss": 0.572,
+      "step": 5155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9125008511353809,
+      "learning_rate": 1.3059108522074373e-05,
+      "loss": 0.5362,
+      "step": 5156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8805060313587676,
+      "learning_rate": 1.3056602143936185e-05,
+      "loss": 0.5868,
+      "step": 5157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8926377582854529,
+      "learning_rate": 1.3054095553978181e-05,
+      "loss": 0.5372,
+      "step": 5158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8709372652470595,
+      "learning_rate": 1.3051588752374067e-05,
+      "loss": 0.5799,
+      "step": 5159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9118354309326878,
+      "learning_rate": 1.3049081739297556e-05,
+      "loss": 0.5731,
+      "step": 5160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8113643930294744,
+      "learning_rate": 1.3046574514922386e-05,
+      "loss": 0.5097,
+      "step": 5161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8474405953530317,
+      "learning_rate": 1.3044067079422304e-05,
+      "loss": 0.5331,
+      "step": 5162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9439939773612503,
+      "learning_rate": 1.304155943297107e-05,
+      "loss": 0.5656,
+      "step": 5163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8037282697874863,
+      "learning_rate": 1.303905157574247e-05,
+      "loss": 0.5426,
+      "step": 5164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.879080006172946,
+      "learning_rate": 1.303654350791029e-05,
+      "loss": 0.6089,
+      "step": 5165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.931439091942248,
+      "learning_rate": 1.3034035229648338e-05,
+      "loss": 0.5938,
+      "step": 5166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.897235722501806,
+      "learning_rate": 1.3031526741130435e-05,
+      "loss": 0.6158,
+      "step": 5167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9778046778570904,
+      "learning_rate": 1.3029018042530421e-05,
+      "loss": 0.616,
+      "step": 5168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8938904176131726,
+      "learning_rate": 1.3026509134022143e-05,
+      "loss": 0.5148,
+      "step": 5169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7637438153298557,
+      "learning_rate": 1.3024000015779462e-05,
+      "loss": 0.5162,
+      "step": 5170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8970184303722865,
+      "learning_rate": 1.3021490687976269e-05,
+      "loss": 0.5411,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9379911755413544,
+      "learning_rate": 1.3018981150786445e-05,
+      "loss": 0.5198,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9159176187237429,
+      "learning_rate": 1.3016471404383907e-05,
+      "loss": 0.5677,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8686715488022813,
+      "learning_rate": 1.3013961448942578e-05,
+      "loss": 0.5728,
+      "step": 5174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0833391331559723,
+      "learning_rate": 1.301145128463639e-05,
+      "loss": 0.6127,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9097788336027876,
+      "learning_rate": 1.3008940911639302e-05,
+      "loss": 0.5475,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9956793165985215,
+      "learning_rate": 1.3006430330125279e-05,
+      "loss": 0.511,
+      "step": 5177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9377355682219023,
+      "learning_rate": 1.30039195402683e-05,
+      "loss": 0.5849,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.875112847377846,
+      "learning_rate": 1.300140854224236e-05,
+      "loss": 0.5154,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8801229189319658,
+      "learning_rate": 1.299889733622147e-05,
+      "loss": 0.4734,
+      "step": 5180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.968504878800029,
+      "learning_rate": 1.2996385922379657e-05,
+      "loss": 0.5744,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7574635362509162,
+      "learning_rate": 1.2993874300890956e-05,
+      "loss": 0.4807,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0783731403307208,
+      "learning_rate": 1.2991362471929421e-05,
+      "loss": 0.6033,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9783506627178276,
+      "learning_rate": 1.2988850435669123e-05,
+      "loss": 0.5147,
+      "step": 5184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9580822811329507,
+      "learning_rate": 1.2986338192284136e-05,
+      "loss": 0.6557,
+      "step": 5185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.953670429295701,
+      "learning_rate": 1.2983825741948564e-05,
+      "loss": 0.5506,
+      "step": 5186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9522694105450515,
+      "learning_rate": 1.2981313084836514e-05,
+      "loss": 0.5054,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8276032219953149,
+      "learning_rate": 1.2978800221122112e-05,
+      "loss": 0.5091,
+      "step": 5188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.915831507762252,
+      "learning_rate": 1.2976287150979497e-05,
+      "loss": 0.5395,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.89869095528428,
+      "learning_rate": 1.297377387458282e-05,
+      "loss": 0.6037,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8638883544541025,
+      "learning_rate": 1.2971260392106255e-05,
+      "loss": 0.5695,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8361552108274168,
+      "learning_rate": 1.296874670372398e-05,
+      "loss": 0.4965,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.78892219953772,
+      "learning_rate": 1.2966232809610189e-05,
+      "loss": 0.459,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0306619602849194,
+      "learning_rate": 1.2963718709939098e-05,
+      "loss": 0.6183,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9698810524692302,
+      "learning_rate": 1.2961204404884928e-05,
+      "loss": 0.6101,
+      "step": 5195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8714946620693301,
+      "learning_rate": 1.2958689894621918e-05,
+      "loss": 0.5364,
+      "step": 5196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9102783335520639,
+      "learning_rate": 1.2956175179324323e-05,
+      "loss": 0.5288,
+      "step": 5197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9513914469151453,
+      "learning_rate": 1.2953660259166413e-05,
+      "loss": 0.5882,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9080645066236228,
+      "learning_rate": 1.2951145134322465e-05,
+      "loss": 0.5207,
+      "step": 5199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9749830303692215,
+      "learning_rate": 1.2948629804966776e-05,
+      "loss": 0.5629,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0249364407270003,
+      "learning_rate": 1.294611427127366e-05,
+      "loss": 0.5952,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9204881999222911,
+      "learning_rate": 1.2943598533417437e-05,
+      "loss": 0.5396,
+      "step": 5202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8352575903111814,
+      "learning_rate": 1.2941082591572443e-05,
+      "loss": 0.494,
+      "step": 5203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0853393787202654,
+      "learning_rate": 1.2938566445913037e-05,
+      "loss": 0.6347,
+      "step": 5204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9185748013824929,
+      "learning_rate": 1.2936050096613584e-05,
+      "loss": 0.5317,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9245617764417169,
+      "learning_rate": 1.2933533543848462e-05,
+      "loss": 0.5411,
+      "step": 5206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0003230199061517,
+      "learning_rate": 1.2931016787792069e-05,
+      "loss": 0.5995,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9833957390509364,
+      "learning_rate": 1.292849982861881e-05,
+      "loss": 0.604,
+      "step": 5208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9909115252326485,
+      "learning_rate": 1.2925982666503111e-05,
+      "loss": 0.5742,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9407003727079608,
+      "learning_rate": 1.2923465301619408e-05,
+      "loss": 0.5392,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9458924993894788,
+      "learning_rate": 1.2920947734142155e-05,
+      "loss": 0.5189,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9561205767333625,
+      "learning_rate": 1.2918429964245813e-05,
+      "loss": 0.5466,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8589889002194464,
+      "learning_rate": 1.2915911992104864e-05,
+      "loss": 0.5233,
+      "step": 5213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8304101043631497,
+      "learning_rate": 1.2913393817893803e-05,
+      "loss": 0.4897,
+      "step": 5214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9015121090208729,
+      "learning_rate": 1.291087544178713e-05,
+      "loss": 0.5162,
+      "step": 5215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8932814665787939,
+      "learning_rate": 1.2908356863959372e-05,
+      "loss": 0.5288,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8710258804583972,
+      "learning_rate": 1.2905838084585066e-05,
+      "loss": 0.4831,
+      "step": 5217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8989285859062396,
+      "learning_rate": 1.2903319103838756e-05,
+      "loss": 0.5019,
+      "step": 5218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9187626899309298,
+      "learning_rate": 1.2900799921895004e-05,
+      "loss": 0.5805,
+      "step": 5219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9328702855013249,
+      "learning_rate": 1.2898280538928396e-05,
+      "loss": 0.5927,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9985233117601223,
+      "learning_rate": 1.2895760955113514e-05,
+      "loss": 0.5583,
+      "step": 5221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9419826483791731,
+      "learning_rate": 1.2893241170624968e-05,
+      "loss": 0.5928,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9519831626411481,
+      "learning_rate": 1.2890721185637376e-05,
+      "loss": 0.5869,
+      "step": 5223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9777805383448769,
+      "learning_rate": 1.2888201000325368e-05,
+      "loss": 0.5647,
+      "step": 5224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8718493503514608,
+      "learning_rate": 1.2885680614863591e-05,
+      "loss": 0.532,
+      "step": 5225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9404180060319389,
+      "learning_rate": 1.2883160029426712e-05,
+      "loss": 0.5258,
+      "step": 5226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9516912980108595,
+      "learning_rate": 1.2880639244189397e-05,
+      "loss": 0.5553,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.920813450561122,
+      "learning_rate": 1.2878118259326335e-05,
+      "loss": 0.5373,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8585172444562801,
+      "learning_rate": 1.2875597075012236e-05,
+      "loss": 0.5208,
+      "step": 5229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8584243934134526,
+      "learning_rate": 1.2873075691421808e-05,
+      "loss": 0.4836,
+      "step": 5230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8292944955247089,
+      "learning_rate": 1.2870554108729783e-05,
+      "loss": 0.5396,
+      "step": 5231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8223913490116747,
+      "learning_rate": 1.2868032327110904e-05,
+      "loss": 0.4598,
+      "step": 5232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8055596776849586,
+      "learning_rate": 1.2865510346739928e-05,
+      "loss": 0.5109,
+      "step": 5233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.817052218239722,
+      "learning_rate": 1.2862988167791627e-05,
+      "loss": 0.5415,
+      "step": 5234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0810268067366446,
+      "learning_rate": 1.2860465790440788e-05,
+      "loss": 0.5753,
+      "step": 5235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8402617458059244,
+      "learning_rate": 1.2857943214862205e-05,
+      "loss": 0.4776,
+      "step": 5236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8276994257844174,
+      "learning_rate": 1.285542044123069e-05,
+      "loss": 0.5087,
+      "step": 5237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8298950139953116,
+      "learning_rate": 1.2852897469721074e-05,
+      "loss": 0.5485,
+      "step": 5238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9031430098759592,
+      "learning_rate": 1.2850374300508195e-05,
+      "loss": 0.5088,
+      "step": 5239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.950323481245847,
+      "learning_rate": 1.2847850933766901e-05,
+      "loss": 0.5373,
+      "step": 5240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.005940982880705,
+      "learning_rate": 1.2845327369672069e-05,
+      "loss": 0.6226,
+      "step": 5241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0628170200163112,
+      "learning_rate": 1.2842803608398568e-05,
+      "loss": 0.5751,
+      "step": 5242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9088578999699106,
+      "learning_rate": 1.2840279650121301e-05,
+      "loss": 0.5189,
+      "step": 5243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9250960010536801,
+      "learning_rate": 1.2837755495015176e-05,
+      "loss": 0.5714,
+      "step": 5244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9393467217764105,
+      "learning_rate": 1.283523114325511e-05,
+      "loss": 0.546,
+      "step": 5245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0273850999637157,
+      "learning_rate": 1.283270659501604e-05,
+      "loss": 0.5864,
+      "step": 5246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8594473911869244,
+      "learning_rate": 1.2830181850472918e-05,
+      "loss": 0.5368,
+      "step": 5247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8807421155459568,
+      "learning_rate": 1.2827656909800701e-05,
+      "loss": 0.5712,
+      "step": 5248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8545283767021522,
+      "learning_rate": 1.2825131773174371e-05,
+      "loss": 0.5173,
+      "step": 5249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8419064676592511,
+      "learning_rate": 1.2822606440768911e-05,
+      "loss": 0.5258,
+      "step": 5250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8260318829592953,
+      "learning_rate": 1.2820080912759334e-05,
+      "loss": 0.4934,
+      "step": 5251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9071781523737736,
+      "learning_rate": 1.2817555189320647e-05,
+      "loss": 0.5202,
+      "step": 5252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9027757793662681,
+      "learning_rate": 1.2815029270627885e-05,
+      "loss": 0.5303,
+      "step": 5253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0212430197446423,
+      "learning_rate": 1.2812503156856093e-05,
+      "loss": 0.6029,
+      "step": 5254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8681516837710173,
+      "learning_rate": 1.2809976848180328e-05,
+      "loss": 0.5677,
+      "step": 5255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.93728669923876,
+      "learning_rate": 1.2807450344775656e-05,
+      "loss": 0.5703,
+      "step": 5256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8472921948892896,
+      "learning_rate": 1.2804923646817169e-05,
+      "loss": 0.4963,
+      "step": 5257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9384456367903224,
+      "learning_rate": 1.2802396754479958e-05,
+      "loss": 0.5729,
+      "step": 5258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9159751306971758,
+      "learning_rate": 1.279986966793914e-05,
+      "loss": 0.5178,
+      "step": 5259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9505984994518845,
+      "learning_rate": 1.2797342387369837e-05,
+      "loss": 0.5697,
+      "step": 5260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8650825044597112,
+      "learning_rate": 1.279481491294719e-05,
+      "loss": 0.5482,
+      "step": 5261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9432578608303249,
+      "learning_rate": 1.2792287244846345e-05,
+      "loss": 0.5622,
+      "step": 5262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9072994162122723,
+      "learning_rate": 1.2789759383242471e-05,
+      "loss": 0.5521,
+      "step": 5263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.927937356858074,
+      "learning_rate": 1.2787231328310744e-05,
+      "loss": 0.5355,
+      "step": 5264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8077970985617918,
+      "learning_rate": 1.2784703080226364e-05,
+      "loss": 0.5006,
+      "step": 5265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9341490586832795,
+      "learning_rate": 1.2782174639164528e-05,
+      "loss": 0.5911,
+      "step": 5266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8463801416391185,
+      "learning_rate": 1.2779646005300457e-05,
+      "loss": 0.5753,
+      "step": 5267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8251393824383413,
+      "learning_rate": 1.2777117178809383e-05,
+      "loss": 0.52,
+      "step": 5268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9756413025958026,
+      "learning_rate": 1.2774588159866554e-05,
+      "loss": 0.5064,
+      "step": 5269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9333713823028544,
+      "learning_rate": 1.2772058948647224e-05,
+      "loss": 0.5589,
+      "step": 5270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8904186468309239,
+      "learning_rate": 1.2769529545326669e-05,
+      "loss": 0.5401,
+      "step": 5271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7971808339481724,
+      "learning_rate": 1.2766999950080172e-05,
+      "loss": 0.5002,
+      "step": 5272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9132419627921222,
+      "learning_rate": 1.2764470163083034e-05,
+      "loss": 0.5205,
+      "step": 5273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9284686013582293,
+      "learning_rate": 1.2761940184510564e-05,
+      "loss": 0.6319,
+      "step": 5274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8241780076529768,
+      "learning_rate": 1.2759410014538092e-05,
+      "loss": 0.5005,
+      "step": 5275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9026954517606373,
+      "learning_rate": 1.275687965334095e-05,
+      "loss": 0.503,
+      "step": 5276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8621089733539753,
+      "learning_rate": 1.2754349101094493e-05,
+      "loss": 0.5055,
+      "step": 5277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.876532065494681,
+      "learning_rate": 1.2751818357974092e-05,
+      "loss": 0.5718,
+      "step": 5278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9111254296857755,
+      "learning_rate": 1.2749287424155114e-05,
+      "loss": 0.5633,
+      "step": 5279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8105020558905384,
+      "learning_rate": 1.2746756299812959e-05,
+      "loss": 0.5318,
+      "step": 5280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8516231862863062,
+      "learning_rate": 1.2744224985123031e-05,
+      "loss": 0.5428,
+      "step": 5281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.89583325399951,
+      "learning_rate": 1.2741693480260742e-05,
+      "loss": 0.56,
+      "step": 5282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9014276458558959,
+      "learning_rate": 1.2739161785401525e-05,
+      "loss": 0.5501,
+      "step": 5283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9709594700786953,
+      "learning_rate": 1.2736629900720832e-05,
+      "loss": 0.6068,
+      "step": 5284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8386333307614217,
+      "learning_rate": 1.273409782639411e-05,
+      "loss": 0.5139,
+      "step": 5285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9843303001964966,
+      "learning_rate": 1.2731565562596833e-05,
+      "loss": 0.5859,
+      "step": 5286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9895222868007922,
+      "learning_rate": 1.2729033109504489e-05,
+      "loss": 0.5086,
+      "step": 5287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8602660281687846,
+      "learning_rate": 1.2726500467292569e-05,
+      "loss": 0.5303,
+      "step": 5288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9087743403337459,
+      "learning_rate": 1.2723967636136582e-05,
+      "loss": 0.5649,
+      "step": 5289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.841464399939494,
+      "learning_rate": 1.272143461621206e-05,
+      "loss": 0.5266,
+      "step": 5290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.945482006218386,
+      "learning_rate": 1.2718901407694529e-05,
+      "loss": 0.5942,
+      "step": 5291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0264055245686943,
+      "learning_rate": 1.2716368010759541e-05,
+      "loss": 0.625,
+      "step": 5292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8925968532334795,
+      "learning_rate": 1.2713834425582665e-05,
+      "loss": 0.5599,
+      "step": 5293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9076586798006714,
+      "learning_rate": 1.2711300652339466e-05,
+      "loss": 0.5792,
+      "step": 5294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8088359146950926,
+      "learning_rate": 1.2708766691205536e-05,
+      "loss": 0.542,
+      "step": 5295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8453489931750089,
+      "learning_rate": 1.270623254235648e-05,
+      "loss": 0.4916,
+      "step": 5296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8287437918320928,
+      "learning_rate": 1.2703698205967907e-05,
+      "loss": 0.4938,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0493865520195256,
+      "learning_rate": 1.2701163682215447e-05,
+      "loss": 0.5401,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8536139904524773,
+      "learning_rate": 1.2698628971274743e-05,
+      "loss": 0.5139,
+      "step": 5299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.89012269240296,
+      "learning_rate": 1.269609407332144e-05,
+      "loss": 0.5295,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9291127537817483,
+      "learning_rate": 1.2693558988531209e-05,
+      "loss": 0.5729,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9126294441308344,
+      "learning_rate": 1.2691023717079735e-05,
+      "loss": 0.5836,
+      "step": 5302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9426586461545383,
+      "learning_rate": 1.26884882591427e-05,
+      "loss": 0.5436,
+      "step": 5303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9323610742444876,
+      "learning_rate": 1.2685952614895813e-05,
+      "loss": 0.5832,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8529103894723162,
+      "learning_rate": 1.2683416784514796e-05,
+      "loss": 0.6131,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8558498414608312,
+      "learning_rate": 1.2680880768175372e-05,
+      "loss": 0.5317,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8504492314757207,
+      "learning_rate": 1.267834456605329e-05,
+      "loss": 0.5051,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8464138513156245,
+      "learning_rate": 1.2675808178324305e-05,
+      "loss": 0.4996,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8419298654246816,
+      "learning_rate": 1.2673271605164189e-05,
+      "loss": 0.5133,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9322900796698717,
+      "learning_rate": 1.2670734846748717e-05,
+      "loss": 0.5407,
+      "step": 5310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8430730919042455,
+      "learning_rate": 1.2668197903253694e-05,
+      "loss": 0.5238,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9579902652026588,
+      "learning_rate": 1.266566077485492e-05,
+      "loss": 0.5807,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9361797772639265,
+      "learning_rate": 1.2663123461728219e-05,
+      "loss": 0.5349,
+      "step": 5313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9216298671386612,
+      "learning_rate": 1.2660585964049425e-05,
+      "loss": 0.5433,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.895439295800641,
+      "learning_rate": 1.2658048281994386e-05,
+      "loss": 0.4908,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8042110357612069,
+      "learning_rate": 1.2655510415738954e-05,
+      "loss": 0.459,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9505226398868347,
+      "learning_rate": 1.2652972365459008e-05,
+      "loss": 0.5766,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0448013742685371,
+      "learning_rate": 1.2650434131330434e-05,
+      "loss": 0.6055,
+      "step": 5318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9055457061171901,
+      "learning_rate": 1.2647895713529119e-05,
+      "loss": 0.4885,
+      "step": 5319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9932989807705419,
+      "learning_rate": 1.2645357112230983e-05,
+      "loss": 0.5372,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8511979663476752,
+      "learning_rate": 1.2642818327611947e-05,
+      "loss": 0.4758,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8423055609806502,
+      "learning_rate": 1.2640279359847942e-05,
+      "loss": 0.4833,
+      "step": 5322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.909840172729127,
+      "learning_rate": 1.2637740209114918e-05,
+      "loss": 0.5762,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8666441465222685,
+      "learning_rate": 1.2635200875588843e-05,
+      "loss": 0.5552,
+      "step": 5324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8048930938774788,
+      "learning_rate": 1.2632661359445682e-05,
+      "loss": 0.4874,
+      "step": 5325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8658373056426494,
+      "learning_rate": 1.2630121660861421e-05,
+      "loss": 0.5751,
+      "step": 5326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.873960183814025,
+      "learning_rate": 1.2627581780012066e-05,
+      "loss": 0.5404,
+      "step": 5327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8694573208813933,
+      "learning_rate": 1.2625041717073623e-05,
+      "loss": 0.5388,
+      "step": 5328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8367705185934514,
+      "learning_rate": 1.2622501472222116e-05,
+      "loss": 0.4857,
+      "step": 5329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.021461286000572,
+      "learning_rate": 1.2619961045633584e-05,
+      "loss": 0.5045,
+      "step": 5330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9150106164172199,
+      "learning_rate": 1.2617420437484076e-05,
+      "loss": 0.5622,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9566331795578452,
+      "learning_rate": 1.2614879647949652e-05,
+      "loss": 0.5835,
+      "step": 5332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.860789775611763,
+      "learning_rate": 1.261233867720639e-05,
+      "loss": 0.5326,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8448972384896664,
+      "learning_rate": 1.2609797525430374e-05,
+      "loss": 0.4775,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.847748902308439,
+      "learning_rate": 1.2607256192797702e-05,
+      "loss": 0.5381,
+      "step": 5335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8675705368591691,
+      "learning_rate": 1.260471467948449e-05,
+      "loss": 0.4952,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8005689733352995,
+      "learning_rate": 1.2602172985666863e-05,
+      "loss": 0.553,
+      "step": 5337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.074567676112369,
+      "learning_rate": 1.2599631111520956e-05,
+      "loss": 0.4664,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8500137656374482,
+      "learning_rate": 1.2597089057222915e-05,
+      "loss": 0.5165,
+      "step": 5339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9387099432086206,
+      "learning_rate": 1.2594546822948909e-05,
+      "loss": 0.549,
+      "step": 5340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8832865858693016,
+      "learning_rate": 1.259200440887511e-05,
+      "loss": 0.5153,
+      "step": 5341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8929926334581008,
+      "learning_rate": 1.2589461815177702e-05,
+      "loss": 0.4826,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0259069016436662,
+      "learning_rate": 1.2586919042032889e-05,
+      "loss": 0.5732,
+      "step": 5343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0308709360387325,
+      "learning_rate": 1.258437608961688e-05,
+      "loss": 0.555,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9248441205920206,
+      "learning_rate": 1.2581832958105902e-05,
+      "loss": 0.5604,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0011778449614308,
+      "learning_rate": 1.257928964767619e-05,
+      "loss": 0.5902,
+      "step": 5346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9170889642768937,
+      "learning_rate": 1.2576746158503992e-05,
+      "loss": 0.5381,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8797931764427254,
+      "learning_rate": 1.257420249076557e-05,
+      "loss": 0.5294,
+      "step": 5348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.835885065953182,
+      "learning_rate": 1.25716586446372e-05,
+      "loss": 0.4715,
+      "step": 5349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8538701496862996,
+      "learning_rate": 1.2569114620295166e-05,
+      "loss": 0.4901,
+      "step": 5350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9695748609946032,
+      "learning_rate": 1.2566570417915769e-05,
+      "loss": 0.6038,
+      "step": 5351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9125029822928001,
+      "learning_rate": 1.2564026037675317e-05,
+      "loss": 0.5378,
+      "step": 5352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8841389114546176,
+      "learning_rate": 1.2561481479750135e-05,
+      "loss": 0.5214,
+      "step": 5353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9362021878627022,
+      "learning_rate": 1.2558936744316561e-05,
+      "loss": 0.5825,
+      "step": 5354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.917462682069863,
+      "learning_rate": 1.2556391831550938e-05,
+      "loss": 0.5607,
+      "step": 5355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9527731739824842,
+      "learning_rate": 1.255384674162963e-05,
+      "loss": 0.5483,
+      "step": 5356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9007801694808039,
+      "learning_rate": 1.2551301474729008e-05,
+      "loss": 0.5005,
+      "step": 5357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8961448090023717,
+      "learning_rate": 1.2548756031025455e-05,
+      "loss": 0.5556,
+      "step": 5358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.999360309272564,
+      "learning_rate": 1.254621041069537e-05,
+      "loss": 0.5881,
+      "step": 5359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8481061045768453,
+      "learning_rate": 1.2543664613915165e-05,
+      "loss": 0.4995,
+      "step": 5360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.947731835355656,
+      "learning_rate": 1.2541118640861255e-05,
+      "loss": 0.539,
+      "step": 5361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8498245240875937,
+      "learning_rate": 1.2538572491710079e-05,
+      "loss": 0.5049,
+      "step": 5362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.769815234398063,
+      "learning_rate": 1.2536026166638082e-05,
+      "loss": 0.4527,
+      "step": 5363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9020547480424845,
+      "learning_rate": 1.2533479665821719e-05,
+      "loss": 0.5664,
+      "step": 5364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9948695645265985,
+      "learning_rate": 1.2530932989437463e-05,
+      "loss": 0.5468,
+      "step": 5365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8738954752780324,
+      "learning_rate": 1.2528386137661797e-05,
+      "loss": 0.5222,
+      "step": 5366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9995522905076301,
+      "learning_rate": 1.2525839110671212e-05,
+      "loss": 0.6052,
+      "step": 5367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0371681322258024,
+      "learning_rate": 1.2523291908642219e-05,
+      "loss": 0.5546,
+      "step": 5368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8721378928964585,
+      "learning_rate": 1.2520744531751334e-05,
+      "loss": 0.51,
+      "step": 5369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8634386973360564,
+      "learning_rate": 1.251819698017509e-05,
+      "loss": 0.5024,
+      "step": 5370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9503165783798369,
+      "learning_rate": 1.2515649254090025e-05,
+      "loss": 0.5597,
+      "step": 5371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8720813638454793,
+      "learning_rate": 1.2513101353672703e-05,
+      "loss": 0.5894,
+      "step": 5372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9695051345654175,
+      "learning_rate": 1.2510553279099684e-05,
+      "loss": 0.5444,
+      "step": 5373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8972858005853297,
+      "learning_rate": 1.250800503054755e-05,
+      "loss": 0.556,
+      "step": 5374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8021491655858767,
+      "learning_rate": 1.2505456608192889e-05,
+      "loss": 0.5289,
+      "step": 5375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9734012631123989,
+      "learning_rate": 1.2502908012212313e-05,
+      "loss": 0.5637,
+      "step": 5376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9441552924743251,
+      "learning_rate": 1.2500359242782429e-05,
+      "loss": 0.6188,
+      "step": 5377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.899529468266094,
+      "learning_rate": 1.2497810300079866e-05,
+      "loss": 0.5009,
+      "step": 5378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8728144939932249,
+      "learning_rate": 1.249526118428127e-05,
+      "loss": 0.5664,
+      "step": 5379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8267144567594343,
+      "learning_rate": 1.2492711895563281e-05,
+      "loss": 0.5369,
+      "step": 5380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.873087845840797,
+      "learning_rate": 1.249016243410257e-05,
+      "loss": 0.5279,
+      "step": 5381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.858248198037963,
+      "learning_rate": 1.2487612800075814e-05,
+      "loss": 0.5452,
+      "step": 5382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9861992051149057,
+      "learning_rate": 1.2485062993659696e-05,
+      "loss": 0.5358,
+      "step": 5383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8845727809369965,
+      "learning_rate": 1.2482513015030915e-05,
+      "loss": 0.4948,
+      "step": 5384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9039360878072947,
+      "learning_rate": 1.2479962864366186e-05,
+      "loss": 0.5673,
+      "step": 5385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8615365882769283,
+      "learning_rate": 1.2477412541842231e-05,
+      "loss": 0.5202,
+      "step": 5386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7911663726867374,
+      "learning_rate": 1.247486204763578e-05,
+      "loss": 0.4393,
+      "step": 5387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9019516455908042,
+      "learning_rate": 1.247231138192359e-05,
+      "loss": 0.5104,
+      "step": 5388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9517601590994649,
+      "learning_rate": 1.246976054488241e-05,
+      "loss": 0.5931,
+      "step": 5389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8783075827940406,
+      "learning_rate": 1.2467209536689016e-05,
+      "loss": 0.512,
+      "step": 5390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9801083493781189,
+      "learning_rate": 1.2464658357520192e-05,
+      "loss": 0.5946,
+      "step": 5391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9940969062462093,
+      "learning_rate": 1.2462107007552726e-05,
+      "loss": 0.6072,
+      "step": 5392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9240346316696333,
+      "learning_rate": 1.2459555486963431e-05,
+      "loss": 0.5583,
+      "step": 5393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9381190189854962,
+      "learning_rate": 1.2457003795929121e-05,
+      "loss": 0.6326,
+      "step": 5394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0090297957220815,
+      "learning_rate": 1.2454451934626628e-05,
+      "loss": 0.5168,
+      "step": 5395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.833215564045747,
+      "learning_rate": 1.2451899903232793e-05,
+      "loss": 0.5483,
+      "step": 5396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9284695649603013,
+      "learning_rate": 1.244934770192447e-05,
+      "loss": 0.596,
+      "step": 5397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8845855956442529,
+      "learning_rate": 1.2446795330878522e-05,
+      "loss": 0.5107,
+      "step": 5398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.002581956309397,
+      "learning_rate": 1.244424279027183e-05,
+      "loss": 0.493,
+      "step": 5399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8829187422959885,
+      "learning_rate": 1.244169008028128e-05,
+      "loss": 0.4969,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9295869401823778,
+      "learning_rate": 1.2439137201083772e-05,
+      "loss": 0.5593,
+      "step": 5401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9731943915057911,
+      "learning_rate": 1.243658415285622e-05,
+      "loss": 0.5267,
+      "step": 5402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9063916835917486,
+      "learning_rate": 1.243403093577555e-05,
+      "loss": 0.5136,
+      "step": 5403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.865307935840072,
+      "learning_rate": 1.2431477550018691e-05,
+      "loss": 0.5174,
+      "step": 5404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9523780286387016,
+      "learning_rate": 1.2428923995762597e-05,
+      "loss": 0.5624,
+      "step": 5405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9867060620553869,
+      "learning_rate": 1.2426370273184226e-05,
+      "loss": 0.5949,
+      "step": 5406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9363530752198639,
+      "learning_rate": 1.2423816382460544e-05,
+      "loss": 0.5715,
+      "step": 5407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8694286672573258,
+      "learning_rate": 1.2421262323768537e-05,
+      "loss": 0.5034,
+      "step": 5408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8299654192646883,
+      "learning_rate": 1.2418708097285202e-05,
+      "loss": 0.4707,
+      "step": 5409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9104100825566318,
+      "learning_rate": 1.2416153703187537e-05,
+      "loss": 0.5178,
+      "step": 5410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9194277608599328,
+      "learning_rate": 1.2413599141652565e-05,
+      "loss": 0.4901,
+      "step": 5411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9041818758818694,
+      "learning_rate": 1.2411044412857317e-05,
+      "loss": 0.54,
+      "step": 5412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.016449033785623,
+      "learning_rate": 1.2408489516978824e-05,
+      "loss": 0.5814,
+      "step": 5413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9805254090857588,
+      "learning_rate": 1.2405934454194146e-05,
+      "loss": 0.5594,
+      "step": 5414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9510567274523963,
+      "learning_rate": 1.2403379224680346e-05,
+      "loss": 0.5203,
+      "step": 5415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9394727691546123,
+      "learning_rate": 1.2400823828614495e-05,
+      "loss": 0.5802,
+      "step": 5416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8411449916095526,
+      "learning_rate": 1.2398268266173683e-05,
+      "loss": 0.4941,
+      "step": 5417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.975070662691036,
+      "learning_rate": 1.239571253753501e-05,
+      "loss": 0.5484,
+      "step": 5418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9109850157859793,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.4947,
+      "step": 5419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8938766973316515,
+      "learning_rate": 1.2390600582372517e-05,
+      "loss": 0.5626,
+      "step": 5420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9179313851288792,
+      "learning_rate": 1.2388044356202958e-05,
+      "loss": 0.572,
+      "step": 5421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8698575621096661,
+      "learning_rate": 1.2385487964544038e-05,
+      "loss": 0.5427,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8987838459695859,
+      "learning_rate": 1.238293140757292e-05,
+      "loss": 0.505,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9269681022844656,
+      "learning_rate": 1.2380374685466772e-05,
+      "loss": 0.537,
+      "step": 5424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9014675995828588,
+      "learning_rate": 1.2377817798402767e-05,
+      "loss": 0.4887,
+      "step": 5425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9120011897073391,
+      "learning_rate": 1.2375260746558098e-05,
+      "loss": 0.5395,
+      "step": 5426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.934333432226395,
+      "learning_rate": 1.2372703530109967e-05,
+      "loss": 0.5583,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.990985907641905,
+      "learning_rate": 1.2370146149235585e-05,
+      "loss": 0.5498,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9623084732384362,
+      "learning_rate": 1.2367588604112177e-05,
+      "loss": 0.5172,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.838520536415113,
+      "learning_rate": 1.236503089491698e-05,
+      "loss": 0.5648,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8518399557849459,
+      "learning_rate": 1.236247302182724e-05,
+      "loss": 0.5575,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8947014247499556,
+      "learning_rate": 1.2359914985020212e-05,
+      "loss": 0.5625,
+      "step": 5432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9468389551013499,
+      "learning_rate": 1.2357356784673171e-05,
+      "loss": 0.6043,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8053680832226762,
+      "learning_rate": 1.2354798420963396e-05,
+      "loss": 0.4899,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9694037917733627,
+      "learning_rate": 1.2352239894068179e-05,
+      "loss": 0.6325,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8632305139677751,
+      "learning_rate": 1.2349681204164823e-05,
+      "loss": 0.5214,
+      "step": 5436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8563381614510859,
+      "learning_rate": 1.2347122351430645e-05,
+      "loss": 0.528,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8623448064496342,
+      "learning_rate": 1.2344563336042967e-05,
+      "loss": 0.5471,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8587003201570016,
+      "learning_rate": 1.2342004158179133e-05,
+      "loss": 0.5409,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8868847939450765,
+      "learning_rate": 1.2339444818016488e-05,
+      "loss": 0.5818,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9258515474662156,
+      "learning_rate": 1.233688531573239e-05,
+      "loss": 0.4862,
+      "step": 5441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8677169571698846,
+      "learning_rate": 1.2334325651504214e-05,
+      "loss": 0.5151,
+      "step": 5442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.870121746459327,
+      "learning_rate": 1.233176582550934e-05,
+      "loss": 0.5004,
+      "step": 5443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8453323949166696,
+      "learning_rate": 1.2329205837925162e-05,
+      "loss": 0.4571,
+      "step": 5444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8392697669827083,
+      "learning_rate": 1.2326645688929087e-05,
+      "loss": 0.4784,
+      "step": 5445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8743550979582336,
+      "learning_rate": 1.2324085378698529e-05,
+      "loss": 0.5231,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8375402911183744,
+      "learning_rate": 1.2321524907410916e-05,
+      "loss": 0.5185,
+      "step": 5447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.90586681912078,
+      "learning_rate": 1.2318964275243683e-05,
+      "loss": 0.5223,
+      "step": 5448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9029561290748873,
+      "learning_rate": 1.2316403482374289e-05,
+      "loss": 0.5531,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9179206919093462,
+      "learning_rate": 1.2313842528980184e-05,
+      "loss": 0.5461,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9248238397243361,
+      "learning_rate": 1.2311281415238842e-05,
+      "loss": 0.5494,
+      "step": 5451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8941957605560482,
+      "learning_rate": 1.2308720141327753e-05,
+      "loss": 0.5751,
+      "step": 5452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8724227657559777,
+      "learning_rate": 1.2306158707424402e-05,
+      "loss": 0.4947,
+      "step": 5453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9910439980193736,
+      "learning_rate": 1.2303597113706301e-05,
+      "loss": 0.5735,
+      "step": 5454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.764935426291101,
+      "learning_rate": 1.2301035360350964e-05,
+      "loss": 0.4984,
+      "step": 5455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9037221761184959,
+      "learning_rate": 1.2298473447535914e-05,
+      "loss": 0.503,
+      "step": 5456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8955191193698375,
+      "learning_rate": 1.2295911375438694e-05,
+      "loss": 0.5683,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8596766302284218,
+      "learning_rate": 1.2293349144236855e-05,
+      "loss": 0.5369,
+      "step": 5458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9915006897890639,
+      "learning_rate": 1.229078675410795e-05,
+      "loss": 0.5778,
+      "step": 5459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8705018308535379,
+      "learning_rate": 1.2288224205229557e-05,
+      "loss": 0.4935,
+      "step": 5460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8800565338043739,
+      "learning_rate": 1.228566149777926e-05,
+      "loss": 0.5087,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9986894910833827,
+      "learning_rate": 1.2283098631934642e-05,
+      "loss": 0.5404,
+      "step": 5462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8912962051173339,
+      "learning_rate": 1.2280535607873318e-05,
+      "loss": 0.5032,
+      "step": 5463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8038436091058793,
+      "learning_rate": 1.22779724257729e-05,
+      "loss": 0.5227,
+      "step": 5464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.938614221484701,
+      "learning_rate": 1.227540908581101e-05,
+      "loss": 0.5578,
+      "step": 5465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8750121548910289,
+      "learning_rate": 1.227284558816529e-05,
+      "loss": 0.5488,
+      "step": 5466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9057698650290454,
+      "learning_rate": 1.2270281933013388e-05,
+      "loss": 0.5608,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8328101242823822,
+      "learning_rate": 1.2267718120532958e-05,
+      "loss": 0.5489,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9559423602996424,
+      "learning_rate": 1.2265154150901677e-05,
+      "loss": 0.5267,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0768815635170421,
+      "learning_rate": 1.2262590024297226e-05,
+      "loss": 0.576,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.017951933254484,
+      "learning_rate": 1.2260025740897286e-05,
+      "loss": 0.5198,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9203291906948043,
+      "learning_rate": 1.225746130087957e-05,
+      "loss": 0.5603,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0045960023358238,
+      "learning_rate": 1.2254896704421789e-05,
+      "loss": 0.5984,
+      "step": 5473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9114512728903359,
+      "learning_rate": 1.2252331951701665e-05,
+      "loss": 0.5265,
+      "step": 5474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8752647547925406,
+      "learning_rate": 1.2249767042896934e-05,
+      "loss": 0.5302,
+      "step": 5475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8801933966164635,
+      "learning_rate": 1.2247201978185346e-05,
+      "loss": 0.5267,
+      "step": 5476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.992872988262125,
+      "learning_rate": 1.224463675774465e-05,
+      "loss": 0.5711,
+      "step": 5477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9604426674753421,
+      "learning_rate": 1.224207138175262e-05,
+      "loss": 0.5276,
+      "step": 5478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8640035052898114,
+      "learning_rate": 1.2239505850387032e-05,
+      "loss": 0.618,
+      "step": 5479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9810432720165322,
+      "learning_rate": 1.2236940163825675e-05,
+      "loss": 0.5615,
+      "step": 5480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8961399600071736,
+      "learning_rate": 1.2234374322246348e-05,
+      "loss": 0.5542,
+      "step": 5481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.879129224347014,
+      "learning_rate": 1.2231808325826862e-05,
+      "loss": 0.548,
+      "step": 5482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8514413978104286,
+      "learning_rate": 1.222924217474504e-05,
+      "loss": 0.5043,
+      "step": 5483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0162949354532185,
+      "learning_rate": 1.2226675869178713e-05,
+      "loss": 0.6186,
+      "step": 5484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9010809840353643,
+      "learning_rate": 1.222410940930572e-05,
+      "loss": 0.4572,
+      "step": 5485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9070348230842781,
+      "learning_rate": 1.2221542795303921e-05,
+      "loss": 0.5338,
+      "step": 5486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.905376480185276,
+      "learning_rate": 1.2218976027351177e-05,
+      "loss": 0.4919,
+      "step": 5487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9351701239895585,
+      "learning_rate": 1.221640910562536e-05,
+      "loss": 0.5618,
+      "step": 5488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.87105543303774,
+      "learning_rate": 1.2213842030304358e-05,
+      "loss": 0.5298,
+      "step": 5489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0073370084141626,
+      "learning_rate": 1.221127480156607e-05,
+      "loss": 0.5373,
+      "step": 5490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.877935805226189,
+      "learning_rate": 1.2208707419588397e-05,
+      "loss": 0.5437,
+      "step": 5491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9243438906000147,
+      "learning_rate": 1.220613988454926e-05,
+      "loss": 0.5324,
+      "step": 5492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9477932541999349,
+      "learning_rate": 1.2203572196626587e-05,
+      "loss": 0.5556,
+      "step": 5493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9184085706505207,
+      "learning_rate": 1.2201004355998312e-05,
+      "loss": 0.5626,
+      "step": 5494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8732092279319581,
+      "learning_rate": 1.2198436362842389e-05,
+      "loss": 0.5584,
+      "step": 5495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8818231916647266,
+      "learning_rate": 1.2195868217336778e-05,
+      "loss": 0.4974,
+      "step": 5496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9253509177877544,
+      "learning_rate": 1.2193299919659444e-05,
+      "loss": 0.507,
+      "step": 5497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9976868248539731,
+      "learning_rate": 1.2190731469988372e-05,
+      "loss": 0.6117,
+      "step": 5498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8668678379356317,
+      "learning_rate": 1.2188162868501557e-05,
+      "loss": 0.5028,
+      "step": 5499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9165726909760753,
+      "learning_rate": 1.2185594115376991e-05,
+      "loss": 0.5579,
+      "step": 5500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9154946875925669,
+      "learning_rate": 1.2183025210792692e-05,
+      "loss": 0.5251,
+      "step": 5501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8404918378432502,
+      "learning_rate": 1.218045615492668e-05,
+      "loss": 0.5286,
+      "step": 5502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8712436361695308,
+      "learning_rate": 1.2177886947956997e-05,
+      "loss": 0.5304,
+      "step": 5503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8286927110237194,
+      "learning_rate": 1.2175317590061676e-05,
+      "loss": 0.5094,
+      "step": 5504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.866565081858643,
+      "learning_rate": 1.2172748081418775e-05,
+      "loss": 0.5399,
+      "step": 5505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9016013035316106,
+      "learning_rate": 1.2170178422206362e-05,
+      "loss": 0.5273,
+      "step": 5506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8433212719566922,
+      "learning_rate": 1.2167608612602507e-05,
+      "loss": 0.4995,
+      "step": 5507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8611469548274269,
+      "learning_rate": 1.2165038652785297e-05,
+      "loss": 0.5434,
+      "step": 5508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9387736783283651,
+      "learning_rate": 1.2162468542932832e-05,
+      "loss": 0.5529,
+      "step": 5509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7884933758412176,
+      "learning_rate": 1.2159898283223213e-05,
+      "loss": 0.4898,
+      "step": 5510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8797535779005713,
+      "learning_rate": 1.2157327873834559e-05,
+      "loss": 0.5357,
+      "step": 5511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7845933710931979,
+      "learning_rate": 1.2154757314944997e-05,
+      "loss": 0.4465,
+      "step": 5512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8716534801191331,
+      "learning_rate": 1.2152186606732665e-05,
+      "loss": 0.4596,
+      "step": 5513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8806006921198266,
+      "learning_rate": 1.2149615749375707e-05,
+      "loss": 0.5238,
+      "step": 5514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9519784857876017,
+      "learning_rate": 1.2147044743052288e-05,
+      "loss": 0.5458,
+      "step": 5515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8809363718552123,
+      "learning_rate": 1.2144473587940573e-05,
+      "loss": 0.5031,
+      "step": 5516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9694116837326615,
+      "learning_rate": 1.2141902284218738e-05,
+      "loss": 0.4856,
+      "step": 5517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.005998546729219,
+      "learning_rate": 1.2139330832064975e-05,
+      "loss": 0.5647,
+      "step": 5518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9363523110506934,
+      "learning_rate": 1.2136759231657485e-05,
+      "loss": 0.514,
+      "step": 5519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8834684878319838,
+      "learning_rate": 1.2134187483174474e-05,
+      "loss": 0.5651,
+      "step": 5520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9199792993437208,
+      "learning_rate": 1.2131615586794162e-05,
+      "loss": 0.5119,
+      "step": 5521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9083253044159368,
+      "learning_rate": 1.2129043542694783e-05,
+      "loss": 0.5347,
+      "step": 5522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9079207141814368,
+      "learning_rate": 1.2126471351054574e-05,
+      "loss": 0.5234,
+      "step": 5523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8996213537502855,
+      "learning_rate": 1.2123899012051785e-05,
+      "loss": 0.5154,
+      "step": 5524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8167759767195947,
+      "learning_rate": 1.212132652586468e-05,
+      "loss": 0.5262,
+      "step": 5525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8599043700048763,
+      "learning_rate": 1.211875389267153e-05,
+      "loss": 0.5309,
+      "step": 5526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8820657466473959,
+      "learning_rate": 1.211618111265061e-05,
+      "loss": 0.4873,
+      "step": 5527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8665427619337003,
+      "learning_rate": 1.2113608185980221e-05,
+      "loss": 0.5374,
+      "step": 5528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8908908247445878,
+      "learning_rate": 1.2111035112838657e-05,
+      "loss": 0.4539,
+      "step": 5529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8837991806660707,
+      "learning_rate": 1.2108461893404231e-05,
+      "loss": 0.5449,
+      "step": 5530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9001160427372054,
+      "learning_rate": 1.210588852785527e-05,
+      "loss": 0.608,
+      "step": 5531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9406825613037999,
+      "learning_rate": 1.2103315016370098e-05,
+      "loss": 0.5248,
+      "step": 5532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9492003060158064,
+      "learning_rate": 1.2100741359127062e-05,
+      "loss": 0.6211,
+      "step": 5533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8165428959386474,
+      "learning_rate": 1.2098167556304514e-05,
+      "loss": 0.5004,
+      "step": 5534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9547917538163686,
+      "learning_rate": 1.2095593608080815e-05,
+      "loss": 0.5788,
+      "step": 5535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8694750633823777,
+      "learning_rate": 1.2093019514634337e-05,
+      "loss": 0.5144,
+      "step": 5536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8783697437957985,
+      "learning_rate": 1.2090445276143466e-05,
+      "loss": 0.4806,
+      "step": 5537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8783928877709485,
+      "learning_rate": 1.2087870892786588e-05,
+      "loss": 0.5376,
+      "step": 5538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.949269064846333,
+      "learning_rate": 1.208529636474211e-05,
+      "loss": 0.5852,
+      "step": 5539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9646333868024378,
+      "learning_rate": 1.2082721692188446e-05,
+      "loss": 0.5591,
+      "step": 5540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.901401626566398,
+      "learning_rate": 1.2080146875304012e-05,
+      "loss": 0.5857,
+      "step": 5541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8454219718524448,
+      "learning_rate": 1.2077571914267248e-05,
+      "loss": 0.5336,
+      "step": 5542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8808345197786775,
+      "learning_rate": 1.2074996809256594e-05,
+      "loss": 0.5246,
+      "step": 5543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8775532655927482,
+      "learning_rate": 1.2072421560450497e-05,
+      "loss": 0.5933,
+      "step": 5544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9150474033054725,
+      "learning_rate": 1.2069846168027427e-05,
+      "loss": 0.5064,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9070112198834746,
+      "learning_rate": 1.2067270632165856e-05,
+      "loss": 0.5646,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8538367633178499,
+      "learning_rate": 1.2064694953044259e-05,
+      "loss": 0.5207,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9279054018383129,
+      "learning_rate": 1.2062119130841135e-05,
+      "loss": 0.6147,
+      "step": 5548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0040818800527798,
+      "learning_rate": 1.2059543165734986e-05,
+      "loss": 0.6258,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9316998246149493,
+      "learning_rate": 1.2056967057904319e-05,
+      "loss": 0.5779,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9190557347896983,
+      "learning_rate": 1.2054390807527661e-05,
+      "loss": 0.5837,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9606163824300413,
+      "learning_rate": 1.2051814414783544e-05,
+      "loss": 0.501,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9640744408170316,
+      "learning_rate": 1.2049237879850506e-05,
+      "loss": 0.535,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8788815642734494,
+      "learning_rate": 1.2046661202907101e-05,
+      "loss": 0.4897,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7727154304321403,
+      "learning_rate": 1.2044084384131891e-05,
+      "loss": 0.5226,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0129556356194376,
+      "learning_rate": 1.2041507423703445e-05,
+      "loss": 0.6603,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9331842751781539,
+      "learning_rate": 1.2038930321800346e-05,
+      "loss": 0.5507,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9206369409583717,
+      "learning_rate": 1.2036353078601187e-05,
+      "loss": 0.5849,
+      "step": 5558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9268985407098875,
+      "learning_rate": 1.2033775694284562e-05,
+      "loss": 0.5296,
+      "step": 5559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8851565422818308,
+      "learning_rate": 1.2031198169029084e-05,
+      "loss": 0.5511,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7522171063517573,
+      "learning_rate": 1.2028620503013377e-05,
+      "loss": 0.5037,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9237884426494588,
+      "learning_rate": 1.2026042696416069e-05,
+      "loss": 0.4642,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8747303264360102,
+      "learning_rate": 1.20234647494158e-05,
+      "loss": 0.5071,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9304691768041673,
+      "learning_rate": 1.2020886662191216e-05,
+      "loss": 0.5432,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.983912008294001,
+      "learning_rate": 1.2018308434920983e-05,
+      "loss": 0.628,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.908796074695498,
+      "learning_rate": 1.201573006778376e-05,
+      "loss": 0.5621,
+      "step": 5566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9866490956856787,
+      "learning_rate": 1.2013151560958233e-05,
+      "loss": 0.5905,
+      "step": 5567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.958082254827949,
+      "learning_rate": 1.2010572914623091e-05,
+      "loss": 0.5013,
+      "step": 5568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9013354679959612,
+      "learning_rate": 1.2007994128957029e-05,
+      "loss": 0.4822,
+      "step": 5569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8861126057246507,
+      "learning_rate": 1.2005415204138753e-05,
+      "loss": 0.5551,
+      "step": 5570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8977421349365613,
+      "learning_rate": 1.2002836140346984e-05,
+      "loss": 0.5817,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9163167902421183,
+      "learning_rate": 1.2000256937760446e-05,
+      "loss": 0.5701,
+      "step": 5572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8061350723006823,
+      "learning_rate": 1.1997677596557875e-05,
+      "loss": 0.4702,
+      "step": 5573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8525210977763026,
+      "learning_rate": 1.1995098116918022e-05,
+      "loss": 0.4862,
+      "step": 5574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9739495505934435,
+      "learning_rate": 1.1992518499019637e-05,
+      "loss": 0.5159,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8864361775922159,
+      "learning_rate": 1.1989938743041487e-05,
+      "loss": 0.5461,
+      "step": 5576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.004216155887953,
+      "learning_rate": 1.1987358849162349e-05,
+      "loss": 0.5517,
+      "step": 5577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9796249965818342,
+      "learning_rate": 1.1984778817561002e-05,
+      "loss": 0.5026,
+      "step": 5578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8592763662836835,
+      "learning_rate": 1.1982198648416245e-05,
+      "loss": 0.5431,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8886989571121534,
+      "learning_rate": 1.1979618341906884e-05,
+      "loss": 0.5394,
+      "step": 5580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9863795713903801,
+      "learning_rate": 1.1977037898211723e-05,
+      "loss": 0.5857,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8654981222560337,
+      "learning_rate": 1.1974457317509591e-05,
+      "loss": 0.5435,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8571963875092631,
+      "learning_rate": 1.197187659997932e-05,
+      "loss": 0.5196,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9245798277364339,
+      "learning_rate": 1.1969295745799746e-05,
+      "loss": 0.548,
+      "step": 5584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8977120983327718,
+      "learning_rate": 1.1966714755149724e-05,
+      "loss": 0.5561,
+      "step": 5585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.911986785371461,
+      "learning_rate": 1.1964133628208116e-05,
+      "loss": 0.5717,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8219590372461145,
+      "learning_rate": 1.196155236515379e-05,
+      "loss": 0.4861,
+      "step": 5587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8681933932768588,
+      "learning_rate": 1.1958970966165622e-05,
+      "loss": 0.512,
+      "step": 5588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0060914497630464,
+      "learning_rate": 1.1956389431422508e-05,
+      "loss": 0.6282,
+      "step": 5589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.028934286984717,
+      "learning_rate": 1.1953807761103338e-05,
+      "loss": 0.5839,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8952935651557266,
+      "learning_rate": 1.1951225955387025e-05,
+      "loss": 0.5875,
+      "step": 5591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1605839344131503,
+      "learning_rate": 1.1948644014452484e-05,
+      "loss": 0.538,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9642560605768177,
+      "learning_rate": 1.1946061938478638e-05,
+      "loss": 0.5382,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9704106218133949,
+      "learning_rate": 1.1943479727644429e-05,
+      "loss": 0.5499,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.940747211343841,
+      "learning_rate": 1.19408973821288e-05,
+      "loss": 0.5168,
+      "step": 5595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9471035641500342,
+      "learning_rate": 1.1938314902110701e-05,
+      "loss": 0.5431,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8100368901211942,
+      "learning_rate": 1.1935732287769099e-05,
+      "loss": 0.5276,
+      "step": 5597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9216492266864472,
+      "learning_rate": 1.193314953928297e-05,
+      "loss": 0.5553,
+      "step": 5598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9298432323689065,
+      "learning_rate": 1.1930566656831288e-05,
+      "loss": 0.5584,
+      "step": 5599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9101378356734033,
+      "learning_rate": 1.1927983640593053e-05,
+      "loss": 0.4693,
+      "step": 5600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.854826184502737,
+      "learning_rate": 1.192540049074726e-05,
+      "loss": 0.5558,
+      "step": 5601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.029647115535759,
+      "learning_rate": 1.1922817207472921e-05,
+      "loss": 0.5951,
+      "step": 5602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8843012858921222,
+      "learning_rate": 1.1920233790949051e-05,
+      "loss": 0.5116,
+      "step": 5603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8733868563514845,
+      "learning_rate": 1.191765024135469e-05,
+      "loss": 0.5402,
+      "step": 5604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8751251911535606,
+      "learning_rate": 1.1915066558868865e-05,
+      "loss": 0.4786,
+      "step": 5605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8603668761491104,
+      "learning_rate": 1.1912482743670624e-05,
+      "loss": 0.4809,
+      "step": 5606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8454520161691437,
+      "learning_rate": 1.1909898795939028e-05,
+      "loss": 0.5249,
+      "step": 5607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9231626415447299,
+      "learning_rate": 1.1907314715853138e-05,
+      "loss": 0.4975,
+      "step": 5608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9305487251057392,
+      "learning_rate": 1.190473050359203e-05,
+      "loss": 0.5702,
+      "step": 5609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9161739295244067,
+      "learning_rate": 1.1902146159334788e-05,
+      "loss": 0.5287,
+      "step": 5610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9175136570243649,
+      "learning_rate": 1.1899561683260506e-05,
+      "loss": 0.5298,
+      "step": 5611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9511351394030586,
+      "learning_rate": 1.189697707554828e-05,
+      "loss": 0.4931,
+      "step": 5612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9041144006702436,
+      "learning_rate": 1.189439233637723e-05,
+      "loss": 0.5236,
+      "step": 5613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9583396369583367,
+      "learning_rate": 1.1891807465926467e-05,
+      "loss": 0.5616,
+      "step": 5614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8930178278090726,
+      "learning_rate": 1.1889222464375127e-05,
+      "loss": 0.5447,
+      "step": 5615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8980531194959175,
+      "learning_rate": 1.1886637331902349e-05,
+      "loss": 0.4901,
+      "step": 5616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8984989642229662,
+      "learning_rate": 1.1884052068687273e-05,
+      "loss": 0.5205,
+      "step": 5617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9111022679999827,
+      "learning_rate": 1.188146667490906e-05,
+      "loss": 0.5738,
+      "step": 5618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9693879175100667,
+      "learning_rate": 1.1878881150746878e-05,
+      "loss": 0.5605,
+      "step": 5619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8351788339485661,
+      "learning_rate": 1.1876295496379894e-05,
+      "loss": 0.5249,
+      "step": 5620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8617744689704865,
+      "learning_rate": 1.18737097119873e-05,
+      "loss": 0.5436,
+      "step": 5621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8774024363332183,
+      "learning_rate": 1.1871123797748285e-05,
+      "loss": 0.5631,
+      "step": 5622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8808367572990808,
+      "learning_rate": 1.1868537753842052e-05,
+      "loss": 0.515,
+      "step": 5623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9095122786929855,
+      "learning_rate": 1.1865951580447805e-05,
+      "loss": 0.5841,
+      "step": 5624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8600399569527147,
+      "learning_rate": 1.1863365277744771e-05,
+      "loss": 0.5722,
+      "step": 5625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0620943004279169,
+      "learning_rate": 1.1860778845912177e-05,
+      "loss": 0.6108,
+      "step": 5626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9001555709351374,
+      "learning_rate": 1.185819228512926e-05,
+      "loss": 0.5784,
+      "step": 5627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8901095149384635,
+      "learning_rate": 1.1855605595575263e-05,
+      "loss": 0.5315,
+      "step": 5628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8751412997045999,
+      "learning_rate": 1.1853018777429449e-05,
+      "loss": 0.5046,
+      "step": 5629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9685659253082625,
+      "learning_rate": 1.1850431830871075e-05,
+      "loss": 0.6441,
+      "step": 5630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8627187869655145,
+      "learning_rate": 1.1847844756079414e-05,
+      "loss": 0.4905,
+      "step": 5631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9491405505682985,
+      "learning_rate": 1.1845257553233753e-05,
+      "loss": 0.4991,
+      "step": 5632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8990900153819141,
+      "learning_rate": 1.1842670222513379e-05,
+      "loss": 0.5544,
+      "step": 5633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8231690363328513,
+      "learning_rate": 1.1840082764097593e-05,
+      "loss": 0.5093,
+      "step": 5634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8340209505703706,
+      "learning_rate": 1.1837495178165706e-05,
+      "loss": 0.5191,
+      "step": 5635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8824202838144878,
+      "learning_rate": 1.183490746489703e-05,
+      "loss": 0.5046,
+      "step": 5636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9022132591342349,
+      "learning_rate": 1.1832319624470895e-05,
+      "loss": 0.5241,
+      "step": 5637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8615562704331575,
+      "learning_rate": 1.1829731657066638e-05,
+      "loss": 0.5016,
+      "step": 5638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8257423829771974,
+      "learning_rate": 1.1827143562863597e-05,
+      "loss": 0.526,
+      "step": 5639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9633143307811811,
+      "learning_rate": 1.1824555342041129e-05,
+      "loss": 0.5566,
+      "step": 5640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8663038774802817,
+      "learning_rate": 1.1821966994778594e-05,
+      "loss": 0.4996,
+      "step": 5641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7574394729770879,
+      "learning_rate": 1.1819378521255362e-05,
+      "loss": 0.4978,
+      "step": 5642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.874450535779544,
+      "learning_rate": 1.181678992165081e-05,
+      "loss": 0.5369,
+      "step": 5643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8685484962413709,
+      "learning_rate": 1.1814201196144332e-05,
+      "loss": 0.5223,
+      "step": 5644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9004285750051082,
+      "learning_rate": 1.181161234491532e-05,
+      "loss": 0.5838,
+      "step": 5645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8438898141326269,
+      "learning_rate": 1.1809023368143178e-05,
+      "loss": 0.5334,
+      "step": 5646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9248851968313923,
+      "learning_rate": 1.180643426600732e-05,
+      "loss": 0.5461,
+      "step": 5647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.937478179130208,
+      "learning_rate": 1.1803845038687171e-05,
+      "loss": 0.5556,
+      "step": 5648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8211910043207973,
+      "learning_rate": 1.1801255686362161e-05,
+      "loss": 0.4998,
+      "step": 5649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9253745602627443,
+      "learning_rate": 1.1798666209211729e-05,
+      "loss": 0.5692,
+      "step": 5650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8931107878011122,
+      "learning_rate": 1.1796076607415324e-05,
+      "loss": 0.5769,
+      "step": 5651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.032722335210612,
+      "learning_rate": 1.1793486881152405e-05,
+      "loss": 0.6256,
+      "step": 5652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8894699649744721,
+      "learning_rate": 1.1790897030602436e-05,
+      "loss": 0.461,
+      "step": 5653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.864199629548567,
+      "learning_rate": 1.1788307055944887e-05,
+      "loss": 0.4806,
+      "step": 5654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9561792929043047,
+      "learning_rate": 1.178571695735925e-05,
+      "loss": 0.527,
+      "step": 5655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0988366775659597,
+      "learning_rate": 1.178312673502501e-05,
+      "loss": 0.5619,
+      "step": 5656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.809626546189783,
+      "learning_rate": 1.1780536389121668e-05,
+      "loss": 0.505,
+      "step": 5657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8827486792369572,
+      "learning_rate": 1.1777945919828735e-05,
+      "loss": 0.4705,
+      "step": 5658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9744348262600504,
+      "learning_rate": 1.1775355327325726e-05,
+      "loss": 0.5681,
+      "step": 5659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9503573694547869,
+      "learning_rate": 1.1772764611792167e-05,
+      "loss": 0.588,
+      "step": 5660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.855926523059355,
+      "learning_rate": 1.1770173773407594e-05,
+      "loss": 0.5295,
+      "step": 5661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8577754564930304,
+      "learning_rate": 1.176758281235155e-05,
+      "loss": 0.5658,
+      "step": 5662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8640342306312921,
+      "learning_rate": 1.1764991728803582e-05,
+      "loss": 0.4698,
+      "step": 5663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8347933853968565,
+      "learning_rate": 1.1762400522943254e-05,
+      "loss": 0.5503,
+      "step": 5664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9123725024563369,
+      "learning_rate": 1.1759809194950134e-05,
+      "loss": 0.5921,
+      "step": 5665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8766709410074994,
+      "learning_rate": 1.1757217745003797e-05,
+      "loss": 0.5239,
+      "step": 5666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0059369165290584,
+      "learning_rate": 1.1754626173283827e-05,
+      "loss": 0.5633,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8504046043846838,
+      "learning_rate": 1.1752034479969822e-05,
+      "loss": 0.5828,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9058726927790431,
+      "learning_rate": 1.1749442665241382e-05,
+      "loss": 0.5252,
+      "step": 5669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.862169907036434,
+      "learning_rate": 1.1746850729278114e-05,
+      "loss": 0.5506,
+      "step": 5670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9637561441997063,
+      "learning_rate": 1.1744258672259642e-05,
+      "loss": 0.5342,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1554538855075298,
+      "learning_rate": 1.174166649436559e-05,
+      "loss": 0.602,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1545314272880411,
+      "learning_rate": 1.1739074195775597e-05,
+      "loss": 0.6155,
+      "step": 5673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8747730766584026,
+      "learning_rate": 1.1736481776669307e-05,
+      "loss": 0.5385,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8670802659680884,
+      "learning_rate": 1.1733889237226364e-05,
+      "loss": 0.5082,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8223266345411775,
+      "learning_rate": 1.1731296577626437e-05,
+      "loss": 0.5325,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.872649722870166,
+      "learning_rate": 1.1728703798049194e-05,
+      "loss": 0.5906,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8882721443314949,
+      "learning_rate": 1.172611089867431e-05,
+      "loss": 0.6095,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8946066834933611,
+      "learning_rate": 1.1723517879681472e-05,
+      "loss": 0.5391,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9320314319421376,
+      "learning_rate": 1.1720924741250378e-05,
+      "loss": 0.5406,
+      "step": 5680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8376780591969047,
+      "learning_rate": 1.1718331483560719e-05,
+      "loss": 0.5445,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0277180139899027,
+      "learning_rate": 1.1715738106792214e-05,
+      "loss": 0.5892,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0513463702008126,
+      "learning_rate": 1.1713144611124583e-05,
+      "loss": 0.573,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9765226894666977,
+      "learning_rate": 1.1710550996737548e-05,
+      "loss": 0.5579,
+      "step": 5684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.860387461075728,
+      "learning_rate": 1.1707957263810845e-05,
+      "loss": 0.5598,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0143344858243026,
+      "learning_rate": 1.170536341252422e-05,
+      "loss": 0.579,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8562956547995043,
+      "learning_rate": 1.1702769443057425e-05,
+      "loss": 0.4799,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8056493289664926,
+      "learning_rate": 1.1700175355590215e-05,
+      "loss": 0.5143,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9927643119593819,
+      "learning_rate": 1.1697581150302362e-05,
+      "loss": 0.5658,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9375696945093996,
+      "learning_rate": 1.1694986827373642e-05,
+      "loss": 0.5483,
+      "step": 5690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9725306054037465,
+      "learning_rate": 1.1692392386983837e-05,
+      "loss": 0.59,
+      "step": 5691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9917244244980913,
+      "learning_rate": 1.1689797829312741e-05,
+      "loss": 0.5406,
+      "step": 5692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9484530592426267,
+      "learning_rate": 1.1687203154540154e-05,
+      "loss": 0.5242,
+      "step": 5693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9394009906033737,
+      "learning_rate": 1.1684608362845886e-05,
+      "loss": 0.5825,
+      "step": 5694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8359560568490447,
+      "learning_rate": 1.168201345440975e-05,
+      "loss": 0.4697,
+      "step": 5695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8926529091421872,
+      "learning_rate": 1.1679418429411577e-05,
+      "loss": 0.5752,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8486560830613838,
+      "learning_rate": 1.1676823288031197e-05,
+      "loss": 0.5451,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8495140880211136,
+      "learning_rate": 1.1674228030448447e-05,
+      "loss": 0.5324,
+      "step": 5698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0059261820258834,
+      "learning_rate": 1.167163265684318e-05,
+      "loss": 0.6143,
+      "step": 5699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9051695413093083,
+      "learning_rate": 1.1669037167395256e-05,
+      "loss": 0.6137,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.988907191699513,
+      "learning_rate": 1.1666441562284534e-05,
+      "loss": 0.5691,
+      "step": 5701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8473370068987643,
+      "learning_rate": 1.166384584169089e-05,
+      "loss": 0.4898,
+      "step": 5702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0623212023511857,
+      "learning_rate": 1.1661250005794206e-05,
+      "loss": 0.5344,
+      "step": 5703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.860164057627655,
+      "learning_rate": 1.1658654054774368e-05,
+      "loss": 0.478,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.97579313832456,
+      "learning_rate": 1.1656057988811278e-05,
+      "loss": 0.5864,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9823768671863748,
+      "learning_rate": 1.1653461808084839e-05,
+      "loss": 0.5735,
+      "step": 5706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9273988931839644,
+      "learning_rate": 1.1650865512774959e-05,
+      "loss": 0.4811,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.955658161290628,
+      "learning_rate": 1.1648269103061567e-05,
+      "loss": 0.5759,
+      "step": 5708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9437176531862602,
+      "learning_rate": 1.1645672579124586e-05,
+      "loss": 0.6016,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8760405278095192,
+      "learning_rate": 1.1643075941143956e-05,
+      "loss": 0.5191,
+      "step": 5710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8942711976734233,
+      "learning_rate": 1.164047918929962e-05,
+      "loss": 0.5556,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.902951537057684,
+      "learning_rate": 1.1637882323771532e-05,
+      "loss": 0.5103,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8573316626768663,
+      "learning_rate": 1.163528534473965e-05,
+      "loss": 0.5217,
+      "step": 5713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0027552259507533,
+      "learning_rate": 1.1632688252383948e-05,
+      "loss": 0.5788,
+      "step": 5714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8367820712338532,
+      "learning_rate": 1.1630091046884394e-05,
+      "loss": 0.5012,
+      "step": 5715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8006736737941973,
+      "learning_rate": 1.1627493728420978e-05,
+      "loss": 0.4689,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9051130579150752,
+      "learning_rate": 1.1624896297173693e-05,
+      "loss": 0.6383,
+      "step": 5717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8939363733073373,
+      "learning_rate": 1.1622298753322531e-05,
+      "loss": 0.5842,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9497539058849697,
+      "learning_rate": 1.1619701097047507e-05,
+      "loss": 0.4973,
+      "step": 5719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8778289148182196,
+      "learning_rate": 1.1617103328528634e-05,
+      "loss": 0.6009,
+      "step": 5720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8498858417717557,
+      "learning_rate": 1.1614505447945935e-05,
+      "loss": 0.4794,
+      "step": 5721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.891479859871484,
+      "learning_rate": 1.1611907455479439e-05,
+      "loss": 0.5115,
+      "step": 5722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8737329204498719,
+      "learning_rate": 1.1609309351309185e-05,
+      "loss": 0.5272,
+      "step": 5723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8746927578986288,
+      "learning_rate": 1.1606711135615223e-05,
+      "loss": 0.5308,
+      "step": 5724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8336535581674581,
+      "learning_rate": 1.1604112808577603e-05,
+      "loss": 0.5128,
+      "step": 5725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8336680782694602,
+      "learning_rate": 1.160151437037639e-05,
+      "loss": 0.5382,
+      "step": 5726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0068426087236186,
+      "learning_rate": 1.159891582119165e-05,
+      "loss": 0.5948,
+      "step": 5727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8861976744403234,
+      "learning_rate": 1.159631716120346e-05,
+      "loss": 0.5187,
+      "step": 5728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8648814082644428,
+      "learning_rate": 1.1593718390591913e-05,
+      "loss": 0.535,
+      "step": 5729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8212720183780278,
+      "learning_rate": 1.159111950953709e-05,
+      "loss": 0.4522,
+      "step": 5730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8500083762981806,
+      "learning_rate": 1.1588520518219095e-05,
+      "loss": 0.5048,
+      "step": 5731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9552823604332356,
+      "learning_rate": 1.1585921416818042e-05,
+      "loss": 0.5432,
+      "step": 5732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9879107801336218,
+      "learning_rate": 1.1583322205514039e-05,
+      "loss": 0.5447,
+      "step": 5733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9424763879235039,
+      "learning_rate": 1.158072288448721e-05,
+      "loss": 0.5644,
+      "step": 5734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9149517629325111,
+      "learning_rate": 1.1578123453917692e-05,
+      "loss": 0.5157,
+      "step": 5735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8057691249709181,
+      "learning_rate": 1.1575523913985614e-05,
+      "loss": 0.5248,
+      "step": 5736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8919661942830389,
+      "learning_rate": 1.1572924264871126e-05,
+      "loss": 0.4596,
+      "step": 5737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9432369794674959,
+      "learning_rate": 1.1570324506754385e-05,
+      "loss": 0.5393,
+      "step": 5738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9501011722952974,
+      "learning_rate": 1.1567724639815546e-05,
+      "loss": 0.597,
+      "step": 5739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0192209841846716,
+      "learning_rate": 1.156512466423478e-05,
+      "loss": 0.5543,
+      "step": 5740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9118793885972419,
+      "learning_rate": 1.1562524580192265e-05,
+      "loss": 0.5442,
+      "step": 5741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8681686671544112,
+      "learning_rate": 1.155992438786818e-05,
+      "loss": 0.5597,
+      "step": 5742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8270316553924332,
+      "learning_rate": 1.1557324087442719e-05,
+      "loss": 0.5338,
+      "step": 5743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8938029173929268,
+      "learning_rate": 1.1554723679096083e-05,
+      "loss": 0.5791,
+      "step": 5744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0399905111718846,
+      "learning_rate": 1.155212316300847e-05,
+      "loss": 0.5945,
+      "step": 5745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8980299501605497,
+      "learning_rate": 1.1549522539360103e-05,
+      "loss": 0.5261,
+      "step": 5746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9107191322138124,
+      "learning_rate": 1.1546921808331196e-05,
+      "loss": 0.5396,
+      "step": 5747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8837694344631738,
+      "learning_rate": 1.1544320970101981e-05,
+      "loss": 0.5647,
+      "step": 5748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1697362226170598,
+      "learning_rate": 1.1541720024852692e-05,
+      "loss": 0.5716,
+      "step": 5749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9325464531274007,
+      "learning_rate": 1.1539118972763572e-05,
+      "loss": 0.549,
+      "step": 5750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.910145331273149,
+      "learning_rate": 1.1536517814014876e-05,
+      "loss": 0.5646,
+      "step": 5751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9314768239082533,
+      "learning_rate": 1.1533916548786856e-05,
+      "loss": 0.5403,
+      "step": 5752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8545528383721573,
+      "learning_rate": 1.153131517725978e-05,
+      "loss": 0.5583,
+      "step": 5753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9390530857052009,
+      "learning_rate": 1.1528713699613921e-05,
+      "loss": 0.5674,
+      "step": 5754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9354357281115419,
+      "learning_rate": 1.1526112116029555e-05,
+      "loss": 0.5262,
+      "step": 5755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8490438361271319,
+      "learning_rate": 1.1523510426686977e-05,
+      "loss": 0.53,
+      "step": 5756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8766802437406609,
+      "learning_rate": 1.1520908631766476e-05,
+      "loss": 0.6403,
+      "step": 5757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8991187659471015,
+      "learning_rate": 1.1518306731448357e-05,
+      "loss": 0.5231,
+      "step": 5758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9545109160032564,
+      "learning_rate": 1.1515704725912926e-05,
+      "loss": 0.5326,
+      "step": 5759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.010803460178925,
+      "learning_rate": 1.1513102615340505e-05,
+      "loss": 0.5481,
+      "step": 5760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9448566756878536,
+      "learning_rate": 1.1510500399911413e-05,
+      "loss": 0.53,
+      "step": 5761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.889478480629088,
+      "learning_rate": 1.1507898079805984e-05,
+      "loss": 0.5635,
+      "step": 5762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7889905138028853,
+      "learning_rate": 1.1505295655204557e-05,
+      "loss": 0.5205,
+      "step": 5763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9482501001206173,
+      "learning_rate": 1.1502693126287473e-05,
+      "loss": 0.5474,
+      "step": 5764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8962124749836796,
+      "learning_rate": 1.1500090493235088e-05,
+      "loss": 0.5721,
+      "step": 5765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9142381851003678,
+      "learning_rate": 1.1497487756227765e-05,
+      "loss": 0.5591,
+      "step": 5766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9397027105849795,
+      "learning_rate": 1.1494884915445867e-05,
+      "loss": 0.5431,
+      "step": 5767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.931798640912055,
+      "learning_rate": 1.1492281971069772e-05,
+      "loss": 0.6008,
+      "step": 5768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8497883566831262,
+      "learning_rate": 1.148967892327986e-05,
+      "loss": 0.4807,
+      "step": 5769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0284837210866395,
+      "learning_rate": 1.1487075772256517e-05,
+      "loss": 0.5645,
+      "step": 5770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9293165286236764,
+      "learning_rate": 1.1484472518180146e-05,
+      "loss": 0.5455,
+      "step": 5771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8389218359292604,
+      "learning_rate": 1.1481869161231146e-05,
+      "loss": 0.5453,
+      "step": 5772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9159749563956316,
+      "learning_rate": 1.1479265701589924e-05,
+      "loss": 0.5282,
+      "step": 5773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9070787010145015,
+      "learning_rate": 1.1476662139436903e-05,
+      "loss": 0.5471,
+      "step": 5774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.010745817868924,
+      "learning_rate": 1.1474058474952505e-05,
+      "loss": 0.4959,
+      "step": 5775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8536666744909506,
+      "learning_rate": 1.1471454708317163e-05,
+      "loss": 0.5393,
+      "step": 5776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9417959392783372,
+      "learning_rate": 1.1468850839711314e-05,
+      "loss": 0.5515,
+      "step": 5777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9056937867738876,
+      "learning_rate": 1.1466246869315407e-05,
+      "loss": 0.5208,
+      "step": 5778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9178472792346383,
+      "learning_rate": 1.1463642797309889e-05,
+      "loss": 0.54,
+      "step": 5779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9127826094034939,
+      "learning_rate": 1.1461038623875224e-05,
+      "loss": 0.5335,
+      "step": 5780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9142747269421215,
+      "learning_rate": 1.145843434919188e-05,
+      "loss": 0.5424,
+      "step": 5781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9521292000322145,
+      "learning_rate": 1.1455829973440328e-05,
+      "loss": 0.4752,
+      "step": 5782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9023297612335568,
+      "learning_rate": 1.1453225496801052e-05,
+      "loss": 0.5146,
+      "step": 5783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8767471354615514,
+      "learning_rate": 1.1450620919454538e-05,
+      "loss": 0.5128,
+      "step": 5784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9506251594156782,
+      "learning_rate": 1.144801624158128e-05,
+      "loss": 0.6153,
+      "step": 5785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8936628203180962,
+      "learning_rate": 1.144541146336178e-05,
+      "loss": 0.5134,
+      "step": 5786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9560096879745319,
+      "learning_rate": 1.1442806584976549e-05,
+      "loss": 0.547,
+      "step": 5787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9683698526031369,
+      "learning_rate": 1.14402016066061e-05,
+      "loss": 0.5176,
+      "step": 5788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8838884747303756,
+      "learning_rate": 1.1437596528430956e-05,
+      "loss": 0.5551,
+      "step": 5789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0051779630328477,
+      "learning_rate": 1.143499135063165e-05,
+      "loss": 0.6128,
+      "step": 5790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9116566361869124,
+      "learning_rate": 1.1432386073388718e-05,
+      "loss": 0.4915,
+      "step": 5791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9302519835107923,
+      "learning_rate": 1.1429780696882697e-05,
+      "loss": 0.5214,
+      "step": 5792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8898580580646048,
+      "learning_rate": 1.1427175221294145e-05,
+      "loss": 0.524,
+      "step": 5793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9166491607210097,
+      "learning_rate": 1.1424569646803616e-05,
+      "loss": 0.5383,
+      "step": 5794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9242787204585731,
+      "learning_rate": 1.1421963973591674e-05,
+      "loss": 0.466,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.896400261572487,
+      "learning_rate": 1.1419358201838888e-05,
+      "loss": 0.5314,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9348660459482668,
+      "learning_rate": 1.1416752331725842e-05,
+      "loss": 0.486,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9553987352915291,
+      "learning_rate": 1.1414146363433112e-05,
+      "loss": 0.4985,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9954151620187609,
+      "learning_rate": 1.1411540297141293e-05,
+      "loss": 0.5526,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8631823698760575,
+      "learning_rate": 1.1408934133030985e-05,
+      "loss": 0.5711,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8611002496151663,
+      "learning_rate": 1.1406327871282792e-05,
+      "loss": 0.5432,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9373509114087041,
+      "learning_rate": 1.1403721512077324e-05,
+      "loss": 0.5104,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8651715229849188,
+      "learning_rate": 1.14011150555952e-05,
+      "loss": 0.5173,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8591973900203775,
+      "learning_rate": 1.1398508502017047e-05,
+      "loss": 0.5324,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9105118528653173,
+      "learning_rate": 1.1395901851523494e-05,
+      "loss": 0.5249,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9613722495446257,
+      "learning_rate": 1.1393295104295178e-05,
+      "loss": 0.5782,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7954537842824811,
+      "learning_rate": 1.1390688260512755e-05,
+      "loss": 0.5114,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9248260565773117,
+      "learning_rate": 1.1388081320356861e-05,
+      "loss": 0.5524,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8469594823476866,
+      "learning_rate": 1.1385474284008167e-05,
+      "loss": 0.4896,
+      "step": 5809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8069975623735556,
+      "learning_rate": 1.1382867151647333e-05,
+      "loss": 0.5172,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9064033476636228,
+      "learning_rate": 1.1380259923455033e-05,
+      "loss": 0.5351,
+      "step": 5811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8459661189916762,
+      "learning_rate": 1.1377652599611942e-05,
+      "loss": 0.5321,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8653354047603307,
+      "learning_rate": 1.1375045180298749e-05,
+      "loss": 0.5899,
+      "step": 5813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9715947135905709,
+      "learning_rate": 1.1372437665696145e-05,
+      "loss": 0.5819,
+      "step": 5814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8273124149020723,
+      "learning_rate": 1.136983005598483e-05,
+      "loss": 0.4911,
+      "step": 5815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9421996751596458,
+      "learning_rate": 1.1367222351345504e-05,
+      "loss": 0.5739,
+      "step": 5816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9107314035512659,
+      "learning_rate": 1.136461455195888e-05,
+      "loss": 0.4639,
+      "step": 5817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8399480002609003,
+      "learning_rate": 1.1362006658005684e-05,
+      "loss": 0.4968,
+      "step": 5818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9050184961256491,
+      "learning_rate": 1.135939866966663e-05,
+      "loss": 0.5679,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8902604436710996,
+      "learning_rate": 1.1356790587122457e-05,
+      "loss": 0.5179,
+      "step": 5820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8722050269762409,
+      "learning_rate": 1.13541824105539e-05,
+      "loss": 0.543,
+      "step": 5821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9447918438261024,
+      "learning_rate": 1.1351574140141701e-05,
+      "loss": 0.5653,
+      "step": 5822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.874738767797678,
+      "learning_rate": 1.1348965776066611e-05,
+      "loss": 0.5667,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.974237757914514,
+      "learning_rate": 1.1346357318509395e-05,
+      "loss": 0.5381,
+      "step": 5824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8495673811717946,
+      "learning_rate": 1.1343748767650806e-05,
+      "loss": 0.5473,
+      "step": 5825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.941721385466632,
+      "learning_rate": 1.1341140123671621e-05,
+      "loss": 0.5479,
+      "step": 5826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9913803444659351,
+      "learning_rate": 1.1338531386752618e-05,
+      "loss": 0.5276,
+      "step": 5827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8792161364554925,
+      "learning_rate": 1.1335922557074572e-05,
+      "loss": 0.5012,
+      "step": 5828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8561321256615947,
+      "learning_rate": 1.133331363481828e-05,
+      "loss": 0.5516,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8714825042279836,
+      "learning_rate": 1.133070462016454e-05,
+      "loss": 0.4823,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8371557202330407,
+      "learning_rate": 1.1328095513294143e-05,
+      "loss": 0.5371,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8041356268035228,
+      "learning_rate": 1.1325486314387908e-05,
+      "loss": 0.4643,
+      "step": 5832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8817325273916904,
+      "learning_rate": 1.1322877023626647e-05,
+      "loss": 0.4926,
+      "step": 5833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8390037601845575,
+      "learning_rate": 1.1320267641191183e-05,
+      "loss": 0.5356,
+      "step": 5834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9465123705146559,
+      "learning_rate": 1.131765816726234e-05,
+      "loss": 0.5277,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8709013251621589,
+      "learning_rate": 1.1315048602020956e-05,
+      "loss": 0.5292,
+      "step": 5836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9327073017044094,
+      "learning_rate": 1.1312438945647873e-05,
+      "loss": 0.5589,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.913930006801104,
+      "learning_rate": 1.1309829198323929e-05,
+      "loss": 0.5097,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9536570039795854,
+      "learning_rate": 1.1307219360229991e-05,
+      "loss": 0.5517,
+      "step": 5839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8559047225531874,
+      "learning_rate": 1.1304609431546905e-05,
+      "loss": 0.4756,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8773984137628708,
+      "learning_rate": 1.1301999412455545e-05,
+      "loss": 0.5227,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.056976234813652,
+      "learning_rate": 1.129938930313678e-05,
+      "loss": 0.5863,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8246643505599478,
+      "learning_rate": 1.129677910377149e-05,
+      "loss": 0.4886,
+      "step": 5843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.949542374607715,
+      "learning_rate": 1.1294168814540554e-05,
+      "loss": 0.5693,
+      "step": 5844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0296550313854473,
+      "learning_rate": 1.1291558435624871e-05,
+      "loss": 0.5254,
+      "step": 5845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8437049883098626,
+      "learning_rate": 1.1288947967205335e-05,
+      "loss": 0.5596,
+      "step": 5846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8585532394054722,
+      "learning_rate": 1.1286337409462844e-05,
+      "loss": 0.5473,
+      "step": 5847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9820576693821286,
+      "learning_rate": 1.1283726762578316e-05,
+      "loss": 0.6178,
+      "step": 5848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9108796836809981,
+      "learning_rate": 1.128111602673266e-05,
+      "loss": 0.5125,
+      "step": 5849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8617572776326249,
+      "learning_rate": 1.1278505202106797e-05,
+      "loss": 0.5172,
+      "step": 5850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.886056947138033,
+      "learning_rate": 1.1275894288881664e-05,
+      "loss": 0.498,
+      "step": 5851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9296123160821662,
+      "learning_rate": 1.1273283287238184e-05,
+      "loss": 0.5488,
+      "step": 5852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9310583665236684,
+      "learning_rate": 1.12706721973573e-05,
+      "loss": 0.5318,
+      "step": 5853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8348396049829269,
+      "learning_rate": 1.1268061019419965e-05,
+      "loss": 0.5751,
+      "step": 5854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8750748024173053,
+      "learning_rate": 1.1265449753607122e-05,
+      "loss": 0.5463,
+      "step": 5855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8049671125117622,
+      "learning_rate": 1.1262838400099733e-05,
+      "loss": 0.4689,
+      "step": 5856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8821628993628409,
+      "learning_rate": 1.1260226959078766e-05,
+      "loss": 0.4792,
+      "step": 5857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8499335731309763,
+      "learning_rate": 1.1257615430725188e-05,
+      "loss": 0.5666,
+      "step": 5858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8016705903840738,
+      "learning_rate": 1.1255003815219973e-05,
+      "loss": 0.4629,
+      "step": 5859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8575923821384105,
+      "learning_rate": 1.1252392112744113e-05,
+      "loss": 0.5538,
+      "step": 5860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9091042763411217,
+      "learning_rate": 1.1249780323478585e-05,
+      "loss": 0.5528,
+      "step": 5861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.879419120033352,
+      "learning_rate": 1.124716844760439e-05,
+      "loss": 0.5088,
+      "step": 5862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9238669229492489,
+      "learning_rate": 1.1244556485302532e-05,
+      "loss": 0.5262,
+      "step": 5863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9622603224438493,
+      "learning_rate": 1.1241944436754008e-05,
+      "loss": 0.6327,
+      "step": 5864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8709572099514622,
+      "learning_rate": 1.1239332302139839e-05,
+      "loss": 0.6088,
+      "step": 5865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8728800808030194,
+      "learning_rate": 1.1236720081641042e-05,
+      "loss": 0.5195,
+      "step": 5866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9157664222436604,
+      "learning_rate": 1.1234107775438637e-05,
+      "loss": 0.521,
+      "step": 5867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9421227582596474,
+      "learning_rate": 1.1231495383713657e-05,
+      "loss": 0.5419,
+      "step": 5868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8996081993865935,
+      "learning_rate": 1.1228882906647142e-05,
+      "loss": 0.4896,
+      "step": 5869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8783857047148886,
+      "learning_rate": 1.1226270344420131e-05,
+      "loss": 0.516,
+      "step": 5870
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7932167932457265,
+      "learning_rate": 1.1223657697213672e-05,
+      "loss": 0.4892,
+      "step": 5871
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9248614707298407,
+      "learning_rate": 1.1221044965208821e-05,
+      "loss": 0.5565,
+      "step": 5872
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8308674630147707,
+      "learning_rate": 1.1218432148586638e-05,
+      "loss": 0.5883,
+      "step": 5873
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8704808487120602,
+      "learning_rate": 1.1215819247528186e-05,
+      "loss": 0.5364,
+      "step": 5874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9057993331161298,
+      "learning_rate": 1.121320626221454e-05,
+      "loss": 0.5342,
+      "step": 5875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9622876164726756,
+      "learning_rate": 1.1210593192826776e-05,
+      "loss": 0.6037,
+      "step": 5876
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8488925867142036,
+      "learning_rate": 1.1207980039545976e-05,
+      "loss": 0.5314,
+      "step": 5877
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8730947745792921,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.5412,
+      "step": 5878
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9506959222264564,
+      "learning_rate": 1.1202753482029639e-05,
+      "loss": 0.6111,
+      "step": 5879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8871307714612382,
+      "learning_rate": 1.1200140078156293e-05,
+      "loss": 0.5414,
+      "step": 5880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9470858331126085,
+      "learning_rate": 1.1197526591114306e-05,
+      "loss": 0.5387,
+      "step": 5881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.968646981582715,
+      "learning_rate": 1.119491302108479e-05,
+      "loss": 0.6078,
+      "step": 5882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9271943209176021,
+      "learning_rate": 1.1192299368248858e-05,
+      "loss": 0.569,
+      "step": 5883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.899014706576815,
+      "learning_rate": 1.1189685632787638e-05,
+      "loss": 0.5318,
+      "step": 5884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8912231576345668,
+      "learning_rate": 1.1187071814882262e-05,
+      "loss": 0.5458,
+      "step": 5885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9244677150779027,
+      "learning_rate": 1.118445791471386e-05,
+      "loss": 0.5491,
+      "step": 5886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7735443209972106,
+      "learning_rate": 1.1181843932463577e-05,
+      "loss": 0.4898,
+      "step": 5887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8876592892316715,
+      "learning_rate": 1.1179229868312555e-05,
+      "loss": 0.4807,
+      "step": 5888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8108220261166168,
+      "learning_rate": 1.117661572244195e-05,
+      "loss": 0.4873,
+      "step": 5889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9695562013180646,
+      "learning_rate": 1.117400149503292e-05,
+      "loss": 0.6291,
+      "step": 5890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0105840919169105,
+      "learning_rate": 1.1171387186266628e-05,
+      "loss": 0.6225,
+      "step": 5891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8261616963980778,
+      "learning_rate": 1.1168772796324241e-05,
+      "loss": 0.4972,
+      "step": 5892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.938537542798799,
+      "learning_rate": 1.1166158325386938e-05,
+      "loss": 0.5204,
+      "step": 5893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8614582593712691,
+      "learning_rate": 1.1163543773635896e-05,
+      "loss": 0.5719,
+      "step": 5894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9374411897404901,
+      "learning_rate": 1.1160929141252303e-05,
+      "loss": 0.491,
+      "step": 5895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9359501939977077,
+      "learning_rate": 1.115831442841735e-05,
+      "loss": 0.4821,
+      "step": 5896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9385210570409233,
+      "learning_rate": 1.1155699635312235e-05,
+      "loss": 0.5466,
+      "step": 5897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8988786435675817,
+      "learning_rate": 1.1153084762118163e-05,
+      "loss": 0.5296,
+      "step": 5898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9405919058361396,
+      "learning_rate": 1.1150469809016336e-05,
+      "loss": 0.5517,
+      "step": 5899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9164295167852962,
+      "learning_rate": 1.1147854776187973e-05,
+      "loss": 0.6041,
+      "step": 5900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0104231088332736,
+      "learning_rate": 1.1145239663814291e-05,
+      "loss": 0.5689,
+      "step": 5901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9624897296538513,
+      "learning_rate": 1.1142624472076518e-05,
+      "loss": 0.643,
+      "step": 5902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9035371445616356,
+      "learning_rate": 1.1140009201155881e-05,
+      "loss": 0.5335,
+      "step": 5903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8552416264952798,
+      "learning_rate": 1.1137393851233618e-05,
+      "loss": 0.5205,
+      "step": 5904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8690331556336678,
+      "learning_rate": 1.1134778422490971e-05,
+      "loss": 0.5088,
+      "step": 5905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8007198775976306,
+      "learning_rate": 1.1132162915109186e-05,
+      "loss": 0.5099,
+      "step": 5906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8398136874931055,
+      "learning_rate": 1.1129547329269512e-05,
+      "loss": 0.4875,
+      "step": 5907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.968349450845184,
+      "learning_rate": 1.1126931665153213e-05,
+      "loss": 0.6041,
+      "step": 5908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8786642078065504,
+      "learning_rate": 1.1124315922941549e-05,
+      "loss": 0.4978,
+      "step": 5909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9614973758016057,
+      "learning_rate": 1.1121700102815787e-05,
+      "loss": 0.5669,
+      "step": 5910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9050042577946916,
+      "learning_rate": 1.1119084204957204e-05,
+      "loss": 0.5286,
+      "step": 5911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8748269695827618,
+      "learning_rate": 1.1116468229547079e-05,
+      "loss": 0.4821,
+      "step": 5912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9758458472696362,
+      "learning_rate": 1.1113852176766695e-05,
+      "loss": 0.5561,
+      "step": 5913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.968997489808725,
+      "learning_rate": 1.1111236046797342e-05,
+      "loss": 0.5863,
+      "step": 5914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8520026964412684,
+      "learning_rate": 1.110861983982032e-05,
+      "loss": 0.5155,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9705553198685964,
+      "learning_rate": 1.1106003556016924e-05,
+      "loss": 0.5326,
+      "step": 5916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.942824468226576,
+      "learning_rate": 1.1103387195568463e-05,
+      "loss": 0.5439,
+      "step": 5917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.002773840554435,
+      "learning_rate": 1.110077075865625e-05,
+      "loss": 0.6232,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9123206687630057,
+      "learning_rate": 1.1098154245461597e-05,
+      "loss": 0.5355,
+      "step": 5919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8978286343392776,
+      "learning_rate": 1.109553765616583e-05,
+      "loss": 0.5299,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9972585079973012,
+      "learning_rate": 1.1092920990950276e-05,
+      "loss": 0.5836,
+      "step": 5921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8419429273808501,
+      "learning_rate": 1.1090304249996264e-05,
+      "loss": 0.4881,
+      "step": 5922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9619008015768903,
+      "learning_rate": 1.1087687433485135e-05,
+      "loss": 0.4746,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9965086736539188,
+      "learning_rate": 1.1085070541598235e-05,
+      "loss": 0.5737,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8223959891987548,
+      "learning_rate": 1.1082453574516907e-05,
+      "loss": 0.5213,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9194358573261691,
+      "learning_rate": 1.1079836532422505e-05,
+      "loss": 0.5685,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8433036241482647,
+      "learning_rate": 1.1077219415496391e-05,
+      "loss": 0.5119,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8369149997767066,
+      "learning_rate": 1.1074602223919925e-05,
+      "loss": 0.5315,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.962608668637937,
+      "learning_rate": 1.107198495787448e-05,
+      "loss": 0.5197,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9299123924527506,
+      "learning_rate": 1.106936761754143e-05,
+      "loss": 0.4922,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9346185712137653,
+      "learning_rate": 1.1066750203102148e-05,
+      "loss": 0.5297,
+      "step": 5931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8796465411194002,
+      "learning_rate": 1.1064132714738024e-05,
+      "loss": 0.5336,
+      "step": 5932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9216539178960147,
+      "learning_rate": 1.1061515152630448e-05,
+      "loss": 0.6052,
+      "step": 5933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9549392231390016,
+      "learning_rate": 1.1058897516960817e-05,
+      "loss": 0.6129,
+      "step": 5934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9376318659921885,
+      "learning_rate": 1.1056279807910522e-05,
+      "loss": 0.6062,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9601200523129949,
+      "learning_rate": 1.1053662025660973e-05,
+      "loss": 0.4865,
+      "step": 5936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9388283502974828,
+      "learning_rate": 1.1051044170393586e-05,
+      "loss": 0.5551,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9367217738098178,
+      "learning_rate": 1.1048426242289767e-05,
+      "loss": 0.5412,
+      "step": 5938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9095242772845795,
+      "learning_rate": 1.1045808241530937e-05,
+      "loss": 0.5014,
+      "step": 5939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9201435611696015,
+      "learning_rate": 1.1043190168298527e-05,
+      "loss": 0.5581,
+      "step": 5940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.942430264175165,
+      "learning_rate": 1.104057202277396e-05,
+      "loss": 0.5376,
+      "step": 5941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9284963591597374,
+      "learning_rate": 1.1037953805138679e-05,
+      "loss": 0.5984,
+      "step": 5942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9806126690798201,
+      "learning_rate": 1.103533551557412e-05,
+      "loss": 0.5994,
+      "step": 5943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9852923229857014,
+      "learning_rate": 1.1032717154261725e-05,
+      "loss": 0.5086,
+      "step": 5944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8857293632409518,
+      "learning_rate": 1.1030098721382947e-05,
+      "loss": 0.5406,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9089245246997715,
+      "learning_rate": 1.1027480217119245e-05,
+      "loss": 0.5156,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8422760311162406,
+      "learning_rate": 1.102486164165207e-05,
+      "loss": 0.5645,
+      "step": 5947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8867799252180668,
+      "learning_rate": 1.1022242995162895e-05,
+      "loss": 0.5605,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8775600612421792,
+      "learning_rate": 1.1019624277833188e-05,
+      "loss": 0.6068,
+      "step": 5949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.919899703948138,
+      "learning_rate": 1.1017005489844424e-05,
+      "loss": 0.4996,
+      "step": 5950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9793631542392197,
+      "learning_rate": 1.1014386631378079e-05,
+      "loss": 0.5301,
+      "step": 5951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8502436939737742,
+      "learning_rate": 1.101176770261564e-05,
+      "loss": 0.5321,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9908848621988976,
+      "learning_rate": 1.1009148703738599e-05,
+      "loss": 0.5246,
+      "step": 5953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8478840248580101,
+      "learning_rate": 1.1006529634928446e-05,
+      "loss": 0.503,
+      "step": 5954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9149829104564555,
+      "learning_rate": 1.1003910496366683e-05,
+      "loss": 0.5766,
+      "step": 5955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3735137099920995,
+      "learning_rate": 1.1001291288234812e-05,
+      "loss": 0.5335,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0109614737297976,
+      "learning_rate": 1.0998672010714344e-05,
+      "loss": 0.5876,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9525877879244522,
+      "learning_rate": 1.0996052663986791e-05,
+      "loss": 0.5306,
+      "step": 5958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.872468204745674,
+      "learning_rate": 1.0993433248233672e-05,
+      "loss": 0.5275,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9800315769595653,
+      "learning_rate": 1.0990813763636511e-05,
+      "loss": 0.5752,
+      "step": 5960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9484721067744775,
+      "learning_rate": 1.0988194210376834e-05,
+      "loss": 0.5565,
+      "step": 5961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.970274479323285,
+      "learning_rate": 1.0985574588636174e-05,
+      "loss": 0.5795,
+      "step": 5962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8814469628462033,
+      "learning_rate": 1.0982954898596072e-05,
+      "loss": 0.4921,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9531102160371265,
+      "learning_rate": 1.0980335140438066e-05,
+      "loss": 0.5659,
+      "step": 5964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8076295124586604,
+      "learning_rate": 1.0977715314343702e-05,
+      "loss": 0.5268,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9189308585416595,
+      "learning_rate": 1.0975095420494537e-05,
+      "loss": 0.529,
+      "step": 5966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8835362326913786,
+      "learning_rate": 1.0972475459072124e-05,
+      "loss": 0.5095,
+      "step": 5967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9200688922262339,
+      "learning_rate": 1.0969855430258022e-05,
+      "loss": 0.5085,
+      "step": 5968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0103589244278846,
+      "learning_rate": 1.0967235334233802e-05,
+      "loss": 0.6143,
+      "step": 5969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8747556366743678,
+      "learning_rate": 1.096461517118103e-05,
+      "loss": 0.5439,
+      "step": 5970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8383655597267304,
+      "learning_rate": 1.096199494128128e-05,
+      "loss": 0.5071,
+      "step": 5971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9352976550371722,
+      "learning_rate": 1.0959374644716137e-05,
+      "loss": 0.5272,
+      "step": 5972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9736793571068132,
+      "learning_rate": 1.0956754281667182e-05,
+      "loss": 0.5717,
+      "step": 5973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8320012054898668,
+      "learning_rate": 1.0954133852316003e-05,
+      "loss": 0.4946,
+      "step": 5974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9549832251089269,
+      "learning_rate": 1.0951513356844192e-05,
+      "loss": 0.5193,
+      "step": 5975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7905276524529273,
+      "learning_rate": 1.0948892795433353e-05,
+      "loss": 0.4973,
+      "step": 5976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9168115233955988,
+      "learning_rate": 1.0946272168265081e-05,
+      "loss": 0.5515,
+      "step": 5977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9139760132751583,
+      "learning_rate": 1.094365147552099e-05,
+      "loss": 0.5441,
+      "step": 5978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9405480838687121,
+      "learning_rate": 1.094103071738269e-05,
+      "loss": 0.5967,
+      "step": 5979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8594264268924838,
+      "learning_rate": 1.0938409894031793e-05,
+      "loss": 0.5401,
+      "step": 5980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8756876468757003,
+      "learning_rate": 1.0935789005649924e-05,
+      "loss": 0.5171,
+      "step": 5981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8974310052441253,
+      "learning_rate": 1.0933168052418708e-05,
+      "loss": 0.5579,
+      "step": 5982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9251336556455075,
+      "learning_rate": 1.0930547034519772e-05,
+      "loss": 0.5695,
+      "step": 5983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9314886487604079,
+      "learning_rate": 1.0927925952134753e-05,
+      "loss": 0.5214,
+      "step": 5984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.869125763237957,
+      "learning_rate": 1.092530480544529e-05,
+      "loss": 0.501,
+      "step": 5985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.915804992331489,
+      "learning_rate": 1.092268359463302e-05,
+      "loss": 0.5217,
+      "step": 5986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8962623903964275,
+      "learning_rate": 1.0920062319879599e-05,
+      "loss": 0.5452,
+      "step": 5987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8486741066180117,
+      "learning_rate": 1.0917440981366677e-05,
+      "loss": 0.5229,
+      "step": 5988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9043838061637083,
+      "learning_rate": 1.0914819579275903e-05,
+      "loss": 0.5257,
+      "step": 5989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8980008090414996,
+      "learning_rate": 1.0912198113788947e-05,
+      "loss": 0.539,
+      "step": 5990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8991497597361532,
+      "learning_rate": 1.0909576585087472e-05,
+      "loss": 0.5007,
+      "step": 5991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8883674978924093,
+      "learning_rate": 1.0906954993353145e-05,
+      "loss": 0.5107,
+      "step": 5992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8998564171217721,
+      "learning_rate": 1.0904333338767641e-05,
+      "loss": 0.5625,
+      "step": 5993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8363802061981552,
+      "learning_rate": 1.090171162151264e-05,
+      "loss": 0.5212,
+      "step": 5994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9017747043223552,
+      "learning_rate": 1.0899089841769824e-05,
+      "loss": 0.5528,
+      "step": 5995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9231025436302872,
+      "learning_rate": 1.0896467999720876e-05,
+      "loss": 0.5525,
+      "step": 5996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8912850378764711,
+      "learning_rate": 1.0893846095547493e-05,
+      "loss": 0.5891,
+      "step": 5997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9500971063396939,
+      "learning_rate": 1.0891224129431368e-05,
+      "loss": 0.5333,
+      "step": 5998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8674435355298813,
+      "learning_rate": 1.0888602101554202e-05,
+      "loss": 0.4728,
+      "step": 5999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8477710168689605,
+      "learning_rate": 1.0885980012097698e-05,
+      "loss": 0.4872,
+      "step": 6000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9681405308198463,
+      "learning_rate": 1.0883357861243567e-05,
+      "loss": 0.5292,
+      "step": 6001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8159825384348709,
+      "learning_rate": 1.0880735649173518e-05,
+      "loss": 0.5195,
+      "step": 6002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8594649461848198,
+      "learning_rate": 1.0878113376069268e-05,
+      "loss": 0.5064,
+      "step": 6003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8755507296784738,
+      "learning_rate": 1.0875491042112543e-05,
+      "loss": 0.4985,
+      "step": 6004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1053674630111627,
+      "learning_rate": 1.0872868647485064e-05,
+      "loss": 0.5157,
+      "step": 6005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8650612305952349,
+      "learning_rate": 1.087024619236856e-05,
+      "loss": 0.5791,
+      "step": 6006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9342489013084267,
+      "learning_rate": 1.0867623676944771e-05,
+      "loss": 0.5481,
+      "step": 6007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8953188110291666,
+      "learning_rate": 1.0865001101395429e-05,
+      "loss": 0.4963,
+      "step": 6008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9266589925253965,
+      "learning_rate": 1.0862378465902276e-05,
+      "loss": 0.5699,
+      "step": 6009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.876106966122857,
+      "learning_rate": 1.0859755770647063e-05,
+      "loss": 0.5398,
+      "step": 6010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9158713362331197,
+      "learning_rate": 1.0857133015811537e-05,
+      "loss": 0.5676,
+      "step": 6011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8439366191962501,
+      "learning_rate": 1.0854510201577451e-05,
+      "loss": 0.4834,
+      "step": 6012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9129278894147207,
+      "learning_rate": 1.0851887328126569e-05,
+      "loss": 0.4796,
+      "step": 6013
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9651114390249649,
+      "learning_rate": 1.084926439564065e-05,
+      "loss": 0.519,
+      "step": 6014
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9133574311556374,
+      "learning_rate": 1.084664140430146e-05,
+      "loss": 0.569,
+      "step": 6015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9137297031652001,
+      "learning_rate": 1.0844018354290776e-05,
+      "loss": 0.5623,
+      "step": 6016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8509472903986172,
+      "learning_rate": 1.0841395245790363e-05,
+      "loss": 0.5037,
+      "step": 6017
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9168976366504523,
+      "learning_rate": 1.0838772078982008e-05,
+      "loss": 0.5698,
+      "step": 6018
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.902456982349151,
+      "learning_rate": 1.0836148854047494e-05,
+      "loss": 0.5128,
+      "step": 6019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.922315991787359,
+      "learning_rate": 1.0833525571168603e-05,
+      "loss": 0.4852,
+      "step": 6020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9847818345055375,
+      "learning_rate": 1.0830902230527129e-05,
+      "loss": 0.5451,
+      "step": 6021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9403531190902872,
+      "learning_rate": 1.082827883230487e-05,
+      "loss": 0.5087,
+      "step": 6022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9536798293761942,
+      "learning_rate": 1.0825655376683621e-05,
+      "loss": 0.5691,
+      "step": 6023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8687923617162312,
+      "learning_rate": 1.0823031863845189e-05,
+      "loss": 0.5028,
+      "step": 6024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9344505134076202,
+      "learning_rate": 1.082040829397138e-05,
+      "loss": 0.5811,
+      "step": 6025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0892474648248411,
+      "learning_rate": 1.0817784667243998e-05,
+      "loss": 0.5767,
+      "step": 6026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9631439527676332,
+      "learning_rate": 1.0815160983844865e-05,
+      "loss": 0.5175,
+      "step": 6027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9986721453630223,
+      "learning_rate": 1.0812537243955804e-05,
+      "loss": 0.5998,
+      "step": 6028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9578778424630721,
+      "learning_rate": 1.0809913447758628e-05,
+      "loss": 0.5775,
+      "step": 6029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9161777185614014,
+      "learning_rate": 1.080728959543517e-05,
+      "loss": 0.5776,
+      "step": 6030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9106523123320278,
+      "learning_rate": 1.0804665687167262e-05,
+      "loss": 0.5458,
+      "step": 6031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8102666418535739,
+      "learning_rate": 1.0802041723136731e-05,
+      "loss": 0.5212,
+      "step": 6032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9620209665623649,
+      "learning_rate": 1.079941770352542e-05,
+      "loss": 0.5381,
+      "step": 6033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9765426156604666,
+      "learning_rate": 1.0796793628515176e-05,
+      "loss": 0.5639,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9410479105661026,
+      "learning_rate": 1.0794169498287837e-05,
+      "loss": 0.5848,
+      "step": 6035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9258676956045075,
+      "learning_rate": 1.0791545313025255e-05,
+      "loss": 0.5677,
+      "step": 6036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9773392276289621,
+      "learning_rate": 1.078892107290929e-05,
+      "loss": 0.634,
+      "step": 6037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9256631034571904,
+      "learning_rate": 1.0786296778121787e-05,
+      "loss": 0.5795,
+      "step": 6038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.933137084813143,
+      "learning_rate": 1.078367242884462e-05,
+      "loss": 0.5872,
+      "step": 6039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9953572621075247,
+      "learning_rate": 1.0781048025259648e-05,
+      "loss": 0.6192,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0050752288603593,
+      "learning_rate": 1.0778423567548739e-05,
+      "loss": 0.5837,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8950958764206185,
+      "learning_rate": 1.0775799055893768e-05,
+      "loss": 0.5146,
+      "step": 6042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9012573457259359,
+      "learning_rate": 1.0773174490476613e-05,
+      "loss": 0.5663,
+      "step": 6043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9568547546834592,
+      "learning_rate": 1.0770549871479149e-05,
+      "loss": 0.5659,
+      "step": 6044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.921074163124563,
+      "learning_rate": 1.0767925199083262e-05,
+      "loss": 0.5878,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8790758009068457,
+      "learning_rate": 1.0765300473470841e-05,
+      "loss": 0.4825,
+      "step": 6046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9692177897991965,
+      "learning_rate": 1.0762675694823777e-05,
+      "loss": 0.5294,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7912159684386474,
+      "learning_rate": 1.0760050863323961e-05,
+      "loss": 0.4587,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.902387535550519,
+      "learning_rate": 1.0757425979153297e-05,
+      "loss": 0.5185,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.90900599928075,
+      "learning_rate": 1.0754801042493683e-05,
+      "loss": 0.5249,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9150820410769867,
+      "learning_rate": 1.0752176053527025e-05,
+      "loss": 0.5712,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9762071829758882,
+      "learning_rate": 1.0749551012435237e-05,
+      "loss": 0.573,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8573612056669551,
+      "learning_rate": 1.0746925919400226e-05,
+      "loss": 0.4773,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.911707052115374,
+      "learning_rate": 1.0744300774603914e-05,
+      "loss": 0.5026,
+      "step": 6054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9979792020943502,
+      "learning_rate": 1.0741675578228216e-05,
+      "loss": 0.6148,
+      "step": 6055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9553815644356218,
+      "learning_rate": 1.073905033045506e-05,
+      "loss": 0.5726,
+      "step": 6056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9136612647564725,
+      "learning_rate": 1.0736425031466369e-05,
+      "loss": 0.5387,
+      "step": 6057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.973117546702641,
+      "learning_rate": 1.0733799681444077e-05,
+      "loss": 0.5942,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9243498996981262,
+      "learning_rate": 1.073117428057012e-05,
+      "loss": 0.5464,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9654957940954922,
+      "learning_rate": 1.0728548829026433e-05,
+      "loss": 0.5123,
+      "step": 6060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8545999154255652,
+      "learning_rate": 1.0725923326994958e-05,
+      "loss": 0.502,
+      "step": 6061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9459939415916986,
+      "learning_rate": 1.0723297774657642e-05,
+      "loss": 0.6115,
+      "step": 6062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8484583387652496,
+      "learning_rate": 1.0720672172196432e-05,
+      "loss": 0.5328,
+      "step": 6063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.827531492547263,
+      "learning_rate": 1.0718046519793276e-05,
+      "loss": 0.4541,
+      "step": 6064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8574412032251646,
+      "learning_rate": 1.0715420817630137e-05,
+      "loss": 0.4967,
+      "step": 6065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8550222014048962,
+      "learning_rate": 1.0712795065888968e-05,
+      "loss": 0.4633,
+      "step": 6066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9021045314787502,
+      "learning_rate": 1.0710169264751733e-05,
+      "loss": 0.5815,
+      "step": 6067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9894222824218805,
+      "learning_rate": 1.0707543414400398e-05,
+      "loss": 0.528,
+      "step": 6068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9728184096575895,
+      "learning_rate": 1.0704917515016933e-05,
+      "loss": 0.5035,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9641583065696216,
+      "learning_rate": 1.0702291566783307e-05,
+      "loss": 0.5558,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8391781018752039,
+      "learning_rate": 1.0699665569881503e-05,
+      "loss": 0.5121,
+      "step": 6071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8621647843592861,
+      "learning_rate": 1.0697039524493492e-05,
+      "loss": 0.5277,
+      "step": 6072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7567533685287966,
+      "learning_rate": 1.069441343080126e-05,
+      "loss": 0.4858,
+      "step": 6073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8648137761316245,
+      "learning_rate": 1.0691787288986795e-05,
+      "loss": 0.532,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9042486379215297,
+      "learning_rate": 1.0689161099232084e-05,
+      "loss": 0.526,
+      "step": 6075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.906851550856305,
+      "learning_rate": 1.0686534861719118e-05,
+      "loss": 0.5236,
+      "step": 6076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9131446706717494,
+      "learning_rate": 1.06839085766299e-05,
+      "loss": 0.4741,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8710047907608442,
+      "learning_rate": 1.068128224414642e-05,
+      "loss": 0.5528,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9268905719223248,
+      "learning_rate": 1.0678655864450684e-05,
+      "loss": 0.5913,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9435191680227256,
+      "learning_rate": 1.0676029437724703e-05,
+      "loss": 0.5458,
+      "step": 6080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8724092493981516,
+      "learning_rate": 1.0673402964150479e-05,
+      "loss": 0.5784,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8601665421495636,
+      "learning_rate": 1.0670776443910024e-05,
+      "loss": 0.5473,
+      "step": 6082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9680710017013635,
+      "learning_rate": 1.0668149877185361e-05,
+      "loss": 0.6032,
+      "step": 6083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9492639664421051,
+      "learning_rate": 1.0665523264158501e-05,
+      "loss": 0.5389,
+      "step": 6084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8455152863436715,
+      "learning_rate": 1.0662896605011472e-05,
+      "loss": 0.566,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.85243477804133,
+      "learning_rate": 1.0660269899926296e-05,
+      "loss": 0.5244,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9435897995784096,
+      "learning_rate": 1.0657643149084999e-05,
+      "loss": 0.5268,
+      "step": 6087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9004172884401473,
+      "learning_rate": 1.0655016352669616e-05,
+      "loss": 0.5125,
+      "step": 6088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8696831235448563,
+      "learning_rate": 1.0652389510862182e-05,
+      "loss": 0.4969,
+      "step": 6089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8757254452087241,
+      "learning_rate": 1.0649762623844733e-05,
+      "loss": 0.5422,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8533339193252824,
+      "learning_rate": 1.064713569179931e-05,
+      "loss": 0.49,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.929141233600302,
+      "learning_rate": 1.064450871490796e-05,
+      "loss": 0.5793,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9853060175197331,
+      "learning_rate": 1.0641881693352724e-05,
+      "loss": 0.5836,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8969091374661617,
+      "learning_rate": 1.0639254627315658e-05,
+      "loss": 0.572,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9445842218624012,
+      "learning_rate": 1.0636627516978815e-05,
+      "loss": 0.5491,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8521417788675616,
+      "learning_rate": 1.0634000362524247e-05,
+      "loss": 0.4567,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8243800789666074,
+      "learning_rate": 1.0631373164134015e-05,
+      "loss": 0.5237,
+      "step": 6097
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8972677822447758,
+      "learning_rate": 1.0628745921990184e-05,
+      "loss": 0.5347,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9096471053000708,
+      "learning_rate": 1.062611863627482e-05,
+      "loss": 0.5355,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8451782292679187,
+      "learning_rate": 1.062349130716999e-05,
+      "loss": 0.4855,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8535857129649261,
+      "learning_rate": 1.0620863934857764e-05,
+      "loss": 0.5551,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.816157477591498,
+      "learning_rate": 1.0618236519520219e-05,
+      "loss": 0.501,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0310068024642354,
+      "learning_rate": 1.0615609061339431e-05,
+      "loss": 0.6114,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9423701264828378,
+      "learning_rate": 1.061298156049748e-05,
+      "loss": 0.5538,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8732552391809214,
+      "learning_rate": 1.061035401717645e-05,
+      "loss": 0.5382,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8797775091115982,
+      "learning_rate": 1.0607726431558431e-05,
+      "loss": 0.5089,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9400596223756281,
+      "learning_rate": 1.060509880382551e-05,
+      "loss": 0.5566,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8467700720411315,
+      "learning_rate": 1.0602471134159773e-05,
+      "loss": 0.5317,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.822881476771488,
+      "learning_rate": 1.0599843422743328e-05,
+      "loss": 0.505,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.945363657168463,
+      "learning_rate": 1.059721566975826e-05,
+      "loss": 0.5576,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8550973491019075,
+      "learning_rate": 1.0594587875386677e-05,
+      "loss": 0.4685,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8978924311350094,
+      "learning_rate": 1.0591960039810684e-05,
+      "loss": 0.5041,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9659817672441282,
+      "learning_rate": 1.0589332163212384e-05,
+      "loss": 0.5497,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9426550282970327,
+      "learning_rate": 1.0586704245773886e-05,
+      "loss": 0.5407,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9234953615378141,
+      "learning_rate": 1.0584076287677307e-05,
+      "loss": 0.6009,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8702469617913521,
+      "learning_rate": 1.0581448289104759e-05,
+      "loss": 0.493,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9994597487559939,
+      "learning_rate": 1.057882025023836e-05,
+      "loss": 0.5883,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.918411318404071,
+      "learning_rate": 1.0576192171260228e-05,
+      "loss": 0.5317,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9373647802629542,
+      "learning_rate": 1.0573564052352496e-05,
+      "loss": 0.5399,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8050241018619839,
+      "learning_rate": 1.0570935893697278e-05,
+      "loss": 0.4442,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9022679662284697,
+      "learning_rate": 1.0568307695476712e-05,
+      "loss": 0.5033,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8846901204793555,
+      "learning_rate": 1.0565679457872928e-05,
+      "loss": 0.5182,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0059532477165203,
+      "learning_rate": 1.0563051181068056e-05,
+      "loss": 0.5464,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9327245880379895,
+      "learning_rate": 1.0560422865244237e-05,
+      "loss": 0.5133,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8496071103162126,
+      "learning_rate": 1.0557794510583611e-05,
+      "loss": 0.4938,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8298240451208877,
+      "learning_rate": 1.0555166117268322e-05,
+      "loss": 0.5385,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.934187978424719,
+      "learning_rate": 1.0552537685480512e-05,
+      "loss": 0.5143,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8674548058152174,
+      "learning_rate": 1.054990921540233e-05,
+      "loss": 0.5794,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0370693329998053,
+      "learning_rate": 1.054728070721593e-05,
+      "loss": 0.5977,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8819244109215885,
+      "learning_rate": 1.0544652161103459e-05,
+      "loss": 0.5054,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8960578144446292,
+      "learning_rate": 1.0542023577247076e-05,
+      "loss": 0.545,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.886580731545749,
+      "learning_rate": 1.0539394955828944e-05,
+      "loss": 0.5429,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9374761540391363,
+      "learning_rate": 1.0536766297031216e-05,
+      "loss": 0.6374,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9103004242142594,
+      "learning_rate": 1.053413760103606e-05,
+      "loss": 0.5159,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8968055272655576,
+      "learning_rate": 1.0531508868025647e-05,
+      "loss": 0.5283,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.970055594503045,
+      "learning_rate": 1.0528880098182136e-05,
+      "loss": 0.5435,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9389369412572518,
+      "learning_rate": 1.0526251291687703e-05,
+      "loss": 0.5747,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8078187197158875,
+      "learning_rate": 1.0523622448724524e-05,
+      "loss": 0.5405,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9190980092900863,
+      "learning_rate": 1.0520993569474773e-05,
+      "loss": 0.5321,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.958215322494185,
+      "learning_rate": 1.051836465412063e-05,
+      "loss": 0.5699,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8721689328952731,
+      "learning_rate": 1.051573570284428e-05,
+      "loss": 0.6109,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0429424389942195,
+      "learning_rate": 1.0513106715827897e-05,
+      "loss": 0.5949,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8826173376580875,
+      "learning_rate": 1.0510477693253676e-05,
+      "loss": 0.4822,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9558074179782857,
+      "learning_rate": 1.0507848635303805e-05,
+      "loss": 0.5409,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0047521941779742,
+      "learning_rate": 1.0505219542160474e-05,
+      "loss": 0.5804,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0319590843517332,
+      "learning_rate": 1.0502590414005875e-05,
+      "loss": 0.5394,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8462689043839046,
+      "learning_rate": 1.0499961251022208e-05,
+      "loss": 0.5235,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9645087832836281,
+      "learning_rate": 1.049733205339167e-05,
+      "loss": 0.5744,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9931263205618819,
+      "learning_rate": 1.0494702821296458e-05,
+      "loss": 0.5615,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8977359164608016,
+      "learning_rate": 1.0492073554918782e-05,
+      "loss": 0.5124,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9580917800214114,
+      "learning_rate": 1.0489444254440846e-05,
+      "loss": 0.556,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9868313720168621,
+      "learning_rate": 1.0486814920044857e-05,
+      "loss": 0.6069,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9786912759940932,
+      "learning_rate": 1.0484185551913027e-05,
+      "loss": 0.5963,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8802270262465605,
+      "learning_rate": 1.0481556150227562e-05,
+      "loss": 0.4737,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9904658437694336,
+      "learning_rate": 1.0478926715170687e-05,
+      "loss": 0.5483,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.922169756232924,
+      "learning_rate": 1.0476297246924619e-05,
+      "loss": 0.5358,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9532812373990939,
+      "learning_rate": 1.047366774567157e-05,
+      "loss": 0.4799,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9210650459076102,
+      "learning_rate": 1.0471038211593764e-05,
+      "loss": 0.5301,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7225895551992169,
+      "learning_rate": 1.0468408644873433e-05,
+      "loss": 0.4377,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9545247474276838,
+      "learning_rate": 1.0465779045692796e-05,
+      "loss": 0.564,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9428918239243667,
+      "learning_rate": 1.0463149414234084e-05,
+      "loss": 0.5187,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9060269955149981,
+      "learning_rate": 1.046051975067953e-05,
+      "loss": 0.5088,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9264511973693403,
+      "learning_rate": 1.0457890055211364e-05,
+      "loss": 0.5766,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.899604789095249,
+      "learning_rate": 1.0455260328011822e-05,
+      "loss": 0.516,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9773994581290497,
+      "learning_rate": 1.0452630569263147e-05,
+      "loss": 0.6503,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9720121111533556,
+      "learning_rate": 1.0450000779147573e-05,
+      "loss": 0.5527,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9737789980852884,
+      "learning_rate": 1.0447370957847343e-05,
+      "loss": 0.5923,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8662034958602722,
+      "learning_rate": 1.0444741105544705e-05,
+      "loss": 0.5748,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9015025314707872,
+      "learning_rate": 1.04421112224219e-05,
+      "loss": 0.5418,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8663857088672641,
+      "learning_rate": 1.0439481308661181e-05,
+      "loss": 0.515,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9607331905556366,
+      "learning_rate": 1.0436851364444798e-05,
+      "loss": 0.5785,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8698730300041885,
+      "learning_rate": 1.0434221389955002e-05,
+      "loss": 0.4768,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9302907035593343,
+      "learning_rate": 1.0431591385374047e-05,
+      "loss": 0.5422,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8590944033288993,
+      "learning_rate": 1.0428961350884194e-05,
+      "loss": 0.5405,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8604175708088144,
+      "learning_rate": 1.0426331286667701e-05,
+      "loss": 0.5365,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8882444903167186,
+      "learning_rate": 1.0423701192906825e-05,
+      "loss": 0.5045,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8505136672359528,
+      "learning_rate": 1.0421071069783834e-05,
+      "loss": 0.4987,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8571941770141829,
+      "learning_rate": 1.0418440917480992e-05,
+      "loss": 0.5007,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9102254227045845,
+      "learning_rate": 1.0415810736180563e-05,
+      "loss": 0.5232,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7864199735738805,
+      "learning_rate": 1.0413180526064824e-05,
+      "loss": 0.4712,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8072441676513875,
+      "learning_rate": 1.0410550287316035e-05,
+      "loss": 0.5144,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.850410157333761,
+      "learning_rate": 1.0407920020116477e-05,
+      "loss": 0.5598,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9830887682458233,
+      "learning_rate": 1.0405289724648425e-05,
+      "loss": 0.5918,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9114690900097432,
+      "learning_rate": 1.0402659401094154e-05,
+      "loss": 0.525,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8534813523391873,
+      "learning_rate": 1.0400029049635942e-05,
+      "loss": 0.4923,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.877842763272256,
+      "learning_rate": 1.039739867045607e-05,
+      "loss": 0.5255,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0236133388921236,
+      "learning_rate": 1.039476826373683e-05,
+      "loss": 0.5916,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8590065023812046,
+      "learning_rate": 1.0392137829660494e-05,
+      "loss": 0.4965,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8639027295230077,
+      "learning_rate": 1.0389507368409356e-05,
+      "loss": 0.5561,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8836484623587807,
+      "learning_rate": 1.0386876880165701e-05,
+      "loss": 0.5385,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8632103809318213,
+      "learning_rate": 1.0384246365111823e-05,
+      "loss": 0.5109,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9279412725843376,
+      "learning_rate": 1.0381615823430012e-05,
+      "loss": 0.4993,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8684010140359392,
+      "learning_rate": 1.0378985255302565e-05,
+      "loss": 0.5207,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8880388535422927,
+      "learning_rate": 1.0376354660911772e-05,
+      "loss": 0.5221,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0301038902267186,
+      "learning_rate": 1.0373724040439936e-05,
+      "loss": 0.5964,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7284968100578956,
+      "learning_rate": 1.0371093394069359e-05,
+      "loss": 0.4624,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.859546061851474,
+      "learning_rate": 1.0368462721982336e-05,
+      "loss": 0.4492,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8540284446590735,
+      "learning_rate": 1.0365832024361173e-05,
+      "loss": 0.5058,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9311871787223283,
+      "learning_rate": 1.0363201301388177e-05,
+      "loss": 0.5237,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8783458652186021,
+      "learning_rate": 1.036057055324565e-05,
+      "loss": 0.5155,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9133086616533669,
+      "learning_rate": 1.0357939780115906e-05,
+      "loss": 0.5134,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.93527568205224,
+      "learning_rate": 1.0355308982181254e-05,
+      "loss": 0.5286,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9748174070640031,
+      "learning_rate": 1.0352678159624e-05,
+      "loss": 0.4804,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8304324868783131,
+      "learning_rate": 1.0350047312626465e-05,
+      "loss": 0.5307,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.737183332887448,
+      "learning_rate": 1.0347416441370963e-05,
+      "loss": 0.4458,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8193594305614337,
+      "learning_rate": 1.0344785546039808e-05,
+      "loss": 0.4981,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9325176197238431,
+      "learning_rate": 1.0342154626815321e-05,
+      "loss": 0.5829,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9374309225936001,
+      "learning_rate": 1.0339523683879824e-05,
+      "loss": 0.5537,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8479987941359138,
+      "learning_rate": 1.0336892717415635e-05,
+      "loss": 0.5356,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7980171217781268,
+      "learning_rate": 1.0334261727605076e-05,
+      "loss": 0.5199,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8531021371446437,
+      "learning_rate": 1.0331630714630481e-05,
+      "loss": 0.5068,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8625990920480998,
+      "learning_rate": 1.032899967867417e-05,
+      "loss": 0.4964,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9672841283305325,
+      "learning_rate": 1.0326368619918471e-05,
+      "loss": 0.5782,
+      "step": 6213
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0735602156729935,
+      "learning_rate": 1.032373753854572e-05,
+      "loss": 0.5248,
+      "step": 6214
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9113137478830905,
+      "learning_rate": 1.0321106434738242e-05,
+      "loss": 0.569,
+      "step": 6215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.811414298943247,
+      "learning_rate": 1.0318475308678374e-05,
+      "loss": 0.5266,
+      "step": 6216
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9225312761086015,
+      "learning_rate": 1.031584416054845e-05,
+      "loss": 0.5044,
+      "step": 6217
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8248517090127305,
+      "learning_rate": 1.0313212990530804e-05,
+      "loss": 0.4785,
+      "step": 6218
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9024948796769673,
+      "learning_rate": 1.0310581798807776e-05,
+      "loss": 0.5658,
+      "step": 6219
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9641619822749372,
+      "learning_rate": 1.0307950585561705e-05,
+      "loss": 0.5675,
+      "step": 6220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8721024784016015,
+      "learning_rate": 1.0305319350974932e-05,
+      "loss": 0.5326,
+      "step": 6221
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9505797840996292,
+      "learning_rate": 1.0302688095229798e-05,
+      "loss": 0.5678,
+      "step": 6222
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8774987959531362,
+      "learning_rate": 1.030005681850865e-05,
+      "loss": 0.5067,
+      "step": 6223
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9024372804219885,
+      "learning_rate": 1.0297425520993829e-05,
+      "loss": 0.5454,
+      "step": 6224
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9296782390888202,
+      "learning_rate": 1.0294794202867681e-05,
+      "loss": 0.5238,
+      "step": 6225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8898639473811754,
+      "learning_rate": 1.029216286431256e-05,
+      "loss": 0.5284,
+      "step": 6226
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9559196004726771,
+      "learning_rate": 1.028953150551081e-05,
+      "loss": 0.606,
+      "step": 6227
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9758923851832058,
+      "learning_rate": 1.0286900126644783e-05,
+      "loss": 0.6068,
+      "step": 6228
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8429290545153291,
+      "learning_rate": 1.0284268727896833e-05,
+      "loss": 0.5196,
+      "step": 6229
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9241137844044234,
+      "learning_rate": 1.028163730944931e-05,
+      "loss": 0.5021,
+      "step": 6230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8844502814224067,
+      "learning_rate": 1.0279005871484572e-05,
+      "loss": 0.5981,
+      "step": 6231
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7940233993428926,
+      "learning_rate": 1.0276374414184977e-05,
+      "loss": 0.4836,
+      "step": 6232
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8540121985206928,
+      "learning_rate": 1.0273742937732877e-05,
+      "loss": 0.5212,
+      "step": 6233
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8411679107015873,
+      "learning_rate": 1.0271111442310638e-05,
+      "loss": 0.4643,
+      "step": 6234
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.843190016884267,
+      "learning_rate": 1.0268479928100615e-05,
+      "loss": 0.4889,
+      "step": 6235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9298687033475067,
+      "learning_rate": 1.026584839528517e-05,
+      "loss": 0.4891,
+      "step": 6236
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8314982900576823,
+      "learning_rate": 1.0263216844046666e-05,
+      "loss": 0.5256,
+      "step": 6237
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9635576321688502,
+      "learning_rate": 1.026058527456747e-05,
+      "loss": 0.5927,
+      "step": 6238
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9399777954113789,
+      "learning_rate": 1.0257953687029945e-05,
+      "loss": 0.5366,
+      "step": 6239
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9062288398497633,
+      "learning_rate": 1.0255322081616456e-05,
+      "loss": 0.4792,
+      "step": 6240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9000858459834027,
+      "learning_rate": 1.025269045850938e-05,
+      "loss": 0.4799,
+      "step": 6241
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9559802820060481,
+      "learning_rate": 1.0250058817891074e-05,
+      "loss": 0.5548,
+      "step": 6242
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9062852807311451,
+      "learning_rate": 1.0247427159943912e-05,
+      "loss": 0.4895,
+      "step": 6243
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9464049844786903,
+      "learning_rate": 1.0244795484850272e-05,
+      "loss": 0.5534,
+      "step": 6244
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8847921545422844,
+      "learning_rate": 1.024216379279252e-05,
+      "loss": 0.5079,
+      "step": 6245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8978136301515469,
+      "learning_rate": 1.0239532083953032e-05,
+      "loss": 0.5608,
+      "step": 6246
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9463582730856492,
+      "learning_rate": 1.0236900358514181e-05,
+      "loss": 0.5277,
+      "step": 6247
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8067583675753313,
+      "learning_rate": 1.023426861665835e-05,
+      "loss": 0.4961,
+      "step": 6248
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8883255946762623,
+      "learning_rate": 1.0231636858567909e-05,
+      "loss": 0.5654,
+      "step": 6249
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9190681551860245,
+      "learning_rate": 1.022900508442524e-05,
+      "loss": 0.562,
+      "step": 6250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9494150603533084,
+      "learning_rate": 1.0226373294412718e-05,
+      "loss": 0.5676,
+      "step": 6251
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9061517922527588,
+      "learning_rate": 1.0223741488712732e-05,
+      "loss": 0.5428,
+      "step": 6252
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9275782262305359,
+      "learning_rate": 1.0221109667507656e-05,
+      "loss": 0.5184,
+      "step": 6253
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9189579629915223,
+      "learning_rate": 1.0218477830979878e-05,
+      "loss": 0.5439,
+      "step": 6254
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9182833448414004,
+      "learning_rate": 1.0215845979311783e-05,
+      "loss": 0.5327,
+      "step": 6255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8902344636216859,
+      "learning_rate": 1.0213214112685747e-05,
+      "loss": 0.4828,
+      "step": 6256
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.897277648623057,
+      "learning_rate": 1.0210582231284165e-05,
+      "loss": 0.5015,
+      "step": 6257
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9182502828960334,
+      "learning_rate": 1.0207950335289423e-05,
+      "loss": 0.552,
+      "step": 6258
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.885910996981317,
+      "learning_rate": 1.0205318424883906e-05,
+      "loss": 0.4874,
+      "step": 6259
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.922634853593885,
+      "learning_rate": 1.0202686500250003e-05,
+      "loss": 0.529,
+      "step": 6260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8720759553847209,
+      "learning_rate": 1.0200054561570108e-05,
+      "loss": 0.5301,
+      "step": 6261
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9239527432654803,
+      "learning_rate": 1.0197422609026606e-05,
+      "loss": 0.5539,
+      "step": 6262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8890671690015547,
+      "learning_rate": 1.0194790642801893e-05,
+      "loss": 0.5476,
+      "step": 6263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8569505481042301,
+      "learning_rate": 1.0192158663078362e-05,
+      "loss": 0.5216,
+      "step": 6264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8738179518705772,
+      "learning_rate": 1.0189526670038407e-05,
+      "loss": 0.5254,
+      "step": 6265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8517879706535724,
+      "learning_rate": 1.0186894663864421e-05,
+      "loss": 0.5613,
+      "step": 6266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9467237412597023,
+      "learning_rate": 1.01842626447388e-05,
+      "loss": 0.565,
+      "step": 6267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9372933446510155,
+      "learning_rate": 1.0181630612843943e-05,
+      "loss": 0.5869,
+      "step": 6268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8379813367737269,
+      "learning_rate": 1.0178998568362243e-05,
+      "loss": 0.4953,
+      "step": 6269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8807616083485671,
+      "learning_rate": 1.0176366511476102e-05,
+      "loss": 0.5605,
+      "step": 6270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.909667415705031,
+      "learning_rate": 1.0173734442367919e-05,
+      "loss": 0.5494,
+      "step": 6271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9180665855259821,
+      "learning_rate": 1.0171102361220093e-05,
+      "loss": 0.5776,
+      "step": 6272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8866368103793839,
+      "learning_rate": 1.0168470268215025e-05,
+      "loss": 0.5162,
+      "step": 6273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9331207432565604,
+      "learning_rate": 1.0165838163535115e-05,
+      "loss": 0.5799,
+      "step": 6274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9299756848302255,
+      "learning_rate": 1.0163206047362773e-05,
+      "loss": 0.5631,
+      "step": 6275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9153501532529058,
+      "learning_rate": 1.016057391988039e-05,
+      "loss": 0.5746,
+      "step": 6276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9106064921432052,
+      "learning_rate": 1.015794178127038e-05,
+      "loss": 0.4931,
+      "step": 6277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8583263808738999,
+      "learning_rate": 1.0155309631715145e-05,
+      "loss": 0.5349,
+      "step": 6278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9011920327335722,
+      "learning_rate": 1.015267747139709e-05,
+      "loss": 0.5075,
+      "step": 6279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8793178650201761,
+      "learning_rate": 1.0150045300498618e-05,
+      "loss": 0.5156,
+      "step": 6280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8115799752291708,
+      "learning_rate": 1.0147413119202145e-05,
+      "loss": 0.4846,
+      "step": 6281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9177489139524344,
+      "learning_rate": 1.0144780927690072e-05,
+      "loss": 0.5457,
+      "step": 6282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8929122588141614,
+      "learning_rate": 1.0142148726144807e-05,
+      "loss": 0.5309,
+      "step": 6283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.804018811198706,
+      "learning_rate": 1.0139516514748767e-05,
+      "loss": 0.4608,
+      "step": 6284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9347609164966086,
+      "learning_rate": 1.013688429368435e-05,
+      "loss": 0.4832,
+      "step": 6285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9217079376541866,
+      "learning_rate": 1.0134252063133976e-05,
+      "loss": 0.5749,
+      "step": 6286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8506565309327903,
+      "learning_rate": 1.0131619823280053e-05,
+      "loss": 0.5394,
+      "step": 6287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8415600271925963,
+      "learning_rate": 1.0128987574304991e-05,
+      "loss": 0.5208,
+      "step": 6288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8758726806842629,
+      "learning_rate": 1.0126355316391206e-05,
+      "loss": 0.5892,
+      "step": 6289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9039543673993381,
+      "learning_rate": 1.012372304972111e-05,
+      "loss": 0.572,
+      "step": 6290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8837831665094975,
+      "learning_rate": 1.0121090774477116e-05,
+      "loss": 0.5164,
+      "step": 6291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7969372582312041,
+      "learning_rate": 1.0118458490841639e-05,
+      "loss": 0.4687,
+      "step": 6292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9144472654901419,
+      "learning_rate": 1.0115826198997094e-05,
+      "loss": 0.4895,
+      "step": 6293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8546462905815698,
+      "learning_rate": 1.0113193899125895e-05,
+      "loss": 0.4768,
+      "step": 6294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8571567901864389,
+      "learning_rate": 1.0110561591410456e-05,
+      "loss": 0.5441,
+      "step": 6295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8062069436489295,
+      "learning_rate": 1.0107929276033204e-05,
+      "loss": 0.4905,
+      "step": 6296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8690171443782525,
+      "learning_rate": 1.0105296953176544e-05,
+      "loss": 0.4848,
+      "step": 6297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8929739364675837,
+      "learning_rate": 1.01026646230229e-05,
+      "loss": 0.5132,
+      "step": 6298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8985021936347624,
+      "learning_rate": 1.010003228575469e-05,
+      "loss": 0.5021,
+      "step": 6299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8734985699165543,
+      "learning_rate": 1.009739994155433e-05,
+      "loss": 0.4902,
+      "step": 6300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9145764779595406,
+      "learning_rate": 1.0094767590604238e-05,
+      "loss": 0.5366,
+      "step": 6301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.938111655768843,
+      "learning_rate": 1.009213523308684e-05,
+      "loss": 0.5352,
+      "step": 6302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9187062301830011,
+      "learning_rate": 1.0089502869184549e-05,
+      "loss": 0.5064,
+      "step": 6303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8909106006968533,
+      "learning_rate": 1.0086870499079791e-05,
+      "loss": 0.526,
+      "step": 6304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9891112088673406,
+      "learning_rate": 1.0084238122954984e-05,
+      "loss": 0.4361,
+      "step": 6305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9986087397097535,
+      "learning_rate": 1.0081605740992548e-05,
+      "loss": 0.5481,
+      "step": 6306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9478586520651313,
+      "learning_rate": 1.0078973353374908e-05,
+      "loss": 0.5567,
+      "step": 6307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.889552514423573,
+      "learning_rate": 1.0076340960284483e-05,
+      "loss": 0.5208,
+      "step": 6308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8643499230620482,
+      "learning_rate": 1.0073708561903702e-05,
+      "loss": 0.5414,
+      "step": 6309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.976399247172449,
+      "learning_rate": 1.0071076158414977e-05,
+      "loss": 0.6492,
+      "step": 6310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9571144105482331,
+      "learning_rate": 1.006844375000074e-05,
+      "loss": 0.6006,
+      "step": 6311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8677112699372111,
+      "learning_rate": 1.0065811336843412e-05,
+      "loss": 0.5269,
+      "step": 6312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8705189403238094,
+      "learning_rate": 1.0063178919125416e-05,
+      "loss": 0.5816,
+      "step": 6313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.936745039273025,
+      "learning_rate": 1.0060546497029178e-05,
+      "loss": 0.519,
+      "step": 6314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2267196954416242,
+      "learning_rate": 1.0057914070737123e-05,
+      "loss": 0.6564,
+      "step": 6315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8408749871775812,
+      "learning_rate": 1.0055281640431669e-05,
+      "loss": 0.5499,
+      "step": 6316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9485173287734637,
+      "learning_rate": 1.005264920629525e-05,
+      "loss": 0.5244,
+      "step": 6317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8744806225444277,
+      "learning_rate": 1.0050016768510288e-05,
+      "loss": 0.4819,
+      "step": 6318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9837730314596066,
+      "learning_rate": 1.0047384327259207e-05,
+      "loss": 0.5324,
+      "step": 6319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9897169442511425,
+      "learning_rate": 1.0044751882724436e-05,
+      "loss": 0.5116,
+      "step": 6320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8355025962859695,
+      "learning_rate": 1.0042119435088397e-05,
+      "loss": 0.5085,
+      "step": 6321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8822374522103449,
+      "learning_rate": 1.003948698453352e-05,
+      "loss": 0.5093,
+      "step": 6322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8868032716019236,
+      "learning_rate": 1.0036854531242228e-05,
+      "loss": 0.5216,
+      "step": 6323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0298087875025894,
+      "learning_rate": 1.0034222075396954e-05,
+      "loss": 0.5794,
+      "step": 6324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.859260370043238,
+      "learning_rate": 1.0031589617180115e-05,
+      "loss": 0.4686,
+      "step": 6325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8158325936763585,
+      "learning_rate": 1.0028957156774146e-05,
+      "loss": 0.4922,
+      "step": 6326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9242530488088277,
+      "learning_rate": 1.0026324694361474e-05,
+      "loss": 0.5227,
+      "step": 6327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0280646267210969,
+      "learning_rate": 1.002369223012452e-05,
+      "loss": 0.5422,
+      "step": 6328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.848472981714665,
+      "learning_rate": 1.0021059764245718e-05,
+      "loss": 0.514,
+      "step": 6329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8569811344917727,
+      "learning_rate": 1.0018427296907494e-05,
+      "loss": 0.5174,
+      "step": 6330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.042013096707966,
+      "learning_rate": 1.001579482829227e-05,
+      "loss": 0.5577,
+      "step": 6331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0091968409285097,
+      "learning_rate": 1.0013162358582483e-05,
+      "loss": 0.5897,
+      "step": 6332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8723519274260123,
+      "learning_rate": 1.0010529887960554e-05,
+      "loss": 0.4901,
+      "step": 6333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9537740574635022,
+      "learning_rate": 1.0007897416608914e-05,
+      "loss": 0.5724,
+      "step": 6334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9789588220022455,
+      "learning_rate": 1.0005264944709989e-05,
+      "loss": 0.5557,
+      "step": 6335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9737723535686139,
+      "learning_rate": 1.000263247244621e-05,
+      "loss": 0.5868,
+      "step": 6336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9126233242598474,
+      "learning_rate": 1e-05,
+      "loss": 0.5436,
+      "step": 6337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9065250250113098,
+      "learning_rate": 9.997367527553795e-06,
+      "loss": 0.5473,
+      "step": 6338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7824478859233489,
+      "learning_rate": 9.994735055290011e-06,
+      "loss": 0.4108,
+      "step": 6339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9394386325341539,
+      "learning_rate": 9.992102583391089e-06,
+      "loss": 0.5323,
+      "step": 6340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9901071026908101,
+      "learning_rate": 9.98947011203945e-06,
+      "loss": 0.5428,
+      "step": 6341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8936489106322111,
+      "learning_rate": 9.986837641417519e-06,
+      "loss": 0.5059,
+      "step": 6342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8990739204643596,
+      "learning_rate": 9.984205171707731e-06,
+      "loss": 0.5022,
+      "step": 6343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9451398467900175,
+      "learning_rate": 9.981572703092513e-06,
+      "loss": 0.5639,
+      "step": 6344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9201894358093854,
+      "learning_rate": 9.978940235754283e-06,
+      "loss": 0.5315,
+      "step": 6345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9482135020604635,
+      "learning_rate": 9.976307769875483e-06,
+      "loss": 0.5399,
+      "step": 6346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9473339865184615,
+      "learning_rate": 9.973675305638531e-06,
+      "loss": 0.5061,
+      "step": 6347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9599685256552569,
+      "learning_rate": 9.971042843225856e-06,
+      "loss": 0.5604,
+      "step": 6348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8559541252886981,
+      "learning_rate": 9.968410382819888e-06,
+      "loss": 0.5574,
+      "step": 6349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8868070586988529,
+      "learning_rate": 9.965777924603053e-06,
+      "loss": 0.5052,
+      "step": 6350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9358125147616042,
+      "learning_rate": 9.963145468757773e-06,
+      "loss": 0.5475,
+      "step": 6351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9305033132415802,
+      "learning_rate": 9.960513015466484e-06,
+      "loss": 0.5354,
+      "step": 6352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9508925128715102,
+      "learning_rate": 9.957880564911608e-06,
+      "loss": 0.5395,
+      "step": 6353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8046306445228953,
+      "learning_rate": 9.955248117275566e-06,
+      "loss": 0.4776,
+      "step": 6354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8810642297736707,
+      "learning_rate": 9.952615672740795e-06,
+      "loss": 0.465,
+      "step": 6355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8271357839574274,
+      "learning_rate": 9.949983231489717e-06,
+      "loss": 0.5456,
+      "step": 6356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9932580719434206,
+      "learning_rate": 9.947350793704751e-06,
+      "loss": 0.5465,
+      "step": 6357
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8911737003658028,
+      "learning_rate": 9.944718359568333e-06,
+      "loss": 0.4734,
+      "step": 6358
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9414596037214545,
+      "learning_rate": 9.942085929262884e-06,
+      "loss": 0.5159,
+      "step": 6359
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9720304123433685,
+      "learning_rate": 9.939453502970824e-06,
+      "loss": 0.5785,
+      "step": 6360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9146124769744249,
+      "learning_rate": 9.936821080874587e-06,
+      "loss": 0.5701,
+      "step": 6361
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9294401348339756,
+      "learning_rate": 9.934188663156592e-06,
+      "loss": 0.5041,
+      "step": 6362
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.896786916401712,
+      "learning_rate": 9.931556249999262e-06,
+      "loss": 0.4878,
+      "step": 6363
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9214512261006776,
+      "learning_rate": 9.928923841585025e-06,
+      "loss": 0.5577,
+      "step": 6364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.952653505975468,
+      "learning_rate": 9.926291438096305e-06,
+      "loss": 0.5745,
+      "step": 6365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.894555231533999,
+      "learning_rate": 9.923659039715517e-06,
+      "loss": 0.538,
+      "step": 6366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9779553837225547,
+      "learning_rate": 9.921026646625094e-06,
+      "loss": 0.5475,
+      "step": 6367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9825468080035322,
+      "learning_rate": 9.918394259007458e-06,
+      "loss": 0.5964,
+      "step": 6368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9816450220972703,
+      "learning_rate": 9.91576187704502e-06,
+      "loss": 0.5483,
+      "step": 6369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9936843969441377,
+      "learning_rate": 9.913129500920214e-06,
+      "loss": 0.5398,
+      "step": 6370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8657007233111466,
+      "learning_rate": 9.910497130815454e-06,
+      "loss": 0.4922,
+      "step": 6371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8313299803368648,
+      "learning_rate": 9.907864766913162e-06,
+      "loss": 0.5318,
+      "step": 6372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9539489317761674,
+      "learning_rate": 9.905232409395764e-06,
+      "loss": 0.5689,
+      "step": 6373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8829599504299749,
+      "learning_rate": 9.902600058445676e-06,
+      "loss": 0.5331,
+      "step": 6374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0323815487828643,
+      "learning_rate": 9.899967714245313e-06,
+      "loss": 0.6064,
+      "step": 6375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8874248857833198,
+      "learning_rate": 9.897335376977104e-06,
+      "loss": 0.5538,
+      "step": 6376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9590342091526441,
+      "learning_rate": 9.894703046823461e-06,
+      "loss": 0.571,
+      "step": 6377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9608727054645129,
+      "learning_rate": 9.8920707239668e-06,
+      "loss": 0.5436,
+      "step": 6378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9303455966981541,
+      "learning_rate": 9.889438408589545e-06,
+      "loss": 0.4842,
+      "step": 6379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9910314393804825,
+      "learning_rate": 9.88680610087411e-06,
+      "loss": 0.528,
+      "step": 6380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9227435341657937,
+      "learning_rate": 9.884173801002909e-06,
+      "loss": 0.5307,
+      "step": 6381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8368401603641216,
+      "learning_rate": 9.881541509158366e-06,
+      "loss": 0.4937,
+      "step": 6382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8543858991076968,
+      "learning_rate": 9.878909225522889e-06,
+      "loss": 0.4891,
+      "step": 6383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9324945143635585,
+      "learning_rate": 9.876276950278893e-06,
+      "loss": 0.5354,
+      "step": 6384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8203979580170194,
+      "learning_rate": 9.873644683608798e-06,
+      "loss": 0.4777,
+      "step": 6385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8715151945909194,
+      "learning_rate": 9.87101242569501e-06,
+      "loss": 0.5123,
+      "step": 6386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.954708756603573,
+      "learning_rate": 9.86838017671995e-06,
+      "loss": 0.5242,
+      "step": 6387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9317365832888832,
+      "learning_rate": 9.865747936866027e-06,
+      "loss": 0.5395,
+      "step": 6388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8770081131433685,
+      "learning_rate": 9.863115706315652e-06,
+      "loss": 0.5174,
+      "step": 6389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8598260750153,
+      "learning_rate": 9.860483485251238e-06,
+      "loss": 0.4591,
+      "step": 6390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9755980531605132,
+      "learning_rate": 9.857851273855195e-06,
+      "loss": 0.5801,
+      "step": 6391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0087368311921079,
+      "learning_rate": 9.855219072309931e-06,
+      "loss": 0.498,
+      "step": 6392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9230860988633534,
+      "learning_rate": 9.852586880797857e-06,
+      "loss": 0.5284,
+      "step": 6393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9058880053248296,
+      "learning_rate": 9.849954699501383e-06,
+      "loss": 0.5242,
+      "step": 6394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9677734450221844,
+      "learning_rate": 9.847322528602913e-06,
+      "loss": 0.5892,
+      "step": 6395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9367373557695239,
+      "learning_rate": 9.844690368284857e-06,
+      "loss": 0.56,
+      "step": 6396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.878603950726151,
+      "learning_rate": 9.842058218729623e-06,
+      "loss": 0.4691,
+      "step": 6397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8616635511939935,
+      "learning_rate": 9.839426080119612e-06,
+      "loss": 0.5389,
+      "step": 6398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9986338396517531,
+      "learning_rate": 9.836793952637232e-06,
+      "loss": 0.552,
+      "step": 6399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9511148684059743,
+      "learning_rate": 9.834161836464888e-06,
+      "loss": 0.5818,
+      "step": 6400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0105896126209348,
+      "learning_rate": 9.831529731784975e-06,
+      "loss": 0.5089,
+      "step": 6401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8717280136431198,
+      "learning_rate": 9.828897638779909e-06,
+      "loss": 0.5305,
+      "step": 6402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9162669287847024,
+      "learning_rate": 9.826265557632083e-06,
+      "loss": 0.5128,
+      "step": 6403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8920982133737825,
+      "learning_rate": 9.823633488523898e-06,
+      "loss": 0.5622,
+      "step": 6404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9054418294786118,
+      "learning_rate": 9.821001431637759e-06,
+      "loss": 0.5298,
+      "step": 6405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8805665336617836,
+      "learning_rate": 9.81836938715606e-06,
+      "loss": 0.5627,
+      "step": 6406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9678033064998812,
+      "learning_rate": 9.815737355261201e-06,
+      "loss": 0.5519,
+      "step": 6407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8733129307408266,
+      "learning_rate": 9.813105336135582e-06,
+      "loss": 0.5598,
+      "step": 6408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9086165997165361,
+      "learning_rate": 9.810473329961595e-06,
+      "loss": 0.524,
+      "step": 6409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9492530342714538,
+      "learning_rate": 9.807841336921639e-06,
+      "loss": 0.4998,
+      "step": 6410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8393818681658488,
+      "learning_rate": 9.80520935719811e-06,
+      "loss": 0.544,
+      "step": 6411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9381276538883282,
+      "learning_rate": 9.802577390973397e-06,
+      "loss": 0.5205,
+      "step": 6412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8978597929624021,
+      "learning_rate": 9.799945438429895e-06,
+      "loss": 0.5191,
+      "step": 6413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9130042053056334,
+      "learning_rate": 9.79731349975e-06,
+      "loss": 0.52,
+      "step": 6414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.909046664883525,
+      "learning_rate": 9.794681575116097e-06,
+      "loss": 0.5285,
+      "step": 6415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9534951232598357,
+      "learning_rate": 9.792049664710579e-06,
+      "loss": 0.5247,
+      "step": 6416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.93802065691835,
+      "learning_rate": 9.789417768715837e-06,
+      "loss": 0.4866,
+      "step": 6417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7893135352036715,
+      "learning_rate": 9.786785887314255e-06,
+      "loss": 0.412,
+      "step": 6418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9111844420461667,
+      "learning_rate": 9.784154020688222e-06,
+      "loss": 0.5623,
+      "step": 6419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9935656482433806,
+      "learning_rate": 9.781522169020125e-06,
+      "loss": 0.5779,
+      "step": 6420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8351481116157773,
+      "learning_rate": 9.778890332492346e-06,
+      "loss": 0.4643,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9588974823514563,
+      "learning_rate": 9.776258511287271e-06,
+      "loss": 0.5176,
+      "step": 6422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.828264983941755,
+      "learning_rate": 9.773626705587283e-06,
+      "loss": 0.4941,
+      "step": 6423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9097675795507923,
+      "learning_rate": 9.770994915574766e-06,
+      "loss": 0.5653,
+      "step": 6424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8263116395093373,
+      "learning_rate": 9.768363141432095e-06,
+      "loss": 0.5402,
+      "step": 6425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8329323971338786,
+      "learning_rate": 9.765731383341654e-06,
+      "loss": 0.4887,
+      "step": 6426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9319485658181705,
+      "learning_rate": 9.76309964148582e-06,
+      "loss": 0.5301,
+      "step": 6427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9109510386316052,
+      "learning_rate": 9.760467916046971e-06,
+      "loss": 0.5702,
+      "step": 6428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9192063800704567,
+      "learning_rate": 9.757836207207483e-06,
+      "loss": 0.5517,
+      "step": 6429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8938122304610931,
+      "learning_rate": 9.755204515149731e-06,
+      "loss": 0.5129,
+      "step": 6430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8885073683990292,
+      "learning_rate": 9.75257284005609e-06,
+      "loss": 0.4888,
+      "step": 6431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9694255751986594,
+      "learning_rate": 9.74994118210893e-06,
+      "loss": 0.5145,
+      "step": 6432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9905287843588529,
+      "learning_rate": 9.747309541490627e-06,
+      "loss": 0.5592,
+      "step": 6433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9468510667397407,
+      "learning_rate": 9.744677918383546e-06,
+      "loss": 0.4493,
+      "step": 6434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8159532262872364,
+      "learning_rate": 9.742046312970058e-06,
+      "loss": 0.4615,
+      "step": 6435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9604668106872334,
+      "learning_rate": 9.739414725432535e-06,
+      "loss": 0.5214,
+      "step": 6436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9556439331531814,
+      "learning_rate": 9.736783155953338e-06,
+      "loss": 0.5053,
+      "step": 6437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.916486932919353,
+      "learning_rate": 9.734151604714834e-06,
+      "loss": 0.5413,
+      "step": 6438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8681676499631316,
+      "learning_rate": 9.73152007189939e-06,
+      "loss": 0.4963,
+      "step": 6439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8529159346600007,
+      "learning_rate": 9.728888557689364e-06,
+      "loss": 0.5299,
+      "step": 6440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8944587393736042,
+      "learning_rate": 9.726257062267124e-06,
+      "loss": 0.523,
+      "step": 6441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8535702354962332,
+      "learning_rate": 9.723625585815028e-06,
+      "loss": 0.5471,
+      "step": 6442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9894346512705576,
+      "learning_rate": 9.720994128515428e-06,
+      "loss": 0.5799,
+      "step": 6443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8912292132663773,
+      "learning_rate": 9.718362690550693e-06,
+      "loss": 0.5419,
+      "step": 6444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8431590559324045,
+      "learning_rate": 9.715731272103172e-06,
+      "loss": 0.5395,
+      "step": 6445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8296842798941493,
+      "learning_rate": 9.713099873355219e-06,
+      "loss": 0.5027,
+      "step": 6446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.049682332775139,
+      "learning_rate": 9.710468494489194e-06,
+      "loss": 0.5357,
+      "step": 6447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9346713890089724,
+      "learning_rate": 9.707837135687444e-06,
+      "loss": 0.5126,
+      "step": 6448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9759963987028532,
+      "learning_rate": 9.705205797132319e-06,
+      "loss": 0.5742,
+      "step": 6449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.878814498064152,
+      "learning_rate": 9.702574479006174e-06,
+      "loss": 0.5266,
+      "step": 6450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8991362672949392,
+      "learning_rate": 9.699943181491355e-06,
+      "loss": 0.4653,
+      "step": 6451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8413370445817436,
+      "learning_rate": 9.697311904770202e-06,
+      "loss": 0.469,
+      "step": 6452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8689595579603493,
+      "learning_rate": 9.69468064902507e-06,
+      "loss": 0.5227,
+      "step": 6453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9697152793274977,
+      "learning_rate": 9.692049414438298e-06,
+      "loss": 0.5844,
+      "step": 6454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8949486024463249,
+      "learning_rate": 9.689418201192226e-06,
+      "loss": 0.5068,
+      "step": 6455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9773141105296431,
+      "learning_rate": 9.6867870094692e-06,
+      "loss": 0.5538,
+      "step": 6456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.857328825009141,
+      "learning_rate": 9.684155839451555e-06,
+      "loss": 0.4758,
+      "step": 6457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9429415025858808,
+      "learning_rate": 9.681524691321628e-06,
+      "loss": 0.6401,
+      "step": 6458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0145411858316762,
+      "learning_rate": 9.678893565261761e-06,
+      "loss": 0.5969,
+      "step": 6459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9661312013835944,
+      "learning_rate": 9.676262461454285e-06,
+      "loss": 0.5137,
+      "step": 6460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9337415348932296,
+      "learning_rate": 9.67363138008153e-06,
+      "loss": 0.5644,
+      "step": 6461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8901844897600891,
+      "learning_rate": 9.671000321325832e-06,
+      "loss": 0.6057,
+      "step": 6462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9016225379646999,
+      "learning_rate": 9.668369285369524e-06,
+      "loss": 0.498,
+      "step": 6463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9297387666303769,
+      "learning_rate": 9.665738272394924e-06,
+      "loss": 0.5294,
+      "step": 6464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9734669398659537,
+      "learning_rate": 9.66310728258437e-06,
+      "loss": 0.5146,
+      "step": 6465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9545143524091929,
+      "learning_rate": 9.660476316120181e-06,
+      "loss": 0.5714,
+      "step": 6466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8922757284332623,
+      "learning_rate": 9.65784537318468e-06,
+      "loss": 0.542,
+      "step": 6467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8573370579926729,
+      "learning_rate": 9.655214453960195e-06,
+      "loss": 0.4861,
+      "step": 6468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8977961285324865,
+      "learning_rate": 9.652583558629042e-06,
+      "loss": 0.5144,
+      "step": 6469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9579689428849033,
+      "learning_rate": 9.649952687373535e-06,
+      "loss": 0.5472,
+      "step": 6470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8805879235281294,
+      "learning_rate": 9.647321840376001e-06,
+      "loss": 0.5174,
+      "step": 6471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.051855667302005,
+      "learning_rate": 9.644691017818752e-06,
+      "loss": 0.5926,
+      "step": 6472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8989815009031404,
+      "learning_rate": 9.642060219884096e-06,
+      "loss": 0.4875,
+      "step": 6473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9211094147866692,
+      "learning_rate": 9.639429446754352e-06,
+      "loss": 0.5151,
+      "step": 6474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9399000526890481,
+      "learning_rate": 9.636798698611828e-06,
+      "loss": 0.554,
+      "step": 6475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8830150022783372,
+      "learning_rate": 9.634167975638828e-06,
+      "loss": 0.5276,
+      "step": 6476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.858616695781131,
+      "learning_rate": 9.631537278017667e-06,
+      "loss": 0.4538,
+      "step": 6477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9342039254589489,
+      "learning_rate": 9.628906605930647e-06,
+      "loss": 0.5417,
+      "step": 6478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8611122539892918,
+      "learning_rate": 9.626275959560064e-06,
+      "loss": 0.536,
+      "step": 6479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.857609366638501,
+      "learning_rate": 9.62364533908823e-06,
+      "loss": 0.489,
+      "step": 6480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.825074690220345,
+      "learning_rate": 9.621014744697442e-06,
+      "loss": 0.4701,
+      "step": 6481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.934085548351372,
+      "learning_rate": 9.61838417656999e-06,
+      "loss": 0.4784,
+      "step": 6482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9527486975769726,
+      "learning_rate": 9.615753634888179e-06,
+      "loss": 0.5408,
+      "step": 6483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.84672859944973,
+      "learning_rate": 9.613123119834304e-06,
+      "loss": 0.5066,
+      "step": 6484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8997458292305717,
+      "learning_rate": 9.610492631590646e-06,
+      "loss": 0.5416,
+      "step": 6485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9626727248227898,
+      "learning_rate": 9.60786217033951e-06,
+      "loss": 0.5657,
+      "step": 6486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.764459498528756,
+      "learning_rate": 9.605231736263176e-06,
+      "loss": 0.4537,
+      "step": 6487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9296132740721262,
+      "learning_rate": 9.602601329543928e-06,
+      "loss": 0.4775,
+      "step": 6488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8278080916393366,
+      "learning_rate": 9.599970950364061e-06,
+      "loss": 0.4814,
+      "step": 6489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9615444489092069,
+      "learning_rate": 9.597340598905851e-06,
+      "loss": 0.591,
+      "step": 6490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9166588125726522,
+      "learning_rate": 9.594710275351577e-06,
+      "loss": 0.5656,
+      "step": 6491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9398183643638113,
+      "learning_rate": 9.592079979883526e-06,
+      "loss": 0.5495,
+      "step": 6492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9101662724536649,
+      "learning_rate": 9.58944971268397e-06,
+      "loss": 0.575,
+      "step": 6493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9027671688552364,
+      "learning_rate": 9.586819473935181e-06,
+      "loss": 0.5345,
+      "step": 6494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9741851024488997,
+      "learning_rate": 9.58418926381944e-06,
+      "loss": 0.6057,
+      "step": 6495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8528086376805016,
+      "learning_rate": 9.581559082519015e-06,
+      "loss": 0.4635,
+      "step": 6496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9158825951760049,
+      "learning_rate": 9.578928930216167e-06,
+      "loss": 0.5513,
+      "step": 6497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9498083322382603,
+      "learning_rate": 9.576298807093177e-06,
+      "loss": 0.5602,
+      "step": 6498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8378768819215866,
+      "learning_rate": 9.573668713332305e-06,
+      "loss": 0.4934,
+      "step": 6499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.849447929846769,
+      "learning_rate": 9.571038649115807e-06,
+      "loss": 0.5089,
+      "step": 6500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9094774639747515,
+      "learning_rate": 9.568408614625956e-06,
+      "loss": 0.5415,
+      "step": 6501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8683994871193035,
+      "learning_rate": 9.565778610045003e-06,
+      "loss": 0.4996,
+      "step": 6502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8416512392944917,
+      "learning_rate": 9.563148635555205e-06,
+      "loss": 0.4769,
+      "step": 6503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8417091367157625,
+      "learning_rate": 9.560518691338822e-06,
+      "loss": 0.5091,
+      "step": 6504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9307177969399728,
+      "learning_rate": 9.557888777578105e-06,
+      "loss": 0.5736,
+      "step": 6505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9148668617918116,
+      "learning_rate": 9.555258894455298e-06,
+      "loss": 0.4671,
+      "step": 6506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.933610510661884,
+      "learning_rate": 9.55262904215266e-06,
+      "loss": 0.5215,
+      "step": 6507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8493681051755907,
+      "learning_rate": 9.549999220852432e-06,
+      "loss": 0.4692,
+      "step": 6508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9008032770859512,
+      "learning_rate": 9.547369430736857e-06,
+      "loss": 0.5602,
+      "step": 6509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8313031359121507,
+      "learning_rate": 9.54473967198818e-06,
+      "loss": 0.4551,
+      "step": 6510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0024548477030657,
+      "learning_rate": 9.542109944788643e-06,
+      "loss": 0.4755,
+      "step": 6511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9424448090376205,
+      "learning_rate": 9.539480249320473e-06,
+      "loss": 0.5853,
+      "step": 6512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8858742800185914,
+      "learning_rate": 9.53685058576592e-06,
+      "loss": 0.5339,
+      "step": 6513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9352935076731403,
+      "learning_rate": 9.53422095430721e-06,
+      "loss": 0.5547,
+      "step": 6514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0227461291170266,
+      "learning_rate": 9.53159135512657e-06,
+      "loss": 0.5612,
+      "step": 6515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9646149787645724,
+      "learning_rate": 9.528961788406237e-06,
+      "loss": 0.5668,
+      "step": 6516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9089296205705552,
+      "learning_rate": 9.526332254328437e-06,
+      "loss": 0.5376,
+      "step": 6517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8838124970501346,
+      "learning_rate": 9.523702753075386e-06,
+      "loss": 0.5065,
+      "step": 6518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9206866517683869,
+      "learning_rate": 9.521073284829315e-06,
+      "loss": 0.4834,
+      "step": 6519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.912784923782488,
+      "learning_rate": 9.518443849772441e-06,
+      "loss": 0.512,
+      "step": 6520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9058344101233732,
+      "learning_rate": 9.515814448086978e-06,
+      "loss": 0.5461,
+      "step": 6521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8793164453057128,
+      "learning_rate": 9.513185079955148e-06,
+      "loss": 0.4845,
+      "step": 6522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8984635639989134,
+      "learning_rate": 9.51055574555916e-06,
+      "loss": 0.5754,
+      "step": 6523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9293547040387503,
+      "learning_rate": 9.50792644508122e-06,
+      "loss": 0.5819,
+      "step": 6524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.784032342081321,
+      "learning_rate": 9.505297178703546e-06,
+      "loss": 0.5305,
+      "step": 6525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.883478994718842,
+      "learning_rate": 9.502667946608332e-06,
+      "loss": 0.556,
+      "step": 6526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8416266967465563,
+      "learning_rate": 9.500038748977794e-06,
+      "loss": 0.5466,
+      "step": 6527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0312877433341645,
+      "learning_rate": 9.497409585994128e-06,
+      "loss": 0.5922,
+      "step": 6528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9451565220427494,
+      "learning_rate": 9.494780457839527e-06,
+      "loss": 0.4781,
+      "step": 6529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9527771032047712,
+      "learning_rate": 9.492151364696196e-06,
+      "loss": 0.5277,
+      "step": 6530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8612304851836948,
+      "learning_rate": 9.489522306746327e-06,
+      "loss": 0.487,
+      "step": 6531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8498362511354313,
+      "learning_rate": 9.486893284172103e-06,
+      "loss": 0.5537,
+      "step": 6532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9141501671369846,
+      "learning_rate": 9.484264297155724e-06,
+      "loss": 0.521,
+      "step": 6533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8471454028075913,
+      "learning_rate": 9.481635345879373e-06,
+      "loss": 0.5362,
+      "step": 6534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8502542534921609,
+      "learning_rate": 9.479006430525227e-06,
+      "loss": 0.5177,
+      "step": 6535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8558504290560964,
+      "learning_rate": 9.476377551275478e-06,
+      "loss": 0.5179,
+      "step": 6536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9149921933622068,
+      "learning_rate": 9.4737487083123e-06,
+      "loss": 0.4804,
+      "step": 6537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8624426856664629,
+      "learning_rate": 9.471119901817866e-06,
+      "loss": 0.5192,
+      "step": 6538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9475697368614927,
+      "learning_rate": 9.468491131974358e-06,
+      "loss": 0.5907,
+      "step": 6539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8597657353172048,
+      "learning_rate": 9.465862398963943e-06,
+      "loss": 0.4512,
+      "step": 6540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.791808646314481,
+      "learning_rate": 9.463233702968784e-06,
+      "loss": 0.4951,
+      "step": 6541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9190383831550103,
+      "learning_rate": 9.46060504417106e-06,
+      "loss": 0.5751,
+      "step": 6542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9219461465709242,
+      "learning_rate": 9.457976422752925e-06,
+      "loss": 0.5326,
+      "step": 6543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8887657703238295,
+      "learning_rate": 9.455347838896541e-06,
+      "loss": 0.4661,
+      "step": 6544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8733624243772576,
+      "learning_rate": 9.452719292784074e-06,
+      "loss": 0.5464,
+      "step": 6545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9112115008584419,
+      "learning_rate": 9.450090784597673e-06,
+      "loss": 0.5647,
+      "step": 6546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7311921806377226,
+      "learning_rate": 9.44746231451949e-06,
+      "loss": 0.5059,
+      "step": 6547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8463909442796727,
+      "learning_rate": 9.444833882731681e-06,
+      "loss": 0.5476,
+      "step": 6548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9461525435875031,
+      "learning_rate": 9.442205489416392e-06,
+      "loss": 0.5637,
+      "step": 6549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8698853022532326,
+      "learning_rate": 9.439577134755763e-06,
+      "loss": 0.476,
+      "step": 6550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8221458505598757,
+      "learning_rate": 9.436948818931947e-06,
+      "loss": 0.4741,
+      "step": 6551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8672724748753476,
+      "learning_rate": 9.434320542127075e-06,
+      "loss": 0.5206,
+      "step": 6552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9586388160464261,
+      "learning_rate": 9.43169230452329e-06,
+      "loss": 0.5331,
+      "step": 6553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9699403966085434,
+      "learning_rate": 9.429064106302724e-06,
+      "loss": 0.5506,
+      "step": 6554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9896460820121302,
+      "learning_rate": 9.426435947647508e-06,
+      "loss": 0.5081,
+      "step": 6555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8609140681024577,
+      "learning_rate": 9.42380782873977e-06,
+      "loss": 0.4972,
+      "step": 6556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9419865663383613,
+      "learning_rate": 9.421179749761643e-06,
+      "loss": 0.5552,
+      "step": 6557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8792446463968906,
+      "learning_rate": 9.418551710895243e-06,
+      "loss": 0.4424,
+      "step": 6558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9401380367951486,
+      "learning_rate": 9.415923712322693e-06,
+      "loss": 0.4913,
+      "step": 6559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9564624625255295,
+      "learning_rate": 9.413295754226115e-06,
+      "loss": 0.4873,
+      "step": 6560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9038480801805764,
+      "learning_rate": 9.410667836787619e-06,
+      "loss": 0.5728,
+      "step": 6561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8070238910164431,
+      "learning_rate": 9.408039960189317e-06,
+      "loss": 0.4674,
+      "step": 6562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9804935597251521,
+      "learning_rate": 9.405412124613325e-06,
+      "loss": 0.5187,
+      "step": 6563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8857730008887388,
+      "learning_rate": 9.402784330241743e-06,
+      "loss": 0.5002,
+      "step": 6564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8739355689761309,
+      "learning_rate": 9.400156577256675e-06,
+      "loss": 0.5278,
+      "step": 6565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9579620085111423,
+      "learning_rate": 9.397528865840229e-06,
+      "loss": 0.5324,
+      "step": 6566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8396241787978719,
+      "learning_rate": 9.394901196174496e-06,
+      "loss": 0.5279,
+      "step": 6567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9116099647277938,
+      "learning_rate": 9.39227356844157e-06,
+      "loss": 0.4807,
+      "step": 6568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9085824599768635,
+      "learning_rate": 9.389645982823552e-06,
+      "loss": 0.5559,
+      "step": 6569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.946189269407832,
+      "learning_rate": 9.387018439502524e-06,
+      "loss": 0.5012,
+      "step": 6570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8718190611058443,
+      "learning_rate": 9.384390938660572e-06,
+      "loss": 0.5061,
+      "step": 6571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8547334865321193,
+      "learning_rate": 9.381763480479784e-06,
+      "loss": 0.5166,
+      "step": 6572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0652378374846714,
+      "learning_rate": 9.379136065142241e-06,
+      "loss": 0.5806,
+      "step": 6573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8143736521167203,
+      "learning_rate": 9.376508692830012e-06,
+      "loss": 0.5379,
+      "step": 6574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9218503896969388,
+      "learning_rate": 9.373881363725182e-06,
+      "loss": 0.5158,
+      "step": 6575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9985452463714473,
+      "learning_rate": 9.371254078009819e-06,
+      "loss": 0.5931,
+      "step": 6576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9356941082925173,
+      "learning_rate": 9.368626835865987e-06,
+      "loss": 0.4867,
+      "step": 6577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9597854538853388,
+      "learning_rate": 9.365999637475756e-06,
+      "loss": 0.609,
+      "step": 6578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8403719685382467,
+      "learning_rate": 9.363372483021191e-06,
+      "loss": 0.4493,
+      "step": 6579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.989244981681324,
+      "learning_rate": 9.360745372684346e-06,
+      "loss": 0.6439,
+      "step": 6580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9416956127576758,
+      "learning_rate": 9.358118306647278e-06,
+      "loss": 0.5829,
+      "step": 6581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9294458423197391,
+      "learning_rate": 9.355491285092045e-06,
+      "loss": 0.597,
+      "step": 6582
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8599793380974954,
+      "learning_rate": 9.352864308200693e-06,
+      "loss": 0.5003,
+      "step": 6583
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8562934086264922,
+      "learning_rate": 9.350237376155269e-06,
+      "loss": 0.4935,
+      "step": 6584
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9499034838164225,
+      "learning_rate": 9.347610489137821e-06,
+      "loss": 0.5657,
+      "step": 6585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9886652459642102,
+      "learning_rate": 9.344983647330386e-06,
+      "loss": 0.4546,
+      "step": 6586
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8408184507746581,
+      "learning_rate": 9.342356850915003e-06,
+      "loss": 0.4653,
+      "step": 6587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9170817507583063,
+      "learning_rate": 9.339730100073709e-06,
+      "loss": 0.5481,
+      "step": 6588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8354998641142983,
+      "learning_rate": 9.33710339498853e-06,
+      "loss": 0.4542,
+      "step": 6589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8346026834004256,
+      "learning_rate": 9.3344767358415e-06,
+      "loss": 0.4872,
+      "step": 6590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9166805271898558,
+      "learning_rate": 9.331850122814644e-06,
+      "loss": 0.5649,
+      "step": 6591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9016451210187113,
+      "learning_rate": 9.329223556089976e-06,
+      "loss": 0.5336,
+      "step": 6592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8541527908291205,
+      "learning_rate": 9.326597035849524e-06,
+      "loss": 0.514,
+      "step": 6593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8920893747093328,
+      "learning_rate": 9.323970562275302e-06,
+      "loss": 0.5288,
+      "step": 6594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.887453410935998,
+      "learning_rate": 9.321344135549316e-06,
+      "loss": 0.5056,
+      "step": 6595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8794368276693035,
+      "learning_rate": 9.318717755853583e-06,
+      "loss": 0.4963,
+      "step": 6596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8652655941315313,
+      "learning_rate": 9.316091423370105e-06,
+      "loss": 0.5329,
+      "step": 6597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9208644302407073,
+      "learning_rate": 9.313465138280882e-06,
+      "loss": 0.5108,
+      "step": 6598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.963722543617233,
+      "learning_rate": 9.31083890076792e-06,
+      "loss": 0.575,
+      "step": 6599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9541286094340493,
+      "learning_rate": 9.30821271101321e-06,
+      "loss": 0.5455,
+      "step": 6600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0065486795180587,
+      "learning_rate": 9.305586569198742e-06,
+      "loss": 0.5707,
+      "step": 6601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9259839776751105,
+      "learning_rate": 9.30296047550651e-06,
+      "loss": 0.4889,
+      "step": 6602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9726764156210098,
+      "learning_rate": 9.300334430118504e-06,
+      "loss": 0.5617,
+      "step": 6603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.884838643280822,
+      "learning_rate": 9.297708433216693e-06,
+      "loss": 0.503,
+      "step": 6604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8280384803844053,
+      "learning_rate": 9.29508248498307e-06,
+      "loss": 0.4996,
+      "step": 6605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8906426897381897,
+      "learning_rate": 9.292456585599607e-06,
+      "loss": 0.5115,
+      "step": 6606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.996972130518505,
+      "learning_rate": 9.289830735248269e-06,
+      "loss": 0.5234,
+      "step": 6607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8291707987913881,
+      "learning_rate": 9.287204934111035e-06,
+      "loss": 0.5578,
+      "step": 6608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.915568083458182,
+      "learning_rate": 9.284579182369868e-06,
+      "loss": 0.5419,
+      "step": 6609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9954268850193402,
+      "learning_rate": 9.281953480206725e-06,
+      "loss": 0.522,
+      "step": 6610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9047193985797033,
+      "learning_rate": 9.279327827803573e-06,
+      "loss": 0.4784,
+      "step": 6611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8701895692193299,
+      "learning_rate": 9.276702225342363e-06,
+      "loss": 0.4848,
+      "step": 6612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9950557491152223,
+      "learning_rate": 9.274076673005042e-06,
+      "loss": 0.5792,
+      "step": 6613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9822586130445328,
+      "learning_rate": 9.271451170973568e-06,
+      "loss": 0.4991,
+      "step": 6614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9107882853370662,
+      "learning_rate": 9.268825719429884e-06,
+      "loss": 0.5307,
+      "step": 6615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9603488113009204,
+      "learning_rate": 9.266200318555923e-06,
+      "loss": 0.5128,
+      "step": 6616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9931853248124778,
+      "learning_rate": 9.263574968533635e-06,
+      "loss": 0.5775,
+      "step": 6617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9562773602607122,
+      "learning_rate": 9.260949669544946e-06,
+      "loss": 0.4898,
+      "step": 6618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9262129155659268,
+      "learning_rate": 9.258324421771785e-06,
+      "loss": 0.5486,
+      "step": 6619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9512134810062653,
+      "learning_rate": 9.255699225396091e-06,
+      "loss": 0.5639,
+      "step": 6620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8827005731153343,
+      "learning_rate": 9.25307408059978e-06,
+      "loss": 0.5139,
+      "step": 6621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8965142343662935,
+      "learning_rate": 9.250448987564765e-06,
+      "loss": 0.5929,
+      "step": 6622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9594934835087743,
+      "learning_rate": 9.247823946472978e-06,
+      "loss": 0.5294,
+      "step": 6623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9948381063542425,
+      "learning_rate": 9.245198957506324e-06,
+      "loss": 0.5583,
+      "step": 6624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9747739230563899,
+      "learning_rate": 9.242574020846706e-06,
+      "loss": 0.596,
+      "step": 6625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9866782142190802,
+      "learning_rate": 9.239949136676042e-06,
+      "loss": 0.5165,
+      "step": 6626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.836617718413994,
+      "learning_rate": 9.23732430517623e-06,
+      "loss": 0.4931,
+      "step": 6627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9969240754051573,
+      "learning_rate": 9.23469952652916e-06,
+      "loss": 0.5709,
+      "step": 6628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.084925424323057,
+      "learning_rate": 9.232074800916741e-06,
+      "loss": 0.5608,
+      "step": 6629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8014165391119704,
+      "learning_rate": 9.229450128520856e-06,
+      "loss": 0.4672,
+      "step": 6630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9275634877793757,
+      "learning_rate": 9.22682550952339e-06,
+      "loss": 0.5796,
+      "step": 6631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9943343152879219,
+      "learning_rate": 9.224200944106234e-06,
+      "loss": 0.5677,
+      "step": 6632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.875222035331371,
+      "learning_rate": 9.221576432451266e-06,
+      "loss": 0.5456,
+      "step": 6633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7647523024721653,
+      "learning_rate": 9.218951974740354e-06,
+      "loss": 0.4952,
+      "step": 6634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8870884362180773,
+      "learning_rate": 9.216327571155384e-06,
+      "loss": 0.4923,
+      "step": 6635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8953762772911025,
+      "learning_rate": 9.213703221878217e-06,
+      "loss": 0.5387,
+      "step": 6636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7947655687162931,
+      "learning_rate": 9.211078927090714e-06,
+      "loss": 0.536,
+      "step": 6637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9651351409257324,
+      "learning_rate": 9.208454686974748e-06,
+      "loss": 0.4714,
+      "step": 6638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9095238031554407,
+      "learning_rate": 9.205830501712168e-06,
+      "loss": 0.5529,
+      "step": 6639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8954728476730376,
+      "learning_rate": 9.203206371484827e-06,
+      "loss": 0.4868,
+      "step": 6640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9236020847975327,
+      "learning_rate": 9.200582296474581e-06,
+      "loss": 0.5553,
+      "step": 6641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.973573370807308,
+      "learning_rate": 9.197958276863274e-06,
+      "loss": 0.6337,
+      "step": 6642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8218184879634992,
+      "learning_rate": 9.195334312832742e-06,
+      "loss": 0.5422,
+      "step": 6643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8703081984309136,
+      "learning_rate": 9.192710404564833e-06,
+      "loss": 0.5375,
+      "step": 6644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.868266999358897,
+      "learning_rate": 9.190086552241375e-06,
+      "loss": 0.5186,
+      "step": 6645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9387964986316296,
+      "learning_rate": 9.187462756044198e-06,
+      "loss": 0.5584,
+      "step": 6646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9318884814426076,
+      "learning_rate": 9.184839016155136e-06,
+      "loss": 0.5393,
+      "step": 6647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.894118676292748,
+      "learning_rate": 9.182215332756003e-06,
+      "loss": 0.5073,
+      "step": 6648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9782985369401567,
+      "learning_rate": 9.179591706028626e-06,
+      "loss": 0.5251,
+      "step": 6649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9914198352405411,
+      "learning_rate": 9.176968136154815e-06,
+      "loss": 0.5461,
+      "step": 6650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8639055680751122,
+      "learning_rate": 9.174344623316377e-06,
+      "loss": 0.5449,
+      "step": 6651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9484262400017193,
+      "learning_rate": 9.171721167695132e-06,
+      "loss": 0.4951,
+      "step": 6652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.897854575462643,
+      "learning_rate": 9.169097769472873e-06,
+      "loss": 0.5389,
+      "step": 6653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9094567346832353,
+      "learning_rate": 9.166474428831399e-06,
+      "loss": 0.5056,
+      "step": 6654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8990709731930566,
+      "learning_rate": 9.16385114595251e-06,
+      "loss": 0.4614,
+      "step": 6655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8562388383715867,
+      "learning_rate": 9.161227921017996e-06,
+      "loss": 0.5157,
+      "step": 6656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8713141936842353,
+      "learning_rate": 9.158604754209637e-06,
+      "loss": 0.4757,
+      "step": 6657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8443143174087063,
+      "learning_rate": 9.15598164570923e-06,
+      "loss": 0.5069,
+      "step": 6658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.92136601553196,
+      "learning_rate": 9.153358595698542e-06,
+      "loss": 0.5045,
+      "step": 6659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.827326202667017,
+      "learning_rate": 9.15073560435935e-06,
+      "loss": 0.5152,
+      "step": 6660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9489168131505338,
+      "learning_rate": 9.148112671873433e-06,
+      "loss": 0.5234,
+      "step": 6661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9148988968297469,
+      "learning_rate": 9.14548979842255e-06,
+      "loss": 0.5703,
+      "step": 6662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9248357145748127,
+      "learning_rate": 9.142866984188465e-06,
+      "loss": 0.5427,
+      "step": 6663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9191461067370301,
+      "learning_rate": 9.140244229352939e-06,
+      "loss": 0.5771,
+      "step": 6664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8586547352456922,
+      "learning_rate": 9.137621534097727e-06,
+      "loss": 0.5078,
+      "step": 6665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9672594179057945,
+      "learning_rate": 9.134998898604573e-06,
+      "loss": 0.5568,
+      "step": 6666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9385908631281702,
+      "learning_rate": 9.13237632305523e-06,
+      "loss": 0.5224,
+      "step": 6667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8288168870807058,
+      "learning_rate": 9.129753807631441e-06,
+      "loss": 0.4726,
+      "step": 6668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.822299693879274,
+      "learning_rate": 9.127131352514936e-06,
+      "loss": 0.4879,
+      "step": 6669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9248585728305219,
+      "learning_rate": 9.124508957887458e-06,
+      "loss": 0.5317,
+      "step": 6670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8571133889927011,
+      "learning_rate": 9.121886623930735e-06,
+      "loss": 0.55,
+      "step": 6671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9419444582984893,
+      "learning_rate": 9.119264350826484e-06,
+      "loss": 0.538,
+      "step": 6672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9999021530257328,
+      "learning_rate": 9.116642138756436e-06,
+      "loss": 0.5301,
+      "step": 6673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9350978949571519,
+      "learning_rate": 9.114019987902305e-06,
+      "loss": 0.5313,
+      "step": 6674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8767162700759692,
+      "learning_rate": 9.111397898445798e-06,
+      "loss": 0.4861,
+      "step": 6675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.811548256992899,
+      "learning_rate": 9.108775870568633e-06,
+      "loss": 0.4757,
+      "step": 6676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8172579337528821,
+      "learning_rate": 9.10615390445251e-06,
+      "loss": 0.4537,
+      "step": 6677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8618640332615486,
+      "learning_rate": 9.103532000279126e-06,
+      "loss": 0.473,
+      "step": 6678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8147714207686093,
+      "learning_rate": 9.100910158230181e-06,
+      "loss": 0.4793,
+      "step": 6679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8892794701271312,
+      "learning_rate": 9.098288378487365e-06,
+      "loss": 0.4554,
+      "step": 6680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9288749782454486,
+      "learning_rate": 9.095666661232359e-06,
+      "loss": 0.5453,
+      "step": 6681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9337429815153883,
+      "learning_rate": 9.093045006646858e-06,
+      "loss": 0.5492,
+      "step": 6682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8679010196336439,
+      "learning_rate": 9.090423414912533e-06,
+      "loss": 0.5516,
+      "step": 6683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0241778177487386,
+      "learning_rate": 9.087801886211054e-06,
+      "loss": 0.5558,
+      "step": 6684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9616016252490783,
+      "learning_rate": 9.085180420724098e-06,
+      "loss": 0.5434,
+      "step": 6685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0637639240727763,
+      "learning_rate": 9.08255901863333e-06,
+      "loss": 0.5941,
+      "step": 6686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8857949184792532,
+      "learning_rate": 9.079937680120403e-06,
+      "loss": 0.513,
+      "step": 6687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8920117605122396,
+      "learning_rate": 9.07731640536698e-06,
+      "loss": 0.5177,
+      "step": 6688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9771864597037313,
+      "learning_rate": 9.074695194554716e-06,
+      "loss": 0.5003,
+      "step": 6689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8469320043425077,
+      "learning_rate": 9.072074047865249e-06,
+      "loss": 0.5068,
+      "step": 6690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9443738087006439,
+      "learning_rate": 9.06945296548023e-06,
+      "loss": 0.5157,
+      "step": 6691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9300489829476154,
+      "learning_rate": 9.066831947581297e-06,
+      "loss": 0.5403,
+      "step": 6692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.053327155730105,
+      "learning_rate": 9.064210994350077e-06,
+      "loss": 0.6007,
+      "step": 6693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4474967110123402,
+      "learning_rate": 9.061590105968208e-06,
+      "loss": 0.575,
+      "step": 6694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8526689997310141,
+      "learning_rate": 9.058969282617314e-06,
+      "loss": 0.5209,
+      "step": 6695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8097094716494864,
+      "learning_rate": 9.056348524479011e-06,
+      "loss": 0.4076,
+      "step": 6696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8654787700636426,
+      "learning_rate": 9.05372783173492e-06,
+      "loss": 0.4703,
+      "step": 6697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8186544312924194,
+      "learning_rate": 9.051107204566652e-06,
+      "loss": 0.4329,
+      "step": 6698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8674267898300603,
+      "learning_rate": 9.04848664315581e-06,
+      "loss": 0.5081,
+      "step": 6699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.925496440387974,
+      "learning_rate": 9.045866147684002e-06,
+      "loss": 0.6014,
+      "step": 6700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.974342361846761,
+      "learning_rate": 9.043245718332821e-06,
+      "loss": 0.5809,
+      "step": 6701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8923396073658344,
+      "learning_rate": 9.040625355283865e-06,
+      "loss": 0.5784,
+      "step": 6702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.861738870749732,
+      "learning_rate": 9.038005058718722e-06,
+      "loss": 0.5445,
+      "step": 6703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8438303457500101,
+      "learning_rate": 9.035384828818974e-06,
+      "loss": 0.5452,
+      "step": 6704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8829006208766645,
+      "learning_rate": 9.0327646657662e-06,
+      "loss": 0.5218,
+      "step": 6705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8418172476199486,
+      "learning_rate": 9.03014456974198e-06,
+      "loss": 0.4835,
+      "step": 6706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.008223396610821,
+      "learning_rate": 9.027524540927878e-06,
+      "loss": 0.5586,
+      "step": 6707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8712725958245666,
+      "learning_rate": 9.024904579505465e-06,
+      "loss": 0.4877,
+      "step": 6708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9502419365005647,
+      "learning_rate": 9.0222846856563e-06,
+      "loss": 0.5817,
+      "step": 6709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.951721220447067,
+      "learning_rate": 9.019664859561938e-06,
+      "loss": 0.5443,
+      "step": 6710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8979820465371918,
+      "learning_rate": 9.01704510140393e-06,
+      "loss": 0.5145,
+      "step": 6711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9145141115880605,
+      "learning_rate": 9.014425411363827e-06,
+      "loss": 0.5198,
+      "step": 6712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.891068921265497,
+      "learning_rate": 9.011805789623168e-06,
+      "loss": 0.5452,
+      "step": 6713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8933266887513471,
+      "learning_rate": 9.00918623636349e-06,
+      "loss": 0.5038,
+      "step": 6714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8983927826301715,
+      "learning_rate": 9.00656675176633e-06,
+      "loss": 0.5516,
+      "step": 6715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.145812936653427,
+      "learning_rate": 9.003947336013212e-06,
+      "loss": 0.5339,
+      "step": 6716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.800954853444465,
+      "learning_rate": 9.001327989285658e-06,
+      "loss": 0.4313,
+      "step": 6717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9599757274661858,
+      "learning_rate": 8.99870871176519e-06,
+      "loss": 0.5393,
+      "step": 6718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8637559886276752,
+      "learning_rate": 8.99608950363332e-06,
+      "loss": 0.4987,
+      "step": 6719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9177154870428762,
+      "learning_rate": 8.993470365071557e-06,
+      "loss": 0.4984,
+      "step": 6720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8671382758403665,
+      "learning_rate": 8.990851296261403e-06,
+      "loss": 0.4806,
+      "step": 6721
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8485265776210158,
+      "learning_rate": 8.988232297384363e-06,
+      "loss": 0.5311,
+      "step": 6722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8996283623951162,
+      "learning_rate": 8.985613368621923e-06,
+      "loss": 0.479,
+      "step": 6723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8339680584043668,
+      "learning_rate": 8.98299451015558e-06,
+      "loss": 0.4592,
+      "step": 6724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.911429171245189,
+      "learning_rate": 8.980375722166816e-06,
+      "loss": 0.5446,
+      "step": 6725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9296998668878459,
+      "learning_rate": 8.977757004837107e-06,
+      "loss": 0.526,
+      "step": 6726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8934152401740832,
+      "learning_rate": 8.975138358347931e-06,
+      "loss": 0.5168,
+      "step": 6727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8699340458300051,
+      "learning_rate": 8.97251978288076e-06,
+      "loss": 0.5115,
+      "step": 6728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8042443273524694,
+      "learning_rate": 8.969901278617056e-06,
+      "loss": 0.493,
+      "step": 6729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9139150214408609,
+      "learning_rate": 8.967282845738278e-06,
+      "loss": 0.5782,
+      "step": 6730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9087431719596738,
+      "learning_rate": 8.964664484425887e-06,
+      "loss": 0.4987,
+      "step": 6731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9863678036432829,
+      "learning_rate": 8.962046194861324e-06,
+      "loss": 0.5613,
+      "step": 6732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8027492964740189,
+      "learning_rate": 8.959427977226041e-06,
+      "loss": 0.5256,
+      "step": 6733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9706281749582049,
+      "learning_rate": 8.956809831701478e-06,
+      "loss": 0.5961,
+      "step": 6734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.837992334486424,
+      "learning_rate": 8.954191758469065e-06,
+      "loss": 0.5812,
+      "step": 6735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9209016693375767,
+      "learning_rate": 8.951573757710237e-06,
+      "loss": 0.5021,
+      "step": 6736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8954100318518413,
+      "learning_rate": 8.948955829606419e-06,
+      "loss": 0.5376,
+      "step": 6737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8414160352133793,
+      "learning_rate": 8.946337974339025e-06,
+      "loss": 0.4799,
+      "step": 6738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9261969375567992,
+      "learning_rate": 8.94372019208948e-06,
+      "loss": 0.5435,
+      "step": 6739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8862438885604225,
+      "learning_rate": 8.941102483039188e-06,
+      "loss": 0.4679,
+      "step": 6740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.998787432571941,
+      "learning_rate": 8.938484847369552e-06,
+      "loss": 0.529,
+      "step": 6741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.933973434550398,
+      "learning_rate": 8.935867285261977e-06,
+      "loss": 0.5466,
+      "step": 6742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8179520104205413,
+      "learning_rate": 8.933249796897857e-06,
+      "loss": 0.5482,
+      "step": 6743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0059179488657666,
+      "learning_rate": 8.930632382458574e-06,
+      "loss": 0.5231,
+      "step": 6744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8559062638771029,
+      "learning_rate": 8.928015042125523e-06,
+      "loss": 0.5056,
+      "step": 6745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9097764966615546,
+      "learning_rate": 8.92539777608008e-06,
+      "loss": 0.5607,
+      "step": 6746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9397337848518229,
+      "learning_rate": 8.92278058450361e-06,
+      "loss": 0.5203,
+      "step": 6747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9255883182156407,
+      "learning_rate": 8.920163467577498e-06,
+      "loss": 0.5321,
+      "step": 6748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0030233998599511,
+      "learning_rate": 8.9175464254831e-06,
+      "loss": 0.5644,
+      "step": 6749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8286218798653401,
+      "learning_rate": 8.914929458401767e-06,
+      "loss": 0.5238,
+      "step": 6750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8400491207198404,
+      "learning_rate": 8.912312566514867e-06,
+      "loss": 0.5077,
+      "step": 6751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0080073977252997,
+      "learning_rate": 8.909695750003741e-06,
+      "loss": 0.5366,
+      "step": 6752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9119849927606701,
+      "learning_rate": 8.907079009049728e-06,
+      "loss": 0.4794,
+      "step": 6753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9238807903194287,
+      "learning_rate": 8.904462343834174e-06,
+      "loss": 0.4932,
+      "step": 6754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8953783979565785,
+      "learning_rate": 8.901845754538408e-06,
+      "loss": 0.5132,
+      "step": 6755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8881835413458181,
+      "learning_rate": 8.899229241343753e-06,
+      "loss": 0.5388,
+      "step": 6756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8059493226002098,
+      "learning_rate": 8.89661280443154e-06,
+      "loss": 0.4756,
+      "step": 6757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8589226325513761,
+      "learning_rate": 8.89399644398308e-06,
+      "loss": 0.5012,
+      "step": 6758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8661157764886962,
+      "learning_rate": 8.891380160179683e-06,
+      "loss": 0.5472,
+      "step": 6759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8802847133080646,
+      "learning_rate": 8.88876395320266e-06,
+      "loss": 0.5278,
+      "step": 6760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.839132635416687,
+      "learning_rate": 8.88614782323331e-06,
+      "loss": 0.5119,
+      "step": 6761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8662673702579419,
+      "learning_rate": 8.883531770452924e-06,
+      "loss": 0.5376,
+      "step": 6762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8629463522649556,
+      "learning_rate": 8.880915795042798e-06,
+      "loss": 0.4509,
+      "step": 6763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9198570897646432,
+      "learning_rate": 8.878299897184218e-06,
+      "loss": 0.5309,
+      "step": 6764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9682401080681828,
+      "learning_rate": 8.875684077058453e-06,
+      "loss": 0.5539,
+      "step": 6765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9131512395292984,
+      "learning_rate": 8.87306833484679e-06,
+      "loss": 0.5804,
+      "step": 6766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9043883538180063,
+      "learning_rate": 8.870452670730491e-06,
+      "loss": 0.5376,
+      "step": 6767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8842007847462771,
+      "learning_rate": 8.867837084890817e-06,
+      "loss": 0.5603,
+      "step": 6768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8917436434495278,
+      "learning_rate": 8.865221577509034e-06,
+      "loss": 0.5391,
+      "step": 6769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8969528174452776,
+      "learning_rate": 8.862606148766386e-06,
+      "loss": 0.5521,
+      "step": 6770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8549046038065036,
+      "learning_rate": 8.85999079884412e-06,
+      "loss": 0.5104,
+      "step": 6771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8370046492826245,
+      "learning_rate": 8.857375527923487e-06,
+      "loss": 0.4694,
+      "step": 6772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9201091933672658,
+      "learning_rate": 8.854760336185709e-06,
+      "loss": 0.5549,
+      "step": 6773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9192420832976912,
+      "learning_rate": 8.85214522381203e-06,
+      "loss": 0.5098,
+      "step": 6774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8566796001040332,
+      "learning_rate": 8.849530190983669e-06,
+      "loss": 0.4605,
+      "step": 6775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8527635064927266,
+      "learning_rate": 8.846915237881838e-06,
+      "loss": 0.5575,
+      "step": 6776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8049186625770314,
+      "learning_rate": 8.844300364687766e-06,
+      "loss": 0.4864,
+      "step": 6777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.874725545010611,
+      "learning_rate": 8.841685571582652e-06,
+      "loss": 0.5308,
+      "step": 6778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9224938208424457,
+      "learning_rate": 8.839070858747697e-06,
+      "loss": 0.5409,
+      "step": 6779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8814961698465333,
+      "learning_rate": 8.836456226364106e-06,
+      "loss": 0.5138,
+      "step": 6780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9087348939049014,
+      "learning_rate": 8.833841674613066e-06,
+      "loss": 0.548,
+      "step": 6781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9069806988718064,
+      "learning_rate": 8.831227203675759e-06,
+      "loss": 0.5685,
+      "step": 6782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9541666664847532,
+      "learning_rate": 8.828612813733375e-06,
+      "loss": 0.5086,
+      "step": 6783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8871536751057871,
+      "learning_rate": 8.825998504967083e-06,
+      "loss": 0.5631,
+      "step": 6784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8795996020980996,
+      "learning_rate": 8.823384277558049e-06,
+      "loss": 0.5313,
+      "step": 6785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9247932552615178,
+      "learning_rate": 8.820770131687447e-06,
+      "loss": 0.5294,
+      "step": 6786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8857003101974371,
+      "learning_rate": 8.818156067536428e-06,
+      "loss": 0.5431,
+      "step": 6787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9214931834118468,
+      "learning_rate": 8.81554208528614e-06,
+      "loss": 0.5022,
+      "step": 6788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9891148743193596,
+      "learning_rate": 8.81292818511774e-06,
+      "loss": 0.5028,
+      "step": 6789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8078361977439683,
+      "learning_rate": 8.810314367212363e-06,
+      "loss": 0.4507,
+      "step": 6790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8869455383110579,
+      "learning_rate": 8.807700631751142e-06,
+      "loss": 0.5118,
+      "step": 6791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9101666591005333,
+      "learning_rate": 8.805086978915215e-06,
+      "loss": 0.5276,
+      "step": 6792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8699626001339014,
+      "learning_rate": 8.802473408885698e-06,
+      "loss": 0.5151,
+      "step": 6793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8984143344812208,
+      "learning_rate": 8.799859921843708e-06,
+      "loss": 0.5043,
+      "step": 6794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8903818770283506,
+      "learning_rate": 8.797246517970365e-06,
+      "loss": 0.4752,
+      "step": 6795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8499060192340044,
+      "learning_rate": 8.79463319744677e-06,
+      "loss": 0.5005,
+      "step": 6796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8716575656378632,
+      "learning_rate": 8.792019960454025e-06,
+      "loss": 0.5114,
+      "step": 6797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9138479375865836,
+      "learning_rate": 8.789406807173226e-06,
+      "loss": 0.5285,
+      "step": 6798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9210455881917426,
+      "learning_rate": 8.786793737785465e-06,
+      "loss": 0.5623,
+      "step": 6799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.059953054962172,
+      "learning_rate": 8.784180752471814e-06,
+      "loss": 0.6021,
+      "step": 6800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9384117653547553,
+      "learning_rate": 8.781567851413363e-06,
+      "loss": 0.4719,
+      "step": 6801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9469162552609387,
+      "learning_rate": 8.77895503479118e-06,
+      "loss": 0.544,
+      "step": 6802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.918780292574788,
+      "learning_rate": 8.776342302786327e-06,
+      "loss": 0.5458,
+      "step": 6803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8163978896404098,
+      "learning_rate": 8.773729655579872e-06,
+      "loss": 0.4747,
+      "step": 6804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8581993144879657,
+      "learning_rate": 8.771117093352861e-06,
+      "loss": 0.4709,
+      "step": 6805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7416031093198225,
+      "learning_rate": 8.768504616286343e-06,
+      "loss": 0.4403,
+      "step": 6806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9415202265596095,
+      "learning_rate": 8.765892224561367e-06,
+      "loss": 0.4747,
+      "step": 6807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8392199576195705,
+      "learning_rate": 8.763279918358965e-06,
+      "loss": 0.4821,
+      "step": 6808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9156867689783389,
+      "learning_rate": 8.760667697860163e-06,
+      "loss": 0.5173,
+      "step": 6809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8526700674513049,
+      "learning_rate": 8.758055563245994e-06,
+      "loss": 0.4999,
+      "step": 6810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9897941428051377,
+      "learning_rate": 8.755443514697475e-06,
+      "loss": 0.5442,
+      "step": 6811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9994090842074854,
+      "learning_rate": 8.75283155239561e-06,
+      "loss": 0.5672,
+      "step": 6812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9643497009203205,
+      "learning_rate": 8.750219676521417e-06,
+      "loss": 0.5924,
+      "step": 6813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9052958275216263,
+      "learning_rate": 8.747607887255892e-06,
+      "loss": 0.4995,
+      "step": 6814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9110981702917157,
+      "learning_rate": 8.744996184780027e-06,
+      "loss": 0.5191,
+      "step": 6815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9418260441700346,
+      "learning_rate": 8.742384569274815e-06,
+      "loss": 0.5403,
+      "step": 6816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9570874072452588,
+      "learning_rate": 8.739773040921237e-06,
+      "loss": 0.5515,
+      "step": 6817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8969656147692997,
+      "learning_rate": 8.737161599900267e-06,
+      "loss": 0.5075,
+      "step": 6818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8998191286869733,
+      "learning_rate": 8.734550246392881e-06,
+      "loss": 0.5951,
+      "step": 6819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8362595923787626,
+      "learning_rate": 8.73193898058004e-06,
+      "loss": 0.4867,
+      "step": 6820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8519217677907318,
+      "learning_rate": 8.7293278026427e-06,
+      "loss": 0.545,
+      "step": 6821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9234059581598991,
+      "learning_rate": 8.726716712761821e-06,
+      "loss": 0.5394,
+      "step": 6822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8558200276209291,
+      "learning_rate": 8.724105711118342e-06,
+      "loss": 0.5496,
+      "step": 6823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8377667534966279,
+      "learning_rate": 8.721494797893201e-06,
+      "loss": 0.5268,
+      "step": 6824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8517243807535965,
+      "learning_rate": 8.718883973267344e-06,
+      "loss": 0.5228,
+      "step": 6825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8573727687125215,
+      "learning_rate": 8.716273237421688e-06,
+      "loss": 0.4513,
+      "step": 6826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8949363228005951,
+      "learning_rate": 8.713662590537155e-06,
+      "loss": 0.5142,
+      "step": 6827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9823612055364075,
+      "learning_rate": 8.711052032794668e-06,
+      "loss": 0.6011,
+      "step": 6828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9043280010894629,
+      "learning_rate": 8.708441564375132e-06,
+      "loss": 0.5257,
+      "step": 6829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8884025036511776,
+      "learning_rate": 8.705831185459446e-06,
+      "loss": 0.5371,
+      "step": 6830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8851961931117828,
+      "learning_rate": 8.703220896228515e-06,
+      "loss": 0.5296,
+      "step": 6831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9496141722730437,
+      "learning_rate": 8.700610696863225e-06,
+      "loss": 0.5642,
+      "step": 6832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8771895138899715,
+      "learning_rate": 8.698000587544457e-06,
+      "loss": 0.5325,
+      "step": 6833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8952605449602975,
+      "learning_rate": 8.695390568453099e-06,
+      "loss": 0.5094,
+      "step": 6834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9544184892925124,
+      "learning_rate": 8.692780639770015e-06,
+      "loss": 0.5366,
+      "step": 6835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9309194420917352,
+      "learning_rate": 8.69017080167607e-06,
+      "loss": 0.541,
+      "step": 6836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8723271018988956,
+      "learning_rate": 8.687561054352132e-06,
+      "loss": 0.4398,
+      "step": 6837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8901099533404632,
+      "learning_rate": 8.684951397979049e-06,
+      "loss": 0.4952,
+      "step": 6838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9158614565675948,
+      "learning_rate": 8.68234183273766e-06,
+      "loss": 0.498,
+      "step": 6839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9034404937210446,
+      "learning_rate": 8.679732358808822e-06,
+      "loss": 0.5241,
+      "step": 6840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.91821176168655,
+      "learning_rate": 8.677122976373356e-06,
+      "loss": 0.5759,
+      "step": 6841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9047712664252271,
+      "learning_rate": 8.674513685612093e-06,
+      "loss": 0.5376,
+      "step": 6842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9452668129583185,
+      "learning_rate": 8.67190448670586e-06,
+      "loss": 0.5454,
+      "step": 6843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9045047109963618,
+      "learning_rate": 8.669295379835467e-06,
+      "loss": 0.4366,
+      "step": 6844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.848893201507573,
+      "learning_rate": 8.66668636518172e-06,
+      "loss": 0.5043,
+      "step": 6845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9522331654771756,
+      "learning_rate": 8.66407744292543e-06,
+      "loss": 0.5705,
+      "step": 6846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8362035815919093,
+      "learning_rate": 8.661468613247387e-06,
+      "loss": 0.528,
+      "step": 6847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8439852570234441,
+      "learning_rate": 8.658859876328379e-06,
+      "loss": 0.5723,
+      "step": 6848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8729192359407367,
+      "learning_rate": 8.656251232349196e-06,
+      "loss": 0.5041,
+      "step": 6849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9365431033561334,
+      "learning_rate": 8.653642681490608e-06,
+      "loss": 0.541,
+      "step": 6850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9090828954084572,
+      "learning_rate": 8.651034223933387e-06,
+      "loss": 0.5001,
+      "step": 6851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9040470072628123,
+      "learning_rate": 8.648425859858302e-06,
+      "loss": 0.5238,
+      "step": 6852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8724924522448079,
+      "learning_rate": 8.645817589446104e-06,
+      "loss": 0.5102,
+      "step": 6853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9094503790788749,
+      "learning_rate": 8.643209412877545e-06,
+      "loss": 0.4854,
+      "step": 6854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8456235151710461,
+      "learning_rate": 8.640601330333372e-06,
+      "loss": 0.4561,
+      "step": 6855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.847154330493836,
+      "learning_rate": 8.637993341994318e-06,
+      "loss": 0.4553,
+      "step": 6856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9621209098974626,
+      "learning_rate": 8.63538544804112e-06,
+      "loss": 0.6182,
+      "step": 6857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8996509203921742,
+      "learning_rate": 8.6327776486545e-06,
+      "loss": 0.557,
+      "step": 6858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9084173699307313,
+      "learning_rate": 8.630169944015175e-06,
+      "loss": 0.5676,
+      "step": 6859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9108279490107368,
+      "learning_rate": 8.627562334303856e-06,
+      "loss": 0.5245,
+      "step": 6860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.951437155317072,
+      "learning_rate": 8.624954819701254e-06,
+      "loss": 0.5329,
+      "step": 6861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8850829837968626,
+      "learning_rate": 8.62234740038806e-06,
+      "loss": 0.5477,
+      "step": 6862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9464888994757814,
+      "learning_rate": 8.61974007654497e-06,
+      "loss": 0.5163,
+      "step": 6863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8720404929484166,
+      "learning_rate": 8.617132848352672e-06,
+      "loss": 0.5445,
+      "step": 6864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8728133761361274,
+      "learning_rate": 8.614525715991838e-06,
+      "loss": 0.4891,
+      "step": 6865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8127871687528759,
+      "learning_rate": 8.61191867964314e-06,
+      "loss": 0.5541,
+      "step": 6866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9500036866001732,
+      "learning_rate": 8.60931173948725e-06,
+      "loss": 0.5357,
+      "step": 6867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9133395294343943,
+      "learning_rate": 8.606704895704824e-06,
+      "loss": 0.5075,
+      "step": 6868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8935830929853792,
+      "learning_rate": 8.60409814847651e-06,
+      "loss": 0.5594,
+      "step": 6869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8317842659008998,
+      "learning_rate": 8.601491497982956e-06,
+      "loss": 0.4818,
+      "step": 6870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8450371891233477,
+      "learning_rate": 8.598884944404803e-06,
+      "loss": 0.5497,
+      "step": 6871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8318739755346117,
+      "learning_rate": 8.59627848792268e-06,
+      "loss": 0.4475,
+      "step": 6872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8811775776785117,
+      "learning_rate": 8.593672128717211e-06,
+      "loss": 0.549,
+      "step": 6873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9664851965942245,
+      "learning_rate": 8.591065866969019e-06,
+      "loss": 0.5226,
+      "step": 6874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.896869716587723,
+      "learning_rate": 8.588459702858709e-06,
+      "loss": 0.5001,
+      "step": 6875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8946735161087283,
+      "learning_rate": 8.585853636566891e-06,
+      "loss": 0.5486,
+      "step": 6876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9655935262527788,
+      "learning_rate": 8.583247668274163e-06,
+      "loss": 0.5092,
+      "step": 6877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8410789878708989,
+      "learning_rate": 8.580641798161114e-06,
+      "loss": 0.544,
+      "step": 6878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8277452055258614,
+      "learning_rate": 8.578036026408329e-06,
+      "loss": 0.4597,
+      "step": 6879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.815244774195561,
+      "learning_rate": 8.575430353196388e-06,
+      "loss": 0.444,
+      "step": 6880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9093735584737653,
+      "learning_rate": 8.572824778705858e-06,
+      "loss": 0.5305,
+      "step": 6881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.824061186847819,
+      "learning_rate": 8.570219303117305e-06,
+      "loss": 0.5214,
+      "step": 6882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8606650396604164,
+      "learning_rate": 8.567613926611287e-06,
+      "loss": 0.5188,
+      "step": 6883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8166098225852793,
+      "learning_rate": 8.565008649368353e-06,
+      "loss": 0.4376,
+      "step": 6884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8347231452093975,
+      "learning_rate": 8.562403471569045e-06,
+      "loss": 0.547,
+      "step": 6885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8097988214796327,
+      "learning_rate": 8.559798393393905e-06,
+      "loss": 0.5447,
+      "step": 6886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9244774758950597,
+      "learning_rate": 8.557193415023453e-06,
+      "loss": 0.5247,
+      "step": 6887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9112277686677142,
+      "learning_rate": 8.554588536638223e-06,
+      "loss": 0.5273,
+      "step": 6888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9646248263987668,
+      "learning_rate": 8.551983758418726e-06,
+      "loss": 0.5838,
+      "step": 6889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9010237227640583,
+      "learning_rate": 8.549379080545465e-06,
+      "loss": 0.5373,
+      "step": 6890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8840857549032455,
+      "learning_rate": 8.546774503198952e-06,
+      "loss": 0.5212,
+      "step": 6891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8703836152904741,
+      "learning_rate": 8.544170026559675e-06,
+      "loss": 0.515,
+      "step": 6892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9110363190568741,
+      "learning_rate": 8.541565650808121e-06,
+      "loss": 0.536,
+      "step": 6893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8667711655341527,
+      "learning_rate": 8.538961376124778e-06,
+      "loss": 0.4941,
+      "step": 6894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8986411060576323,
+      "learning_rate": 8.536357202690115e-06,
+      "loss": 0.5344,
+      "step": 6895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8943813528753555,
+      "learning_rate": 8.533753130684596e-06,
+      "loss": 0.5101,
+      "step": 6896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7868350464166751,
+      "learning_rate": 8.531149160288689e-06,
+      "loss": 0.4988,
+      "step": 6897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8812378069066342,
+      "learning_rate": 8.528545291682839e-06,
+      "loss": 0.4645,
+      "step": 6898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8133762449585381,
+      "learning_rate": 8.525941525047497e-06,
+      "loss": 0.4826,
+      "step": 6899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9905503171322216,
+      "learning_rate": 8.5233378605631e-06,
+      "loss": 0.5461,
+      "step": 6900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9258311041322348,
+      "learning_rate": 8.520734298410078e-06,
+      "loss": 0.5587,
+      "step": 6901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8495388917943362,
+      "learning_rate": 8.51813083876886e-06,
+      "loss": 0.5298,
+      "step": 6902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9334580528775865,
+      "learning_rate": 8.51552748181986e-06,
+      "loss": 0.4874,
+      "step": 6903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8939242695409655,
+      "learning_rate": 8.512924227743482e-06,
+      "loss": 0.5229,
+      "step": 6904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8807139695796883,
+      "learning_rate": 8.510321076720143e-06,
+      "loss": 0.5246,
+      "step": 6905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8559006349127389,
+      "learning_rate": 8.507718028930232e-06,
+      "loss": 0.4901,
+      "step": 6906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0000213916080734,
+      "learning_rate": 8.505115084554133e-06,
+      "loss": 0.5512,
+      "step": 6907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8732842596234691,
+      "learning_rate": 8.502512243772238e-06,
+      "loss": 0.5052,
+      "step": 6908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.883491030341863,
+      "learning_rate": 8.499909506764914e-06,
+      "loss": 0.4797,
+      "step": 6909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8306560563324287,
+      "learning_rate": 8.497306873712529e-06,
+      "loss": 0.5345,
+      "step": 6910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9538067482071597,
+      "learning_rate": 8.494704344795447e-06,
+      "loss": 0.5573,
+      "step": 6911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8159645915961217,
+      "learning_rate": 8.49210192019402e-06,
+      "loss": 0.4658,
+      "step": 6912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8417783364151031,
+      "learning_rate": 8.489499600088587e-06,
+      "loss": 0.4694,
+      "step": 6913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9480484593522762,
+      "learning_rate": 8.486897384659496e-06,
+      "loss": 0.5814,
+      "step": 6914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9328104862498665,
+      "learning_rate": 8.484295274087077e-06,
+      "loss": 0.5421,
+      "step": 6915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9775408046991969,
+      "learning_rate": 8.481693268551645e-06,
+      "loss": 0.5591,
+      "step": 6916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9216249536740916,
+      "learning_rate": 8.479091368233527e-06,
+      "loss": 0.5456,
+      "step": 6917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8600506186216511,
+      "learning_rate": 8.476489573313026e-06,
+      "loss": 0.4522,
+      "step": 6918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9424247242351319,
+      "learning_rate": 8.473887883970444e-06,
+      "loss": 0.5218,
+      "step": 6919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9545641299517474,
+      "learning_rate": 8.471286300386084e-06,
+      "loss": 0.5171,
+      "step": 6920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9165517930847911,
+      "learning_rate": 8.468684822740226e-06,
+      "loss": 0.554,
+      "step": 6921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8727313173381859,
+      "learning_rate": 8.466083451213145e-06,
+      "loss": 0.4691,
+      "step": 6922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8684377015353146,
+      "learning_rate": 8.463482185985127e-06,
+      "loss": 0.5302,
+      "step": 6923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.85865382887651,
+      "learning_rate": 8.46088102723643e-06,
+      "loss": 0.5485,
+      "step": 6924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9178264914623333,
+      "learning_rate": 8.458279975147308e-06,
+      "loss": 0.5013,
+      "step": 6925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8364225548664053,
+      "learning_rate": 8.45567902989802e-06,
+      "loss": 0.4822,
+      "step": 6926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8802465512047095,
+      "learning_rate": 8.453078191668806e-06,
+      "loss": 0.4854,
+      "step": 6927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9392464025402688,
+      "learning_rate": 8.450477460639898e-06,
+      "loss": 0.5663,
+      "step": 6928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9006510750755666,
+      "learning_rate": 8.447876836991531e-06,
+      "loss": 0.5447,
+      "step": 6929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8320821905591601,
+      "learning_rate": 8.445276320903922e-06,
+      "loss": 0.4939,
+      "step": 6930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8839694642333762,
+      "learning_rate": 8.442675912557281e-06,
+      "loss": 0.506,
+      "step": 6931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9395330291731121,
+      "learning_rate": 8.440075612131823e-06,
+      "loss": 0.5072,
+      "step": 6932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8635009440795924,
+      "learning_rate": 8.437475419807742e-06,
+      "loss": 0.5385,
+      "step": 6933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9849878980592961,
+      "learning_rate": 8.434875335765222e-06,
+      "loss": 0.6095,
+      "step": 6934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9255841893242462,
+      "learning_rate": 8.432275360184458e-06,
+      "loss": 0.5174,
+      "step": 6935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9090364558178271,
+      "learning_rate": 8.42967549324562e-06,
+      "loss": 0.5509,
+      "step": 6936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0063817065902216,
+      "learning_rate": 8.427075735128874e-06,
+      "loss": 0.5363,
+      "step": 6937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9496204700450661,
+      "learning_rate": 8.42447608601439e-06,
+      "loss": 0.4875,
+      "step": 6938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7609053421255205,
+      "learning_rate": 8.421876546082315e-06,
+      "loss": 0.4664,
+      "step": 6939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.915322118497608,
+      "learning_rate": 8.419277115512791e-06,
+      "loss": 0.5448,
+      "step": 6940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9743354645399194,
+      "learning_rate": 8.416677794485965e-06,
+      "loss": 0.483,
+      "step": 6941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8938767929650255,
+      "learning_rate": 8.414078583181963e-06,
+      "loss": 0.4847,
+      "step": 6942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8624235874313937,
+      "learning_rate": 8.411479481780904e-06,
+      "loss": 0.4361,
+      "step": 6943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8460237552503656,
+      "learning_rate": 8.408880490462914e-06,
+      "loss": 0.5013,
+      "step": 6944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9505478379663955,
+      "learning_rate": 8.406281609408094e-06,
+      "loss": 0.5043,
+      "step": 6945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0443733816143181,
+      "learning_rate": 8.403682838796539e-06,
+      "loss": 0.6496,
+      "step": 6946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8187946528356171,
+      "learning_rate": 8.401084178808353e-06,
+      "loss": 0.529,
+      "step": 6947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0118247077047577,
+      "learning_rate": 8.398485629623613e-06,
+      "loss": 0.6163,
+      "step": 6948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9438440213096889,
+      "learning_rate": 8.395887191422397e-06,
+      "loss": 0.5318,
+      "step": 6949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8133948811582256,
+      "learning_rate": 8.393288864384778e-06,
+      "loss": 0.491,
+      "step": 6950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8383840381164659,
+      "learning_rate": 8.390690648690818e-06,
+      "loss": 0.4626,
+      "step": 6951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9563077923567581,
+      "learning_rate": 8.388092544520563e-06,
+      "loss": 0.4867,
+      "step": 6952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9220258557852953,
+      "learning_rate": 8.385494552054069e-06,
+      "loss": 0.4902,
+      "step": 6953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8819023030230523,
+      "learning_rate": 8.38289667147137e-06,
+      "loss": 0.5122,
+      "step": 6954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9896701536921275,
+      "learning_rate": 8.380298902952493e-06,
+      "loss": 0.4423,
+      "step": 6955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8947119579675652,
+      "learning_rate": 8.37770124667747e-06,
+      "loss": 0.4862,
+      "step": 6956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.884022859472568,
+      "learning_rate": 8.375103702826313e-06,
+      "loss": 0.5322,
+      "step": 6957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8643526429122288,
+      "learning_rate": 8.372506271579022e-06,
+      "loss": 0.4419,
+      "step": 6958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.836333281072999,
+      "learning_rate": 8.369908953115609e-06,
+      "loss": 0.4877,
+      "step": 6959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8772669558464992,
+      "learning_rate": 8.367311747616057e-06,
+      "loss": 0.5038,
+      "step": 6960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.902382786223706,
+      "learning_rate": 8.36471465526035e-06,
+      "loss": 0.5299,
+      "step": 6961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7776272809054607,
+      "learning_rate": 8.362117676228471e-06,
+      "loss": 0.4358,
+      "step": 6962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8457410241726842,
+      "learning_rate": 8.359520810700384e-06,
+      "loss": 0.5294,
+      "step": 6963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8787627448587005,
+      "learning_rate": 8.356924058856046e-06,
+      "loss": 0.5382,
+      "step": 6964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8641307163943069,
+      "learning_rate": 8.354327420875416e-06,
+      "loss": 0.5088,
+      "step": 6965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9338969195084026,
+      "learning_rate": 8.351730896938438e-06,
+      "loss": 0.5405,
+      "step": 6966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9571853366920003,
+      "learning_rate": 8.349134487225041e-06,
+      "loss": 0.5288,
+      "step": 6967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9484888179387783,
+      "learning_rate": 8.346538191915166e-06,
+      "loss": 0.5038,
+      "step": 6968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9092391023046498,
+      "learning_rate": 8.343942011188726e-06,
+      "loss": 0.5155,
+      "step": 6969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9628755470864837,
+      "learning_rate": 8.341345945225632e-06,
+      "loss": 0.5141,
+      "step": 6970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8925177180147768,
+      "learning_rate": 8.338749994205797e-06,
+      "loss": 0.5277,
+      "step": 6971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8013471751031944,
+      "learning_rate": 8.336154158309114e-06,
+      "loss": 0.4862,
+      "step": 6972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9451259573313375,
+      "learning_rate": 8.333558437715468e-06,
+      "loss": 0.6033,
+      "step": 6973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9367395661597856,
+      "learning_rate": 8.330962832604747e-06,
+      "loss": 0.5249,
+      "step": 6974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9432032175641137,
+      "learning_rate": 8.328367343156823e-06,
+      "loss": 0.5391,
+      "step": 6975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8479723797015235,
+      "learning_rate": 8.325771969551553e-06,
+      "loss": 0.5147,
+      "step": 6976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.893248762027732,
+      "learning_rate": 8.323176711968807e-06,
+      "loss": 0.5,
+      "step": 6977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8770357987840728,
+      "learning_rate": 8.320581570588426e-06,
+      "loss": 0.5198,
+      "step": 6978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9602752488044517,
+      "learning_rate": 8.31798654559025e-06,
+      "loss": 0.4744,
+      "step": 6979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9535097302063359,
+      "learning_rate": 8.315391637154116e-06,
+      "loss": 0.4729,
+      "step": 6980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9705726891103388,
+      "learning_rate": 8.31279684545985e-06,
+      "loss": 0.5509,
+      "step": 6981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8861562625840504,
+      "learning_rate": 8.31020217068726e-06,
+      "loss": 0.5101,
+      "step": 6982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9006660475549908,
+      "learning_rate": 8.307607613016166e-06,
+      "loss": 0.505,
+      "step": 6983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9894974040605742,
+      "learning_rate": 8.305013172626363e-06,
+      "loss": 0.4981,
+      "step": 6984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8933836425811029,
+      "learning_rate": 8.30241884969764e-06,
+      "loss": 0.5277,
+      "step": 6985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8515700939920732,
+      "learning_rate": 8.299824644409787e-06,
+      "loss": 0.521,
+      "step": 6986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.941090038946153,
+      "learning_rate": 8.29723055694258e-06,
+      "loss": 0.5387,
+      "step": 6987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.896364200664156,
+      "learning_rate": 8.294636587475781e-06,
+      "loss": 0.609,
+      "step": 6988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.03393686900745,
+      "learning_rate": 8.292042736189156e-06,
+      "loss": 0.5685,
+      "step": 6989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9231576064543081,
+      "learning_rate": 8.289449003262457e-06,
+      "loss": 0.5319,
+      "step": 6990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9946567275629944,
+      "learning_rate": 8.286855388875418e-06,
+      "loss": 0.5518,
+      "step": 6991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.886001255462769,
+      "learning_rate": 8.284261893207788e-06,
+      "loss": 0.5029,
+      "step": 6992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9284596693800119,
+      "learning_rate": 8.281668516439286e-06,
+      "loss": 0.5258,
+      "step": 6993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8278539413461194,
+      "learning_rate": 8.279075258749627e-06,
+      "loss": 0.4514,
+      "step": 6994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9555577959427803,
+      "learning_rate": 8.276482120318532e-06,
+      "loss": 0.6038,
+      "step": 6995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8387302642220866,
+      "learning_rate": 8.273889101325693e-06,
+      "loss": 0.4807,
+      "step": 6996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8962507721279136,
+      "learning_rate": 8.271296201950809e-06,
+      "loss": 0.5234,
+      "step": 6997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8712383728560765,
+      "learning_rate": 8.268703422373564e-06,
+      "loss": 0.4448,
+      "step": 6998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0123153230611763,
+      "learning_rate": 8.266110762773638e-06,
+      "loss": 0.643,
+      "step": 6999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.935456479745211,
+      "learning_rate": 8.263518223330698e-06,
+      "loss": 0.4819,
+      "step": 7000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8840558016114327,
+      "learning_rate": 8.260925804224406e-06,
+      "loss": 0.5238,
+      "step": 7001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9027702652673076,
+      "learning_rate": 8.25833350563441e-06,
+      "loss": 0.4374,
+      "step": 7002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9705104517200848,
+      "learning_rate": 8.25574132774036e-06,
+      "loss": 0.5247,
+      "step": 7003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.93427552403838,
+      "learning_rate": 8.253149270721889e-06,
+      "loss": 0.5646,
+      "step": 7004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9734270631354048,
+      "learning_rate": 8.250557334758623e-06,
+      "loss": 0.5581,
+      "step": 7005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8741399160481841,
+      "learning_rate": 8.24796552003018e-06,
+      "loss": 0.5323,
+      "step": 7006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9273933255218394,
+      "learning_rate": 8.245373826716176e-06,
+      "loss": 0.4716,
+      "step": 7007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8434637395989865,
+      "learning_rate": 8.242782254996207e-06,
+      "loss": 0.5463,
+      "step": 7008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9336333731635436,
+      "learning_rate": 8.240190805049868e-06,
+      "loss": 0.548,
+      "step": 7009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8929443364357782,
+      "learning_rate": 8.23759947705675e-06,
+      "loss": 0.5514,
+      "step": 7010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9201455293723961,
+      "learning_rate": 8.235008271196421e-06,
+      "loss": 0.559,
+      "step": 7011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9102701414183574,
+      "learning_rate": 8.232417187648454e-06,
+      "loss": 0.509,
+      "step": 7012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9264031657521252,
+      "learning_rate": 8.22982622659241e-06,
+      "loss": 0.5409,
+      "step": 7013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0630764564563944,
+      "learning_rate": 8.227235388207835e-06,
+      "loss": 0.4508,
+      "step": 7014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8822532387155293,
+      "learning_rate": 8.224644672674276e-06,
+      "loss": 0.5393,
+      "step": 7015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9174219199288024,
+      "learning_rate": 8.222054080171267e-06,
+      "loss": 0.5642,
+      "step": 7016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.864390451108849,
+      "learning_rate": 8.219463610878336e-06,
+      "loss": 0.4691,
+      "step": 7017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8587854032206618,
+      "learning_rate": 8.216873264974993e-06,
+      "loss": 0.4795,
+      "step": 7018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9363362878742622,
+      "learning_rate": 8.214283042640752e-06,
+      "loss": 0.5474,
+      "step": 7019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9144074402642741,
+      "learning_rate": 8.211692944055116e-06,
+      "loss": 0.5136,
+      "step": 7020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.032086763222279,
+      "learning_rate": 8.20910296939757e-06,
+      "loss": 0.4673,
+      "step": 7021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.917707524800078,
+      "learning_rate": 8.206513118847598e-06,
+      "loss": 0.5868,
+      "step": 7022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.922625636653506,
+      "learning_rate": 8.203923392584676e-06,
+      "loss": 0.4888,
+      "step": 7023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8764509529934924,
+      "learning_rate": 8.201333790788275e-06,
+      "loss": 0.5328,
+      "step": 7024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.859695349317226,
+      "learning_rate": 8.198744313637842e-06,
+      "loss": 0.5474,
+      "step": 7025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0344486716589807,
+      "learning_rate": 8.19615496131283e-06,
+      "loss": 0.5053,
+      "step": 7026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8837125709121174,
+      "learning_rate": 8.193565733992684e-06,
+      "loss": 0.5018,
+      "step": 7027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8819851186881112,
+      "learning_rate": 8.190976631856827e-06,
+      "loss": 0.5101,
+      "step": 7028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.798697172712139,
+      "learning_rate": 8.188387655084684e-06,
+      "loss": 0.4678,
+      "step": 7029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8911852719708158,
+      "learning_rate": 8.185798803855671e-06,
+      "loss": 0.5498,
+      "step": 7030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9231967644330206,
+      "learning_rate": 8.183210078349191e-06,
+      "loss": 0.5421,
+      "step": 7031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8819993157677812,
+      "learning_rate": 8.180621478744641e-06,
+      "loss": 0.5196,
+      "step": 7032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9075262050015743,
+      "learning_rate": 8.178033005221412e-06,
+      "loss": 0.4782,
+      "step": 7033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8737630080614599,
+      "learning_rate": 8.175444657958875e-06,
+      "loss": 0.487,
+      "step": 7034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9154320954425816,
+      "learning_rate": 8.172856437136407e-06,
+      "loss": 0.5353,
+      "step": 7035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9871118216349929,
+      "learning_rate": 8.170268342933365e-06,
+      "loss": 0.5067,
+      "step": 7036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8265507952904249,
+      "learning_rate": 8.167680375529108e-06,
+      "loss": 0.531,
+      "step": 7037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9496077446671609,
+      "learning_rate": 8.165092535102972e-06,
+      "loss": 0.5114,
+      "step": 7038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8783081821659814,
+      "learning_rate": 8.162504821834296e-06,
+      "loss": 0.4619,
+      "step": 7039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8717227626487613,
+      "learning_rate": 8.159917235902409e-06,
+      "loss": 0.5163,
+      "step": 7040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8932338295571399,
+      "learning_rate": 8.157329777486623e-06,
+      "loss": 0.5319,
+      "step": 7041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9199211873496858,
+      "learning_rate": 8.154742446766249e-06,
+      "loss": 0.498,
+      "step": 7042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.891227190168023,
+      "learning_rate": 8.15215524392059e-06,
+      "loss": 0.4943,
+      "step": 7043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9412199201992494,
+      "learning_rate": 8.14956816912893e-06,
+      "loss": 0.5129,
+      "step": 7044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8556676533700766,
+      "learning_rate": 8.146981222570553e-06,
+      "loss": 0.5195,
+      "step": 7045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9710161830339895,
+      "learning_rate": 8.144394404424739e-06,
+      "loss": 0.5701,
+      "step": 7046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9812013727215241,
+      "learning_rate": 8.141807714870743e-06,
+      "loss": 0.5025,
+      "step": 7047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6689600561603462,
+      "learning_rate": 8.139221154087825e-06,
+      "loss": 0.4416,
+      "step": 7048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8059420540732499,
+      "learning_rate": 8.136634722255232e-06,
+      "loss": 0.4449,
+      "step": 7049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9333593303331403,
+      "learning_rate": 8.134048419552197e-06,
+      "loss": 0.568,
+      "step": 7050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8472084290656892,
+      "learning_rate": 8.131462246157953e-06,
+      "loss": 0.5321,
+      "step": 7051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9126154074678705,
+      "learning_rate": 8.128876202251719e-06,
+      "loss": 0.5109,
+      "step": 7052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8763009264392355,
+      "learning_rate": 8.126290288012701e-06,
+      "loss": 0.5216,
+      "step": 7053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8264540746375304,
+      "learning_rate": 8.123704503620107e-06,
+      "loss": 0.485,
+      "step": 7054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8936189165840086,
+      "learning_rate": 8.121118849253127e-06,
+      "loss": 0.5393,
+      "step": 7055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8849486877355847,
+      "learning_rate": 8.11853332509094e-06,
+      "loss": 0.5256,
+      "step": 7056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.874095324515099,
+      "learning_rate": 8.11594793131273e-06,
+      "loss": 0.5495,
+      "step": 7057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8441694949243681,
+      "learning_rate": 8.113362668097658e-06,
+      "loss": 0.4437,
+      "step": 7058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.902798856936396,
+      "learning_rate": 8.110777535624875e-06,
+      "loss": 0.5449,
+      "step": 7059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8255384853373445,
+      "learning_rate": 8.108192534073534e-06,
+      "loss": 0.5014,
+      "step": 7060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8517134480254044,
+      "learning_rate": 8.105607663622775e-06,
+      "loss": 0.5029,
+      "step": 7061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8067488124368722,
+      "learning_rate": 8.10302292445172e-06,
+      "loss": 0.452,
+      "step": 7062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9532445683932202,
+      "learning_rate": 8.100438316739499e-06,
+      "loss": 0.5633,
+      "step": 7063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8412583157335193,
+      "learning_rate": 8.097853840665217e-06,
+      "loss": 0.4861,
+      "step": 7064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9635981222701807,
+      "learning_rate": 8.095269496407972e-06,
+      "loss": 0.5369,
+      "step": 7065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9723101800664249,
+      "learning_rate": 8.092685284146865e-06,
+      "loss": 0.5529,
+      "step": 7066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.882965623304785,
+      "learning_rate": 8.090101204060977e-06,
+      "loss": 0.548,
+      "step": 7067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8898000120431133,
+      "learning_rate": 8.087517256329376e-06,
+      "loss": 0.5369,
+      "step": 7068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9130973916070677,
+      "learning_rate": 8.08493344113114e-06,
+      "loss": 0.557,
+      "step": 7069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8234264567707733,
+      "learning_rate": 8.082349758645316e-06,
+      "loss": 0.525,
+      "step": 7070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7937641653928239,
+      "learning_rate": 8.079766209050947e-06,
+      "loss": 0.4841,
+      "step": 7071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.875824564412514,
+      "learning_rate": 8.077182792527082e-06,
+      "loss": 0.4847,
+      "step": 7072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9159490950540292,
+      "learning_rate": 8.074599509252745e-06,
+      "loss": 0.5131,
+      "step": 7073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8557200095988629,
+      "learning_rate": 8.072016359406949e-06,
+      "loss": 0.511,
+      "step": 7074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.860909775478447,
+      "learning_rate": 8.069433343168713e-06,
+      "loss": 0.4982,
+      "step": 7075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9084085645848757,
+      "learning_rate": 8.066850460717035e-06,
+      "loss": 0.4887,
+      "step": 7076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9552739657622884,
+      "learning_rate": 8.0642677122309e-06,
+      "loss": 0.5326,
+      "step": 7077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8301615599182511,
+      "learning_rate": 8.0616850978893e-06,
+      "loss": 0.4942,
+      "step": 7078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8730009479479265,
+      "learning_rate": 8.059102617871203e-06,
+      "loss": 0.4616,
+      "step": 7079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9619992822577931,
+      "learning_rate": 8.056520272355571e-06,
+      "loss": 0.492,
+      "step": 7080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9328794065066321,
+      "learning_rate": 8.053938061521363e-06,
+      "loss": 0.5819,
+      "step": 7081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8940979839317937,
+      "learning_rate": 8.05135598554752e-06,
+      "loss": 0.5396,
+      "step": 7082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9114092619314533,
+      "learning_rate": 8.048774044612977e-06,
+      "loss": 0.4878,
+      "step": 7083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9038805767427703,
+      "learning_rate": 8.046192238896665e-06,
+      "loss": 0.5246,
+      "step": 7084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9281520876936268,
+      "learning_rate": 8.043610568577497e-06,
+      "loss": 0.5246,
+      "step": 7085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8755547684924566,
+      "learning_rate": 8.041029033834378e-06,
+      "loss": 0.5387,
+      "step": 7086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0413527560484352,
+      "learning_rate": 8.038447634846214e-06,
+      "loss": 0.4924,
+      "step": 7087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9365148241845588,
+      "learning_rate": 8.035866371791889e-06,
+      "loss": 0.5573,
+      "step": 7088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8583013464946238,
+      "learning_rate": 8.033285244850276e-06,
+      "loss": 0.4931,
+      "step": 7089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9798987553897932,
+      "learning_rate": 8.030704254200256e-06,
+      "loss": 0.5488,
+      "step": 7090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8481462707609014,
+      "learning_rate": 8.028123400020686e-06,
+      "loss": 0.5436,
+      "step": 7091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8908057979277204,
+      "learning_rate": 8.02554268249041e-06,
+      "loss": 0.5143,
+      "step": 7092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9223070040189832,
+      "learning_rate": 8.02296210178828e-06,
+      "loss": 0.5308,
+      "step": 7093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8816963595479083,
+      "learning_rate": 8.02038165809312e-06,
+      "loss": 0.5327,
+      "step": 7094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8929250749843264,
+      "learning_rate": 8.017801351583753e-06,
+      "loss": 0.4613,
+      "step": 7095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7983790901220064,
+      "learning_rate": 8.015221182439e-06,
+      "loss": 0.4302,
+      "step": 7096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8384414175374417,
+      "learning_rate": 8.012641150837656e-06,
+      "loss": 0.431,
+      "step": 7097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8930745141481806,
+      "learning_rate": 8.010061256958515e-06,
+      "loss": 0.517,
+      "step": 7098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9267017949110046,
+      "learning_rate": 8.007481500980366e-06,
+      "loss": 0.4581,
+      "step": 7099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8493740892753433,
+      "learning_rate": 8.004901883081983e-06,
+      "loss": 0.4934,
+      "step": 7100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.885590654760843,
+      "learning_rate": 8.002322403442125e-06,
+      "loss": 0.5524,
+      "step": 7101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9324619489542131,
+      "learning_rate": 7.999743062239557e-06,
+      "loss": 0.5681,
+      "step": 7102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9608122058291328,
+      "learning_rate": 7.99716385965302e-06,
+      "loss": 0.5449,
+      "step": 7103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8549468729838979,
+      "learning_rate": 7.994584795861248e-06,
+      "loss": 0.5272,
+      "step": 7104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9065242963654259,
+      "learning_rate": 7.992005871042975e-06,
+      "loss": 0.5188,
+      "step": 7105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8642241252773025,
+      "learning_rate": 7.989427085376914e-06,
+      "loss": 0.4868,
+      "step": 7106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8191305164843538,
+      "learning_rate": 7.986848439041767e-06,
+      "loss": 0.5231,
+      "step": 7107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9345831959630203,
+      "learning_rate": 7.984269932216241e-06,
+      "loss": 0.5122,
+      "step": 7108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0591280413854145,
+      "learning_rate": 7.981691565079024e-06,
+      "loss": 0.5526,
+      "step": 7109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0004936234543034,
+      "learning_rate": 7.979113337808786e-06,
+      "loss": 0.4942,
+      "step": 7110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.816574338598874,
+      "learning_rate": 7.976535250584204e-06,
+      "loss": 0.5098,
+      "step": 7111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8664437742943795,
+      "learning_rate": 7.973957303583936e-06,
+      "loss": 0.5368,
+      "step": 7112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.91862171457887,
+      "learning_rate": 7.971379496986625e-06,
+      "loss": 0.5826,
+      "step": 7113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9417244042887671,
+      "learning_rate": 7.968801830970917e-06,
+      "loss": 0.5015,
+      "step": 7114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9634260358824619,
+      "learning_rate": 7.966224305715443e-06,
+      "loss": 0.5266,
+      "step": 7115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8893242016477338,
+      "learning_rate": 7.963646921398818e-06,
+      "loss": 0.4669,
+      "step": 7116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9680052559082771,
+      "learning_rate": 7.961069678199658e-06,
+      "loss": 0.5176,
+      "step": 7117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8865734175274613,
+      "learning_rate": 7.95849257629656e-06,
+      "loss": 0.5046,
+      "step": 7118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9284345145026058,
+      "learning_rate": 7.95591561586811e-06,
+      "loss": 0.5279,
+      "step": 7119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8619392433080499,
+      "learning_rate": 7.953338797092902e-06,
+      "loss": 0.4835,
+      "step": 7120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0180767658901928,
+      "learning_rate": 7.950762120149499e-06,
+      "loss": 0.5636,
+      "step": 7121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8686763777204012,
+      "learning_rate": 7.94818558521646e-06,
+      "loss": 0.5189,
+      "step": 7122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.93918029784971,
+      "learning_rate": 7.94560919247234e-06,
+      "loss": 0.4968,
+      "step": 7123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9442326161843043,
+      "learning_rate": 7.943032942095685e-06,
+      "loss": 0.5392,
+      "step": 7124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9935891972254188,
+      "learning_rate": 7.940456834265017e-06,
+      "loss": 0.5106,
+      "step": 7125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8666080937052979,
+      "learning_rate": 7.937880869158868e-06,
+      "loss": 0.4962,
+      "step": 7126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.849261195585096,
+      "learning_rate": 7.935305046955746e-06,
+      "loss": 0.5083,
+      "step": 7127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8464975330500006,
+      "learning_rate": 7.93272936783415e-06,
+      "loss": 0.5107,
+      "step": 7128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9556219695413329,
+      "learning_rate": 7.930153831972575e-06,
+      "loss": 0.5619,
+      "step": 7129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9412515314641338,
+      "learning_rate": 7.927578439549506e-06,
+      "loss": 0.5491,
+      "step": 7130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8079216989915801,
+      "learning_rate": 7.92500319074341e-06,
+      "loss": 0.5057,
+      "step": 7131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9025189985626063,
+      "learning_rate": 7.922428085732755e-06,
+      "loss": 0.5127,
+      "step": 7132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9482220895231857,
+      "learning_rate": 7.919853124695993e-06,
+      "loss": 0.4783,
+      "step": 7133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8388943639401142,
+      "learning_rate": 7.917278307811557e-06,
+      "loss": 0.4902,
+      "step": 7134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8968964756970629,
+      "learning_rate": 7.914703635257892e-06,
+      "loss": 0.4887,
+      "step": 7135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8833263700487893,
+      "learning_rate": 7.912129107213417e-06,
+      "loss": 0.5087,
+      "step": 7136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8963174813537905,
+      "learning_rate": 7.909554723856537e-06,
+      "loss": 0.5032,
+      "step": 7137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9074153912009865,
+      "learning_rate": 7.906980485365665e-06,
+      "loss": 0.5232,
+      "step": 7138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8283767399063273,
+      "learning_rate": 7.90440639191919e-06,
+      "loss": 0.4602,
+      "step": 7139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8903407901604158,
+      "learning_rate": 7.901832443695487e-06,
+      "loss": 0.5257,
+      "step": 7140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9091972597395084,
+      "learning_rate": 7.899258640872942e-06,
+      "loss": 0.5404,
+      "step": 7141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0976786999602899,
+      "learning_rate": 7.896684983629907e-06,
+      "loss": 0.5162,
+      "step": 7142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8003995289691711,
+      "learning_rate": 7.894111472144733e-06,
+      "loss": 0.4649,
+      "step": 7143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8420642146592251,
+      "learning_rate": 7.89153810659577e-06,
+      "loss": 0.5088,
+      "step": 7144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0426511293723346,
+      "learning_rate": 7.888964887161348e-06,
+      "loss": 0.576,
+      "step": 7145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9027421728823426,
+      "learning_rate": 7.886391814019782e-06,
+      "loss": 0.5263,
+      "step": 7146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9367759837162823,
+      "learning_rate": 7.883818887349391e-06,
+      "loss": 0.5385,
+      "step": 7147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8860146610850669,
+      "learning_rate": 7.881246107328472e-06,
+      "loss": 0.4902,
+      "step": 7148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9387255568652029,
+      "learning_rate": 7.878673474135321e-06,
+      "loss": 0.5192,
+      "step": 7149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9864023313146227,
+      "learning_rate": 7.876100987948217e-06,
+      "loss": 0.588,
+      "step": 7150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.994880359320626,
+      "learning_rate": 7.873528648945428e-06,
+      "loss": 0.5507,
+      "step": 7151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8755218895848427,
+      "learning_rate": 7.87095645730522e-06,
+      "loss": 0.5207,
+      "step": 7152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8424039240765442,
+      "learning_rate": 7.868384413205842e-06,
+      "loss": 0.5149,
+      "step": 7153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9298680000054024,
+      "learning_rate": 7.865812516825528e-06,
+      "loss": 0.4513,
+      "step": 7154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9378542402912848,
+      "learning_rate": 7.863240768342518e-06,
+      "loss": 0.56,
+      "step": 7155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.941553399591606,
+      "learning_rate": 7.860669167935028e-06,
+      "loss": 0.5182,
+      "step": 7156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9906634051079232,
+      "learning_rate": 7.858097715781264e-06,
+      "loss": 0.5827,
+      "step": 7157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8681871023458261,
+      "learning_rate": 7.85552641205943e-06,
+      "loss": 0.5365,
+      "step": 7158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8604262563644228,
+      "learning_rate": 7.852955256947717e-06,
+      "loss": 0.4941,
+      "step": 7159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9035245788633994,
+      "learning_rate": 7.850384250624293e-06,
+      "loss": 0.4967,
+      "step": 7160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9448437606155281,
+      "learning_rate": 7.847813393267338e-06,
+      "loss": 0.5439,
+      "step": 7161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8518107595401718,
+      "learning_rate": 7.845242685055008e-06,
+      "loss": 0.4615,
+      "step": 7162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9074697637525523,
+      "learning_rate": 7.842672126165443e-06,
+      "loss": 0.545,
+      "step": 7163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8956320738778516,
+      "learning_rate": 7.84010171677679e-06,
+      "loss": 0.5295,
+      "step": 7164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9930685920080908,
+      "learning_rate": 7.837531457067171e-06,
+      "loss": 0.5653,
+      "step": 7165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9672444191250434,
+      "learning_rate": 7.834961347214704e-06,
+      "loss": 0.496,
+      "step": 7166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9470240702819805,
+      "learning_rate": 7.832391387397495e-06,
+      "loss": 0.5262,
+      "step": 7167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.890014574596134,
+      "learning_rate": 7.829821577793642e-06,
+      "loss": 0.5246,
+      "step": 7168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9901783082879074,
+      "learning_rate": 7.827251918581225e-06,
+      "loss": 0.6147,
+      "step": 7169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8810342910057687,
+      "learning_rate": 7.824682409938328e-06,
+      "loss": 0.5329,
+      "step": 7170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9177398262818202,
+      "learning_rate": 7.822113052043007e-06,
+      "loss": 0.5194,
+      "step": 7171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9514360356116424,
+      "learning_rate": 7.819543845073319e-06,
+      "loss": 0.5121,
+      "step": 7172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9172146908279984,
+      "learning_rate": 7.816974789207311e-06,
+      "loss": 0.4944,
+      "step": 7173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8682178687211466,
+      "learning_rate": 7.814405884623012e-06,
+      "loss": 0.5039,
+      "step": 7174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8708333791973569,
+      "learning_rate": 7.811837131498448e-06,
+      "loss": 0.5385,
+      "step": 7175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8474069428549672,
+      "learning_rate": 7.80926853001163e-06,
+      "loss": 0.4809,
+      "step": 7176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9659314208908955,
+      "learning_rate": 7.806700080340558e-06,
+      "loss": 0.5581,
+      "step": 7177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0070230808243348,
+      "learning_rate": 7.804131782663224e-06,
+      "loss": 0.5707,
+      "step": 7178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9345471374137605,
+      "learning_rate": 7.801563637157614e-06,
+      "loss": 0.5631,
+      "step": 7179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8770225209942285,
+      "learning_rate": 7.79899564400169e-06,
+      "loss": 0.5538,
+      "step": 7180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9830268920995888,
+      "learning_rate": 7.796427803373416e-06,
+      "loss": 0.5664,
+      "step": 7181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9248689772553825,
+      "learning_rate": 7.793860115450744e-06,
+      "loss": 0.5054,
+      "step": 7182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9004370145255842,
+      "learning_rate": 7.791292580411606e-06,
+      "loss": 0.4964,
+      "step": 7183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8716515992647562,
+      "learning_rate": 7.788725198433933e-06,
+      "loss": 0.4919,
+      "step": 7184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8169934980297398,
+      "learning_rate": 7.786157969695643e-06,
+      "loss": 0.4821,
+      "step": 7185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8647980879138384,
+      "learning_rate": 7.783590894374642e-06,
+      "loss": 0.5286,
+      "step": 7186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8614129816503232,
+      "learning_rate": 7.781023972648826e-06,
+      "loss": 0.4908,
+      "step": 7187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9324921763996864,
+      "learning_rate": 7.778457204696082e-06,
+      "loss": 0.5547,
+      "step": 7188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9111081523908714,
+      "learning_rate": 7.775890590694283e-06,
+      "loss": 0.4669,
+      "step": 7189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9610565102640668,
+      "learning_rate": 7.77332413082129e-06,
+      "loss": 0.5295,
+      "step": 7190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9293235516497907,
+      "learning_rate": 7.770757825254962e-06,
+      "loss": 0.5041,
+      "step": 7191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9121050505057278,
+      "learning_rate": 7.768191674173142e-06,
+      "loss": 0.5348,
+      "step": 7192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7854722354449046,
+      "learning_rate": 7.765625677753656e-06,
+      "loss": 0.4922,
+      "step": 7193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9010995980188043,
+      "learning_rate": 7.763059836174329e-06,
+      "loss": 0.5002,
+      "step": 7194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8549641547968623,
+      "learning_rate": 7.760494149612971e-06,
+      "loss": 0.4898,
+      "step": 7195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8698049730542157,
+      "learning_rate": 7.757928618247384e-06,
+      "loss": 0.4968,
+      "step": 7196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9109125156970205,
+      "learning_rate": 7.755363242255352e-06,
+      "loss": 0.5399,
+      "step": 7197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9878217858400418,
+      "learning_rate": 7.752798021814659e-06,
+      "loss": 0.56,
+      "step": 7198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9468859034587981,
+      "learning_rate": 7.750232957103068e-06,
+      "loss": 0.4919,
+      "step": 7199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9303700707294289,
+      "learning_rate": 7.747668048298338e-06,
+      "loss": 0.572,
+      "step": 7200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9098675738960599,
+      "learning_rate": 7.745103295578216e-06,
+      "loss": 0.5466,
+      "step": 7201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9343802851406701,
+      "learning_rate": 7.74253869912043e-06,
+      "loss": 0.5386,
+      "step": 7202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9476829141758246,
+      "learning_rate": 7.739974259102716e-06,
+      "loss": 0.5189,
+      "step": 7203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7783194872808666,
+      "learning_rate": 7.73740997570278e-06,
+      "loss": 0.4538,
+      "step": 7204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1009354505492426,
+      "learning_rate": 7.734845849098324e-06,
+      "loss": 0.5546,
+      "step": 7205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9266294852249874,
+      "learning_rate": 7.732281879467043e-06,
+      "loss": 0.5376,
+      "step": 7206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8711521298481469,
+      "learning_rate": 7.729718066986617e-06,
+      "loss": 0.496,
+      "step": 7207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8892162259369395,
+      "learning_rate": 7.727154411834712e-06,
+      "loss": 0.483,
+      "step": 7208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9028078827519032,
+      "learning_rate": 7.724590914188994e-06,
+      "loss": 0.5044,
+      "step": 7209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9815763452473252,
+      "learning_rate": 7.722027574227107e-06,
+      "loss": 0.5594,
+      "step": 7210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8967294073556648,
+      "learning_rate": 7.719464392126684e-06,
+      "loss": 0.5169,
+      "step": 7211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9065002695688444,
+      "learning_rate": 7.71690136806536e-06,
+      "loss": 0.5255,
+      "step": 7212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9222558351013072,
+      "learning_rate": 7.714338502220746e-06,
+      "loss": 0.5538,
+      "step": 7213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8671006891727058,
+      "learning_rate": 7.711775794770443e-06,
+      "loss": 0.5086,
+      "step": 7214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0232282025989965,
+      "learning_rate": 7.709213245892051e-06,
+      "loss": 0.548,
+      "step": 7215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9529100879092692,
+      "learning_rate": 7.70665085576315e-06,
+      "loss": 0.5342,
+      "step": 7216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9734892020915824,
+      "learning_rate": 7.704088624561306e-06,
+      "loss": 0.5127,
+      "step": 7217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9270597288383303,
+      "learning_rate": 7.701526552464087e-06,
+      "loss": 0.5173,
+      "step": 7218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8720870889009098,
+      "learning_rate": 7.698964639649041e-06,
+      "loss": 0.4809,
+      "step": 7219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.923839288776839,
+      "learning_rate": 7.6964028862937e-06,
+      "loss": 0.5046,
+      "step": 7220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.02966338890196,
+      "learning_rate": 7.6938412925756e-06,
+      "loss": 0.5881,
+      "step": 7221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9594420898093694,
+      "learning_rate": 7.691279858672252e-06,
+      "loss": 0.5048,
+      "step": 7222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9115683816055996,
+      "learning_rate": 7.688718584761158e-06,
+      "loss": 0.4904,
+      "step": 7223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9054139733472941,
+      "learning_rate": 7.68615747101982e-06,
+      "loss": 0.5241,
+      "step": 7224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7437088104314823,
+      "learning_rate": 7.683596517625716e-06,
+      "loss": 0.4348,
+      "step": 7225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8783371843681202,
+      "learning_rate": 7.681035724756317e-06,
+      "loss": 0.5548,
+      "step": 7226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9830512757805387,
+      "learning_rate": 7.678475092589088e-06,
+      "loss": 0.5338,
+      "step": 7227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8493114744220931,
+      "learning_rate": 7.675914621301476e-06,
+      "loss": 0.4998,
+      "step": 7228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9992033936519503,
+      "learning_rate": 7.673354311070914e-06,
+      "loss": 0.5458,
+      "step": 7229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8176998664379581,
+      "learning_rate": 7.67079416207484e-06,
+      "loss": 0.4448,
+      "step": 7230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8844593418859039,
+      "learning_rate": 7.668234174490664e-06,
+      "loss": 0.508,
+      "step": 7231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8417371432578538,
+      "learning_rate": 7.665674348495788e-06,
+      "loss": 0.4771,
+      "step": 7232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9956057436389267,
+      "learning_rate": 7.663114684267612e-06,
+      "loss": 0.5216,
+      "step": 7233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9428429486313148,
+      "learning_rate": 7.660555181983517e-06,
+      "loss": 0.4689,
+      "step": 7234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9375806767617715,
+      "learning_rate": 7.657995841820869e-06,
+      "loss": 0.5296,
+      "step": 7235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8450786498494435,
+      "learning_rate": 7.655436663957035e-06,
+      "loss": 0.4782,
+      "step": 7236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9589117881742945,
+      "learning_rate": 7.65287764856936e-06,
+      "loss": 0.5631,
+      "step": 7237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9600384970168856,
+      "learning_rate": 7.650318795835179e-06,
+      "loss": 0.5248,
+      "step": 7238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9766270180046419,
+      "learning_rate": 7.647760105931825e-06,
+      "loss": 0.5195,
+      "step": 7239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0108489371038865,
+      "learning_rate": 7.64520157903661e-06,
+      "loss": 0.5331,
+      "step": 7240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9379986049299248,
+      "learning_rate": 7.64264321532683e-06,
+      "loss": 0.4585,
+      "step": 7241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9131980397882025,
+      "learning_rate": 7.640085014979792e-06,
+      "loss": 0.5732,
+      "step": 7242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9529997529146212,
+      "learning_rate": 7.637526978172767e-06,
+      "loss": 0.5263,
+      "step": 7243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8851984451396883,
+      "learning_rate": 7.634969105083023e-06,
+      "loss": 0.5099,
+      "step": 7244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9883433976165381,
+      "learning_rate": 7.632411395887826e-06,
+      "loss": 0.553,
+      "step": 7245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9934441008382426,
+      "learning_rate": 7.62985385076442e-06,
+      "loss": 0.543,
+      "step": 7246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9492112256678701,
+      "learning_rate": 7.6272964698900356e-06,
+      "loss": 0.5126,
+      "step": 7247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8564922898626104,
+      "learning_rate": 7.624739253441905e-06,
+      "loss": 0.5169,
+      "step": 7248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0115163408161787,
+      "learning_rate": 7.622182201597238e-06,
+      "loss": 0.5686,
+      "step": 7249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9392761542496402,
+      "learning_rate": 7.619625314533231e-06,
+      "loss": 0.5365,
+      "step": 7250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8739873176329055,
+      "learning_rate": 7.6170685924270815e-06,
+      "loss": 0.5418,
+      "step": 7251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8558631122985492,
+      "learning_rate": 7.6145120354559666e-06,
+      "loss": 0.5435,
+      "step": 7252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9014633753616382,
+      "learning_rate": 7.611955643797046e-06,
+      "loss": 0.5492,
+      "step": 7253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0011582929357539,
+      "learning_rate": 7.609399417627486e-06,
+      "loss": 0.6088,
+      "step": 7254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.875694028910975,
+      "learning_rate": 7.606843357124426e-06,
+      "loss": 0.534,
+      "step": 7255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8900901373620113,
+      "learning_rate": 7.604287462464995e-06,
+      "loss": 0.5299,
+      "step": 7256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8280069475864995,
+      "learning_rate": 7.60173173382632e-06,
+      "loss": 0.4966,
+      "step": 7257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.880686332922401,
+      "learning_rate": 7.599176171385509e-06,
+      "loss": 0.5073,
+      "step": 7258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8840388220644066,
+      "learning_rate": 7.5966207753196574e-06,
+      "loss": 0.5394,
+      "step": 7259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8961429378494427,
+      "learning_rate": 7.5940655458058575e-06,
+      "loss": 0.4807,
+      "step": 7260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9821194016091872,
+      "learning_rate": 7.59151048302118e-06,
+      "loss": 0.5458,
+      "step": 7261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9752484428211466,
+      "learning_rate": 7.588955587142688e-06,
+      "loss": 0.516,
+      "step": 7262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9843824414210166,
+      "learning_rate": 7.586400858347438e-06,
+      "loss": 0.5626,
+      "step": 7263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8707842938023541,
+      "learning_rate": 7.583846296812467e-06,
+      "loss": 0.4788,
+      "step": 7264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8926431378262193,
+      "learning_rate": 7.581291902714801e-06,
+      "loss": 0.4816,
+      "step": 7265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8237824939681814,
+      "learning_rate": 7.578737676231466e-06,
+      "loss": 0.508,
+      "step": 7266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9437865429552902,
+      "learning_rate": 7.576183617539461e-06,
+      "loss": 0.5317,
+      "step": 7267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9322937825461017,
+      "learning_rate": 7.573629726815778e-06,
+      "loss": 0.5568,
+      "step": 7268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8211316207432359,
+      "learning_rate": 7.5710760042374056e-06,
+      "loss": 0.4647,
+      "step": 7269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8637140877102417,
+      "learning_rate": 7.568522449981308e-06,
+      "loss": 0.4874,
+      "step": 7270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8855791028344088,
+      "learning_rate": 7.565969064224453e-06,
+      "loss": 0.5189,
+      "step": 7271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.014135661037069,
+      "learning_rate": 7.563415847143782e-06,
+      "loss": 0.5775,
+      "step": 7272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9497931596687522,
+      "learning_rate": 7.560862798916229e-06,
+      "loss": 0.5442,
+      "step": 7273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9961094952611982,
+      "learning_rate": 7.558309919718723e-06,
+      "loss": 0.5498,
+      "step": 7274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8910273913576999,
+      "learning_rate": 7.555757209728174e-06,
+      "loss": 0.5109,
+      "step": 7275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8401278429671708,
+      "learning_rate": 7.553204669121478e-06,
+      "loss": 0.5287,
+      "step": 7276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0258306091345222,
+      "learning_rate": 7.550652298075532e-06,
+      "loss": 0.5642,
+      "step": 7277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0388928356690446,
+      "learning_rate": 7.54810009676721e-06,
+      "loss": 0.5273,
+      "step": 7278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.894137035048821,
+      "learning_rate": 7.545548065373372e-06,
+      "loss": 0.5418,
+      "step": 7279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8520592226061066,
+      "learning_rate": 7.54299620407088e-06,
+      "loss": 0.5039,
+      "step": 7280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9289674724415184,
+      "learning_rate": 7.540444513036572e-06,
+      "loss": 0.5908,
+      "step": 7281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8688189308375063,
+      "learning_rate": 7.5378929924472735e-06,
+      "loss": 0.5631,
+      "step": 7282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9200999552756037,
+      "learning_rate": 7.535341642479811e-06,
+      "loss": 0.5082,
+      "step": 7283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9210146899679073,
+      "learning_rate": 7.532790463310986e-06,
+      "loss": 0.5396,
+      "step": 7284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9137799113440104,
+      "learning_rate": 7.530239455117589e-06,
+      "loss": 0.5432,
+      "step": 7285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9363505809898554,
+      "learning_rate": 7.527688618076413e-06,
+      "loss": 0.5339,
+      "step": 7286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9763535749511509,
+      "learning_rate": 7.525137952364222e-06,
+      "loss": 0.5225,
+      "step": 7287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8645863285917511,
+      "learning_rate": 7.522587458157771e-06,
+      "loss": 0.518,
+      "step": 7288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8407841151958855,
+      "learning_rate": 7.520037135633817e-06,
+      "loss": 0.4463,
+      "step": 7289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9285178402786097,
+      "learning_rate": 7.517486984969088e-06,
+      "loss": 0.5605,
+      "step": 7290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8945561346406458,
+      "learning_rate": 7.514937006340306e-06,
+      "loss": 0.5643,
+      "step": 7291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8963075832585806,
+      "learning_rate": 7.512387199924189e-06,
+      "loss": 0.5199,
+      "step": 7292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9563988735599971,
+      "learning_rate": 7.509837565897432e-06,
+      "loss": 0.5658,
+      "step": 7293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9423292581170268,
+      "learning_rate": 7.507288104436719e-06,
+      "loss": 0.5352,
+      "step": 7294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.992898315830082,
+      "learning_rate": 7.504738815718734e-06,
+      "loss": 0.5415,
+      "step": 7295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9020731368058453,
+      "learning_rate": 7.502189699920136e-06,
+      "loss": 0.5312,
+      "step": 7296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9363841774035592,
+      "learning_rate": 7.499640757217572e-06,
+      "loss": 0.5127,
+      "step": 7297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8631830942622,
+      "learning_rate": 7.497091987787689e-06,
+      "loss": 0.4652,
+      "step": 7298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9029696039606694,
+      "learning_rate": 7.494543391807112e-06,
+      "loss": 0.469,
+      "step": 7299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0395322395699418,
+      "learning_rate": 7.4919949694524506e-06,
+      "loss": 0.5787,
+      "step": 7300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9576457627425172,
+      "learning_rate": 7.489446720900319e-06,
+      "loss": 0.541,
+      "step": 7301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9116151576314977,
+      "learning_rate": 7.486898646327301e-06,
+      "loss": 0.514,
+      "step": 7302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9491659654888721,
+      "learning_rate": 7.484350745909974e-06,
+      "loss": 0.497,
+      "step": 7303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9257407304664244,
+      "learning_rate": 7.481803019824914e-06,
+      "loss": 0.5212,
+      "step": 7304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.830204046262912,
+      "learning_rate": 7.47925546824867e-06,
+      "loss": 0.5119,
+      "step": 7305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9099262225306468,
+      "learning_rate": 7.476708091357783e-06,
+      "loss": 0.5123,
+      "step": 7306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9658265435756943,
+      "learning_rate": 7.47416088932879e-06,
+      "loss": 0.5334,
+      "step": 7307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9570047689135044,
+      "learning_rate": 7.471613862338207e-06,
+      "loss": 0.5118,
+      "step": 7308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8938157106055836,
+      "learning_rate": 7.469067010562538e-06,
+      "loss": 0.5066,
+      "step": 7309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9216259117988101,
+      "learning_rate": 7.466520334178284e-06,
+      "loss": 0.4809,
+      "step": 7310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9052948098015442,
+      "learning_rate": 7.463973833361923e-06,
+      "loss": 0.5192,
+      "step": 7311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9010973814962012,
+      "learning_rate": 7.461427508289922e-06,
+      "loss": 0.4918,
+      "step": 7312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8954927965727761,
+      "learning_rate": 7.458881359138746e-06,
+      "loss": 0.4777,
+      "step": 7313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9163133645892921,
+      "learning_rate": 7.4563353860848375e-06,
+      "loss": 0.5013,
+      "step": 7314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9584330383366784,
+      "learning_rate": 7.453789589304629e-06,
+      "loss": 0.5731,
+      "step": 7315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8238821401967524,
+      "learning_rate": 7.451243968974547e-06,
+      "loss": 0.4768,
+      "step": 7316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9103939120287774,
+      "learning_rate": 7.448698525270995e-06,
+      "loss": 0.54,
+      "step": 7317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8802397289849867,
+      "learning_rate": 7.446153258370372e-06,
+      "loss": 0.489,
+      "step": 7318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9323534952947754,
+      "learning_rate": 7.443608168449063e-06,
+      "loss": 0.5387,
+      "step": 7319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9041005321484384,
+      "learning_rate": 7.44106325568344e-06,
+      "loss": 0.5025,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9351837814636498,
+      "learning_rate": 7.438518520249865e-06,
+      "loss": 0.5344,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9896753959713007,
+      "learning_rate": 7.435973962324685e-06,
+      "loss": 0.509,
+      "step": 7322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9217534565109436,
+      "learning_rate": 7.433429582084233e-06,
+      "loss": 0.527,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8924443816109849,
+      "learning_rate": 7.4308853797048355e-06,
+      "loss": 0.493,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.790774098801519,
+      "learning_rate": 7.428341355362803e-06,
+      "loss": 0.5064,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.895239464915871,
+      "learning_rate": 7.425797509234433e-06,
+      "loss": 0.5247,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8337451552078113,
+      "learning_rate": 7.423253841496011e-06,
+      "loss": 0.4385,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9388895050134273,
+      "learning_rate": 7.420710352323814e-06,
+      "loss": 0.5099,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8729235534099536,
+      "learning_rate": 7.418167041894101e-06,
+      "loss": 0.5312,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8780973606573884,
+      "learning_rate": 7.415623910383121e-06,
+      "loss": 0.5222,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8900258292874804,
+      "learning_rate": 7.413080957967114e-06,
+      "loss": 0.5311,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8735285883101643,
+      "learning_rate": 7.4105381848223005e-06,
+      "loss": 0.4678,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9108584288618684,
+      "learning_rate": 7.407995591124892e-06,
+      "loss": 0.5694,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8409007811031063,
+      "learning_rate": 7.405453177051092e-06,
+      "loss": 0.5019,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9763197336086359,
+      "learning_rate": 7.4029109427770875e-06,
+      "loss": 0.5829,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8358779487127997,
+      "learning_rate": 7.400368888479048e-06,
+      "loss": 0.4701,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9256093445064898,
+      "learning_rate": 7.3978270143331386e-06,
+      "loss": 0.4672,
+      "step": 7337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8703907207903545,
+      "learning_rate": 7.395285320515513e-06,
+      "loss": 0.4854,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8341471873689289,
+      "learning_rate": 7.392743807202301e-06,
+      "loss": 0.4957,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9060027701674533,
+      "learning_rate": 7.39020247456963e-06,
+      "loss": 0.5042,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8764931059475367,
+      "learning_rate": 7.3876613227936145e-06,
+      "loss": 0.4678,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0120630716408041,
+      "learning_rate": 7.38512035205035e-06,
+      "loss": 0.5496,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8493623592295444,
+      "learning_rate": 7.382579562515926e-06,
+      "loss": 0.4452,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7699481233724054,
+      "learning_rate": 7.38003895436642e-06,
+      "loss": 0.4266,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8349390281702176,
+      "learning_rate": 7.377498527777887e-06,
+      "loss": 0.5024,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8471489722827498,
+      "learning_rate": 7.374958282926381e-06,
+      "loss": 0.5316,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9221474469176514,
+      "learning_rate": 7.372418219987938e-06,
+      "loss": 0.5334,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8966209510688218,
+      "learning_rate": 7.369878339138581e-06,
+      "loss": 0.5412,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9422004357531978,
+      "learning_rate": 7.367338640554322e-06,
+      "loss": 0.5665,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9077745442266718,
+      "learning_rate": 7.364799124411162e-06,
+      "loss": 0.4843,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8639359821414474,
+      "learning_rate": 7.36225979088508e-06,
+      "loss": 0.5359,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9410610289961132,
+      "learning_rate": 7.359720640152061e-06,
+      "loss": 0.5008,
+      "step": 7352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7901229733572144,
+      "learning_rate": 7.357181672388059e-06,
+      "loss": 0.5045,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9425112865880872,
+      "learning_rate": 7.354642887769018e-06,
+      "loss": 0.5639,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8782766252811707,
+      "learning_rate": 7.3521042864708825e-06,
+      "loss": 0.4998,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9183737268629781,
+      "learning_rate": 7.349565868669573e-06,
+      "loss": 0.5099,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9348792780089465,
+      "learning_rate": 7.347027634540993e-06,
+      "loss": 0.5533,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9344656577875735,
+      "learning_rate": 7.344489584261047e-06,
+      "loss": 0.5298,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8642903378823802,
+      "learning_rate": 7.34195171800562e-06,
+      "loss": 0.4594,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.981632696955239,
+      "learning_rate": 7.339414035950576e-06,
+      "loss": 0.5647,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9097634668291024,
+      "learning_rate": 7.3368765382717835e-06,
+      "loss": 0.5358,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8220863835173063,
+      "learning_rate": 7.334339225145084e-06,
+      "loss": 0.5158,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8641456239026523,
+      "learning_rate": 7.331802096746309e-06,
+      "loss": 0.484,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8618023808624063,
+      "learning_rate": 7.329265153251285e-06,
+      "loss": 0.4845,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8504294393521689,
+      "learning_rate": 7.326728394835818e-06,
+      "loss": 0.4895,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8267207028527193,
+      "learning_rate": 7.324191821675697e-06,
+      "loss": 0.4765,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8853464997118534,
+      "learning_rate": 7.321655433946714e-06,
+      "loss": 0.4995,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8718350723134894,
+      "learning_rate": 7.319119231824633e-06,
+      "loss": 0.5163,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9512318446047324,
+      "learning_rate": 7.316583215485208e-06,
+      "loss": 0.5818,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8815002766107952,
+      "learning_rate": 7.314047385104189e-06,
+      "loss": 0.5589,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8286685998102985,
+      "learning_rate": 7.311511740857304e-06,
+      "loss": 0.4836,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9162307895735335,
+      "learning_rate": 7.308976282920268e-06,
+      "loss": 0.5594,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0199676957259531,
+      "learning_rate": 7.306441011468792e-06,
+      "loss": 0.5708,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8913322632852547,
+      "learning_rate": 7.303905926678565e-06,
+      "loss": 0.5483,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8926702359397981,
+      "learning_rate": 7.301371028725261e-06,
+      "loss": 0.6005,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8996125260292345,
+      "learning_rate": 7.298836317784556e-06,
+      "loss": 0.5087,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8145218099302595,
+      "learning_rate": 7.296301794032097e-06,
+      "loss": 0.5221,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9077749787424425,
+      "learning_rate": 7.293767457643523e-06,
+      "loss": 0.4831,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.011921708629939,
+      "learning_rate": 7.291233308794467e-06,
+      "loss": 0.5276,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8568360633757683,
+      "learning_rate": 7.28869934766054e-06,
+      "loss": 0.4962,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8952349900165589,
+      "learning_rate": 7.286165574417339e-06,
+      "loss": 0.5249,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.944935733661351,
+      "learning_rate": 7.283631989240461e-06,
+      "loss": 0.5662,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9099120859981927,
+      "learning_rate": 7.281098592305475e-06,
+      "loss": 0.4988,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8401807741972126,
+      "learning_rate": 7.2785653837879435e-06,
+      "loss": 0.4836,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9144781640878747,
+      "learning_rate": 7.276032363863419e-06,
+      "loss": 0.551,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8775209203998569,
+      "learning_rate": 7.273499532707438e-06,
+      "loss": 0.5175,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9255115313293777,
+      "learning_rate": 7.270966890495515e-06,
+      "loss": 0.5155,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7901310404074847,
+      "learning_rate": 7.268434437403169e-06,
+      "loss": 0.3836,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.911734446879483,
+      "learning_rate": 7.2659021736058966e-06,
+      "loss": 0.5051,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8213031260282923,
+      "learning_rate": 7.263370099279173e-06,
+      "loss": 0.4969,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8756849290520115,
+      "learning_rate": 7.260838214598475e-06,
+      "loss": 0.5081,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9154177602786717,
+      "learning_rate": 7.258306519739263e-06,
+      "loss": 0.5225,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8133074922872229,
+      "learning_rate": 7.2557750148769724e-06,
+      "loss": 0.4734,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9995526440978737,
+      "learning_rate": 7.253243700187043e-06,
+      "loss": 0.5527,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9495255233213361,
+      "learning_rate": 7.250712575844885e-06,
+      "loss": 0.5328,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8704164475758905,
+      "learning_rate": 7.248181642025911e-06,
+      "loss": 0.5269,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9126078437554483,
+      "learning_rate": 7.245650898905507e-06,
+      "loss": 0.5635,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.852380419372046,
+      "learning_rate": 7.243120346659049e-06,
+      "loss": 0.4722,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9106044444571901,
+      "learning_rate": 7.240589985461911e-06,
+      "loss": 0.4851,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.850767141256381,
+      "learning_rate": 7.238059815489439e-06,
+      "loss": 0.4833,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9655676492891794,
+      "learning_rate": 7.235529836916968e-06,
+      "loss": 0.4686,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9151038100861985,
+      "learning_rate": 7.233000049919829e-06,
+      "loss": 0.5022,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8971978008944801,
+      "learning_rate": 7.230470454673335e-06,
+      "loss": 0.5701,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9944557059493082,
+      "learning_rate": 7.227941051352777e-06,
+      "loss": 0.5367,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9504825452251051,
+      "learning_rate": 7.225411840133449e-06,
+      "loss": 0.5686,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0065511172114605,
+      "learning_rate": 7.22288282119062e-06,
+      "loss": 0.5309,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9250351141978376,
+      "learning_rate": 7.2203539946995435e-06,
+      "loss": 0.4719,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8965608857066834,
+      "learning_rate": 7.217825360835475e-06,
+      "loss": 0.5639,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.877393505298854,
+      "learning_rate": 7.21529691977364e-06,
+      "loss": 0.4718,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9556307511449684,
+      "learning_rate": 7.212768671689255e-06,
+      "loss": 0.4729,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9377000629117607,
+      "learning_rate": 7.210240616757531e-06,
+      "loss": 0.4944,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8263931357979006,
+      "learning_rate": 7.20771275515366e-06,
+      "loss": 0.4764,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9111012936356778,
+      "learning_rate": 7.205185087052813e-06,
+      "loss": 0.5189,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3651652953137372,
+      "learning_rate": 7.202657612630165e-06,
+      "loss": 0.5264,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9200714563305669,
+      "learning_rate": 7.200130332060864e-06,
+      "loss": 0.5478,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9002547184998756,
+      "learning_rate": 7.197603245520042e-06,
+      "loss": 0.5124,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.013949035151007,
+      "learning_rate": 7.195076353182834e-06,
+      "loss": 0.4858,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.824976407945111,
+      "learning_rate": 7.192549655224346e-06,
+      "loss": 0.4967,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8838276237041307,
+      "learning_rate": 7.190023151819674e-06,
+      "loss": 0.5368,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8606988929514742,
+      "learning_rate": 7.187496843143908e-06,
+      "loss": 0.4799,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8872973984439431,
+      "learning_rate": 7.184970729372118e-06,
+      "loss": 0.4783,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9381768183168121,
+      "learning_rate": 7.182444810679354e-06,
+      "loss": 0.5491,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9379227502504542,
+      "learning_rate": 7.17991908724067e-06,
+      "loss": 0.556,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9137485386338625,
+      "learning_rate": 7.17739355923109e-06,
+      "loss": 0.51,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2644241726683547,
+      "learning_rate": 7.174868226825631e-06,
+      "loss": 0.5416,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8958335672404645,
+      "learning_rate": 7.172343090199301e-06,
+      "loss": 0.5207,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8590210296481723,
+      "learning_rate": 7.169818149527087e-06,
+      "loss": 0.5337,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8384171957758327,
+      "learning_rate": 7.167293404983962e-06,
+      "loss": 0.4613,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.954130281660937,
+      "learning_rate": 7.164768856744893e-06,
+      "loss": 0.5646,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0030607106468707,
+      "learning_rate": 7.1622445049848286e-06,
+      "loss": 0.5386,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8139381902091727,
+      "learning_rate": 7.159720349878698e-06,
+      "loss": 0.4671,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8966599226454,
+      "learning_rate": 7.157196391601433e-06,
+      "loss": 0.5602,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9522741156653227,
+      "learning_rate": 7.154672630327937e-06,
+      "loss": 0.5023,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9261395171564893,
+      "learning_rate": 7.152149066233098e-06,
+      "loss": 0.5739,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9771922952299095,
+      "learning_rate": 7.149625699491809e-06,
+      "loss": 0.5441,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8530678821927267,
+      "learning_rate": 7.147102530278929e-06,
+      "loss": 0.4926,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.926699880774957,
+      "learning_rate": 7.14457955876931e-06,
+      "loss": 0.5174,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9079063732597161,
+      "learning_rate": 7.142056785137799e-06,
+      "loss": 0.475,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0016805486137383,
+      "learning_rate": 7.139534209559217e-06,
+      "loss": 0.6123,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.926003799746655,
+      "learning_rate": 7.137011832208374e-06,
+      "loss": 0.5129,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8901597233303454,
+      "learning_rate": 7.134489653260075e-06,
+      "loss": 0.4922,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9039581597904545,
+      "learning_rate": 7.131967672889101e-06,
+      "loss": 0.494,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7942695446448926,
+      "learning_rate": 7.129445891270219e-06,
+      "loss": 0.4672,
+      "step": 7443
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8833734306634364,
+      "learning_rate": 7.126924308578196e-06,
+      "loss": 0.5106,
+      "step": 7444
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8939098610259423,
+      "learning_rate": 7.124402924987767e-06,
+      "loss": 0.5146,
+      "step": 7445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8382061476262879,
+      "learning_rate": 7.121881740673664e-06,
+      "loss": 0.4823,
+      "step": 7446
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7778752928864608,
+      "learning_rate": 7.119360755810607e-06,
+      "loss": 0.4548,
+      "step": 7447
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8011529749679672,
+      "learning_rate": 7.116839970573292e-06,
+      "loss": 0.4825,
+      "step": 7448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9036417498542302,
+      "learning_rate": 7.114319385136408e-06,
+      "loss": 0.5491,
+      "step": 7449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8761401707218572,
+      "learning_rate": 7.111798999674635e-06,
+      "loss": 0.4814,
+      "step": 7450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8523161458806872,
+      "learning_rate": 7.109278814362629e-06,
+      "loss": 0.543,
+      "step": 7451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9528197182425885,
+      "learning_rate": 7.106758829375033e-06,
+      "loss": 0.5567,
+      "step": 7452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8899638209596322,
+      "learning_rate": 7.104239044886487e-06,
+      "loss": 0.5432,
+      "step": 7453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.870345394746601,
+      "learning_rate": 7.101719461071608e-06,
+      "loss": 0.4923,
+      "step": 7454
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9432026321304644,
+      "learning_rate": 7.099200078104995e-06,
+      "loss": 0.4855,
+      "step": 7455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.909353097814008,
+      "learning_rate": 7.0966808961612475e-06,
+      "loss": 0.5445,
+      "step": 7456
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8222116703513183,
+      "learning_rate": 7.094161915414939e-06,
+      "loss": 0.5132,
+      "step": 7457
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8998287656530938,
+      "learning_rate": 7.091643136040629e-06,
+      "loss": 0.5051,
+      "step": 7458
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.909655368823903,
+      "learning_rate": 7.089124558212872e-06,
+      "loss": 0.5384,
+      "step": 7459
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9122072478770556,
+      "learning_rate": 7.0866061821062025e-06,
+      "loss": 0.4496,
+      "step": 7460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9958858222651449,
+      "learning_rate": 7.084088007895136e-06,
+      "loss": 0.5486,
+      "step": 7461
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8786152819765978,
+      "learning_rate": 7.081570035754189e-06,
+      "loss": 0.4973,
+      "step": 7462
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.895279929473395,
+      "learning_rate": 7.079052265857847e-06,
+      "loss": 0.5163,
+      "step": 7463
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9016037453401397,
+      "learning_rate": 7.0765346983805925e-06,
+      "loss": 0.5418,
+      "step": 7464
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9594257408814107,
+      "learning_rate": 7.074017333496892e-06,
+      "loss": 0.4872,
+      "step": 7465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9198839250326495,
+      "learning_rate": 7.071500171381193e-06,
+      "loss": 0.5639,
+      "step": 7466
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9113176502875616,
+      "learning_rate": 7.068983212207934e-06,
+      "loss": 0.5025,
+      "step": 7467
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8920345474434166,
+      "learning_rate": 7.066466456151541e-06,
+      "loss": 0.5027,
+      "step": 7468
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.926683305043765,
+      "learning_rate": 7.063949903386419e-06,
+      "loss": 0.478,
+      "step": 7469
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8747561455025346,
+      "learning_rate": 7.061433554086964e-06,
+      "loss": 0.5093,
+      "step": 7470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8233419034552449,
+      "learning_rate": 7.058917408427559e-06,
+      "loss": 0.5061,
+      "step": 7471
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.968703122551061,
+      "learning_rate": 7.056401466582567e-06,
+      "loss": 0.5174,
+      "step": 7472
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.98020612835273,
+      "learning_rate": 7.053885728726343e-06,
+      "loss": 0.4846,
+      "step": 7473
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9245074098344465,
+      "learning_rate": 7.051370195033227e-06,
+      "loss": 0.5086,
+      "step": 7474
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8498280153892569,
+      "learning_rate": 7.048854865677538e-06,
+      "loss": 0.4974,
+      "step": 7475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9584978657105867,
+      "learning_rate": 7.04633974083359e-06,
+      "loss": 0.553,
+      "step": 7476
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9813074133240363,
+      "learning_rate": 7.04382482067568e-06,
+      "loss": 0.5338,
+      "step": 7477
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8597723255573146,
+      "learning_rate": 7.041310105378085e-06,
+      "loss": 0.5018,
+      "step": 7478
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9700494913363266,
+      "learning_rate": 7.038795595115076e-06,
+      "loss": 0.5172,
+      "step": 7479
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9476074585160819,
+      "learning_rate": 7.036281290060907e-06,
+      "loss": 0.5126,
+      "step": 7480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9883722573627236,
+      "learning_rate": 7.033767190389814e-06,
+      "loss": 0.5522,
+      "step": 7481
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8750535173574526,
+      "learning_rate": 7.031253296276024e-06,
+      "loss": 0.4966,
+      "step": 7482
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8808199103635561,
+      "learning_rate": 7.028739607893746e-06,
+      "loss": 0.4345,
+      "step": 7483
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9292995866899433,
+      "learning_rate": 7.026226125417182e-06,
+      "loss": 0.5172,
+      "step": 7484
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8682760191868,
+      "learning_rate": 7.023712849020506e-06,
+      "loss": 0.5119,
+      "step": 7485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9130486211387956,
+      "learning_rate": 7.021199778877891e-06,
+      "loss": 0.508,
+      "step": 7486
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8989745028238334,
+      "learning_rate": 7.01868691516349e-06,
+      "loss": 0.5389,
+      "step": 7487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8711288089074521,
+      "learning_rate": 7.016174258051441e-06,
+      "loss": 0.5308,
+      "step": 7488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8887074836595515,
+      "learning_rate": 7.013661807715866e-06,
+      "loss": 0.5039,
+      "step": 7489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.915092893637171,
+      "learning_rate": 7.0111495643308836e-06,
+      "loss": 0.5086,
+      "step": 7490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9629412883487775,
+      "learning_rate": 7.008637528070583e-06,
+      "loss": 0.5252,
+      "step": 7491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8843063699819966,
+      "learning_rate": 7.006125699109048e-06,
+      "loss": 0.4488,
+      "step": 7492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9509376929761509,
+      "learning_rate": 7.003614077620348e-06,
+      "loss": 0.525,
+      "step": 7493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8383261728774117,
+      "learning_rate": 7.001102663778533e-06,
+      "loss": 0.5241,
+      "step": 7494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8631840695371249,
+      "learning_rate": 6.998591457757643e-06,
+      "loss": 0.4911,
+      "step": 7495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8323660725814972,
+      "learning_rate": 6.9960804597317045e-06,
+      "loss": 0.421,
+      "step": 7496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7882803380832858,
+      "learning_rate": 6.993569669874724e-06,
+      "loss": 0.4342,
+      "step": 7497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9324318208303033,
+      "learning_rate": 6.9910590883607e-06,
+      "loss": 0.477,
+      "step": 7498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8485127165181465,
+      "learning_rate": 6.9885487153636125e-06,
+      "loss": 0.4926,
+      "step": 7499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9585992901038118,
+      "learning_rate": 6.986038551057426e-06,
+      "loss": 0.4797,
+      "step": 7500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9353390797402146,
+      "learning_rate": 6.983528595616096e-06,
+      "loss": 0.505,
+      "step": 7501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8813237756519612,
+      "learning_rate": 6.98101884921356e-06,
+      "loss": 0.4647,
+      "step": 7502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8437093853065601,
+      "learning_rate": 6.978509312023736e-06,
+      "loss": 0.4573,
+      "step": 7503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8652761631005595,
+      "learning_rate": 6.975999984220541e-06,
+      "loss": 0.4656,
+      "step": 7504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9284862732710436,
+      "learning_rate": 6.9734908659778636e-06,
+      "loss": 0.5179,
+      "step": 7505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8519841381992954,
+      "learning_rate": 6.97098195746958e-06,
+      "loss": 0.4691,
+      "step": 7506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9432823008748794,
+      "learning_rate": 6.968473258869566e-06,
+      "loss": 0.5224,
+      "step": 7507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9314459449139519,
+      "learning_rate": 6.965964770351665e-06,
+      "loss": 0.5125,
+      "step": 7508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9017714642380549,
+      "learning_rate": 6.963456492089711e-06,
+      "loss": 0.5316,
+      "step": 7509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9322043966660608,
+      "learning_rate": 6.960948424257532e-06,
+      "loss": 0.4958,
+      "step": 7510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0163089194858204,
+      "learning_rate": 6.9584405670289326e-06,
+      "loss": 0.5121,
+      "step": 7511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9067873188043927,
+      "learning_rate": 6.955932920577699e-06,
+      "loss": 0.576,
+      "step": 7512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8910120630030297,
+      "learning_rate": 6.953425485077618e-06,
+      "loss": 0.5537,
+      "step": 7513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9099219769405817,
+      "learning_rate": 6.950918260702449e-06,
+      "loss": 0.5537,
+      "step": 7514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9177750168038247,
+      "learning_rate": 6.948411247625937e-06,
+      "loss": 0.5237,
+      "step": 7515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8678464941195639,
+      "learning_rate": 6.9459044460218205e-06,
+      "loss": 0.5501,
+      "step": 7516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8816036797700124,
+      "learning_rate": 6.943397856063818e-06,
+      "loss": 0.5234,
+      "step": 7517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9188057188168129,
+      "learning_rate": 6.9408914779256285e-06,
+      "loss": 0.5016,
+      "step": 7518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8878831599490864,
+      "learning_rate": 6.938385311780951e-06,
+      "loss": 0.4565,
+      "step": 7519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9888578876329767,
+      "learning_rate": 6.935879357803453e-06,
+      "loss": 0.5835,
+      "step": 7520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.993637491505397,
+      "learning_rate": 6.933373616166799e-06,
+      "loss": 0.5403,
+      "step": 7521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8068368590318049,
+      "learning_rate": 6.930868087044634e-06,
+      "loss": 0.4891,
+      "step": 7522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8173237797717122,
+      "learning_rate": 6.9283627706105836e-06,
+      "loss": 0.4959,
+      "step": 7523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8868903526963838,
+      "learning_rate": 6.925857667038274e-06,
+      "loss": 0.5011,
+      "step": 7524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.882363978650632,
+      "learning_rate": 6.923352776501302e-06,
+      "loss": 0.5676,
+      "step": 7525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8759337087339624,
+      "learning_rate": 6.920848099173247e-06,
+      "loss": 0.5372,
+      "step": 7526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9091568143961809,
+      "learning_rate": 6.918343635227694e-06,
+      "loss": 0.5224,
+      "step": 7527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9492937811993646,
+      "learning_rate": 6.915839384838192e-06,
+      "loss": 0.5301,
+      "step": 7528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9466146485903585,
+      "learning_rate": 6.913335348178283e-06,
+      "loss": 0.4828,
+      "step": 7529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8734794349166198,
+      "learning_rate": 6.910831525421499e-06,
+      "loss": 0.5246,
+      "step": 7530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9151217895547622,
+      "learning_rate": 6.90832791674135e-06,
+      "loss": 0.5339,
+      "step": 7531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8742496126873698,
+      "learning_rate": 6.905824522311331e-06,
+      "loss": 0.4919,
+      "step": 7532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9385429607671967,
+      "learning_rate": 6.90332134230493e-06,
+      "loss": 0.5338,
+      "step": 7533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9194053424538637,
+      "learning_rate": 6.900818376895615e-06,
+      "loss": 0.5629,
+      "step": 7534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9612350119647755,
+      "learning_rate": 6.898315626256833e-06,
+      "loss": 0.5202,
+      "step": 7535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9415097629294374,
+      "learning_rate": 6.895813090562031e-06,
+      "loss": 0.5008,
+      "step": 7536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8088669541501505,
+      "learning_rate": 6.893310769984629e-06,
+      "loss": 0.5369,
+      "step": 7537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.826545654887671,
+      "learning_rate": 6.890808664698031e-06,
+      "loss": 0.468,
+      "step": 7538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8972849914206232,
+      "learning_rate": 6.888306774875638e-06,
+      "loss": 0.4912,
+      "step": 7539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8815680247579892,
+      "learning_rate": 6.885805100690825e-06,
+      "loss": 0.5126,
+      "step": 7540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9884025324541118,
+      "learning_rate": 6.883303642316954e-06,
+      "loss": 0.5923,
+      "step": 7541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8249574853963665,
+      "learning_rate": 6.8808023999273784e-06,
+      "loss": 0.5241,
+      "step": 7542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8637034169688512,
+      "learning_rate": 6.878301373695431e-06,
+      "loss": 0.488,
+      "step": 7543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.930755552156985,
+      "learning_rate": 6.8758005637944245e-06,
+      "loss": 0.5284,
+      "step": 7544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9909722910723238,
+      "learning_rate": 6.873299970397672e-06,
+      "loss": 0.5452,
+      "step": 7545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9544814719075744,
+      "learning_rate": 6.870799593678459e-06,
+      "loss": 0.5359,
+      "step": 7546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8329864765301852,
+      "learning_rate": 6.868299433810053e-06,
+      "loss": 0.4407,
+      "step": 7547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9008973754613913,
+      "learning_rate": 6.8657994909657235e-06,
+      "loss": 0.5463,
+      "step": 7548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8973104532423516,
+      "learning_rate": 6.86329976531871e-06,
+      "loss": 0.5159,
+      "step": 7549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8675670232532094,
+      "learning_rate": 6.860800257042235e-06,
+      "loss": 0.4741,
+      "step": 7550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9491734938139588,
+      "learning_rate": 6.8583009663095215e-06,
+      "loss": 0.5078,
+      "step": 7551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.036963655326414,
+      "learning_rate": 6.855801893293765e-06,
+      "loss": 0.6484,
+      "step": 7552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9071145539618047,
+      "learning_rate": 6.853303038168144e-06,
+      "loss": 0.5096,
+      "step": 7553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0233484930312005,
+      "learning_rate": 6.8508044011058375e-06,
+      "loss": 0.5484,
+      "step": 7554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.941587849993408,
+      "learning_rate": 6.84830598227999e-06,
+      "loss": 0.5463,
+      "step": 7555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9071870766619743,
+      "learning_rate": 6.845807781863739e-06,
+      "loss": 0.5312,
+      "step": 7556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8802657720651078,
+      "learning_rate": 6.8433098000302155e-06,
+      "loss": 0.5114,
+      "step": 7557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8716345359948411,
+      "learning_rate": 6.840812036952522e-06,
+      "loss": 0.43,
+      "step": 7558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9806567088859193,
+      "learning_rate": 6.83831449280375e-06,
+      "loss": 0.4978,
+      "step": 7559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.927207704508905,
+      "learning_rate": 6.8358171677569814e-06,
+      "loss": 0.5402,
+      "step": 7560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9100483563196377,
+      "learning_rate": 6.833320061985278e-06,
+      "loss": 0.5295,
+      "step": 7561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9003269433900096,
+      "learning_rate": 6.830823175661681e-06,
+      "loss": 0.5654,
+      "step": 7562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8768709859948077,
+      "learning_rate": 6.828326508959229e-06,
+      "loss": 0.4975,
+      "step": 7563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9148195812629236,
+      "learning_rate": 6.825830062050939e-06,
+      "loss": 0.5429,
+      "step": 7564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8353913853922079,
+      "learning_rate": 6.823333835109805e-06,
+      "loss": 0.5226,
+      "step": 7565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9198354109819284,
+      "learning_rate": 6.820837828308823e-06,
+      "loss": 0.519,
+      "step": 7566
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9656566943225391,
+      "learning_rate": 6.818342041820959e-06,
+      "loss": 0.5811,
+      "step": 7567
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9093452706591472,
+      "learning_rate": 6.815846475819166e-06,
+      "loss": 0.5197,
+      "step": 7568
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9813198921313696,
+      "learning_rate": 6.813351130476391e-06,
+      "loss": 0.4577,
+      "step": 7569
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0110714074684641,
+      "learning_rate": 6.810856005965558e-06,
+      "loss": 0.5877,
+      "step": 7570
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9041099671510405,
+      "learning_rate": 6.808361102459568e-06,
+      "loss": 0.5148,
+      "step": 7571
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8462419013566584,
+      "learning_rate": 6.80586642013133e-06,
+      "loss": 0.4424,
+      "step": 7572
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.954344805260286,
+      "learning_rate": 6.803371959153714e-06,
+      "loss": 0.5495,
+      "step": 7573
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8968157028231019,
+      "learning_rate": 6.800877719699581e-06,
+      "loss": 0.5173,
+      "step": 7574
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8759713185900999,
+      "learning_rate": 6.798383701941791e-06,
+      "loss": 0.5261,
+      "step": 7575
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9532005915781344,
+      "learning_rate": 6.795889906053168e-06,
+      "loss": 0.5458,
+      "step": 7576
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.958076507362562,
+      "learning_rate": 6.79339633220653e-06,
+      "loss": 0.4898,
+      "step": 7577
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9654309823547377,
+      "learning_rate": 6.7909029805746855e-06,
+      "loss": 0.4927,
+      "step": 7578
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8362198322373752,
+      "learning_rate": 6.788409851330419e-06,
+      "loss": 0.4931,
+      "step": 7579
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8885475948416547,
+      "learning_rate": 6.7859169446464955e-06,
+      "loss": 0.464,
+      "step": 7580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8621341356587938,
+      "learning_rate": 6.783424260695681e-06,
+      "loss": 0.5048,
+      "step": 7581
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9836189609295446,
+      "learning_rate": 6.780931799650714e-06,
+      "loss": 0.5756,
+      "step": 7582
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9043346316869034,
+      "learning_rate": 6.778439561684311e-06,
+      "loss": 0.5557,
+      "step": 7583
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9260466410600587,
+      "learning_rate": 6.775947546969195e-06,
+      "loss": 0.5034,
+      "step": 7584
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9557467442673623,
+      "learning_rate": 6.773455755678054e-06,
+      "loss": 0.5385,
+      "step": 7585
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8211146342579827,
+      "learning_rate": 6.770964187983563e-06,
+      "loss": 0.4589,
+      "step": 7586
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9688902980547851,
+      "learning_rate": 6.7684728440583934e-06,
+      "loss": 0.5368,
+      "step": 7587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9309423999874568,
+      "learning_rate": 6.7659817240751906e-06,
+      "loss": 0.5372,
+      "step": 7588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8931467289113518,
+      "learning_rate": 6.76349082820658e-06,
+      "loss": 0.4692,
+      "step": 7589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8857663662886142,
+      "learning_rate": 6.7610001566251885e-06,
+      "loss": 0.4539,
+      "step": 7590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8738241729867862,
+      "learning_rate": 6.758509709503614e-06,
+      "loss": 0.5054,
+      "step": 7591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.85924221496039,
+      "learning_rate": 6.756019487014437e-06,
+      "loss": 0.4874,
+      "step": 7592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8547022449248107,
+      "learning_rate": 6.753529489330235e-06,
+      "loss": 0.5146,
+      "step": 7593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7745590974676932,
+      "learning_rate": 6.751039716623562e-06,
+      "loss": 0.4518,
+      "step": 7594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9500988557366958,
+      "learning_rate": 6.7485501690669495e-06,
+      "loss": 0.5643,
+      "step": 7595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9158231254949393,
+      "learning_rate": 6.74606084683293e-06,
+      "loss": 0.4598,
+      "step": 7596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9059197379989498,
+      "learning_rate": 6.743571750094009e-06,
+      "loss": 0.534,
+      "step": 7597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9930882474280531,
+      "learning_rate": 6.741082879022671e-06,
+      "loss": 0.6028,
+      "step": 7598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9398046854732968,
+      "learning_rate": 6.738594233791405e-06,
+      "loss": 0.5218,
+      "step": 7599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9324822671783136,
+      "learning_rate": 6.7361058145726645e-06,
+      "loss": 0.5388,
+      "step": 7600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9652578493993295,
+      "learning_rate": 6.733617621538893e-06,
+      "loss": 0.6204,
+      "step": 7601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9326608597566178,
+      "learning_rate": 6.731129654862526e-06,
+      "loss": 0.5349,
+      "step": 7602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.874818064011275,
+      "learning_rate": 6.7286419147159745e-06,
+      "loss": 0.5201,
+      "step": 7603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8026460981939355,
+      "learning_rate": 6.726154401271633e-06,
+      "loss": 0.452,
+      "step": 7604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9147355097876252,
+      "learning_rate": 6.723667114701892e-06,
+      "loss": 0.512,
+      "step": 7605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8944621682572889,
+      "learning_rate": 6.721180055179113e-06,
+      "loss": 0.4557,
+      "step": 7606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0302155219494231,
+      "learning_rate": 6.718693222875644e-06,
+      "loss": 0.4877,
+      "step": 7607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8921531524431237,
+      "learning_rate": 6.7162066179638286e-06,
+      "loss": 0.5354,
+      "step": 7608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8902878385006948,
+      "learning_rate": 6.713720240615982e-06,
+      "loss": 0.4891,
+      "step": 7609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8666985623979152,
+      "learning_rate": 6.711234091004404e-06,
+      "loss": 0.4978,
+      "step": 7610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8367718524021635,
+      "learning_rate": 6.708748169301389e-06,
+      "loss": 0.5615,
+      "step": 7611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9411993131581817,
+      "learning_rate": 6.706262475679205e-06,
+      "loss": 0.5926,
+      "step": 7612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9335725128206275,
+      "learning_rate": 6.703777010310111e-06,
+      "loss": 0.5449,
+      "step": 7613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9219406710574435,
+      "learning_rate": 6.701291773366347e-06,
+      "loss": 0.5564,
+      "step": 7614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9183979061357251,
+      "learning_rate": 6.698806765020136e-06,
+      "loss": 0.5195,
+      "step": 7615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9554532988160243,
+      "learning_rate": 6.696321985443688e-06,
+      "loss": 0.5166,
+      "step": 7616
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8800060318983304,
+      "learning_rate": 6.693837434809199e-06,
+      "loss": 0.4841,
+      "step": 7617
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9066257671231671,
+      "learning_rate": 6.691353113288839e-06,
+      "loss": 0.4821,
+      "step": 7618
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8696221704487445,
+      "learning_rate": 6.688869021054773e-06,
+      "loss": 0.4696,
+      "step": 7619
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.810305875465549,
+      "learning_rate": 6.686385158279151e-06,
+      "loss": 0.4334,
+      "step": 7620
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9236041534274468,
+      "learning_rate": 6.683901525134096e-06,
+      "loss": 0.5369,
+      "step": 7621
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9437786609905792,
+      "learning_rate": 6.681418121791725e-06,
+      "loss": 0.525,
+      "step": 7622
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9462694496156745,
+      "learning_rate": 6.678934948424134e-06,
+      "loss": 0.4898,
+      "step": 7623
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9412840607698836,
+      "learning_rate": 6.6764520052034054e-06,
+      "loss": 0.5509,
+      "step": 7624
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8845126699011401,
+      "learning_rate": 6.673969292301604e-06,
+      "loss": 0.4877,
+      "step": 7625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8569357062286083,
+      "learning_rate": 6.6714868098907825e-06,
+      "loss": 0.5046,
+      "step": 7626
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9405002337042057,
+      "learning_rate": 6.6690045581429705e-06,
+      "loss": 0.5373,
+      "step": 7627
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8880358306525056,
+      "learning_rate": 6.666522537230189e-06,
+      "loss": 0.5047,
+      "step": 7628
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9598858005780319,
+      "learning_rate": 6.664040747324437e-06,
+      "loss": 0.5166,
+      "step": 7629
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8742479010742386,
+      "learning_rate": 6.661559188597706e-06,
+      "loss": 0.5402,
+      "step": 7630
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.894172077733656,
+      "learning_rate": 6.659077861221959e-06,
+      "loss": 0.4528,
+      "step": 7631
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8451101514996602,
+      "learning_rate": 6.656596765369153e-06,
+      "loss": 0.4539,
+      "step": 7632
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9342223311005059,
+      "learning_rate": 6.654115901211229e-06,
+      "loss": 0.5558,
+      "step": 7633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9352876875867336,
+      "learning_rate": 6.651635268920101e-06,
+      "loss": 0.5217,
+      "step": 7634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9387669753786152,
+      "learning_rate": 6.64915486866768e-06,
+      "loss": 0.4883,
+      "step": 7635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7909913569761613,
+      "learning_rate": 6.646674700625857e-06,
+      "loss": 0.4587,
+      "step": 7636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.940002808704583,
+      "learning_rate": 6.644194764966499e-06,
+      "loss": 0.5573,
+      "step": 7637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8798200273870914,
+      "learning_rate": 6.641715061861469e-06,
+      "loss": 0.5314,
+      "step": 7638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9118097465638171,
+      "learning_rate": 6.639235591482608e-06,
+      "loss": 0.4951,
+      "step": 7639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9265393540958253,
+      "learning_rate": 6.636756354001737e-06,
+      "loss": 0.5746,
+      "step": 7640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8996667634118599,
+      "learning_rate": 6.6342773495906675e-06,
+      "loss": 0.5058,
+      "step": 7641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9338778709815375,
+      "learning_rate": 6.631798578421195e-06,
+      "loss": 0.5082,
+      "step": 7642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8415158595072395,
+      "learning_rate": 6.62932004066509e-06,
+      "loss": 0.467,
+      "step": 7643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9406207341615731,
+      "learning_rate": 6.626841736494119e-06,
+      "loss": 0.551,
+      "step": 7644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9956040687979807,
+      "learning_rate": 6.624363666080021e-06,
+      "loss": 0.5501,
+      "step": 7645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.984348172818565,
+      "learning_rate": 6.62188582959453e-06,
+      "loss": 0.5825,
+      "step": 7646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8693191073068617,
+      "learning_rate": 6.619408227209352e-06,
+      "loss": 0.5317,
+      "step": 7647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9489546286652094,
+      "learning_rate": 6.616930859096185e-06,
+      "loss": 0.5389,
+      "step": 7648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9362386586362307,
+      "learning_rate": 6.61445372542671e-06,
+      "loss": 0.4964,
+      "step": 7649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8275225693781739,
+      "learning_rate": 6.61197682637259e-06,
+      "loss": 0.5144,
+      "step": 7650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8801001351782364,
+      "learning_rate": 6.609500162105469e-06,
+      "loss": 0.5059,
+      "step": 7651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8904305060344985,
+      "learning_rate": 6.60702373279698e-06,
+      "loss": 0.5316,
+      "step": 7652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8565058285424657,
+      "learning_rate": 6.6045475386187376e-06,
+      "loss": 0.4461,
+      "step": 7653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9101220847706999,
+      "learning_rate": 6.602071579742337e-06,
+      "loss": 0.5361,
+      "step": 7654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9587323995861344,
+      "learning_rate": 6.599595856339363e-06,
+      "loss": 0.4869,
+      "step": 7655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.976018429112491,
+      "learning_rate": 6.597120368581382e-06,
+      "loss": 0.518,
+      "step": 7656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7956454072227547,
+      "learning_rate": 6.594645116639939e-06,
+      "loss": 0.4691,
+      "step": 7657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9794199560151343,
+      "learning_rate": 6.592170100686568e-06,
+      "loss": 0.5548,
+      "step": 7658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9088047741398635,
+      "learning_rate": 6.5896953208927886e-06,
+      "loss": 0.5546,
+      "step": 7659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8844084202928726,
+      "learning_rate": 6.587220777430097e-06,
+      "loss": 0.5246,
+      "step": 7660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9737946962118639,
+      "learning_rate": 6.584746470469978e-06,
+      "loss": 0.4907,
+      "step": 7661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8894751424608862,
+      "learning_rate": 6.582272400183901e-06,
+      "loss": 0.5086,
+      "step": 7662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9085894752314807,
+      "learning_rate": 6.579798566743314e-06,
+      "loss": 0.4848,
+      "step": 7663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8982538680909244,
+      "learning_rate": 6.577324970319652e-06,
+      "loss": 0.5328,
+      "step": 7664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8584185228163657,
+      "learning_rate": 6.574851611084335e-06,
+      "loss": 0.5389,
+      "step": 7665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9102759729627776,
+      "learning_rate": 6.572378489208762e-06,
+      "loss": 0.5237,
+      "step": 7666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8677201197600535,
+      "learning_rate": 6.569905604864319e-06,
+      "loss": 0.4592,
+      "step": 7667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2005174417117501,
+      "learning_rate": 6.567432958222379e-06,
+      "loss": 0.537,
+      "step": 7668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8504096257899675,
+      "learning_rate": 6.564960549454285e-06,
+      "loss": 0.5129,
+      "step": 7669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9084205797953231,
+      "learning_rate": 6.562488378731381e-06,
+      "loss": 0.5082,
+      "step": 7670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.961813736450623,
+      "learning_rate": 6.560016446224983e-06,
+      "loss": 0.5459,
+      "step": 7671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8893401843439549,
+      "learning_rate": 6.557544752106392e-06,
+      "loss": 0.5062,
+      "step": 7672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8730833336776954,
+      "learning_rate": 6.5550732965468985e-06,
+      "loss": 0.4991,
+      "step": 7673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.96633050431716,
+      "learning_rate": 6.552602079717772e-06,
+      "loss": 0.5265,
+      "step": 7674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9610131291886921,
+      "learning_rate": 6.550131101790258e-06,
+      "loss": 0.5188,
+      "step": 7675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9082109466001308,
+      "learning_rate": 6.547660362935603e-06,
+      "loss": 0.542,
+      "step": 7676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9664922066319224,
+      "learning_rate": 6.545189863325023e-06,
+      "loss": 0.5302,
+      "step": 7677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8572812606217288,
+      "learning_rate": 6.542719603129716e-06,
+      "loss": 0.5131,
+      "step": 7678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0805563703213321,
+      "learning_rate": 6.540249582520879e-06,
+      "loss": 0.5196,
+      "step": 7679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9355592376625326,
+      "learning_rate": 6.537779801669677e-06,
+      "loss": 0.51,
+      "step": 7680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0094796867058717,
+      "learning_rate": 6.535310260747259e-06,
+      "loss": 0.5075,
+      "step": 7681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9183654901634732,
+      "learning_rate": 6.5328409599247715e-06,
+      "loss": 0.4783,
+      "step": 7682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.918630109918347,
+      "learning_rate": 6.530371899373329e-06,
+      "loss": 0.5179,
+      "step": 7683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8697336708812505,
+      "learning_rate": 6.527903079264033e-06,
+      "loss": 0.4734,
+      "step": 7684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0264393557260938,
+      "learning_rate": 6.525434499767978e-06,
+      "loss": 0.5394,
+      "step": 7685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8638538677983133,
+      "learning_rate": 6.52296616105623e-06,
+      "loss": 0.4703,
+      "step": 7686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9552971643499772,
+      "learning_rate": 6.5204980632998394e-06,
+      "loss": 0.5097,
+      "step": 7687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8846975559947114,
+      "learning_rate": 6.5180302066698495e-06,
+      "loss": 0.5423,
+      "step": 7688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8952890587377771,
+      "learning_rate": 6.515562591337279e-06,
+      "loss": 0.5103,
+      "step": 7689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9432984337915332,
+      "learning_rate": 6.513095217473127e-06,
+      "loss": 0.5504,
+      "step": 7690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9411424831312857,
+      "learning_rate": 6.510628085248385e-06,
+      "loss": 0.4815,
+      "step": 7691
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.93391599034221,
+      "learning_rate": 6.508161194834024e-06,
+      "loss": 0.5194,
+      "step": 7692
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8209924490621792,
+      "learning_rate": 6.505694546400989e-06,
+      "loss": 0.4406,
+      "step": 7693
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9632496572429818,
+      "learning_rate": 6.503228140120228e-06,
+      "loss": 0.5211,
+      "step": 7694
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9636864179124074,
+      "learning_rate": 6.500761976162655e-06,
+      "loss": 0.544,
+      "step": 7695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9628458753847853,
+      "learning_rate": 6.498296054699169e-06,
+      "loss": 0.4921,
+      "step": 7696
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8605164541806277,
+      "learning_rate": 6.495830375900665e-06,
+      "loss": 0.5045,
+      "step": 7697
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8357418989520136,
+      "learning_rate": 6.493364939938007e-06,
+      "loss": 0.5184,
+      "step": 7698
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8775988577494073,
+      "learning_rate": 6.490899746982045e-06,
+      "loss": 0.5773,
+      "step": 7699
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8507036855449437,
+      "learning_rate": 6.48843479720362e-06,
+      "loss": 0.5532,
+      "step": 7700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.885247279447743,
+      "learning_rate": 6.48597009077355e-06,
+      "loss": 0.4917,
+      "step": 7701
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8213151157416119,
+      "learning_rate": 6.483505627862632e-06,
+      "loss": 0.4146,
+      "step": 7702
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9521487130066696,
+      "learning_rate": 6.481041408641659e-06,
+      "loss": 0.4853,
+      "step": 7703
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9047442954374255,
+      "learning_rate": 6.478577433281394e-06,
+      "loss": 0.501,
+      "step": 7704
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9123953950706234,
+      "learning_rate": 6.476113701952587e-06,
+      "loss": 0.505,
+      "step": 7705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8918355614489641,
+      "learning_rate": 6.473650214825979e-06,
+      "loss": 0.5052,
+      "step": 7706
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8460692223732004,
+      "learning_rate": 6.4711869720722804e-06,
+      "loss": 0.5007,
+      "step": 7707
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9514163634580783,
+      "learning_rate": 6.468723973862194e-06,
+      "loss": 0.4888,
+      "step": 7708
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8407870858699927,
+      "learning_rate": 6.466261220366406e-06,
+      "loss": 0.5123,
+      "step": 7709
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0438409207420924,
+      "learning_rate": 6.463798711755582e-06,
+      "loss": 0.5079,
+      "step": 7710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8393621192093546,
+      "learning_rate": 6.461336448200366e-06,
+      "loss": 0.5016,
+      "step": 7711
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8504075861887,
+      "learning_rate": 6.458874429871399e-06,
+      "loss": 0.5265,
+      "step": 7712
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8860413988426402,
+      "learning_rate": 6.456412656939293e-06,
+      "loss": 0.5535,
+      "step": 7713
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0156180242304687,
+      "learning_rate": 6.453951129574644e-06,
+      "loss": 0.5815,
+      "step": 7714
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0120525970975776,
+      "learning_rate": 6.451489847948039e-06,
+      "loss": 0.5458,
+      "step": 7715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9642455156600728,
+      "learning_rate": 6.44902881223004e-06,
+      "loss": 0.5594,
+      "step": 7716
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8316070911321247,
+      "learning_rate": 6.446568022591192e-06,
+      "loss": 0.4743,
+      "step": 7717
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9448781886763189,
+      "learning_rate": 6.4441074792020305e-06,
+      "loss": 0.5924,
+      "step": 7718
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8313716764000463,
+      "learning_rate": 6.4416471822330684e-06,
+      "loss": 0.4905,
+      "step": 7719
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9063248870520652,
+      "learning_rate": 6.439187131854796e-06,
+      "loss": 0.5645,
+      "step": 7720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9106218132909022,
+      "learning_rate": 6.436727328237699e-06,
+      "loss": 0.5104,
+      "step": 7721
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9383859538169318,
+      "learning_rate": 6.43426777155224e-06,
+      "loss": 0.5514,
+      "step": 7722
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8098969358101791,
+      "learning_rate": 6.431808461968856e-06,
+      "loss": 0.4979,
+      "step": 7723
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9013036702084408,
+      "learning_rate": 6.429349399657985e-06,
+      "loss": 0.4835,
+      "step": 7724
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.939866858909316,
+      "learning_rate": 6.4268905847900335e-06,
+      "loss": 0.505,
+      "step": 7725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.84843342037553,
+      "learning_rate": 6.424432017535391e-06,
+      "loss": 0.4922,
+      "step": 7726
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9328549244243041,
+      "learning_rate": 6.421973698064443e-06,
+      "loss": 0.5338,
+      "step": 7727
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9802417980606807,
+      "learning_rate": 6.419515626547543e-06,
+      "loss": 0.5675,
+      "step": 7728
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8478739144926289,
+      "learning_rate": 6.41705780315503e-06,
+      "loss": 0.4788,
+      "step": 7729
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8969066595806814,
+      "learning_rate": 6.414600228057237e-06,
+      "loss": 0.5265,
+      "step": 7730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.961301932687817,
+      "learning_rate": 6.41214290142447e-06,
+      "loss": 0.5468,
+      "step": 7731
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9124040441320103,
+      "learning_rate": 6.409685823427012e-06,
+      "loss": 0.5271,
+      "step": 7732
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9247331515893926,
+      "learning_rate": 6.407228994235146e-06,
+      "loss": 0.4851,
+      "step": 7733
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8764915994820545,
+      "learning_rate": 6.404772414019124e-06,
+      "loss": 0.5494,
+      "step": 7734
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8456362246353507,
+      "learning_rate": 6.40231608294918e-06,
+      "loss": 0.4967,
+      "step": 7735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8166377742541707,
+      "learning_rate": 6.399860001195546e-06,
+      "loss": 0.4253,
+      "step": 7736
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8533838510419213,
+      "learning_rate": 6.397404168928418e-06,
+      "loss": 0.5479,
+      "step": 7737
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.85629970661516,
+      "learning_rate": 6.394948586317984e-06,
+      "loss": 0.4668,
+      "step": 7738
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9226611146431447,
+      "learning_rate": 6.392493253534418e-06,
+      "loss": 0.4656,
+      "step": 7739
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9290966735283805,
+      "learning_rate": 6.39003817074787e-06,
+      "loss": 0.5628,
+      "step": 7740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9762580395364672,
+      "learning_rate": 6.387583338128471e-06,
+      "loss": 0.5663,
+      "step": 7741
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9107305782629377,
+      "learning_rate": 6.385128755846346e-06,
+      "loss": 0.5754,
+      "step": 7742
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8881321158916198,
+      "learning_rate": 6.382674424071593e-06,
+      "loss": 0.4683,
+      "step": 7743
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8133542061284578,
+      "learning_rate": 6.3802203429742884e-06,
+      "loss": 0.4527,
+      "step": 7744
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8277818013192928,
+      "learning_rate": 6.377766512724508e-06,
+      "loss": 0.4517,
+      "step": 7745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8749312544643651,
+      "learning_rate": 6.375312933492295e-06,
+      "loss": 0.493,
+      "step": 7746
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8712980229852738,
+      "learning_rate": 6.372859605447677e-06,
+      "loss": 0.4727,
+      "step": 7747
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8990200778286532,
+      "learning_rate": 6.370406528760675e-06,
+      "loss": 0.4291,
+      "step": 7748
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9048743821787182,
+      "learning_rate": 6.367953703601282e-06,
+      "loss": 0.5218,
+      "step": 7749
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8505330860159265,
+      "learning_rate": 6.36550113013947e-06,
+      "loss": 0.4798,
+      "step": 7750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9402198572741421,
+      "learning_rate": 6.3630488085452105e-06,
+      "loss": 0.5336,
+      "step": 7751
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9729586284896738,
+      "learning_rate": 6.360596738988443e-06,
+      "loss": 0.5871,
+      "step": 7752
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8969176170147588,
+      "learning_rate": 6.358144921639089e-06,
+      "loss": 0.549,
+      "step": 7753
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8885249861439682,
+      "learning_rate": 6.3556933566670656e-06,
+      "loss": 0.5263,
+      "step": 7754
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9578092824346056,
+      "learning_rate": 6.353242044242261e-06,
+      "loss": 0.5295,
+      "step": 7755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8080942797341487,
+      "learning_rate": 6.350790984534543e-06,
+      "loss": 0.4604,
+      "step": 7756
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9077725823304328,
+      "learning_rate": 6.348340177713776e-06,
+      "loss": 0.4722,
+      "step": 7757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9109428713389269,
+      "learning_rate": 6.3458896239497965e-06,
+      "loss": 0.5195,
+      "step": 7758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9463265119734536,
+      "learning_rate": 6.343439323412422e-06,
+      "loss": 0.4988,
+      "step": 7759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9059028499902411,
+      "learning_rate": 6.340989276271462e-06,
+      "loss": 0.4988,
+      "step": 7760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8568743425472681,
+      "learning_rate": 6.3385394826966975e-06,
+      "loss": 0.5055,
+      "step": 7761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.942925223145349,
+      "learning_rate": 6.336089942857899e-06,
+      "loss": 0.519,
+      "step": 7762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9032602036621082,
+      "learning_rate": 6.33364065692482e-06,
+      "loss": 0.4703,
+      "step": 7763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8792975094902231,
+      "learning_rate": 6.33119162506719e-06,
+      "loss": 0.4999,
+      "step": 7764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9501182269012853,
+      "learning_rate": 6.3287428474547256e-06,
+      "loss": 0.5199,
+      "step": 7765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9448877264739883,
+      "learning_rate": 6.326294324257127e-06,
+      "loss": 0.4954,
+      "step": 7766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8565240003077423,
+      "learning_rate": 6.32384605564407e-06,
+      "loss": 0.4208,
+      "step": 7767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8486055449790287,
+      "learning_rate": 6.321398041785225e-06,
+      "loss": 0.5115,
+      "step": 7768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9117738135586135,
+      "learning_rate": 6.318950282850231e-06,
+      "loss": 0.5145,
+      "step": 7769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8913146224618086,
+      "learning_rate": 6.3165027790087156e-06,
+      "loss": 0.4775,
+      "step": 7770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7365979070890599,
+      "learning_rate": 6.3140555304302915e-06,
+      "loss": 0.3933,
+      "step": 7771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9768804673095871,
+      "learning_rate": 6.311608537284553e-06,
+      "loss": 0.4881,
+      "step": 7772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.016542017230885,
+      "learning_rate": 6.309161799741064e-06,
+      "loss": 0.526,
+      "step": 7773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8925076055403189,
+      "learning_rate": 6.306715317969394e-06,
+      "loss": 0.4922,
+      "step": 7774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8267041800491832,
+      "learning_rate": 6.304269092139077e-06,
+      "loss": 0.4777,
+      "step": 7775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0650334644803787,
+      "learning_rate": 6.3018231224196305e-06,
+      "loss": 0.5939,
+      "step": 7776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8943182460099235,
+      "learning_rate": 6.299377408980563e-06,
+      "loss": 0.5152,
+      "step": 7777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.888006560881075,
+      "learning_rate": 6.296931951991358e-06,
+      "loss": 0.4804,
+      "step": 7778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9545552190672068,
+      "learning_rate": 6.2944867516214845e-06,
+      "loss": 0.5142,
+      "step": 7779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9929717798897725,
+      "learning_rate": 6.292041808040393e-06,
+      "loss": 0.5967,
+      "step": 7780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8714173763064299,
+      "learning_rate": 6.289597121417514e-06,
+      "loss": 0.4446,
+      "step": 7781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9078441721534038,
+      "learning_rate": 6.287152691922264e-06,
+      "loss": 0.5224,
+      "step": 7782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.996162174356343,
+      "learning_rate": 6.284708519724041e-06,
+      "loss": 0.5057,
+      "step": 7783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8774138005700622,
+      "learning_rate": 6.2822646049922185e-06,
+      "loss": 0.4563,
+      "step": 7784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8856112477861796,
+      "learning_rate": 6.279820947896163e-06,
+      "loss": 0.5064,
+      "step": 7785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9295493307030581,
+      "learning_rate": 6.277377548605217e-06,
+      "loss": 0.5247,
+      "step": 7786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8765121315853294,
+      "learning_rate": 6.274934407288704e-06,
+      "loss": 0.4566,
+      "step": 7787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9399025860311087,
+      "learning_rate": 6.2724915241159315e-06,
+      "loss": 0.5375,
+      "step": 7788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8557716593867082,
+      "learning_rate": 6.2700488992561925e-06,
+      "loss": 0.4419,
+      "step": 7789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9658120983337019,
+      "learning_rate": 6.267606532878754e-06,
+      "loss": 0.5243,
+      "step": 7790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.958679780460244,
+      "learning_rate": 6.265164425152872e-06,
+      "loss": 0.5179,
+      "step": 7791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9658173133446795,
+      "learning_rate": 6.262722576247785e-06,
+      "loss": 0.5464,
+      "step": 7792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8716878684699769,
+      "learning_rate": 6.260280986332707e-06,
+      "loss": 0.4832,
+      "step": 7793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9349477383809048,
+      "learning_rate": 6.257839655576839e-06,
+      "loss": 0.4898,
+      "step": 7794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9180139743849978,
+      "learning_rate": 6.255398584149366e-06,
+      "loss": 0.55,
+      "step": 7795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9231079645761651,
+      "learning_rate": 6.252957772219446e-06,
+      "loss": 0.5134,
+      "step": 7796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9223827123007751,
+      "learning_rate": 6.25051721995623e-06,
+      "loss": 0.5552,
+      "step": 7797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0099215371331756,
+      "learning_rate": 6.248076927528845e-06,
+      "loss": 0.5718,
+      "step": 7798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8269613087050963,
+      "learning_rate": 6.245636895106403e-06,
+      "loss": 0.4471,
+      "step": 7799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8713953691636022,
+      "learning_rate": 6.243197122857991e-06,
+      "loss": 0.5009,
+      "step": 7800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1097324395029895,
+      "learning_rate": 6.240757610952688e-06,
+      "loss": 0.4947,
+      "step": 7801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9103097554000824,
+      "learning_rate": 6.238318359559548e-06,
+      "loss": 0.4672,
+      "step": 7802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9089885001686926,
+      "learning_rate": 6.2358793688476085e-06,
+      "loss": 0.4802,
+      "step": 7803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9365668706441963,
+      "learning_rate": 6.233440638985889e-06,
+      "loss": 0.509,
+      "step": 7804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0305363078095187,
+      "learning_rate": 6.231002170143395e-06,
+      "loss": 0.5831,
+      "step": 7805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8677823936257921,
+      "learning_rate": 6.228563962489106e-06,
+      "loss": 0.5054,
+      "step": 7806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8956157333665741,
+      "learning_rate": 6.226126016191989e-06,
+      "loss": 0.4789,
+      "step": 7807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8960203160039782,
+      "learning_rate": 6.223688331420992e-06,
+      "loss": 0.4975,
+      "step": 7808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.809149890697131,
+      "learning_rate": 6.221250908345043e-06,
+      "loss": 0.4941,
+      "step": 7809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.943687691157147,
+      "learning_rate": 6.218813747133054e-06,
+      "loss": 0.4947,
+      "step": 7810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8036480414231428,
+      "learning_rate": 6.2163768479539224e-06,
+      "loss": 0.4487,
+      "step": 7811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9941996127619828,
+      "learning_rate": 6.2139402109765145e-06,
+      "loss": 0.4801,
+      "step": 7812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9008355927023047,
+      "learning_rate": 6.211503836369695e-06,
+      "loss": 0.4602,
+      "step": 7813
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9154622606923055,
+      "learning_rate": 6.209067724302298e-06,
+      "loss": 0.5009,
+      "step": 7814
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8444425530184236,
+      "learning_rate": 6.206631874943142e-06,
+      "loss": 0.4654,
+      "step": 7815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8723556650932853,
+      "learning_rate": 6.204196288461037e-06,
+      "loss": 0.4781,
+      "step": 7816
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.868682917707013,
+      "learning_rate": 6.2017609650247616e-06,
+      "loss": 0.4906,
+      "step": 7817
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9050595246873593,
+      "learning_rate": 6.19932590480308e-06,
+      "loss": 0.5166,
+      "step": 7818
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9716180252733098,
+      "learning_rate": 6.196891107964744e-06,
+      "loss": 0.5476,
+      "step": 7819
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8566243321600501,
+      "learning_rate": 6.194456574678481e-06,
+      "loss": 0.5152,
+      "step": 7820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9824046976752693,
+      "learning_rate": 6.192022305112999e-06,
+      "loss": 0.544,
+      "step": 7821
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9305953427523077,
+      "learning_rate": 6.189588299436997e-06,
+      "loss": 0.5713,
+      "step": 7822
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8070106072056493,
+      "learning_rate": 6.187154557819146e-06,
+      "loss": 0.4906,
+      "step": 7823
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9498681934374452,
+      "learning_rate": 6.184721080428098e-06,
+      "loss": 0.5671,
+      "step": 7824
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.820186983805444,
+      "learning_rate": 6.1822878674324995e-06,
+      "loss": 0.4313,
+      "step": 7825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9463991323254897,
+      "learning_rate": 6.179854919000965e-06,
+      "loss": 0.507,
+      "step": 7826
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9640186853955188,
+      "learning_rate": 6.177422235302093e-06,
+      "loss": 0.5476,
+      "step": 7827
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9492051998384852,
+      "learning_rate": 6.174989816504472e-06,
+      "loss": 0.4895,
+      "step": 7828
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8160052673463781,
+      "learning_rate": 6.172557662776665e-06,
+      "loss": 0.4485,
+      "step": 7829
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.832072497682488,
+      "learning_rate": 6.170125774287212e-06,
+      "loss": 0.4871,
+      "step": 7830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9821687629742838,
+      "learning_rate": 6.167694151204651e-06,
+      "loss": 0.5267,
+      "step": 7831
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8475254419513637,
+      "learning_rate": 6.165262793697486e-06,
+      "loss": 0.4286,
+      "step": 7832
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0397427201090619,
+      "learning_rate": 6.162831701934203e-06,
+      "loss": 0.5719,
+      "step": 7833
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9163483237401832,
+      "learning_rate": 6.160400876083283e-06,
+      "loss": 0.4926,
+      "step": 7834
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8156428618841962,
+      "learning_rate": 6.157970316313179e-06,
+      "loss": 0.4452,
+      "step": 7835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8687730375391427,
+      "learning_rate": 6.155540022792319e-06,
+      "loss": 0.543,
+      "step": 7836
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7983037783374483,
+      "learning_rate": 6.153109995689129e-06,
+      "loss": 0.4492,
+      "step": 7837
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9766388620578832,
+      "learning_rate": 6.150680235172004e-06,
+      "loss": 0.4917,
+      "step": 7838
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8463802416443318,
+      "learning_rate": 6.148250741409321e-06,
+      "loss": 0.4741,
+      "step": 7839
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.894432537815772,
+      "learning_rate": 6.145821514569449e-06,
+      "loss": 0.4996,
+      "step": 7840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8991657851225787,
+      "learning_rate": 6.143392554820726e-06,
+      "loss": 0.4578,
+      "step": 7841
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9054030277853536,
+      "learning_rate": 6.140963862331476e-06,
+      "loss": 0.4742,
+      "step": 7842
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8797646494203251,
+      "learning_rate": 6.13853543727001e-06,
+      "loss": 0.5034,
+      "step": 7843
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9484979833940284,
+      "learning_rate": 6.1361072798046155e-06,
+      "loss": 0.513,
+      "step": 7844
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9038890293909475,
+      "learning_rate": 6.1336793901035526e-06,
+      "loss": 0.4885,
+      "step": 7845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9380943990182781,
+      "learning_rate": 6.131251768335083e-06,
+      "loss": 0.5606,
+      "step": 7846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8525723409370285,
+      "learning_rate": 6.128824414667436e-06,
+      "loss": 0.481,
+      "step": 7847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8900496438478617,
+      "learning_rate": 6.126397329268817e-06,
+      "loss": 0.5184,
+      "step": 7848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.924435430812342,
+      "learning_rate": 6.123970512307433e-06,
+      "loss": 0.5249,
+      "step": 7849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0436953649832952,
+      "learning_rate": 6.121543963951453e-06,
+      "loss": 0.5956,
+      "step": 7850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.967753485718807,
+      "learning_rate": 6.119117684369033e-06,
+      "loss": 0.5847,
+      "step": 7851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9268366254061351,
+      "learning_rate": 6.116691673728319e-06,
+      "loss": 0.5063,
+      "step": 7852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0184125987174175,
+      "learning_rate": 6.114265932197427e-06,
+      "loss": 0.5789,
+      "step": 7853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9061381924635158,
+      "learning_rate": 6.111840459944456e-06,
+      "loss": 0.4939,
+      "step": 7854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8112643985533705,
+      "learning_rate": 6.109415257137496e-06,
+      "loss": 0.4545,
+      "step": 7855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8822508528963398,
+      "learning_rate": 6.1069903239446085e-06,
+      "loss": 0.4922,
+      "step": 7856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9376464388034186,
+      "learning_rate": 6.104565660533834e-06,
+      "loss": 0.5353,
+      "step": 7857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1579114301426976,
+      "learning_rate": 6.102141267073207e-06,
+      "loss": 0.586,
+      "step": 7858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9165096943932053,
+      "learning_rate": 6.099717143730735e-06,
+      "loss": 0.4738,
+      "step": 7859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9121514176411212,
+      "learning_rate": 6.0972932906744e-06,
+      "loss": 0.5212,
+      "step": 7860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9866894805187648,
+      "learning_rate": 6.094869708072182e-06,
+      "loss": 0.5184,
+      "step": 7861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9143118932789693,
+      "learning_rate": 6.092446396092029e-06,
+      "loss": 0.5419,
+      "step": 7862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8796210889513394,
+      "learning_rate": 6.0900233549018715e-06,
+      "loss": 0.5014,
+      "step": 7863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9600479927730894,
+      "learning_rate": 6.087600584669631e-06,
+      "loss": 0.5166,
+      "step": 7864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8497125065287534,
+      "learning_rate": 6.0851780855632005e-06,
+      "loss": 0.4715,
+      "step": 7865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8294242835797552,
+      "learning_rate": 6.082755857750451e-06,
+      "loss": 0.4082,
+      "step": 7866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8411002825465396,
+      "learning_rate": 6.080333901399252e-06,
+      "loss": 0.4416,
+      "step": 7867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0470602635891193,
+      "learning_rate": 6.077912216677435e-06,
+      "loss": 0.5246,
+      "step": 7868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9070343840890479,
+      "learning_rate": 6.075490803752818e-06,
+      "loss": 0.4806,
+      "step": 7869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9051386050346829,
+      "learning_rate": 6.073069662793213e-06,
+      "loss": 0.4905,
+      "step": 7870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8752448696259255,
+      "learning_rate": 6.070648793966396e-06,
+      "loss": 0.4699,
+      "step": 7871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8925887090586246,
+      "learning_rate": 6.068228197440129e-06,
+      "loss": 0.5167,
+      "step": 7872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.964748588002339,
+      "learning_rate": 6.065807873382163e-06,
+      "loss": 0.5169,
+      "step": 7873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9275308110947283,
+      "learning_rate": 6.063387821960224e-06,
+      "loss": 0.5562,
+      "step": 7874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8429438517721336,
+      "learning_rate": 6.060968043342013e-06,
+      "loss": 0.4673,
+      "step": 7875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8850378488006732,
+      "learning_rate": 6.058548537695225e-06,
+      "loss": 0.4752,
+      "step": 7876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9524802145799047,
+      "learning_rate": 6.056129305187528e-06,
+      "loss": 0.5744,
+      "step": 7877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9361706523073291,
+      "learning_rate": 6.053710345986568e-06,
+      "loss": 0.6133,
+      "step": 7878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9839367384874946,
+      "learning_rate": 6.051291660259984e-06,
+      "loss": 0.5047,
+      "step": 7879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7923563256741162,
+      "learning_rate": 6.048873248175387e-06,
+      "loss": 0.4301,
+      "step": 7880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8936789644649209,
+      "learning_rate": 6.046455109900364e-06,
+      "loss": 0.4995,
+      "step": 7881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.868032478797101,
+      "learning_rate": 6.044037245602498e-06,
+      "loss": 0.5239,
+      "step": 7882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.98929414338559,
+      "learning_rate": 6.041619655449345e-06,
+      "loss": 0.5709,
+      "step": 7883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8506950731385884,
+      "learning_rate": 6.039202339608432e-06,
+      "loss": 0.5399,
+      "step": 7884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8806249635423963,
+      "learning_rate": 6.03678529824729e-06,
+      "loss": 0.4679,
+      "step": 7885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8525185718459264,
+      "learning_rate": 6.03436853153341e-06,
+      "loss": 0.4969,
+      "step": 7886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9305266634188015,
+      "learning_rate": 6.031952039634269e-06,
+      "loss": 0.4984,
+      "step": 7887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.902035425375862,
+      "learning_rate": 6.029535822717336e-06,
+      "loss": 0.443,
+      "step": 7888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8470784591374255,
+      "learning_rate": 6.0271198809500495e-06,
+      "loss": 0.5191,
+      "step": 7889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.974557691623203,
+      "learning_rate": 6.024704214499828e-06,
+      "loss": 0.5018,
+      "step": 7890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.787087408132767,
+      "learning_rate": 6.02228882353408e-06,
+      "loss": 0.5078,
+      "step": 7891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8542335974387377,
+      "learning_rate": 6.019873708220187e-06,
+      "loss": 0.4908,
+      "step": 7892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9474423848025191,
+      "learning_rate": 6.0174588687255175e-06,
+      "loss": 0.5605,
+      "step": 7893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8575669120750463,
+      "learning_rate": 6.0150443052174165e-06,
+      "loss": 0.5251,
+      "step": 7894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8612373472694415,
+      "learning_rate": 6.012630017863207e-06,
+      "loss": 0.4617,
+      "step": 7895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8586947151074552,
+      "learning_rate": 6.010216006830204e-06,
+      "loss": 0.4587,
+      "step": 7896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8632841159755757,
+      "learning_rate": 6.007802272285693e-06,
+      "loss": 0.5079,
+      "step": 7897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8379444445979347,
+      "learning_rate": 6.0053888143969395e-06,
+      "loss": 0.4695,
+      "step": 7898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0050526074923891,
+      "learning_rate": 6.002975633331202e-06,
+      "loss": 0.4807,
+      "step": 7899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9495555894351163,
+      "learning_rate": 6.000562729255708e-06,
+      "loss": 0.5104,
+      "step": 7900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8893561306967496,
+      "learning_rate": 5.998150102337665e-06,
+      "loss": 0.4987,
+      "step": 7901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9049187018469663,
+      "learning_rate": 5.995737752744274e-06,
+      "loss": 0.3982,
+      "step": 7902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9645879156768508,
+      "learning_rate": 5.9933256806427056e-06,
+      "loss": 0.52,
+      "step": 7903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8483830405355759,
+      "learning_rate": 5.990913886200109e-06,
+      "loss": 0.5376,
+      "step": 7904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8176645223893816,
+      "learning_rate": 5.9885023695836285e-06,
+      "loss": 0.4675,
+      "step": 7905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8683650406947293,
+      "learning_rate": 5.986091130960374e-06,
+      "loss": 0.4619,
+      "step": 7906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8638758477258518,
+      "learning_rate": 5.983680170497441e-06,
+      "loss": 0.4918,
+      "step": 7907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9072700579393461,
+      "learning_rate": 5.981269488361915e-06,
+      "loss": 0.5199,
+      "step": 7908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8330566847213845,
+      "learning_rate": 5.978859084720847e-06,
+      "loss": 0.4817,
+      "step": 7909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.853580850835999,
+      "learning_rate": 5.9764489597412744e-06,
+      "loss": 0.4865,
+      "step": 7910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9350521083805324,
+      "learning_rate": 5.974039113590224e-06,
+      "loss": 0.5252,
+      "step": 7911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8370042228867535,
+      "learning_rate": 5.971629546434692e-06,
+      "loss": 0.4613,
+      "step": 7912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9258955551707068,
+      "learning_rate": 5.969220258441656e-06,
+      "loss": 0.464,
+      "step": 7913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.004101425937656,
+      "learning_rate": 5.966811249778084e-06,
+      "loss": 0.5862,
+      "step": 7914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0387617616671656,
+      "learning_rate": 5.964402520610915e-06,
+      "loss": 0.5051,
+      "step": 7915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9856894626875013,
+      "learning_rate": 5.961994071107067e-06,
+      "loss": 0.5665,
+      "step": 7916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9105291820348035,
+      "learning_rate": 5.959585901433453e-06,
+      "loss": 0.4826,
+      "step": 7917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8617063645840858,
+      "learning_rate": 5.957178011756952e-06,
+      "loss": 0.4357,
+      "step": 7918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9170271796296493,
+      "learning_rate": 5.954770402244425e-06,
+      "loss": 0.5619,
+      "step": 7919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9411333802566632,
+      "learning_rate": 5.9523630730627255e-06,
+      "loss": 0.4756,
+      "step": 7920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7903009111273711,
+      "learning_rate": 5.949956024378673e-06,
+      "loss": 0.5194,
+      "step": 7921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8558134713436262,
+      "learning_rate": 5.947549256359074e-06,
+      "loss": 0.493,
+      "step": 7922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8332628028313226,
+      "learning_rate": 5.94514276917072e-06,
+      "loss": 0.488,
+      "step": 7923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9005806018388123,
+      "learning_rate": 5.9427365629803756e-06,
+      "loss": 0.5588,
+      "step": 7924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9971919710537726,
+      "learning_rate": 5.940330637954783e-06,
+      "loss": 0.478,
+      "step": 7925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8456390472224813,
+      "learning_rate": 5.937924994260682e-06,
+      "loss": 0.4745,
+      "step": 7926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8541660700367373,
+      "learning_rate": 5.9355196320647745e-06,
+      "loss": 0.4937,
+      "step": 7927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8553699609380877,
+      "learning_rate": 5.933114551533749e-06,
+      "loss": 0.4747,
+      "step": 7928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9374042630967032,
+      "learning_rate": 5.930709752834281e-06,
+      "loss": 0.4991,
+      "step": 7929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9049318778548395,
+      "learning_rate": 5.928305236133016e-06,
+      "loss": 0.5303,
+      "step": 7930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9678237556822508,
+      "learning_rate": 5.925901001596586e-06,
+      "loss": 0.5392,
+      "step": 7931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8766681404745948,
+      "learning_rate": 5.923497049391605e-06,
+      "loss": 0.4992,
+      "step": 7932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0277711153649378,
+      "learning_rate": 5.9210933796846616e-06,
+      "loss": 0.5993,
+      "step": 7933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9128322651045907,
+      "learning_rate": 5.918689992642328e-06,
+      "loss": 0.5476,
+      "step": 7934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.835028262630942,
+      "learning_rate": 5.9162868884311596e-06,
+      "loss": 0.448,
+      "step": 7935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9091933878596423,
+      "learning_rate": 5.913884067217686e-06,
+      "loss": 0.4761,
+      "step": 7936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8631763279005022,
+      "learning_rate": 5.911481529168421e-06,
+      "loss": 0.4818,
+      "step": 7937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9481818857199612,
+      "learning_rate": 5.9090792744498625e-06,
+      "loss": 0.5641,
+      "step": 7938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8160053258398908,
+      "learning_rate": 5.9066773032284804e-06,
+      "loss": 0.448,
+      "step": 7939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9511732542896206,
+      "learning_rate": 5.9042756156707295e-06,
+      "loss": 0.5564,
+      "step": 7940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.929582203893334,
+      "learning_rate": 5.901874211943048e-06,
+      "loss": 0.5347,
+      "step": 7941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8946384425547035,
+      "learning_rate": 5.899473092211847e-06,
+      "loss": 0.5131,
+      "step": 7942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8740876502947456,
+      "learning_rate": 5.897072256643522e-06,
+      "loss": 0.4932,
+      "step": 7943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8753668882927068,
+      "learning_rate": 5.894671705404453e-06,
+      "loss": 0.495,
+      "step": 7944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9465339505107648,
+      "learning_rate": 5.892271438660993e-06,
+      "loss": 0.5105,
+      "step": 7945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8751984352332609,
+      "learning_rate": 5.889871456579477e-06,
+      "loss": 0.5559,
+      "step": 7946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0290060841753401,
+      "learning_rate": 5.887471759326223e-06,
+      "loss": 0.5856,
+      "step": 7947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8780230357458085,
+      "learning_rate": 5.885072347067531e-06,
+      "loss": 0.5009,
+      "step": 7948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.880023849643744,
+      "learning_rate": 5.882673219969673e-06,
+      "loss": 0.5137,
+      "step": 7949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.865486218156423,
+      "learning_rate": 5.880274378198909e-06,
+      "loss": 0.4774,
+      "step": 7950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8907322492856702,
+      "learning_rate": 5.877875821921479e-06,
+      "loss": 0.4895,
+      "step": 7951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8918944661195479,
+      "learning_rate": 5.875477551303596e-06,
+      "loss": 0.5237,
+      "step": 7952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9662505920258414,
+      "learning_rate": 5.873079566511459e-06,
+      "loss": 0.5327,
+      "step": 7953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9372929223735276,
+      "learning_rate": 5.870681867711252e-06,
+      "loss": 0.4707,
+      "step": 7954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9801069798337949,
+      "learning_rate": 5.868284455069124e-06,
+      "loss": 0.4903,
+      "step": 7955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8265909750429461,
+      "learning_rate": 5.865887328751221e-06,
+      "loss": 0.5053,
+      "step": 7956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8773827747816701,
+      "learning_rate": 5.86349048892366e-06,
+      "loss": 0.5323,
+      "step": 7957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9125462407522994,
+      "learning_rate": 5.8610939357525365e-06,
+      "loss": 0.457,
+      "step": 7958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9009283341814897,
+      "learning_rate": 5.8586976694039325e-06,
+      "loss": 0.4907,
+      "step": 7959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8862610572080457,
+      "learning_rate": 5.856301690043909e-06,
+      "loss": 0.4928,
+      "step": 7960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8271397349442127,
+      "learning_rate": 5.8539059978385e-06,
+      "loss": 0.4825,
+      "step": 7961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0027140353575406,
+      "learning_rate": 5.851510592953729e-06,
+      "loss": 0.5539,
+      "step": 7962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9227278978093961,
+      "learning_rate": 5.849115475555596e-06,
+      "loss": 0.4464,
+      "step": 7963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.872548999788522,
+      "learning_rate": 5.846720645810073e-06,
+      "loss": 0.4953,
+      "step": 7964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8580393334113791,
+      "learning_rate": 5.844326103883131e-06,
+      "loss": 0.4775,
+      "step": 7965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0315902936759194,
+      "learning_rate": 5.841931849940704e-06,
+      "loss": 0.494,
+      "step": 7966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0272260657495227,
+      "learning_rate": 5.839537884148707e-06,
+      "loss": 0.5617,
+      "step": 7967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8998302210454385,
+      "learning_rate": 5.837144206673049e-06,
+      "loss": 0.4609,
+      "step": 7968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9546931742771414,
+      "learning_rate": 5.834750817679606e-06,
+      "loss": 0.5295,
+      "step": 7969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9181160691313232,
+      "learning_rate": 5.832357717334229e-06,
+      "loss": 0.5299,
+      "step": 7970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.866694492393031,
+      "learning_rate": 5.829964905802774e-06,
+      "loss": 0.4898,
+      "step": 7971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.829496100447018,
+      "learning_rate": 5.827572383251048e-06,
+      "loss": 0.5281,
+      "step": 7972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8751261025322412,
+      "learning_rate": 5.825180149844856e-06,
+      "loss": 0.4339,
+      "step": 7973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8704829205465353,
+      "learning_rate": 5.822788205749974e-06,
+      "loss": 0.4777,
+      "step": 7974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.849680191816542,
+      "learning_rate": 5.82039655113217e-06,
+      "loss": 0.4442,
+      "step": 7975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8376653229107791,
+      "learning_rate": 5.8180051861571695e-06,
+      "loss": 0.4852,
+      "step": 7976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8389294368790021,
+      "learning_rate": 5.815614110990708e-06,
+      "loss": 0.462,
+      "step": 7977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8807585814155449,
+      "learning_rate": 5.813223325798473e-06,
+      "loss": 0.5042,
+      "step": 7978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.000691488112414,
+      "learning_rate": 5.810832830746147e-06,
+      "loss": 0.4713,
+      "step": 7979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8532742444248469,
+      "learning_rate": 5.8084426259993905e-06,
+      "loss": 0.4491,
+      "step": 7980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9100918922284562,
+      "learning_rate": 5.8060527117238475e-06,
+      "loss": 0.5095,
+      "step": 7981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8929274143229312,
+      "learning_rate": 5.803663088085123e-06,
+      "loss": 0.5257,
+      "step": 7982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8455248193661312,
+      "learning_rate": 5.801273755248831e-06,
+      "loss": 0.477,
+      "step": 7983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8811988221328768,
+      "learning_rate": 5.798884713380542e-06,
+      "loss": 0.5042,
+      "step": 7984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.967260471731767,
+      "learning_rate": 5.796495962645814e-06,
+      "loss": 0.5802,
+      "step": 7985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.98700331385454,
+      "learning_rate": 5.794107503210187e-06,
+      "loss": 0.5205,
+      "step": 7986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.902464167796891,
+      "learning_rate": 5.791719335239185e-06,
+      "loss": 0.5258,
+      "step": 7987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.852618831364867,
+      "learning_rate": 5.7893314588982905e-06,
+      "loss": 0.4333,
+      "step": 7988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9179412089940903,
+      "learning_rate": 5.7869438743529994e-06,
+      "loss": 0.5615,
+      "step": 7989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9335365738514484,
+      "learning_rate": 5.784556581768757e-06,
+      "loss": 0.5715,
+      "step": 7990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7787335792129625,
+      "learning_rate": 5.782169581311006e-06,
+      "loss": 0.478,
+      "step": 7991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9464878407470988,
+      "learning_rate": 5.7797828731451596e-06,
+      "loss": 0.4992,
+      "step": 7992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9479587984475946,
+      "learning_rate": 5.7773964574366185e-06,
+      "loss": 0.5334,
+      "step": 7993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8369870158085331,
+      "learning_rate": 5.7750103343507565e-06,
+      "loss": 0.4922,
+      "step": 7994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8797202532805338,
+      "learning_rate": 5.772624504052935e-06,
+      "loss": 0.5185,
+      "step": 7995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8525431115226239,
+      "learning_rate": 5.770238966708482e-06,
+      "loss": 0.4396,
+      "step": 7996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8919277953759788,
+      "learning_rate": 5.767853722482717e-06,
+      "loss": 0.4865,
+      "step": 7997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.924505401141746,
+      "learning_rate": 5.765468771540934e-06,
+      "loss": 0.5014,
+      "step": 7998
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9132203193112512,
+      "learning_rate": 5.763084114048409e-06,
+      "loss": 0.4474,
+      "step": 7999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9793395084081031,
+      "learning_rate": 5.7606997501703975e-06,
+      "loss": 0.5358,
+      "step": 8000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9415267497989768,
+      "learning_rate": 5.758315680072137e-06,
+      "loss": 0.4871,
+      "step": 8001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.899446816115677,
+      "learning_rate": 5.755931903918835e-06,
+      "loss": 0.4913,
+      "step": 8002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8433015727474659,
+      "learning_rate": 5.753548421875686e-06,
+      "loss": 0.4995,
+      "step": 8003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.928114233175902,
+      "learning_rate": 5.751165234107864e-06,
+      "loss": 0.5294,
+      "step": 8004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7961058992253185,
+      "learning_rate": 5.748782340780523e-06,
+      "loss": 0.4393,
+      "step": 8005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9787936172926667,
+      "learning_rate": 5.746399742058796e-06,
+      "loss": 0.5387,
+      "step": 8006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8751511295227653,
+      "learning_rate": 5.744017438107796e-06,
+      "loss": 0.4633,
+      "step": 8007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9077415186497616,
+      "learning_rate": 5.741635429092611e-06,
+      "loss": 0.5377,
+      "step": 8008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8952564586058097,
+      "learning_rate": 5.7392537151783125e-06,
+      "loss": 0.5049,
+      "step": 8009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8646877275625106,
+      "learning_rate": 5.736872296529952e-06,
+      "loss": 0.4616,
+      "step": 8010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.022240202103064,
+      "learning_rate": 5.734491173312559e-06,
+      "loss": 0.5524,
+      "step": 8011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8915803775655604,
+      "learning_rate": 5.732110345691146e-06,
+      "loss": 0.511,
+      "step": 8012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8479179390848613,
+      "learning_rate": 5.7297298138307e-06,
+      "loss": 0.4507,
+      "step": 8013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9476076758828081,
+      "learning_rate": 5.727349577896194e-06,
+      "loss": 0.4865,
+      "step": 8014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8676077389176223,
+      "learning_rate": 5.724969638052569e-06,
+      "loss": 0.4834,
+      "step": 8015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9255473966779431,
+      "learning_rate": 5.722589994464754e-06,
+      "loss": 0.5681,
+      "step": 8016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8470769391341288,
+      "learning_rate": 5.72021064729766e-06,
+      "loss": 0.5295,
+      "step": 8017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8572510515489944,
+      "learning_rate": 5.717831596716173e-06,
+      "loss": 0.487,
+      "step": 8018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0118769672041892,
+      "learning_rate": 5.715452842885157e-06,
+      "loss": 0.5201,
+      "step": 8019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9132906756275612,
+      "learning_rate": 5.713074385969457e-06,
+      "loss": 0.5172,
+      "step": 8020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8406249617881854,
+      "learning_rate": 5.710696226133905e-06,
+      "loss": 0.4563,
+      "step": 8021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9288896517388993,
+      "learning_rate": 5.708318363543297e-06,
+      "loss": 0.4847,
+      "step": 8022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9714000918656338,
+      "learning_rate": 5.705940798362417e-06,
+      "loss": 0.5423,
+      "step": 8023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7738297426860237,
+      "learning_rate": 5.703563530756033e-06,
+      "loss": 0.4568,
+      "step": 8024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9015248782470349,
+      "learning_rate": 5.701186560888885e-06,
+      "loss": 0.4834,
+      "step": 8025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9615545542578514,
+      "learning_rate": 5.698809888925696e-06,
+      "loss": 0.4691,
+      "step": 8026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8640647567153519,
+      "learning_rate": 5.696433515031169e-06,
+      "loss": 0.4209,
+      "step": 8027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8546615520462587,
+      "learning_rate": 5.694057439369979e-06,
+      "loss": 0.4583,
+      "step": 8028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8921814771572696,
+      "learning_rate": 5.6916816621067895e-06,
+      "loss": 0.5259,
+      "step": 8029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8644158011998065,
+      "learning_rate": 5.689306183406238e-06,
+      "loss": 0.5409,
+      "step": 8030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9855746769335014,
+      "learning_rate": 5.686931003432945e-06,
+      "loss": 0.498,
+      "step": 8031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8982014572232642,
+      "learning_rate": 5.684556122351508e-06,
+      "loss": 0.4861,
+      "step": 8032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.925016604280845,
+      "learning_rate": 5.682181540326503e-06,
+      "loss": 0.4698,
+      "step": 8033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9717081180975732,
+      "learning_rate": 5.679807257522493e-06,
+      "loss": 0.4723,
+      "step": 8034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8842071042921347,
+      "learning_rate": 5.677433274104003e-06,
+      "loss": 0.546,
+      "step": 8035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9502961755883366,
+      "learning_rate": 5.675059590235553e-06,
+      "loss": 0.5469,
+      "step": 8036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9981482851768781,
+      "learning_rate": 5.672686206081638e-06,
+      "loss": 0.5291,
+      "step": 8037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9835649935761642,
+      "learning_rate": 5.67031312180673e-06,
+      "loss": 0.5477,
+      "step": 8038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8502872797613393,
+      "learning_rate": 5.6679403375752816e-06,
+      "loss": 0.5428,
+      "step": 8039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9344665940312018,
+      "learning_rate": 5.66556785355173e-06,
+      "loss": 0.5195,
+      "step": 8040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9192075507681627,
+      "learning_rate": 5.663195669900479e-06,
+      "loss": 0.4879,
+      "step": 8041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.986064051073003,
+      "learning_rate": 5.6608237867859184e-06,
+      "loss": 0.5824,
+      "step": 8042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8066618730198377,
+      "learning_rate": 5.6584522043724226e-06,
+      "loss": 0.4839,
+      "step": 8043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9818638450965491,
+      "learning_rate": 5.656080922824337e-06,
+      "loss": 0.5366,
+      "step": 8044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8365237997024432,
+      "learning_rate": 5.65370994230599e-06,
+      "loss": 0.4576,
+      "step": 8045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9445525770789083,
+      "learning_rate": 5.651339262981694e-06,
+      "loss": 0.5432,
+      "step": 8046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8473995922041537,
+      "learning_rate": 5.648968885015726e-06,
+      "loss": 0.4734,
+      "step": 8047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9149444795504126,
+      "learning_rate": 5.646598808572355e-06,
+      "loss": 0.5067,
+      "step": 8048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9382812760027435,
+      "learning_rate": 5.644229033815823e-06,
+      "loss": 0.5029,
+      "step": 8049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9131763915953596,
+      "learning_rate": 5.641859560910356e-06,
+      "loss": 0.5207,
+      "step": 8050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9083095218043912,
+      "learning_rate": 5.639490390020158e-06,
+      "loss": 0.4829,
+      "step": 8051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9217744465084563,
+      "learning_rate": 5.637121521309411e-06,
+      "loss": 0.5367,
+      "step": 8052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8950244905018768,
+      "learning_rate": 5.634752954942264e-06,
+      "loss": 0.4848,
+      "step": 8053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9220101284527779,
+      "learning_rate": 5.632384691082874e-06,
+      "loss": 0.4886,
+      "step": 8054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8510786318647385,
+      "learning_rate": 5.630016729895346e-06,
+      "loss": 0.4604,
+      "step": 8055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0006639059946054,
+      "learning_rate": 5.627649071543784e-06,
+      "loss": 0.5222,
+      "step": 8056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9580343958505725,
+      "learning_rate": 5.6252817161922616e-06,
+      "loss": 0.5535,
+      "step": 8057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9379452625036485,
+      "learning_rate": 5.6229146640048415e-06,
+      "loss": 0.5168,
+      "step": 8058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.872992477532868,
+      "learning_rate": 5.620547915145542e-06,
+      "loss": 0.5281,
+      "step": 8059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7956540694611413,
+      "learning_rate": 5.618181469778399e-06,
+      "loss": 0.4395,
+      "step": 8060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8921079894021173,
+      "learning_rate": 5.615815328067387e-06,
+      "loss": 0.4747,
+      "step": 8061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8262492322669378,
+      "learning_rate": 5.613449490176484e-06,
+      "loss": 0.4567,
+      "step": 8062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8583148399648072,
+      "learning_rate": 5.6110839562696404e-06,
+      "loss": 0.4694,
+      "step": 8063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9328997683673895,
+      "learning_rate": 5.608718726510791e-06,
+      "loss": 0.5111,
+      "step": 8064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8097909207413837,
+      "learning_rate": 5.60635380106383e-06,
+      "loss": 0.4543,
+      "step": 8065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.863117123650207,
+      "learning_rate": 5.603989180092661e-06,
+      "loss": 0.4795,
+      "step": 8066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9068547480911131,
+      "learning_rate": 5.6016248637611395e-06,
+      "loss": 0.4863,
+      "step": 8067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8251512618816547,
+      "learning_rate": 5.5992608522331126e-06,
+      "loss": 0.4601,
+      "step": 8068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9337837799868443,
+      "learning_rate": 5.596897145672407e-06,
+      "loss": 0.537,
+      "step": 8069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8894318690847395,
+      "learning_rate": 5.594533744242825e-06,
+      "loss": 0.4685,
+      "step": 8070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8633650959858853,
+      "learning_rate": 5.5921706481081405e-06,
+      "loss": 0.536,
+      "step": 8071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9743246388023774,
+      "learning_rate": 5.589807857432128e-06,
+      "loss": 0.5396,
+      "step": 8072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9548473977649928,
+      "learning_rate": 5.587445372378515e-06,
+      "loss": 0.5391,
+      "step": 8073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8957082635019312,
+      "learning_rate": 5.585083193111025e-06,
+      "loss": 0.4466,
+      "step": 8074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9149137254325227,
+      "learning_rate": 5.582721319793351e-06,
+      "loss": 0.4543,
+      "step": 8075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9338244908287366,
+      "learning_rate": 5.580359752589178e-06,
+      "loss": 0.4991,
+      "step": 8076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9282335254284353,
+      "learning_rate": 5.5779984916621455e-06,
+      "loss": 0.5303,
+      "step": 8077
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9862482780242752,
+      "learning_rate": 5.575637537175902e-06,
+      "loss": 0.4957,
+      "step": 8078
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7528909278795783,
+      "learning_rate": 5.57327688929405e-06,
+      "loss": 0.4195,
+      "step": 8079
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9557139196501997,
+      "learning_rate": 5.570916548180183e-06,
+      "loss": 0.4914,
+      "step": 8080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8315133258006441,
+      "learning_rate": 5.568556513997869e-06,
+      "loss": 0.4918,
+      "step": 8081
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0828275890412324,
+      "learning_rate": 5.566196786910665e-06,
+      "loss": 0.564,
+      "step": 8082
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9521834237171527,
+      "learning_rate": 5.56383736708208e-06,
+      "loss": 0.4365,
+      "step": 8083
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8505018319958945,
+      "learning_rate": 5.561478254675639e-06,
+      "loss": 0.4845,
+      "step": 8084
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9605289024212322,
+      "learning_rate": 5.559119449854815e-06,
+      "loss": 0.5306,
+      "step": 8085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0344634196486486,
+      "learning_rate": 5.556760952783073e-06,
+      "loss": 0.4886,
+      "step": 8086
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9630866882564898,
+      "learning_rate": 5.554402763623857e-06,
+      "loss": 0.4603,
+      "step": 8087
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8845755841128609,
+      "learning_rate": 5.55204488254059e-06,
+      "loss": 0.4963,
+      "step": 8088
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.927336189724174,
+      "learning_rate": 5.549687309696658e-06,
+      "loss": 0.5091,
+      "step": 8089
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9453205589847619,
+      "learning_rate": 5.547330045255458e-06,
+      "loss": 0.5375,
+      "step": 8090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9241048097039017,
+      "learning_rate": 5.5449730893803326e-06,
+      "loss": 0.4709,
+      "step": 8091
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8701604477982022,
+      "learning_rate": 5.542616442234618e-06,
+      "loss": 0.4896,
+      "step": 8092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0247372373451231,
+      "learning_rate": 5.5402601039816315e-06,
+      "loss": 0.5078,
+      "step": 8093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9368300927082287,
+      "learning_rate": 5.537904074784668e-06,
+      "loss": 0.5267,
+      "step": 8094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8953371671011336,
+      "learning_rate": 5.5355483548069866e-06,
+      "loss": 0.4929,
+      "step": 8095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8632314786605364,
+      "learning_rate": 5.533192944211852e-06,
+      "loss": 0.4403,
+      "step": 8096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9249710661785733,
+      "learning_rate": 5.53083784316248e-06,
+      "loss": 0.5305,
+      "step": 8097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9478487611876848,
+      "learning_rate": 5.528483051822082e-06,
+      "loss": 0.485,
+      "step": 8098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1189668918359563,
+      "learning_rate": 5.526128570353842e-06,
+      "loss": 0.5108,
+      "step": 8099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.011156273732329,
+      "learning_rate": 5.523774398920927e-06,
+      "loss": 0.5158,
+      "step": 8100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8582598832505969,
+      "learning_rate": 5.521420537686468e-06,
+      "loss": 0.5109,
+      "step": 8101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.890431587765962,
+      "learning_rate": 5.519066986813602e-06,
+      "loss": 0.5406,
+      "step": 8102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9234301484449461,
+      "learning_rate": 5.5167137464654155e-06,
+      "loss": 0.5037,
+      "step": 8103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8827748934064557,
+      "learning_rate": 5.514360816804989e-06,
+      "loss": 0.5078,
+      "step": 8104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9646521521805869,
+      "learning_rate": 5.512008197995379e-06,
+      "loss": 0.5017,
+      "step": 8105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9690885937124468,
+      "learning_rate": 5.5096558901996235e-06,
+      "loss": 0.5432,
+      "step": 8106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8448217293276843,
+      "learning_rate": 5.507303893580724e-06,
+      "loss": 0.4992,
+      "step": 8107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9631252639189662,
+      "learning_rate": 5.5049522083016895e-06,
+      "loss": 0.5089,
+      "step": 8108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9643236838561009,
+      "learning_rate": 5.502600834525475e-06,
+      "loss": 0.5508,
+      "step": 8109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9953018292367835,
+      "learning_rate": 5.500249772415033e-06,
+      "loss": 0.5355,
+      "step": 8110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8863125003391487,
+      "learning_rate": 5.49789902213329e-06,
+      "loss": 0.543,
+      "step": 8111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8268057293234355,
+      "learning_rate": 5.495548583843155e-06,
+      "loss": 0.4978,
+      "step": 8112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8258851400525931,
+      "learning_rate": 5.493198457707503e-06,
+      "loss": 0.4619,
+      "step": 8113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9138777270458119,
+      "learning_rate": 5.490848643889205e-06,
+      "loss": 0.4819,
+      "step": 8114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8789485096300262,
+      "learning_rate": 5.488499142551094e-06,
+      "loss": 0.4584,
+      "step": 8115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9067069193255343,
+      "learning_rate": 5.48614995385599e-06,
+      "loss": 0.5312,
+      "step": 8116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9639711693883865,
+      "learning_rate": 5.48380107796669e-06,
+      "loss": 0.474,
+      "step": 8117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9871386684830713,
+      "learning_rate": 5.481452515045974e-06,
+      "loss": 0.5496,
+      "step": 8118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.856246256936417,
+      "learning_rate": 5.479104265256583e-06,
+      "loss": 0.4469,
+      "step": 8119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9666843510279574,
+      "learning_rate": 5.476756328761264e-06,
+      "loss": 0.4889,
+      "step": 8120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8589689713812814,
+      "learning_rate": 5.474408705722716e-06,
+      "loss": 0.4909,
+      "step": 8121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8871792197882404,
+      "learning_rate": 5.47206139630363e-06,
+      "loss": 0.4712,
+      "step": 8122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8684447485140943,
+      "learning_rate": 5.469714400666673e-06,
+      "loss": 0.4664,
+      "step": 8123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9112182769895082,
+      "learning_rate": 5.467367718974492e-06,
+      "loss": 0.5288,
+      "step": 8124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8856249596176711,
+      "learning_rate": 5.465021351389702e-06,
+      "loss": 0.5061,
+      "step": 8125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0244913612133435,
+      "learning_rate": 5.462675298074918e-06,
+      "loss": 0.5295,
+      "step": 8126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8781385330939273,
+      "learning_rate": 5.460329559192705e-06,
+      "loss": 0.4622,
+      "step": 8127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9235091415547136,
+      "learning_rate": 5.4579841349056285e-06,
+      "loss": 0.548,
+      "step": 8128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8837638906811883,
+      "learning_rate": 5.455639025376223e-06,
+      "loss": 0.4918,
+      "step": 8129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9490852904400812,
+      "learning_rate": 5.453294230767005e-06,
+      "loss": 0.5258,
+      "step": 8130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8936978817795787,
+      "learning_rate": 5.450949751240456e-06,
+      "loss": 0.5531,
+      "step": 8131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8136657702496216,
+      "learning_rate": 5.448605586959063e-06,
+      "loss": 0.419,
+      "step": 8132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8734233583149454,
+      "learning_rate": 5.446261738085261e-06,
+      "loss": 0.4688,
+      "step": 8133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9808667220509426,
+      "learning_rate": 5.443918204781482e-06,
+      "loss": 0.5776,
+      "step": 8134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9728198474455235,
+      "learning_rate": 5.44157498721013e-06,
+      "loss": 0.4703,
+      "step": 8135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8547436340944682,
+      "learning_rate": 5.439232085533592e-06,
+      "loss": 0.4995,
+      "step": 8136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.037319351927126,
+      "learning_rate": 5.436889499914218e-06,
+      "loss": 0.5527,
+      "step": 8137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9336300856188522,
+      "learning_rate": 5.43454723051436e-06,
+      "loss": 0.5334,
+      "step": 8138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.91337608051943,
+      "learning_rate": 5.432205277496327e-06,
+      "loss": 0.4993,
+      "step": 8139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9486914224331876,
+      "learning_rate": 5.429863641022416e-06,
+      "loss": 0.5073,
+      "step": 8140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9595276005602759,
+      "learning_rate": 5.427522321254901e-06,
+      "loss": 0.5485,
+      "step": 8141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9541951421866232,
+      "learning_rate": 5.425181318356035e-06,
+      "loss": 0.5148,
+      "step": 8142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8624392948766714,
+      "learning_rate": 5.4228406324880434e-06,
+      "loss": 0.503,
+      "step": 8143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9007306472206367,
+      "learning_rate": 5.420500263813141e-06,
+      "loss": 0.5872,
+      "step": 8144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9184533974463028,
+      "learning_rate": 5.418160212493501e-06,
+      "loss": 0.5086,
+      "step": 8145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9301255956721959,
+      "learning_rate": 5.415820478691301e-06,
+      "loss": 0.6116,
+      "step": 8146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8777358719757851,
+      "learning_rate": 5.413481062568672e-06,
+      "loss": 0.4959,
+      "step": 8147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.944307831775758,
+      "learning_rate": 5.411141964287737e-06,
+      "loss": 0.5362,
+      "step": 8148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9121351906003211,
+      "learning_rate": 5.408803184010593e-06,
+      "loss": 0.544,
+      "step": 8149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8415452275491403,
+      "learning_rate": 5.406464721899323e-06,
+      "loss": 0.497,
+      "step": 8150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9062392888284309,
+      "learning_rate": 5.404126578115962e-06,
+      "loss": 0.5131,
+      "step": 8151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.98075689579646,
+      "learning_rate": 5.401788752822562e-06,
+      "loss": 0.5472,
+      "step": 8152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.970371527601638,
+      "learning_rate": 5.399451246181118e-06,
+      "loss": 0.5397,
+      "step": 8153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9526438115570307,
+      "learning_rate": 5.397114058353623e-06,
+      "loss": 0.5376,
+      "step": 8154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8734375021626799,
+      "learning_rate": 5.39477718950204e-06,
+      "loss": 0.4983,
+      "step": 8155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9938007613913521,
+      "learning_rate": 5.3924406397883174e-06,
+      "loss": 0.536,
+      "step": 8156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0030332858642188,
+      "learning_rate": 5.390104409374364e-06,
+      "loss": 0.5175,
+      "step": 8157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9435118882877143,
+      "learning_rate": 5.3877684984220945e-06,
+      "loss": 0.5013,
+      "step": 8158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9512703874552919,
+      "learning_rate": 5.385432907093371e-06,
+      "loss": 0.5388,
+      "step": 8159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.873374334708962,
+      "learning_rate": 5.383097635550057e-06,
+      "loss": 0.5086,
+      "step": 8160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0269739332825232,
+      "learning_rate": 5.380762683953978e-06,
+      "loss": 0.464,
+      "step": 8161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8972458312913809,
+      "learning_rate": 5.37842805246695e-06,
+      "loss": 0.5318,
+      "step": 8162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.987479872006018,
+      "learning_rate": 5.376093741250758e-06,
+      "loss": 0.522,
+      "step": 8163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8746417642079718,
+      "learning_rate": 5.373759750467173e-06,
+      "loss": 0.4816,
+      "step": 8164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9936009159063266,
+      "learning_rate": 5.371426080277928e-06,
+      "loss": 0.5823,
+      "step": 8165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8921448928229553,
+      "learning_rate": 5.369092730844752e-06,
+      "loss": 0.5245,
+      "step": 8166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8386162750995261,
+      "learning_rate": 5.366759702329343e-06,
+      "loss": 0.4778,
+      "step": 8167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8330787084771702,
+      "learning_rate": 5.364426994893375e-06,
+      "loss": 0.483,
+      "step": 8168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9387651680510244,
+      "learning_rate": 5.362094608698505e-06,
+      "loss": 0.5306,
+      "step": 8169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8747296879461391,
+      "learning_rate": 5.3597625439063685e-06,
+      "loss": 0.486,
+      "step": 8170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9628154040176343,
+      "learning_rate": 5.357430800678568e-06,
+      "loss": 0.5462,
+      "step": 8171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8612974069576437,
+      "learning_rate": 5.3550993791766955e-06,
+      "loss": 0.4286,
+      "step": 8172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.026048945583325,
+      "learning_rate": 5.352768279562315e-06,
+      "loss": 0.5305,
+      "step": 8173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.852733927613396,
+      "learning_rate": 5.350437501996972e-06,
+      "loss": 0.5464,
+      "step": 8174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9309496207123703,
+      "learning_rate": 5.348107046642186e-06,
+      "loss": 0.535,
+      "step": 8175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9011431585683592,
+      "learning_rate": 5.345776913659458e-06,
+      "loss": 0.4597,
+      "step": 8176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9514530227899883,
+      "learning_rate": 5.343447103210257e-06,
+      "loss": 0.4818,
+      "step": 8177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9556001417671066,
+      "learning_rate": 5.341117615456044e-06,
+      "loss": 0.482,
+      "step": 8178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9221886068743322,
+      "learning_rate": 5.338788450558246e-06,
+      "loss": 0.4494,
+      "step": 8179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9579423098386334,
+      "learning_rate": 5.336459608678275e-06,
+      "loss": 0.5036,
+      "step": 8180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9695762173510455,
+      "learning_rate": 5.334131089977516e-06,
+      "loss": 0.5367,
+      "step": 8181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9761452544087317,
+      "learning_rate": 5.331802894617333e-06,
+      "loss": 0.5363,
+      "step": 8182
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9141851363212332,
+      "learning_rate": 5.329475022759074e-06,
+      "loss": 0.5249,
+      "step": 8183
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9093173516423304,
+      "learning_rate": 5.32714747456405e-06,
+      "loss": 0.4814,
+      "step": 8184
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8240847030191094,
+      "learning_rate": 5.324820250193559e-06,
+      "loss": 0.4586,
+      "step": 8185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9576069962900918,
+      "learning_rate": 5.322493349808878e-06,
+      "loss": 0.5164,
+      "step": 8186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8550129467805008,
+      "learning_rate": 5.32016677357126e-06,
+      "loss": 0.4515,
+      "step": 8187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.989239559909075,
+      "learning_rate": 5.3178405216419325e-06,
+      "loss": 0.5694,
+      "step": 8188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8402449147196775,
+      "learning_rate": 5.31551459418211e-06,
+      "loss": 0.4815,
+      "step": 8189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.929650535088433,
+      "learning_rate": 5.313188991352964e-06,
+      "loss": 0.4851,
+      "step": 8190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8608747172187374,
+      "learning_rate": 5.310863713315666e-06,
+      "loss": 0.5316,
+      "step": 8191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8601414518555403,
+      "learning_rate": 5.308538760231352e-06,
+      "loss": 0.48,
+      "step": 8192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8621843116643283,
+      "learning_rate": 5.306214132261141e-06,
+      "loss": 0.5108,
+      "step": 8193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9873625363606519,
+      "learning_rate": 5.303889829566128e-06,
+      "loss": 0.5105,
+      "step": 8194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9533788876320213,
+      "learning_rate": 5.301565852307388e-06,
+      "loss": 0.562,
+      "step": 8195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9941344908746977,
+      "learning_rate": 5.299242200645959e-06,
+      "loss": 0.5008,
+      "step": 8196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8709593447651172,
+      "learning_rate": 5.296918874742882e-06,
+      "loss": 0.3924,
+      "step": 8197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.889954834129633,
+      "learning_rate": 5.294595874759154e-06,
+      "loss": 0.483,
+      "step": 8198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8876593394083251,
+      "learning_rate": 5.292273200855758e-06,
+      "loss": 0.4491,
+      "step": 8199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8524918627473157,
+      "learning_rate": 5.2899508531936526e-06,
+      "loss": 0.5124,
+      "step": 8200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9321934019395298,
+      "learning_rate": 5.2876288319337785e-06,
+      "loss": 0.5354,
+      "step": 8201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9706937675078717,
+      "learning_rate": 5.285307137237039e-06,
+      "loss": 0.5392,
+      "step": 8202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.888202310462275,
+      "learning_rate": 5.282985769264342e-06,
+      "loss": 0.4573,
+      "step": 8203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9661150146788169,
+      "learning_rate": 5.280664728176542e-06,
+      "loss": 0.5915,
+      "step": 8204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8687295099552259,
+      "learning_rate": 5.278344014134491e-06,
+      "loss": 0.4291,
+      "step": 8205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9305137050651606,
+      "learning_rate": 5.276023627299011e-06,
+      "loss": 0.545,
+      "step": 8206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9249269297925938,
+      "learning_rate": 5.273703567830908e-06,
+      "loss": 0.5775,
+      "step": 8207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8719197817568556,
+      "learning_rate": 5.271383835890947e-06,
+      "loss": 0.5072,
+      "step": 8208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.035278658434689,
+      "learning_rate": 5.269064431639901e-06,
+      "loss": 0.5533,
+      "step": 8209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8978940252138314,
+      "learning_rate": 5.266745355238489e-06,
+      "loss": 0.5206,
+      "step": 8210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9912583159987712,
+      "learning_rate": 5.264426606847426e-06,
+      "loss": 0.5126,
+      "step": 8211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9517846552457596,
+      "learning_rate": 5.262108186627397e-06,
+      "loss": 0.5587,
+      "step": 8212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8812396536972811,
+      "learning_rate": 5.259790094739073e-06,
+      "loss": 0.4878,
+      "step": 8213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9195417195160543,
+      "learning_rate": 5.257472331343083e-06,
+      "loss": 0.5044,
+      "step": 8214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8872421603933094,
+      "learning_rate": 5.2551548966000635e-06,
+      "loss": 0.5249,
+      "step": 8215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9310784935184504,
+      "learning_rate": 5.252837790670595e-06,
+      "loss": 0.5033,
+      "step": 8216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.997219302064515,
+      "learning_rate": 5.250521013715257e-06,
+      "loss": 0.5841,
+      "step": 8217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9227493754424385,
+      "learning_rate": 5.2482045658946e-06,
+      "loss": 0.5427,
+      "step": 8218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8252507468536766,
+      "learning_rate": 5.245888447369157e-06,
+      "loss": 0.4472,
+      "step": 8219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8750538145933262,
+      "learning_rate": 5.243572658299418e-06,
+      "loss": 0.5236,
+      "step": 8220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8899187586334513,
+      "learning_rate": 5.241257198845884e-06,
+      "loss": 0.497,
+      "step": 8221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9948214678414993,
+      "learning_rate": 5.238942069169e-06,
+      "loss": 0.5945,
+      "step": 8222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9692887420393211,
+      "learning_rate": 5.236627269429208e-06,
+      "loss": 0.535,
+      "step": 8223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.990515988397188,
+      "learning_rate": 5.234312799786921e-06,
+      "loss": 0.4657,
+      "step": 8224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9534115892704664,
+      "learning_rate": 5.231998660402535e-06,
+      "loss": 0.5732,
+      "step": 8225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0066116250736143,
+      "learning_rate": 5.229684851436403e-06,
+      "loss": 0.4618,
+      "step": 8226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0698888040079477,
+      "learning_rate": 5.2273713730488886e-06,
+      "loss": 0.4996,
+      "step": 8227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9414620310160213,
+      "learning_rate": 5.2250582254003016e-06,
+      "loss": 0.4974,
+      "step": 8228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9323123866798388,
+      "learning_rate": 5.222745408650942e-06,
+      "loss": 0.5358,
+      "step": 8229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9696890580161104,
+      "learning_rate": 5.220432922961089e-06,
+      "loss": 0.552,
+      "step": 8230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7697669490610657,
+      "learning_rate": 5.218120768491e-06,
+      "loss": 0.3773,
+      "step": 8231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.822697108368092,
+      "learning_rate": 5.215808945400891e-06,
+      "loss": 0.5099,
+      "step": 8232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9367819561071206,
+      "learning_rate": 5.213497453850986e-06,
+      "loss": 0.4753,
+      "step": 8233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9265513771122531,
+      "learning_rate": 5.21118629400146e-06,
+      "loss": 0.4978,
+      "step": 8234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8839751231544198,
+      "learning_rate": 5.208875466012475e-06,
+      "loss": 0.4512,
+      "step": 8235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.853117385808319,
+      "learning_rate": 5.20656497004417e-06,
+      "loss": 0.4687,
+      "step": 8236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8691450740963526,
+      "learning_rate": 5.2042548062566654e-06,
+      "loss": 0.4995,
+      "step": 8237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9239210980343396,
+      "learning_rate": 5.2019449748100405e-06,
+      "loss": 0.4635,
+      "step": 8238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8869835721887451,
+      "learning_rate": 5.19963547586438e-06,
+      "loss": 0.4768,
+      "step": 8239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8739268679710483,
+      "learning_rate": 5.197326309579721e-06,
+      "loss": 0.5038,
+      "step": 8240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7983910758111474,
+      "learning_rate": 5.195017476116089e-06,
+      "loss": 0.4672,
+      "step": 8241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9395625394613684,
+      "learning_rate": 5.192708975633483e-06,
+      "loss": 0.5245,
+      "step": 8242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8647830622368983,
+      "learning_rate": 5.190400808291884e-06,
+      "loss": 0.4765,
+      "step": 8243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8477915105418694,
+      "learning_rate": 5.1880929742512355e-06,
+      "loss": 0.4986,
+      "step": 8244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9637514843118539,
+      "learning_rate": 5.185785473671484e-06,
+      "loss": 0.5349,
+      "step": 8245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8844263857942123,
+      "learning_rate": 5.183478306712525e-06,
+      "loss": 0.4652,
+      "step": 8246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9204972924084194,
+      "learning_rate": 5.181171473534248e-06,
+      "loss": 0.5243,
+      "step": 8247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.987097280515057,
+      "learning_rate": 5.178864974296511e-06,
+      "loss": 0.5505,
+      "step": 8248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8715329192132406,
+      "learning_rate": 5.176558809159161e-06,
+      "loss": 0.473,
+      "step": 8249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7745699602973506,
+      "learning_rate": 5.174252978281999e-06,
+      "loss": 0.449,
+      "step": 8250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9377200872321961,
+      "learning_rate": 5.171947481824832e-06,
+      "loss": 0.4785,
+      "step": 8251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0250797101577,
+      "learning_rate": 5.16964231994742e-06,
+      "loss": 0.527,
+      "step": 8252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9173853321475266,
+      "learning_rate": 5.16733749280951e-06,
+      "loss": 0.5303,
+      "step": 8253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8971832660446327,
+      "learning_rate": 5.165033000570825e-06,
+      "loss": 0.5138,
+      "step": 8254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9360969938703607,
+      "learning_rate": 5.162728843391067e-06,
+      "loss": 0.532,
+      "step": 8255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8989978474119811,
+      "learning_rate": 5.160425021429904e-06,
+      "loss": 0.4812,
+      "step": 8256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0203611815836584,
+      "learning_rate": 5.158121534847e-06,
+      "loss": 0.4886,
+      "step": 8257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9387756661800578,
+      "learning_rate": 5.155818383801976e-06,
+      "loss": 0.4807,
+      "step": 8258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9140901137656047,
+      "learning_rate": 5.153515568454441e-06,
+      "loss": 0.4608,
+      "step": 8259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8470184521435462,
+      "learning_rate": 5.1512130889639785e-06,
+      "loss": 0.4343,
+      "step": 8260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9058316582478934,
+      "learning_rate": 5.148910945490152e-06,
+      "loss": 0.5013,
+      "step": 8261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9659521040360318,
+      "learning_rate": 5.1466091381924864e-06,
+      "loss": 0.5635,
+      "step": 8262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8385312611403691,
+      "learning_rate": 5.144307667230511e-06,
+      "loss": 0.4924,
+      "step": 8263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8757450526978969,
+      "learning_rate": 5.142006532763698e-06,
+      "loss": 0.4793,
+      "step": 8264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9538110402970061,
+      "learning_rate": 5.139705734951532e-06,
+      "loss": 0.5155,
+      "step": 8265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9180442618190727,
+      "learning_rate": 5.137405273953443e-06,
+      "loss": 0.4495,
+      "step": 8266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9031867790761441,
+      "learning_rate": 5.1351051499288565e-06,
+      "loss": 0.462,
+      "step": 8267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8721581631393269,
+      "learning_rate": 5.1328053630371656e-06,
+      "loss": 0.4448,
+      "step": 8268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9379064954460689,
+      "learning_rate": 5.130505913437751e-06,
+      "loss": 0.4893,
+      "step": 8269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.914567607388558,
+      "learning_rate": 5.12820680128995e-06,
+      "loss": 0.4544,
+      "step": 8270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9786738451139384,
+      "learning_rate": 5.125908026753105e-06,
+      "loss": 0.5393,
+      "step": 8271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9384087179971159,
+      "learning_rate": 5.123609589986505e-06,
+      "loss": 0.5539,
+      "step": 8272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.874501938753754,
+      "learning_rate": 5.121311491149437e-06,
+      "loss": 0.4775,
+      "step": 8273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9130947014954597,
+      "learning_rate": 5.119013730401152e-06,
+      "loss": 0.4939,
+      "step": 8274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8959183379839931,
+      "learning_rate": 5.116716307900893e-06,
+      "loss": 0.4896,
+      "step": 8275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8765212972837662,
+      "learning_rate": 5.114419223807854e-06,
+      "loss": 0.4999,
+      "step": 8276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9158320029828727,
+      "learning_rate": 5.112122478281236e-06,
+      "loss": 0.4995,
+      "step": 8277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9464418970680727,
+      "learning_rate": 5.109826071480191e-06,
+      "loss": 0.4647,
+      "step": 8278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9579586965573048,
+      "learning_rate": 5.107530003563862e-06,
+      "loss": 0.4866,
+      "step": 8279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8705674475250073,
+      "learning_rate": 5.105234274691364e-06,
+      "loss": 0.4887,
+      "step": 8280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8972048362169255,
+      "learning_rate": 5.1029388850217935e-06,
+      "loss": 0.4719,
+      "step": 8281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9114081984949848,
+      "learning_rate": 5.100643834714206e-06,
+      "loss": 0.461,
+      "step": 8282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0612571737985634,
+      "learning_rate": 5.098349123927664e-06,
+      "loss": 0.5402,
+      "step": 8283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9419682913709696,
+      "learning_rate": 5.096054752821174e-06,
+      "loss": 0.4755,
+      "step": 8284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9728192014206865,
+      "learning_rate": 5.093760721553742e-06,
+      "loss": 0.4933,
+      "step": 8285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9092408942177812,
+      "learning_rate": 5.091467030284339e-06,
+      "loss": 0.5053,
+      "step": 8286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9806559724074531,
+      "learning_rate": 5.089173679171922e-06,
+      "loss": 0.5024,
+      "step": 8287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8201955970089637,
+      "learning_rate": 5.086880668375404e-06,
+      "loss": 0.44,
+      "step": 8288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8804686334163445,
+      "learning_rate": 5.084587998053706e-06,
+      "loss": 0.4512,
+      "step": 8289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9055907430616575,
+      "learning_rate": 5.082295668365695e-06,
+      "loss": 0.5823,
+      "step": 8290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9471393593487665,
+      "learning_rate": 5.080003679470234e-06,
+      "loss": 0.5007,
+      "step": 8291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9592584808386907,
+      "learning_rate": 5.077712031526153e-06,
+      "loss": 0.4993,
+      "step": 8292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.969919058632812,
+      "learning_rate": 5.075420724692266e-06,
+      "loss": 0.5404,
+      "step": 8293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9528343829077746,
+      "learning_rate": 5.073129759127346e-06,
+      "loss": 0.4887,
+      "step": 8294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9123961547566919,
+      "learning_rate": 5.070839134990173e-06,
+      "loss": 0.5215,
+      "step": 8295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9159772154630276,
+      "learning_rate": 5.0685488524394725e-06,
+      "loss": 0.5017,
+      "step": 8296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0016846413150464,
+      "learning_rate": 5.06625891163396e-06,
+      "loss": 0.5566,
+      "step": 8297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8644745935064114,
+      "learning_rate": 5.063969312732331e-06,
+      "loss": 0.484,
+      "step": 8298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9849624018056626,
+      "learning_rate": 5.0616800558932525e-06,
+      "loss": 0.5766,
+      "step": 8299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.740789537902068,
+      "learning_rate": 5.059391141275358e-06,
+      "loss": 0.4458,
+      "step": 8300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9474555248146016,
+      "learning_rate": 5.057102569037284e-06,
+      "loss": 0.4613,
+      "step": 8301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9662314646834963,
+      "learning_rate": 5.054814339337613e-06,
+      "loss": 0.5123,
+      "step": 8302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8568849751085164,
+      "learning_rate": 5.052526452334923e-06,
+      "loss": 0.4924,
+      "step": 8303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8870863599394765,
+      "learning_rate": 5.050238908187759e-06,
+      "loss": 0.5371,
+      "step": 8304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9210411495904444,
+      "learning_rate": 5.047951707054655e-06,
+      "loss": 0.5217,
+      "step": 8305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8181478158838038,
+      "learning_rate": 5.0456648490940966e-06,
+      "loss": 0.4833,
+      "step": 8306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0159089504437258,
+      "learning_rate": 5.043378334464576e-06,
+      "loss": 0.5152,
+      "step": 8307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9093263342954216,
+      "learning_rate": 5.041092163324537e-06,
+      "loss": 0.4997,
+      "step": 8308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7751798119901546,
+      "learning_rate": 5.038806335832414e-06,
+      "loss": 0.4602,
+      "step": 8309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.94501594818902,
+      "learning_rate": 5.036520852146609e-06,
+      "loss": 0.481,
+      "step": 8310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8825734086585589,
+      "learning_rate": 5.034235712425508e-06,
+      "loss": 0.4882,
+      "step": 8311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9411659546887424,
+      "learning_rate": 5.031950916827467e-06,
+      "loss": 0.5628,
+      "step": 8312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7861532836500162,
+      "learning_rate": 5.029666465510825e-06,
+      "loss": 0.4508,
+      "step": 8313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8417521207534142,
+      "learning_rate": 5.027382358633884e-06,
+      "loss": 0.4638,
+      "step": 8314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9312435206636756,
+      "learning_rate": 5.0250985963549356e-06,
+      "loss": 0.5199,
+      "step": 8315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8892045252922159,
+      "learning_rate": 5.0228151788322414e-06,
+      "loss": 0.454,
+      "step": 8316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8297226180785294,
+      "learning_rate": 5.020532106224041e-06,
+      "loss": 0.4403,
+      "step": 8317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8955990912624906,
+      "learning_rate": 5.018249378688547e-06,
+      "loss": 0.4802,
+      "step": 8318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8838311608237159,
+      "learning_rate": 5.0159669963839575e-06,
+      "loss": 0.4684,
+      "step": 8319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9126526422811626,
+      "learning_rate": 5.01368495946843e-06,
+      "loss": 0.4477,
+      "step": 8320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9647439971830661,
+      "learning_rate": 5.011403268100112e-06,
+      "loss": 0.5284,
+      "step": 8321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9468468051520715,
+      "learning_rate": 5.009121922437124e-06,
+      "loss": 0.492,
+      "step": 8322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9635052588870198,
+      "learning_rate": 5.006840922637559e-06,
+      "loss": 0.5155,
+      "step": 8323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.858159908275672,
+      "learning_rate": 5.004560268859488e-06,
+      "loss": 0.4753,
+      "step": 8324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8783926198443746,
+      "learning_rate": 5.002279961260965e-06,
+      "loss": 0.463,
+      "step": 8325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9444190542938617,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.5454,
+      "step": 8326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8049607873816147,
+      "learning_rate": 4.9977203852346054e-06,
+      "loss": 0.4467,
+      "step": 8327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8703985735991734,
+      "learning_rate": 4.995441117122749e-06,
+      "loss": 0.4628,
+      "step": 8328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8775384023172903,
+      "learning_rate": 4.993162195822383e-06,
+      "loss": 0.5367,
+      "step": 8329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9282200217981799,
+      "learning_rate": 4.990883621491437e-06,
+      "loss": 0.5325,
+      "step": 8330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9705331260749077,
+      "learning_rate": 4.988605394287813e-06,
+      "loss": 0.5236,
+      "step": 8331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8734560259788998,
+      "learning_rate": 4.986327514369393e-06,
+      "loss": 0.5211,
+      "step": 8332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8882083363563135,
+      "learning_rate": 4.9840499818940255e-06,
+      "loss": 0.478,
+      "step": 8333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9391766553099113,
+      "learning_rate": 4.981772797019546e-06,
+      "loss": 0.4778,
+      "step": 8334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8743280481822453,
+      "learning_rate": 4.979495959903759e-06,
+      "loss": 0.5005,
+      "step": 8335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0280600118997267,
+      "learning_rate": 4.977219470704451e-06,
+      "loss": 0.5358,
+      "step": 8336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8990860486061055,
+      "learning_rate": 4.974943329579377e-06,
+      "loss": 0.4788,
+      "step": 8337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8991675776632321,
+      "learning_rate": 4.972667536686276e-06,
+      "loss": 0.4737,
+      "step": 8338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0160340949846591,
+      "learning_rate": 4.970392092182853e-06,
+      "loss": 0.5754,
+      "step": 8339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8822072783767629,
+      "learning_rate": 4.9681169962267975e-06,
+      "loss": 0.4731,
+      "step": 8340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8928340994522213,
+      "learning_rate": 4.96584224897577e-06,
+      "loss": 0.5183,
+      "step": 8341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8628952774436416,
+      "learning_rate": 4.963567850587408e-06,
+      "loss": 0.483,
+      "step": 8342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.894242138765183,
+      "learning_rate": 4.961293801219328e-06,
+      "loss": 0.5149,
+      "step": 8343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9176497901536859,
+      "learning_rate": 4.959020101029122e-06,
+      "loss": 0.5161,
+      "step": 8344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9679310502922771,
+      "learning_rate": 4.956746750174344e-06,
+      "loss": 0.5017,
+      "step": 8345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9750828264808324,
+      "learning_rate": 4.954473748812551e-06,
+      "loss": 0.4805,
+      "step": 8346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8381740441734827,
+      "learning_rate": 4.9522010971012465e-06,
+      "loss": 0.4963,
+      "step": 8347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8706129568474391,
+      "learning_rate": 4.949928795197931e-06,
+      "loss": 0.5171,
+      "step": 8348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8983833831441445,
+      "learning_rate": 4.947656843260068e-06,
+      "loss": 0.5004,
+      "step": 8349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.893544769279935,
+      "learning_rate": 4.9453852414451085e-06,
+      "loss": 0.454,
+      "step": 8350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9605367288494783,
+      "learning_rate": 4.943113989910462e-06,
+      "loss": 0.5191,
+      "step": 8351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8768862838630231,
+      "learning_rate": 4.940843088813537e-06,
+      "loss": 0.4355,
+      "step": 8352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9872960457295382,
+      "learning_rate": 4.938572538311696e-06,
+      "loss": 0.4885,
+      "step": 8353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9488391552128462,
+      "learning_rate": 4.936302338562288e-06,
+      "loss": 0.5248,
+      "step": 8354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8430325210263615,
+      "learning_rate": 4.934032489722637e-06,
+      "loss": 0.4987,
+      "step": 8355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9887525415033993,
+      "learning_rate": 4.931762991950043e-06,
+      "loss": 0.5418,
+      "step": 8356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8868405672461215,
+      "learning_rate": 4.929493845401772e-06,
+      "loss": 0.512,
+      "step": 8357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9305561141732096,
+      "learning_rate": 4.927225050235087e-06,
+      "loss": 0.4856,
+      "step": 8358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9093606578240055,
+      "learning_rate": 4.924956606607203e-06,
+      "loss": 0.4728,
+      "step": 8359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9255114452581292,
+      "learning_rate": 4.922688514675325e-06,
+      "loss": 0.5543,
+      "step": 8360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9108840803155676,
+      "learning_rate": 4.9204207745966285e-06,
+      "loss": 0.4961,
+      "step": 8361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9470093265948771,
+      "learning_rate": 4.918153386528271e-06,
+      "loss": 0.5079,
+      "step": 8362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9337221774130737,
+      "learning_rate": 4.915886350627368e-06,
+      "loss": 0.5819,
+      "step": 8363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8493640279177445,
+      "learning_rate": 4.91361966705104e-06,
+      "loss": 0.4399,
+      "step": 8364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.896803616770482,
+      "learning_rate": 4.911353335956353e-06,
+      "loss": 0.4911,
+      "step": 8365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9681357654854811,
+      "learning_rate": 4.909087357500366e-06,
+      "loss": 0.5298,
+      "step": 8366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9894347696243155,
+      "learning_rate": 4.906821731840109e-06,
+      "loss": 0.5156,
+      "step": 8367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0480262385449974,
+      "learning_rate": 4.904556459132593e-06,
+      "loss": 0.5048,
+      "step": 8368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9904266056570253,
+      "learning_rate": 4.902291539534787e-06,
+      "loss": 0.5019,
+      "step": 8369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8826399168957701,
+      "learning_rate": 4.900026973203663e-06,
+      "loss": 0.4766,
+      "step": 8370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9321355444330982,
+      "learning_rate": 4.897762760296143e-06,
+      "loss": 0.4984,
+      "step": 8371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9711002397447057,
+      "learning_rate": 4.895498900969138e-06,
+      "loss": 0.5301,
+      "step": 8372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8176926757979365,
+      "learning_rate": 4.893235395379531e-06,
+      "loss": 0.4271,
+      "step": 8373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9430190210323978,
+      "learning_rate": 4.890972243684185e-06,
+      "loss": 0.4597,
+      "step": 8374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8833110729017636,
+      "learning_rate": 4.888709446039923e-06,
+      "loss": 0.4718,
+      "step": 8375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.93005358908339,
+      "learning_rate": 4.8864470026035715e-06,
+      "loss": 0.5218,
+      "step": 8376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8287871369359837,
+      "learning_rate": 4.8841849135319015e-06,
+      "loss": 0.4328,
+      "step": 8377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9779055683662576,
+      "learning_rate": 4.881923178981681e-06,
+      "loss": 0.4185,
+      "step": 8378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0164807398807587,
+      "learning_rate": 4.879661799109644e-06,
+      "loss": 0.4969,
+      "step": 8379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9724315338703285,
+      "learning_rate": 4.877400774072506e-06,
+      "loss": 0.5074,
+      "step": 8380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.039312544558339,
+      "learning_rate": 4.875140104026943e-06,
+      "loss": 0.5116,
+      "step": 8381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9431211589333425,
+      "learning_rate": 4.872879789129632e-06,
+      "loss": 0.5218,
+      "step": 8382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8790106929243705,
+      "learning_rate": 4.870619829537201e-06,
+      "loss": 0.4876,
+      "step": 8383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0114065824703244,
+      "learning_rate": 4.8683602254062665e-06,
+      "loss": 0.5155,
+      "step": 8384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8819030896745794,
+      "learning_rate": 4.866100976893416e-06,
+      "loss": 0.5177,
+      "step": 8385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9055194541725219,
+      "learning_rate": 4.863842084155217e-06,
+      "loss": 0.4772,
+      "step": 8386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8911414355791749,
+      "learning_rate": 4.8615835473482e-06,
+      "loss": 0.5097,
+      "step": 8387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8504547122828845,
+      "learning_rate": 4.859325366628892e-06,
+      "loss": 0.4843,
+      "step": 8388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8963887036025582,
+      "learning_rate": 4.8570675421537685e-06,
+      "loss": 0.5021,
+      "step": 8389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9273190455880552,
+      "learning_rate": 4.854810074079311e-06,
+      "loss": 0.5287,
+      "step": 8390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8755536442112973,
+      "learning_rate": 4.852552962561946e-06,
+      "loss": 0.5171,
+      "step": 8391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8631816572604099,
+      "learning_rate": 4.8502962077580965e-06,
+      "loss": 0.4588,
+      "step": 8392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7915483046368798,
+      "learning_rate": 4.848039809824151e-06,
+      "loss": 0.4609,
+      "step": 8393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8926376451506014,
+      "learning_rate": 4.845783768916482e-06,
+      "loss": 0.5605,
+      "step": 8394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8692178271144266,
+      "learning_rate": 4.843528085191418e-06,
+      "loss": 0.4615,
+      "step": 8395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8715520588557414,
+      "learning_rate": 4.841272758805291e-06,
+      "loss": 0.4289,
+      "step": 8396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8749040524844592,
+      "learning_rate": 4.839017789914382e-06,
+      "loss": 0.4879,
+      "step": 8397
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8892874514883294,
+      "learning_rate": 4.836763178674963e-06,
+      "loss": 0.5299,
+      "step": 8398
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9678418266925191,
+      "learning_rate": 4.8345089252432765e-06,
+      "loss": 0.4949,
+      "step": 8399
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0388938109902268,
+      "learning_rate": 4.832255029775542e-06,
+      "loss": 0.5974,
+      "step": 8400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.883887908245481,
+      "learning_rate": 4.830001492427943e-06,
+      "loss": 0.4758,
+      "step": 8401
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8568929953037968,
+      "learning_rate": 4.827748313356664e-06,
+      "loss": 0.4778,
+      "step": 8402
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.84966750040475,
+      "learning_rate": 4.825495492717833e-06,
+      "loss": 0.433,
+      "step": 8403
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8802543271044627,
+      "learning_rate": 4.823243030667576e-06,
+      "loss": 0.5092,
+      "step": 8404
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8394079105908062,
+      "learning_rate": 4.8209909273619845e-06,
+      "loss": 0.448,
+      "step": 8405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9688607104012111,
+      "learning_rate": 4.818739182957131e-06,
+      "loss": 0.5156,
+      "step": 8406
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9174760388846913,
+      "learning_rate": 4.816487797609051e-06,
+      "loss": 0.5143,
+      "step": 8407
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8252282392609436,
+      "learning_rate": 4.814236771473774e-06,
+      "loss": 0.4832,
+      "step": 8408
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9581235332847905,
+      "learning_rate": 4.811986104707288e-06,
+      "loss": 0.4909,
+      "step": 8409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9153656263449683,
+      "learning_rate": 4.8097357974655615e-06,
+      "loss": 0.5051,
+      "step": 8410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8799875134221885,
+      "learning_rate": 4.8074858499045405e-06,
+      "loss": 0.4633,
+      "step": 8411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9366573540901902,
+      "learning_rate": 4.8052362621801484e-06,
+      "loss": 0.5065,
+      "step": 8412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8184587541847153,
+      "learning_rate": 4.802987034448267e-06,
+      "loss": 0.4333,
+      "step": 8413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8854510262214775,
+      "learning_rate": 4.800738166864784e-06,
+      "loss": 0.5147,
+      "step": 8414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8934024712409966,
+      "learning_rate": 4.798489659585529e-06,
+      "loss": 0.4496,
+      "step": 8415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8190161995760137,
+      "learning_rate": 4.7962415127663265e-06,
+      "loss": 0.4901,
+      "step": 8416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9948716753673862,
+      "learning_rate": 4.7939937265629725e-06,
+      "loss": 0.5582,
+      "step": 8417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9359470085761421,
+      "learning_rate": 4.79174630113124e-06,
+      "loss": 0.5093,
+      "step": 8418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8856683904774265,
+      "learning_rate": 4.789499236626859e-06,
+      "loss": 0.4876,
+      "step": 8419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8698069429816057,
+      "learning_rate": 4.7872525332055685e-06,
+      "loss": 0.4606,
+      "step": 8420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.951276069893926,
+      "learning_rate": 4.7850061910230495e-06,
+      "loss": 0.5241,
+      "step": 8421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9954898094083914,
+      "learning_rate": 4.782760210234976e-06,
+      "loss": 0.5543,
+      "step": 8422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8824398322853885,
+      "learning_rate": 4.780514590996992e-06,
+      "loss": 0.4602,
+      "step": 8423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8618284654301703,
+      "learning_rate": 4.778269333464721e-06,
+      "loss": 0.4473,
+      "step": 8424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8492717540748159,
+      "learning_rate": 4.776024437793746e-06,
+      "loss": 0.4279,
+      "step": 8425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9246816255673895,
+      "learning_rate": 4.773779904139652e-06,
+      "loss": 0.4564,
+      "step": 8426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.919388251924666,
+      "learning_rate": 4.7715357326579705e-06,
+      "loss": 0.4705,
+      "step": 8427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8265756691259849,
+      "learning_rate": 4.769291923504226e-06,
+      "loss": 0.4794,
+      "step": 8428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8666817154876048,
+      "learning_rate": 4.76704847683391e-06,
+      "loss": 0.4728,
+      "step": 8429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9208247524740416,
+      "learning_rate": 4.764805392802497e-06,
+      "loss": 0.4782,
+      "step": 8430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8166297428420718,
+      "learning_rate": 4.7625626715654205e-06,
+      "loss": 0.5053,
+      "step": 8431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9197309838095528,
+      "learning_rate": 4.760320313278112e-06,
+      "loss": 0.4857,
+      "step": 8432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8730053148337936,
+      "learning_rate": 4.758078318095953e-06,
+      "loss": 0.5088,
+      "step": 8433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9183706154217698,
+      "learning_rate": 4.755836686174319e-06,
+      "loss": 0.4954,
+      "step": 8434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7651626575366501,
+      "learning_rate": 4.753595417668551e-06,
+      "loss": 0.4731,
+      "step": 8435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8658156966163371,
+      "learning_rate": 4.75135451273397e-06,
+      "loss": 0.4705,
+      "step": 8436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8977597260123354,
+      "learning_rate": 4.749113971525858e-06,
+      "loss": 0.515,
+      "step": 8437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0343449984371662,
+      "learning_rate": 4.746873794199498e-06,
+      "loss": 0.569,
+      "step": 8438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8206778279220518,
+      "learning_rate": 4.744633980910122e-06,
+      "loss": 0.4751,
+      "step": 8439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9921478178561506,
+      "learning_rate": 4.742394531812949e-06,
+      "loss": 0.5699,
+      "step": 8440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9236598458945292,
+      "learning_rate": 4.74015544706317e-06,
+      "loss": 0.5102,
+      "step": 8441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9029506730016744,
+      "learning_rate": 4.737916726815958e-06,
+      "loss": 0.4348,
+      "step": 8442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8466552613302887,
+      "learning_rate": 4.7356783712264405e-06,
+      "loss": 0.503,
+      "step": 8443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8989842391438065,
+      "learning_rate": 4.733440380449752e-06,
+      "loss": 0.4691,
+      "step": 8444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9437254308309423,
+      "learning_rate": 4.731202754640969e-06,
+      "loss": 0.4926,
+      "step": 8445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8748975753828981,
+      "learning_rate": 4.728965493955162e-06,
+      "loss": 0.4764,
+      "step": 8446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8641293354033068,
+      "learning_rate": 4.72672859854737e-06,
+      "loss": 0.4868,
+      "step": 8447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9008577718286481,
+      "learning_rate": 4.724492068572614e-06,
+      "loss": 0.5056,
+      "step": 8448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.947620089849962,
+      "learning_rate": 4.722255904185869e-06,
+      "loss": 0.5184,
+      "step": 8449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9941821471275278,
+      "learning_rate": 4.720020105542117e-06,
+      "loss": 0.5347,
+      "step": 8450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8432679680464656,
+      "learning_rate": 4.717784672796285e-06,
+      "loss": 0.5014,
+      "step": 8451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8928988386363494,
+      "learning_rate": 4.715549606103289e-06,
+      "loss": 0.4781,
+      "step": 8452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0339335317195717,
+      "learning_rate": 4.7133149056180185e-06,
+      "loss": 0.5745,
+      "step": 8453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8003162042838572,
+      "learning_rate": 4.7110805714953385e-06,
+      "loss": 0.4694,
+      "step": 8454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8461276414515637,
+      "learning_rate": 4.708846603890077e-06,
+      "loss": 0.4899,
+      "step": 8455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8958768222352202,
+      "learning_rate": 4.70661300295706e-06,
+      "loss": 0.4872,
+      "step": 8456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9505494510294067,
+      "learning_rate": 4.704379768851063e-06,
+      "loss": 0.5211,
+      "step": 8457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8700792444455523,
+      "learning_rate": 4.702146901726851e-06,
+      "loss": 0.4902,
+      "step": 8458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9200429978046982,
+      "learning_rate": 4.69991440173916e-06,
+      "loss": 0.5211,
+      "step": 8459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8938147931028365,
+      "learning_rate": 4.697682269042698e-06,
+      "loss": 0.5195,
+      "step": 8460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8550642884636459,
+      "learning_rate": 4.695450503792153e-06,
+      "loss": 0.4923,
+      "step": 8461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0662928591320688,
+      "learning_rate": 4.693219106142186e-06,
+      "loss": 0.577,
+      "step": 8462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8924981782282854,
+      "learning_rate": 4.690988076247425e-06,
+      "loss": 0.4816,
+      "step": 8463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.903820363407162,
+      "learning_rate": 4.68875741426248e-06,
+      "loss": 0.5177,
+      "step": 8464
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0387163445050478,
+      "learning_rate": 4.686527120341936e-06,
+      "loss": 0.5164,
+      "step": 8465
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9481820813313033,
+      "learning_rate": 4.68429719464035e-06,
+      "loss": 0.5185,
+      "step": 8466
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8594449777348103,
+      "learning_rate": 4.6820676373122535e-06,
+      "loss": 0.4656,
+      "step": 8467
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8502875164783491,
+      "learning_rate": 4.679838448512155e-06,
+      "loss": 0.4476,
+      "step": 8468
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9414262894799519,
+      "learning_rate": 4.677609628394529e-06,
+      "loss": 0.4923,
+      "step": 8469
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8068791737650673,
+      "learning_rate": 4.675381177113837e-06,
+      "loss": 0.4124,
+      "step": 8470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.909949833574802,
+      "learning_rate": 4.673153094824505e-06,
+      "loss": 0.4937,
+      "step": 8471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9170348465696564,
+      "learning_rate": 4.670925381680938e-06,
+      "loss": 0.5086,
+      "step": 8472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9315108992160059,
+      "learning_rate": 4.668698037837517e-06,
+      "loss": 0.5393,
+      "step": 8473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.873085812567832,
+      "learning_rate": 4.666471063448595e-06,
+      "loss": 0.5344,
+      "step": 8474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0079423729232555,
+      "learning_rate": 4.664244458668496e-06,
+      "loss": 0.5647,
+      "step": 8475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8207324962622179,
+      "learning_rate": 4.662018223651521e-06,
+      "loss": 0.4896,
+      "step": 8476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9446632078134728,
+      "learning_rate": 4.65979235855195e-06,
+      "loss": 0.5198,
+      "step": 8477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8965188809159295,
+      "learning_rate": 4.65756686352403e-06,
+      "loss": 0.467,
+      "step": 8478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8649659172553711,
+      "learning_rate": 4.655341738721989e-06,
+      "loss": 0.4583,
+      "step": 8479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9028168849235375,
+      "learning_rate": 4.653116984300024e-06,
+      "loss": 0.5207,
+      "step": 8480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8951444423118171,
+      "learning_rate": 4.6508926004123145e-06,
+      "loss": 0.4682,
+      "step": 8481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8852410562883132,
+      "learning_rate": 4.648668587212998e-06,
+      "loss": 0.4784,
+      "step": 8482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9167657976916679,
+      "learning_rate": 4.646444944856202e-06,
+      "loss": 0.5031,
+      "step": 8483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9260651434554181,
+      "learning_rate": 4.644221673496023e-06,
+      "loss": 0.5693,
+      "step": 8484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0223575682469483,
+      "learning_rate": 4.641998773286531e-06,
+      "loss": 0.4936,
+      "step": 8485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8672998551519863,
+      "learning_rate": 4.639776244381772e-06,
+      "loss": 0.491,
+      "step": 8486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8243916096199051,
+      "learning_rate": 4.637554086935768e-06,
+      "loss": 0.49,
+      "step": 8487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8811146797443113,
+      "learning_rate": 4.635332301102507e-06,
+      "loss": 0.4524,
+      "step": 8488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0189452727877433,
+      "learning_rate": 4.633110887035957e-06,
+      "loss": 0.5636,
+      "step": 8489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9832362112575355,
+      "learning_rate": 4.630889844890063e-06,
+      "loss": 0.4915,
+      "step": 8490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9609507067698836,
+      "learning_rate": 4.628669174818741e-06,
+      "loss": 0.5044,
+      "step": 8491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8886354571444008,
+      "learning_rate": 4.6264488769758795e-06,
+      "loss": 0.4868,
+      "step": 8492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9127612427732726,
+      "learning_rate": 4.6242289515153495e-06,
+      "loss": 0.5326,
+      "step": 8493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9293163321558532,
+      "learning_rate": 4.622009398590976e-06,
+      "loss": 0.4987,
+      "step": 8494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8595017051048842,
+      "learning_rate": 4.619790218356589e-06,
+      "loss": 0.4478,
+      "step": 8495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.932634641983612,
+      "learning_rate": 4.617571410965964e-06,
+      "loss": 0.5015,
+      "step": 8496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9429550807269602,
+      "learning_rate": 4.615352976572867e-06,
+      "loss": 0.4838,
+      "step": 8497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9593082113581994,
+      "learning_rate": 4.613134915331031e-06,
+      "loss": 0.4767,
+      "step": 8498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8782623572538939,
+      "learning_rate": 4.610917227394172e-06,
+      "loss": 0.4841,
+      "step": 8499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9823091850599716,
+      "learning_rate": 4.608699912915961e-06,
+      "loss": 0.5373,
+      "step": 8500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9240946447050321,
+      "learning_rate": 4.606482972050072e-06,
+      "loss": 0.493,
+      "step": 8501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9995884422788283,
+      "learning_rate": 4.604266404950124e-06,
+      "loss": 0.4942,
+      "step": 8502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9315818382316459,
+      "learning_rate": 4.60205021176973e-06,
+      "loss": 0.5339,
+      "step": 8503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9313904947633198,
+      "learning_rate": 4.599834392662467e-06,
+      "loss": 0.5309,
+      "step": 8504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9234364248702774,
+      "learning_rate": 4.5976189477818945e-06,
+      "loss": 0.5381,
+      "step": 8505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9642841600311677,
+      "learning_rate": 4.5954038772815305e-06,
+      "loss": 0.5149,
+      "step": 8506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9758857960131119,
+      "learning_rate": 4.5931891813148895e-06,
+      "loss": 0.5616,
+      "step": 8507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9939874774025128,
+      "learning_rate": 4.5909748600354395e-06,
+      "loss": 0.5817,
+      "step": 8508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.883662600989568,
+      "learning_rate": 4.588760913596635e-06,
+      "loss": 0.4887,
+      "step": 8509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0144586260482555,
+      "learning_rate": 4.586547342151898e-06,
+      "loss": 0.5689,
+      "step": 8510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9213786833800294,
+      "learning_rate": 4.584334145854633e-06,
+      "loss": 0.5044,
+      "step": 8511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8710235090027155,
+      "learning_rate": 4.582121324858201e-06,
+      "loss": 0.475,
+      "step": 8512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9984563099572079,
+      "learning_rate": 4.579908879315962e-06,
+      "loss": 0.5219,
+      "step": 8513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.907665700346533,
+      "learning_rate": 4.577696809381222e-06,
+      "loss": 0.5097,
+      "step": 8514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8598035995585397,
+      "learning_rate": 4.5754851152072935e-06,
+      "loss": 0.4853,
+      "step": 8515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9093082298469464,
+      "learning_rate": 4.573273796947429e-06,
+      "loss": 0.5327,
+      "step": 8516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8808884117504743,
+      "learning_rate": 4.571062854754878e-06,
+      "loss": 0.5258,
+      "step": 8517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8533527701048154,
+      "learning_rate": 4.568852288782854e-06,
+      "loss": 0.4881,
+      "step": 8518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9163197742017514,
+      "learning_rate": 4.566642099184555e-06,
+      "loss": 0.5018,
+      "step": 8519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8215761882741016,
+      "learning_rate": 4.564432286113131e-06,
+      "loss": 0.4834,
+      "step": 8520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7933050125474624,
+      "learning_rate": 4.562222849721735e-06,
+      "loss": 0.4094,
+      "step": 8521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8897154668309718,
+      "learning_rate": 4.560013790163469e-06,
+      "loss": 0.5105,
+      "step": 8522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.855198442809487,
+      "learning_rate": 4.557805107591421e-06,
+      "loss": 0.455,
+      "step": 8523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9524596667633722,
+      "learning_rate": 4.555596802158653e-06,
+      "loss": 0.5091,
+      "step": 8524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.950194938204538,
+      "learning_rate": 4.553388874018201e-06,
+      "loss": 0.5583,
+      "step": 8525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9400851744243489,
+      "learning_rate": 4.551181323323062e-06,
+      "loss": 0.4941,
+      "step": 8526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9146266111978486,
+      "learning_rate": 4.548974150226231e-06,
+      "loss": 0.4986,
+      "step": 8527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9660153299413448,
+      "learning_rate": 4.546767354880653e-06,
+      "loss": 0.5393,
+      "step": 8528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9297055371592443,
+      "learning_rate": 4.544560937439258e-06,
+      "loss": 0.4953,
+      "step": 8529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9696637231242728,
+      "learning_rate": 4.542354898054953e-06,
+      "loss": 0.5172,
+      "step": 8530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9218306554310379,
+      "learning_rate": 4.540149236880616e-06,
+      "loss": 0.4773,
+      "step": 8531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9486063599967517,
+      "learning_rate": 4.537943954069088e-06,
+      "loss": 0.546,
+      "step": 8532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9924791760385479,
+      "learning_rate": 4.535739049773206e-06,
+      "loss": 0.5536,
+      "step": 8533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8468685421929781,
+      "learning_rate": 4.533534524145756e-06,
+      "loss": 0.4782,
+      "step": 8534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9512118915905358,
+      "learning_rate": 4.531330377339516e-06,
+      "loss": 0.4956,
+      "step": 8535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9716515137110536,
+      "learning_rate": 4.529126609507229e-06,
+      "loss": 0.5597,
+      "step": 8536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8938265448169714,
+      "learning_rate": 4.52692322080162e-06,
+      "loss": 0.4546,
+      "step": 8537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8921292523190358,
+      "learning_rate": 4.52472021137537e-06,
+      "loss": 0.4938,
+      "step": 8538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.977254514957181,
+      "learning_rate": 4.5225175813811585e-06,
+      "loss": 0.4989,
+      "step": 8539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9901426180974761,
+      "learning_rate": 4.520315330971617e-06,
+      "loss": 0.523,
+      "step": 8540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9762476438745803,
+      "learning_rate": 4.518113460299364e-06,
+      "loss": 0.5033,
+      "step": 8541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9511493492501386,
+      "learning_rate": 4.515911969516985e-06,
+      "loss": 0.5201,
+      "step": 8542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9544821754504096,
+      "learning_rate": 4.513710858777045e-06,
+      "loss": 0.5167,
+      "step": 8543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0477136437004189,
+      "learning_rate": 4.5115101282320695e-06,
+      "loss": 0.5641,
+      "step": 8544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9212158833631946,
+      "learning_rate": 4.509309778034582e-06,
+      "loss": 0.4567,
+      "step": 8545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9859534975875366,
+      "learning_rate": 4.507109808337051e-06,
+      "loss": 0.5537,
+      "step": 8546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8521980359414838,
+      "learning_rate": 4.504910219291941e-06,
+      "loss": 0.4825,
+      "step": 8547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8726401564149688,
+      "learning_rate": 4.502711011051676e-06,
+      "loss": 0.4641,
+      "step": 8548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8850970445322484,
+      "learning_rate": 4.500512183768666e-06,
+      "loss": 0.4365,
+      "step": 8549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8528034962174915,
+      "learning_rate": 4.498313737595277e-06,
+      "loss": 0.4914,
+      "step": 8550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8943513341597491,
+      "learning_rate": 4.4961156726838725e-06,
+      "loss": 0.5079,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8238291990363437,
+      "learning_rate": 4.493917989186768e-06,
+      "loss": 0.5233,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.935896226406557,
+      "learning_rate": 4.491720687256261e-06,
+      "loss": 0.492,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8779507327285243,
+      "learning_rate": 4.489523767044625e-06,
+      "loss": 0.4805,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.998133203070941,
+      "learning_rate": 4.487327228704108e-06,
+      "loss": 0.5005,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9016653665148767,
+      "learning_rate": 4.4851310723869176e-06,
+      "loss": 0.4611,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9136562820143127,
+      "learning_rate": 4.482935298245259e-06,
+      "loss": 0.4689,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.990290997862627,
+      "learning_rate": 4.480739906431287e-06,
+      "loss": 0.5298,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9447926331667892,
+      "learning_rate": 4.478544897097144e-06,
+      "loss": 0.521,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8593961200741635,
+      "learning_rate": 4.476350270394942e-06,
+      "loss": 0.4391,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8770524066755371,
+      "learning_rate": 4.47415602647677e-06,
+      "loss": 0.5104,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9107096575839747,
+      "learning_rate": 4.471962165494678e-06,
+      "loss": 0.5393,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9230544531802135,
+      "learning_rate": 4.469768687600711e-06,
+      "loss": 0.4659,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9017244828244761,
+      "learning_rate": 4.467575592946865e-06,
+      "loss": 0.5099,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9072309195504163,
+      "learning_rate": 4.465382881685122e-06,
+      "loss": 0.4922,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9127834367641511,
+      "learning_rate": 4.463190553967438e-06,
+      "loss": 0.531,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9630119394437752,
+      "learning_rate": 4.46099860994574e-06,
+      "loss": 0.4919,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.402515032284407,
+      "learning_rate": 4.458807049771918e-06,
+      "loss": 0.5254,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.979726220736822,
+      "learning_rate": 4.4566158735978605e-06,
+      "loss": 0.5463,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9475301973882784,
+      "learning_rate": 4.454425081575402e-06,
+      "loss": 0.5353,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.939575507723578,
+      "learning_rate": 4.452234673856366e-06,
+      "loss": 0.5089,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9487876723711443,
+      "learning_rate": 4.450044650592547e-06,
+      "loss": 0.5164,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9522260285452261,
+      "learning_rate": 4.447855011935714e-06,
+      "loss": 0.5307,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8633787236123219,
+      "learning_rate": 4.4456657580375966e-06,
+      "loss": 0.4582,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9049310398365067,
+      "learning_rate": 4.443476889049923e-06,
+      "loss": 0.5123,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9042445194944134,
+      "learning_rate": 4.441288405124368e-06,
+      "loss": 0.4971,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9061722854430694,
+      "learning_rate": 4.439100306412596e-06,
+      "loss": 0.5561,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9681196865332937,
+      "learning_rate": 4.436912593066241e-06,
+      "loss": 0.5445,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9045333714159708,
+      "learning_rate": 4.434725265236912e-06,
+      "loss": 0.4744,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9791358579263623,
+      "learning_rate": 4.4325383230761785e-06,
+      "loss": 0.5647,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9190795854611128,
+      "learning_rate": 4.430351766735609e-06,
+      "loss": 0.4574,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9443589825222185,
+      "learning_rate": 4.428165596366717e-06,
+      "loss": 0.4894,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.010699088324618,
+      "learning_rate": 4.425979812121008e-06,
+      "loss": 0.5239,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0005524797056764,
+      "learning_rate": 4.423794414149953e-06,
+      "loss": 0.5694,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9028625873596521,
+      "learning_rate": 4.421609402605003e-06,
+      "loss": 0.5101,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9555413133270083,
+      "learning_rate": 4.419424777637565e-06,
+      "loss": 0.4983,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.018225022093693,
+      "learning_rate": 4.4172405393990495e-06,
+      "loss": 0.4761,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9019231468353708,
+      "learning_rate": 4.415056688040807e-06,
+      "loss": 0.509,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8944092264707802,
+      "learning_rate": 4.412873223714184e-06,
+      "loss": 0.4492,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0205877351895996,
+      "learning_rate": 4.41069014657049e-06,
+      "loss": 0.5679,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9175443585321001,
+      "learning_rate": 4.408507456761014e-06,
+      "loss": 0.4681,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.941166738333424,
+      "learning_rate": 4.4063251544370055e-06,
+      "loss": 0.5679,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9257517601389404,
+      "learning_rate": 4.404143239749709e-06,
+      "loss": 0.5314,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9041427478022955,
+      "learning_rate": 4.401961712850318e-06,
+      "loss": 0.4892,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8909288731891772,
+      "learning_rate": 4.399780573890016e-06,
+      "loss": 0.4575,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9788352851978374,
+      "learning_rate": 4.397599823019953e-06,
+      "loss": 0.4932,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9628780507902193,
+      "learning_rate": 4.395419460391256e-06,
+      "loss": 0.5098,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9626038196010682,
+      "learning_rate": 4.393239486155011e-06,
+      "loss": 0.5357,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9024459623483713,
+      "learning_rate": 4.391059900462305e-06,
+      "loss": 0.5328,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9620848702073997,
+      "learning_rate": 4.3888807034641686e-06,
+      "loss": 0.4771,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.973843994968412,
+      "learning_rate": 4.386701895311622e-06,
+      "loss": 0.4848,
+      "step": 8601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9457771770198199,
+      "learning_rate": 4.384523476155657e-06,
+      "loss": 0.5132,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.929037612513199,
+      "learning_rate": 4.382345446147236e-06,
+      "loss": 0.5633,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8646828981519963,
+      "learning_rate": 4.380167805437285e-06,
+      "loss": 0.4819,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9185236496323183,
+      "learning_rate": 4.377990554176729e-06,
+      "loss": 0.499,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9021456393842846,
+      "learning_rate": 4.375813692516437e-06,
+      "loss": 0.493,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9828541683033288,
+      "learning_rate": 4.3736372206072666e-06,
+      "loss": 0.5731,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9002309994632519,
+      "learning_rate": 4.371461138600047e-06,
+      "loss": 0.554,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.960763045179828,
+      "learning_rate": 4.369285446645578e-06,
+      "loss": 0.5608,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9058200343498363,
+      "learning_rate": 4.367110144894633e-06,
+      "loss": 0.5084,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9395399780793992,
+      "learning_rate": 4.364935233497962e-06,
+      "loss": 0.5184,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9215093750756229,
+      "learning_rate": 4.362760712606278e-06,
+      "loss": 0.4846,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8610037956075174,
+      "learning_rate": 4.360586582370275e-06,
+      "loss": 0.4386,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.882746044843972,
+      "learning_rate": 4.35841284294062e-06,
+      "loss": 0.4848,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8953735482621343,
+      "learning_rate": 4.356239494467952e-06,
+      "loss": 0.4674,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.943203974943108,
+      "learning_rate": 4.35406653710288e-06,
+      "loss": 0.492,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8983064274161381,
+      "learning_rate": 4.351893970995994e-06,
+      "loss": 0.4969,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9973530547501277,
+      "learning_rate": 4.349721796297841e-06,
+      "loss": 0.5303,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9117438767939997,
+      "learning_rate": 4.347550013158956e-06,
+      "loss": 0.4307,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0095913131582344,
+      "learning_rate": 4.345378621729842e-06,
+      "loss": 0.5226,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8285377532732942,
+      "learning_rate": 4.343207622160973e-06,
+      "loss": 0.4997,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9936775538811486,
+      "learning_rate": 4.341037014602799e-06,
+      "loss": 0.5532,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8750442052146093,
+      "learning_rate": 4.338866799205744e-06,
+      "loss": 0.4762,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.012281747560933,
+      "learning_rate": 4.3366969761201935e-06,
+      "loss": 0.5108,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9792624934234849,
+      "learning_rate": 4.334527545496521e-06,
+      "loss": 0.5235,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9781064179748855,
+      "learning_rate": 4.332358507485064e-06,
+      "loss": 0.5202,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9132213170356942,
+      "learning_rate": 4.330189862236134e-06,
+      "loss": 0.5103,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8954379103279083,
+      "learning_rate": 4.328021609900018e-06,
+      "loss": 0.5055,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9164004427300606,
+      "learning_rate": 4.3258537506269735e-06,
+      "loss": 0.4858,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.951325657558114,
+      "learning_rate": 4.3236862845672355e-06,
+      "loss": 0.5028,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8636461833908563,
+      "learning_rate": 4.3215192118709984e-06,
+      "loss": 0.5028,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1844359900170125,
+      "learning_rate": 4.319352532688444e-06,
+      "loss": 0.4569,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9495900125986644,
+      "learning_rate": 4.317186247169719e-06,
+      "loss": 0.5037,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9685745342821419,
+      "learning_rate": 4.315020355464947e-06,
+      "loss": 0.5319,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8690042344461989,
+      "learning_rate": 4.312854857724222e-06,
+      "loss": 0.4636,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9177292653684901,
+      "learning_rate": 4.3106897540976154e-06,
+      "loss": 0.5345,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9735025676432103,
+      "learning_rate": 4.308525044735158e-06,
+      "loss": 0.5682,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9660018093666137,
+      "learning_rate": 4.306360729786867e-06,
+      "loss": 0.5515,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9595466170331142,
+      "learning_rate": 4.304196809402726e-06,
+      "loss": 0.4858,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9330526134964297,
+      "learning_rate": 4.302033283732695e-06,
+      "loss": 0.4753,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9228600220437362,
+      "learning_rate": 4.299870152926703e-06,
+      "loss": 0.4527,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.946059941678235,
+      "learning_rate": 4.297707417134653e-06,
+      "loss": 0.5222,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9401672436336672,
+      "learning_rate": 4.295545076506422e-06,
+      "loss": 0.5587,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9230695826194234,
+      "learning_rate": 4.293383131191861e-06,
+      "loss": 0.5095,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9264723424024507,
+      "learning_rate": 4.291221581340783e-06,
+      "loss": 0.5024,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7676869641906288,
+      "learning_rate": 4.2890604271029855e-06,
+      "loss": 0.444,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9593290265798053,
+      "learning_rate": 4.286899668628235e-06,
+      "loss": 0.5213,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9010995858897504,
+      "learning_rate": 4.28473930606627e-06,
+      "loss": 0.5134,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9418499445234443,
+      "learning_rate": 4.282579339566802e-06,
+      "loss": 0.509,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9286350494558199,
+      "learning_rate": 4.280419769279518e-06,
+      "loss": 0.5012,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8983920940568924,
+      "learning_rate": 4.278260595354067e-06,
+      "loss": 0.5336,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8294771280671885,
+      "learning_rate": 4.276101817940082e-06,
+      "loss": 0.4598,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8855833075909199,
+      "learning_rate": 4.273943437187163e-06,
+      "loss": 0.4714,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9034299559561075,
+      "learning_rate": 4.271785453244886e-06,
+      "loss": 0.4397,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9608987870079996,
+      "learning_rate": 4.269627866262794e-06,
+      "loss": 0.5578,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8949249243243168,
+      "learning_rate": 4.267470676390414e-06,
+      "loss": 0.4699,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8901792843107094,
+      "learning_rate": 4.2653138837772265e-06,
+      "loss": 0.5048,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9389581626819746,
+      "learning_rate": 4.2631574885727e-06,
+      "loss": 0.5082,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9186421948955802,
+      "learning_rate": 4.261001490926272e-06,
+      "loss": 0.52,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9401889077750639,
+      "learning_rate": 4.25884589098735e-06,
+      "loss": 0.4897,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9039686624889018,
+      "learning_rate": 4.256690688905315e-06,
+      "loss": 0.5232,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9773294716052752,
+      "learning_rate": 4.254535884829524e-06,
+      "loss": 0.5253,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7950138750683371,
+      "learning_rate": 4.252381478909293e-06,
+      "loss": 0.4714,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0925343616353147,
+      "learning_rate": 4.2502274712939355e-06,
+      "loss": 0.5992,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8961489935419249,
+      "learning_rate": 4.24807386213271e-06,
+      "loss": 0.4854,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8461885044508616,
+      "learning_rate": 4.245920651574864e-06,
+      "loss": 0.4316,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8998814032194048,
+      "learning_rate": 4.243767839769612e-06,
+      "loss": 0.4834,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8852003798729836,
+      "learning_rate": 4.241615426866148e-06,
+      "loss": 0.4867,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9462793357999046,
+      "learning_rate": 4.239463413013619e-06,
+      "loss": 0.5006,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8870614184576234,
+      "learning_rate": 4.237311798361175e-06,
+      "loss": 0.5095,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8429600883646559,
+      "learning_rate": 4.235160583057905e-06,
+      "loss": 0.5015,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.948909062706813,
+      "learning_rate": 4.233009767252896e-06,
+      "loss": 0.5001,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9127038362884151,
+      "learning_rate": 4.230859351095193e-06,
+      "loss": 0.5077,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9305603826139454,
+      "learning_rate": 4.2287093347338245e-06,
+      "loss": 0.5122,
+      "step": 8674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8877865608484397,
+      "learning_rate": 4.226559718317773e-06,
+      "loss": 0.4166,
+      "step": 8675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9078446718113264,
+      "learning_rate": 4.224410501996018e-06,
+      "loss": 0.5177,
+      "step": 8676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8757235975774688,
+      "learning_rate": 4.222261685917489e-06,
+      "loss": 0.4597,
+      "step": 8677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9691724908290217,
+      "learning_rate": 4.220113270231101e-06,
+      "loss": 0.5702,
+      "step": 8678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9633912836541412,
+      "learning_rate": 4.217965255085737e-06,
+      "loss": 0.5597,
+      "step": 8679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8084054151127114,
+      "learning_rate": 4.215817640630254e-06,
+      "loss": 0.4092,
+      "step": 8680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9928820511774611,
+      "learning_rate": 4.2136704270134725e-06,
+      "loss": 0.5033,
+      "step": 8681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9771886848828312,
+      "learning_rate": 4.2115236143842046e-06,
+      "loss": 0.522,
+      "step": 8682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9075441822519663,
+      "learning_rate": 4.209377202891212e-06,
+      "loss": 0.4654,
+      "step": 8683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9183635386993099,
+      "learning_rate": 4.207231192683243e-06,
+      "loss": 0.5257,
+      "step": 8684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8588744030939618,
+      "learning_rate": 4.205085583909014e-06,
+      "loss": 0.4934,
+      "step": 8685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9081390911988166,
+      "learning_rate": 4.2029403767172175e-06,
+      "loss": 0.4586,
+      "step": 8686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8660306671818201,
+      "learning_rate": 4.200795571256504e-06,
+      "loss": 0.4527,
+      "step": 8687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9407829774136649,
+      "learning_rate": 4.19865116767552e-06,
+      "loss": 0.528,
+      "step": 8688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8590314558726129,
+      "learning_rate": 4.196507166122862e-06,
+      "loss": 0.4732,
+      "step": 8689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8731436134748567,
+      "learning_rate": 4.1943635667471095e-06,
+      "loss": 0.4915,
+      "step": 8690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8511358601352428,
+      "learning_rate": 4.192220369696811e-06,
+      "loss": 0.4785,
+      "step": 8691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.068335180497299,
+      "learning_rate": 4.190077575120493e-06,
+      "loss": 0.5777,
+      "step": 8692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0273366727449977,
+      "learning_rate": 4.187935183166641e-06,
+      "loss": 0.5265,
+      "step": 8693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9019956570633716,
+      "learning_rate": 4.1857931939837305e-06,
+      "loss": 0.4879,
+      "step": 8694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8957925039835313,
+      "learning_rate": 4.18365160772019e-06,
+      "loss": 0.4607,
+      "step": 8695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8908531131530597,
+      "learning_rate": 4.1815104245244364e-06,
+      "loss": 0.5171,
+      "step": 8696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9426170847860915,
+      "learning_rate": 4.179369644544849e-06,
+      "loss": 0.4745,
+      "step": 8697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9223070695579061,
+      "learning_rate": 4.177229267929785e-06,
+      "loss": 0.5021,
+      "step": 8698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0669590657342318,
+      "learning_rate": 4.17508929482756e-06,
+      "loss": 0.5615,
+      "step": 8699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9170041445449919,
+      "learning_rate": 4.172949725386488e-06,
+      "loss": 0.4175,
+      "step": 8700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8346287732255635,
+      "learning_rate": 4.170810559754829e-06,
+      "loss": 0.4755,
+      "step": 8701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.825095119982811,
+      "learning_rate": 4.168671798080826e-06,
+      "loss": 0.4639,
+      "step": 8702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9463481324513929,
+      "learning_rate": 4.166533440512696e-06,
+      "loss": 0.4778,
+      "step": 8703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1810244437305535,
+      "learning_rate": 4.164395487198628e-06,
+      "loss": 0.5336,
+      "step": 8704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0187738412895564,
+      "learning_rate": 4.1622579382867686e-06,
+      "loss": 0.5458,
+      "step": 8705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9659152585832806,
+      "learning_rate": 4.160120793925264e-06,
+      "loss": 0.5135,
+      "step": 8706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9033350716520201,
+      "learning_rate": 4.157984054262205e-06,
+      "loss": 0.5468,
+      "step": 8707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.01327241202795,
+      "learning_rate": 4.155847719445669e-06,
+      "loss": 0.5685,
+      "step": 8708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8969467151084107,
+      "learning_rate": 4.1537117896237026e-06,
+      "loss": 0.4953,
+      "step": 8709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8190720523141264,
+      "learning_rate": 4.151576264944326e-06,
+      "loss": 0.5048,
+      "step": 8710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8853197781726923,
+      "learning_rate": 4.14944114555552e-06,
+      "loss": 0.511,
+      "step": 8711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9285788886413802,
+      "learning_rate": 4.14730643160526e-06,
+      "loss": 0.5235,
+      "step": 8712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9257368811326977,
+      "learning_rate": 4.14517212324147e-06,
+      "loss": 0.4874,
+      "step": 8713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.881845676784708,
+      "learning_rate": 4.143038220612058e-06,
+      "loss": 0.4813,
+      "step": 8714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9902780238114853,
+      "learning_rate": 4.140904723864903e-06,
+      "loss": 0.5445,
+      "step": 8715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.04868441240058,
+      "learning_rate": 4.138771633147856e-06,
+      "loss": 0.508,
+      "step": 8716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8659370475684836,
+      "learning_rate": 4.13663894860873e-06,
+      "loss": 0.5388,
+      "step": 8717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8893485951346831,
+      "learning_rate": 4.13450667039533e-06,
+      "loss": 0.4901,
+      "step": 8718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8054311327775409,
+      "learning_rate": 4.132374798655413e-06,
+      "loss": 0.4823,
+      "step": 8719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0250186548566913,
+      "learning_rate": 4.130243333536718e-06,
+      "loss": 0.5844,
+      "step": 8720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9198453007877438,
+      "learning_rate": 4.128112275186952e-06,
+      "loss": 0.5061,
+      "step": 8721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9542831606083931,
+      "learning_rate": 4.125981623753801e-06,
+      "loss": 0.5177,
+      "step": 8722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9138929453969917,
+      "learning_rate": 4.1238513793849065e-06,
+      "loss": 0.4643,
+      "step": 8723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.008733247589399,
+      "learning_rate": 4.121721542227906e-06,
+      "loss": 0.4846,
+      "step": 8724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9343475870206165,
+      "learning_rate": 4.1195921124303864e-06,
+      "loss": 0.5228,
+      "step": 8725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9956146315203084,
+      "learning_rate": 4.117463090139916e-06,
+      "loss": 0.5167,
+      "step": 8726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9457657258776898,
+      "learning_rate": 4.1153344755040355e-06,
+      "loss": 0.5152,
+      "step": 8727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8830292824760735,
+      "learning_rate": 4.11320626867026e-06,
+      "loss": 0.4626,
+      "step": 8728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9490427326245271,
+      "learning_rate": 4.111078469786062e-06,
+      "loss": 0.4836,
+      "step": 8729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9092871656552273,
+      "learning_rate": 4.10895107899891e-06,
+      "loss": 0.458,
+      "step": 8730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.93005774304165,
+      "learning_rate": 4.106824096456217e-06,
+      "loss": 0.4641,
+      "step": 8731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8860861090356404,
+      "learning_rate": 4.104697522305388e-06,
+      "loss": 0.5386,
+      "step": 8732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.846867975757357,
+      "learning_rate": 4.102571356693793e-06,
+      "loss": 0.4506,
+      "step": 8733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9016018547200599,
+      "learning_rate": 4.100445599768774e-06,
+      "loss": 0.4755,
+      "step": 8734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9656100627050762,
+      "learning_rate": 4.098320251677637e-06,
+      "loss": 0.5042,
+      "step": 8735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8760587893762447,
+      "learning_rate": 4.096195312567677e-06,
+      "loss": 0.5041,
+      "step": 8736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9051519692958128,
+      "learning_rate": 4.094070782586141e-06,
+      "loss": 0.4741,
+      "step": 8737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9101216832063238,
+      "learning_rate": 4.091946661880262e-06,
+      "loss": 0.5247,
+      "step": 8738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9338495810228892,
+      "learning_rate": 4.089822950597239e-06,
+      "loss": 0.4601,
+      "step": 8739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9623678853244182,
+      "learning_rate": 4.087699648884248e-06,
+      "loss": 0.5257,
+      "step": 8740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3681873907347593,
+      "learning_rate": 4.085576756888418e-06,
+      "loss": 0.49,
+      "step": 8741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8947410753937106,
+      "learning_rate": 4.083454274756881e-06,
+      "loss": 0.5437,
+      "step": 8742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8560471468894103,
+      "learning_rate": 4.081332202636711e-06,
+      "loss": 0.553,
+      "step": 8743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8614050583496167,
+      "learning_rate": 4.07921054067497e-06,
+      "loss": 0.5101,
+      "step": 8744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8858829368669646,
+      "learning_rate": 4.0770892890186854e-06,
+      "loss": 0.5019,
+      "step": 8745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8358142772584893,
+      "learning_rate": 4.074968447814865e-06,
+      "loss": 0.4682,
+      "step": 8746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.008198929727442,
+      "learning_rate": 4.072848017210467e-06,
+      "loss": 0.4819,
+      "step": 8747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.831564582604135,
+      "learning_rate": 4.070727997352451e-06,
+      "loss": 0.4379,
+      "step": 8748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9825779377624262,
+      "learning_rate": 4.068608388387722e-06,
+      "loss": 0.5053,
+      "step": 8749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0133746817461498,
+      "learning_rate": 4.066489190463171e-06,
+      "loss": 0.4938,
+      "step": 8750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0040275358581465,
+      "learning_rate": 4.0643704037256556e-06,
+      "loss": 0.5035,
+      "step": 8751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8856134943574181,
+      "learning_rate": 4.0622520283220115e-06,
+      "loss": 0.491,
+      "step": 8752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8965321999288322,
+      "learning_rate": 4.060134064399026e-06,
+      "loss": 0.4345,
+      "step": 8753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8186774390161647,
+      "learning_rate": 4.05801651210349e-06,
+      "loss": 0.4117,
+      "step": 8754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0365583157244873,
+      "learning_rate": 4.0558993715821335e-06,
+      "loss": 0.5259,
+      "step": 8755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8711117721422899,
+      "learning_rate": 4.053782642981679e-06,
+      "loss": 0.4699,
+      "step": 8756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8808213604888707,
+      "learning_rate": 4.0516663264488145e-06,
+      "loss": 0.4718,
+      "step": 8757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9377918559791827,
+      "learning_rate": 4.049550422130196e-06,
+      "loss": 0.4345,
+      "step": 8758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8839879672109973,
+      "learning_rate": 4.047434930172456e-06,
+      "loss": 0.5011,
+      "step": 8759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8693425125764165,
+      "learning_rate": 4.045319850722198e-06,
+      "loss": 0.4625,
+      "step": 8760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9119674025694322,
+      "learning_rate": 4.04320518392599e-06,
+      "loss": 0.5342,
+      "step": 8761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7183687244172643,
+      "learning_rate": 4.041090929930378e-06,
+      "loss": 0.5476,
+      "step": 8762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9880235645483642,
+      "learning_rate": 4.03897708888188e-06,
+      "loss": 0.5248,
+      "step": 8763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.871714760467819,
+      "learning_rate": 4.036863660926982e-06,
+      "loss": 0.4717,
+      "step": 8764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9458534642219324,
+      "learning_rate": 4.0347506462121434e-06,
+      "loss": 0.5022,
+      "step": 8765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.893834833687743,
+      "learning_rate": 4.032638044883796e-06,
+      "loss": 0.4513,
+      "step": 8766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9164768678852234,
+      "learning_rate": 4.0305258570883336e-06,
+      "loss": 0.4428,
+      "step": 8767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9575711407404988,
+      "learning_rate": 4.028414082972141e-06,
+      "loss": 0.4718,
+      "step": 8768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1116004691640229,
+      "learning_rate": 4.026302722681551e-06,
+      "loss": 0.498,
+      "step": 8769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8936242693860171,
+      "learning_rate": 4.024191776362884e-06,
+      "loss": 0.4714,
+      "step": 8770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9006703505518066,
+      "learning_rate": 4.022081244162428e-06,
+      "loss": 0.5088,
+      "step": 8771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9518530865834943,
+      "learning_rate": 4.019971126226442e-06,
+      "loss": 0.5148,
+      "step": 8772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8983615821416384,
+      "learning_rate": 4.017861422701144e-06,
+      "loss": 0.5111,
+      "step": 8773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9018554754900465,
+      "learning_rate": 4.015752133732752e-06,
+      "loss": 0.5303,
+      "step": 8774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9615507125371773,
+      "learning_rate": 4.013643259467426e-06,
+      "loss": 0.5036,
+      "step": 8775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9221345131864341,
+      "learning_rate": 4.011534800051311e-06,
+      "loss": 0.509,
+      "step": 8776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9346443887263562,
+      "learning_rate": 4.0094267556305236e-06,
+      "loss": 0.5485,
+      "step": 8777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8919882717782461,
+      "learning_rate": 4.0073191263511475e-06,
+      "loss": 0.4993,
+      "step": 8778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9302969938319966,
+      "learning_rate": 4.005211912359241e-06,
+      "loss": 0.4702,
+      "step": 8779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9522390605496788,
+      "learning_rate": 4.003105113800835e-06,
+      "loss": 0.5209,
+      "step": 8780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8053579275810232,
+      "learning_rate": 4.000998730821922e-06,
+      "loss": 0.454,
+      "step": 8781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9899449056444444,
+      "learning_rate": 3.998892763568476e-06,
+      "loss": 0.5019,
+      "step": 8782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9943263088056388,
+      "learning_rate": 3.996787212186438e-06,
+      "loss": 0.5298,
+      "step": 8783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9242626375286663,
+      "learning_rate": 3.994682076821721e-06,
+      "loss": 0.501,
+      "step": 8784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9675664947615202,
+      "learning_rate": 3.99257735762021e-06,
+      "loss": 0.515,
+      "step": 8785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8811697102442329,
+      "learning_rate": 3.990473054727764e-06,
+      "loss": 0.4832,
+      "step": 8786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9777855258544456,
+      "learning_rate": 3.988369168290199e-06,
+      "loss": 0.5334,
+      "step": 8787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0989247905108661,
+      "learning_rate": 3.98626569845332e-06,
+      "loss": 0.5236,
+      "step": 8788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0143168723303095,
+      "learning_rate": 3.984162645362893e-06,
+      "loss": 0.5851,
+      "step": 8789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8706688945273519,
+      "learning_rate": 3.98206000916466e-06,
+      "loss": 0.5129,
+      "step": 8790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0117501135253928,
+      "learning_rate": 3.97995779000433e-06,
+      "loss": 0.4835,
+      "step": 8791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9361724235863483,
+      "learning_rate": 3.977855988027585e-06,
+      "loss": 0.5167,
+      "step": 8792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9441390871206049,
+      "learning_rate": 3.975754603380082e-06,
+      "loss": 0.52,
+      "step": 8793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9325045022867406,
+      "learning_rate": 3.973653636207437e-06,
+      "loss": 0.501,
+      "step": 8794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8737033160786428,
+      "learning_rate": 3.971553086655251e-06,
+      "loss": 0.5012,
+      "step": 8795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8631823070347903,
+      "learning_rate": 3.969452954869089e-06,
+      "loss": 0.5286,
+      "step": 8796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8817665001884151,
+      "learning_rate": 3.967353240994487e-06,
+      "loss": 0.4844,
+      "step": 8797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9297465007856393,
+      "learning_rate": 3.9652539451769554e-06,
+      "loss": 0.4972,
+      "step": 8798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0959093161662927,
+      "learning_rate": 3.963155067561976e-06,
+      "loss": 0.4475,
+      "step": 8799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8978766746333792,
+      "learning_rate": 3.961056608294992e-06,
+      "loss": 0.4925,
+      "step": 8800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9327436321983399,
+      "learning_rate": 3.958958567521428e-06,
+      "loss": 0.4779,
+      "step": 8801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.923248650172493,
+      "learning_rate": 3.956860945386677e-06,
+      "loss": 0.4828,
+      "step": 8802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9605785520965624,
+      "learning_rate": 3.954763742036103e-06,
+      "loss": 0.5256,
+      "step": 8803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8488022308694663,
+      "learning_rate": 3.952666957615039e-06,
+      "loss": 0.4822,
+      "step": 8804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.012607013833751,
+      "learning_rate": 3.950570592268794e-06,
+      "loss": 0.4769,
+      "step": 8805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8621331876085766,
+      "learning_rate": 3.948474646142638e-06,
+      "loss": 0.4088,
+      "step": 8806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9353129675040686,
+      "learning_rate": 3.946379119381822e-06,
+      "loss": 0.4666,
+      "step": 8807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.928181887357087,
+      "learning_rate": 3.9442840121315625e-06,
+      "loss": 0.5957,
+      "step": 8808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8097859720618114,
+      "learning_rate": 3.94218932453705e-06,
+      "loss": 0.3988,
+      "step": 8809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8555817269549301,
+      "learning_rate": 3.940095056743444e-06,
+      "loss": 0.4944,
+      "step": 8810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8688428830218954,
+      "learning_rate": 3.938001208895878e-06,
+      "loss": 0.5357,
+      "step": 8811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8943758455626842,
+      "learning_rate": 3.935907781139446e-06,
+      "loss": 0.495,
+      "step": 8812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9526621111614739,
+      "learning_rate": 3.933814773619232e-06,
+      "loss": 0.4879,
+      "step": 8813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8697916155567965,
+      "learning_rate": 3.93172218648027e-06,
+      "loss": 0.474,
+      "step": 8814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.009523382502148,
+      "learning_rate": 3.929630019867579e-06,
+      "loss": 0.4285,
+      "step": 8815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0902489236515984,
+      "learning_rate": 3.927538273926141e-06,
+      "loss": 0.4748,
+      "step": 8816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9224848262996478,
+      "learning_rate": 3.92544694880092e-06,
+      "loss": 0.4925,
+      "step": 8817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.928690974890286,
+      "learning_rate": 3.923356044636829e-06,
+      "loss": 0.449,
+      "step": 8818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0201833658121,
+      "learning_rate": 3.921265561578781e-06,
+      "loss": 0.548,
+      "step": 8819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9215773732856094,
+      "learning_rate": 3.919175499771635e-06,
+      "loss": 0.4728,
+      "step": 8820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9183799946338584,
+      "learning_rate": 3.917085859360234e-06,
+      "loss": 0.4792,
+      "step": 8821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9020914043322761,
+      "learning_rate": 3.9149966404893854e-06,
+      "loss": 0.4599,
+      "step": 8822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0179660059920808,
+      "learning_rate": 3.912907843303877e-06,
+      "loss": 0.5157,
+      "step": 8823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8934013813007406,
+      "learning_rate": 3.910819467948448e-06,
+      "loss": 0.4946,
+      "step": 8824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9137460578656958,
+      "learning_rate": 3.908731514567836e-06,
+      "loss": 0.4284,
+      "step": 8825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8983840220601542,
+      "learning_rate": 3.906643983306724e-06,
+      "loss": 0.4622,
+      "step": 8826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8299218289902504,
+      "learning_rate": 3.904556874309779e-06,
+      "loss": 0.4446,
+      "step": 8827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8824793357068349,
+      "learning_rate": 3.902470187721636e-06,
+      "loss": 0.4705,
+      "step": 8828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9958803430976835,
+      "learning_rate": 3.900383923686905e-06,
+      "loss": 0.5064,
+      "step": 8829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.908448281142226,
+      "learning_rate": 3.898298082350149e-06,
+      "loss": 0.5247,
+      "step": 8830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9807265834771793,
+      "learning_rate": 3.896212663855932e-06,
+      "loss": 0.5566,
+      "step": 8831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.918857203642734,
+      "learning_rate": 3.894127668348759e-06,
+      "loss": 0.4687,
+      "step": 8832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9019988377097433,
+      "learning_rate": 3.892043095973123e-06,
+      "loss": 0.4991,
+      "step": 8833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9049263772974447,
+      "learning_rate": 3.889958946873482e-06,
+      "loss": 0.4594,
+      "step": 8834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1347811419110758,
+      "learning_rate": 3.887875221194271e-06,
+      "loss": 0.5158,
+      "step": 8835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.827489198544077,
+      "learning_rate": 3.885791919079878e-06,
+      "loss": 0.4375,
+      "step": 8836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8444464095083356,
+      "learning_rate": 3.883709040674688e-06,
+      "loss": 0.4174,
+      "step": 8837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9260301120795823,
+      "learning_rate": 3.881626586123034e-06,
+      "loss": 0.4922,
+      "step": 8838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9377917778876699,
+      "learning_rate": 3.8795445555692305e-06,
+      "loss": 0.5823,
+      "step": 8839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9395718084838521,
+      "learning_rate": 3.87746294915756e-06,
+      "loss": 0.5049,
+      "step": 8840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9314406826381335,
+      "learning_rate": 3.87538176703228e-06,
+      "loss": 0.4928,
+      "step": 8841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9542956906849946,
+      "learning_rate": 3.873301009337604e-06,
+      "loss": 0.4691,
+      "step": 8842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8688695931633011,
+      "learning_rate": 3.871220676217742e-06,
+      "loss": 0.4676,
+      "step": 8843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.986603792422116,
+      "learning_rate": 3.869140767816846e-06,
+      "loss": 0.5367,
+      "step": 8844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9600355737891579,
+      "learning_rate": 3.867061284279058e-06,
+      "loss": 0.5112,
+      "step": 8845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9573543631323532,
+      "learning_rate": 3.864982225748481e-06,
+      "loss": 0.5086,
+      "step": 8846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8153846929151147,
+      "learning_rate": 3.862903592369199e-06,
+      "loss": 0.455,
+      "step": 8847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9317872841841982,
+      "learning_rate": 3.860825384285247e-06,
+      "loss": 0.537,
+      "step": 8848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9100160868925283,
+      "learning_rate": 3.858747601640658e-06,
+      "loss": 0.4817,
+      "step": 8849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8958109692096803,
+      "learning_rate": 3.856670244579409e-06,
+      "loss": 0.4736,
+      "step": 8850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8956550140399889,
+      "learning_rate": 3.854593313245463e-06,
+      "loss": 0.4674,
+      "step": 8851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9936883431069679,
+      "learning_rate": 3.852516807782749e-06,
+      "loss": 0.5141,
+      "step": 8852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9475637588463383,
+      "learning_rate": 3.850440728335171e-06,
+      "loss": 0.4938,
+      "step": 8853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9297870299422121,
+      "learning_rate": 3.848365075046589e-06,
+      "loss": 0.526,
+      "step": 8854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8497992598755991,
+      "learning_rate": 3.846289848060858e-06,
+      "loss": 0.4832,
+      "step": 8855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9046064998492238,
+      "learning_rate": 3.844215047521779e-06,
+      "loss": 0.5173,
+      "step": 8856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0890530082437015,
+      "learning_rate": 3.842140673573136e-06,
+      "loss": 0.5348,
+      "step": 8857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.002209227608526,
+      "learning_rate": 3.840066726358683e-06,
+      "loss": 0.4426,
+      "step": 8858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8119903241278402,
+      "learning_rate": 3.837993206022146e-06,
+      "loss": 0.4799,
+      "step": 8859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9668063740346428,
+      "learning_rate": 3.8359201127072065e-06,
+      "loss": 0.524,
+      "step": 8860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8836622588003586,
+      "learning_rate": 3.8338474465575425e-06,
+      "loss": 0.5279,
+      "step": 8861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9555312989641757,
+      "learning_rate": 3.831775207716778e-06,
+      "loss": 0.4856,
+      "step": 8862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9065472201631909,
+      "learning_rate": 3.82970339632852e-06,
+      "loss": 0.5127,
+      "step": 8863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8870497018587068,
+      "learning_rate": 3.827632012536344e-06,
+      "loss": 0.4403,
+      "step": 8864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8824030645406618,
+      "learning_rate": 3.825561056483798e-06,
+      "loss": 0.5422,
+      "step": 8865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8760984152703735,
+      "learning_rate": 3.823490528314387e-06,
+      "loss": 0.4916,
+      "step": 8866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9023598070151023,
+      "learning_rate": 3.821420428171611e-06,
+      "loss": 0.5103,
+      "step": 8867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9606187506813286,
+      "learning_rate": 3.819350756198915e-06,
+      "loss": 0.5248,
+      "step": 8868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9106250445068397,
+      "learning_rate": 3.81728151253973e-06,
+      "loss": 0.3984,
+      "step": 8869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8406825255854373,
+      "learning_rate": 3.815212697337451e-06,
+      "loss": 0.533,
+      "step": 8870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8989819504597925,
+      "learning_rate": 3.8131443107354503e-06,
+      "loss": 0.5037,
+      "step": 8871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7929612535636658,
+      "learning_rate": 3.8110763528770543e-06,
+      "loss": 0.3884,
+      "step": 8872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.814849985403055,
+      "learning_rate": 3.8090088239055843e-06,
+      "loss": 0.4299,
+      "step": 8873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.857194615734218,
+      "learning_rate": 3.8069417239643082e-06,
+      "loss": 0.4467,
+      "step": 8874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8907631497574886,
+      "learning_rate": 3.804875053196477e-06,
+      "loss": 0.4932,
+      "step": 8875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8830049019992933,
+      "learning_rate": 3.80280881174531e-06,
+      "loss": 0.486,
+      "step": 8876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9476380674101876,
+      "learning_rate": 3.800742999753999e-06,
+      "loss": 0.4813,
+      "step": 8877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9612638714238807,
+      "learning_rate": 3.7986776173656927e-06,
+      "loss": 0.4955,
+      "step": 8878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9275579157713626,
+      "learning_rate": 3.7966126647235326e-06,
+      "loss": 0.4744,
+      "step": 8879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9605038116462703,
+      "learning_rate": 3.79454814197061e-06,
+      "loss": 0.4821,
+      "step": 8880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.871832545094725,
+      "learning_rate": 3.792484049249996e-06,
+      "loss": 0.5148,
+      "step": 8881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0045689213531521,
+      "learning_rate": 3.790420386704733e-06,
+      "loss": 0.5913,
+      "step": 8882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8229711090759823,
+      "learning_rate": 3.788357154477831e-06,
+      "loss": 0.4665,
+      "step": 8883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.92957969362573,
+      "learning_rate": 3.786294352712262e-06,
+      "loss": 0.4311,
+      "step": 8884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8949166464230011,
+      "learning_rate": 3.784231981550991e-06,
+      "loss": 0.4892,
+      "step": 8885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9962607277158397,
+      "learning_rate": 3.782170041136922e-06,
+      "loss": 0.5495,
+      "step": 8886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.958883598870349,
+      "learning_rate": 3.7801085316129615e-06,
+      "loss": 0.5129,
+      "step": 8887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9222815789807826,
+      "learning_rate": 3.778047453121958e-06,
+      "loss": 0.5116,
+      "step": 8888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.030209109733231,
+      "learning_rate": 3.7759868058067483e-06,
+      "loss": 0.5329,
+      "step": 8889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0035102607125828,
+      "learning_rate": 3.773926589810133e-06,
+      "loss": 0.4956,
+      "step": 8890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8946525169233979,
+      "learning_rate": 3.7718668052748842e-06,
+      "loss": 0.4211,
+      "step": 8891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7977790198384506,
+      "learning_rate": 3.7698074523437355e-06,
+      "loss": 0.4643,
+      "step": 8892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0180235740982717,
+      "learning_rate": 3.7677485311594107e-06,
+      "loss": 0.5146,
+      "step": 8893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9956941872006838,
+      "learning_rate": 3.7656900418645826e-06,
+      "loss": 0.5284,
+      "step": 8894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8680531550447771,
+      "learning_rate": 3.763631984601903e-06,
+      "loss": 0.4795,
+      "step": 8895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8941870294776818,
+      "learning_rate": 3.7615743595139965e-06,
+      "loss": 0.4795,
+      "step": 8896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.911033857549809,
+      "learning_rate": 3.759517166743456e-06,
+      "loss": 0.5178,
+      "step": 8897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9492113344650935,
+      "learning_rate": 3.7574604064328336e-06,
+      "loss": 0.4921,
+      "step": 8898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.877478301058727,
+      "learning_rate": 3.7554040787246746e-06,
+      "loss": 0.4545,
+      "step": 8899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9549658963774811,
+      "learning_rate": 3.7533481837614717e-06,
+      "loss": 0.5422,
+      "step": 8900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8911457825497752,
+      "learning_rate": 3.7512927216856987e-06,
+      "loss": 0.5134,
+      "step": 8901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8854910809568246,
+      "learning_rate": 3.7492376926397966e-06,
+      "loss": 0.4683,
+      "step": 8902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9407344486002542,
+      "learning_rate": 3.7471830967661815e-06,
+      "loss": 0.489,
+      "step": 8903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8656497076194637,
+      "learning_rate": 3.745128934207225e-06,
+      "loss": 0.4724,
+      "step": 8904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8441131056004215,
+      "learning_rate": 3.743075205105292e-06,
+      "loss": 0.4291,
+      "step": 8905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9530959715206083,
+      "learning_rate": 3.7410219096026944e-06,
+      "loss": 0.5696,
+      "step": 8906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9786935110086086,
+      "learning_rate": 3.7389690478417273e-06,
+      "loss": 0.4892,
+      "step": 8907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9564410548228618,
+      "learning_rate": 3.7369166199646502e-06,
+      "loss": 0.492,
+      "step": 8908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9658684056753355,
+      "learning_rate": 3.7348646261137e-06,
+      "loss": 0.5069,
+      "step": 8909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8958230878728628,
+      "learning_rate": 3.732813066431068e-06,
+      "loss": 0.4948,
+      "step": 8910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9869591315367217,
+      "learning_rate": 3.730761941058938e-06,
+      "loss": 0.5447,
+      "step": 8911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.92782933804282,
+      "learning_rate": 3.7287112501394406e-06,
+      "loss": 0.5225,
+      "step": 8912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9217797291013208,
+      "learning_rate": 3.7266609938146912e-06,
+      "loss": 0.4856,
+      "step": 8913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9822645901036696,
+      "learning_rate": 3.724611172226771e-06,
+      "loss": 0.457,
+      "step": 8914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8908297362107798,
+      "learning_rate": 3.722561785517732e-06,
+      "loss": 0.4382,
+      "step": 8915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0721806990810707,
+      "learning_rate": 3.7205128338295884e-06,
+      "loss": 0.4757,
+      "step": 8916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8454484712366738,
+      "learning_rate": 3.718464317304341e-06,
+      "loss": 0.455,
+      "step": 8917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8733536681919967,
+      "learning_rate": 3.716416236083942e-06,
+      "loss": 0.4972,
+      "step": 8918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9702493580649202,
+      "learning_rate": 3.7143685903103242e-06,
+      "loss": 0.5163,
+      "step": 8919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8858387912709786,
+      "learning_rate": 3.7123213801253876e-06,
+      "loss": 0.4982,
+      "step": 8920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9127452804890583,
+      "learning_rate": 3.7102746056710025e-06,
+      "loss": 0.4876,
+      "step": 8921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8992842307594613,
+      "learning_rate": 3.708228267089008e-06,
+      "loss": 0.5092,
+      "step": 8922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8843837452095896,
+      "learning_rate": 3.706182364521217e-06,
+      "loss": 0.4668,
+      "step": 8923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0186044909447514,
+      "learning_rate": 3.704136898109403e-06,
+      "loss": 0.5457,
+      "step": 8924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9663204881717411,
+      "learning_rate": 3.7020918679953166e-06,
+      "loss": 0.5576,
+      "step": 8925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9004867321863538,
+      "learning_rate": 3.7000472743206773e-06,
+      "loss": 0.4949,
+      "step": 8926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8848182522515886,
+      "learning_rate": 3.698003117227175e-06,
+      "loss": 0.5097,
+      "step": 8927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9150649446475405,
+      "learning_rate": 3.6959593968564654e-06,
+      "loss": 0.454,
+      "step": 8928
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8919034910686628,
+      "learning_rate": 3.6939161133501823e-06,
+      "loss": 0.5015,
+      "step": 8929
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.953865934457871,
+      "learning_rate": 3.691873266849916e-06,
+      "loss": 0.5266,
+      "step": 8930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9061766912229933,
+      "learning_rate": 3.6898308574972365e-06,
+      "loss": 0.5061,
+      "step": 8931
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1822247801890384,
+      "learning_rate": 3.6877888854336808e-06,
+      "loss": 0.5964,
+      "step": 8932
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8468554061541979,
+      "learning_rate": 3.6857473508007567e-06,
+      "loss": 0.4568,
+      "step": 8933
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9397863791139718,
+      "learning_rate": 3.6837062537399414e-06,
+      "loss": 0.5087,
+      "step": 8934
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9095373378132872,
+      "learning_rate": 3.6816655943926825e-06,
+      "loss": 0.498,
+      "step": 8935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.885010856329452,
+      "learning_rate": 3.6796253729003905e-06,
+      "loss": 0.5331,
+      "step": 8936
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0204269253203586,
+      "learning_rate": 3.6775855894044543e-06,
+      "loss": 0.5541,
+      "step": 8937
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.947069376220401,
+      "learning_rate": 3.6755462440462288e-06,
+      "loss": 0.4936,
+      "step": 8938
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9047457184730575,
+      "learning_rate": 3.673507336967038e-06,
+      "loss": 0.5206,
+      "step": 8939
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.045100952717416,
+      "learning_rate": 3.6714688683081778e-06,
+      "loss": 0.4995,
+      "step": 8940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9339554233023288,
+      "learning_rate": 3.669430838210911e-06,
+      "loss": 0.4846,
+      "step": 8941
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9311600655124447,
+      "learning_rate": 3.6673932468164763e-06,
+      "loss": 0.5243,
+      "step": 8942
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9600652156787634,
+      "learning_rate": 3.6653560942660694e-06,
+      "loss": 0.5025,
+      "step": 8943
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.875907664889498,
+      "learning_rate": 3.663319380700865e-06,
+      "loss": 0.4649,
+      "step": 8944
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9481821720808001,
+      "learning_rate": 3.661283106262008e-06,
+      "loss": 0.4747,
+      "step": 8945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0168558120772304,
+      "learning_rate": 3.659247271090609e-06,
+      "loss": 0.4481,
+      "step": 8946
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9162316079435071,
+      "learning_rate": 3.6572118753277495e-06,
+      "loss": 0.4684,
+      "step": 8947
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9225979331389788,
+      "learning_rate": 3.655176919114485e-06,
+      "loss": 0.4973,
+      "step": 8948
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9014094862874904,
+      "learning_rate": 3.6531424025918284e-06,
+      "loss": 0.5053,
+      "step": 8949
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.970179655664484,
+      "learning_rate": 3.651108325900773e-06,
+      "loss": 0.466,
+      "step": 8950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.032032574074345,
+      "learning_rate": 3.6490746891822806e-06,
+      "loss": 0.5263,
+      "step": 8951
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9129906619640558,
+      "learning_rate": 3.647041492577278e-06,
+      "loss": 0.5223,
+      "step": 8952
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8550106008853208,
+      "learning_rate": 3.645008736226664e-06,
+      "loss": 0.4532,
+      "step": 8953
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.801384404024512,
+      "learning_rate": 3.6429764202713124e-06,
+      "loss": 0.4851,
+      "step": 8954
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9889555676924645,
+      "learning_rate": 3.6409445448520533e-06,
+      "loss": 0.5049,
+      "step": 8955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9403531883891941,
+      "learning_rate": 3.6389131101096953e-06,
+      "loss": 0.536,
+      "step": 8956
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0502677111502992,
+      "learning_rate": 3.6368821161850176e-06,
+      "loss": 0.5614,
+      "step": 8957
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8630171779379783,
+      "learning_rate": 3.6348515632187643e-06,
+      "loss": 0.4034,
+      "step": 8958
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9047027156195153,
+      "learning_rate": 3.6328214513516523e-06,
+      "loss": 0.4744,
+      "step": 8959
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.916757086523125,
+      "learning_rate": 3.6307917807243697e-06,
+      "loss": 0.4842,
+      "step": 8960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9940495108617984,
+      "learning_rate": 3.6287625514775602e-06,
+      "loss": 0.5406,
+      "step": 8961
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9253795932773297,
+      "learning_rate": 3.626733763751861e-06,
+      "loss": 0.4978,
+      "step": 8962
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9150118745581817,
+      "learning_rate": 3.624705417687856e-06,
+      "loss": 0.4843,
+      "step": 8963
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8946597973175461,
+      "learning_rate": 3.6226775134261106e-06,
+      "loss": 0.455,
+      "step": 8964
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8485633591778271,
+      "learning_rate": 3.620650051107156e-06,
+      "loss": 0.5026,
+      "step": 8965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8989951788879078,
+      "learning_rate": 3.6186230308714985e-06,
+      "loss": 0.4281,
+      "step": 8966
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9932172971365646,
+      "learning_rate": 3.6165964528595988e-06,
+      "loss": 0.4487,
+      "step": 8967
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9419138818036944,
+      "learning_rate": 3.6145703172119085e-06,
+      "loss": 0.5094,
+      "step": 8968
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0242264441554165,
+      "learning_rate": 3.6125446240688276e-06,
+      "loss": 0.533,
+      "step": 8969
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9880175132776277,
+      "learning_rate": 3.61051937357074e-06,
+      "loss": 0.5336,
+      "step": 8970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9623432157150716,
+      "learning_rate": 3.6084945658579918e-06,
+      "loss": 0.5216,
+      "step": 8971
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9547697834135409,
+      "learning_rate": 3.606470201070904e-06,
+      "loss": 0.4688,
+      "step": 8972
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9583453259256163,
+      "learning_rate": 3.6044462793497526e-06,
+      "loss": 0.5612,
+      "step": 8973
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9720258126741251,
+      "learning_rate": 3.6024228008348096e-06,
+      "loss": 0.4992,
+      "step": 8974
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9474217582347544,
+      "learning_rate": 3.600399765666287e-06,
+      "loss": 0.5311,
+      "step": 8975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9902690886560699,
+      "learning_rate": 3.5983771739843855e-06,
+      "loss": 0.5334,
+      "step": 8976
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9613087078021019,
+      "learning_rate": 3.596355025929267e-06,
+      "loss": 0.4961,
+      "step": 8977
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9757830970674216,
+      "learning_rate": 3.594333321641068e-06,
+      "loss": 0.5468,
+      "step": 8978
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9347073297449352,
+      "learning_rate": 3.5923120612598828e-06,
+      "loss": 0.4568,
+      "step": 8979
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9029241676866938,
+      "learning_rate": 3.590291244925793e-06,
+      "loss": 0.4688,
+      "step": 8980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8488345626658037,
+      "learning_rate": 3.588270872778833e-06,
+      "loss": 0.4678,
+      "step": 8981
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9033708830753971,
+      "learning_rate": 3.5862509449590135e-06,
+      "loss": 0.4801,
+      "step": 8982
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9888788145320088,
+      "learning_rate": 3.5842314616063134e-06,
+      "loss": 0.5105,
+      "step": 8983
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8813975859723068,
+      "learning_rate": 3.582212422860687e-06,
+      "loss": 0.4764,
+      "step": 8984
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0035285428927048,
+      "learning_rate": 3.5801938288620395e-06,
+      "loss": 0.4795,
+      "step": 8985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9402018886597426,
+      "learning_rate": 3.5781756797502733e-06,
+      "loss": 0.4551,
+      "step": 8986
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9275798940011335,
+      "learning_rate": 3.576157975665232e-06,
+      "loss": 0.5709,
+      "step": 8987
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.88546274771075,
+      "learning_rate": 3.5741407167467444e-06,
+      "loss": 0.4388,
+      "step": 8988
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8964212413484605,
+      "learning_rate": 3.5721239031346067e-06,
+      "loss": 0.4633,
+      "step": 8989
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8716362942550768,
+      "learning_rate": 3.5701075349685842e-06,
+      "loss": 0.4732,
+      "step": 8990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.872765870527491,
+      "learning_rate": 3.568091612388399e-06,
+      "loss": 0.4433,
+      "step": 8991
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8806035542732507,
+      "learning_rate": 3.566076135533767e-06,
+      "loss": 0.4762,
+      "step": 8992
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9063579768886735,
+      "learning_rate": 3.5640611045443485e-06,
+      "loss": 0.5312,
+      "step": 8993
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1252243359435834,
+      "learning_rate": 3.5620465195597865e-06,
+      "loss": 0.4534,
+      "step": 8994
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9411728999230131,
+      "learning_rate": 3.5600323807196912e-06,
+      "loss": 0.4822,
+      "step": 8995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9028802160371076,
+      "learning_rate": 3.5580186881636414e-06,
+      "loss": 0.519,
+      "step": 8996
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.918337987256864,
+      "learning_rate": 3.5560054420311776e-06,
+      "loss": 0.4788,
+      "step": 8997
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9989192754355186,
+      "learning_rate": 3.5539926424618265e-06,
+      "loss": 0.5386,
+      "step": 8998
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.838803831436354,
+      "learning_rate": 3.551980289595064e-06,
+      "loss": 0.4483,
+      "step": 8999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8445468941893491,
+      "learning_rate": 3.5499683835703493e-06,
+      "loss": 0.4687,
+      "step": 9000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.848857920123972,
+      "learning_rate": 3.547956924527103e-06,
+      "loss": 0.5277,
+      "step": 9001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8680873155519754,
+      "learning_rate": 3.5459459126047226e-06,
+      "loss": 0.4873,
+      "step": 9002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9469250103942338,
+      "learning_rate": 3.5439353479425597e-06,
+      "loss": 0.5117,
+      "step": 9003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0264133283896932,
+      "learning_rate": 3.5419252306799567e-06,
+      "loss": 0.5101,
+      "step": 9004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8621328186046997,
+      "learning_rate": 3.539915560956204e-06,
+      "loss": 0.4558,
+      "step": 9005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8616535313317472,
+      "learning_rate": 3.5379063389105727e-06,
+      "loss": 0.4193,
+      "step": 9006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8940689750239936,
+      "learning_rate": 3.5358975646823002e-06,
+      "loss": 0.5044,
+      "step": 9007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9983698752868682,
+      "learning_rate": 3.533889238410596e-06,
+      "loss": 0.5353,
+      "step": 9008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1563379381969516,
+      "learning_rate": 3.5318813602346257e-06,
+      "loss": 0.5731,
+      "step": 9009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8336786353501896,
+      "learning_rate": 3.529873930293546e-06,
+      "loss": 0.457,
+      "step": 9010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9314020854618265,
+      "learning_rate": 3.5278669487264583e-06,
+      "loss": 0.5305,
+      "step": 9011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9373727014930834,
+      "learning_rate": 3.525860415672456e-06,
+      "loss": 0.5118,
+      "step": 9012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.036601676300659,
+      "learning_rate": 3.523854331270582e-06,
+      "loss": 0.5551,
+      "step": 9013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9470780237472647,
+      "learning_rate": 3.5218486956598573e-06,
+      "loss": 0.4952,
+      "step": 9014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9371280542011889,
+      "learning_rate": 3.5198435089792726e-06,
+      "loss": 0.5261,
+      "step": 9015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9707691354773395,
+      "learning_rate": 3.5178387713677886e-06,
+      "loss": 0.4942,
+      "step": 9016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9204704483209045,
+      "learning_rate": 3.515834482964321e-06,
+      "loss": 0.5014,
+      "step": 9017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8861801504688388,
+      "learning_rate": 3.5138306439077784e-06,
+      "loss": 0.4825,
+      "step": 9018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8744753752055189,
+      "learning_rate": 3.5118272543370157e-06,
+      "loss": 0.4632,
+      "step": 9019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9304866500057495,
+      "learning_rate": 3.5098243143908694e-06,
+      "loss": 0.5482,
+      "step": 9020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9796429237511085,
+      "learning_rate": 3.507821824208142e-06,
+      "loss": 0.5174,
+      "step": 9021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8162021792821621,
+      "learning_rate": 3.5058197839276064e-06,
+      "loss": 0.4494,
+      "step": 9022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8470310632620339,
+      "learning_rate": 3.5038181936879932e-06,
+      "loss": 0.4983,
+      "step": 9023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9326230422525369,
+      "learning_rate": 3.5018170536280237e-06,
+      "loss": 0.4876,
+      "step": 9024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.081302621542103,
+      "learning_rate": 3.4998163638863646e-06,
+      "loss": 0.5806,
+      "step": 9025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8398561267240168,
+      "learning_rate": 3.4978161246016664e-06,
+      "loss": 0.5022,
+      "step": 9026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.008510871462587,
+      "learning_rate": 3.495816335912543e-06,
+      "loss": 0.5308,
+      "step": 9027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.813490241806617,
+      "learning_rate": 3.493816997957582e-06,
+      "loss": 0.3922,
+      "step": 9028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8522965620275039,
+      "learning_rate": 3.4918181108753247e-06,
+      "loss": 0.447,
+      "step": 9029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0595297523859928,
+      "learning_rate": 3.4898196748043065e-06,
+      "loss": 0.4964,
+      "step": 9030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9004845069622794,
+      "learning_rate": 3.4878216898830074e-06,
+      "loss": 0.4881,
+      "step": 9031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9295585565321477,
+      "learning_rate": 3.4858241562498884e-06,
+      "loss": 0.5394,
+      "step": 9032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9078302483781503,
+      "learning_rate": 3.4838270740433776e-06,
+      "loss": 0.4755,
+      "step": 9033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.891665135254238,
+      "learning_rate": 3.4818304434018734e-06,
+      "loss": 0.4798,
+      "step": 9034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8578424459037337,
+      "learning_rate": 3.4798342644637327e-06,
+      "loss": 0.5138,
+      "step": 9035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8970354823693978,
+      "learning_rate": 3.4778385373672996e-06,
+      "loss": 0.4402,
+      "step": 9036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8832316717638571,
+      "learning_rate": 3.4758432622508677e-06,
+      "loss": 0.4979,
+      "step": 9037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8935845014449028,
+      "learning_rate": 3.4738484392527107e-06,
+      "loss": 0.4677,
+      "step": 9038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9709035988811506,
+      "learning_rate": 3.471854068511068e-06,
+      "loss": 0.4925,
+      "step": 9039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.99935079521221,
+      "learning_rate": 3.4698601501641517e-06,
+      "loss": 0.4836,
+      "step": 9040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9496127903701702,
+      "learning_rate": 3.4678666843501276e-06,
+      "loss": 0.4915,
+      "step": 9041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8638931125885806,
+      "learning_rate": 3.465873671207155e-06,
+      "loss": 0.5133,
+      "step": 9042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8278711857326407,
+      "learning_rate": 3.4638811108733383e-06,
+      "loss": 0.4489,
+      "step": 9043
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9039654110449227,
+      "learning_rate": 3.4618890034867626e-06,
+      "loss": 0.5066,
+      "step": 9044
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9173666067331488,
+      "learning_rate": 3.4598973491854804e-06,
+      "loss": 0.473,
+      "step": 9045
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0228975566013427,
+      "learning_rate": 3.4579061481075137e-06,
+      "loss": 0.5098,
+      "step": 9046
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8442960933660839,
+      "learning_rate": 3.455915400390841e-06,
+      "loss": 0.4608,
+      "step": 9047
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9276664713771727,
+      "learning_rate": 3.4539251061734337e-06,
+      "loss": 0.5207,
+      "step": 9048
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8647425352503317,
+      "learning_rate": 3.451935265593207e-06,
+      "loss": 0.5113,
+      "step": 9049
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9182509306051623,
+      "learning_rate": 3.449945878788058e-06,
+      "loss": 0.4752,
+      "step": 9050
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9667703242017811,
+      "learning_rate": 3.4479569458958494e-06,
+      "loss": 0.5247,
+      "step": 9051
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.890490488057399,
+      "learning_rate": 3.4459684670544157e-06,
+      "loss": 0.4682,
+      "step": 9052
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8894062584761423,
+      "learning_rate": 3.4439804424015486e-06,
+      "loss": 0.4623,
+      "step": 9053
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1678266481326605,
+      "learning_rate": 3.4419928720750274e-06,
+      "loss": 0.5092,
+      "step": 9054
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9180142808067063,
+      "learning_rate": 3.44000575621258e-06,
+      "loss": 0.4445,
+      "step": 9055
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9747151736279898,
+      "learning_rate": 3.4380190949519155e-06,
+      "loss": 0.5326,
+      "step": 9056
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8864171290008381,
+      "learning_rate": 3.4360328884307058e-06,
+      "loss": 0.4892,
+      "step": 9057
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9903037454617845,
+      "learning_rate": 3.4340471367865992e-06,
+      "loss": 0.5742,
+      "step": 9058
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8513222506288766,
+      "learning_rate": 3.432061840157196e-06,
+      "loss": 0.5226,
+      "step": 9059
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0307240084812919,
+      "learning_rate": 3.4300769986800863e-06,
+      "loss": 0.5663,
+      "step": 9060
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9968384356366111,
+      "learning_rate": 3.4280926124928115e-06,
+      "loss": 0.5288,
+      "step": 9061
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9519071916218125,
+      "learning_rate": 3.4261086817328882e-06,
+      "loss": 0.4699,
+      "step": 9062
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8450654754196741,
+      "learning_rate": 3.424125206537803e-06,
+      "loss": 0.4618,
+      "step": 9063
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8698269817828128,
+      "learning_rate": 3.422142187045011e-06,
+      "loss": 0.4738,
+      "step": 9064
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8777962455889311,
+      "learning_rate": 3.4201596233919243e-06,
+      "loss": 0.455,
+      "step": 9065
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9974300937915439,
+      "learning_rate": 3.418177515715947e-06,
+      "loss": 0.6206,
+      "step": 9066
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9838505746206851,
+      "learning_rate": 3.416195864154426e-06,
+      "loss": 0.525,
+      "step": 9067
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0055224851762832,
+      "learning_rate": 3.414214668844691e-06,
+      "loss": 0.4833,
+      "step": 9068
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.012178258142545,
+      "learning_rate": 3.4122339299240383e-06,
+      "loss": 0.5122,
+      "step": 9069
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9477903204487034,
+      "learning_rate": 3.410253647529731e-06,
+      "loss": 0.5254,
+      "step": 9070
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9954868910750313,
+      "learning_rate": 3.408273821799001e-06,
+      "loss": 0.4825,
+      "step": 9071
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9441065536533816,
+      "learning_rate": 3.4062944528690512e-06,
+      "loss": 0.4791,
+      "step": 9072
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9837226393405486,
+      "learning_rate": 3.4043155408770435e-06,
+      "loss": 0.5217,
+      "step": 9073
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.883534359457053,
+      "learning_rate": 3.4023370859601192e-06,
+      "loss": 0.4569,
+      "step": 9074
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8829348070197951,
+      "learning_rate": 3.400359088255383e-06,
+      "loss": 0.4665,
+      "step": 9075
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9242326361724409,
+      "learning_rate": 3.3983815478999073e-06,
+      "loss": 0.4844,
+      "step": 9076
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8751730630958149,
+      "learning_rate": 3.396404465030735e-06,
+      "loss": 0.5159,
+      "step": 9077
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9524856623633947,
+      "learning_rate": 3.3944278397848797e-06,
+      "loss": 0.4986,
+      "step": 9078
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9643769255032821,
+      "learning_rate": 3.3924516722993115e-06,
+      "loss": 0.5314,
+      "step": 9079
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9979175564097665,
+      "learning_rate": 3.3904759627109828e-06,
+      "loss": 0.5332,
+      "step": 9080
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9073140095491039,
+      "learning_rate": 3.388500711156807e-06,
+      "loss": 0.4793,
+      "step": 9081
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8235618593206219,
+      "learning_rate": 3.3865259177736663e-06,
+      "loss": 0.4428,
+      "step": 9082
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.945033382039681,
+      "learning_rate": 3.3845515826984143e-06,
+      "loss": 0.576,
+      "step": 9083
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8513977017652784,
+      "learning_rate": 3.3825777060678734e-06,
+      "loss": 0.4843,
+      "step": 9084
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.938115084213264,
+      "learning_rate": 3.380604288018824e-06,
+      "loss": 0.481,
+      "step": 9085
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9257780529818783,
+      "learning_rate": 3.3786313286880257e-06,
+      "loss": 0.4521,
+      "step": 9086
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9387396351436141,
+      "learning_rate": 3.3766588282122037e-06,
+      "loss": 0.458,
+      "step": 9087
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9144270665316344,
+      "learning_rate": 3.3746867867280496e-06,
+      "loss": 0.5025,
+      "step": 9088
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8722399311563757,
+      "learning_rate": 3.3727152043722257e-06,
+      "loss": 0.4524,
+      "step": 9089
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.955180665187194,
+      "learning_rate": 3.3707440812813584e-06,
+      "loss": 0.515,
+      "step": 9090
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0856676436543693,
+      "learning_rate": 3.3687734175920505e-06,
+      "loss": 0.5452,
+      "step": 9091
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9367234887166933,
+      "learning_rate": 3.366803213440859e-06,
+      "loss": 0.5332,
+      "step": 9092
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9966112933019008,
+      "learning_rate": 3.3648334689643214e-06,
+      "loss": 0.4963,
+      "step": 9093
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9522114454224588,
+      "learning_rate": 3.3628641842989384e-06,
+      "loss": 0.4583,
+      "step": 9094
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8942989570626828,
+      "learning_rate": 3.36089535958118e-06,
+      "loss": 0.5099,
+      "step": 9095
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.924628658722338,
+      "learning_rate": 3.3589269949474856e-06,
+      "loss": 0.4925,
+      "step": 9096
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.890733017360491,
+      "learning_rate": 3.356959090534262e-06,
+      "loss": 0.4837,
+      "step": 9097
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9854852437262702,
+      "learning_rate": 3.3549916464778787e-06,
+      "loss": 0.4994,
+      "step": 9098
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9036658514233938,
+      "learning_rate": 3.35302466291468e-06,
+      "loss": 0.5012,
+      "step": 9099
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9687654855435422,
+      "learning_rate": 3.3510581399809762e-06,
+      "loss": 0.4729,
+      "step": 9100
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0067087474395098,
+      "learning_rate": 3.3490920778130455e-06,
+      "loss": 0.5084,
+      "step": 9101
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9463734582922495,
+      "learning_rate": 3.3471264765471346e-06,
+      "loss": 0.5594,
+      "step": 9102
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0202413779029968,
+      "learning_rate": 3.3451613363194603e-06,
+      "loss": 0.5247,
+      "step": 9103
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8987810339759174,
+      "learning_rate": 3.3431966572662e-06,
+      "loss": 0.5141,
+      "step": 9104
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8590651835433493,
+      "learning_rate": 3.341232439523506e-06,
+      "loss": 0.5129,
+      "step": 9105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9656965528739612,
+      "learning_rate": 3.339268683227499e-06,
+      "loss": 0.5474,
+      "step": 9106
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8168953824209993,
+      "learning_rate": 3.3373053885142636e-06,
+      "loss": 0.4938,
+      "step": 9107
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9425801989390152,
+      "learning_rate": 3.335342555519855e-06,
+      "loss": 0.5095,
+      "step": 9108
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0167412583769833,
+      "learning_rate": 3.3333801843802994e-06,
+      "loss": 0.5314,
+      "step": 9109
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8834281023977991,
+      "learning_rate": 3.331418275231576e-06,
+      "loss": 0.4723,
+      "step": 9110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9367658540260484,
+      "learning_rate": 3.3294568282096586e-06,
+      "loss": 0.4902,
+      "step": 9111
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0125504424161789,
+      "learning_rate": 3.3274958434504625e-06,
+      "loss": 0.5003,
+      "step": 9112
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8158336140834846,
+      "learning_rate": 3.3255353210898866e-06,
+      "loss": 0.4647,
+      "step": 9113
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8216839414525001,
+      "learning_rate": 3.3235752612637917e-06,
+      "loss": 0.496,
+      "step": 9114
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.98806789784039,
+      "learning_rate": 3.3216156641080134e-06,
+      "loss": 0.4611,
+      "step": 9115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8943357113254146,
+      "learning_rate": 3.319656529758339e-06,
+      "loss": 0.5236,
+      "step": 9116
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9066027966449727,
+      "learning_rate": 3.317697858350548e-06,
+      "loss": 0.4432,
+      "step": 9117
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8431741448497101,
+      "learning_rate": 3.3157396500203655e-06,
+      "loss": 0.4948,
+      "step": 9118
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9503603522033455,
+      "learning_rate": 3.3137819049034957e-06,
+      "loss": 0.4786,
+      "step": 9119
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9381882710223372,
+      "learning_rate": 3.31182462313561e-06,
+      "loss": 0.5003,
+      "step": 9120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8472589492270928,
+      "learning_rate": 3.309867804852348e-06,
+      "loss": 0.4728,
+      "step": 9121
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9965341501914862,
+      "learning_rate": 3.3079114501893063e-06,
+      "loss": 0.5548,
+      "step": 9122
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9186851430217279,
+      "learning_rate": 3.3059555592820726e-06,
+      "loss": 0.4724,
+      "step": 9123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9803717640615625,
+      "learning_rate": 3.3040001322661772e-06,
+      "loss": 0.5643,
+      "step": 9124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.011552371194042,
+      "learning_rate": 3.3020451692771337e-06,
+      "loss": 0.537,
+      "step": 9125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9319617645191561,
+      "learning_rate": 3.3000906704504176e-06,
+      "loss": 0.5305,
+      "step": 9126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.98808305950421,
+      "learning_rate": 3.2981366359214806e-06,
+      "loss": 0.5001,
+      "step": 9127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.028054694369749,
+      "learning_rate": 3.296183065825722e-06,
+      "loss": 0.5232,
+      "step": 9128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9378023811252546,
+      "learning_rate": 3.294229960298537e-06,
+      "loss": 0.5227,
+      "step": 9129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8934671402668811,
+      "learning_rate": 3.2922773194752653e-06,
+      "loss": 0.4626,
+      "step": 9130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8791588094117161,
+      "learning_rate": 3.2903251434912265e-06,
+      "loss": 0.4498,
+      "step": 9131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9675333837599392,
+      "learning_rate": 3.288373432481703e-06,
+      "loss": 0.5532,
+      "step": 9132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9734440759086094,
+      "learning_rate": 3.28642218658195e-06,
+      "loss": 0.4815,
+      "step": 9133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9726973591593845,
+      "learning_rate": 3.2844714059271788e-06,
+      "loss": 0.5354,
+      "step": 9134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8416426012871814,
+      "learning_rate": 3.2825210906525885e-06,
+      "loss": 0.4334,
+      "step": 9135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.854242262963165,
+      "learning_rate": 3.2805712408933223e-06,
+      "loss": 0.428,
+      "step": 9136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9023628780098923,
+      "learning_rate": 3.278621856784514e-06,
+      "loss": 0.4703,
+      "step": 9137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8885723574629798,
+      "learning_rate": 3.2766729384612473e-06,
+      "loss": 0.4343,
+      "step": 9138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0417131985652142,
+      "learning_rate": 3.2747244860585823e-06,
+      "loss": 0.5197,
+      "step": 9139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8936134291252534,
+      "learning_rate": 3.272776499711545e-06,
+      "loss": 0.4929,
+      "step": 9140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0327830703736356,
+      "learning_rate": 3.270828979555133e-06,
+      "loss": 0.5942,
+      "step": 9141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.950809911721351,
+      "learning_rate": 3.2688819257242963e-06,
+      "loss": 0.4976,
+      "step": 9142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9363799676811368,
+      "learning_rate": 3.266935338353978e-06,
+      "loss": 0.572,
+      "step": 9143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.893436845428354,
+      "learning_rate": 3.2649892175790667e-06,
+      "loss": 0.4879,
+      "step": 9144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9416979646344061,
+      "learning_rate": 3.2630435635344283e-06,
+      "loss": 0.4463,
+      "step": 9145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9346917071558514,
+      "learning_rate": 3.261098376354894e-06,
+      "loss": 0.5248,
+      "step": 9146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9589850809495569,
+      "learning_rate": 3.259153656175269e-06,
+      "loss": 0.4937,
+      "step": 9147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8429765478620248,
+      "learning_rate": 3.2572094031303103e-06,
+      "loss": 0.4373,
+      "step": 9148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8736249174868056,
+      "learning_rate": 3.255265617354766e-06,
+      "loss": 0.4693,
+      "step": 9149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8879306103161134,
+      "learning_rate": 3.253322298983327e-06,
+      "loss": 0.4574,
+      "step": 9150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9695827910172448,
+      "learning_rate": 3.25137944815067e-06,
+      "loss": 0.5195,
+      "step": 9151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8645467946324216,
+      "learning_rate": 3.2494370649914296e-06,
+      "loss": 0.4675,
+      "step": 9152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0658706154975952,
+      "learning_rate": 3.2474951496402175e-06,
+      "loss": 0.567,
+      "step": 9153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8517465790920219,
+      "learning_rate": 3.245553702231595e-06,
+      "loss": 0.4684,
+      "step": 9154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0026469376848923,
+      "learning_rate": 3.243612722900117e-06,
+      "loss": 0.4907,
+      "step": 9155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9615501023379072,
+      "learning_rate": 3.2416722117802803e-06,
+      "loss": 0.5011,
+      "step": 9156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8908905189634679,
+      "learning_rate": 3.2397321690065643e-06,
+      "loss": 0.4729,
+      "step": 9157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9676487859660207,
+      "learning_rate": 3.2377925947134137e-06,
+      "loss": 0.4864,
+      "step": 9158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9666007142870987,
+      "learning_rate": 3.235853489035241e-06,
+      "loss": 0.5508,
+      "step": 9159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0605811965729126,
+      "learning_rate": 3.2339148521064146e-06,
+      "loss": 0.5257,
+      "step": 9160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0201172769043783,
+      "learning_rate": 3.2319766840612954e-06,
+      "loss": 0.4734,
+      "step": 9161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8410082186099735,
+      "learning_rate": 3.230038985034184e-06,
+      "loss": 0.4574,
+      "step": 9162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0832135483157863,
+      "learning_rate": 3.2281017551593665e-06,
+      "loss": 0.4434,
+      "step": 9163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0581813148791377,
+      "learning_rate": 3.2261649945710916e-06,
+      "loss": 0.5907,
+      "step": 9164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.908821683888088,
+      "learning_rate": 3.2242287034035756e-06,
+      "loss": 0.4716,
+      "step": 9165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8791225901790715,
+      "learning_rate": 3.222292881790996e-06,
+      "loss": 0.4546,
+      "step": 9166
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9342783962101158,
+      "learning_rate": 3.2203575298675126e-06,
+      "loss": 0.5162,
+      "step": 9167
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9890614420190501,
+      "learning_rate": 3.2184226477672366e-06,
+      "loss": 0.5029,
+      "step": 9168
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9082676055397952,
+      "learning_rate": 3.2164882356242555e-06,
+      "loss": 0.4896,
+      "step": 9169
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9268421651069393,
+      "learning_rate": 3.2145542935726224e-06,
+      "loss": 0.4767,
+      "step": 9170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8961329360018759,
+      "learning_rate": 3.212620821746362e-06,
+      "loss": 0.4751,
+      "step": 9171
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8984912218661776,
+      "learning_rate": 3.2106878202794513e-06,
+      "loss": 0.5043,
+      "step": 9172
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9263734336016696,
+      "learning_rate": 3.2087552893058594e-06,
+      "loss": 0.4616,
+      "step": 9173
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.992989819188374,
+      "learning_rate": 3.206823228959498e-06,
+      "loss": 0.5415,
+      "step": 9174
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.96993392568458,
+      "learning_rate": 3.2048916393742622e-06,
+      "loss": 0.5095,
+      "step": 9175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9049184006651029,
+      "learning_rate": 3.2029605206840088e-06,
+      "loss": 0.4609,
+      "step": 9176
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8877311555149989,
+      "learning_rate": 3.201029873022565e-06,
+      "loss": 0.4444,
+      "step": 9177
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8661027546548056,
+      "learning_rate": 3.1990996965237143e-06,
+      "loss": 0.408,
+      "step": 9178
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9130267941869538,
+      "learning_rate": 3.1971699913212272e-06,
+      "loss": 0.4907,
+      "step": 9179
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9096717689460737,
+      "learning_rate": 3.1952407575488243e-06,
+      "loss": 0.5102,
+      "step": 9180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8384692208379614,
+      "learning_rate": 3.1933119953402e-06,
+      "loss": 0.4654,
+      "step": 9181
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.916298509249038,
+      "learning_rate": 3.1913837048290176e-06,
+      "loss": 0.4712,
+      "step": 9182
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0794714537379706,
+      "learning_rate": 3.189455886148908e-06,
+      "loss": 0.509,
+      "step": 9183
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9145557666072167,
+      "learning_rate": 3.1875285394334575e-06,
+      "loss": 0.5139,
+      "step": 9184
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9588017745471442,
+      "learning_rate": 3.1856016648162435e-06,
+      "loss": 0.511,
+      "step": 9185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.93977533850814,
+      "learning_rate": 3.1836752624307878e-06,
+      "loss": 0.4925,
+      "step": 9186
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9905180889851486,
+      "learning_rate": 3.1817493324105884e-06,
+      "loss": 0.5343,
+      "step": 9187
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9626382947667614,
+      "learning_rate": 3.179823874889113e-06,
+      "loss": 0.5192,
+      "step": 9188
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.008409226761721,
+      "learning_rate": 3.1778988899997977e-06,
+      "loss": 0.576,
+      "step": 9189
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9254016781809032,
+      "learning_rate": 3.175974377876031e-06,
+      "loss": 0.5201,
+      "step": 9190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8827104917802043,
+      "learning_rate": 3.1740503386511933e-06,
+      "loss": 0.5004,
+      "step": 9191
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9195045711783788,
+      "learning_rate": 3.17212677245861e-06,
+      "loss": 0.4723,
+      "step": 9192
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9178177294623786,
+      "learning_rate": 3.1702036794315837e-06,
+      "loss": 0.52,
+      "step": 9193
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.930776420999054,
+      "learning_rate": 3.1682810597033853e-06,
+      "loss": 0.5447,
+      "step": 9194
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8603670066733564,
+      "learning_rate": 3.1663589134072537e-06,
+      "loss": 0.539,
+      "step": 9195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9266579862289025,
+      "learning_rate": 3.16443724067638e-06,
+      "loss": 0.4666,
+      "step": 9196
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8913826793152476,
+      "learning_rate": 3.1625160416439503e-06,
+      "loss": 0.4906,
+      "step": 9197
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8448070499801077,
+      "learning_rate": 3.1605953164430904e-06,
+      "loss": 0.4222,
+      "step": 9198
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8777398534519062,
+      "learning_rate": 3.1586750652069077e-06,
+      "loss": 0.5065,
+      "step": 9199
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8780703673114545,
+      "learning_rate": 3.156755288068475e-06,
+      "loss": 0.494,
+      "step": 9200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.002260390936769,
+      "learning_rate": 3.1548359851608344e-06,
+      "loss": 0.5149,
+      "step": 9201
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9969350471910058,
+      "learning_rate": 3.1529171566169825e-06,
+      "loss": 0.5127,
+      "step": 9202
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8782140061192855,
+      "learning_rate": 3.1509988025699046e-06,
+      "loss": 0.4238,
+      "step": 9203
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9204540185363628,
+      "learning_rate": 3.14908092315253e-06,
+      "loss": 0.5102,
+      "step": 9204
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8708557332916683,
+      "learning_rate": 3.147163518497772e-06,
+      "loss": 0.4824,
+      "step": 9205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9061520741038792,
+      "learning_rate": 3.145246588738503e-06,
+      "loss": 0.4841,
+      "step": 9206
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9419277444337851,
+      "learning_rate": 3.1433301340075694e-06,
+      "loss": 0.4981,
+      "step": 9207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0182335668415685,
+      "learning_rate": 3.1414141544377686e-06,
+      "loss": 0.5721,
+      "step": 9208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9456494049006637,
+      "learning_rate": 3.1394986501618897e-06,
+      "loss": 0.4902,
+      "step": 9209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9234880089818398,
+      "learning_rate": 3.1375836213126653e-06,
+      "loss": 0.5534,
+      "step": 9210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8922445520981266,
+      "learning_rate": 3.135669068022811e-06,
+      "loss": 0.4862,
+      "step": 9211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0138485298431035,
+      "learning_rate": 3.1337549904249996e-06,
+      "loss": 0.5239,
+      "step": 9212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8779972974608878,
+      "learning_rate": 3.1318413886518804e-06,
+      "loss": 0.4709,
+      "step": 9213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9511817289467243,
+      "learning_rate": 3.129928262836055e-06,
+      "loss": 0.5514,
+      "step": 9214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9160031996320639,
+      "learning_rate": 3.1280156131101136e-06,
+      "loss": 0.4835,
+      "step": 9215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9136188010687992,
+      "learning_rate": 3.1261034396065924e-06,
+      "loss": 0.4356,
+      "step": 9216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9672584722550528,
+      "learning_rate": 3.1241917424580047e-06,
+      "loss": 0.5193,
+      "step": 9217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8269303323982298,
+      "learning_rate": 3.122280521796831e-06,
+      "loss": 0.4713,
+      "step": 9218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9173435143365514,
+      "learning_rate": 3.1203697777555163e-06,
+      "loss": 0.5053,
+      "step": 9219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8772695738063876,
+      "learning_rate": 3.1184595104664726e-06,
+      "loss": 0.5159,
+      "step": 9220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8953996339227956,
+      "learning_rate": 3.1165497200620863e-06,
+      "loss": 0.4851,
+      "step": 9221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9178968916849854,
+      "learning_rate": 3.114640406674694e-06,
+      "loss": 0.4623,
+      "step": 9222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9705439786022239,
+      "learning_rate": 3.1127315704366144e-06,
+      "loss": 0.4735,
+      "step": 9223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0537737416285637,
+      "learning_rate": 3.1108232114801283e-06,
+      "loss": 0.5305,
+      "step": 9224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.006997561164009,
+      "learning_rate": 3.108915329937483e-06,
+      "loss": 0.4649,
+      "step": 9225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9484263275516899,
+      "learning_rate": 3.1070079259408934e-06,
+      "loss": 0.5857,
+      "step": 9226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8931141053668399,
+      "learning_rate": 3.1051009996225434e-06,
+      "loss": 0.433,
+      "step": 9227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9955212573195641,
+      "learning_rate": 3.1031945511145744e-06,
+      "loss": 0.4869,
+      "step": 9228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9445210297345634,
+      "learning_rate": 3.101288580549107e-06,
+      "loss": 0.5201,
+      "step": 9229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8845297197394515,
+      "learning_rate": 3.09938308805822e-06,
+      "loss": 0.4468,
+      "step": 9230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8677616975176144,
+      "learning_rate": 3.0974780737739653e-06,
+      "loss": 0.4696,
+      "step": 9231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8440699624266844,
+      "learning_rate": 3.095573537828357e-06,
+      "loss": 0.4719,
+      "step": 9232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0118764043381994,
+      "learning_rate": 3.0936694803533817e-06,
+      "loss": 0.5053,
+      "step": 9233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.010118539516588,
+      "learning_rate": 3.091765901480983e-06,
+      "loss": 0.5169,
+      "step": 9234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8893225432422138,
+      "learning_rate": 3.0898628013430787e-06,
+      "loss": 0.4511,
+      "step": 9235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9298359284103193,
+      "learning_rate": 3.087960180071553e-06,
+      "loss": 0.5121,
+      "step": 9236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.899101273379794,
+      "learning_rate": 3.0860580377982563e-06,
+      "loss": 0.5377,
+      "step": 9237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9002401268153558,
+      "learning_rate": 3.084156374655005e-06,
+      "loss": 0.4642,
+      "step": 9238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9113145964585156,
+      "learning_rate": 3.0822551907735833e-06,
+      "loss": 0.5161,
+      "step": 9239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8565877087319691,
+      "learning_rate": 3.080354486285743e-06,
+      "loss": 0.4561,
+      "step": 9240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9296165703485347,
+      "learning_rate": 3.078454261323196e-06,
+      "loss": 0.4638,
+      "step": 9241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9871738649945768,
+      "learning_rate": 3.076554516017629e-06,
+      "loss": 0.5422,
+      "step": 9242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9845398816258983,
+      "learning_rate": 3.074655250500693e-06,
+      "loss": 0.5418,
+      "step": 9243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8907187812517292,
+      "learning_rate": 3.0727564649040066e-06,
+      "loss": 0.4575,
+      "step": 9244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9005922717320375,
+      "learning_rate": 3.0708581593591513e-06,
+      "loss": 0.4844,
+      "step": 9245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8811211339067954,
+      "learning_rate": 3.068960333997684e-06,
+      "loss": 0.4851,
+      "step": 9246
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8297490979293182,
+      "learning_rate": 3.0670629889511128e-06,
+      "loss": 0.4663,
+      "step": 9247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9965530095100212,
+      "learning_rate": 3.0651661243509277e-06,
+      "loss": 0.465,
+      "step": 9248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9304897978087547,
+      "learning_rate": 3.063269740328579e-06,
+      "loss": 0.5054,
+      "step": 9249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.247859611749849,
+      "learning_rate": 3.0613738370154853e-06,
+      "loss": 0.5959,
+      "step": 9250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8559408643557311,
+      "learning_rate": 3.059478414543029e-06,
+      "loss": 0.4542,
+      "step": 9251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9487040433145538,
+      "learning_rate": 3.0575834730425658e-06,
+      "loss": 0.443,
+      "step": 9252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0418904384512588,
+      "learning_rate": 3.0556890126454075e-06,
+      "loss": 0.5321,
+      "step": 9253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9142524061866827,
+      "learning_rate": 3.0537950334828405e-06,
+      "loss": 0.5283,
+      "step": 9254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9849848109141258,
+      "learning_rate": 3.051901535686116e-06,
+      "loss": 0.5214,
+      "step": 9255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9474074774382758,
+      "learning_rate": 3.0500085193864525e-06,
+      "loss": 0.4811,
+      "step": 9256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9814885554674336,
+      "learning_rate": 3.0481159847150343e-06,
+      "loss": 0.5196,
+      "step": 9257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9491077327581923,
+      "learning_rate": 3.046223931803015e-06,
+      "loss": 0.4665,
+      "step": 9258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9047139713494868,
+      "learning_rate": 3.044332360781502e-06,
+      "loss": 0.4961,
+      "step": 9259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9322500438401069,
+      "learning_rate": 3.0424412717815943e-06,
+      "loss": 0.4989,
+      "step": 9260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.048631041096126,
+      "learning_rate": 3.040550664934332e-06,
+      "loss": 0.5397,
+      "step": 9261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9549270032708144,
+      "learning_rate": 3.0386605403707347e-06,
+      "loss": 0.5324,
+      "step": 9262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9334118783548825,
+      "learning_rate": 3.036770898221787e-06,
+      "loss": 0.4805,
+      "step": 9263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8760160616514998,
+      "learning_rate": 3.0348817386184403e-06,
+      "loss": 0.523,
+      "step": 9264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8602677591239977,
+      "learning_rate": 3.0329930616916114e-06,
+      "loss": 0.4556,
+      "step": 9265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.942749923235785,
+      "learning_rate": 3.0311048675721865e-06,
+      "loss": 0.4959,
+      "step": 9266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.967553145530537,
+      "learning_rate": 3.02921715639101e-06,
+      "loss": 0.5025,
+      "step": 9267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.875163075806297,
+      "learning_rate": 3.0273299282789004e-06,
+      "loss": 0.4791,
+      "step": 9268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9610020745516507,
+      "learning_rate": 3.025443183366643e-06,
+      "loss": 0.5296,
+      "step": 9269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0015074057092335,
+      "learning_rate": 3.023556921784987e-06,
+      "loss": 0.5247,
+      "step": 9270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.93960616711754,
+      "learning_rate": 3.021671143664647e-06,
+      "loss": 0.481,
+      "step": 9271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9447586684544488,
+      "learning_rate": 3.019785849136311e-06,
+      "loss": 0.5358,
+      "step": 9272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9784661202493571,
+      "learning_rate": 3.0179010383306208e-06,
+      "loss": 0.4865,
+      "step": 9273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9071899625936168,
+      "learning_rate": 3.0160167113781945e-06,
+      "loss": 0.5029,
+      "step": 9274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.839349146751607,
+      "learning_rate": 3.014132868409617e-06,
+      "loss": 0.5041,
+      "step": 9275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8927128357241134,
+      "learning_rate": 3.012249509555435e-06,
+      "loss": 0.4717,
+      "step": 9276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9196859640341792,
+      "learning_rate": 3.0103666349461624e-06,
+      "loss": 0.5086,
+      "step": 9277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0067864835070655,
+      "learning_rate": 3.008484244712286e-06,
+      "loss": 0.4523,
+      "step": 9278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.933143092486881,
+      "learning_rate": 3.0066023389842446e-06,
+      "loss": 0.482,
+      "step": 9279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9386096596080008,
+      "learning_rate": 3.004720917892464e-06,
+      "loss": 0.4981,
+      "step": 9280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9007385757794655,
+      "learning_rate": 3.0028399815673147e-06,
+      "loss": 0.4798,
+      "step": 9281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.080130380478513,
+      "learning_rate": 3.0009595301391494e-06,
+      "loss": 0.5987,
+      "step": 9282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0126379378704609,
+      "learning_rate": 2.999079563738281e-06,
+      "loss": 0.494,
+      "step": 9283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9617860043838014,
+      "learning_rate": 2.9972000824949908e-06,
+      "loss": 0.5165,
+      "step": 9284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8559597205644912,
+      "learning_rate": 2.9953210865395176e-06,
+      "loss": 0.4558,
+      "step": 9285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8985111470687281,
+      "learning_rate": 2.9934425760020857e-06,
+      "loss": 0.4772,
+      "step": 9286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9475458320896335,
+      "learning_rate": 2.9915645510128666e-06,
+      "loss": 0.454,
+      "step": 9287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.935781669540675,
+      "learning_rate": 2.9896870117020073e-06,
+      "loss": 0.4811,
+      "step": 9288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9664864293034208,
+      "learning_rate": 2.987809958199619e-06,
+      "loss": 0.577,
+      "step": 9289
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8836746825941467,
+      "learning_rate": 2.9859333906357845e-06,
+      "loss": 0.5378,
+      "step": 9290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9582144787011051,
+      "learning_rate": 2.984057309140539e-06,
+      "loss": 0.478,
+      "step": 9291
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8878155885088148,
+      "learning_rate": 2.9821817138439036e-06,
+      "loss": 0.5188,
+      "step": 9292
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8990987481542069,
+      "learning_rate": 2.980306604875849e-06,
+      "loss": 0.4875,
+      "step": 9293
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.927337620844019,
+      "learning_rate": 2.9784319823663188e-06,
+      "loss": 0.5161,
+      "step": 9294
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8955209176918927,
+      "learning_rate": 2.976557846445225e-06,
+      "loss": 0.5026,
+      "step": 9295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9350116459240718,
+      "learning_rate": 2.9746841972424456e-06,
+      "loss": 0.5127,
+      "step": 9296
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9022083479959,
+      "learning_rate": 2.9728110348878135e-06,
+      "loss": 0.4831,
+      "step": 9297
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8905880164735378,
+      "learning_rate": 2.9709383595111506e-06,
+      "loss": 0.4788,
+      "step": 9298
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9840888834902731,
+      "learning_rate": 2.969066171242221e-06,
+      "loss": 0.5257,
+      "step": 9299
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8482427926265703,
+      "learning_rate": 2.967194470210769e-06,
+      "loss": 0.5342,
+      "step": 9300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8943824965526529,
+      "learning_rate": 2.9653232565465017e-06,
+      "loss": 0.4111,
+      "step": 9301
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8255995194547043,
+      "learning_rate": 2.9634525303790973e-06,
+      "loss": 0.4847,
+      "step": 9302
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9574954648262838,
+      "learning_rate": 2.9615822918381844e-06,
+      "loss": 0.5338,
+      "step": 9303
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9660797147778306,
+      "learning_rate": 2.959712541053381e-06,
+      "loss": 0.4644,
+      "step": 9304
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9252816124540626,
+      "learning_rate": 2.9578432781542523e-06,
+      "loss": 0.5135,
+      "step": 9305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9478822301015203,
+      "learning_rate": 2.955974503270337e-06,
+      "loss": 0.4625,
+      "step": 9306
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.962311200737902,
+      "learning_rate": 2.954106216531141e-06,
+      "loss": 0.5013,
+      "step": 9307
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.838888152056901,
+      "learning_rate": 2.952238418066137e-06,
+      "loss": 0.4996,
+      "step": 9308
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.98241779392544,
+      "learning_rate": 2.9503711080047535e-06,
+      "loss": 0.4344,
+      "step": 9309
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0205651361476762,
+      "learning_rate": 2.9485042864764047e-06,
+      "loss": 0.5232,
+      "step": 9310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9279477006987505,
+      "learning_rate": 2.9466379536104518e-06,
+      "loss": 0.4812,
+      "step": 9311
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8593933307119445,
+      "learning_rate": 2.9447721095362325e-06,
+      "loss": 0.4849,
+      "step": 9312
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9690665581476634,
+      "learning_rate": 2.942906754383048e-06,
+      "loss": 0.4679,
+      "step": 9313
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8951759511066802,
+      "learning_rate": 2.9410418882801682e-06,
+      "loss": 0.5178,
+      "step": 9314
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9530573012082669,
+      "learning_rate": 2.939177511356819e-06,
+      "loss": 0.5129,
+      "step": 9315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0308139688157876,
+      "learning_rate": 2.9373136237422107e-06,
+      "loss": 0.5144,
+      "step": 9316
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9297141613759037,
+      "learning_rate": 2.9354502255655002e-06,
+      "loss": 0.5619,
+      "step": 9317
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8485557605262156,
+      "learning_rate": 2.9335873169558236e-06,
+      "loss": 0.4499,
+      "step": 9318
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9753919721602036,
+      "learning_rate": 2.9317248980422785e-06,
+      "loss": 0.5164,
+      "step": 9319
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9329553550167934,
+      "learning_rate": 2.9298629689539315e-06,
+      "loss": 0.4638,
+      "step": 9320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9864133114751812,
+      "learning_rate": 2.9280015298198026e-06,
+      "loss": 0.4754,
+      "step": 9321
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9716573961646281,
+      "learning_rate": 2.9261405807689014e-06,
+      "loss": 0.5036,
+      "step": 9322
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9333844851764159,
+      "learning_rate": 2.9242801219301797e-06,
+      "loss": 0.4928,
+      "step": 9323
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8484772389666334,
+      "learning_rate": 2.9224201534325703e-06,
+      "loss": 0.4605,
+      "step": 9324
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.887435635454266,
+      "learning_rate": 2.9205606754049667e-06,
+      "loss": 0.4935,
+      "step": 9325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9015840885486097,
+      "learning_rate": 2.918701687976231e-06,
+      "loss": 0.4776,
+      "step": 9326
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9225799390724025,
+      "learning_rate": 2.9168431912751805e-06,
+      "loss": 0.489,
+      "step": 9327
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9216293988583227,
+      "learning_rate": 2.914985185430621e-06,
+      "loss": 0.4783,
+      "step": 9328
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9834546685153829,
+      "learning_rate": 2.9131276705713008e-06,
+      "loss": 0.5459,
+      "step": 9329
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9503774917454169,
+      "learning_rate": 2.9112706468259478e-06,
+      "loss": 0.5112,
+      "step": 9330
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8558042581003334,
+      "learning_rate": 2.90941411432325e-06,
+      "loss": 0.4642,
+      "step": 9331
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8959538771990759,
+      "learning_rate": 2.9075580731918684e-06,
+      "loss": 0.4843,
+      "step": 9332
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9834208135406659,
+      "learning_rate": 2.905702523560415e-06,
+      "loss": 0.4608,
+      "step": 9333
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.926465065803759,
+      "learning_rate": 2.90384746555749e-06,
+      "loss": 0.4842,
+      "step": 9334
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8753341930421762,
+      "learning_rate": 2.9019928993116388e-06,
+      "loss": 0.4617,
+      "step": 9335
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.87421173569085,
+      "learning_rate": 2.900138824951383e-06,
+      "loss": 0.5181,
+      "step": 9336
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9219804198644024,
+      "learning_rate": 2.89828524260521e-06,
+      "loss": 0.4836,
+      "step": 9337
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9697470723187763,
+      "learning_rate": 2.8964321524015725e-06,
+      "loss": 0.5239,
+      "step": 9338
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9569066941173024,
+      "learning_rate": 2.8945795544688814e-06,
+      "loss": 0.544,
+      "step": 9339
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.982711465105853,
+      "learning_rate": 2.8927274489355296e-06,
+      "loss": 0.5016,
+      "step": 9340
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9280691158631974,
+      "learning_rate": 2.890875835929858e-06,
+      "loss": 0.5355,
+      "step": 9341
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9580433080173101,
+      "learning_rate": 2.8890247155801864e-06,
+      "loss": 0.5485,
+      "step": 9342
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9684247483206608,
+      "learning_rate": 2.8871740880147935e-06,
+      "loss": 0.5603,
+      "step": 9343
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8835115268579586,
+      "learning_rate": 2.8853239533619314e-06,
+      "loss": 0.5186,
+      "step": 9344
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2424654229776397,
+      "learning_rate": 2.883474311749802e-06,
+      "loss": 0.5008,
+      "step": 9345
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.842275429217379,
+      "learning_rate": 2.8816251633065963e-06,
+      "loss": 0.4049,
+      "step": 9346
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9070179199220347,
+      "learning_rate": 2.87977650816045e-06,
+      "loss": 0.5289,
+      "step": 9347
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.851064073144121,
+      "learning_rate": 2.877928346439476e-06,
+      "loss": 0.5164,
+      "step": 9348
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9271901856488475,
+      "learning_rate": 2.87608067827175e-06,
+      "loss": 0.4729,
+      "step": 9349
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9421025894776374,
+      "learning_rate": 2.8742335037853173e-06,
+      "loss": 0.4991,
+      "step": 9350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9507336913105746,
+      "learning_rate": 2.8723868231081762e-06,
+      "loss": 0.4617,
+      "step": 9351
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8159297374452942,
+      "learning_rate": 2.870540636368312e-06,
+      "loss": 0.4482,
+      "step": 9352
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0030324921252392,
+      "learning_rate": 2.868694943693655e-06,
+      "loss": 0.4944,
+      "step": 9353
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9596171869334853,
+      "learning_rate": 2.8668497452121137e-06,
+      "loss": 0.5073,
+      "step": 9354
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9262832524325116,
+      "learning_rate": 2.8650050410515573e-06,
+      "loss": 0.4648,
+      "step": 9355
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9137971459810631,
+      "learning_rate": 2.8631608313398252e-06,
+      "loss": 0.4643,
+      "step": 9356
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0017381152896594,
+      "learning_rate": 2.8613171162047116e-06,
+      "loss": 0.5519,
+      "step": 9357
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0135694589387556,
+      "learning_rate": 2.8594738957739964e-06,
+      "loss": 0.4904,
+      "step": 9358
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9381392192800052,
+      "learning_rate": 2.8576311701754033e-06,
+      "loss": 0.4843,
+      "step": 9359
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9194632348387407,
+      "learning_rate": 2.8557889395366344e-06,
+      "loss": 0.4958,
+      "step": 9360
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9099550661414307,
+      "learning_rate": 2.8539472039853557e-06,
+      "loss": 0.4452,
+      "step": 9361
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9598811542751433,
+      "learning_rate": 2.8521059636492e-06,
+      "loss": 0.5969,
+      "step": 9362
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9889306022941791,
+      "learning_rate": 2.8502652186557546e-06,
+      "loss": 0.5026,
+      "step": 9363
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9599089363640353,
+      "learning_rate": 2.8484249691325936e-06,
+      "loss": 0.5357,
+      "step": 9364
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8071106061246606,
+      "learning_rate": 2.846585215207236e-06,
+      "loss": 0.4196,
+      "step": 9365
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8789960512738482,
+      "learning_rate": 2.844745957007178e-06,
+      "loss": 0.4865,
+      "step": 9366
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.004915994614177,
+      "learning_rate": 2.8429071946598784e-06,
+      "loss": 0.4525,
+      "step": 9367
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.949052188692021,
+      "learning_rate": 2.841068928292762e-06,
+      "loss": 0.515,
+      "step": 9368
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8423956562674407,
+      "learning_rate": 2.839231158033219e-06,
+      "loss": 0.3948,
+      "step": 9369
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8096642908837642,
+      "learning_rate": 2.837393884008608e-06,
+      "loss": 0.4039,
+      "step": 9370
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8576437588563828,
+      "learning_rate": 2.835557106346244e-06,
+      "loss": 0.4479,
+      "step": 9371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9191350794963664,
+      "learning_rate": 2.8337208251734183e-06,
+      "loss": 0.4786,
+      "step": 9372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9644820042800843,
+      "learning_rate": 2.8318850406173827e-06,
+      "loss": 0.4845,
+      "step": 9373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9416056292197382,
+      "learning_rate": 2.830049752805356e-06,
+      "loss": 0.4415,
+      "step": 9374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9614306850353689,
+      "learning_rate": 2.8282149618645215e-06,
+      "loss": 0.5065,
+      "step": 9375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.919458526211513,
+      "learning_rate": 2.826380667922032e-06,
+      "loss": 0.4998,
+      "step": 9376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9254357372660033,
+      "learning_rate": 2.824546871104996e-06,
+      "loss": 0.4881,
+      "step": 9377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.940488112437101,
+      "learning_rate": 2.8227135715404975e-06,
+      "loss": 0.5278,
+      "step": 9378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9064149146132878,
+      "learning_rate": 2.820880769355582e-06,
+      "loss": 0.4802,
+      "step": 9379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9971665290177097,
+      "learning_rate": 2.819048464677261e-06,
+      "loss": 0.5214,
+      "step": 9380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9594023669258108,
+      "learning_rate": 2.817216657632512e-06,
+      "loss": 0.5152,
+      "step": 9381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9912338435984925,
+      "learning_rate": 2.8153853483482817e-06,
+      "loss": 0.4847,
+      "step": 9382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.918143957172683,
+      "learning_rate": 2.813554536951466e-06,
+      "loss": 0.5439,
+      "step": 9383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9542177968776675,
+      "learning_rate": 2.8117242235689546e-06,
+      "loss": 0.5633,
+      "step": 9384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9676961660664399,
+      "learning_rate": 2.8098944083275735e-06,
+      "loss": 0.4745,
+      "step": 9385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8463354068720405,
+      "learning_rate": 2.8080650913541343e-06,
+      "loss": 0.4333,
+      "step": 9386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9071559562682353,
+      "learning_rate": 2.8062362727754034e-06,
+      "loss": 0.4202,
+      "step": 9387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0620416420487766,
+      "learning_rate": 2.804407952718119e-06,
+      "loss": 0.5192,
+      "step": 9388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9700130342590119,
+      "learning_rate": 2.8025801313089808e-06,
+      "loss": 0.4732,
+      "step": 9389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.905026639613572,
+      "learning_rate": 2.8007528086746574e-06,
+      "loss": 0.4955,
+      "step": 9390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8949781812267134,
+      "learning_rate": 2.798925984941776e-06,
+      "loss": 0.4971,
+      "step": 9391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0893057417483687,
+      "learning_rate": 2.797099660236937e-06,
+      "loss": 0.5203,
+      "step": 9392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9056671218015045,
+      "learning_rate": 2.7952738346867026e-06,
+      "loss": 0.5157,
+      "step": 9393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9267444438266536,
+      "learning_rate": 2.7934485084176012e-06,
+      "loss": 0.4864,
+      "step": 9394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9428738338535183,
+      "learning_rate": 2.791623681556125e-06,
+      "loss": 0.4714,
+      "step": 9395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9593348069639677,
+      "learning_rate": 2.789799354228737e-06,
+      "loss": 0.4895,
+      "step": 9396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0047888951482236,
+      "learning_rate": 2.7879755265618558e-06,
+      "loss": 0.4577,
+      "step": 9397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0235280843992327,
+      "learning_rate": 2.786152198681874e-06,
+      "loss": 0.5219,
+      "step": 9398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9472089839708605,
+      "learning_rate": 2.7843293707151455e-06,
+      "loss": 0.5328,
+      "step": 9399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9448776068653041,
+      "learning_rate": 2.782507042787991e-06,
+      "loss": 0.4486,
+      "step": 9400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0305616862250657,
+      "learning_rate": 2.7806852150266974e-06,
+      "loss": 0.5552,
+      "step": 9401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8278570108296879,
+      "learning_rate": 2.778863887557517e-06,
+      "loss": 0.4195,
+      "step": 9402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9276535441346776,
+      "learning_rate": 2.777043060506661e-06,
+      "loss": 0.4551,
+      "step": 9403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9487354552538336,
+      "learning_rate": 2.7752227340003145e-06,
+      "loss": 0.5439,
+      "step": 9404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9832271050557329,
+      "learning_rate": 2.773402908164625e-06,
+      "loss": 0.5789,
+      "step": 9405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9550858429413105,
+      "learning_rate": 2.771583583125703e-06,
+      "loss": 0.4835,
+      "step": 9406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0062011570686709,
+      "learning_rate": 2.7697647590096277e-06,
+      "loss": 0.5558,
+      "step": 9407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.868966115835517,
+      "learning_rate": 2.76794643594244e-06,
+      "loss": 0.4517,
+      "step": 9408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9535790756363516,
+      "learning_rate": 2.766128614050154e-06,
+      "loss": 0.4804,
+      "step": 9409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9308323891682322,
+      "learning_rate": 2.7643112934587346e-06,
+      "loss": 0.4971,
+      "step": 9410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0176657701318508,
+      "learning_rate": 2.7624944742941253e-06,
+      "loss": 0.5166,
+      "step": 9411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9501763096451348,
+      "learning_rate": 2.760678156682229e-06,
+      "loss": 0.4569,
+      "step": 9412
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9236459156836029,
+      "learning_rate": 2.7588623407489158e-06,
+      "loss": 0.5252,
+      "step": 9413
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8433794190856604,
+      "learning_rate": 2.7570470266200177e-06,
+      "loss": 0.4018,
+      "step": 9414
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.915106529891791,
+      "learning_rate": 2.7552322144213405e-06,
+      "loss": 0.4964,
+      "step": 9415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8850114828887584,
+      "learning_rate": 2.753417904278641e-06,
+      "loss": 0.5148,
+      "step": 9416
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9452199835968088,
+      "learning_rate": 2.751604096317655e-06,
+      "loss": 0.4584,
+      "step": 9417
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.94196405618762,
+      "learning_rate": 2.749790790664074e-06,
+      "loss": 0.5041,
+      "step": 9418
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9554860308445395,
+      "learning_rate": 2.7479779874435607e-06,
+      "loss": 0.4885,
+      "step": 9419
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8251822970851799,
+      "learning_rate": 2.7461656867817397e-06,
+      "loss": 0.4526,
+      "step": 9420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9104051781210106,
+      "learning_rate": 2.7443538888042065e-06,
+      "loss": 0.4844,
+      "step": 9421
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9173725950469258,
+      "learning_rate": 2.742542593636509e-06,
+      "loss": 0.4853,
+      "step": 9422
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9440247777740798,
+      "learning_rate": 2.7407318014041727e-06,
+      "loss": 0.4653,
+      "step": 9423
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9456146477999366,
+      "learning_rate": 2.738921512232684e-06,
+      "loss": 0.4939,
+      "step": 9424
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8871216035281972,
+      "learning_rate": 2.7371117262474945e-06,
+      "loss": 0.4852,
+      "step": 9425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8784139534365634,
+      "learning_rate": 2.7353024435740194e-06,
+      "loss": 0.5012,
+      "step": 9426
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9562799444064749,
+      "learning_rate": 2.7334936643376443e-06,
+      "loss": 0.5088,
+      "step": 9427
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8914075216309687,
+      "learning_rate": 2.7316853886637075e-06,
+      "loss": 0.4959,
+      "step": 9428
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8622154551462666,
+      "learning_rate": 2.729877616677531e-06,
+      "loss": 0.5002,
+      "step": 9429
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8689274219881318,
+      "learning_rate": 2.7280703485043846e-06,
+      "loss": 0.445,
+      "step": 9430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8775797672203981,
+      "learning_rate": 2.726263584269513e-06,
+      "loss": 0.477,
+      "step": 9431
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9325799673463466,
+      "learning_rate": 2.724457324098123e-06,
+      "loss": 0.5041,
+      "step": 9432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9626934216789831,
+      "learning_rate": 2.7226515681153907e-06,
+      "loss": 0.5497,
+      "step": 9433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9011603340435773,
+      "learning_rate": 2.720846316446443e-06,
+      "loss": 0.48,
+      "step": 9434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0063613488595504,
+      "learning_rate": 2.7190415692163954e-06,
+      "loss": 0.4801,
+      "step": 9435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9004373985017571,
+      "learning_rate": 2.717237326550306e-06,
+      "loss": 0.4982,
+      "step": 9436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9216341366680187,
+      "learning_rate": 2.71543358857321e-06,
+      "loss": 0.5106,
+      "step": 9437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0024610759720805,
+      "learning_rate": 2.713630355410104e-06,
+      "loss": 0.4884,
+      "step": 9438
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9858367768202332,
+      "learning_rate": 2.7118276271859555e-06,
+      "loss": 0.5171,
+      "step": 9439
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9180550622016094,
+      "learning_rate": 2.7100254040256813e-06,
+      "loss": 0.5102,
+      "step": 9440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.839648095978644,
+      "learning_rate": 2.7082236860541867e-06,
+      "loss": 0.4548,
+      "step": 9441
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.965488004574052,
+      "learning_rate": 2.7064224733963197e-06,
+      "loss": 0.5141,
+      "step": 9442
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9617855033127427,
+      "learning_rate": 2.704621766176905e-06,
+      "loss": 0.5163,
+      "step": 9443
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9178157408943385,
+      "learning_rate": 2.702821564520732e-06,
+      "loss": 0.4896,
+      "step": 9444
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9110526638080855,
+      "learning_rate": 2.7010218685525545e-06,
+      "loss": 0.462,
+      "step": 9445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9618892367926404,
+      "learning_rate": 2.699222678397082e-06,
+      "loss": 0.4735,
+      "step": 9446
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0025662808262406,
+      "learning_rate": 2.697423994179007e-06,
+      "loss": 0.4799,
+      "step": 9447
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9728581194952428,
+      "learning_rate": 2.69562581602297e-06,
+      "loss": 0.4956,
+      "step": 9448
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8845833875493481,
+      "learning_rate": 2.693828144053584e-06,
+      "loss": 0.4821,
+      "step": 9449
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9953067711145615,
+      "learning_rate": 2.6920309783954277e-06,
+      "loss": 0.5049,
+      "step": 9450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9240944266957233,
+      "learning_rate": 2.690234319173045e-06,
+      "loss": 0.4273,
+      "step": 9451
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8056754390924662,
+      "learning_rate": 2.688438166510935e-06,
+      "loss": 0.4646,
+      "step": 9452
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.953057999448235,
+      "learning_rate": 2.68664252053358e-06,
+      "loss": 0.5191,
+      "step": 9453
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8365867448944447,
+      "learning_rate": 2.6848473813654087e-06,
+      "loss": 0.4236,
+      "step": 9454
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.889545304390988,
+      "learning_rate": 2.6830527491308257e-06,
+      "loss": 0.459,
+      "step": 9455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9223982552953418,
+      "learning_rate": 2.681258623954196e-06,
+      "loss": 0.4824,
+      "step": 9456
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9608557480794483,
+      "learning_rate": 2.679465005959856e-06,
+      "loss": 0.5088,
+      "step": 9457
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0188180580969628,
+      "learning_rate": 2.6776718952720903e-06,
+      "loss": 0.458,
+      "step": 9458
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8616425434444939,
+      "learning_rate": 2.6758792920151745e-06,
+      "loss": 0.4269,
+      "step": 9459
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0145444858253545,
+      "learning_rate": 2.6740871963133243e-06,
+      "loss": 0.4662,
+      "step": 9460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9901094833309612,
+      "learning_rate": 2.6722956082907334e-06,
+      "loss": 0.4941,
+      "step": 9461
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.90990267972996,
+      "learning_rate": 2.670504528071557e-06,
+      "loss": 0.4996,
+      "step": 9462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8671739375567898,
+      "learning_rate": 2.668713955779918e-06,
+      "loss": 0.4801,
+      "step": 9463
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8410079626584746,
+      "learning_rate": 2.6669238915398943e-06,
+      "loss": 0.4196,
+      "step": 9464
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9790064700550006,
+      "learning_rate": 2.6651343354755453e-06,
+      "loss": 0.5025,
+      "step": 9465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.989725553659738,
+      "learning_rate": 2.663345287710878e-06,
+      "loss": 0.4847,
+      "step": 9466
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0120797707377107,
+      "learning_rate": 2.6615567483698746e-06,
+      "loss": 0.5814,
+      "step": 9467
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0247258994347321,
+      "learning_rate": 2.65976871757648e-06,
+      "loss": 0.5506,
+      "step": 9468
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9353296234328144,
+      "learning_rate": 2.6579811954546054e-06,
+      "loss": 0.4999,
+      "step": 9469
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.873564732962634,
+      "learning_rate": 2.6561941821281145e-06,
+      "loss": 0.4801,
+      "step": 9470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9463813696150183,
+      "learning_rate": 2.6544076777208603e-06,
+      "loss": 0.4883,
+      "step": 9471
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9701094798832111,
+      "learning_rate": 2.6526216823566342e-06,
+      "loss": 0.4821,
+      "step": 9472
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9290121340326827,
+      "learning_rate": 2.65083619615921e-06,
+      "loss": 0.5389,
+      "step": 9473
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9520983634733708,
+      "learning_rate": 2.6490512192523175e-06,
+      "loss": 0.4464,
+      "step": 9474
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8680418166970947,
+      "learning_rate": 2.6472667517596584e-06,
+      "loss": 0.461,
+      "step": 9475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8481399018890614,
+      "learning_rate": 2.6454827938048855e-06,
+      "loss": 0.4635,
+      "step": 9476
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8970798829121703,
+      "learning_rate": 2.643699345511638e-06,
+      "loss": 0.4517,
+      "step": 9477
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0208441448818784,
+      "learning_rate": 2.6419164070034974e-06,
+      "loss": 0.517,
+      "step": 9478
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8962370261002454,
+      "learning_rate": 2.6401339784040226e-06,
+      "loss": 0.5019,
+      "step": 9479
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8456090475029107,
+      "learning_rate": 2.6383520598367363e-06,
+      "loss": 0.4345,
+      "step": 9480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8804027202694467,
+      "learning_rate": 2.6365706514251244e-06,
+      "loss": 0.4929,
+      "step": 9481
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9137295020872604,
+      "learning_rate": 2.6347897532926293e-06,
+      "loss": 0.5308,
+      "step": 9482
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9212246006767905,
+      "learning_rate": 2.6330093655626777e-06,
+      "loss": 0.5352,
+      "step": 9483
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.042293489524842,
+      "learning_rate": 2.6312294883586385e-06,
+      "loss": 0.5123,
+      "step": 9484
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8841512558148169,
+      "learning_rate": 2.6294501218038603e-06,
+      "loss": 0.446,
+      "step": 9485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8420400723533827,
+      "learning_rate": 2.627671266021652e-06,
+      "loss": 0.4708,
+      "step": 9486
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9083137612965833,
+      "learning_rate": 2.625892921135288e-06,
+      "loss": 0.4783,
+      "step": 9487
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9938593713070609,
+      "learning_rate": 2.6241150872679968e-06,
+      "loss": 0.5045,
+      "step": 9488
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9119577080481137,
+      "learning_rate": 2.6223377645429948e-06,
+      "loss": 0.4987,
+      "step": 9489
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0419405547070333,
+      "learning_rate": 2.6205609530834388e-06,
+      "loss": 0.4877,
+      "step": 9490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9126973386049844,
+      "learning_rate": 2.6187846530124615e-06,
+      "loss": 0.5284,
+      "step": 9491
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9253444650389665,
+      "learning_rate": 2.6170088644531623e-06,
+      "loss": 0.5231,
+      "step": 9492
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8321810930190079,
+      "learning_rate": 2.6152335875286027e-06,
+      "loss": 0.4274,
+      "step": 9493
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9589537397942688,
+      "learning_rate": 2.6134588223617995e-06,
+      "loss": 0.5199,
+      "step": 9494
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9208045079162227,
+      "learning_rate": 2.6116845690757533e-06,
+      "loss": 0.4873,
+      "step": 9495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9568307857077881,
+      "learning_rate": 2.6099108277934105e-06,
+      "loss": 0.5072,
+      "step": 9496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0430924603891203,
+      "learning_rate": 2.6081375986376924e-06,
+      "loss": 0.5863,
+      "step": 9497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8728274230320672,
+      "learning_rate": 2.6063648817314825e-06,
+      "loss": 0.4295,
+      "step": 9498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9260757150277996,
+      "learning_rate": 2.6045926771976306e-06,
+      "loss": 0.4946,
+      "step": 9499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9797267215825568,
+      "learning_rate": 2.6028209851589403e-06,
+      "loss": 0.5314,
+      "step": 9500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8235502974705514,
+      "learning_rate": 2.6010498057382005e-06,
+      "loss": 0.4063,
+      "step": 9501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9412399480520027,
+      "learning_rate": 2.599279139058143e-06,
+      "loss": 0.5289,
+      "step": 9502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9456699294244987,
+      "learning_rate": 2.597508985241477e-06,
+      "loss": 0.4849,
+      "step": 9503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8843131800728579,
+      "learning_rate": 2.5957393444108724e-06,
+      "loss": 0.462,
+      "step": 9504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9220872425336669,
+      "learning_rate": 2.593970216688967e-06,
+      "loss": 0.5103,
+      "step": 9505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9703291160659552,
+      "learning_rate": 2.592201602198351e-06,
+      "loss": 0.4847,
+      "step": 9506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0437899523711882,
+      "learning_rate": 2.5904335010615976e-06,
+      "loss": 0.4661,
+      "step": 9507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8931272353575075,
+      "learning_rate": 2.588665913401226e-06,
+      "loss": 0.4767,
+      "step": 9508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8971499854589432,
+      "learning_rate": 2.5868988393397376e-06,
+      "loss": 0.4886,
+      "step": 9509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9114204447284524,
+      "learning_rate": 2.5851322789995815e-06,
+      "loss": 0.4812,
+      "step": 9510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9433208349417566,
+      "learning_rate": 2.5833662325031816e-06,
+      "loss": 0.4965,
+      "step": 9511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9745841552592122,
+      "learning_rate": 2.5816006999729225e-06,
+      "loss": 0.5398,
+      "step": 9512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1338912061174495,
+      "learning_rate": 2.5798356815311587e-06,
+      "loss": 0.4673,
+      "step": 9513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9102039459425405,
+      "learning_rate": 2.5780711773001943e-06,
+      "loss": 0.4687,
+      "step": 9514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8972357697454645,
+      "learning_rate": 2.5763071874023205e-06,
+      "loss": 0.442,
+      "step": 9515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9881441739800304,
+      "learning_rate": 2.5745437119597704e-06,
+      "loss": 0.4863,
+      "step": 9516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9675120535718381,
+      "learning_rate": 2.5727807510947545e-06,
+      "loss": 0.5134,
+      "step": 9517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9015375721815215,
+      "learning_rate": 2.5710183049294445e-06,
+      "loss": 0.4743,
+      "step": 9518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9734237840850912,
+      "learning_rate": 2.56925637358598e-06,
+      "loss": 0.5198,
+      "step": 9519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9644298697485151,
+      "learning_rate": 2.5674949571864517e-06,
+      "loss": 0.4779,
+      "step": 9520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9969693321534621,
+      "learning_rate": 2.5657340558529353e-06,
+      "loss": 0.4861,
+      "step": 9521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9181947952380234,
+      "learning_rate": 2.5639736697074525e-06,
+      "loss": 0.4836,
+      "step": 9522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9606810410171697,
+      "learning_rate": 2.5622137988719985e-06,
+      "loss": 0.5083,
+      "step": 9523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.940512663001103,
+      "learning_rate": 2.5604544434685307e-06,
+      "loss": 0.4637,
+      "step": 9524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9785135388690958,
+      "learning_rate": 2.558695603618975e-06,
+      "loss": 0.4966,
+      "step": 9525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9331808129454607,
+      "learning_rate": 2.5569372794452063e-06,
+      "loss": 0.5326,
+      "step": 9526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9294999568518234,
+      "learning_rate": 2.555179471069089e-06,
+      "loss": 0.48,
+      "step": 9527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.824147125924437,
+      "learning_rate": 2.553422178612427e-06,
+      "loss": 0.4187,
+      "step": 9528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8936141587086273,
+      "learning_rate": 2.5516654021970035e-06,
+      "loss": 0.5049,
+      "step": 9529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9344279939268219,
+      "learning_rate": 2.549909141944561e-06,
+      "loss": 0.479,
+      "step": 9530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9227147305713791,
+      "learning_rate": 2.5481533979768092e-06,
+      "loss": 0.4666,
+      "step": 9531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9643823288376145,
+      "learning_rate": 2.546398170415412e-06,
+      "loss": 0.4633,
+      "step": 9532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9135184290568981,
+      "learning_rate": 2.5446434593820156e-06,
+      "loss": 0.4838,
+      "step": 9533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9140822658348559,
+      "learning_rate": 2.5428892649982117e-06,
+      "loss": 0.4997,
+      "step": 9534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9113683203067207,
+      "learning_rate": 2.5411355873855683e-06,
+      "loss": 0.4603,
+      "step": 9535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8322288199107464,
+      "learning_rate": 2.539382426665611e-06,
+      "loss": 0.4867,
+      "step": 9536
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9147448880750836,
+      "learning_rate": 2.537629782959835e-06,
+      "loss": 0.4898,
+      "step": 9537
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.868170086743266,
+      "learning_rate": 2.5358776563896957e-06,
+      "loss": 0.5012,
+      "step": 9538
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9781015542954844,
+      "learning_rate": 2.5341260470766173e-06,
+      "loss": 0.4501,
+      "step": 9539
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.933814696684243,
+      "learning_rate": 2.5323749551419775e-06,
+      "loss": 0.4684,
+      "step": 9540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8989376433156853,
+      "learning_rate": 2.5306243807071305e-06,
+      "loss": 0.5363,
+      "step": 9541
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8773684970723838,
+      "learning_rate": 2.5288743238933887e-06,
+      "loss": 0.4537,
+      "step": 9542
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9498142865358258,
+      "learning_rate": 2.5271247848220294e-06,
+      "loss": 0.5679,
+      "step": 9543
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0550451747440102,
+      "learning_rate": 2.525375763614294e-06,
+      "loss": 0.5219,
+      "step": 9544
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9255261798407214,
+      "learning_rate": 2.5236272603913915e-06,
+      "loss": 0.5036,
+      "step": 9545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9255632591648001,
+      "learning_rate": 2.5218792752744847e-06,
+      "loss": 0.5053,
+      "step": 9546
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9945535840087542,
+      "learning_rate": 2.5201318083847105e-06,
+      "loss": 0.5119,
+      "step": 9547
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9160402986044256,
+      "learning_rate": 2.518384859843168e-06,
+      "loss": 0.4775,
+      "step": 9548
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9130901844914137,
+      "learning_rate": 2.516638429770919e-06,
+      "loss": 0.4579,
+      "step": 9549
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8844580873802608,
+      "learning_rate": 2.514892518288988e-06,
+      "loss": 0.4504,
+      "step": 9550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9584403151191385,
+      "learning_rate": 2.5131471255183705e-06,
+      "loss": 0.525,
+      "step": 9551
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9838552272295368,
+      "learning_rate": 2.511402251580013e-06,
+      "loss": 0.5415,
+      "step": 9552
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9275636019742892,
+      "learning_rate": 2.509657896594837e-06,
+      "loss": 0.4796,
+      "step": 9553
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9780448168288274,
+      "learning_rate": 2.507914060683725e-06,
+      "loss": 0.4797,
+      "step": 9554
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0122611059232476,
+      "learning_rate": 2.5061707439675222e-06,
+      "loss": 0.4896,
+      "step": 9555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.98051994871729,
+      "learning_rate": 2.5044279465670408e-06,
+      "loss": 0.5126,
+      "step": 9556
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9005729862836364,
+      "learning_rate": 2.502685668603053e-06,
+      "loss": 0.5542,
+      "step": 9557
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9593751701329121,
+      "learning_rate": 2.5009439101963027e-06,
+      "loss": 0.5163,
+      "step": 9558
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9336429381380371,
+      "learning_rate": 2.499202671467483e-06,
+      "loss": 0.553,
+      "step": 9559
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0034038590580017,
+      "learning_rate": 2.497461952537267e-06,
+      "loss": 0.5383,
+      "step": 9560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8763757759226454,
+      "learning_rate": 2.4957217535262824e-06,
+      "loss": 0.4981,
+      "step": 9561
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8586560124179438,
+      "learning_rate": 2.4939820745551235e-06,
+      "loss": 0.4485,
+      "step": 9562
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9065661965736719,
+      "learning_rate": 2.4922429157443484e-06,
+      "loss": 0.4697,
+      "step": 9563
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8918877224490259,
+      "learning_rate": 2.490504277214484e-06,
+      "loss": 0.4276,
+      "step": 9564
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8620501304929654,
+      "learning_rate": 2.488766159086009e-06,
+      "loss": 0.4448,
+      "step": 9565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9677343569180769,
+      "learning_rate": 2.4870285614793764e-06,
+      "loss": 0.5492,
+      "step": 9566
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.962363169336885,
+      "learning_rate": 2.485291484515e-06,
+      "loss": 0.5051,
+      "step": 9567
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.959581770332624,
+      "learning_rate": 2.4835549283132597e-06,
+      "loss": 0.5402,
+      "step": 9568
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8514647633831238,
+      "learning_rate": 2.4818188929944942e-06,
+      "loss": 0.4795,
+      "step": 9569
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9092107495200746,
+      "learning_rate": 2.4800833786790145e-06,
+      "loss": 0.4939,
+      "step": 9570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9327251184011716,
+      "learning_rate": 2.478348385487084e-06,
+      "loss": 0.4533,
+      "step": 9571
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8735652553810023,
+      "learning_rate": 2.476613913538938e-06,
+      "loss": 0.4457,
+      "step": 9572
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0708948579788935,
+      "learning_rate": 2.474879962954775e-06,
+      "loss": 0.5,
+      "step": 9573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9015351195751412,
+      "learning_rate": 2.4731465338547556e-06,
+      "loss": 0.5406,
+      "step": 9574
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9339400308308337,
+      "learning_rate": 2.471413626359005e-06,
+      "loss": 0.4519,
+      "step": 9575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9899716519567556,
+      "learning_rate": 2.4696812405876147e-06,
+      "loss": 0.4821,
+      "step": 9576
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9336255754909419,
+      "learning_rate": 2.46794937666063e-06,
+      "loss": 0.5054,
+      "step": 9577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9350326299793474,
+      "learning_rate": 2.466218034698078e-06,
+      "loss": 0.4756,
+      "step": 9578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.896737867789691,
+      "learning_rate": 2.4644872148199316e-06,
+      "loss": 0.4517,
+      "step": 9579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9455232413075647,
+      "learning_rate": 2.4627569171461363e-06,
+      "loss": 0.5259,
+      "step": 9580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8955076931033673,
+      "learning_rate": 2.4610271417966014e-06,
+      "loss": 0.4832,
+      "step": 9581
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.866670299293872,
+      "learning_rate": 2.4592978888912013e-06,
+      "loss": 0.477,
+      "step": 9582
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8821913909176877,
+      "learning_rate": 2.457569158549763e-06,
+      "loss": 0.4958,
+      "step": 9583
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9863938976553523,
+      "learning_rate": 2.455840950892099e-06,
+      "loss": 0.5138,
+      "step": 9584
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9181851836678736,
+      "learning_rate": 2.4541132660379606e-06,
+      "loss": 0.5052,
+      "step": 9585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.87896082720906,
+      "learning_rate": 2.4523861041070806e-06,
+      "loss": 0.4395,
+      "step": 9586
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9452044022637043,
+      "learning_rate": 2.4506594652191485e-06,
+      "loss": 0.4585,
+      "step": 9587
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9581124622435845,
+      "learning_rate": 2.448933349493823e-06,
+      "loss": 0.4749,
+      "step": 9588
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9126209879683963,
+      "learning_rate": 2.4472077570507124e-06,
+      "loss": 0.5034,
+      "step": 9589
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8679783835342433,
+      "learning_rate": 2.4454826880094106e-06,
+      "loss": 0.5103,
+      "step": 9590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8997725986140164,
+      "learning_rate": 2.443758142489454e-06,
+      "loss": 0.4383,
+      "step": 9591
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9282741810737473,
+      "learning_rate": 2.442034120610357e-06,
+      "loss": 0.4888,
+      "step": 9592
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9426438609355607,
+      "learning_rate": 2.4403106224915908e-06,
+      "loss": 0.4972,
+      "step": 9593
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9193151062029296,
+      "learning_rate": 2.438587648252596e-06,
+      "loss": 0.4282,
+      "step": 9594
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8577142049418723,
+      "learning_rate": 2.4368651980127644e-06,
+      "loss": 0.4865,
+      "step": 9595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9277333955003407,
+      "learning_rate": 2.4351432718914727e-06,
+      "loss": 0.4682,
+      "step": 9596
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8571646694128863,
+      "learning_rate": 2.433421870008038e-06,
+      "loss": 0.4404,
+      "step": 9597
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9350419161187955,
+      "learning_rate": 2.431700992481757e-06,
+      "loss": 0.468,
+      "step": 9598
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7483768064636855,
+      "learning_rate": 2.4299806394318837e-06,
+      "loss": 0.3972,
+      "step": 9599
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0325912865114175,
+      "learning_rate": 2.428260810977641e-06,
+      "loss": 0.5122,
+      "step": 9600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9058356524191399,
+      "learning_rate": 2.4265415072382016e-06,
+      "loss": 0.5053,
+      "step": 9601
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.942026873185667,
+      "learning_rate": 2.424822728332724e-06,
+      "loss": 0.5308,
+      "step": 9602
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9747815459252755,
+      "learning_rate": 2.423104474380309e-06,
+      "loss": 0.4839,
+      "step": 9603
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8950242754724765,
+      "learning_rate": 2.421386745500034e-06,
+      "loss": 0.476,
+      "step": 9604
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8927261857002168,
+      "learning_rate": 2.419669541810934e-06,
+      "loss": 0.4962,
+      "step": 9605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.031794549609845,
+      "learning_rate": 2.417952863432015e-06,
+      "loss": 0.4638,
+      "step": 9606
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.941660614300367,
+      "learning_rate": 2.4162367104822313e-06,
+      "loss": 0.4732,
+      "step": 9607
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9447255800757794,
+      "learning_rate": 2.414521083080523e-06,
+      "loss": 0.4594,
+      "step": 9608
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0667354565333285,
+      "learning_rate": 2.4128059813457716e-06,
+      "loss": 0.5123,
+      "step": 9609
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8808929346517816,
+      "learning_rate": 2.411091405396836e-06,
+      "loss": 0.4741,
+      "step": 9610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9070711715732838,
+      "learning_rate": 2.409377355352536e-06,
+      "loss": 0.4799,
+      "step": 9611
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9931652854573912,
+      "learning_rate": 2.4076638313316537e-06,
+      "loss": 0.5847,
+      "step": 9612
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9773303592488506,
+      "learning_rate": 2.405950833452928e-06,
+      "loss": 0.438,
+      "step": 9613
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8959575285566649,
+      "learning_rate": 2.4042383618350795e-06,
+      "loss": 0.4453,
+      "step": 9614
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.907181588357818,
+      "learning_rate": 2.402526416596772e-06,
+      "loss": 0.4528,
+      "step": 9615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9749348938822829,
+      "learning_rate": 2.400814997856645e-06,
+      "loss": 0.515,
+      "step": 9616
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.901830383250741,
+      "learning_rate": 2.399104105733299e-06,
+      "loss": 0.4434,
+      "step": 9617
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8893553496072407,
+      "learning_rate": 2.3973937403452983e-06,
+      "loss": 0.4025,
+      "step": 9618
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9038113038070881,
+      "learning_rate": 2.3956839018111634e-06,
+      "loss": 0.4542,
+      "step": 9619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.992165409961388,
+      "learning_rate": 2.393974590249394e-06,
+      "loss": 0.5901,
+      "step": 9620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8849862750002848,
+      "learning_rate": 2.3922658057784355e-06,
+      "loss": 0.5163,
+      "step": 9621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0192890227700766,
+      "learning_rate": 2.3905575485167098e-06,
+      "loss": 0.4513,
+      "step": 9622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8242519230545053,
+      "learning_rate": 2.388849818582596e-06,
+      "loss": 0.4359,
+      "step": 9623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.91491375213291,
+      "learning_rate": 2.387142616094441e-06,
+      "loss": 0.4939,
+      "step": 9624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9370320013929132,
+      "learning_rate": 2.385435941170544e-06,
+      "loss": 0.534,
+      "step": 9625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0038811640559153,
+      "learning_rate": 2.3837297939291893e-06,
+      "loss": 0.5309,
+      "step": 9626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8885292541660332,
+      "learning_rate": 2.382024174488601e-06,
+      "loss": 0.4796,
+      "step": 9627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9018924053530425,
+      "learning_rate": 2.38031908296698e-06,
+      "loss": 0.4756,
+      "step": 9628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9536886642846818,
+      "learning_rate": 2.378614519482487e-06,
+      "loss": 0.476,
+      "step": 9629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8693603792931843,
+      "learning_rate": 2.376910484153252e-06,
+      "loss": 0.4338,
+      "step": 9630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9319492597355499,
+      "learning_rate": 2.375206977097353e-06,
+      "loss": 0.4931,
+      "step": 9631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8780828804269664,
+      "learning_rate": 2.373503998432852e-06,
+      "loss": 0.4885,
+      "step": 9632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.869729081783994,
+      "learning_rate": 2.3718015482777535e-06,
+      "loss": 0.4,
+      "step": 9633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9881607907408684,
+      "learning_rate": 2.3700996267500486e-06,
+      "loss": 0.4615,
+      "step": 9634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8627314779565832,
+      "learning_rate": 2.368398233967668e-06,
+      "loss": 0.4861,
+      "step": 9635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9965660580842806,
+      "learning_rate": 2.3666973700485207e-06,
+      "loss": 0.4252,
+      "step": 9636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8821204773558616,
+      "learning_rate": 2.3649970351104744e-06,
+      "loss": 0.4828,
+      "step": 9637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8515620291830194,
+      "learning_rate": 2.363297229271365e-06,
+      "loss": 0.4645,
+      "step": 9638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9134141990975033,
+      "learning_rate": 2.3615979526489773e-06,
+      "loss": 0.4702,
+      "step": 9639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.045250600324794,
+      "learning_rate": 2.3598992053610826e-06,
+      "loss": 0.514,
+      "step": 9640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9719330232433728,
+      "learning_rate": 2.358200987525393e-06,
+      "loss": 0.4843,
+      "step": 9641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9890798245689238,
+      "learning_rate": 2.356503299259597e-06,
+      "loss": 0.5332,
+      "step": 9642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.951865022725942,
+      "learning_rate": 2.354806140681343e-06,
+      "loss": 0.477,
+      "step": 9643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9038312048330461,
+      "learning_rate": 2.353109511908245e-06,
+      "loss": 0.4452,
+      "step": 9644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9811003847716361,
+      "learning_rate": 2.351413413057868e-06,
+      "loss": 0.5233,
+      "step": 9645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9479546781437039,
+      "learning_rate": 2.349717844247764e-06,
+      "loss": 0.4685,
+      "step": 9646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8049590726514992,
+      "learning_rate": 2.3480228055954246e-06,
+      "loss": 0.4421,
+      "step": 9647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9733781462938889,
+      "learning_rate": 2.3463282972183176e-06,
+      "loss": 0.5053,
+      "step": 9648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9600108414024033,
+      "learning_rate": 2.3446343192338705e-06,
+      "loss": 0.5029,
+      "step": 9649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.908930633454571,
+      "learning_rate": 2.342940871759477e-06,
+      "loss": 0.51,
+      "step": 9650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8563848690253233,
+      "learning_rate": 2.3412479549124843e-06,
+      "loss": 0.4161,
+      "step": 9651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9269137705402973,
+      "learning_rate": 2.339555568810221e-06,
+      "loss": 0.4778,
+      "step": 9652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9345994871074489,
+      "learning_rate": 2.337863713569959e-06,
+      "loss": 0.4064,
+      "step": 9653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8561141204612545,
+      "learning_rate": 2.336172389308945e-06,
+      "loss": 0.4461,
+      "step": 9654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9263354207130737,
+      "learning_rate": 2.334481596144387e-06,
+      "loss": 0.4931,
+      "step": 9655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9646139588123631,
+      "learning_rate": 2.3327913341934573e-06,
+      "loss": 0.4798,
+      "step": 9656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9821076222973479,
+      "learning_rate": 2.3311016035732825e-06,
+      "loss": 0.5305,
+      "step": 9657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8870591207772264,
+      "learning_rate": 2.329412404400969e-06,
+      "loss": 0.4557,
+      "step": 9658
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8341473641915862,
+      "learning_rate": 2.3277237367935702e-06,
+      "loss": 0.3986,
+      "step": 9659
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0667231560326589,
+      "learning_rate": 2.3260356008681107e-06,
+      "loss": 0.5143,
+      "step": 9660
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9747798801186214,
+      "learning_rate": 2.3243479967415773e-06,
+      "loss": 0.4896,
+      "step": 9661
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.934749581531243,
+      "learning_rate": 2.322660924530922e-06,
+      "loss": 0.4454,
+      "step": 9662
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9290559957629269,
+      "learning_rate": 2.32097438435305e-06,
+      "loss": 0.4966,
+      "step": 9663
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0453163615005325,
+      "learning_rate": 2.319288376324846e-06,
+      "loss": 0.4847,
+      "step": 9664
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9662583465504331,
+      "learning_rate": 2.317602900563143e-06,
+      "loss": 0.5262,
+      "step": 9665
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0136411008586397,
+      "learning_rate": 2.3159179571847446e-06,
+      "loss": 0.5118,
+      "step": 9666
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9338870766327807,
+      "learning_rate": 2.314233546306416e-06,
+      "loss": 0.4899,
+      "step": 9667
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.472138551273479,
+      "learning_rate": 2.3125496680448877e-06,
+      "loss": 0.5472,
+      "step": 9668
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.809799494625484,
+      "learning_rate": 2.3108663225168436e-06,
+      "loss": 0.424,
+      "step": 9669
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9614148731201514,
+      "learning_rate": 2.3091835098389493e-06,
+      "loss": 0.4957,
+      "step": 9670
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0110403015033504,
+      "learning_rate": 2.307501230127812e-06,
+      "loss": 0.4923,
+      "step": 9671
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9439389881727236,
+      "learning_rate": 2.3058194835000167e-06,
+      "loss": 0.4792,
+      "step": 9672
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9739089259424125,
+      "learning_rate": 2.3041382700721073e-06,
+      "loss": 0.5028,
+      "step": 9673
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9288763116072336,
+      "learning_rate": 2.3024575899605906e-06,
+      "loss": 0.5383,
+      "step": 9674
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0166464866010603,
+      "learning_rate": 2.3007774432819308e-06,
+      "loss": 0.4986,
+      "step": 9675
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9510888787436993,
+      "learning_rate": 2.2990978301525702e-06,
+      "loss": 0.4687,
+      "step": 9676
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.015865022970959,
+      "learning_rate": 2.297418750688897e-06,
+      "loss": 0.4786,
+      "step": 9677
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8650789735205615,
+      "learning_rate": 2.2957402050072717e-06,
+      "loss": 0.4507,
+      "step": 9678
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0254599114877216,
+      "learning_rate": 2.294062193224016e-06,
+      "loss": 0.4999,
+      "step": 9679
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9169938351661332,
+      "learning_rate": 2.292384715455419e-06,
+      "loss": 0.4526,
+      "step": 9680
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0437066091326381,
+      "learning_rate": 2.2907077718177183e-06,
+      "loss": 0.5488,
+      "step": 9681
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9602945871231334,
+      "learning_rate": 2.2890313624271363e-06,
+      "loss": 0.4523,
+      "step": 9682
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9967843968821193,
+      "learning_rate": 2.2873554873998393e-06,
+      "loss": 0.5272,
+      "step": 9683
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9266329545721437,
+      "learning_rate": 2.285680146851965e-06,
+      "loss": 0.4719,
+      "step": 9684
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9560564901228441,
+      "learning_rate": 2.2840053408996154e-06,
+      "loss": 0.5117,
+      "step": 9685
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9832147229170937,
+      "learning_rate": 2.28233106965885e-06,
+      "loss": 0.4882,
+      "step": 9686
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9488749898311409,
+      "learning_rate": 2.2806573332456973e-06,
+      "loss": 0.4401,
+      "step": 9687
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9169415250097036,
+      "learning_rate": 2.278984131776145e-06,
+      "loss": 0.4657,
+      "step": 9688
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8472737030846431,
+      "learning_rate": 2.2773114653661433e-06,
+      "loss": 0.4264,
+      "step": 9689
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9309930762749372,
+      "learning_rate": 2.2756393341316065e-06,
+      "loss": 0.4762,
+      "step": 9690
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9340556544258505,
+      "learning_rate": 2.2739677381884117e-06,
+      "loss": 0.4586,
+      "step": 9691
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8572938512075292,
+      "learning_rate": 2.272296677652399e-06,
+      "loss": 0.4299,
+      "step": 9692
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9464483530843911,
+      "learning_rate": 2.2706261526393734e-06,
+      "loss": 0.4628,
+      "step": 9693
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9382340497970053,
+      "learning_rate": 2.2689561632651024e-06,
+      "loss": 0.4493,
+      "step": 9694
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9729695950363514,
+      "learning_rate": 2.267286709645309e-06,
+      "loss": 0.4547,
+      "step": 9695
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9290966977858953,
+      "learning_rate": 2.2656177918956867e-06,
+      "loss": 0.499,
+      "step": 9696
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9104737506886303,
+      "learning_rate": 2.2639494101318914e-06,
+      "loss": 0.4894,
+      "step": 9697
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8565759329384154,
+      "learning_rate": 2.262281564469541e-06,
+      "loss": 0.472,
+      "step": 9698
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9881347553725216,
+      "learning_rate": 2.260614255024214e-06,
+      "loss": 0.5355,
+      "step": 9699
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8516162680336561,
+      "learning_rate": 2.2589474819114564e-06,
+      "loss": 0.4418,
+      "step": 9700
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0157475120079382,
+      "learning_rate": 2.2572812452467708e-06,
+      "loss": 0.5232,
+      "step": 9701
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9035265623515684,
+      "learning_rate": 2.255615545145626e-06,
+      "loss": 0.4444,
+      "step": 9702
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9245265400864687,
+      "learning_rate": 2.2539503817234553e-06,
+      "loss": 0.5143,
+      "step": 9703
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9118533686851575,
+      "learning_rate": 2.252285755095652e-06,
+      "loss": 0.5552,
+      "step": 9704
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8264811753969631,
+      "learning_rate": 2.2506216653775736e-06,
+      "loss": 0.5094,
+      "step": 9705
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8832554789712772,
+      "learning_rate": 2.2489581126845408e-06,
+      "loss": 0.4546,
+      "step": 9706
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8941894011676278,
+      "learning_rate": 2.2472950971318377e-06,
+      "loss": 0.4624,
+      "step": 9707
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9404419890377869,
+      "learning_rate": 2.2456326188347045e-06,
+      "loss": 0.5373,
+      "step": 9708
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8772512482369041,
+      "learning_rate": 2.2439706779083538e-06,
+      "loss": 0.431,
+      "step": 9709
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9547698813593285,
+      "learning_rate": 2.2423092744679553e-06,
+      "loss": 0.5533,
+      "step": 9710
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.93032348439044,
+      "learning_rate": 2.240648408628643e-06,
+      "loss": 0.4784,
+      "step": 9711
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9153025168867757,
+      "learning_rate": 2.238988080505513e-06,
+      "loss": 0.4683,
+      "step": 9712
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9114927690819346,
+      "learning_rate": 2.2373282902136273e-06,
+      "loss": 0.4661,
+      "step": 9713
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8452718104433999,
+      "learning_rate": 2.2356690378680036e-06,
+      "loss": 0.4077,
+      "step": 9714
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8560275740625997,
+      "learning_rate": 2.2340103235836286e-06,
+      "loss": 0.5184,
+      "step": 9715
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0038013957013665,
+      "learning_rate": 2.2323521474754508e-06,
+      "loss": 0.5474,
+      "step": 9716
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.946783776261924,
+      "learning_rate": 2.2306945096583775e-06,
+      "loss": 0.4912,
+      "step": 9717
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9489969255455457,
+      "learning_rate": 2.2290374102472846e-06,
+      "loss": 0.4942,
+      "step": 9718
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9521301748260809,
+      "learning_rate": 2.2273808493570082e-06,
+      "loss": 0.4411,
+      "step": 9719
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9577675331331204,
+      "learning_rate": 2.2257248271023424e-06,
+      "loss": 0.4104,
+      "step": 9720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0214226660713241,
+      "learning_rate": 2.22406934359805e-06,
+      "loss": 0.4496,
+      "step": 9721
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8994988307598074,
+      "learning_rate": 2.2224143989588545e-06,
+      "loss": 0.4956,
+      "step": 9722
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8937085342620766,
+      "learning_rate": 2.2207599932994427e-06,
+      "loss": 0.4643,
+      "step": 9723
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9139353897263605,
+      "learning_rate": 2.2191061267344636e-06,
+      "loss": 0.4705,
+      "step": 9724
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8709930127962902,
+      "learning_rate": 2.217452799378531e-06,
+      "loss": 0.4819,
+      "step": 9725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.995164754817237,
+      "learning_rate": 2.215800011346211e-06,
+      "loss": 0.5095,
+      "step": 9726
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8863722636206753,
+      "learning_rate": 2.2141477627520504e-06,
+      "loss": 0.4836,
+      "step": 9727
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0011157201630954,
+      "learning_rate": 2.212496053710541e-06,
+      "loss": 0.5673,
+      "step": 9728
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9278185762841324,
+      "learning_rate": 2.2108448843361487e-06,
+      "loss": 0.4849,
+      "step": 9729
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9013886456943522,
+      "learning_rate": 2.209194254743295e-06,
+      "loss": 0.4775,
+      "step": 9730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9261698216325288,
+      "learning_rate": 2.2075441650463734e-06,
+      "loss": 0.4504,
+      "step": 9731
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0073436550756663,
+      "learning_rate": 2.205894615359724e-06,
+      "loss": 0.5411,
+      "step": 9732
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1251748379561548,
+      "learning_rate": 2.2042456057976693e-06,
+      "loss": 0.5031,
+      "step": 9733
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.012572403508521,
+      "learning_rate": 2.2025971364744758e-06,
+      "loss": 0.506,
+      "step": 9734
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.93453836437706,
+      "learning_rate": 2.2009492075043847e-06,
+      "loss": 0.5359,
+      "step": 9735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9697596467071894,
+      "learning_rate": 2.199301819001597e-06,
+      "loss": 0.5473,
+      "step": 9736
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.071892038087734,
+      "learning_rate": 2.1976549710802754e-06,
+      "loss": 0.4695,
+      "step": 9737
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0109392674982929,
+      "learning_rate": 2.1960086638545385e-06,
+      "loss": 0.4593,
+      "step": 9738
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8912902858755715,
+      "learning_rate": 2.1943628974384858e-06,
+      "loss": 0.4292,
+      "step": 9739
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9169358523666661,
+      "learning_rate": 2.192717671946156e-06,
+      "loss": 0.5011,
+      "step": 9740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9412617557898649,
+      "learning_rate": 2.191072987491567e-06,
+      "loss": 0.479,
+      "step": 9741
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9641902627726724,
+      "learning_rate": 2.1894288441886946e-06,
+      "loss": 0.484,
+      "step": 9742
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9835215942561565,
+      "learning_rate": 2.1877852421514767e-06,
+      "loss": 0.5121,
+      "step": 9743
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8855674381525747,
+      "learning_rate": 2.1861421814938076e-06,
+      "loss": 0.3979,
+      "step": 9744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9306720357373683,
+      "learning_rate": 2.18449966232956e-06,
+      "loss": 0.487,
+      "step": 9745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9644690013564537,
+      "learning_rate": 2.18285768477255e-06,
+      "loss": 0.4636,
+      "step": 9746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9276625157351951,
+      "learning_rate": 2.1812162489365686e-06,
+      "loss": 0.5162,
+      "step": 9747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9515009929127202,
+      "learning_rate": 2.179575354935366e-06,
+      "loss": 0.5239,
+      "step": 9748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9622644087270956,
+      "learning_rate": 2.1779350028826584e-06,
+      "loss": 0.5131,
+      "step": 9749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9126946605383349,
+      "learning_rate": 2.1762951928921105e-06,
+      "loss": 0.4957,
+      "step": 9750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0443250368096582,
+      "learning_rate": 2.174655925077371e-06,
+      "loss": 0.5188,
+      "step": 9751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8529784079548666,
+      "learning_rate": 2.1730171995520334e-06,
+      "loss": 0.452,
+      "step": 9752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9125903642765387,
+      "learning_rate": 2.171379016429661e-06,
+      "loss": 0.4801,
+      "step": 9753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2263550367912526,
+      "learning_rate": 2.1697413758237785e-06,
+      "loss": 0.5218,
+      "step": 9754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8015976310048045,
+      "learning_rate": 2.1681042778478755e-06,
+      "loss": 0.4424,
+      "step": 9755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0733405439258612,
+      "learning_rate": 2.166467722615394e-06,
+      "loss": 0.5043,
+      "step": 9756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9307547619272879,
+      "learning_rate": 2.1648317102397565e-06,
+      "loss": 0.4392,
+      "step": 9757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9033183363968029,
+      "learning_rate": 2.1631962408343264e-06,
+      "loss": 0.5032,
+      "step": 9758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0066450643953964,
+      "learning_rate": 2.1615613145124514e-06,
+      "loss": 0.5493,
+      "step": 9759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9444646156175104,
+      "learning_rate": 2.1599269313874217e-06,
+      "loss": 0.488,
+      "step": 9760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.014755699670725,
+      "learning_rate": 2.158293091572501e-06,
+      "loss": 0.5632,
+      "step": 9761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.022504963163021,
+      "learning_rate": 2.156659795180913e-06,
+      "loss": 0.5,
+      "step": 9762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8841065130850687,
+      "learning_rate": 2.155027042325848e-06,
+      "loss": 0.5152,
+      "step": 9763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8805351611241143,
+      "learning_rate": 2.1533948331204445e-06,
+      "loss": 0.4458,
+      "step": 9764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1318116744956215,
+      "learning_rate": 2.151763167677825e-06,
+      "loss": 0.4929,
+      "step": 9765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9358085439942705,
+      "learning_rate": 2.150132046111054e-06,
+      "loss": 0.5011,
+      "step": 9766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9357248606524596,
+      "learning_rate": 2.1485014685331684e-06,
+      "loss": 0.5508,
+      "step": 9767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9589413280149912,
+      "learning_rate": 2.1468714350571683e-06,
+      "loss": 0.46,
+      "step": 9768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9224443985715425,
+      "learning_rate": 2.145241945796014e-06,
+      "loss": 0.4801,
+      "step": 9769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8847139248412932,
+      "learning_rate": 2.14361300086262e-06,
+      "loss": 0.5088,
+      "step": 9770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.962732416595997,
+      "learning_rate": 2.141984600369882e-06,
+      "loss": 0.4688,
+      "step": 9771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8677824980011563,
+      "learning_rate": 2.1403567444306384e-06,
+      "loss": 0.4751,
+      "step": 9772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9071070373566746,
+      "learning_rate": 2.1387294331577e-06,
+      "loss": 0.4645,
+      "step": 9773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9406593469392999,
+      "learning_rate": 2.1371026666638404e-06,
+      "loss": 0.4911,
+      "step": 9774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9851660325459202,
+      "learning_rate": 2.1354764450617937e-06,
+      "loss": 0.4579,
+      "step": 9775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9575769497626949,
+      "learning_rate": 2.1338507684642483e-06,
+      "loss": 0.4854,
+      "step": 9776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9043458981275626,
+      "learning_rate": 2.1322256369838723e-06,
+      "loss": 0.4312,
+      "step": 9777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9136899357618365,
+      "learning_rate": 2.1306010507332787e-06,
+      "loss": 0.4731,
+      "step": 9778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9489654635266751,
+      "learning_rate": 2.128977009825052e-06,
+      "loss": 0.4946,
+      "step": 9779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.886178646999231,
+      "learning_rate": 2.1273535143717372e-06,
+      "loss": 0.4779,
+      "step": 9780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9274155006249445,
+      "learning_rate": 2.125730564485844e-06,
+      "loss": 0.4499,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0206326818075697,
+      "learning_rate": 2.124108160279832e-06,
+      "loss": 0.5138,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9908500142071262,
+      "learning_rate": 2.1224863018661435e-06,
+      "loss": 0.5542,
+      "step": 9783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9421564240045448,
+      "learning_rate": 2.1208649893571653e-06,
+      "loss": 0.472,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9529297223093769,
+      "learning_rate": 2.119244222865253e-06,
+      "loss": 0.5093,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.983152381687885,
+      "learning_rate": 2.117624002502727e-06,
+      "loss": 0.5081,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9269249520941745,
+      "learning_rate": 2.1160043283818697e-06,
+      "loss": 0.4889,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9311810373511374,
+      "learning_rate": 2.114385200614912e-06,
+      "loss": 0.4725,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9330190902902091,
+      "learning_rate": 2.112766619314072e-06,
+      "loss": 0.5116,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9001553215481263,
+      "learning_rate": 2.111148584591506e-06,
+      "loss": 0.5007,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0008529581665582,
+      "learning_rate": 2.1095310965593463e-06,
+      "loss": 0.5086,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.969189216671511,
+      "learning_rate": 2.107914155329682e-06,
+      "loss": 0.542,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9746252765484869,
+      "learning_rate": 2.1062977610145697e-06,
+      "loss": 0.4838,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9223795804782231,
+      "learning_rate": 2.1046819137260155e-06,
+      "loss": 0.4721,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9041889187855691,
+      "learning_rate": 2.103066613576007e-06,
+      "loss": 0.476,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9913638606757014,
+      "learning_rate": 2.1014518606764744e-06,
+      "loss": 0.4791,
+      "step": 9796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.97369286260393,
+      "learning_rate": 2.0998376551393218e-06,
+      "loss": 0.5438,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8437686252604286,
+      "learning_rate": 2.0982239970764127e-06,
+      "loss": 0.3833,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8703143734179483,
+      "learning_rate": 2.096610886599575e-06,
+      "loss": 0.4263,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.035672950025825,
+      "learning_rate": 2.0949983238205863e-06,
+      "loss": 0.542,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9362809465383324,
+      "learning_rate": 2.0933863088512076e-06,
+      "loss": 0.4959,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9017408168518072,
+      "learning_rate": 2.0917748418031415e-06,
+      "loss": 0.4708,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0087137373104949,
+      "learning_rate": 2.0901639227880643e-06,
+      "loss": 0.4836,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8750531971104142,
+      "learning_rate": 2.0885535519176115e-06,
+      "loss": 0.4628,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9406350767966913,
+      "learning_rate": 2.0869437293033835e-06,
+      "loss": 0.4725,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9721435535796548,
+      "learning_rate": 2.08533445505693e-06,
+      "loss": 0.4899,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9351073095710531,
+      "learning_rate": 2.083725729289784e-06,
+      "loss": 0.462,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9166700362208122,
+      "learning_rate": 2.0821175521134208e-06,
+      "loss": 0.4798,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9199003461402707,
+      "learning_rate": 2.080509923639288e-06,
+      "loss": 0.4413,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9284407199731455,
+      "learning_rate": 2.078902843978792e-06,
+      "loss": 0.4966,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.882987750650245,
+      "learning_rate": 2.0772963132433065e-06,
+      "loss": 0.4945,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9658350421424726,
+      "learning_rate": 2.0756903315441535e-06,
+      "loss": 0.5078,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9092385679023566,
+      "learning_rate": 2.0740848989926365e-06,
+      "loss": 0.4854,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0058460963863638,
+      "learning_rate": 2.0724800157000034e-06,
+      "loss": 0.5533,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8937420971510914,
+      "learning_rate": 2.0708756817774743e-06,
+      "loss": 0.401,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0443858846384912,
+      "learning_rate": 2.069271897336227e-06,
+      "loss": 0.5096,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0056854570600682,
+      "learning_rate": 2.0676686624874054e-06,
+      "loss": 0.5215,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9695529148626671,
+      "learning_rate": 2.066065977342103e-06,
+      "loss": 0.5354,
+      "step": 9818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8368337032103081,
+      "learning_rate": 2.064463842011397e-06,
+      "loss": 0.4196,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8703363610587038,
+      "learning_rate": 2.0628622566063063e-06,
+      "loss": 0.481,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8656780500628107,
+      "learning_rate": 2.06126122123782e-06,
+      "loss": 0.431,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.89463213892815,
+      "learning_rate": 2.0596607360168897e-06,
+      "loss": 0.5256,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8925044512010746,
+      "learning_rate": 2.058060801054429e-06,
+      "loss": 0.4504,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8702517722379945,
+      "learning_rate": 2.0564614164613064e-06,
+      "loss": 0.4574,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9533218505721511,
+      "learning_rate": 2.054862582348366e-06,
+      "loss": 0.5221,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9865782669574705,
+      "learning_rate": 2.0532642988263994e-06,
+      "loss": 0.5369,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.086952839581523,
+      "learning_rate": 2.0516665660061675e-06,
+      "loss": 0.5524,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9065297806646161,
+      "learning_rate": 2.050069383998393e-06,
+      "loss": 0.4822,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9407575239281395,
+      "learning_rate": 2.0484727529137616e-06,
+      "loss": 0.4421,
+      "step": 9829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0282959759015617,
+      "learning_rate": 2.0468766728629084e-06,
+      "loss": 0.5221,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9230027591994061,
+      "learning_rate": 2.045281143956455e-06,
+      "loss": 0.4707,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9706643303868061,
+      "learning_rate": 2.0436861663049577e-06,
+      "loss": 0.5287,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8624615421512666,
+      "learning_rate": 2.0420917400189532e-06,
+      "loss": 0.4672,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.019489997515685,
+      "learning_rate": 2.0404978652089325e-06,
+      "loss": 0.5163,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8595574720850923,
+      "learning_rate": 2.0389045419853483e-06,
+      "loss": 0.4102,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0556470578827364,
+      "learning_rate": 2.037311770458619e-06,
+      "loss": 0.4998,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9005746593140518,
+      "learning_rate": 2.0357195507391237e-06,
+      "loss": 0.4978,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9848758102170952,
+      "learning_rate": 2.034127882937197e-06,
+      "loss": 0.5397,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0766503341546003,
+      "learning_rate": 2.032536767163141e-06,
+      "loss": 0.5174,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0545814409469874,
+      "learning_rate": 2.0309462035272207e-06,
+      "loss": 0.5166,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9315275679597564,
+      "learning_rate": 2.02935619213966e-06,
+      "loss": 0.4328,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9159068671624757,
+      "learning_rate": 2.0277667331106456e-06,
+      "loss": 0.4934,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9956840495068574,
+      "learning_rate": 2.026177826550326e-06,
+      "loss": 0.5241,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9763647677692302,
+      "learning_rate": 2.0245894725688097e-06,
+      "loss": 0.509,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8663904354062106,
+      "learning_rate": 2.023001671276168e-06,
+      "loss": 0.476,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9302822464015351,
+      "learning_rate": 2.021414422782435e-06,
+      "loss": 0.526,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9151958944350562,
+      "learning_rate": 2.019827727197605e-06,
+      "loss": 0.4823,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.871270534613334,
+      "learning_rate": 2.018241584631636e-06,
+      "loss": 0.471,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9485613903404995,
+      "learning_rate": 2.0166559951944477e-06,
+      "loss": 0.5482,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0218535933745767,
+      "learning_rate": 2.015070958995915e-06,
+      "loss": 0.5702,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9870914624477832,
+      "learning_rate": 2.0134864761458815e-06,
+      "loss": 0.5596,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9066468027478812,
+      "learning_rate": 2.011902546754152e-06,
+      "loss": 0.4668,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.929133236447092,
+      "learning_rate": 2.01031917093049e-06,
+      "loss": 0.5044,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9296472004568087,
+      "learning_rate": 2.0087363487846236e-06,
+      "loss": 0.464,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8409541167939985,
+      "learning_rate": 2.007154080426239e-06,
+      "loss": 0.4338,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8973893618611448,
+      "learning_rate": 2.0055723659649907e-06,
+      "loss": 0.4885,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8864057016507616,
+      "learning_rate": 2.0039912055104826e-06,
+      "loss": 0.412,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9103346867478691,
+      "learning_rate": 2.002410599172292e-06,
+      "loss": 0.4636,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9040653111981451,
+      "learning_rate": 2.0008305470599533e-06,
+      "loss": 0.3976,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9571850512409811,
+      "learning_rate": 1.999251049282962e-06,
+      "loss": 0.4924,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9831040082324778,
+      "learning_rate": 1.9976721059507766e-06,
+      "loss": 0.5063,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9674607752432008,
+      "learning_rate": 1.996093717172819e-06,
+      "loss": 0.4871,
+      "step": 9862
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8735604631321405,
+      "learning_rate": 1.994515883058464e-06,
+      "loss": 0.443,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8727484466187492,
+      "learning_rate": 1.9929386037170574e-06,
+      "loss": 0.4173,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.893569677723104,
+      "learning_rate": 1.9913618792579037e-06,
+      "loss": 0.4256,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.79604340010775,
+      "learning_rate": 1.9897857097902683e-06,
+      "loss": 0.4173,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9552205687203998,
+      "learning_rate": 1.9882100954233786e-06,
+      "loss": 0.449,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9992518118991707,
+      "learning_rate": 1.9866350362664243e-06,
+      "loss": 0.4373,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9556895391027411,
+      "learning_rate": 1.98506053242855e-06,
+      "loss": 0.529,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9899771953409056,
+      "learning_rate": 1.9834865840188767e-06,
+      "loss": 0.4856,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9073343077121624,
+      "learning_rate": 1.9819131911464682e-06,
+      "loss": 0.4587,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8637435483735759,
+      "learning_rate": 1.9803403539203657e-06,
+      "loss": 0.4847,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8414267955167837,
+      "learning_rate": 1.9787680724495617e-06,
+      "loss": 0.4817,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8750397831401274,
+      "learning_rate": 1.977196346843019e-06,
+      "loss": 0.4716,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.986056259393715,
+      "learning_rate": 1.975625177209648e-06,
+      "loss": 0.5231,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9361734808532571,
+      "learning_rate": 1.9740545636583397e-06,
+      "loss": 0.5589,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8785889300977264,
+      "learning_rate": 1.9724845062979283e-06,
+      "loss": 0.4577,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9376677737805472,
+      "learning_rate": 1.9709150052372206e-06,
+      "loss": 0.4759,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9065893288202759,
+      "learning_rate": 1.96934606058498e-06,
+      "loss": 0.5017,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9532402399142695,
+      "learning_rate": 1.9677776724499354e-06,
+      "loss": 0.4666,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.956789992815111,
+      "learning_rate": 1.9662098409407737e-06,
+      "loss": 0.489,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9697198209953076,
+      "learning_rate": 1.964642566166146e-06,
+      "loss": 0.4756,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9394385645475665,
+      "learning_rate": 1.963075848234659e-06,
+      "loss": 0.4685,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8343886173938521,
+      "learning_rate": 1.9615096872548865e-06,
+      "loss": 0.3947,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8882866642715643,
+      "learning_rate": 1.9599440833353624e-06,
+      "loss": 0.468,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9117902743086408,
+      "learning_rate": 1.9583790365845823e-06,
+      "loss": 0.5095,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.919139555747959,
+      "learning_rate": 1.9568145471110024e-06,
+      "loss": 0.4837,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9313516499344854,
+      "learning_rate": 1.955250615023042e-06,
+      "loss": 0.4367,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9586301700560018,
+      "learning_rate": 1.953687240429073e-06,
+      "loss": 0.5063,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9736966037058312,
+      "learning_rate": 1.952124423437447e-06,
+      "loss": 0.5151,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8183733956732533,
+      "learning_rate": 1.9505621641564567e-06,
+      "loss": 0.4253,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9848703923333934,
+      "learning_rate": 1.9490004626943693e-06,
+      "loss": 0.5306,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9515937598579888,
+      "learning_rate": 1.9474393191594076e-06,
+      "loss": 0.5488,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9331171965430829,
+      "learning_rate": 1.9458787336597617e-06,
+      "loss": 0.5142,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9951701520003015,
+      "learning_rate": 1.9443187063035707e-06,
+      "loss": 0.4915,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0027301745746888,
+      "learning_rate": 1.9427592371989533e-06,
+      "loss": 0.4976,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0164921417024477,
+      "learning_rate": 1.94120032645397e-06,
+      "loss": 0.4775,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9411467082573,
+      "learning_rate": 1.939641974176658e-06,
+      "loss": 0.4446,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9271568784067858,
+      "learning_rate": 1.9380841804750063e-06,
+      "loss": 0.4688,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9318796970360436,
+      "learning_rate": 1.9365269454569724e-06,
+      "loss": 0.4144,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9201807476325126,
+      "learning_rate": 1.934970269230464e-06,
+      "loss": 0.4805,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9016938331848386,
+      "learning_rate": 1.9334141519033676e-06,
+      "loss": 0.4726,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.944460903658482,
+      "learning_rate": 1.931858593583513e-06,
+      "loss": 0.4517,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9583119094345428,
+      "learning_rate": 1.9303035943787017e-06,
+      "loss": 0.4548,
+      "step": 9904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9361054748728511,
+      "learning_rate": 1.928749154396693e-06,
+      "loss": 0.4618,
+      "step": 9905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9730274138362711,
+      "learning_rate": 1.9271952737452116e-06,
+      "loss": 0.5136,
+      "step": 9906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9661630484610683,
+      "learning_rate": 1.9256419525319316e-06,
+      "loss": 0.5029,
+      "step": 9907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0001036297857349,
+      "learning_rate": 1.9240891908645075e-06,
+      "loss": 0.4867,
+      "step": 9908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.928526629453305,
+      "learning_rate": 1.9225369888505364e-06,
+      "loss": 0.4901,
+      "step": 9909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9721509399919,
+      "learning_rate": 1.920985346597588e-06,
+      "loss": 0.469,
+      "step": 9910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9657540152337845,
+      "learning_rate": 1.919434264213188e-06,
+      "loss": 0.4857,
+      "step": 9911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8993769666472281,
+      "learning_rate": 1.917883741804829e-06,
+      "loss": 0.4626,
+      "step": 9912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9028837295832993,
+      "learning_rate": 1.916333779479953e-06,
+      "loss": 0.4623,
+      "step": 9913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9350439033146903,
+      "learning_rate": 1.914784377345982e-06,
+      "loss": 0.4892,
+      "step": 9914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9658679451666188,
+      "learning_rate": 1.9132355355102772e-06,
+      "loss": 0.5008,
+      "step": 9915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0958571373836872,
+      "learning_rate": 1.911687254080179e-06,
+      "loss": 0.5447,
+      "step": 9916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9567367250311967,
+      "learning_rate": 1.910139533162978e-06,
+      "loss": 0.5503,
+      "step": 9917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9283114470381654,
+      "learning_rate": 1.908592372865935e-06,
+      "loss": 0.4855,
+      "step": 9918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.902042369443858,
+      "learning_rate": 1.907045773296259e-06,
+      "loss": 0.4982,
+      "step": 9919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.904011334270541,
+      "learning_rate": 1.905499734561137e-06,
+      "loss": 0.4867,
+      "step": 9920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8789908542582429,
+      "learning_rate": 1.9039542567677005e-06,
+      "loss": 0.4767,
+      "step": 9921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9020209635226962,
+      "learning_rate": 1.9024093400230537e-06,
+      "loss": 0.4793,
+      "step": 9922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9024267922140833,
+      "learning_rate": 1.9008649844342563e-06,
+      "loss": 0.4655,
+      "step": 9923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8526771021123333,
+      "learning_rate": 1.8993211901083353e-06,
+      "loss": 0.4777,
+      "step": 9924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9247623057643407,
+      "learning_rate": 1.8977779571522648e-06,
+      "loss": 0.4666,
+      "step": 9925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.904002456384038,
+      "learning_rate": 1.8962352856729994e-06,
+      "loss": 0.4571,
+      "step": 9926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9967168916175249,
+      "learning_rate": 1.894693175777439e-06,
+      "loss": 0.4864,
+      "step": 9927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9685133596114146,
+      "learning_rate": 1.8931516275724527e-06,
+      "loss": 0.4826,
+      "step": 9928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9063216564545956,
+      "learning_rate": 1.8916106411648671e-06,
+      "loss": 0.4927,
+      "step": 9929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9582577486773548,
+      "learning_rate": 1.8900702166614748e-06,
+      "loss": 0.5338,
+      "step": 9930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9567418434865032,
+      "learning_rate": 1.888530354169017e-06,
+      "loss": 0.5355,
+      "step": 9931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.003573912510184,
+      "learning_rate": 1.886991053794217e-06,
+      "loss": 0.5383,
+      "step": 9932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9694975334450654,
+      "learning_rate": 1.8854523156437378e-06,
+      "loss": 0.484,
+      "step": 9933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.098245385864126,
+      "learning_rate": 1.8839141398242145e-06,
+      "loss": 0.488,
+      "step": 9934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8761128268936527,
+      "learning_rate": 1.8823765264422433e-06,
+      "loss": 0.4723,
+      "step": 9935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9493800786184847,
+      "learning_rate": 1.8808394756043813e-06,
+      "loss": 0.4919,
+      "step": 9936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3088263147613775,
+      "learning_rate": 1.8793029874171365e-06,
+      "loss": 0.5155,
+      "step": 9937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9159160082187201,
+      "learning_rate": 1.877767061986997e-06,
+      "loss": 0.4776,
+      "step": 9938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9965361309784785,
+      "learning_rate": 1.8762316994203933e-06,
+      "loss": 0.5508,
+      "step": 9939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9365775609977088,
+      "learning_rate": 1.874696899823727e-06,
+      "loss": 0.4507,
+      "step": 9940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9352189676628262,
+      "learning_rate": 1.8731626633033573e-06,
+      "loss": 0.4177,
+      "step": 9941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.001110662973336,
+      "learning_rate": 1.8716289899656104e-06,
+      "loss": 0.4951,
+      "step": 9942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9350494135836109,
+      "learning_rate": 1.870095879916759e-06,
+      "loss": 0.4889,
+      "step": 9943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.897962644434671,
+      "learning_rate": 1.868563333263057e-06,
+      "loss": 0.4506,
+      "step": 9944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9040067942876323,
+      "learning_rate": 1.8670313501107007e-06,
+      "loss": 0.4252,
+      "step": 9945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9556567090230454,
+      "learning_rate": 1.8654999305658584e-06,
+      "loss": 0.4749,
+      "step": 9946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0508522995130318,
+      "learning_rate": 1.863969074734655e-06,
+      "loss": 0.5285,
+      "step": 9947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9940303655091299,
+      "learning_rate": 1.8624387827231815e-06,
+      "loss": 0.4676,
+      "step": 9948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0131111844858838,
+      "learning_rate": 1.8609090546374764e-06,
+      "loss": 0.4565,
+      "step": 9949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0139059658204395,
+      "learning_rate": 1.8593798905835602e-06,
+      "loss": 0.5247,
+      "step": 9950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0027110042791425,
+      "learning_rate": 1.857851290667394e-06,
+      "loss": 0.5166,
+      "step": 9951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9092026195676937,
+      "learning_rate": 1.8563232549949107e-06,
+      "loss": 0.4741,
+      "step": 9952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9812192496856625,
+      "learning_rate": 1.8547957836720032e-06,
+      "loss": 0.5114,
+      "step": 9953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9298750878339397,
+      "learning_rate": 1.853268876804526e-06,
+      "loss": 0.4892,
+      "step": 9954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9908936671001775,
+      "learning_rate": 1.8517425344982831e-06,
+      "loss": 0.4935,
+      "step": 9955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8566091038180487,
+      "learning_rate": 1.8502167568590611e-06,
+      "loss": 0.4603,
+      "step": 9956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.857918949746223,
+      "learning_rate": 1.8486915439925857e-06,
+      "loss": 0.4569,
+      "step": 9957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9498102863681523,
+      "learning_rate": 1.8471668960045575e-06,
+      "loss": 0.4881,
+      "step": 9958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.020434986705423,
+      "learning_rate": 1.845642813000631e-06,
+      "loss": 0.5073,
+      "step": 9959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8760589828007601,
+      "learning_rate": 1.8441192950864273e-06,
+      "loss": 0.4469,
+      "step": 9960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9536048276057314,
+      "learning_rate": 1.8425963423675164e-06,
+      "loss": 0.4702,
+      "step": 9961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9686770758998478,
+      "learning_rate": 1.8410739549494494e-06,
+      "loss": 0.5042,
+      "step": 9962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9767156015980288,
+      "learning_rate": 1.8395521329377175e-06,
+      "loss": 0.5231,
+      "step": 9963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8311337343768164,
+      "learning_rate": 1.8380308764377841e-06,
+      "loss": 0.4839,
+      "step": 9964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8986677240710658,
+      "learning_rate": 1.8365101855550716e-06,
+      "loss": 0.564,
+      "step": 9965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0756588962698206,
+      "learning_rate": 1.8349900603949644e-06,
+      "loss": 0.4841,
+      "step": 9966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8578985450159758,
+      "learning_rate": 1.8334705010627996e-06,
+      "loss": 0.4791,
+      "step": 9967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9483676107078279,
+      "learning_rate": 1.8319515076638893e-06,
+      "loss": 0.4844,
+      "step": 9968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8803822299638585,
+      "learning_rate": 1.8304330803034932e-06,
+      "loss": 0.4359,
+      "step": 9969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9990550895909772,
+      "learning_rate": 1.8289152190868376e-06,
+      "loss": 0.4617,
+      "step": 9970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0814653391866287,
+      "learning_rate": 1.8273979241191087e-06,
+      "loss": 0.5006,
+      "step": 9971
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8372147293790334,
+      "learning_rate": 1.8258811955054578e-06,
+      "loss": 0.4535,
+      "step": 9972
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.90090598405615,
+      "learning_rate": 1.8243650333509854e-06,
+      "loss": 0.4859,
+      "step": 9973
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8238375354825378,
+      "learning_rate": 1.8228494377607686e-06,
+      "loss": 0.3835,
+      "step": 9974
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9275498090759045,
+      "learning_rate": 1.82133440883983e-06,
+      "loss": 0.4434,
+      "step": 9975
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9679147524700789,
+      "learning_rate": 1.819819946693162e-06,
+      "loss": 0.4931,
+      "step": 9976
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9938967763690725,
+      "learning_rate": 1.8183060514257167e-06,
+      "loss": 0.506,
+      "step": 9977
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0073247438748036,
+      "learning_rate": 1.8167927231424077e-06,
+      "loss": 0.492,
+      "step": 9978
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9441319716629862,
+      "learning_rate": 1.8152799619480986e-06,
+      "loss": 0.4797,
+      "step": 9979
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8386503280748313,
+      "learning_rate": 1.813767767947634e-06,
+      "loss": 0.446,
+      "step": 9980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9339502064681194,
+      "learning_rate": 1.8122561412457984e-06,
+      "loss": 0.4854,
+      "step": 9981
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9193184258857796,
+      "learning_rate": 1.8107450819473505e-06,
+      "loss": 0.4928,
+      "step": 9982
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8856714395591396,
+      "learning_rate": 1.8092345901570053e-06,
+      "loss": 0.4528,
+      "step": 9983
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9024058082729695,
+      "learning_rate": 1.8077246659794368e-06,
+      "loss": 0.4922,
+      "step": 9984
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9992991290855968,
+      "learning_rate": 1.8062153095192826e-06,
+      "loss": 0.4805,
+      "step": 9985
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9208122168752765,
+      "learning_rate": 1.8047065208811421e-06,
+      "loss": 0.5148,
+      "step": 9986
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8837422771578278,
+      "learning_rate": 1.8031983001695674e-06,
+      "loss": 0.4828,
+      "step": 9987
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9831378843029416,
+      "learning_rate": 1.8016906474890805e-06,
+      "loss": 0.5306,
+      "step": 9988
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9996761669814016,
+      "learning_rate": 1.80018356294416e-06,
+      "loss": 0.4487,
+      "step": 9989
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9537374004354636,
+      "learning_rate": 1.7986770466392445e-06,
+      "loss": 0.5528,
+      "step": 9990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9596722342287717,
+      "learning_rate": 1.797171098678736e-06,
+      "loss": 0.486,
+      "step": 9991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0236408628013114,
+      "learning_rate": 1.7956657191669969e-06,
+      "loss": 0.5186,
+      "step": 9992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8524064003382771,
+      "learning_rate": 1.7941609082083434e-06,
+      "loss": 0.4421,
+      "step": 9993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8894376608304295,
+      "learning_rate": 1.792656665907061e-06,
+      "loss": 0.4256,
+      "step": 9994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8876867329730918,
+      "learning_rate": 1.7911529923673908e-06,
+      "loss": 0.4742,
+      "step": 9995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0183024170797783,
+      "learning_rate": 1.7896498876935374e-06,
+      "loss": 0.5163,
+      "step": 9996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9581560806374572,
+      "learning_rate": 1.7881473519896642e-06,
+      "loss": 0.4665,
+      "step": 9997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8977743384520495,
+      "learning_rate": 1.7866453853598985e-06,
+      "loss": 0.497,
+      "step": 9998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.967090112841617,
+      "learning_rate": 1.7851439879083188e-06,
+      "loss": 0.4658,
+      "step": 9999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9733260797622603,
+      "learning_rate": 1.7836431597389758e-06,
+      "loss": 0.5535,
+      "step": 10000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9988413406779859,
+      "learning_rate": 1.7821429009558723e-06,
+      "loss": 0.5021,
+      "step": 10001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9240963254362822,
+      "learning_rate": 1.7806432116629779e-06,
+      "loss": 0.4725,
+      "step": 10002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9397626741846741,
+      "learning_rate": 1.7791440919642178e-06,
+      "loss": 0.5162,
+      "step": 10003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0066418221971942,
+      "learning_rate": 1.7776455419634797e-06,
+      "loss": 0.4792,
+      "step": 10004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.007951156924576,
+      "learning_rate": 1.776147561764613e-06,
+      "loss": 0.4439,
+      "step": 10005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.900720562872108,
+      "learning_rate": 1.7746501514714277e-06,
+      "loss": 0.4627,
+      "step": 10006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8383180580559433,
+      "learning_rate": 1.7731533111876887e-06,
+      "loss": 0.4472,
+      "step": 10007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.922154067229698,
+      "learning_rate": 1.7716570410171285e-06,
+      "loss": 0.4728,
+      "step": 10008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9071711336797791,
+      "learning_rate": 1.7701613410634367e-06,
+      "loss": 0.4782,
+      "step": 10009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9574378588931637,
+      "learning_rate": 1.7686662114302633e-06,
+      "loss": 0.4943,
+      "step": 10010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.974259591928751,
+      "learning_rate": 1.7671716522212212e-06,
+      "loss": 0.4604,
+      "step": 10011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8729156997285382,
+      "learning_rate": 1.7656776635398832e-06,
+      "loss": 0.4835,
+      "step": 10012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9662586084214412,
+      "learning_rate": 1.7641842454897772e-06,
+      "loss": 0.472,
+      "step": 10013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8912794524180124,
+      "learning_rate": 1.7626913981743975e-06,
+      "loss": 0.4739,
+      "step": 10014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9871749963458629,
+      "learning_rate": 1.761199121697197e-06,
+      "loss": 0.4227,
+      "step": 10015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9448824397611246,
+      "learning_rate": 1.75970741616159e-06,
+      "loss": 0.5074,
+      "step": 10016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9480742488796361,
+      "learning_rate": 1.7582162816709503e-06,
+      "loss": 0.451,
+      "step": 10017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.090097064319148,
+      "learning_rate": 1.7567257183286113e-06,
+      "loss": 0.5197,
+      "step": 10018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9388593712796945,
+      "learning_rate": 1.7552357262378705e-06,
+      "loss": 0.4594,
+      "step": 10019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9233057993892104,
+      "learning_rate": 1.7537463055019788e-06,
+      "loss": 0.4664,
+      "step": 10020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9478652667485742,
+      "learning_rate": 1.7522574562241535e-06,
+      "loss": 0.4949,
+      "step": 10021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.880962159646419,
+      "learning_rate": 1.750769178507571e-06,
+      "loss": 0.5246,
+      "step": 10022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0028321411856618,
+      "learning_rate": 1.7492814724553664e-06,
+      "loss": 0.4973,
+      "step": 10023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9466321446941334,
+      "learning_rate": 1.7477943381706386e-06,
+      "loss": 0.4993,
+      "step": 10024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9446176667725231,
+      "learning_rate": 1.7463077757564452e-06,
+      "loss": 0.4882,
+      "step": 10025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8584036715987096,
+      "learning_rate": 1.7448217853158e-06,
+      "loss": 0.4834,
+      "step": 10026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9167014228298731,
+      "learning_rate": 1.7433363669516823e-06,
+      "loss": 0.4908,
+      "step": 10027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8828091074105395,
+      "learning_rate": 1.7418515207670306e-06,
+      "loss": 0.4757,
+      "step": 10028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9526950744639585,
+      "learning_rate": 1.7403672468647436e-06,
+      "loss": 0.5107,
+      "step": 10029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.949876811338634,
+      "learning_rate": 1.7388835453476805e-06,
+      "loss": 0.5153,
+      "step": 10030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8093822357593842,
+      "learning_rate": 1.737400416318663e-06,
+      "loss": 0.4581,
+      "step": 10031
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9160558834764048,
+      "learning_rate": 1.7359178598804637e-06,
+      "loss": 0.5068,
+      "step": 10032
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9326020932997082,
+      "learning_rate": 1.7344358761358283e-06,
+      "loss": 0.5048,
+      "step": 10033
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9103459982568537,
+      "learning_rate": 1.7329544651874542e-06,
+      "loss": 0.511,
+      "step": 10034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0039486942702247,
+      "learning_rate": 1.7314736271380029e-06,
+      "loss": 0.4903,
+      "step": 10035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9825062782601638,
+      "learning_rate": 1.7299933620900945e-06,
+      "loss": 0.5486,
+      "step": 10036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9997925465861273,
+      "learning_rate": 1.7285136701463134e-06,
+      "loss": 0.4759,
+      "step": 10037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9244469095884913,
+      "learning_rate": 1.7270345514091936e-06,
+      "loss": 0.4788,
+      "step": 10038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.033859739615693,
+      "learning_rate": 1.725556005981246e-06,
+      "loss": 0.5205,
+      "step": 10039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0181678912579275,
+      "learning_rate": 1.7240780339649255e-06,
+      "loss": 0.5217,
+      "step": 10040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9828476813834534,
+      "learning_rate": 1.7226006354626567e-06,
+      "loss": 0.5503,
+      "step": 10041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9345978924819633,
+      "learning_rate": 1.7211238105768213e-06,
+      "loss": 0.5011,
+      "step": 10042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9078361413963695,
+      "learning_rate": 1.719647559409765e-06,
+      "loss": 0.4669,
+      "step": 10043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9774706241629969,
+      "learning_rate": 1.7181718820637839e-06,
+      "loss": 0.4891,
+      "step": 10044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8467688352719807,
+      "learning_rate": 1.7166967786411493e-06,
+      "loss": 0.456,
+      "step": 10045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9516124394986565,
+      "learning_rate": 1.7152222492440796e-06,
+      "loss": 0.444,
+      "step": 10046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.99074343373286,
+      "learning_rate": 1.7137482939747574e-06,
+      "loss": 0.5475,
+      "step": 10047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9347063175331168,
+      "learning_rate": 1.71227491293533e-06,
+      "loss": 0.4965,
+      "step": 10048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9351374387744484,
+      "learning_rate": 1.7108021062279023e-06,
+      "loss": 0.5165,
+      "step": 10049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8743820993543441,
+      "learning_rate": 1.7093298739545305e-06,
+      "loss": 0.4837,
+      "step": 10050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.964425055640185,
+      "learning_rate": 1.7078582162172509e-06,
+      "loss": 0.4377,
+      "step": 10051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.993934678759803,
+      "learning_rate": 1.7063871331180382e-06,
+      "loss": 0.4785,
+      "step": 10052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9542850256323199,
+      "learning_rate": 1.704916624758841e-06,
+      "loss": 0.5095,
+      "step": 10053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.003853157296499,
+      "learning_rate": 1.7034466912415638e-06,
+      "loss": 0.4687,
+      "step": 10054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9517152267431251,
+      "learning_rate": 1.7019773326680745e-06,
+      "loss": 0.5356,
+      "step": 10055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.895878308211716,
+      "learning_rate": 1.7005085491401908e-06,
+      "loss": 0.4277,
+      "step": 10056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9082358180531004,
+      "learning_rate": 1.6990403407597078e-06,
+      "loss": 0.4525,
+      "step": 10057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9908955989458099,
+      "learning_rate": 1.6975727076283642e-06,
+      "loss": 0.5181,
+      "step": 10058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8707024973610689,
+      "learning_rate": 1.6961056498478666e-06,
+      "loss": 0.4356,
+      "step": 10059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9765181578225404,
+      "learning_rate": 1.6946391675198838e-06,
+      "loss": 0.5583,
+      "step": 10060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.958357904820011,
+      "learning_rate": 1.6931732607460405e-06,
+      "loss": 0.52,
+      "step": 10061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0838303700246354,
+      "learning_rate": 1.6917079296279181e-06,
+      "loss": 0.5312,
+      "step": 10062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7970114790011827,
+      "learning_rate": 1.690243174267071e-06,
+      "loss": 0.4548,
+      "step": 10063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9075527996205083,
+      "learning_rate": 1.6887789947649991e-06,
+      "loss": 0.4227,
+      "step": 10064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9414766364746932,
+      "learning_rate": 1.687315391223171e-06,
+      "loss": 0.4893,
+      "step": 10065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9237635388169918,
+      "learning_rate": 1.6858523637430136e-06,
+      "loss": 0.4995,
+      "step": 10066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9143611340266095,
+      "learning_rate": 1.6843899124259133e-06,
+      "loss": 0.4894,
+      "step": 10067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9646461705809476,
+      "learning_rate": 1.6829280373732126e-06,
+      "loss": 0.4884,
+      "step": 10068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9470992263257096,
+      "learning_rate": 1.681466738686227e-06,
+      "loss": 0.5024,
+      "step": 10069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9202676930536429,
+      "learning_rate": 1.6800060164662146e-06,
+      "loss": 0.4791,
+      "step": 10070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9064573173443682,
+      "learning_rate": 1.6785458708144053e-06,
+      "loss": 0.517,
+      "step": 10071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9259280196769716,
+      "learning_rate": 1.677086301831986e-06,
+      "loss": 0.511,
+      "step": 10072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9364034530429901,
+      "learning_rate": 1.675627309620107e-06,
+      "loss": 0.5445,
+      "step": 10073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8440682150423323,
+      "learning_rate": 1.6741688942798663e-06,
+      "loss": 0.4847,
+      "step": 10074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.94140642798752,
+      "learning_rate": 1.6727110559123405e-06,
+      "loss": 0.4846,
+      "step": 10075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9448124346191853,
+      "learning_rate": 1.6712537946185503e-06,
+      "loss": 0.4815,
+      "step": 10076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9390309876950016,
+      "learning_rate": 1.6697971104994847e-06,
+      "loss": 0.4681,
+      "step": 10077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5830720212508986,
+      "learning_rate": 1.6683410036560899e-06,
+      "loss": 0.4805,
+      "step": 10078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9370005137394446,
+      "learning_rate": 1.666885474189276e-06,
+      "loss": 0.463,
+      "step": 10079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8736685269958002,
+      "learning_rate": 1.6654305221999035e-06,
+      "loss": 0.465,
+      "step": 10080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.006200032644295,
+      "learning_rate": 1.663976147788806e-06,
+      "loss": 0.4743,
+      "step": 10081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0545813247664437,
+      "learning_rate": 1.6625223510567667e-06,
+      "loss": 0.5324,
+      "step": 10082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0119780993174157,
+      "learning_rate": 1.6610691321045325e-06,
+      "loss": 0.5058,
+      "step": 10083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9609372264458037,
+      "learning_rate": 1.6596164910328106e-06,
+      "loss": 0.5189,
+      "step": 10084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.937127977991446,
+      "learning_rate": 1.6581644279422705e-06,
+      "loss": 0.4489,
+      "step": 10085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9913881125396772,
+      "learning_rate": 1.6567129429335316e-06,
+      "loss": 0.4661,
+      "step": 10086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9612288385089394,
+      "learning_rate": 1.6552620361071903e-06,
+      "loss": 0.432,
+      "step": 10087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.895201885032304,
+      "learning_rate": 1.6538117075637849e-06,
+      "loss": 0.4727,
+      "step": 10088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8722790693420445,
+      "learning_rate": 1.6523619574038264e-06,
+      "loss": 0.4433,
+      "step": 10089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8694586279318163,
+      "learning_rate": 1.6509127857277784e-06,
+      "loss": 0.4481,
+      "step": 10090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9235188229616803,
+      "learning_rate": 1.6494641926360722e-06,
+      "loss": 0.4391,
+      "step": 10091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.035054049678654,
+      "learning_rate": 1.6480161782290849e-06,
+      "loss": 0.5072,
+      "step": 10092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9909914510220467,
+      "learning_rate": 1.6465687426071741e-06,
+      "loss": 0.5084,
+      "step": 10093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9044075400050785,
+      "learning_rate": 1.6451218858706374e-06,
+      "loss": 0.4573,
+      "step": 10094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9300078914155345,
+      "learning_rate": 1.6436756081197426e-06,
+      "loss": 0.5399,
+      "step": 10095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9039526767115124,
+      "learning_rate": 1.6422299094547156e-06,
+      "loss": 0.4677,
+      "step": 10096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9784978778245498,
+      "learning_rate": 1.6407847899757468e-06,
+      "loss": 0.4883,
+      "step": 10097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0152703015299072,
+      "learning_rate": 1.6393402497829713e-06,
+      "loss": 0.5826,
+      "step": 10098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9145163251531081,
+      "learning_rate": 1.6378962889765048e-06,
+      "loss": 0.4215,
+      "step": 10099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8961781171614732,
+      "learning_rate": 1.6364529076564072e-06,
+      "loss": 0.4545,
+      "step": 10100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9460249075728607,
+      "learning_rate": 1.635010105922704e-06,
+      "loss": 0.5136,
+      "step": 10101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9431549507296044,
+      "learning_rate": 1.633567883875381e-06,
+      "loss": 0.4878,
+      "step": 10102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.966938572007313,
+      "learning_rate": 1.6321262416143856e-06,
+      "loss": 0.555,
+      "step": 10103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9395240377305119,
+      "learning_rate": 1.6306851792396138e-06,
+      "loss": 0.4301,
+      "step": 10104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9624464563853393,
+      "learning_rate": 1.62924469685094e-06,
+      "loss": 0.52,
+      "step": 10105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8621026113911863,
+      "learning_rate": 1.6278047945481823e-06,
+      "loss": 0.4814,
+      "step": 10106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8405037196377936,
+      "learning_rate": 1.626365472431125e-06,
+      "loss": 0.4594,
+      "step": 10107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9767441682840015,
+      "learning_rate": 1.6249267305995141e-06,
+      "loss": 0.496,
+      "step": 10108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8078450200011851,
+      "learning_rate": 1.6234885691530543e-06,
+      "loss": 0.4337,
+      "step": 10109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9512121505034593,
+      "learning_rate": 1.6220509881914015e-06,
+      "loss": 0.489,
+      "step": 10110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9531062549342812,
+      "learning_rate": 1.620613987814189e-06,
+      "loss": 0.5192,
+      "step": 10111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9742192433408755,
+      "learning_rate": 1.6191775681209932e-06,
+      "loss": 0.4612,
+      "step": 10112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.922332170634465,
+      "learning_rate": 1.6177417292113572e-06,
+      "loss": 0.4994,
+      "step": 10113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8630218836631168,
+      "learning_rate": 1.6163064711847842e-06,
+      "loss": 0.4881,
+      "step": 10114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0168371457320893,
+      "learning_rate": 1.6148717941407387e-06,
+      "loss": 0.4818,
+      "step": 10115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0180019521693955,
+      "learning_rate": 1.613437698178636e-06,
+      "loss": 0.4638,
+      "step": 10116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9034584910700775,
+      "learning_rate": 1.6120041833978662e-06,
+      "loss": 0.4181,
+      "step": 10117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.026478894740036,
+      "learning_rate": 1.6105712498977644e-06,
+      "loss": 0.5086,
+      "step": 10118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9523384413831647,
+      "learning_rate": 1.6091388977776334e-06,
+      "loss": 0.4532,
+      "step": 10119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8911241632838921,
+      "learning_rate": 1.607707127136734e-06,
+      "loss": 0.4532,
+      "step": 10120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9356572273897498,
+      "learning_rate": 1.6062759380742898e-06,
+      "loss": 0.4329,
+      "step": 10121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9142388310791838,
+      "learning_rate": 1.6048453306894719e-06,
+      "loss": 0.4623,
+      "step": 10122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8648947239779669,
+      "learning_rate": 1.6034153050814315e-06,
+      "loss": 0.4531,
+      "step": 10123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9380175538325769,
+      "learning_rate": 1.601985861349261e-06,
+      "loss": 0.5184,
+      "step": 10124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9619274961680155,
+      "learning_rate": 1.60055699959202e-06,
+      "loss": 0.5017,
+      "step": 10125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0598349113443915,
+      "learning_rate": 1.599128719908729e-06,
+      "loss": 0.5305,
+      "step": 10126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0604631061633714,
+      "learning_rate": 1.5977010223983692e-06,
+      "loss": 0.5205,
+      "step": 10127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.932318399473858,
+      "learning_rate": 1.5962739071598709e-06,
+      "loss": 0.5136,
+      "step": 10128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0945709533633685,
+      "learning_rate": 1.594847374292141e-06,
+      "loss": 0.4599,
+      "step": 10129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9370398617218378,
+      "learning_rate": 1.5934214238940282e-06,
+      "loss": 0.4661,
+      "step": 10130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9230957099531405,
+      "learning_rate": 1.5919960560643589e-06,
+      "loss": 0.4422,
+      "step": 10131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9601615361009249,
+      "learning_rate": 1.5905712709019017e-06,
+      "loss": 0.5276,
+      "step": 10132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8953985465198216,
+      "learning_rate": 1.589147068505398e-06,
+      "loss": 0.4196,
+      "step": 10133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8861250406008324,
+      "learning_rate": 1.5877234489735405e-06,
+      "loss": 0.4784,
+      "step": 10134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8392301724786144,
+      "learning_rate": 1.5863004124049897e-06,
+      "loss": 0.4503,
+      "step": 10135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8732452868340083,
+      "learning_rate": 1.584877958898352e-06,
+      "loss": 0.476,
+      "step": 10136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0485661019215877,
+      "learning_rate": 1.583456088552212e-06,
+      "loss": 0.4975,
+      "step": 10137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9059777099372927,
+      "learning_rate": 1.5820348014650977e-06,
+      "loss": 0.5057,
+      "step": 10138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.975352945782583,
+      "learning_rate": 1.5806140977355056e-06,
+      "loss": 0.5405,
+      "step": 10139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9568187340392101,
+      "learning_rate": 1.579193977461887e-06,
+      "loss": 0.4768,
+      "step": 10140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8368006214403227,
+      "learning_rate": 1.5777744407426598e-06,
+      "loss": 0.4269,
+      "step": 10141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9323506413597026,
+      "learning_rate": 1.5763554876761888e-06,
+      "loss": 0.4785,
+      "step": 10142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0340089929500242,
+      "learning_rate": 1.5749371183608154e-06,
+      "loss": 0.4363,
+      "step": 10143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8927867883701817,
+      "learning_rate": 1.573519332894824e-06,
+      "loss": 0.4133,
+      "step": 10144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8767356770638337,
+      "learning_rate": 1.5721021313764684e-06,
+      "loss": 0.5238,
+      "step": 10145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9740716993726325,
+      "learning_rate": 1.57068551390396e-06,
+      "loss": 0.5013,
+      "step": 10146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8677354945105021,
+      "learning_rate": 1.5692694805754716e-06,
+      "loss": 0.4156,
+      "step": 10147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.031162555983234,
+      "learning_rate": 1.5678540314891243e-06,
+      "loss": 0.5158,
+      "step": 10148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9945018198037907,
+      "learning_rate": 1.566439166743019e-06,
+      "loss": 0.5158,
+      "step": 10149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0414349625809685,
+      "learning_rate": 1.5650248864351957e-06,
+      "loss": 0.5062,
+      "step": 10150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9229593872417223,
+      "learning_rate": 1.5636111906636665e-06,
+      "loss": 0.4811,
+      "step": 10151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9315096534678732,
+      "learning_rate": 1.5621980795263981e-06,
+      "loss": 0.4762,
+      "step": 10152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9962588003282389,
+      "learning_rate": 1.560785553121319e-06,
+      "loss": 0.5632,
+      "step": 10153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.001513498736759,
+      "learning_rate": 1.5593736115463154e-06,
+      "loss": 0.4853,
+      "step": 10154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9768955139754243,
+      "learning_rate": 1.5579622548992356e-06,
+      "loss": 0.4419,
+      "step": 10155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8768967199113278,
+      "learning_rate": 1.5565514832778816e-06,
+      "loss": 0.4644,
+      "step": 10156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8558221097469616,
+      "learning_rate": 1.5551412967800206e-06,
+      "loss": 0.4429,
+      "step": 10157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.881466729197389,
+      "learning_rate": 1.5537316955033766e-06,
+      "loss": 0.432,
+      "step": 10158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.051284596756194,
+      "learning_rate": 1.5523226795456349e-06,
+      "loss": 0.5281,
+      "step": 10159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9155417122416348,
+      "learning_rate": 1.5509142490044382e-06,
+      "loss": 0.4846,
+      "step": 10160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0411739639118625,
+      "learning_rate": 1.5495064039773921e-06,
+      "loss": 0.5852,
+      "step": 10161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8887895632464202,
+      "learning_rate": 1.5480991445620541e-06,
+      "loss": 0.5016,
+      "step": 10162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9360948530755474,
+      "learning_rate": 1.5466924708559483e-06,
+      "loss": 0.4885,
+      "step": 10163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.945720706242638,
+      "learning_rate": 1.5452863829565568e-06,
+      "loss": 0.5147,
+      "step": 10164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9668172070220067,
+      "learning_rate": 1.5438808809613193e-06,
+      "loss": 0.5072,
+      "step": 10165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8863857853165308,
+      "learning_rate": 1.5424759649676357e-06,
+      "loss": 0.5039,
+      "step": 10166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9821899836139799,
+      "learning_rate": 1.5410716350728671e-06,
+      "loss": 0.5183,
+      "step": 10167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.869172883665788,
+      "learning_rate": 1.5396678913743324e-06,
+      "loss": 0.4712,
+      "step": 10168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8392357457508518,
+      "learning_rate": 1.5382647339693068e-06,
+      "loss": 0.4401,
+      "step": 10169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.887219175964945,
+      "learning_rate": 1.5368621629550295e-06,
+      "loss": 0.4847,
+      "step": 10170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8524345364458238,
+      "learning_rate": 1.535460178428697e-06,
+      "loss": 0.4431,
+      "step": 10171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9630643844671861,
+      "learning_rate": 1.5340587804874662e-06,
+      "loss": 0.4392,
+      "step": 10172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9904964289364382,
+      "learning_rate": 1.5326579692284537e-06,
+      "loss": 0.4567,
+      "step": 10173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9282806035313729,
+      "learning_rate": 1.5312577447487342e-06,
+      "loss": 0.5784,
+      "step": 10174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0558418299723358,
+      "learning_rate": 1.52985810714534e-06,
+      "loss": 0.4927,
+      "step": 10175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1063856478240854,
+      "learning_rate": 1.5284590565152658e-06,
+      "loss": 0.4835,
+      "step": 10176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9117316773353915,
+      "learning_rate": 1.527060592955464e-06,
+      "loss": 0.5542,
+      "step": 10177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8713169752157789,
+      "learning_rate": 1.525662716562849e-06,
+      "loss": 0.4683,
+      "step": 10178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9073438743697106,
+      "learning_rate": 1.5242654274342895e-06,
+      "loss": 0.4708,
+      "step": 10179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8966897278534155,
+      "learning_rate": 1.5228687256666209e-06,
+      "loss": 0.4603,
+      "step": 10180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9327987256824378,
+      "learning_rate": 1.521472611356628e-06,
+      "loss": 0.5207,
+      "step": 10181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9177579625453101,
+      "learning_rate": 1.5200770846010626e-06,
+      "loss": 0.4408,
+      "step": 10182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8501020411606455,
+      "learning_rate": 1.518682145496634e-06,
+      "loss": 0.4893,
+      "step": 10183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9254580734831974,
+      "learning_rate": 1.51728779414001e-06,
+      "loss": 0.4761,
+      "step": 10184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8867520160877831,
+      "learning_rate": 1.515894030627817e-06,
+      "loss": 0.4256,
+      "step": 10185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8189185529115245,
+      "learning_rate": 1.5145008550566454e-06,
+      "loss": 0.4715,
+      "step": 10186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0093508277724619,
+      "learning_rate": 1.5131082675230325e-06,
+      "loss": 0.5081,
+      "step": 10187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9998210644300601,
+      "learning_rate": 1.5117162681234932e-06,
+      "loss": 0.4606,
+      "step": 10188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9591548783203413,
+      "learning_rate": 1.510324856954486e-06,
+      "loss": 0.4548,
+      "step": 10189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8416750074760201,
+      "learning_rate": 1.5089340341124348e-06,
+      "loss": 0.4433,
+      "step": 10190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8237614999712479,
+      "learning_rate": 1.5075437996937248e-06,
+      "loss": 0.458,
+      "step": 10191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8947058468112148,
+      "learning_rate": 1.5061541537946979e-06,
+      "loss": 0.4556,
+      "step": 10192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8301616530169087,
+      "learning_rate": 1.50476509651165e-06,
+      "loss": 0.4465,
+      "step": 10193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9188082792789959,
+      "learning_rate": 1.5033766279408502e-06,
+      "loss": 0.4811,
+      "step": 10194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0092638616018894,
+      "learning_rate": 1.5019887481785112e-06,
+      "loss": 0.5283,
+      "step": 10195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9427040337991623,
+      "learning_rate": 1.500601457320814e-06,
+      "loss": 0.4537,
+      "step": 10196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9217148432993036,
+      "learning_rate": 1.499214755463898e-06,
+      "loss": 0.4822,
+      "step": 10197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0911851164229525,
+      "learning_rate": 1.4978286427038602e-06,
+      "loss": 0.6013,
+      "step": 10198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9671555249802585,
+      "learning_rate": 1.4964431191367524e-06,
+      "loss": 0.4105,
+      "step": 10199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8743030992042353,
+      "learning_rate": 1.4950581848585977e-06,
+      "loss": 0.4241,
+      "step": 10200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.88143266868202,
+      "learning_rate": 1.493673839965365e-06,
+      "loss": 0.4976,
+      "step": 10201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0492592623073465,
+      "learning_rate": 1.4922900845529898e-06,
+      "loss": 0.4855,
+      "step": 10202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9898697168016961,
+      "learning_rate": 1.4909069187173652e-06,
+      "loss": 0.4903,
+      "step": 10203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8645242451171236,
+      "learning_rate": 1.4895243425543459e-06,
+      "loss": 0.4411,
+      "step": 10204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.877275663461517,
+      "learning_rate": 1.4881423561597374e-06,
+      "loss": 0.4829,
+      "step": 10205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8964463282505074,
+      "learning_rate": 1.4867609596293165e-06,
+      "loss": 0.4686,
+      "step": 10206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0755508778310003,
+      "learning_rate": 1.4853801530588085e-06,
+      "loss": 0.5166,
+      "step": 10207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9564994262029834,
+      "learning_rate": 1.4839999365439039e-06,
+      "loss": 0.4424,
+      "step": 10208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9465572798859154,
+      "learning_rate": 1.4826203101802494e-06,
+      "loss": 0.4262,
+      "step": 10209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8966023133247082,
+      "learning_rate": 1.4812412740634552e-06,
+      "loss": 0.4824,
+      "step": 10210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8745563853905437,
+      "learning_rate": 1.4798628282890793e-06,
+      "loss": 0.4788,
+      "step": 10211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0027440224193567,
+      "learning_rate": 1.4784849729526573e-06,
+      "loss": 0.5189,
+      "step": 10212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8894999547435524,
+      "learning_rate": 1.4771077081496654e-06,
+      "loss": 0.4733,
+      "step": 10213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8973434912002435,
+      "learning_rate": 1.4757310339755494e-06,
+      "loss": 0.4915,
+      "step": 10214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9867159710071842,
+      "learning_rate": 1.4743549505257126e-06,
+      "loss": 0.4814,
+      "step": 10215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9946006345156738,
+      "learning_rate": 1.472979457895517e-06,
+      "loss": 0.5211,
+      "step": 10216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9379457171519671,
+      "learning_rate": 1.4716045561802772e-06,
+      "loss": 0.4925,
+      "step": 10217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8979517459470215,
+      "learning_rate": 1.4702302454752815e-06,
+      "loss": 0.4941,
+      "step": 10218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.018496605305982,
+      "learning_rate": 1.4688565258757615e-06,
+      "loss": 0.4438,
+      "step": 10219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9519939488958427,
+      "learning_rate": 1.4674833974769166e-06,
+      "loss": 0.4612,
+      "step": 10220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8882005768328872,
+      "learning_rate": 1.466110860373905e-06,
+      "loss": 0.428,
+      "step": 10221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9357707331145971,
+      "learning_rate": 1.464738914661843e-06,
+      "loss": 0.5477,
+      "step": 10222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0508828769107081,
+      "learning_rate": 1.4633675604357988e-06,
+      "loss": 0.5031,
+      "step": 10223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0270950208906446,
+      "learning_rate": 1.4619967977908157e-06,
+      "loss": 0.4976,
+      "step": 10224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.987460185788436,
+      "learning_rate": 1.4606266268218783e-06,
+      "loss": 0.5491,
+      "step": 10225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9871122718704699,
+      "learning_rate": 1.4592570476239421e-06,
+      "loss": 0.4953,
+      "step": 10226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9820611895502117,
+      "learning_rate": 1.4578880602919165e-06,
+      "loss": 0.5188,
+      "step": 10227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9576539109395478,
+      "learning_rate": 1.4565196649206737e-06,
+      "loss": 0.5063,
+      "step": 10228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8461603678292988,
+      "learning_rate": 1.4551518616050352e-06,
+      "loss": 0.4568,
+      "step": 10229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9297742997758425,
+      "learning_rate": 1.453784650439798e-06,
+      "loss": 0.5646,
+      "step": 10230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9075481777747914,
+      "learning_rate": 1.4524180315197023e-06,
+      "loss": 0.4459,
+      "step": 10231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9125842697656487,
+      "learning_rate": 1.451052004939455e-06,
+      "loss": 0.4746,
+      "step": 10232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.862288669193609,
+      "learning_rate": 1.4496865707937201e-06,
+      "loss": 0.4309,
+      "step": 10233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9072812931637467,
+      "learning_rate": 1.4483217291771257e-06,
+      "loss": 0.4439,
+      "step": 10234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8165505502618535,
+      "learning_rate": 1.4469574801842445e-06,
+      "loss": 0.4725,
+      "step": 10235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.869627256256952,
+      "learning_rate": 1.445593823909628e-06,
+      "loss": 0.4678,
+      "step": 10236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9151138992538269,
+      "learning_rate": 1.444230760447769e-06,
+      "loss": 0.4892,
+      "step": 10237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2319204202974274,
+      "learning_rate": 1.44286828989313e-06,
+      "loss": 0.4873,
+      "step": 10238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9497741178270789,
+      "learning_rate": 1.441506412340129e-06,
+      "loss": 0.482,
+      "step": 10239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8981045686296287,
+      "learning_rate": 1.4401451278831435e-06,
+      "loss": 0.4408,
+      "step": 10240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9088762480381548,
+      "learning_rate": 1.4387844366165038e-06,
+      "loss": 0.5165,
+      "step": 10241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8337177904896431,
+      "learning_rate": 1.4374243386345132e-06,
+      "loss": 0.4659,
+      "step": 10242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0151898625090054,
+      "learning_rate": 1.4360648340314188e-06,
+      "loss": 0.5337,
+      "step": 10243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0033732105045368,
+      "learning_rate": 1.4347059229014359e-06,
+      "loss": 0.5321,
+      "step": 10244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9061509243075366,
+      "learning_rate": 1.433347605338734e-06,
+      "loss": 0.4419,
+      "step": 10245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0150086633032052,
+      "learning_rate": 1.4319898814374477e-06,
+      "loss": 0.4895,
+      "step": 10246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9799337144189451,
+      "learning_rate": 1.4306327512916574e-06,
+      "loss": 0.5107,
+      "step": 10247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9820232899095072,
+      "learning_rate": 1.429276214995421e-06,
+      "loss": 0.5189,
+      "step": 10248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1088600675559581,
+      "learning_rate": 1.4279202726427387e-06,
+      "loss": 0.5836,
+      "step": 10249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0022309856003913,
+      "learning_rate": 1.4265649243275782e-06,
+      "loss": 0.4901,
+      "step": 10250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8821887053125698,
+      "learning_rate": 1.4252101701438636e-06,
+      "loss": 0.4594,
+      "step": 10251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0464708818785635,
+      "learning_rate": 1.4238560101854815e-06,
+      "loss": 0.5232,
+      "step": 10252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8163985577430741,
+      "learning_rate": 1.4225024445462654e-06,
+      "loss": 0.4529,
+      "step": 10253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9243481894091906,
+      "learning_rate": 1.421149473320026e-06,
+      "loss": 0.4898,
+      "step": 10254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8764138643875491,
+      "learning_rate": 1.4197970966005148e-06,
+      "loss": 0.4318,
+      "step": 10255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.92803108348694,
+      "learning_rate": 1.418445314481458e-06,
+      "loss": 0.5064,
+      "step": 10256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8426680871851131,
+      "learning_rate": 1.4170941270565275e-06,
+      "loss": 0.4832,
+      "step": 10257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9808605193966883,
+      "learning_rate": 1.4157435344193605e-06,
+      "loss": 0.4438,
+      "step": 10258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9243613521277586,
+      "learning_rate": 1.4143935366635531e-06,
+      "loss": 0.4952,
+      "step": 10259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9816508493061582,
+      "learning_rate": 1.4130441338826595e-06,
+      "loss": 0.454,
+      "step": 10260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9007035804833416,
+      "learning_rate": 1.411695326170187e-06,
+      "loss": 0.4572,
+      "step": 10261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9111091371833596,
+      "learning_rate": 1.4103471136196145e-06,
+      "loss": 0.4367,
+      "step": 10262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8489740540487186,
+      "learning_rate": 1.4089994963243658e-06,
+      "loss": 0.3963,
+      "step": 10263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9519934832289844,
+      "learning_rate": 1.407652474377832e-06,
+      "loss": 0.5052,
+      "step": 10264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9604546986315811,
+      "learning_rate": 1.4063060478733604e-06,
+      "loss": 0.4527,
+      "step": 10265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8780119658630006,
+      "learning_rate": 1.4049602169042598e-06,
+      "loss": 0.4459,
+      "step": 10266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8712225686758214,
+      "learning_rate": 1.4036149815637866e-06,
+      "loss": 0.4109,
+      "step": 10267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8832135456375555,
+      "learning_rate": 1.4022703419451755e-06,
+      "loss": 0.4775,
+      "step": 10268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8817635322244807,
+      "learning_rate": 1.4009262981416016e-06,
+      "loss": 0.4529,
+      "step": 10269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.978087555119488,
+      "learning_rate": 1.3995828502462072e-06,
+      "loss": 0.4929,
+      "step": 10270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9898534652973374,
+      "learning_rate": 1.3982399983520934e-06,
+      "loss": 0.4752,
+      "step": 10271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8987085477583058,
+      "learning_rate": 1.39689774255232e-06,
+      "loss": 0.4729,
+      "step": 10272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9777419186698381,
+      "learning_rate": 1.3955560829398974e-06,
+      "loss": 0.5074,
+      "step": 10273
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9044479183395163,
+      "learning_rate": 1.3942150196078108e-06,
+      "loss": 0.4735,
+      "step": 10274
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.955084130118838,
+      "learning_rate": 1.3928745526489874e-06,
+      "loss": 0.4806,
+      "step": 10275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0172347146907044,
+      "learning_rate": 1.3915346821563235e-06,
+      "loss": 0.5519,
+      "step": 10276
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.084576363094847,
+      "learning_rate": 1.3901954082226698e-06,
+      "loss": 0.4468,
+      "step": 10277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.952157339824633,
+      "learning_rate": 1.3888567309408396e-06,
+      "loss": 0.5082,
+      "step": 10278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8740904269122499,
+      "learning_rate": 1.3875186504035965e-06,
+      "loss": 0.5169,
+      "step": 10279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9403765835118713,
+      "learning_rate": 1.386181166703675e-06,
+      "loss": 0.4676,
+      "step": 10280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8786768842079513,
+      "learning_rate": 1.384844279933757e-06,
+      "loss": 0.4222,
+      "step": 10281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9925467422216124,
+      "learning_rate": 1.3835079901864878e-06,
+      "loss": 0.5285,
+      "step": 10282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9809311190484911,
+      "learning_rate": 1.3821722975544727e-06,
+      "loss": 0.5496,
+      "step": 10283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9394321489302069,
+      "learning_rate": 1.3808372021302752e-06,
+      "loss": 0.4923,
+      "step": 10284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9952986041724703,
+      "learning_rate": 1.37950270400641e-06,
+      "loss": 0.493,
+      "step": 10285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.934921142448034,
+      "learning_rate": 1.378168803275366e-06,
+      "loss": 0.4647,
+      "step": 10286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9848679192787415,
+      "learning_rate": 1.376835500029573e-06,
+      "loss": 0.4982,
+      "step": 10287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9792804880817738,
+      "learning_rate": 1.375502794361432e-06,
+      "loss": 0.5542,
+      "step": 10288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9281471341149863,
+      "learning_rate": 1.3741706863632976e-06,
+      "loss": 0.5077,
+      "step": 10289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9541313032337133,
+      "learning_rate": 1.372839176127485e-06,
+      "loss": 0.4716,
+      "step": 10290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8903951337293582,
+      "learning_rate": 1.3715082637462607e-06,
+      "loss": 0.514,
+      "step": 10291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.926295498225597,
+      "learning_rate": 1.370177949311866e-06,
+      "loss": 0.4976,
+      "step": 10292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8380071746620592,
+      "learning_rate": 1.368848232916481e-06,
+      "loss": 0.4658,
+      "step": 10293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9757498423299027,
+      "learning_rate": 1.3675191146522593e-06,
+      "loss": 0.5021,
+      "step": 10294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8970210924053779,
+      "learning_rate": 1.366190594611304e-06,
+      "loss": 0.4657,
+      "step": 10295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8435500538474133,
+      "learning_rate": 1.3648626728856862e-06,
+      "loss": 0.4304,
+      "step": 10296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8889349239109522,
+      "learning_rate": 1.3635353495674208e-06,
+      "loss": 0.5115,
+      "step": 10297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8356644523131306,
+      "learning_rate": 1.3622086247484989e-06,
+      "loss": 0.4342,
+      "step": 10298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8973942966517006,
+      "learning_rate": 1.3608824985208569e-06,
+      "loss": 0.4747,
+      "step": 10299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9808056088548277,
+      "learning_rate": 1.3595569709763934e-06,
+      "loss": 0.4904,
+      "step": 10300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9684079981307847,
+      "learning_rate": 1.3582320422069684e-06,
+      "loss": 0.4608,
+      "step": 10301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9465905249500525,
+      "learning_rate": 1.3569077123043973e-06,
+      "loss": 0.4564,
+      "step": 10302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9502539835566461,
+      "learning_rate": 1.3555839813604555e-06,
+      "loss": 0.5308,
+      "step": 10303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8392779707666963,
+      "learning_rate": 1.3542608494668785e-06,
+      "loss": 0.4489,
+      "step": 10304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9518759760255114,
+      "learning_rate": 1.3529383167153543e-06,
+      "loss": 0.5635,
+      "step": 10305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9503279328972887,
+      "learning_rate": 1.3516163831975337e-06,
+      "loss": 0.446,
+      "step": 10306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0188594497418193,
+      "learning_rate": 1.350295049005027e-06,
+      "loss": 0.5087,
+      "step": 10307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9222776847545859,
+      "learning_rate": 1.348974314229401e-06,
+      "loss": 0.4797,
+      "step": 10308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8973687429402823,
+      "learning_rate": 1.3476541789621822e-06,
+      "loss": 0.4761,
+      "step": 10309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.924539907796562,
+      "learning_rate": 1.3463346432948555e-06,
+      "loss": 0.4248,
+      "step": 10310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.913652622748875,
+      "learning_rate": 1.3450157073188608e-06,
+      "loss": 0.4537,
+      "step": 10311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9662074046160266,
+      "learning_rate": 1.3436973711256006e-06,
+      "loss": 0.4851,
+      "step": 10312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8981108949974423,
+      "learning_rate": 1.3423796348064343e-06,
+      "loss": 0.4598,
+      "step": 10313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9353563433744401,
+      "learning_rate": 1.34106249845268e-06,
+      "loss": 0.5112,
+      "step": 10314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9043064502707608,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.4663,
+      "step": 10315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0139357510706588,
+      "learning_rate": 1.338430026006471e-06,
+      "loss": 0.4875,
+      "step": 10316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.891505665915497,
+      "learning_rate": 1.337114690096446e-06,
+      "loss": 0.4168,
+      "step": 10317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3950374138336772,
+      "learning_rate": 1.3357999545166878e-06,
+      "loss": 0.5183,
+      "step": 10318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9195164034412875,
+      "learning_rate": 1.3344858193583076e-06,
+      "loss": 0.4681,
+      "step": 10319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9308858780333419,
+      "learning_rate": 1.333172284712373e-06,
+      "loss": 0.453,
+      "step": 10320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0416797789084544,
+      "learning_rate": 1.3318593506699129e-06,
+      "loss": 0.5003,
+      "step": 10321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.041236641456775,
+      "learning_rate": 1.3305470173219104e-06,
+      "loss": 0.5134,
+      "step": 10322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8013652571175578,
+      "learning_rate": 1.3292352847593115e-06,
+      "loss": 0.4064,
+      "step": 10323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9611643454044503,
+      "learning_rate": 1.3279241530730147e-06,
+      "loss": 0.4909,
+      "step": 10324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8897865375319298,
+      "learning_rate": 1.3266136223538827e-06,
+      "loss": 0.4491,
+      "step": 10325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9901942315736499,
+      "learning_rate": 1.325303692692732e-06,
+      "loss": 0.4868,
+      "step": 10326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9971852320774585,
+      "learning_rate": 1.323994364180342e-06,
+      "loss": 0.5157,
+      "step": 10327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8725406497712802,
+      "learning_rate": 1.322685636907447e-06,
+      "loss": 0.4606,
+      "step": 10328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0280428865888989,
+      "learning_rate": 1.321377510964742e-06,
+      "loss": 0.526,
+      "step": 10329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9565540414208218,
+      "learning_rate": 1.3200699864428757e-06,
+      "loss": 0.5046,
+      "step": 10330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0174627243449055,
+      "learning_rate": 1.31876306343246e-06,
+      "loss": 0.5204,
+      "step": 10331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9857465322789754,
+      "learning_rate": 1.3174567420240647e-06,
+      "loss": 0.4811,
+      "step": 10332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.938586286054605,
+      "learning_rate": 1.3161510223082152e-06,
+      "loss": 0.4589,
+      "step": 10333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0344397790703828,
+      "learning_rate": 1.314845904375397e-06,
+      "loss": 0.5439,
+      "step": 10334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9056043167338509,
+      "learning_rate": 1.3135413883160564e-06,
+      "loss": 0.4655,
+      "step": 10335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9769940959605952,
+      "learning_rate": 1.3122374742205878e-06,
+      "loss": 0.4396,
+      "step": 10336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9450823830694117,
+      "learning_rate": 1.3109341621793614e-06,
+      "loss": 0.4665,
+      "step": 10337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9852144192111556,
+      "learning_rate": 1.309631452282688e-06,
+      "loss": 0.5602,
+      "step": 10338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9717639282246133,
+      "learning_rate": 1.3083293446208467e-06,
+      "loss": 0.452,
+      "step": 10339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9232201932196542,
+      "learning_rate": 1.3070278392840718e-06,
+      "loss": 0.5083,
+      "step": 10340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9227017022484458,
+      "learning_rate": 1.305726936362559e-06,
+      "loss": 0.4331,
+      "step": 10341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0087015750858035,
+      "learning_rate": 1.3044266359464542e-06,
+      "loss": 0.5254,
+      "step": 10342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9414831130487048,
+      "learning_rate": 1.3031269381258737e-06,
+      "loss": 0.4517,
+      "step": 10343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9980339567237606,
+      "learning_rate": 1.3018278429908815e-06,
+      "loss": 0.5105,
+      "step": 10344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8688966377834313,
+      "learning_rate": 1.3005293506315042e-06,
+      "loss": 0.4611,
+      "step": 10345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.945465953881035,
+      "learning_rate": 1.2992314611377255e-06,
+      "loss": 0.5041,
+      "step": 10346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9456803456419439,
+      "learning_rate": 1.2979341745994922e-06,
+      "loss": 0.4881,
+      "step": 10347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9546926099393971,
+      "learning_rate": 1.296637491106697e-06,
+      "loss": 0.4886,
+      "step": 10348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0014373713572333,
+      "learning_rate": 1.295341410749208e-06,
+      "loss": 0.5031,
+      "step": 10349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9578354382404265,
+      "learning_rate": 1.2940459336168366e-06,
+      "loss": 0.4757,
+      "step": 10350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9225824045489499,
+      "learning_rate": 1.2927510597993598e-06,
+      "loss": 0.4755,
+      "step": 10351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0158769338187628,
+      "learning_rate": 1.2914567893865103e-06,
+      "loss": 0.5099,
+      "step": 10352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9259783971877905,
+      "learning_rate": 1.2901631224679844e-06,
+      "loss": 0.4483,
+      "step": 10353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0430204010462292,
+      "learning_rate": 1.2888700591334225e-06,
+      "loss": 0.5591,
+      "step": 10354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9358738496499883,
+      "learning_rate": 1.2875775994724448e-06,
+      "loss": 0.4656,
+      "step": 10355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.999405660521046,
+      "learning_rate": 1.2862857435746078e-06,
+      "loss": 0.5058,
+      "step": 10356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0669907720051155,
+      "learning_rate": 1.284994491529441e-06,
+      "loss": 0.5028,
+      "step": 10357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0632046162659259,
+      "learning_rate": 1.283703843426425e-06,
+      "loss": 0.4915,
+      "step": 10358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9965790286536484,
+      "learning_rate": 1.2824137993550033e-06,
+      "loss": 0.5406,
+      "step": 10359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9001904962554457,
+      "learning_rate": 1.2811243594045697e-06,
+      "loss": 0.5098,
+      "step": 10360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0280818029750178,
+      "learning_rate": 1.2798355236644876e-06,
+      "loss": 0.5532,
+      "step": 10361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.931773872431827,
+      "learning_rate": 1.278547292224067e-06,
+      "loss": 0.4716,
+      "step": 10362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8939526381099582,
+      "learning_rate": 1.2772596651725833e-06,
+      "loss": 0.4572,
+      "step": 10363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9181306232702845,
+      "learning_rate": 1.275972642599268e-06,
+      "loss": 0.4845,
+      "step": 10364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9712261562760381,
+      "learning_rate": 1.274686224593311e-06,
+      "loss": 0.4899,
+      "step": 10365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.011891784830918,
+      "learning_rate": 1.273400411243857e-06,
+      "loss": 0.4992,
+      "step": 10366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9281749249363781,
+      "learning_rate": 1.2721152026400174e-06,
+      "loss": 0.4891,
+      "step": 10367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0111059017453472,
+      "learning_rate": 1.2708305988708502e-06,
+      "loss": 0.4853,
+      "step": 10368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0195681428784356,
+      "learning_rate": 1.2695466000253798e-06,
+      "loss": 0.512,
+      "step": 10369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9951959453513336,
+      "learning_rate": 1.268263206192587e-06,
+      "loss": 0.5128,
+      "step": 10370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0410562492605302,
+      "learning_rate": 1.2669804174614097e-06,
+      "loss": 0.5366,
+      "step": 10371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8844639541632079,
+      "learning_rate": 1.2656982339207401e-06,
+      "loss": 0.4934,
+      "step": 10372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9036479544451441,
+      "learning_rate": 1.2644166556594396e-06,
+      "loss": 0.5045,
+      "step": 10373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0051379719537108,
+      "learning_rate": 1.2631356827663144e-06,
+      "loss": 0.5172,
+      "step": 10374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9048632205778424,
+      "learning_rate": 1.2618553153301361e-06,
+      "loss": 0.421,
+      "step": 10375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0617817877623916,
+      "learning_rate": 1.2605755534396347e-06,
+      "loss": 0.542,
+      "step": 10376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8293522861775061,
+      "learning_rate": 1.259296397183497e-06,
+      "loss": 0.427,
+      "step": 10377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0283502189905431,
+      "learning_rate": 1.2580178466503623e-06,
+      "loss": 0.5244,
+      "step": 10378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8610204224832774,
+      "learning_rate": 1.2567399019288406e-06,
+      "loss": 0.5077,
+      "step": 10379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9043700711845449,
+      "learning_rate": 1.2554625631074846e-06,
+      "loss": 0.4689,
+      "step": 10380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9362311134421188,
+      "learning_rate": 1.2541858302748199e-06,
+      "loss": 0.4184,
+      "step": 10381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9579535314078547,
+      "learning_rate": 1.2529097035193183e-06,
+      "loss": 0.5297,
+      "step": 10382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0046460874542875,
+      "learning_rate": 1.2516341829294155e-06,
+      "loss": 0.5359,
+      "step": 10383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.859887174800114,
+      "learning_rate": 1.2503592685935039e-06,
+      "loss": 0.3874,
+      "step": 10384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1239221563293567,
+      "learning_rate": 1.2490849605999355e-06,
+      "loss": 0.5216,
+      "step": 10385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8692722620485404,
+      "learning_rate": 1.2478112590370139e-06,
+      "loss": 0.3922,
+      "step": 10386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9322363876829223,
+      "learning_rate": 1.246538163993013e-06,
+      "loss": 0.4898,
+      "step": 10387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.944415903299271,
+      "learning_rate": 1.2452656755561509e-06,
+      "loss": 0.4815,
+      "step": 10388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.914130294902286,
+      "learning_rate": 1.2439937938146118e-06,
+      "loss": 0.4974,
+      "step": 10389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8514860899931579,
+      "learning_rate": 1.2427225188565362e-06,
+      "loss": 0.4952,
+      "step": 10390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9783298964863831,
+      "learning_rate": 1.2414518507700247e-06,
+      "loss": 0.5409,
+      "step": 10391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9345235574956751,
+      "learning_rate": 1.2401817896431268e-06,
+      "loss": 0.5215,
+      "step": 10392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9557235832363349,
+      "learning_rate": 1.2389123355638655e-06,
+      "loss": 0.4642,
+      "step": 10393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9882934130814738,
+      "learning_rate": 1.237643488620206e-06,
+      "loss": 0.4993,
+      "step": 10394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9229014262843889,
+      "learning_rate": 1.2363752489000802e-06,
+      "loss": 0.5369,
+      "step": 10395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8677797825145028,
+      "learning_rate": 1.2351076164913767e-06,
+      "loss": 0.5179,
+      "step": 10396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9079183428436072,
+      "learning_rate": 1.2338405914819428e-06,
+      "loss": 0.4528,
+      "step": 10397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0616692072686915,
+      "learning_rate": 1.2325741739595753e-06,
+      "loss": 0.4966,
+      "step": 10398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.017928690459014,
+      "learning_rate": 1.2313083640120461e-06,
+      "loss": 0.4986,
+      "step": 10399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0044277617256578,
+      "learning_rate": 1.2300431617270669e-06,
+      "loss": 0.5175,
+      "step": 10400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9618426399337734,
+      "learning_rate": 1.228778567192318e-06,
+      "loss": 0.5235,
+      "step": 10401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9374152649408706,
+      "learning_rate": 1.2275145804954347e-06,
+      "loss": 0.4962,
+      "step": 10402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8818312190067937,
+      "learning_rate": 1.2262512017240113e-06,
+      "loss": 0.4294,
+      "step": 10403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8987846617538389,
+      "learning_rate": 1.2249884309655935e-06,
+      "loss": 0.4726,
+      "step": 10404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8476644702669166,
+      "learning_rate": 1.2237262683076979e-06,
+      "loss": 0.4813,
+      "step": 10405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9203532221043269,
+      "learning_rate": 1.2224647138377854e-06,
+      "loss": 0.508,
+      "step": 10406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9444794075657026,
+      "learning_rate": 1.221203767643282e-06,
+      "loss": 0.494,
+      "step": 10407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9873844191944869,
+      "learning_rate": 1.219943429811571e-06,
+      "loss": 0.4555,
+      "step": 10408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9150066558016045,
+      "learning_rate": 1.2186837004299957e-06,
+      "loss": 0.4924,
+      "step": 10409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8881440228122434,
+      "learning_rate": 1.2174245795858454e-06,
+      "loss": 0.4637,
+      "step": 10410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9006863395728119,
+      "learning_rate": 1.2161660673663855e-06,
+      "loss": 0.4692,
+      "step": 10411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9597572812479213,
+      "learning_rate": 1.2149081638588246e-06,
+      "loss": 0.4673,
+      "step": 10412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8682495435123212,
+      "learning_rate": 1.2136508691503357e-06,
+      "loss": 0.4435,
+      "step": 10413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9486120778717507,
+      "learning_rate": 1.2123941833280472e-06,
+      "loss": 0.5288,
+      "step": 10414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9196492823660393,
+      "learning_rate": 1.2111381064790506e-06,
+      "loss": 0.5112,
+      "step": 10415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8973238884603776,
+      "learning_rate": 1.2098826386903829e-06,
+      "loss": 0.4586,
+      "step": 10416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9190498993137616,
+      "learning_rate": 1.2086277800490554e-06,
+      "loss": 0.4639,
+      "step": 10417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8949475889209494,
+      "learning_rate": 1.207373530642022e-06,
+      "loss": 0.4672,
+      "step": 10418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9286970525038918,
+      "learning_rate": 1.2061198905562043e-06,
+      "loss": 0.489,
+      "step": 10419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0214400088266034,
+      "learning_rate": 1.2048668598784785e-06,
+      "loss": 0.5502,
+      "step": 10420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9371816622979336,
+      "learning_rate": 1.2036144386956805e-06,
+      "loss": 0.5025,
+      "step": 10421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8994263459924541,
+      "learning_rate": 1.2023626270945943e-06,
+      "loss": 0.4472,
+      "step": 10422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0525060037546006,
+      "learning_rate": 1.2011114251619792e-06,
+      "loss": 0.5032,
+      "step": 10423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1654049372519806,
+      "learning_rate": 1.1998608329845362e-06,
+      "loss": 0.5165,
+      "step": 10424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0218906819426945,
+      "learning_rate": 1.1986108506489314e-06,
+      "loss": 0.5377,
+      "step": 10425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9387363153022205,
+      "learning_rate": 1.1973614782417874e-06,
+      "loss": 0.4903,
+      "step": 10426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8818776810322095,
+      "learning_rate": 1.1961127158496866e-06,
+      "loss": 0.4606,
+      "step": 10427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9745946305449849,
+      "learning_rate": 1.1948645635591627e-06,
+      "loss": 0.4892,
+      "step": 10428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1204945414201928,
+      "learning_rate": 1.1936170214567177e-06,
+      "loss": 0.5149,
+      "step": 10429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.000795057220358,
+      "learning_rate": 1.1923700896288004e-06,
+      "loss": 0.5303,
+      "step": 10430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9685116179114811,
+      "learning_rate": 1.1911237681618226e-06,
+      "loss": 0.4936,
+      "step": 10431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9088048618312514,
+      "learning_rate": 1.1898780571421554e-06,
+      "loss": 0.4575,
+      "step": 10432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8908139699595279,
+      "learning_rate": 1.1886329566561262e-06,
+      "loss": 0.4623,
+      "step": 10433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9185747240565809,
+      "learning_rate": 1.1873884667900125e-06,
+      "loss": 0.4527,
+      "step": 10434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0019140346220445,
+      "learning_rate": 1.186144587630066e-06,
+      "loss": 0.5332,
+      "step": 10435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9822650328652477,
+      "learning_rate": 1.184901319262479e-06,
+      "loss": 0.5219,
+      "step": 10436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9366252276807113,
+      "learning_rate": 1.1836586617734114e-06,
+      "loss": 0.4705,
+      "step": 10437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9925375074413082,
+      "learning_rate": 1.1824166152489791e-06,
+      "loss": 0.493,
+      "step": 10438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9522970269239825,
+      "learning_rate": 1.181175179775257e-06,
+      "loss": 0.4728,
+      "step": 10439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9604610233217911,
+      "learning_rate": 1.179934355438267e-06,
+      "loss": 0.4928,
+      "step": 10440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8529289245362368,
+      "learning_rate": 1.1786941423240072e-06,
+      "loss": 0.4278,
+      "step": 10441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0000596664112409,
+      "learning_rate": 1.1774545405184178e-06,
+      "loss": 0.4782,
+      "step": 10442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9101942517684707,
+      "learning_rate": 1.1762155501074024e-06,
+      "loss": 0.5051,
+      "step": 10443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9253558360388284,
+      "learning_rate": 1.1749771711768233e-06,
+      "loss": 0.492,
+      "step": 10444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.919619505554477,
+      "learning_rate": 1.1737394038124994e-06,
+      "loss": 0.4892,
+      "step": 10445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9461661472433754,
+      "learning_rate": 1.1725022481002024e-06,
+      "loss": 0.4437,
+      "step": 10446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9388303265636923,
+      "learning_rate": 1.1712657041256737e-06,
+      "loss": 0.4766,
+      "step": 10447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.983735419601325,
+      "learning_rate": 1.170029771974599e-06,
+      "loss": 0.5595,
+      "step": 10448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9821181424572664,
+      "learning_rate": 1.1687944517326289e-06,
+      "loss": 0.4663,
+      "step": 10449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9464613846937909,
+      "learning_rate": 1.1675597434853692e-06,
+      "loss": 0.4726,
+      "step": 10450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9347967181397874,
+      "learning_rate": 1.1663256473183858e-06,
+      "loss": 0.4676,
+      "step": 10451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1063597249501547,
+      "learning_rate": 1.1650921633171985e-06,
+      "loss": 0.5331,
+      "step": 10452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8829825883905474,
+      "learning_rate": 1.1638592915672908e-06,
+      "loss": 0.4304,
+      "step": 10453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0064095451786466,
+      "learning_rate": 1.1626270321540945e-06,
+      "loss": 0.497,
+      "step": 10454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9513480613132469,
+      "learning_rate": 1.1613953851630055e-06,
+      "loss": 0.483,
+      "step": 10455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9381733547839735,
+      "learning_rate": 1.160164350679377e-06,
+      "loss": 0.4916,
+      "step": 10456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9758864109716439,
+      "learning_rate": 1.158933928788518e-06,
+      "loss": 0.4989,
+      "step": 10457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9612788161193682,
+      "learning_rate": 1.1577041195756954e-06,
+      "loss": 0.4806,
+      "step": 10458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8620183622668911,
+      "learning_rate": 1.1564749231261364e-06,
+      "loss": 0.4454,
+      "step": 10459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.984063658349551,
+      "learning_rate": 1.155246339525019e-06,
+      "loss": 0.546,
+      "step": 10460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9123286243679032,
+      "learning_rate": 1.1540183688574847e-06,
+      "loss": 0.4465,
+      "step": 10461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9541839139858166,
+      "learning_rate": 1.1527910112086315e-06,
+      "loss": 0.4698,
+      "step": 10462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0870922549213533,
+      "learning_rate": 1.151564266663514e-06,
+      "loss": 0.5383,
+      "step": 10463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9692547053719215,
+      "learning_rate": 1.150338135307144e-06,
+      "loss": 0.4782,
+      "step": 10464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8367552594366106,
+      "learning_rate": 1.1491126172244915e-06,
+      "loss": 0.4565,
+      "step": 10465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9307612165354187,
+      "learning_rate": 1.147887712500486e-06,
+      "loss": 0.4579,
+      "step": 10466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8715305493419948,
+      "learning_rate": 1.1466634212200079e-06,
+      "loss": 0.4211,
+      "step": 10467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9695418907066585,
+      "learning_rate": 1.1454397434679022e-06,
+      "loss": 0.5061,
+      "step": 10468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9374443923373761,
+      "learning_rate": 1.1442166793289677e-06,
+      "loss": 0.566,
+      "step": 10469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9450478119481537,
+      "learning_rate": 1.1429942288879626e-06,
+      "loss": 0.5034,
+      "step": 10470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.983148186080234,
+      "learning_rate": 1.1417723922296008e-06,
+      "loss": 0.492,
+      "step": 10471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8598251224417236,
+      "learning_rate": 1.1405511694385584e-06,
+      "loss": 0.4438,
+      "step": 10472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9462950488597034,
+      "learning_rate": 1.1393305605994587e-06,
+      "loss": 0.4756,
+      "step": 10473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9482638073194297,
+      "learning_rate": 1.1381105657968916e-06,
+      "loss": 0.4996,
+      "step": 10474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.860704600578187,
+      "learning_rate": 1.1368911851154019e-06,
+      "loss": 0.5384,
+      "step": 10475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9570375760818484,
+      "learning_rate": 1.1356724186394918e-06,
+      "loss": 0.4809,
+      "step": 10476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9634767434215831,
+      "learning_rate": 1.1344542664536196e-06,
+      "loss": 0.5581,
+      "step": 10477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9795070369522804,
+      "learning_rate": 1.1332367286422064e-06,
+      "loss": 0.5486,
+      "step": 10478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8677691664078212,
+      "learning_rate": 1.1320198052896203e-06,
+      "loss": 0.474,
+      "step": 10479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9434054008110407,
+      "learning_rate": 1.130803496480195e-06,
+      "loss": 0.524,
+      "step": 10480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9553611331557014,
+      "learning_rate": 1.129587802298222e-06,
+      "loss": 0.4732,
+      "step": 10481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8139743634257981,
+      "learning_rate": 1.128372722827945e-06,
+      "loss": 0.4421,
+      "step": 10482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8881985728660753,
+      "learning_rate": 1.12715825815357e-06,
+      "loss": 0.4171,
+      "step": 10483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9326228528390237,
+      "learning_rate": 1.1259444083592585e-06,
+      "loss": 0.4373,
+      "step": 10484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.958442471726399,
+      "learning_rate": 1.1247311735291255e-06,
+      "loss": 0.4941,
+      "step": 10485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9091981400958642,
+      "learning_rate": 1.1235185537472537e-06,
+      "loss": 0.5158,
+      "step": 10486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9715206218825831,
+      "learning_rate": 1.1223065490976692e-06,
+      "loss": 0.5159,
+      "step": 10487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8629851659158863,
+      "learning_rate": 1.1210951596643682e-06,
+      "loss": 0.4602,
+      "step": 10488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9510654460373925,
+      "learning_rate": 1.1198843855312958e-06,
+      "loss": 0.5051,
+      "step": 10489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9736674492886307,
+      "learning_rate": 1.1186742267823614e-06,
+      "loss": 0.5066,
+      "step": 10490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.041070879920255,
+      "learning_rate": 1.1174646835014213e-06,
+      "loss": 0.5287,
+      "step": 10491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9975479878026994,
+      "learning_rate": 1.1162557557723042e-06,
+      "loss": 0.4458,
+      "step": 10492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9844734373597187,
+      "learning_rate": 1.1150474436787806e-06,
+      "loss": 0.5017,
+      "step": 10493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9716899197564739,
+      "learning_rate": 1.113839747304588e-06,
+      "loss": 0.4888,
+      "step": 10494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9370907891923271,
+      "learning_rate": 1.1126326667334196e-06,
+      "loss": 0.4597,
+      "step": 10495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9241490676999198,
+      "learning_rate": 1.1114262020489264e-06,
+      "loss": 0.456,
+      "step": 10496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8374254769318923,
+      "learning_rate": 1.1102203533347089e-06,
+      "loss": 0.4738,
+      "step": 10497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8877474626590348,
+      "learning_rate": 1.1090151206743393e-06,
+      "loss": 0.4866,
+      "step": 10498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9475470780816515,
+      "learning_rate": 1.1078105041513343e-06,
+      "loss": 0.4918,
+      "step": 10499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9275589409212524,
+      "learning_rate": 1.1066065038491735e-06,
+      "loss": 0.4926,
+      "step": 10500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8795569509194492,
+      "learning_rate": 1.1054031198512938e-06,
+      "loss": 0.4436,
+      "step": 10501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8691998892476426,
+      "learning_rate": 1.1042003522410882e-06,
+      "loss": 0.4589,
+      "step": 10502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9802431120077391,
+      "learning_rate": 1.102998201101908e-06,
+      "loss": 0.4924,
+      "step": 10503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9531451083089679,
+      "learning_rate": 1.1017966665170632e-06,
+      "loss": 0.4823,
+      "step": 10504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8918469153039225,
+      "learning_rate": 1.1005957485698115e-06,
+      "loss": 0.4249,
+      "step": 10505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8277939680529098,
+      "learning_rate": 1.0993954473433854e-06,
+      "loss": 0.474,
+      "step": 10506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9217263067232511,
+      "learning_rate": 1.0981957629209584e-06,
+      "loss": 0.4761,
+      "step": 10507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8490029541683193,
+      "learning_rate": 1.096996695385668e-06,
+      "loss": 0.4925,
+      "step": 10508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9751680603334909,
+      "learning_rate": 1.0957982448206105e-06,
+      "loss": 0.5118,
+      "step": 10509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9325847643538916,
+      "learning_rate": 1.0946004113088381e-06,
+      "loss": 0.4822,
+      "step": 10510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8462218962246554,
+      "learning_rate": 1.0934031949333546e-06,
+      "loss": 0.4943,
+      "step": 10511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8824927665558341,
+      "learning_rate": 1.0922065957771332e-06,
+      "loss": 0.4399,
+      "step": 10512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9460646657620165,
+      "learning_rate": 1.0910106139230913e-06,
+      "loss": 0.4845,
+      "step": 10513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.010466471062037,
+      "learning_rate": 1.0898152494541124e-06,
+      "loss": 0.5382,
+      "step": 10514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0629758525690058,
+      "learning_rate": 1.0886205024530327e-06,
+      "loss": 0.4585,
+      "step": 10515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9760959154365196,
+      "learning_rate": 1.0874263730026502e-06,
+      "loss": 0.4848,
+      "step": 10516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9352196398984858,
+      "learning_rate": 1.0862328611857109e-06,
+      "loss": 0.4978,
+      "step": 10517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9499481473322007,
+      "learning_rate": 1.085039967084931e-06,
+      "loss": 0.4877,
+      "step": 10518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8820445238550193,
+      "learning_rate": 1.083847690782972e-06,
+      "loss": 0.4984,
+      "step": 10519
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8875572183370747,
+      "learning_rate": 1.0826560323624591e-06,
+      "loss": 0.4319,
+      "step": 10520
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9909807008910421,
+      "learning_rate": 1.081464991905975e-06,
+      "loss": 0.4924,
+      "step": 10521
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.923524185885788,
+      "learning_rate": 1.080274569496057e-06,
+      "loss": 0.4533,
+      "step": 10522
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.933531295118856,
+      "learning_rate": 1.079084765215196e-06,
+      "loss": 0.4548,
+      "step": 10523
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9891970105285252,
+      "learning_rate": 1.0778955791458513e-06,
+      "loss": 0.4882,
+      "step": 10524
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9041791644495576,
+      "learning_rate": 1.076707011370427e-06,
+      "loss": 0.4721,
+      "step": 10525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8494593233155678,
+      "learning_rate": 1.075519061971293e-06,
+      "loss": 0.4562,
+      "step": 10526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9632722345680252,
+      "learning_rate": 1.074331731030771e-06,
+      "loss": 0.5007,
+      "step": 10527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9903562734018743,
+      "learning_rate": 1.0731450186311454e-06,
+      "loss": 0.5376,
+      "step": 10528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8708142022778427,
+      "learning_rate": 1.0719589248546469e-06,
+      "loss": 0.4663,
+      "step": 10529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9039712102106406,
+      "learning_rate": 1.070773449783481e-06,
+      "loss": 0.4545,
+      "step": 10530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9787575342806338,
+      "learning_rate": 1.069588593499793e-06,
+      "loss": 0.4533,
+      "step": 10531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.906339996163714,
+      "learning_rate": 1.0684043560856928e-06,
+      "loss": 0.4621,
+      "step": 10532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9995109838408539,
+      "learning_rate": 1.067220737623249e-06,
+      "loss": 0.4951,
+      "step": 10533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9337158432010438,
+      "learning_rate": 1.0660377381944876e-06,
+      "loss": 0.4803,
+      "step": 10534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8939043705730141,
+      "learning_rate": 1.0648553578813813e-06,
+      "loss": 0.5323,
+      "step": 10535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9014274053264499,
+      "learning_rate": 1.0636735967658785e-06,
+      "loss": 0.4608,
+      "step": 10536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9596380740959803,
+      "learning_rate": 1.0624924549298666e-06,
+      "loss": 0.5336,
+      "step": 10537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9184750871004159,
+      "learning_rate": 1.061311932455199e-06,
+      "loss": 0.472,
+      "step": 10538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9820976331082859,
+      "learning_rate": 1.0601320294236872e-06,
+      "loss": 0.4911,
+      "step": 10539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9766999400801963,
+      "learning_rate": 1.0589527459170967e-06,
+      "loss": 0.5862,
+      "step": 10540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9341904993019802,
+      "learning_rate": 1.0577740820171468e-06,
+      "loss": 0.4902,
+      "step": 10541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8466239721311413,
+      "learning_rate": 1.0565960378055263e-06,
+      "loss": 0.4735,
+      "step": 10542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9136573595418811,
+      "learning_rate": 1.0554186133638643e-06,
+      "loss": 0.4461,
+      "step": 10543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8845945755514436,
+      "learning_rate": 1.0542418087737593e-06,
+      "loss": 0.4808,
+      "step": 10544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9664680081992767,
+      "learning_rate": 1.0530656241167613e-06,
+      "loss": 0.4904,
+      "step": 10545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9369665727003559,
+      "learning_rate": 1.051890059474382e-06,
+      "loss": 0.4683,
+      "step": 10546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9122493649287006,
+      "learning_rate": 1.0507151149280804e-06,
+      "loss": 0.4651,
+      "step": 10547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9006968892963719,
+      "learning_rate": 1.049540790559288e-06,
+      "loss": 0.5234,
+      "step": 10548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.894069435907734,
+      "learning_rate": 1.0483670864493777e-06,
+      "loss": 0.4758,
+      "step": 10549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9225468427309412,
+      "learning_rate": 1.0471940026796878e-06,
+      "loss": 0.3915,
+      "step": 10550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9415756739526433,
+      "learning_rate": 1.046021539331512e-06,
+      "loss": 0.5204,
+      "step": 10551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9793688442119916,
+      "learning_rate": 1.0448496964861044e-06,
+      "loss": 0.443,
+      "step": 10552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9143015605359969,
+      "learning_rate": 1.0436784742246652e-06,
+      "loss": 0.44,
+      "step": 10553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9957353280803909,
+      "learning_rate": 1.0425078726283667e-06,
+      "loss": 0.5254,
+      "step": 10554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.095502882617657,
+      "learning_rate": 1.0413378917783267e-06,
+      "loss": 0.5872,
+      "step": 10555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.018574131899571,
+      "learning_rate": 1.0401685317556232e-06,
+      "loss": 0.5307,
+      "step": 10556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0398118534675804,
+      "learning_rate": 1.0389997926412942e-06,
+      "loss": 0.5394,
+      "step": 10557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9024083075890255,
+      "learning_rate": 1.037831674516332e-06,
+      "loss": 0.4886,
+      "step": 10558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9127674671846514,
+      "learning_rate": 1.0366641774616826e-06,
+      "loss": 0.504,
+      "step": 10559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8940640858916057,
+      "learning_rate": 1.0354973015582582e-06,
+      "loss": 0.4301,
+      "step": 10560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9477172972392047,
+      "learning_rate": 1.0343310468869171e-06,
+      "loss": 0.4708,
+      "step": 10561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9215397595796041,
+      "learning_rate": 1.033165413528483e-06,
+      "loss": 0.4648,
+      "step": 10562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9870791421156618,
+      "learning_rate": 1.0320004015637319e-06,
+      "loss": 0.4537,
+      "step": 10563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9182962662979826,
+      "learning_rate": 1.0308360110733994e-06,
+      "loss": 0.4726,
+      "step": 10564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9498092146289201,
+      "learning_rate": 1.0296722421381733e-06,
+      "loss": 0.4866,
+      "step": 10565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.003002994594343,
+      "learning_rate": 1.0285090948387065e-06,
+      "loss": 0.4608,
+      "step": 10566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8831435494424852,
+      "learning_rate": 1.0273465692556006e-06,
+      "loss": 0.4632,
+      "step": 10567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9964518745942484,
+      "learning_rate": 1.0261846654694184e-06,
+      "loss": 0.4706,
+      "step": 10568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8837425792652869,
+      "learning_rate": 1.0250233835606805e-06,
+      "loss": 0.4616,
+      "step": 10569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8938867215972366,
+      "learning_rate": 1.0238627236098619e-06,
+      "loss": 0.4166,
+      "step": 10570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8980332943128998,
+      "learning_rate": 1.0227026856973909e-06,
+      "loss": 0.5217,
+      "step": 10571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9194033443879248,
+      "learning_rate": 1.0215432699036643e-06,
+      "loss": 0.4979,
+      "step": 10572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0607340177473767,
+      "learning_rate": 1.0203844763090243e-06,
+      "loss": 0.4971,
+      "step": 10573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7714088041106679,
+      "learning_rate": 1.0192263049937745e-06,
+      "loss": 0.4175,
+      "step": 10574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8751825274589954,
+      "learning_rate": 1.0180687560381764e-06,
+      "loss": 0.4855,
+      "step": 10575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9285605238440512,
+      "learning_rate": 1.0169118295224488e-06,
+      "loss": 0.4911,
+      "step": 10576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8986266092966625,
+      "learning_rate": 1.0157555255267581e-06,
+      "loss": 0.4886,
+      "step": 10577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0638855808898067,
+      "learning_rate": 1.0145998441312455e-06,
+      "loss": 0.5261,
+      "step": 10578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9644109630050147,
+      "learning_rate": 1.0134447854159913e-06,
+      "loss": 0.4931,
+      "step": 10579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9806403030365124,
+      "learning_rate": 1.0122903494610426e-06,
+      "loss": 0.4277,
+      "step": 10580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9415593828325067,
+      "learning_rate": 1.011136536346401e-06,
+      "loss": 0.4543,
+      "step": 10581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8746620984571902,
+      "learning_rate": 1.009983346152026e-06,
+      "loss": 0.4319,
+      "step": 10582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9425651262758186,
+      "learning_rate": 1.0088307789578266e-06,
+      "loss": 0.4588,
+      "step": 10583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.001027698974047,
+      "learning_rate": 1.0076788348436827e-06,
+      "loss": 0.5128,
+      "step": 10584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9521648979712655,
+      "learning_rate": 1.0065275138894182e-06,
+      "loss": 0.4982,
+      "step": 10585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8767109127856868,
+      "learning_rate": 1.00537681617482e-06,
+      "loss": 0.4291,
+      "step": 10586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9233767299328738,
+      "learning_rate": 1.0042267417796292e-06,
+      "loss": 0.4564,
+      "step": 10587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8818885464968629,
+      "learning_rate": 1.0030772907835484e-06,
+      "loss": 0.4689,
+      "step": 10588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8781990150711255,
+      "learning_rate": 1.0019284632662274e-06,
+      "loss": 0.4965,
+      "step": 10589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9439405869978958,
+      "learning_rate": 1.000780259307287e-06,
+      "loss": 0.5321,
+      "step": 10590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0273924686655207,
+      "learning_rate": 9.996326789862897e-07,
+      "loss": 0.5068,
+      "step": 10591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9148259332012102,
+      "learning_rate": 9.984857223827637e-07,
+      "loss": 0.5091,
+      "step": 10592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9510707852522677,
+      "learning_rate": 9.97339389576194e-07,
+      "loss": 0.4678,
+      "step": 10593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9856242720669215,
+      "learning_rate": 9.961936806460194e-07,
+      "loss": 0.4869,
+      "step": 10594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9662211813822499,
+      "learning_rate": 9.950485956716349e-07,
+      "loss": 0.478,
+      "step": 10595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9995700843908029,
+      "learning_rate": 9.939041347323986e-07,
+      "loss": 0.4696,
+      "step": 10596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9763678246865335,
+      "learning_rate": 9.927602979076146e-07,
+      "loss": 0.496,
+      "step": 10597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9359566163419666,
+      "learning_rate": 9.91617085276554e-07,
+      "loss": 0.5372,
+      "step": 10598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9833137658610944,
+      "learning_rate": 9.904744969184377e-07,
+      "loss": 0.5355,
+      "step": 10599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8636732279612003,
+      "learning_rate": 9.89332532912447e-07,
+      "loss": 0.4137,
+      "step": 10600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.017749469597937,
+      "learning_rate": 9.881911933377197e-07,
+      "loss": 0.5332,
+      "step": 10601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9590708469565017,
+      "learning_rate": 9.870504782733515e-07,
+      "loss": 0.4546,
+      "step": 10602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.875187921212929,
+      "learning_rate": 9.85910387798389e-07,
+      "loss": 0.4611,
+      "step": 10603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9268857981738736,
+      "learning_rate": 9.8477092199184e-07,
+      "loss": 0.4842,
+      "step": 10604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9375699374447519,
+      "learning_rate": 9.836320809326704e-07,
+      "loss": 0.524,
+      "step": 10605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0500442909628693,
+      "learning_rate": 9.824938646998005e-07,
+      "loss": 0.5259,
+      "step": 10606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9722789226492576,
+      "learning_rate": 9.813562733721072e-07,
+      "loss": 0.4629,
+      "step": 10607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8146063568062981,
+      "learning_rate": 9.80219307028426e-07,
+      "loss": 0.4079,
+      "step": 10608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0053294621074196,
+      "learning_rate": 9.790829657475443e-07,
+      "loss": 0.5076,
+      "step": 10609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9424850866724473,
+      "learning_rate": 9.77947249608211e-07,
+      "loss": 0.4957,
+      "step": 10610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9350508643965163,
+      "learning_rate": 9.768121586891322e-07,
+      "loss": 0.5018,
+      "step": 10611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9163531937596598,
+      "learning_rate": 9.75677693068966e-07,
+      "loss": 0.4881,
+      "step": 10612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0296132540056735,
+      "learning_rate": 9.745438528263319e-07,
+      "loss": 0.5175,
+      "step": 10613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.907944549759089,
+      "learning_rate": 9.734106380398022e-07,
+      "loss": 0.4734,
+      "step": 10614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9589597429120368,
+      "learning_rate": 9.722780487879124e-07,
+      "loss": 0.4598,
+      "step": 10615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9015626629553082,
+      "learning_rate": 9.711460851491427e-07,
+      "loss": 0.4868,
+      "step": 10616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9564663604750182,
+      "learning_rate": 9.700147472019416e-07,
+      "loss": 0.5047,
+      "step": 10617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9176161448635084,
+      "learning_rate": 9.688840350247085e-07,
+      "loss": 0.464,
+      "step": 10618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9724339749225807,
+      "learning_rate": 9.67753948695801e-07,
+      "loss": 0.4263,
+      "step": 10619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0027820480186806,
+      "learning_rate": 9.666244882935339e-07,
+      "loss": 0.5469,
+      "step": 10620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0048726193719604,
+      "learning_rate": 9.65495653896179e-07,
+      "loss": 0.508,
+      "step": 10621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0317181478099673,
+      "learning_rate": 9.643674455819597e-07,
+      "loss": 0.4928,
+      "step": 10622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8361262764620936,
+      "learning_rate": 9.632398634290607e-07,
+      "loss": 0.4458,
+      "step": 10623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0278494857565195,
+      "learning_rate": 9.621129075156256e-07,
+      "loss": 0.4697,
+      "step": 10624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0175732831964581,
+      "learning_rate": 9.60986577919748e-07,
+      "loss": 0.5281,
+      "step": 10625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8834953084316942,
+      "learning_rate": 9.598608747194826e-07,
+      "loss": 0.4439,
+      "step": 10626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9985919876888131,
+      "learning_rate": 9.587357979928414e-07,
+      "loss": 0.5225,
+      "step": 10627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9658488397988649,
+      "learning_rate": 9.576113478177905e-07,
+      "loss": 0.4927,
+      "step": 10628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9535489105894337,
+      "learning_rate": 9.564875242722516e-07,
+      "loss": 0.472,
+      "step": 10629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0003753547033694,
+      "learning_rate": 9.55364327434105e-07,
+      "loss": 0.4753,
+      "step": 10630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8012825219021779,
+      "learning_rate": 9.54241757381188e-07,
+      "loss": 0.4108,
+      "step": 10631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0029078064696209,
+      "learning_rate": 9.531198141912945e-07,
+      "loss": 0.5254,
+      "step": 10632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0070371903454114,
+      "learning_rate": 9.519984979421725e-07,
+      "loss": 0.5345,
+      "step": 10633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9225172268554366,
+      "learning_rate": 9.508778087115289e-07,
+      "loss": 0.4823,
+      "step": 10634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9962315849223458,
+      "learning_rate": 9.497577465770292e-07,
+      "loss": 0.5062,
+      "step": 10635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9698496145273163,
+      "learning_rate": 9.486383116162878e-07,
+      "loss": 0.4988,
+      "step": 10636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9541529632180376,
+      "learning_rate": 9.475195039068818e-07,
+      "loss": 0.4303,
+      "step": 10637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9132257518289799,
+      "learning_rate": 9.464013235263458e-07,
+      "loss": 0.46,
+      "step": 10638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9288673980028853,
+      "learning_rate": 9.452837705521678e-07,
+      "loss": 0.4592,
+      "step": 10639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8436898988686481,
+      "learning_rate": 9.441668450617924e-07,
+      "loss": 0.4092,
+      "step": 10640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9952227583228156,
+      "learning_rate": 9.430505471326246e-07,
+      "loss": 0.503,
+      "step": 10641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9311932065408992,
+      "learning_rate": 9.419348768420178e-07,
+      "loss": 0.4727,
+      "step": 10642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9032988698297807,
+      "learning_rate": 9.408198342672903e-07,
+      "loss": 0.4627,
+      "step": 10643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8884037877380864,
+      "learning_rate": 9.397054194857125e-07,
+      "loss": 0.4278,
+      "step": 10644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.923458523901903,
+      "learning_rate": 9.385916325745115e-07,
+      "loss": 0.4998,
+      "step": 10645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8204070866581368,
+      "learning_rate": 9.374784736108744e-07,
+      "loss": 0.4271,
+      "step": 10646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9939294169077609,
+      "learning_rate": 9.363659426719418e-07,
+      "loss": 0.5363,
+      "step": 10647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.869244868759126,
+      "learning_rate": 9.352540398348087e-07,
+      "loss": 0.4782,
+      "step": 10648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0670999004444721,
+      "learning_rate": 9.3414276517653e-07,
+      "loss": 0.4762,
+      "step": 10649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9443594709032453,
+      "learning_rate": 9.330321187741154e-07,
+      "loss": 0.511,
+      "step": 10650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8889307971922469,
+      "learning_rate": 9.319221007045331e-07,
+      "loss": 0.5285,
+      "step": 10651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8981113533125855,
+      "learning_rate": 9.308127110447063e-07,
+      "loss": 0.4873,
+      "step": 10652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9017810176305928,
+      "learning_rate": 9.297039498715155e-07,
+      "loss": 0.4455,
+      "step": 10653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0546082206019811,
+      "learning_rate": 9.285958172617926e-07,
+      "loss": 0.5087,
+      "step": 10654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0022472235181799,
+      "learning_rate": 9.274883132923362e-07,
+      "loss": 0.4739,
+      "step": 10655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9002821987327692,
+      "learning_rate": 9.263814380398917e-07,
+      "loss": 0.4679,
+      "step": 10656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0302589475729587,
+      "learning_rate": 9.252751915811642e-07,
+      "loss": 0.5019,
+      "step": 10657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9775376726614513,
+      "learning_rate": 9.241695739928169e-07,
+      "loss": 0.5532,
+      "step": 10658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0537623712647077,
+      "learning_rate": 9.230645853514697e-07,
+      "loss": 0.4992,
+      "step": 10659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8776796418795295,
+      "learning_rate": 9.219602257336913e-07,
+      "loss": 0.4862,
+      "step": 10660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.962781037418971,
+      "learning_rate": 9.208564952160215e-07,
+      "loss": 0.493,
+      "step": 10661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0364614771872267,
+      "learning_rate": 9.197533938749414e-07,
+      "loss": 0.5501,
+      "step": 10662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9556984026805189,
+      "learning_rate": 9.186509217868966e-07,
+      "loss": 0.4743,
+      "step": 10663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9616683878097968,
+      "learning_rate": 9.175490790282882e-07,
+      "loss": 0.4714,
+      "step": 10664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9021918453873632,
+      "learning_rate": 9.164478656754739e-07,
+      "loss": 0.4759,
+      "step": 10665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9790575486335737,
+      "learning_rate": 9.153472818047627e-07,
+      "loss": 0.5063,
+      "step": 10666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8468403295990357,
+      "learning_rate": 9.142473274924291e-07,
+      "loss": 0.4937,
+      "step": 10667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9145902214127684,
+      "learning_rate": 9.131480028146955e-07,
+      "loss": 0.4602,
+      "step": 10668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9168147252846369,
+      "learning_rate": 9.120493078477455e-07,
+      "loss": 0.4463,
+      "step": 10669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8323131927350101,
+      "learning_rate": 9.109512426677169e-07,
+      "loss": 0.4231,
+      "step": 10670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9473418109738616,
+      "learning_rate": 9.09853807350708e-07,
+      "loss": 0.5073,
+      "step": 10671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9973434046450055,
+      "learning_rate": 9.08757001972762e-07,
+      "loss": 0.5461,
+      "step": 10672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9805885913133771,
+      "learning_rate": 9.076608266098974e-07,
+      "loss": 0.4762,
+      "step": 10673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9568985687028682,
+      "learning_rate": 9.065652813380699e-07,
+      "loss": 0.5519,
+      "step": 10674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.92717019860699,
+      "learning_rate": 9.054703662332021e-07,
+      "loss": 0.477,
+      "step": 10675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8976396587311715,
+      "learning_rate": 9.04376081371171e-07,
+      "loss": 0.4784,
+      "step": 10676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8976010389778204,
+      "learning_rate": 9.032824268278129e-07,
+      "loss": 0.411,
+      "step": 10677
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9962401365447401,
+      "learning_rate": 9.021894026789091e-07,
+      "loss": 0.5412,
+      "step": 10678
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9492064342678247,
+      "learning_rate": 9.010970090002135e-07,
+      "loss": 0.4523,
+      "step": 10679
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8457178835348091,
+      "learning_rate": 9.000052458674224e-07,
+      "loss": 0.4362,
+      "step": 10680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1243935410473074,
+      "learning_rate": 8.989141133561974e-07,
+      "loss": 0.5494,
+      "step": 10681
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9755791044220888,
+      "learning_rate": 8.978236115421501e-07,
+      "loss": 0.5408,
+      "step": 10682
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9845268895265248,
+      "learning_rate": 8.967337405008558e-07,
+      "loss": 0.4561,
+      "step": 10683
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8520746587525816,
+      "learning_rate": 8.956445003078351e-07,
+      "loss": 0.4592,
+      "step": 10684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9837026512736496,
+      "learning_rate": 8.945558910385776e-07,
+      "loss": 0.4943,
+      "step": 10685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9434157332007918,
+      "learning_rate": 8.934679127685197e-07,
+      "loss": 0.4327,
+      "step": 10686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0156228938382437,
+      "learning_rate": 8.923805655730577e-07,
+      "loss": 0.5323,
+      "step": 10687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0581383360974685,
+      "learning_rate": 8.912938495275436e-07,
+      "loss": 0.4858,
+      "step": 10688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.989591810706298,
+      "learning_rate": 8.902077647072883e-07,
+      "loss": 0.4476,
+      "step": 10689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9795339881955004,
+      "learning_rate": 8.891223111875513e-07,
+      "loss": 0.438,
+      "step": 10690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8201781263068513,
+      "learning_rate": 8.880374890435595e-07,
+      "loss": 0.4644,
+      "step": 10691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9322419967945786,
+      "learning_rate": 8.869532983504859e-07,
+      "loss": 0.4709,
+      "step": 10692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8780674815289228,
+      "learning_rate": 8.858697391834658e-07,
+      "loss": 0.4736,
+      "step": 10693
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0073703685587576,
+      "learning_rate": 8.847868116175883e-07,
+      "loss": 0.5376,
+      "step": 10694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8760429696909396,
+      "learning_rate": 8.837045157279023e-07,
+      "loss": 0.4444,
+      "step": 10695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8900133031097341,
+      "learning_rate": 8.82622851589402e-07,
+      "loss": 0.516,
+      "step": 10696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0236461508169916,
+      "learning_rate": 8.815418192770553e-07,
+      "loss": 0.5519,
+      "step": 10697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9442346465952964,
+      "learning_rate": 8.804614188657712e-07,
+      "loss": 0.5174,
+      "step": 10698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0651480720040911,
+      "learning_rate": 8.793816504304209e-07,
+      "loss": 0.4848,
+      "step": 10699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9156691302175118,
+      "learning_rate": 8.783025140458334e-07,
+      "loss": 0.4594,
+      "step": 10700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9191454740517091,
+      "learning_rate": 8.772240097867912e-07,
+      "loss": 0.4513,
+      "step": 10701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.842071309153491,
+      "learning_rate": 8.761461377280311e-07,
+      "loss": 0.4486,
+      "step": 10702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.001306254241895,
+      "learning_rate": 8.750688979442534e-07,
+      "loss": 0.5239,
+      "step": 10703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8248599393774404,
+      "learning_rate": 8.739922905101051e-07,
+      "loss": 0.4204,
+      "step": 10704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9491459913050245,
+      "learning_rate": 8.729163155001975e-07,
+      "loss": 0.5073,
+      "step": 10705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2972173703532743,
+      "learning_rate": 8.71840972989092e-07,
+      "loss": 0.4949,
+      "step": 10706
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8906322183288898,
+      "learning_rate": 8.707662630513136e-07,
+      "loss": 0.4899,
+      "step": 10707
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.976107320047754,
+      "learning_rate": 8.696921857613317e-07,
+      "loss": 0.521,
+      "step": 10708
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9772816546434508,
+      "learning_rate": 8.686187411935854e-07,
+      "loss": 0.4138,
+      "step": 10709
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9880903316140534,
+      "learning_rate": 8.675459294224597e-07,
+      "loss": 0.5624,
+      "step": 10710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9692471090732485,
+      "learning_rate": 8.664737505223009e-07,
+      "loss": 0.4838,
+      "step": 10711
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9192191189203573,
+      "learning_rate": 8.654022045674093e-07,
+      "loss": 0.4744,
+      "step": 10712
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8948526661133622,
+      "learning_rate": 8.643312916320446e-07,
+      "loss": 0.4988,
+      "step": 10713
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9549774148056884,
+      "learning_rate": 8.632610117904139e-07,
+      "loss": 0.4531,
+      "step": 10714
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8601975926054979,
+      "learning_rate": 8.621913651166947e-07,
+      "loss": 0.4734,
+      "step": 10715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9024200533857866,
+      "learning_rate": 8.611223516850076e-07,
+      "loss": 0.4966,
+      "step": 10716
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8210537135329794,
+      "learning_rate": 8.600539715694344e-07,
+      "loss": 0.4491,
+      "step": 10717
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0435012893610338,
+      "learning_rate": 8.58986224844014e-07,
+      "loss": 0.492,
+      "step": 10718
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.905447549472864,
+      "learning_rate": 8.579191115827423e-07,
+      "loss": 0.4656,
+      "step": 10719
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9702528122983963,
+      "learning_rate": 8.568526318595638e-07,
+      "loss": 0.5027,
+      "step": 10720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9028108539852475,
+      "learning_rate": 8.557867857483915e-07,
+      "loss": 0.4443,
+      "step": 10721
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9828446495002682,
+      "learning_rate": 8.547215733230818e-07,
+      "loss": 0.4912,
+      "step": 10722
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1314179435472174,
+      "learning_rate": 8.536569946574546e-07,
+      "loss": 0.5257,
+      "step": 10723
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0384549778518128,
+      "learning_rate": 8.525930498252855e-07,
+      "loss": 0.5085,
+      "step": 10724
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0259094490019722,
+      "learning_rate": 8.515297389003063e-07,
+      "loss": 0.4782,
+      "step": 10725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8670612352353387,
+      "learning_rate": 8.504670619561983e-07,
+      "loss": 0.4417,
+      "step": 10726
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9064697350293002,
+      "learning_rate": 8.494050190666103e-07,
+      "loss": 0.4855,
+      "step": 10727
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9116173287893347,
+      "learning_rate": 8.483436103051357e-07,
+      "loss": 0.4691,
+      "step": 10728
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9671185448154255,
+      "learning_rate": 8.472828357453323e-07,
+      "loss": 0.5249,
+      "step": 10729
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9073917664980375,
+      "learning_rate": 8.4622269546071e-07,
+      "loss": 0.4435,
+      "step": 10730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9893708961286303,
+      "learning_rate": 8.45163189524737e-07,
+      "loss": 0.5178,
+      "step": 10731
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8057742513191817,
+      "learning_rate": 8.441043180108299e-07,
+      "loss": 0.4719,
+      "step": 10732
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9424113091164239,
+      "learning_rate": 8.430460809923768e-07,
+      "loss": 0.4641,
+      "step": 10733
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0577210260964864,
+      "learning_rate": 8.419884785427068e-07,
+      "loss": 0.5043,
+      "step": 10734
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.017229119006042,
+      "learning_rate": 8.409315107351112e-07,
+      "loss": 0.4864,
+      "step": 10735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9160592998089141,
+      "learning_rate": 8.398751776428393e-07,
+      "loss": 0.483,
+      "step": 10736
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8600297093969423,
+      "learning_rate": 8.388194793390924e-07,
+      "loss": 0.4128,
+      "step": 10737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8505559599580625,
+      "learning_rate": 8.377644158970277e-07,
+      "loss": 0.4854,
+      "step": 10738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9108433349901007,
+      "learning_rate": 8.367099873897644e-07,
+      "loss": 0.541,
+      "step": 10739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9776569227810032,
+      "learning_rate": 8.356561938903707e-07,
+      "loss": 0.5111,
+      "step": 10740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9698205117012005,
+      "learning_rate": 8.346030354718727e-07,
+      "loss": 0.4707,
+      "step": 10741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9094940212730316,
+      "learning_rate": 8.335505122072551e-07,
+      "loss": 0.4906,
+      "step": 10742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9679125119758466,
+      "learning_rate": 8.324986241694566e-07,
+      "loss": 0.4892,
+      "step": 10743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9341757886112941,
+      "learning_rate": 8.31447371431372e-07,
+      "loss": 0.4337,
+      "step": 10744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8865816044503693,
+      "learning_rate": 8.303967540658531e-07,
+      "loss": 0.4954,
+      "step": 10745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0152423724483588,
+      "learning_rate": 8.293467721457038e-07,
+      "loss": 0.5048,
+      "step": 10746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0778633743848953,
+      "learning_rate": 8.282974257436904e-07,
+      "loss": 0.4798,
+      "step": 10747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9664900899900147,
+      "learning_rate": 8.272487149325281e-07,
+      "loss": 0.504,
+      "step": 10748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8752434555018468,
+      "learning_rate": 8.262006397848954e-07,
+      "loss": 0.4622,
+      "step": 10749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9312621506287185,
+      "learning_rate": 8.251532003734197e-07,
+      "loss": 0.4799,
+      "step": 10750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9890867364594251,
+      "learning_rate": 8.24106396770692e-07,
+      "loss": 0.4658,
+      "step": 10751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7633149847762869,
+      "learning_rate": 8.230602290492485e-07,
+      "loss": 0.4521,
+      "step": 10752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8720687155099099,
+      "learning_rate": 8.220146972815946e-07,
+      "loss": 0.4463,
+      "step": 10753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9359147457315855,
+      "learning_rate": 8.209698015401791e-07,
+      "loss": 0.4866,
+      "step": 10754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.958715743537337,
+      "learning_rate": 8.19925541897415e-07,
+      "loss": 0.5369,
+      "step": 10755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0407037298595796,
+      "learning_rate": 8.188819184256669e-07,
+      "loss": 0.5594,
+      "step": 10756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9444534547445447,
+      "learning_rate": 8.178389311972612e-07,
+      "loss": 0.5005,
+      "step": 10757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8801319988525427,
+      "learning_rate": 8.167965802844691e-07,
+      "loss": 0.4225,
+      "step": 10758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9247134193102192,
+      "learning_rate": 8.157548657595327e-07,
+      "loss": 0.4827,
+      "step": 10759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9280867544437789,
+      "learning_rate": 8.147137876946354e-07,
+      "loss": 0.427,
+      "step": 10760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9742366224152857,
+      "learning_rate": 8.136733461619251e-07,
+      "loss": 0.4937,
+      "step": 10761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9089698408617726,
+      "learning_rate": 8.12633541233504e-07,
+      "loss": 0.4541,
+      "step": 10762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9281472717487014,
+      "learning_rate": 8.11594372981429e-07,
+      "loss": 0.4826,
+      "step": 10763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8543302871144066,
+      "learning_rate": 8.105558414777137e-07,
+      "loss": 0.4622,
+      "step": 10764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8930667784194314,
+      "learning_rate": 8.095179467943293e-07,
+      "loss": 0.4421,
+      "step": 10765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0264206424491298,
+      "learning_rate": 8.084806890031982e-07,
+      "loss": 0.5471,
+      "step": 10766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8440403105832182,
+      "learning_rate": 8.074440681762019e-07,
+      "loss": 0.4352,
+      "step": 10767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9838278479205321,
+      "learning_rate": 8.064080843851785e-07,
+      "loss": 0.5155,
+      "step": 10768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9654762427606757,
+      "learning_rate": 8.053727377019194e-07,
+      "loss": 0.5047,
+      "step": 10769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9212823039571989,
+      "learning_rate": 8.043380281981739e-07,
+      "loss": 0.4555,
+      "step": 10770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9962002914394492,
+      "learning_rate": 8.03303955945649e-07,
+      "loss": 0.4671,
+      "step": 10771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9551262436521267,
+      "learning_rate": 8.022705210159997e-07,
+      "loss": 0.4842,
+      "step": 10772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0214526838284614,
+      "learning_rate": 8.012377234808455e-07,
+      "loss": 0.4988,
+      "step": 10773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8715570587673264,
+      "learning_rate": 8.002055634117578e-07,
+      "loss": 0.4734,
+      "step": 10774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9266589001004784,
+      "learning_rate": 7.991740408802651e-07,
+      "loss": 0.5139,
+      "step": 10775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0169422673348818,
+      "learning_rate": 7.98143155957849e-07,
+      "loss": 0.4969,
+      "step": 10776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0076439713948204,
+      "learning_rate": 7.971129087159524e-07,
+      "loss": 0.5081,
+      "step": 10777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9002053839249186,
+      "learning_rate": 7.960832992259671e-07,
+      "loss": 0.4021,
+      "step": 10778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9091524190557226,
+      "learning_rate": 7.950543275592449e-07,
+      "loss": 0.4644,
+      "step": 10779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0333079754352386,
+      "learning_rate": 7.94025993787092e-07,
+      "loss": 0.514,
+      "step": 10780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9823506631557626,
+      "learning_rate": 7.929982979807738e-07,
+      "loss": 0.5355,
+      "step": 10781
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9000299533512557,
+      "learning_rate": 7.919712402115054e-07,
+      "loss": 0.422,
+      "step": 10782
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0293709792663799,
+      "learning_rate": 7.909448205504633e-07,
+      "loss": 0.536,
+      "step": 10783
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9110279158047219,
+      "learning_rate": 7.899190390687783e-07,
+      "loss": 0.4659,
+      "step": 10784
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9492518309308532,
+      "learning_rate": 7.888938958375325e-07,
+      "loss": 0.5341,
+      "step": 10785
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0609084182976722,
+      "learning_rate": 7.878693909277702e-07,
+      "loss": 0.5124,
+      "step": 10786
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0009821719336907,
+      "learning_rate": 7.868455244104878e-07,
+      "loss": 0.4542,
+      "step": 10787
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8843299754398897,
+      "learning_rate": 7.858222963566386e-07,
+      "loss": 0.4965,
+      "step": 10788
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8676271737361042,
+      "learning_rate": 7.847997068371305e-07,
+      "loss": 0.4224,
+      "step": 10789
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9568586304329556,
+      "learning_rate": 7.83777755922831e-07,
+      "loss": 0.4841,
+      "step": 10790
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9568289014292817,
+      "learning_rate": 7.827564436845569e-07,
+      "loss": 0.4305,
+      "step": 10791
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9007924177189064,
+      "learning_rate": 7.81735770193085e-07,
+      "loss": 0.4777,
+      "step": 10792
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0356572627261904,
+      "learning_rate": 7.807157355191475e-07,
+      "loss": 0.4975,
+      "step": 10793
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9153262416116543,
+      "learning_rate": 7.796963397334323e-07,
+      "loss": 0.4281,
+      "step": 10794
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9513672979130858,
+      "learning_rate": 7.786775829065829e-07,
+      "loss": 0.5165,
+      "step": 10795
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9632117362919128,
+      "learning_rate": 7.776594651091995e-07,
+      "loss": 0.487,
+      "step": 10796
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9401247195442137,
+      "learning_rate": 7.766419864118325e-07,
+      "loss": 0.4191,
+      "step": 10797
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9176285200237425,
+      "learning_rate": 7.756251468849951e-07,
+      "loss": 0.4436,
+      "step": 10798
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8611310925495618,
+      "learning_rate": 7.746089465991525e-07,
+      "loss": 0.4971,
+      "step": 10799
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9305324159779296,
+      "learning_rate": 7.735933856247269e-07,
+      "loss": 0.4685,
+      "step": 10800
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9107641464674445,
+      "learning_rate": 7.725784640320966e-07,
+      "loss": 0.4931,
+      "step": 10801
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9806018240071644,
+      "learning_rate": 7.715641818915953e-07,
+      "loss": 0.5458,
+      "step": 10802
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9257241322988893,
+      "learning_rate": 7.70550539273508e-07,
+      "loss": 0.4582,
+      "step": 10803
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8949923140018721,
+      "learning_rate": 7.695375362480839e-07,
+      "loss": 0.4443,
+      "step": 10804
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0077955943455112,
+      "learning_rate": 7.685251728855203e-07,
+      "loss": 0.4722,
+      "step": 10805
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8865636886945156,
+      "learning_rate": 7.675134492559733e-07,
+      "loss": 0.5064,
+      "step": 10806
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0190987369655813,
+      "learning_rate": 7.665023654295556e-07,
+      "loss": 0.4824,
+      "step": 10807
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9215282547212966,
+      "learning_rate": 7.654919214763357e-07,
+      "loss": 0.4839,
+      "step": 10808
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9555787188807393,
+      "learning_rate": 7.644821174663308e-07,
+      "loss": 0.4842,
+      "step": 10809
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9429919388180836,
+      "learning_rate": 7.634729534695273e-07,
+      "loss": 0.4577,
+      "step": 10810
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8458292575931327,
+      "learning_rate": 7.624644295558525e-07,
+      "loss": 0.4069,
+      "step": 10811
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8661683221124166,
+      "learning_rate": 7.614565457952005e-07,
+      "loss": 0.4705,
+      "step": 10812
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9333855764767786,
+      "learning_rate": 7.604493022574144e-07,
+      "loss": 0.5074,
+      "step": 10813
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9956363320104379,
+      "learning_rate": 7.594426990122972e-07,
+      "loss": 0.5214,
+      "step": 10814
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9448774180808857,
+      "learning_rate": 7.58436736129603e-07,
+      "loss": 0.4514,
+      "step": 10815
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9361065086180625,
+      "learning_rate": 7.574314136790472e-07,
+      "loss": 0.4398,
+      "step": 10816
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9122194932866072,
+      "learning_rate": 7.564267317302965e-07,
+      "loss": 0.4273,
+      "step": 10817
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9753113897717414,
+      "learning_rate": 7.554226903529726e-07,
+      "loss": 0.5101,
+      "step": 10818
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.914937109857151,
+      "learning_rate": 7.544192896166569e-07,
+      "loss": 0.4923,
+      "step": 10819
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9201451087784769,
+      "learning_rate": 7.534165295908857e-07,
+      "loss": 0.4767,
+      "step": 10820
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9650755957588284,
+      "learning_rate": 7.524144103451436e-07,
+      "loss": 0.4629,
+      "step": 10821
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9972999225253916,
+      "learning_rate": 7.514129319488839e-07,
+      "loss": 0.5472,
+      "step": 10822
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9633231367308103,
+      "learning_rate": 7.504120944715021e-07,
+      "loss": 0.4474,
+      "step": 10823
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9708221061585173,
+      "learning_rate": 7.494118979823584e-07,
+      "loss": 0.4915,
+      "step": 10824
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9672049816339459,
+      "learning_rate": 7.48412342550765e-07,
+      "loss": 0.5398,
+      "step": 10825
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8233298296033982,
+      "learning_rate": 7.47413428245991e-07,
+      "loss": 0.4419,
+      "step": 10826
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9569979497855227,
+      "learning_rate": 7.464151551372567e-07,
+      "loss": 0.4963,
+      "step": 10827
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9372064091819019,
+      "learning_rate": 7.454175232937478e-07,
+      "loss": 0.557,
+      "step": 10828
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9857819829803073,
+      "learning_rate": 7.444205327845932e-07,
+      "loss": 0.5188,
+      "step": 10829
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8485091146600624,
+      "learning_rate": 7.43424183678887e-07,
+      "loss": 0.445,
+      "step": 10830
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9093988479334427,
+      "learning_rate": 7.424284760456734e-07,
+      "loss": 0.4602,
+      "step": 10831
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8697372110048283,
+      "learning_rate": 7.414334099539577e-07,
+      "loss": 0.3928,
+      "step": 10832
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8500637707605192,
+      "learning_rate": 7.404389854726901e-07,
+      "loss": 0.4605,
+      "step": 10833
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9450728224522065,
+      "learning_rate": 7.39445202670791e-07,
+      "loss": 0.495,
+      "step": 10834
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9922786195641765,
+      "learning_rate": 7.384520616171232e-07,
+      "loss": 0.5178,
+      "step": 10835
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9618426009645406,
+      "learning_rate": 7.374595623805137e-07,
+      "loss": 0.4966,
+      "step": 10836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9462859467040637,
+      "learning_rate": 7.364677050297398e-07,
+      "loss": 0.4739,
+      "step": 10837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0227678882461722,
+      "learning_rate": 7.354764896335398e-07,
+      "loss": 0.5576,
+      "step": 10838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9980851593236497,
+      "learning_rate": 7.344859162605966e-07,
+      "loss": 0.4946,
+      "step": 10839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9237865868634112,
+      "learning_rate": 7.334959849795653e-07,
+      "loss": 0.4839,
+      "step": 10840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9178993647703343,
+      "learning_rate": 7.32506695859041e-07,
+      "loss": 0.4641,
+      "step": 10841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9519479260351561,
+      "learning_rate": 7.315180489675822e-07,
+      "loss": 0.5189,
+      "step": 10842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.960758292084623,
+      "learning_rate": 7.305300443737018e-07,
+      "loss": 0.5433,
+      "step": 10843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9801777325896154,
+      "learning_rate": 7.295426821458684e-07,
+      "loss": 0.5057,
+      "step": 10844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9286446248352245,
+      "learning_rate": 7.285559623525018e-07,
+      "loss": 0.4899,
+      "step": 10845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9399120756588593,
+      "learning_rate": 7.275698850619861e-07,
+      "loss": 0.5041,
+      "step": 10846
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8796425834099239,
+      "learning_rate": 7.265844503426512e-07,
+      "loss": 0.4751,
+      "step": 10847
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9478441410449189,
+      "learning_rate": 7.255996582627878e-07,
+      "loss": 0.4731,
+      "step": 10848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9602936406666883,
+      "learning_rate": 7.246155088906426e-07,
+      "loss": 0.4869,
+      "step": 10849
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8380944081829118,
+      "learning_rate": 7.236320022944166e-07,
+      "loss": 0.4285,
+      "step": 10850
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0144426336924712,
+      "learning_rate": 7.226491385422618e-07,
+      "loss": 0.5351,
+      "step": 10851
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9866428772036293,
+      "learning_rate": 7.21666917702295e-07,
+      "loss": 0.4802,
+      "step": 10852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0833877967513377,
+      "learning_rate": 7.206853398425806e-07,
+      "loss": 0.4986,
+      "step": 10853
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9699105979255526,
+      "learning_rate": 7.197044050311408e-07,
+      "loss": 0.495,
+      "step": 10854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8461947804009808,
+      "learning_rate": 7.187241133359535e-07,
+      "loss": 0.4701,
+      "step": 10855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0087068121306357,
+      "learning_rate": 7.177444648249554e-07,
+      "loss": 0.5119,
+      "step": 10856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.927842032367001,
+      "learning_rate": 7.167654595660279e-07,
+      "loss": 0.4756,
+      "step": 10857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0954507707339836,
+      "learning_rate": 7.157870976270243e-07,
+      "loss": 0.5273,
+      "step": 10858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9649063436888304,
+      "learning_rate": 7.148093790757371e-07,
+      "loss": 0.5375,
+      "step": 10859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8600627356479995,
+      "learning_rate": 7.138323039799256e-07,
+      "loss": 0.4304,
+      "step": 10860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0326935851609784,
+      "learning_rate": 7.128558724072976e-07,
+      "loss": 0.489,
+      "step": 10861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8968902380842844,
+      "learning_rate": 7.118800844255214e-07,
+      "loss": 0.5305,
+      "step": 10862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9482496308476169,
+      "learning_rate": 7.109049401022139e-07,
+      "loss": 0.5105,
+      "step": 10863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0149047865729435,
+      "learning_rate": 7.099304395049566e-07,
+      "loss": 0.554,
+      "step": 10864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9135791307678812,
+      "learning_rate": 7.089565827012801e-07,
+      "loss": 0.5188,
+      "step": 10865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8845742531981535,
+      "learning_rate": 7.079833697586702e-07,
+      "loss": 0.4493,
+      "step": 10866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0187836100942078,
+      "learning_rate": 7.070108007445708e-07,
+      "loss": 0.5208,
+      "step": 10867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9641153712470332,
+      "learning_rate": 7.060388757263815e-07,
+      "loss": 0.4721,
+      "step": 10868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9580963683240067,
+      "learning_rate": 7.050675947714514e-07,
+      "loss": 0.525,
+      "step": 10869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9812100492397047,
+      "learning_rate": 7.040969579470947e-07,
+      "loss": 0.4649,
+      "step": 10870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9725844565798529,
+      "learning_rate": 7.03126965320573e-07,
+      "loss": 0.5216,
+      "step": 10871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0468445203398065,
+      "learning_rate": 7.02157616959106e-07,
+      "loss": 0.4856,
+      "step": 10872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.938268266463854,
+      "learning_rate": 7.011889129298688e-07,
+      "loss": 0.4724,
+      "step": 10873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9344094084559463,
+      "learning_rate": 7.002208532999933e-07,
+      "loss": 0.4824,
+      "step": 10874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0943727000052077,
+      "learning_rate": 6.992534381365612e-07,
+      "loss": 0.4795,
+      "step": 10875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9361214667194621,
+      "learning_rate": 6.98286667506618e-07,
+      "loss": 0.46,
+      "step": 10876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9742787944574746,
+      "learning_rate": 6.973205414771567e-07,
+      "loss": 0.4995,
+      "step": 10877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9823861996132334,
+      "learning_rate": 6.963550601151326e-07,
+      "loss": 0.4549,
+      "step": 10878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8710597836786871,
+      "learning_rate": 6.953902234874488e-07,
+      "loss": 0.4224,
+      "step": 10879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9640184246732854,
+      "learning_rate": 6.944260316609696e-07,
+      "loss": 0.4772,
+      "step": 10880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9443593319000072,
+      "learning_rate": 6.934624847025117e-07,
+      "loss": 0.4872,
+      "step": 10881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9102355113413119,
+      "learning_rate": 6.924995826788516e-07,
+      "loss": 0.4886,
+      "step": 10882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8545741021936641,
+      "learning_rate": 6.915373256567104e-07,
+      "loss": 0.4124,
+      "step": 10883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8849429457331734,
+      "learning_rate": 6.905757137027791e-07,
+      "loss": 0.5111,
+      "step": 10884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9274095690131027,
+      "learning_rate": 6.896147468836923e-07,
+      "loss": 0.4492,
+      "step": 10885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9868491348465521,
+      "learning_rate": 6.886544252660455e-07,
+      "loss": 0.4611,
+      "step": 10886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9705796300856186,
+      "learning_rate": 6.876947489163877e-07,
+      "loss": 0.5528,
+      "step": 10887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8459196623914732,
+      "learning_rate": 6.867357179012257e-07,
+      "loss": 0.4765,
+      "step": 10888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9718706349770099,
+      "learning_rate": 6.857773322870132e-07,
+      "loss": 0.5085,
+      "step": 10889
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0231757989355228,
+      "learning_rate": 6.848195921401745e-07,
+      "loss": 0.5608,
+      "step": 10890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9926785324714231,
+      "learning_rate": 6.838624975270724e-07,
+      "loss": 0.556,
+      "step": 10891
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8648754199032377,
+      "learning_rate": 6.829060485140371e-07,
+      "loss": 0.4641,
+      "step": 10892
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9565565653738822,
+      "learning_rate": 6.819502451673477e-07,
+      "loss": 0.4483,
+      "step": 10893
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9210008914980209,
+      "learning_rate": 6.809950875532434e-07,
+      "loss": 0.4526,
+      "step": 10894
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9462179584061005,
+      "learning_rate": 6.800405757379103e-07,
+      "loss": 0.458,
+      "step": 10895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9380696791158867,
+      "learning_rate": 6.790867097875009e-07,
+      "loss": 0.4847,
+      "step": 10896
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8742693957607353,
+      "learning_rate": 6.781334897681136e-07,
+      "loss": 0.4233,
+      "step": 10897
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9531969331088594,
+      "learning_rate": 6.771809157458076e-07,
+      "loss": 0.4273,
+      "step": 10898
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9797729723207618,
+      "learning_rate": 6.762289877865946e-07,
+      "loss": 0.5301,
+      "step": 10899
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9302885601403638,
+      "learning_rate": 6.752777059564431e-07,
+      "loss": 0.498,
+      "step": 10900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9756389310003124,
+      "learning_rate": 6.743270703212734e-07,
+      "loss": 0.5063,
+      "step": 10901
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8954492108953634,
+      "learning_rate": 6.733770809469686e-07,
+      "loss": 0.4318,
+      "step": 10902
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9195584154690314,
+      "learning_rate": 6.724277378993582e-07,
+      "loss": 0.463,
+      "step": 10903
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8998700767219938,
+      "learning_rate": 6.714790412442318e-07,
+      "loss": 0.4945,
+      "step": 10904
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9843924532722907,
+      "learning_rate": 6.705309910473334e-07,
+      "loss": 0.5438,
+      "step": 10905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8902657261878095,
+      "learning_rate": 6.695835873743639e-07,
+      "loss": 0.4437,
+      "step": 10906
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9797169219451506,
+      "learning_rate": 6.686368302909729e-07,
+      "loss": 0.5289,
+      "step": 10907
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.074011462740834,
+      "learning_rate": 6.676907198627758e-07,
+      "loss": 0.5319,
+      "step": 10908
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8932783969377371,
+      "learning_rate": 6.667452561553312e-07,
+      "loss": 0.4721,
+      "step": 10909
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.007817520165382,
+      "learning_rate": 6.658004392341633e-07,
+      "loss": 0.4764,
+      "step": 10910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9753415517502435,
+      "learning_rate": 6.648562691647443e-07,
+      "loss": 0.5611,
+      "step": 10911
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8876119878062856,
+      "learning_rate": 6.639127460125061e-07,
+      "loss": 0.4456,
+      "step": 10912
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8825191330874997,
+      "learning_rate": 6.629698698428333e-07,
+      "loss": 0.4839,
+      "step": 10913
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8801074389167012,
+      "learning_rate": 6.620276407210691e-07,
+      "loss": 0.435,
+      "step": 10914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.907815847163721,
+      "learning_rate": 6.610860587125046e-07,
+      "loss": 0.4488,
+      "step": 10915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8857465115695343,
+      "learning_rate": 6.60145123882392e-07,
+      "loss": 0.4601,
+      "step": 10916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0385760901105907,
+      "learning_rate": 6.59204836295938e-07,
+      "loss": 0.5081,
+      "step": 10917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8568978773404449,
+      "learning_rate": 6.582651960183039e-07,
+      "loss": 0.4356,
+      "step": 10918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.988066516943962,
+      "learning_rate": 6.573262031146055e-07,
+      "loss": 0.4722,
+      "step": 10919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8581986976490972,
+      "learning_rate": 6.563878576499161e-07,
+      "loss": 0.4413,
+      "step": 10920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7929721822827478,
+      "learning_rate": 6.554501596892582e-07,
+      "loss": 0.4118,
+      "step": 10921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9392852715664914,
+      "learning_rate": 6.545131092976165e-07,
+      "loss": 0.5108,
+      "step": 10922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8630871527253957,
+      "learning_rate": 6.535767065399268e-07,
+      "loss": 0.4296,
+      "step": 10923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9950951675346889,
+      "learning_rate": 6.526409514810805e-07,
+      "loss": 0.4728,
+      "step": 10924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9887197876289022,
+      "learning_rate": 6.517058441859248e-07,
+      "loss": 0.4998,
+      "step": 10925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9341212052113047,
+      "learning_rate": 6.507713847192643e-07,
+      "loss": 0.488,
+      "step": 10926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9132481462189752,
+      "learning_rate": 6.498375731458529e-07,
+      "loss": 0.4746,
+      "step": 10927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9152190176417002,
+      "learning_rate": 6.489044095304031e-07,
+      "loss": 0.4698,
+      "step": 10928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.90130991738369,
+      "learning_rate": 6.479718939375846e-07,
+      "loss": 0.4477,
+      "step": 10929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.91942422370572,
+      "learning_rate": 6.470400264320176e-07,
+      "loss": 0.5178,
+      "step": 10930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0685961246932574,
+      "learning_rate": 6.461088070782806e-07,
+      "loss": 0.5362,
+      "step": 10931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8916758914035164,
+      "learning_rate": 6.451782359409076e-07,
+      "loss": 0.4689,
+      "step": 10932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.872337058372946,
+      "learning_rate": 6.442483130843857e-07,
+      "loss": 0.4741,
+      "step": 10933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9024681459121816,
+      "learning_rate": 6.433190385731558e-07,
+      "loss": 0.4499,
+      "step": 10934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9169060857895663,
+      "learning_rate": 6.423904124716174e-07,
+      "loss": 0.4825,
+      "step": 10935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9640055684093934,
+      "learning_rate": 6.414624348441223e-07,
+      "loss": 0.4722,
+      "step": 10936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9129689547723554,
+      "learning_rate": 6.405351057549803e-07,
+      "loss": 0.4385,
+      "step": 10937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0296726654633954,
+      "learning_rate": 6.396084252684532e-07,
+      "loss": 0.4515,
+      "step": 10938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9326947083487568,
+      "learning_rate": 6.386823934487619e-07,
+      "loss": 0.5332,
+      "step": 10939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9960249084282743,
+      "learning_rate": 6.377570103600749e-07,
+      "loss": 0.4267,
+      "step": 10940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8830118434965389,
+      "learning_rate": 6.368322760665235e-07,
+      "loss": 0.467,
+      "step": 10941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9758683488076908,
+      "learning_rate": 6.359081906321896e-07,
+      "loss": 0.4982,
+      "step": 10942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9488295398679685,
+      "learning_rate": 6.349847541211119e-07,
+      "loss": 0.4827,
+      "step": 10943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9750251787938423,
+      "learning_rate": 6.340619665972847e-07,
+      "loss": 0.46,
+      "step": 10944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0234616225302677,
+      "learning_rate": 6.33139828124657e-07,
+      "loss": 0.4685,
+      "step": 10945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0455466925977317,
+      "learning_rate": 6.322183387671299e-07,
+      "loss": 0.4855,
+      "step": 10946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9754433315723913,
+      "learning_rate": 6.312974985885612e-07,
+      "loss": 0.4757,
+      "step": 10947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9357797846488201,
+      "learning_rate": 6.303773076527663e-07,
+      "loss": 0.5077,
+      "step": 10948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9014709161998172,
+      "learning_rate": 6.294577660235146e-07,
+      "loss": 0.4161,
+      "step": 10949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.968122195587275,
+      "learning_rate": 6.28538873764526e-07,
+      "loss": 0.413,
+      "step": 10950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9321219084135877,
+      "learning_rate": 6.276206309394839e-07,
+      "loss": 0.4747,
+      "step": 10951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9857430350810453,
+      "learning_rate": 6.267030376120154e-07,
+      "loss": 0.4877,
+      "step": 10952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9257006654534886,
+      "learning_rate": 6.25786093845715e-07,
+      "loss": 0.5431,
+      "step": 10953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7813395537425469,
+      "learning_rate": 6.248697997041219e-07,
+      "loss": 0.3824,
+      "step": 10954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9357085513534893,
+      "learning_rate": 6.239541552507367e-07,
+      "loss": 0.5074,
+      "step": 10955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9781819541718835,
+      "learning_rate": 6.230391605490105e-07,
+      "loss": 0.5044,
+      "step": 10956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0309509193357862,
+      "learning_rate": 6.22124815662356e-07,
+      "loss": 0.5121,
+      "step": 10957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0824573399034947,
+      "learning_rate": 6.212111206541305e-07,
+      "loss": 0.5119,
+      "step": 10958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0262200241800767,
+      "learning_rate": 6.202980755876575e-07,
+      "loss": 0.4679,
+      "step": 10959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9675571655609572,
+      "learning_rate": 6.193856805262078e-07,
+      "loss": 0.483,
+      "step": 10960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9213920082584263,
+      "learning_rate": 6.184739355330083e-07,
+      "loss": 0.5297,
+      "step": 10961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.052124394282512,
+      "learning_rate": 6.175628406712452e-07,
+      "loss": 0.4136,
+      "step": 10962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9134209043859236,
+      "learning_rate": 6.166523960040549e-07,
+      "loss": 0.4796,
+      "step": 10963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0368144330888827,
+      "learning_rate": 6.157426015945289e-07,
+      "loss": 0.5225,
+      "step": 10964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0664674535151522,
+      "learning_rate": 6.148334575057191e-07,
+      "loss": 0.5076,
+      "step": 10965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0175371255955596,
+      "learning_rate": 6.139249638006251e-07,
+      "loss": 0.5053,
+      "step": 10966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0233060979369943,
+      "learning_rate": 6.130171205422053e-07,
+      "loss": 0.4673,
+      "step": 10967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9631441370141315,
+      "learning_rate": 6.121099277933728e-07,
+      "loss": 0.5094,
+      "step": 10968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0057396520217246,
+      "learning_rate": 6.112033856169974e-07,
+      "loss": 0.5465,
+      "step": 10969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9561951029569171,
+      "learning_rate": 6.102974940758954e-07,
+      "loss": 0.4974,
+      "step": 10970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9866723996930895,
+      "learning_rate": 6.093922532328522e-07,
+      "loss": 0.542,
+      "step": 10971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.052357825958954,
+      "learning_rate": 6.084876631505943e-07,
+      "loss": 0.5748,
+      "step": 10972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8457013897677658,
+      "learning_rate": 6.075837238918114e-07,
+      "loss": 0.4256,
+      "step": 10973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9843655807602537,
+      "learning_rate": 6.066804355191458e-07,
+      "loss": 0.5103,
+      "step": 10974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9114613086221101,
+      "learning_rate": 6.057777980951951e-07,
+      "loss": 0.3927,
+      "step": 10975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0110177998447731,
+      "learning_rate": 6.048758116825071e-07,
+      "loss": 0.5294,
+      "step": 10976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9426252423474122,
+      "learning_rate": 6.039744763435951e-07,
+      "loss": 0.4731,
+      "step": 10977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9087938185398369,
+      "learning_rate": 6.030737921409169e-07,
+      "loss": 0.4269,
+      "step": 10978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8901526154211044,
+      "learning_rate": 6.021737591368892e-07,
+      "loss": 0.483,
+      "step": 10979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.929054298926748,
+      "learning_rate": 6.012743773938845e-07,
+      "loss": 0.4684,
+      "step": 10980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9998305319817405,
+      "learning_rate": 6.003756469742294e-07,
+      "loss": 0.5043,
+      "step": 10981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9654263645064304,
+      "learning_rate": 5.994775679402021e-07,
+      "loss": 0.4799,
+      "step": 10982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8812144427476013,
+      "learning_rate": 5.985801403540436e-07,
+      "loss": 0.4791,
+      "step": 10983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0297036512654214,
+      "learning_rate": 5.976833642779422e-07,
+      "loss": 0.4931,
+      "step": 10984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0959168985236405,
+      "learning_rate": 5.967872397740427e-07,
+      "loss": 0.5191,
+      "step": 10985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9548852649181496,
+      "learning_rate": 5.958917669044472e-07,
+      "loss": 0.4726,
+      "step": 10986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8860636695272686,
+      "learning_rate": 5.949969457312122e-07,
+      "loss": 0.4133,
+      "step": 10987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9551069857313534,
+      "learning_rate": 5.941027763163432e-07,
+      "loss": 0.4779,
+      "step": 10988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9587762764673151,
+      "learning_rate": 5.93209258721812e-07,
+      "loss": 0.4449,
+      "step": 10989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9764154005326386,
+      "learning_rate": 5.923163930095344e-07,
+      "loss": 0.5057,
+      "step": 10990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8694608537022765,
+      "learning_rate": 5.914241792413855e-07,
+      "loss": 0.433,
+      "step": 10991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8938395567692832,
+      "learning_rate": 5.905326174791959e-07,
+      "loss": 0.4547,
+      "step": 10992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8700693290069454,
+      "learning_rate": 5.896417077847505e-07,
+      "loss": 0.443,
+      "step": 10993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9365351744115846,
+      "learning_rate": 5.887514502197855e-07,
+      "loss": 0.5051,
+      "step": 10994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.953988434394032,
+      "learning_rate": 5.878618448460005e-07,
+      "loss": 0.4766,
+      "step": 10995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0255533396332113,
+      "learning_rate": 5.869728917250394e-07,
+      "loss": 0.4966,
+      "step": 10996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9351106480000976,
+      "learning_rate": 5.860845909185076e-07,
+      "loss": 0.4979,
+      "step": 10997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9267295875852679,
+      "learning_rate": 5.851969424879633e-07,
+      "loss": 0.5132,
+      "step": 10998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.044249424301423,
+      "learning_rate": 5.843099464949198e-07,
+      "loss": 0.5185,
+      "step": 10999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.064022787085525,
+      "learning_rate": 5.834236030008455e-07,
+      "loss": 0.5138,
+      "step": 11000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8423978228749698,
+      "learning_rate": 5.825379120671649e-07,
+      "loss": 0.4502,
+      "step": 11001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8854906110383404,
+      "learning_rate": 5.816528737552496e-07,
+      "loss": 0.4603,
+      "step": 11002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9662435120986029,
+      "learning_rate": 5.807684881264397e-07,
+      "loss": 0.4929,
+      "step": 11003
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8918757606853558,
+      "learning_rate": 5.798847552420184e-07,
+      "loss": 0.4759,
+      "step": 11004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9781143034468275,
+      "learning_rate": 5.790016751632266e-07,
+      "loss": 0.529,
+      "step": 11005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9408427539253995,
+      "learning_rate": 5.781192479512621e-07,
+      "loss": 0.4402,
+      "step": 11006
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9169600366711169,
+      "learning_rate": 5.772374736672793e-07,
+      "loss": 0.4747,
+      "step": 11007
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9420777741526181,
+      "learning_rate": 5.763563523723769e-07,
+      "loss": 0.4748,
+      "step": 11008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9565403624864097,
+      "learning_rate": 5.754758841276243e-07,
+      "loss": 0.4469,
+      "step": 11009
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.991885868237849,
+      "learning_rate": 5.745960689940322e-07,
+      "loss": 0.4439,
+      "step": 11010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9151544363228036,
+      "learning_rate": 5.73716907032572e-07,
+      "loss": 0.4787,
+      "step": 11011
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8132533514402628,
+      "learning_rate": 5.728383983041696e-07,
+      "loss": 0.5255,
+      "step": 11012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9310339443213517,
+      "learning_rate": 5.719605428697051e-07,
+      "loss": 0.5024,
+      "step": 11013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8574936208717879,
+      "learning_rate": 5.710833407900096e-07,
+      "loss": 0.433,
+      "step": 11014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8741036577067072,
+      "learning_rate": 5.70206792125878e-07,
+      "loss": 0.4794,
+      "step": 11015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.967796820842037,
+      "learning_rate": 5.693308969380495e-07,
+      "loss": 0.479,
+      "step": 11016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9932310538013736,
+      "learning_rate": 5.684556552872256e-07,
+      "loss": 0.4961,
+      "step": 11017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9871267665159187,
+      "learning_rate": 5.675810672340587e-07,
+      "loss": 0.4598,
+      "step": 11018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.986056563416071,
+      "learning_rate": 5.667071328391593e-07,
+      "loss": 0.532,
+      "step": 11019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9476270127764385,
+      "learning_rate": 5.658338521630846e-07,
+      "loss": 0.4759,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9657676173085136,
+      "learning_rate": 5.649612252663583e-07,
+      "loss": 0.5025,
+      "step": 11021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8858286395635273,
+      "learning_rate": 5.640892522094499e-07,
+      "loss": 0.434,
+      "step": 11022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9552774366100201,
+      "learning_rate": 5.632179330527865e-07,
+      "loss": 0.4871,
+      "step": 11023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9143189633790539,
+      "learning_rate": 5.623472678567498e-07,
+      "loss": 0.4927,
+      "step": 11024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8714387741551404,
+      "learning_rate": 5.614772566816773e-07,
+      "loss": 0.4832,
+      "step": 11025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8858930867826196,
+      "learning_rate": 5.606078995878562e-07,
+      "loss": 0.4106,
+      "step": 11026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9568935096034147,
+      "learning_rate": 5.597391966355381e-07,
+      "loss": 0.4394,
+      "step": 11027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0728162873381937,
+      "learning_rate": 5.588711478849185e-07,
+      "loss": 0.5342,
+      "step": 11028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9180312152603842,
+      "learning_rate": 5.580037533961546e-07,
+      "loss": 0.4655,
+      "step": 11029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9517370360129557,
+      "learning_rate": 5.571370132293552e-07,
+      "loss": 0.4426,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9945553716176935,
+      "learning_rate": 5.562709274445866e-07,
+      "loss": 0.5017,
+      "step": 11031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8730934042078565,
+      "learning_rate": 5.554054961018628e-07,
+      "loss": 0.4803,
+      "step": 11032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9385185317818835,
+      "learning_rate": 5.54540719261164e-07,
+      "loss": 0.441,
+      "step": 11033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9249336081996793,
+      "learning_rate": 5.536765969824132e-07,
+      "loss": 0.4769,
+      "step": 11034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9050719472637894,
+      "learning_rate": 5.528131293254957e-07,
+      "loss": 0.4702,
+      "step": 11035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9136625775583729,
+      "learning_rate": 5.519503163502493e-07,
+      "loss": 0.4389,
+      "step": 11036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9086878888383371,
+      "learning_rate": 5.510881581164662e-07,
+      "loss": 0.3929,
+      "step": 11037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.92276744581274,
+      "learning_rate": 5.502266546838897e-07,
+      "loss": 0.5428,
+      "step": 11038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8805940036916493,
+      "learning_rate": 5.493658061122276e-07,
+      "loss": 0.4426,
+      "step": 11039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9771763240276137,
+      "learning_rate": 5.48505612461131e-07,
+      "loss": 0.4668,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8734235333940492,
+      "learning_rate": 5.476460737902111e-07,
+      "loss": 0.4037,
+      "step": 11041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0466353923440146,
+      "learning_rate": 5.467871901590349e-07,
+      "loss": 0.5033,
+      "step": 11042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9785685435203199,
+      "learning_rate": 5.459289616271224e-07,
+      "loss": 0.4902,
+      "step": 11043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9503035291633423,
+      "learning_rate": 5.450713882539449e-07,
+      "loss": 0.4766,
+      "step": 11044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0429553331877732,
+      "learning_rate": 5.44214470098936e-07,
+      "loss": 0.5244,
+      "step": 11045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.97021397045449,
+      "learning_rate": 5.43358207221476e-07,
+      "loss": 0.5031,
+      "step": 11046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9445804788065731,
+      "learning_rate": 5.425025996809042e-07,
+      "loss": 0.4727,
+      "step": 11047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9205907205100843,
+      "learning_rate": 5.416476475365129e-07,
+      "loss": 0.4863,
+      "step": 11048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3195746670830866,
+      "learning_rate": 5.407933508475515e-07,
+      "loss": 0.5233,
+      "step": 11049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9794249235949587,
+      "learning_rate": 5.399397096732184e-07,
+      "loss": 0.5045,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.894756142856474,
+      "learning_rate": 5.39086724072675e-07,
+      "loss": 0.3943,
+      "step": 11051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9204342660541922,
+      "learning_rate": 5.382343941050272e-07,
+      "loss": 0.471,
+      "step": 11052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9181966409748372,
+      "learning_rate": 5.373827198293446e-07,
+      "loss": 0.4943,
+      "step": 11053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9282760705343676,
+      "learning_rate": 5.365317013046456e-07,
+      "loss": 0.4279,
+      "step": 11054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9893124938884263,
+      "learning_rate": 5.356813385899074e-07,
+      "loss": 0.5161,
+      "step": 11055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9524158665869225,
+      "learning_rate": 5.348316317440549e-07,
+      "loss": 0.4846,
+      "step": 11056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9733809276947557,
+      "learning_rate": 5.339825808259779e-07,
+      "loss": 0.4934,
+      "step": 11057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9904644382388924,
+      "learning_rate": 5.331341858945094e-07,
+      "loss": 0.5697,
+      "step": 11058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9217495342550637,
+      "learning_rate": 5.322864470084455e-07,
+      "loss": 0.4578,
+      "step": 11059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.864877204773377,
+      "learning_rate": 5.314393642265314e-07,
+      "loss": 0.4253,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.981346430166337,
+      "learning_rate": 5.305929376074725e-07,
+      "loss": 0.5363,
+      "step": 11061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.972993618315643,
+      "learning_rate": 5.29747167209923e-07,
+      "loss": 0.4663,
+      "step": 11062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8667091636509169,
+      "learning_rate": 5.28902053092496e-07,
+      "loss": 0.4178,
+      "step": 11063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8573274759690804,
+      "learning_rate": 5.280575953137545e-07,
+      "loss": 0.4511,
+      "step": 11064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9976398282977128,
+      "learning_rate": 5.272137939322208e-07,
+      "loss": 0.487,
+      "step": 11065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9493121201956117,
+      "learning_rate": 5.26370649006368e-07,
+      "loss": 0.4588,
+      "step": 11066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8258826117106118,
+      "learning_rate": 5.255281605946261e-07,
+      "loss": 0.4348,
+      "step": 11067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9915494613066878,
+      "learning_rate": 5.246863287553794e-07,
+      "loss": 0.4855,
+      "step": 11068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9091035541846134,
+      "learning_rate": 5.238451535469658e-07,
+      "loss": 0.4527,
+      "step": 11069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8570386649899411,
+      "learning_rate": 5.230046350276774e-07,
+      "loss": 0.4344,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9304912751627228,
+      "learning_rate": 5.221647732557611e-07,
+      "loss": 0.5141,
+      "step": 11071
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0122734919257559,
+      "learning_rate": 5.213255682894192e-07,
+      "loss": 0.483,
+      "step": 11072
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9107296385416488,
+      "learning_rate": 5.204870201868084e-07,
+      "loss": 0.4708,
+      "step": 11073
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9662778251197135,
+      "learning_rate": 5.19649129006039e-07,
+      "loss": 0.5313,
+      "step": 11074
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9539025751341297,
+      "learning_rate": 5.188118948051768e-07,
+      "loss": 0.4911,
+      "step": 11075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9027565677142386,
+      "learning_rate": 5.179753176422386e-07,
+      "loss": 0.4842,
+      "step": 11076
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.100122923406752,
+      "learning_rate": 5.171393975752015e-07,
+      "loss": 0.5235,
+      "step": 11077
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9943585472020804,
+      "learning_rate": 5.163041346619913e-07,
+      "loss": 0.4866,
+      "step": 11078
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9150676597442815,
+      "learning_rate": 5.154695289604938e-07,
+      "loss": 0.4607,
+      "step": 11079
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8694083085530301,
+      "learning_rate": 5.146355805285452e-07,
+      "loss": 0.465,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.134853604728181,
+      "learning_rate": 5.138022894239369e-07,
+      "loss": 0.5574,
+      "step": 11081
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8853817334786239,
+      "learning_rate": 5.129696557044173e-07,
+      "loss": 0.4293,
+      "step": 11082
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.98048161606491,
+      "learning_rate": 5.121376794276834e-07,
+      "loss": 0.5062,
+      "step": 11083
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9456961383840721,
+      "learning_rate": 5.113063606513935e-07,
+      "loss": 0.4381,
+      "step": 11084
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9519112022167273,
+      "learning_rate": 5.104756994331561e-07,
+      "loss": 0.4312,
+      "step": 11085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.948613829358276,
+      "learning_rate": 5.096456958305351e-07,
+      "loss": 0.4807,
+      "step": 11086
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9351289986800121,
+      "learning_rate": 5.088163499010502e-07,
+      "loss": 0.5065,
+      "step": 11087
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8956220061900297,
+      "learning_rate": 5.07987661702174e-07,
+      "loss": 0.4606,
+      "step": 11088
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9545684634715169,
+      "learning_rate": 5.071596312913329e-07,
+      "loss": 0.4764,
+      "step": 11089
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.012292041949086,
+      "learning_rate": 5.06332258725909e-07,
+      "loss": 0.4834,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9546015528272861,
+      "learning_rate": 5.055055440632384e-07,
+      "loss": 0.4927,
+      "step": 11091
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8922578203787093,
+      "learning_rate": 5.046794873606131e-07,
+      "loss": 0.464,
+      "step": 11092
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.002581031774825,
+      "learning_rate": 5.038540886752752e-07,
+      "loss": 0.5291,
+      "step": 11093
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.956594306713076,
+      "learning_rate": 5.030293480644289e-07,
+      "loss": 0.5018,
+      "step": 11094
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0593410366903622,
+      "learning_rate": 5.022052655852228e-07,
+      "loss": 0.554,
+      "step": 11095
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9620155031743906,
+      "learning_rate": 5.013818412947669e-07,
+      "loss": 0.4683,
+      "step": 11096
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9369348023729992,
+      "learning_rate": 5.005590752501244e-07,
+      "loss": 0.4979,
+      "step": 11097
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9417310972907301,
+      "learning_rate": 4.997369675083131e-07,
+      "loss": 0.4617,
+      "step": 11098
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.934368976889229,
+      "learning_rate": 4.989155181263017e-07,
+      "loss": 0.487,
+      "step": 11099
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9313517760063982,
+      "learning_rate": 4.980947271610192e-07,
+      "loss": 0.4887,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9351666968503773,
+      "learning_rate": 4.972745946693414e-07,
+      "loss": 0.4626,
+      "step": 11101
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9262804697487227,
+      "learning_rate": 4.964551207081081e-07,
+      "loss": 0.4518,
+      "step": 11102
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9592488579633498,
+      "learning_rate": 4.95636305334104e-07,
+      "loss": 0.5235,
+      "step": 11103
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.99004950050435,
+      "learning_rate": 4.948181486040737e-07,
+      "loss": 0.4683,
+      "step": 11104
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9846886660796365,
+      "learning_rate": 4.940006505747142e-07,
+      "loss": 0.513,
+      "step": 11105
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9454931036011615,
+      "learning_rate": 4.931838113026798e-07,
+      "loss": 0.541,
+      "step": 11106
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.920001078733727,
+      "learning_rate": 4.923676308445713e-07,
+      "loss": 0.5109,
+      "step": 11107
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9192704813594433,
+      "learning_rate": 4.915521092569553e-07,
+      "loss": 0.4743,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.955057808863621,
+      "learning_rate": 4.907372465963434e-07,
+      "loss": 0.4339,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9254482314436216,
+      "learning_rate": 4.899230429192059e-07,
+      "loss": 0.466,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8747271004206972,
+      "learning_rate": 4.891094982819656e-07,
+      "loss": 0.4546,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9775214852283891,
+      "learning_rate": 4.882966127410016e-07,
+      "loss": 0.4646,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0915192222629713,
+      "learning_rate": 4.874843863526435e-07,
+      "loss": 0.5919,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.990407126815721,
+      "learning_rate": 4.866728191731829e-07,
+      "loss": 0.4546,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9709797991935482,
+      "learning_rate": 4.858619112588559e-07,
+      "loss": 0.4967,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9241489813877177,
+      "learning_rate": 4.850516626658585e-07,
+      "loss": 0.4624,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9445547169009616,
+      "learning_rate": 4.842420734503428e-07,
+      "loss": 0.48,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9599734450815728,
+      "learning_rate": 4.834331436684114e-07,
+      "loss": 0.4661,
+      "step": 11118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9320491380802918,
+      "learning_rate": 4.826248733761185e-07,
+      "loss": 0.5014,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.964224195814961,
+      "learning_rate": 4.818172626294837e-07,
+      "loss": 0.4735,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9579047810051492,
+      "learning_rate": 4.810103114844688e-07,
+      "loss": 0.4381,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8961668308564504,
+      "learning_rate": 4.802040199969959e-07,
+      "loss": 0.462,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9865262747521256,
+      "learning_rate": 4.793983882229402e-07,
+      "loss": 0.4953,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.106608268670239,
+      "learning_rate": 4.785934162181305e-07,
+      "loss": 0.572,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.928105720357525,
+      "learning_rate": 4.777891040383531e-07,
+      "loss": 0.4759,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9034973838841738,
+      "learning_rate": 4.769854517393447e-07,
+      "loss": 0.4502,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9056139419125202,
+      "learning_rate": 4.761824593767961e-07,
+      "loss": 0.4717,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.91672933848264,
+      "learning_rate": 4.753801270063574e-07,
+      "loss": 0.4453,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9868633642457375,
+      "learning_rate": 4.7457845468362627e-07,
+      "loss": 0.522,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9925761450136723,
+      "learning_rate": 4.7377744246415837e-07,
+      "loss": 0.5019,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8302684218548028,
+      "learning_rate": 4.7297709040346474e-07,
+      "loss": 0.466,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8809889972200473,
+      "learning_rate": 4.7217739855700995e-07,
+      "loss": 0.4608,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.949125194133086,
+      "learning_rate": 4.7137836698020747e-07,
+      "loss": 0.5073,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9828922076326778,
+      "learning_rate": 4.7057999572843516e-07,
+      "loss": 0.5245,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9807408745494204,
+      "learning_rate": 4.6978228485701437e-07,
+      "loss": 0.4997,
+      "step": 11135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8783668016785334,
+      "learning_rate": 4.689852344212287e-07,
+      "loss": 0.416,
+      "step": 11136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8545384196573856,
+      "learning_rate": 4.681888444763116e-07,
+      "loss": 0.4716,
+      "step": 11137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.006594980432949,
+      "learning_rate": 4.673931150774547e-07,
+      "loss": 0.5269,
+      "step": 11138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9954424066558223,
+      "learning_rate": 4.6659804627979697e-07,
+      "loss": 0.5066,
+      "step": 11139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9063227086751301,
+      "learning_rate": 4.658036381384412e-07,
+      "loss": 0.4942,
+      "step": 11140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.871552804947988,
+      "learning_rate": 4.650098907084355e-07,
+      "loss": 0.4441,
+      "step": 11141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0236407870819795,
+      "learning_rate": 4.6421680404478587e-07,
+      "loss": 0.5124,
+      "step": 11142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.824417457432014,
+      "learning_rate": 4.634243782024539e-07,
+      "loss": 0.4079,
+      "step": 11143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.133583472750711,
+      "learning_rate": 4.6263261323635455e-07,
+      "loss": 0.5321,
+      "step": 11144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8843697218228761,
+      "learning_rate": 4.6184150920135395e-07,
+      "loss": 0.4356,
+      "step": 11145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9000063294594467,
+      "learning_rate": 4.610510661522782e-07,
+      "loss": 0.4835,
+      "step": 11146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9434582772507041,
+      "learning_rate": 4.602612841439014e-07,
+      "loss": 0.5203,
+      "step": 11147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9692831264058703,
+      "learning_rate": 4.594721632309551e-07,
+      "loss": 0.4553,
+      "step": 11148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9482907368711121,
+      "learning_rate": 4.5868370346812685e-07,
+      "loss": 0.4821,
+      "step": 11149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.848294913758441,
+      "learning_rate": 4.5789590491005507e-07,
+      "loss": 0.433,
+      "step": 11150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9580137905855243,
+      "learning_rate": 4.571087676113306e-07,
+      "loss": 0.4262,
+      "step": 11151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9885469522737916,
+      "learning_rate": 4.5632229162650755e-07,
+      "loss": 0.4879,
+      "step": 11152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9540913556936466,
+      "learning_rate": 4.555364770100823e-07,
+      "loss": 0.4256,
+      "step": 11153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.901228943473374,
+      "learning_rate": 4.5475132381651356e-07,
+      "loss": 0.5005,
+      "step": 11154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8633625027678152,
+      "learning_rate": 4.5396683210021107e-07,
+      "loss": 0.4495,
+      "step": 11155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.975255504720242,
+      "learning_rate": 4.531830019155425e-07,
+      "loss": 0.5066,
+      "step": 11156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9191865185057262,
+      "learning_rate": 4.5239983331682e-07,
+      "loss": 0.4824,
+      "step": 11157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9105743658096604,
+      "learning_rate": 4.516173263583234e-07,
+      "loss": 0.475,
+      "step": 11158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0033483384909352,
+      "learning_rate": 4.50835481094275e-07,
+      "loss": 0.518,
+      "step": 11159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9174121202867608,
+      "learning_rate": 4.50054297578858e-07,
+      "loss": 0.5079,
+      "step": 11160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9138216995553686,
+      "learning_rate": 4.492737758662069e-07,
+      "loss": 0.4926,
+      "step": 11161
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8698376707960717,
+      "learning_rate": 4.484939160104129e-07,
+      "loss": 0.478,
+      "step": 11162
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0407956612094187,
+      "learning_rate": 4.4771471806551614e-07,
+      "loss": 0.5134,
+      "step": 11163
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9911654660963026,
+      "learning_rate": 4.469361820855189e-07,
+      "loss": 0.4979,
+      "step": 11164
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8965315448269694,
+      "learning_rate": 4.4615830812437035e-07,
+      "loss": 0.5161,
+      "step": 11165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9476143911761191,
+      "learning_rate": 4.4538109623597617e-07,
+      "loss": 0.4384,
+      "step": 11166
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9106293415533426,
+      "learning_rate": 4.446045464741966e-07,
+      "loss": 0.4845,
+      "step": 11167
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9679582344937153,
+      "learning_rate": 4.4382865889284756e-07,
+      "loss": 0.5291,
+      "step": 11168
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.003241904827064,
+      "learning_rate": 4.4305343354569483e-07,
+      "loss": 0.5008,
+      "step": 11169
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9200873262321915,
+      "learning_rate": 4.4227887048646335e-07,
+      "loss": 0.4728,
+      "step": 11170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8862449874934346,
+      "learning_rate": 4.4150496976882783e-07,
+      "loss": 0.4437,
+      "step": 11171
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8813859946917624,
+      "learning_rate": 4.407317314464199e-07,
+      "loss": 0.4773,
+      "step": 11172
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9116257945503873,
+      "learning_rate": 4.399591555728233e-07,
+      "loss": 0.457,
+      "step": 11173
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8952447845796265,
+      "learning_rate": 4.391872422015786e-07,
+      "loss": 0.4996,
+      "step": 11174
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0264398330218911,
+      "learning_rate": 4.384159913861752e-07,
+      "loss": 0.5262,
+      "step": 11175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8986090002767066,
+      "learning_rate": 4.3764540318006587e-07,
+      "loss": 0.5164,
+      "step": 11176
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8649641214293282,
+      "learning_rate": 4.368754776366457e-07,
+      "loss": 0.4174,
+      "step": 11177
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.880916340853826,
+      "learning_rate": 4.3610621480927315e-07,
+      "loss": 0.4489,
+      "step": 11178
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9090094229803269,
+      "learning_rate": 4.3533761475125666e-07,
+      "loss": 0.4797,
+      "step": 11179
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9788562410334294,
+      "learning_rate": 4.3456967751586143e-07,
+      "loss": 0.4673,
+      "step": 11180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0350007495095783,
+      "learning_rate": 4.338024031562993e-07,
+      "loss": 0.4968,
+      "step": 11181
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9862582443544196,
+      "learning_rate": 4.3303579172574884e-07,
+      "loss": 0.5074,
+      "step": 11182
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9358507276894988,
+      "learning_rate": 4.3226984327733093e-07,
+      "loss": 0.4914,
+      "step": 11183
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9753101507137873,
+      "learning_rate": 4.3150455786412526e-07,
+      "loss": 0.4711,
+      "step": 11184
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.944450308660901,
+      "learning_rate": 4.3073993553916726e-07,
+      "loss": 0.4919,
+      "step": 11185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8749586270650181,
+      "learning_rate": 4.2997597635544563e-07,
+      "loss": 0.4242,
+      "step": 11186
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9996020749611297,
+      "learning_rate": 4.292126803658969e-07,
+      "loss": 0.4821,
+      "step": 11187
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.863509502257761,
+      "learning_rate": 4.2845004762342325e-07,
+      "loss": 0.459,
+      "step": 11188
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8681959186479662,
+      "learning_rate": 4.27688078180869e-07,
+      "loss": 0.4579,
+      "step": 11189
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8916377298074757,
+      "learning_rate": 4.269267720910419e-07,
+      "loss": 0.4552,
+      "step": 11190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.913625486076061,
+      "learning_rate": 4.2616612940669657e-07,
+      "loss": 0.4839,
+      "step": 11191
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9608375155236945,
+      "learning_rate": 4.254061501805484e-07,
+      "loss": 0.5321,
+      "step": 11192
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9390229983486676,
+      "learning_rate": 4.246468344652599e-07,
+      "loss": 0.5257,
+      "step": 11193
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.028975729366445,
+      "learning_rate": 4.238881823134533e-07,
+      "loss": 0.4979,
+      "step": 11194
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.873398927746758,
+      "learning_rate": 4.2313019377770104e-07,
+      "loss": 0.435,
+      "step": 11195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9497373872488999,
+      "learning_rate": 4.223728689105322e-07,
+      "loss": 0.5039,
+      "step": 11196
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9519321022576744,
+      "learning_rate": 4.216162077644281e-07,
+      "loss": 0.5677,
+      "step": 11197
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8908282880627073,
+      "learning_rate": 4.208602103918258e-07,
+      "loss": 0.4711,
+      "step": 11198
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0498824109635534,
+      "learning_rate": 4.2010487684511105e-07,
+      "loss": 0.5055,
+      "step": 11199
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9252361422177733,
+      "learning_rate": 4.1935020717663423e-07,
+      "loss": 0.4992,
+      "step": 11200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9402917514441262,
+      "learning_rate": 4.1859620143868793e-07,
+      "loss": 0.4984,
+      "step": 11201
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0012578522166926,
+      "learning_rate": 4.178428596835271e-07,
+      "loss": 0.4871,
+      "step": 11202
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.872894891355588,
+      "learning_rate": 4.170901819633566e-07,
+      "loss": 0.4665,
+      "step": 11203
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.977140130196723,
+      "learning_rate": 4.1633816833033804e-07,
+      "loss": 0.497,
+      "step": 11204
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0056578141335417,
+      "learning_rate": 4.155868188365808e-07,
+      "loss": 0.4943,
+      "step": 11205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9423785526526659,
+      "learning_rate": 4.1483613353415775e-07,
+      "loss": 0.4813,
+      "step": 11206
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0023662451744835,
+      "learning_rate": 4.1408611247508723e-07,
+      "loss": 0.491,
+      "step": 11207
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0159130710754287,
+      "learning_rate": 4.133367557113477e-07,
+      "loss": 0.5373,
+      "step": 11208
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8312726422881775,
+      "learning_rate": 4.1258806329486644e-07,
+      "loss": 0.4175,
+      "step": 11209
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8091738993603446,
+      "learning_rate": 4.118400352775287e-07,
+      "loss": 0.4436,
+      "step": 11210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8804197075085122,
+      "learning_rate": 4.1109267171117184e-07,
+      "loss": 0.468,
+      "step": 11211
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9020380820861756,
+      "learning_rate": 4.103459726475889e-07,
+      "loss": 0.4497,
+      "step": 11212
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9761918417426805,
+      "learning_rate": 4.095999381385229e-07,
+      "loss": 0.5338,
+      "step": 11213
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8782319385368993,
+      "learning_rate": 4.088545682356748e-07,
+      "loss": 0.433,
+      "step": 11214
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.950696236898299,
+      "learning_rate": 4.0810986299069656e-07,
+      "loss": 0.436,
+      "step": 11215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9189322841765247,
+      "learning_rate": 4.0736582245519795e-07,
+      "loss": 0.4815,
+      "step": 11216
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.884327380500117,
+      "learning_rate": 4.066224466807389e-07,
+      "loss": 0.4702,
+      "step": 11217
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9787955704244248,
+      "learning_rate": 4.0587973571883596e-07,
+      "loss": 0.5555,
+      "step": 11218
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9858389648304431,
+      "learning_rate": 4.051376896209558e-07,
+      "loss": 0.5023,
+      "step": 11219
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9994956181583511,
+      "learning_rate": 4.0439630843852383e-07,
+      "loss": 0.5066,
+      "step": 11220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0228356795696085,
+      "learning_rate": 4.0365559222291684e-07,
+      "loss": 0.4589,
+      "step": 11221
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9459753139302353,
+      "learning_rate": 4.029155410254637e-07,
+      "loss": 0.4816,
+      "step": 11222
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.091750919560779,
+      "learning_rate": 4.021761548974523e-07,
+      "loss": 0.5559,
+      "step": 11223
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9724075012713982,
+      "learning_rate": 4.014374338901206e-07,
+      "loss": 0.4748,
+      "step": 11224
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.006268228578613,
+      "learning_rate": 4.0069937805466084e-07,
+      "loss": 0.4083,
+      "step": 11225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9153690149268117,
+      "learning_rate": 3.999619874422178e-07,
+      "loss": 0.4788,
+      "step": 11226
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9338564214866565,
+      "learning_rate": 3.99225262103895e-07,
+      "loss": 0.4906,
+      "step": 11227
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0277728033158247,
+      "learning_rate": 3.9848920209074606e-07,
+      "loss": 0.5594,
+      "step": 11228
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9886771525159102,
+      "learning_rate": 3.977538074537779e-07,
+      "loss": 0.494,
+      "step": 11229
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9988823203761128,
+      "learning_rate": 3.9701907824395426e-07,
+      "loss": 0.4573,
+      "step": 11230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9668732903658148,
+      "learning_rate": 3.9628501451219106e-07,
+      "loss": 0.5663,
+      "step": 11231
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0333664090017052,
+      "learning_rate": 3.9555161630935756e-07,
+      "loss": 0.5688,
+      "step": 11232
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.870329084161462,
+      "learning_rate": 3.9481888368627764e-07,
+      "loss": 0.4793,
+      "step": 11233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.926284708098634,
+      "learning_rate": 3.940868166937295e-07,
+      "loss": 0.4408,
+      "step": 11234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9604043534785126,
+      "learning_rate": 3.933554153824437e-07,
+      "loss": 0.5178,
+      "step": 11235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9889579675663253,
+      "learning_rate": 3.9262467980310747e-07,
+      "loss": 0.4689,
+      "step": 11236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.899283890650655,
+      "learning_rate": 3.918946100063603e-07,
+      "loss": 0.4952,
+      "step": 11237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0361637174713196,
+      "learning_rate": 3.9116520604279285e-07,
+      "loss": 0.4996,
+      "step": 11238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9260930406702779,
+      "learning_rate": 3.904364679629535e-07,
+      "loss": 0.442,
+      "step": 11239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9405497553842672,
+      "learning_rate": 3.897083958173431e-07,
+      "loss": 0.4688,
+      "step": 11240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.039178970007914,
+      "learning_rate": 3.889809896564167e-07,
+      "loss": 0.5402,
+      "step": 11241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8742429461856939,
+      "learning_rate": 3.88254249530583e-07,
+      "loss": 0.5077,
+      "step": 11242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9717294505105695,
+      "learning_rate": 3.8752817549020494e-07,
+      "loss": 0.4859,
+      "step": 11243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9651825344259987,
+      "learning_rate": 3.868027675855968e-07,
+      "loss": 0.5584,
+      "step": 11244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.949190123858509,
+      "learning_rate": 3.8607802586703045e-07,
+      "loss": 0.479,
+      "step": 11245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0633748862786825,
+      "learning_rate": 3.853539503847292e-07,
+      "loss": 0.5105,
+      "step": 11246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.907406015353476,
+      "learning_rate": 3.8463054118887064e-07,
+      "loss": 0.4661,
+      "step": 11247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9235434459247656,
+      "learning_rate": 3.8390779832958804e-07,
+      "loss": 0.4984,
+      "step": 11248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.871393155044762,
+      "learning_rate": 3.831857218569646e-07,
+      "loss": 0.4764,
+      "step": 11249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.002044778848067,
+      "learning_rate": 3.824643118210403e-07,
+      "loss": 0.4984,
+      "step": 11250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8642745302575283,
+      "learning_rate": 3.817435682718096e-07,
+      "loss": 0.4246,
+      "step": 11251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8914467494927035,
+      "learning_rate": 3.810234912592181e-07,
+      "loss": 0.4409,
+      "step": 11252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.916483430239664,
+      "learning_rate": 3.803040808331659e-07,
+      "loss": 0.481,
+      "step": 11253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.898092244042881,
+      "learning_rate": 3.7958533704350763e-07,
+      "loss": 0.4732,
+      "step": 11254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9126761753612629,
+      "learning_rate": 3.788672599400534e-07,
+      "loss": 0.4555,
+      "step": 11255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8453746843953241,
+      "learning_rate": 3.7814984957256327e-07,
+      "loss": 0.4511,
+      "step": 11256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9432585840238175,
+      "learning_rate": 3.7743310599075543e-07,
+      "loss": 0.5324,
+      "step": 11257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8631479672202498,
+      "learning_rate": 3.7671702924429677e-07,
+      "loss": 0.4059,
+      "step": 11258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.003445796403605,
+      "learning_rate": 3.7600161938281196e-07,
+      "loss": 0.4981,
+      "step": 11259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8847485019144521,
+      "learning_rate": 3.7528687645587924e-07,
+      "loss": 0.4643,
+      "step": 11260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9521515676316561,
+      "learning_rate": 3.745728005130289e-07,
+      "loss": 0.4825,
+      "step": 11261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9943019297322592,
+      "learning_rate": 3.7385939160374476e-07,
+      "loss": 0.4789,
+      "step": 11262
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0443854704039524,
+      "learning_rate": 3.731466497774683e-07,
+      "loss": 0.5328,
+      "step": 11263
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9224820193567518,
+      "learning_rate": 3.7243457508358784e-07,
+      "loss": 0.4358,
+      "step": 11264
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9718348345724628,
+      "learning_rate": 3.717231675714539e-07,
+      "loss": 0.5314,
+      "step": 11265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9328422043792436,
+      "learning_rate": 3.710124272903626e-07,
+      "loss": 0.5085,
+      "step": 11266
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8573445232214344,
+      "learning_rate": 3.7030235428956895e-07,
+      "loss": 0.4688,
+      "step": 11267
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8719481532054565,
+      "learning_rate": 3.6959294861828145e-07,
+      "loss": 0.464,
+      "step": 11268
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.022972115265359,
+      "learning_rate": 3.688842103256607e-07,
+      "loss": 0.5125,
+      "step": 11269
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.907652029442627,
+      "learning_rate": 3.681761394608197e-07,
+      "loss": 0.4468,
+      "step": 11270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9742138500191758,
+      "learning_rate": 3.674687360728313e-07,
+      "loss": 0.4637,
+      "step": 11271
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0138898185358978,
+      "learning_rate": 3.667620002107142e-07,
+      "loss": 0.4617,
+      "step": 11272
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9212464426233573,
+      "learning_rate": 3.660559319234447e-07,
+      "loss": 0.4477,
+      "step": 11273
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9109639141362548,
+      "learning_rate": 3.653505312599548e-07,
+      "loss": 0.4575,
+      "step": 11274
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9821477248340027,
+      "learning_rate": 3.646457982691287e-07,
+      "loss": 0.5487,
+      "step": 11275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9427187757868164,
+      "learning_rate": 3.639417329997996e-07,
+      "loss": 0.4833,
+      "step": 11276
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9306861527841653,
+      "learning_rate": 3.632383355007629e-07,
+      "loss": 0.4816,
+      "step": 11277
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9649502106841656,
+      "learning_rate": 3.6253560582076075e-07,
+      "loss": 0.4746,
+      "step": 11278
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9670849975896374,
+      "learning_rate": 3.6183354400849304e-07,
+      "loss": 0.5206,
+      "step": 11279
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0014901452286138,
+      "learning_rate": 3.6113215011261194e-07,
+      "loss": 0.5014,
+      "step": 11280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9168158272006056,
+      "learning_rate": 3.604314241817242e-07,
+      "loss": 0.4216,
+      "step": 11281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.011810912246533,
+      "learning_rate": 3.5973136626438644e-07,
+      "loss": 0.4856,
+      "step": 11282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.070538404968705,
+      "learning_rate": 3.5903197640911546e-07,
+      "loss": 0.5257,
+      "step": 11283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9009972646160386,
+      "learning_rate": 3.5833325466437697e-07,
+      "loss": 0.4777,
+      "step": 11284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8799512675526913,
+      "learning_rate": 3.576352010785911e-07,
+      "loss": 0.4593,
+      "step": 11285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9844801336103844,
+      "learning_rate": 3.5693781570013243e-07,
+      "loss": 0.4648,
+      "step": 11286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9068441044013625,
+      "learning_rate": 3.5624109857733234e-07,
+      "loss": 0.4397,
+      "step": 11287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.049042656285058,
+      "learning_rate": 3.555450497584667e-07,
+      "loss": 0.4826,
+      "step": 11288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9362852634910507,
+      "learning_rate": 3.548496692917769e-07,
+      "loss": 0.4907,
+      "step": 11289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9580446529346488,
+      "learning_rate": 3.541549572254488e-07,
+      "loss": 0.4849,
+      "step": 11290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9345954545496868,
+      "learning_rate": 3.5346091360762615e-07,
+      "loss": 0.513,
+      "step": 11291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9650252641059298,
+      "learning_rate": 3.52767538486406e-07,
+      "loss": 0.4729,
+      "step": 11292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9106790175650638,
+      "learning_rate": 3.520748319098399e-07,
+      "loss": 0.4588,
+      "step": 11293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8666227556153863,
+      "learning_rate": 3.513827939259273e-07,
+      "loss": 0.446,
+      "step": 11294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8980952689189213,
+      "learning_rate": 3.5069142458263093e-07,
+      "loss": 0.4815,
+      "step": 11295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0014655996836292,
+      "learning_rate": 3.500007239278591e-07,
+      "loss": 0.506,
+      "step": 11296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.985359380024748,
+      "learning_rate": 3.49310692009478e-07,
+      "loss": 0.5175,
+      "step": 11297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9953536043633036,
+      "learning_rate": 3.4862132887530485e-07,
+      "loss": 0.5532,
+      "step": 11298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8713096817640018,
+      "learning_rate": 3.4793263457311487e-07,
+      "loss": 0.4345,
+      "step": 11299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9010164983973311,
+      "learning_rate": 3.4724460915062874e-07,
+      "loss": 0.4683,
+      "step": 11300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8829612221593353,
+      "learning_rate": 3.4655725265553276e-07,
+      "loss": 0.4497,
+      "step": 11301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9468093798201783,
+      "learning_rate": 3.458705651354544e-07,
+      "loss": 0.4913,
+      "step": 11302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8907019359980703,
+      "learning_rate": 3.451845466379833e-07,
+      "loss": 0.4794,
+      "step": 11303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9557898281434373,
+      "learning_rate": 3.4449919721065815e-07,
+      "loss": 0.4513,
+      "step": 11304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8546099421797825,
+      "learning_rate": 3.4381451690097653e-07,
+      "loss": 0.4572,
+      "step": 11305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8776209132717654,
+      "learning_rate": 3.4313050575638164e-07,
+      "loss": 0.4674,
+      "step": 11306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0620298487621678,
+      "learning_rate": 3.4244716382427876e-07,
+      "loss": 0.4877,
+      "step": 11307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9600773262046887,
+      "learning_rate": 3.417644911520202e-07,
+      "loss": 0.4858,
+      "step": 11308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0172245807626483,
+      "learning_rate": 3.410824877869157e-07,
+      "loss": 0.5105,
+      "step": 11309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0641696939999097,
+      "learning_rate": 3.404011537762275e-07,
+      "loss": 0.5477,
+      "step": 11310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9414882028257976,
+      "learning_rate": 3.3972048916717127e-07,
+      "loss": 0.4423,
+      "step": 11311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9345099908540642,
+      "learning_rate": 3.3904049400691585e-07,
+      "loss": 0.5102,
+      "step": 11312
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0032701922871918,
+      "learning_rate": 3.3836116834258583e-07,
+      "loss": 0.491,
+      "step": 11313
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2780296973011878,
+      "learning_rate": 3.376825122212568e-07,
+      "loss": 0.5556,
+      "step": 11314
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0067969180363423,
+      "learning_rate": 3.37004525689959e-07,
+      "loss": 0.4622,
+      "step": 11315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9048930187275969,
+      "learning_rate": 3.3632720879567594e-07,
+      "loss": 0.5202,
+      "step": 11316
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9190783750855306,
+      "learning_rate": 3.356505615853478e-07,
+      "loss": 0.4788,
+      "step": 11317
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8700597041027842,
+      "learning_rate": 3.349745841058605e-07,
+      "loss": 0.4796,
+      "step": 11318
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9765173966674195,
+      "learning_rate": 3.3429927640406425e-07,
+      "loss": 0.4858,
+      "step": 11319
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0247910175502006,
+      "learning_rate": 3.336246385267528e-07,
+      "loss": 0.4862,
+      "step": 11320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8672758790970835,
+      "learning_rate": 3.3295067052068086e-07,
+      "loss": 0.4239,
+      "step": 11321
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9803320123619911,
+      "learning_rate": 3.322773724325523e-07,
+      "loss": 0.4698,
+      "step": 11322
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8993180142347001,
+      "learning_rate": 3.3160474430902756e-07,
+      "loss": 0.4318,
+      "step": 11323
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.022883190886677,
+      "learning_rate": 3.30932786196716e-07,
+      "loss": 0.5367,
+      "step": 11324
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8970949373639794,
+      "learning_rate": 3.302614981421881e-07,
+      "loss": 0.4721,
+      "step": 11325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9270048427254719,
+      "learning_rate": 3.2959088019196005e-07,
+      "loss": 0.4475,
+      "step": 11326
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0993593442790337,
+      "learning_rate": 3.2892093239250686e-07,
+      "loss": 0.5155,
+      "step": 11327
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9603091910994954,
+      "learning_rate": 3.282516547902548e-07,
+      "loss": 0.4498,
+      "step": 11328
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9606031793318405,
+      "learning_rate": 3.2758304743158554e-07,
+      "loss": 0.4966,
+      "step": 11329
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.990904867623026,
+      "learning_rate": 3.2691511036282875e-07,
+      "loss": 0.4707,
+      "step": 11330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9704452825773399,
+      "learning_rate": 3.262478436302774e-07,
+      "loss": 0.4255,
+      "step": 11331
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9581276075878125,
+      "learning_rate": 3.255812472801689e-07,
+      "loss": 0.5424,
+      "step": 11332
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9665561902448118,
+      "learning_rate": 3.2491532135869865e-07,
+      "loss": 0.5122,
+      "step": 11333
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8347893223320174,
+      "learning_rate": 3.24250065912014e-07,
+      "loss": 0.4394,
+      "step": 11334
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8701345936737639,
+      "learning_rate": 3.235854809862193e-07,
+      "loss": 0.4746,
+      "step": 11335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.982254247110148,
+      "learning_rate": 3.2292156662736554e-07,
+      "loss": 0.4859,
+      "step": 11336
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4109891620559623,
+      "learning_rate": 3.2225832288146577e-07,
+      "loss": 0.4721,
+      "step": 11337
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8886097420198746,
+      "learning_rate": 3.2159574979447996e-07,
+      "loss": 0.4585,
+      "step": 11338
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0076090882264177,
+      "learning_rate": 3.209338474123225e-07,
+      "loss": 0.4674,
+      "step": 11339
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8590134875014034,
+      "learning_rate": 3.2027261578086443e-07,
+      "loss": 0.4567,
+      "step": 11340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.041771758544706,
+      "learning_rate": 3.1961205494593027e-07,
+      "loss": 0.467,
+      "step": 11341
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9674539778793504,
+      "learning_rate": 3.1895216495329116e-07,
+      "loss": 0.4917,
+      "step": 11342
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9700115735131218,
+      "learning_rate": 3.1829294584868166e-07,
+      "loss": 0.4876,
+      "step": 11343
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9355995499943076,
+      "learning_rate": 3.1763439767778293e-07,
+      "loss": 0.4811,
+      "step": 11344
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9956711944594656,
+      "learning_rate": 3.1697652048623185e-07,
+      "loss": 0.5596,
+      "step": 11345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8720294164515175,
+      "learning_rate": 3.163193143196197e-07,
+      "loss": 0.3914,
+      "step": 11346
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9699170744496788,
+      "learning_rate": 3.156627792234901e-07,
+      "loss": 0.4809,
+      "step": 11347
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.922892098617445,
+      "learning_rate": 3.150069152433377e-07,
+      "loss": 0.4439,
+      "step": 11348
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9107802722920333,
+      "learning_rate": 3.143517224246184e-07,
+      "loss": 0.4472,
+      "step": 11349
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8922634635118099,
+      "learning_rate": 3.1369720081273147e-07,
+      "loss": 0.4981,
+      "step": 11350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9435902052682902,
+      "learning_rate": 3.1304335045303724e-07,
+      "loss": 0.4766,
+      "step": 11351
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8859616279603214,
+      "learning_rate": 3.1239017139084725e-07,
+      "loss": 0.4254,
+      "step": 11352
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8330376400127653,
+      "learning_rate": 3.1173766367142534e-07,
+      "loss": 0.4531,
+      "step": 11353
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.936319051976937,
+      "learning_rate": 3.110858273399886e-07,
+      "loss": 0.4438,
+      "step": 11354
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9737792784501018,
+      "learning_rate": 3.1043466244171204e-07,
+      "loss": 0.5205,
+      "step": 11355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9971314327878192,
+      "learning_rate": 3.097841690217174e-07,
+      "loss": 0.5359,
+      "step": 11356
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.938727495682044,
+      "learning_rate": 3.0913434712508406e-07,
+      "loss": 0.4843,
+      "step": 11357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9856444957341504,
+      "learning_rate": 3.0848519679684606e-07,
+      "loss": 0.5354,
+      "step": 11358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8538657256871016,
+      "learning_rate": 3.078367180819863e-07,
+      "loss": 0.4759,
+      "step": 11359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.956572623856092,
+      "learning_rate": 3.0718891102544556e-07,
+      "loss": 0.5073,
+      "step": 11360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8913295893657242,
+      "learning_rate": 3.0654177567211675e-07,
+      "loss": 0.4138,
+      "step": 11361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9338857544751735,
+      "learning_rate": 3.0589531206684397e-07,
+      "loss": 0.4947,
+      "step": 11362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9580188168889444,
+      "learning_rate": 3.05249520254427e-07,
+      "loss": 0.4949,
+      "step": 11363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9029820242373844,
+      "learning_rate": 3.046044002796189e-07,
+      "loss": 0.4515,
+      "step": 11364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0361265416089567,
+      "learning_rate": 3.039599521871273e-07,
+      "loss": 0.5229,
+      "step": 11365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9361873322359499,
+      "learning_rate": 3.0331617602160965e-07,
+      "loss": 0.509,
+      "step": 11366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9951052754963993,
+      "learning_rate": 3.026730718276805e-07,
+      "loss": 0.5192,
+      "step": 11367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9165161306116084,
+      "learning_rate": 3.020306396499062e-07,
+      "loss": 0.4799,
+      "step": 11368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9247733833045126,
+      "learning_rate": 3.0138887953280573e-07,
+      "loss": 0.4666,
+      "step": 11369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0017284450473003,
+      "learning_rate": 3.0074779152085345e-07,
+      "loss": 0.4935,
+      "step": 11370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8499168034190486,
+      "learning_rate": 3.00107375658476e-07,
+      "loss": 0.4418,
+      "step": 11371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9375159709601683,
+      "learning_rate": 2.9946763199005356e-07,
+      "loss": 0.5188,
+      "step": 11372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9849719113347157,
+      "learning_rate": 2.988285605599206e-07,
+      "loss": 0.4752,
+      "step": 11373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9466116530237808,
+      "learning_rate": 2.981901614123617e-07,
+      "loss": 0.513,
+      "step": 11374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9249846932225795,
+      "learning_rate": 2.9755243459162144e-07,
+      "loss": 0.475,
+      "step": 11375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1001174601861894,
+      "learning_rate": 2.9691538014189005e-07,
+      "loss": 0.5434,
+      "step": 11376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8908442139851055,
+      "learning_rate": 2.9627899810731666e-07,
+      "loss": 0.4922,
+      "step": 11377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.006276155178277,
+      "learning_rate": 2.956432885320004e-07,
+      "loss": 0.5509,
+      "step": 11378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.958713801071645,
+      "learning_rate": 2.9500825145999723e-07,
+      "loss": 0.4641,
+      "step": 11379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.925297448457726,
+      "learning_rate": 2.943738869353141e-07,
+      "loss": 0.4948,
+      "step": 11380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9610141487552996,
+      "learning_rate": 2.9374019500191255e-07,
+      "loss": 0.5046,
+      "step": 11381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0200439829001273,
+      "learning_rate": 2.9310717570370516e-07,
+      "loss": 0.5139,
+      "step": 11382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8904018710302055,
+      "learning_rate": 2.9247482908456027e-07,
+      "loss": 0.4644,
+      "step": 11383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9344880041590723,
+      "learning_rate": 2.918431551882994e-07,
+      "loss": 0.465,
+      "step": 11384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9531718916462352,
+      "learning_rate": 2.9121215405869653e-07,
+      "loss": 0.5138,
+      "step": 11385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8398539963659245,
+      "learning_rate": 2.905818257394799e-07,
+      "loss": 0.4241,
+      "step": 11386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8422635777860722,
+      "learning_rate": 2.899521702743313e-07,
+      "loss": 0.4459,
+      "step": 11387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9634311386263213,
+      "learning_rate": 2.8932318770688364e-07,
+      "loss": 0.5256,
+      "step": 11388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9472397347375905,
+      "learning_rate": 2.886948780807253e-07,
+      "loss": 0.4538,
+      "step": 11389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9592197845636875,
+      "learning_rate": 2.8806724143939814e-07,
+      "loss": 0.5103,
+      "step": 11390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.018974190459149,
+      "learning_rate": 2.874402778263974e-07,
+      "loss": 0.476,
+      "step": 11391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0259783487545626,
+      "learning_rate": 2.868139872851694e-07,
+      "loss": 0.5273,
+      "step": 11392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9251938911299553,
+      "learning_rate": 2.8618836985911837e-07,
+      "loss": 0.4512,
+      "step": 11393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9463515132362895,
+      "learning_rate": 2.8556342559159513e-07,
+      "loss": 0.519,
+      "step": 11394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9191038400094846,
+      "learning_rate": 2.849391545259106e-07,
+      "loss": 0.4614,
+      "step": 11395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0263031963280673,
+      "learning_rate": 2.843155567053246e-07,
+      "loss": 0.5306,
+      "step": 11396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8546711846548672,
+      "learning_rate": 2.8369263217305374e-07,
+      "loss": 0.442,
+      "step": 11397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9869733608981098,
+      "learning_rate": 2.830703809722646e-07,
+      "loss": 0.45,
+      "step": 11398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8344249591406251,
+      "learning_rate": 2.8244880314607924e-07,
+      "loss": 0.3998,
+      "step": 11399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.027015109040339,
+      "learning_rate": 2.8182789873757334e-07,
+      "loss": 0.4416,
+      "step": 11400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9066407958880655,
+      "learning_rate": 2.812076677897735e-07,
+      "loss": 0.4828,
+      "step": 11401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9570317674310846,
+      "learning_rate": 2.8058811034566094e-07,
+      "loss": 0.5149,
+      "step": 11402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9276312174699367,
+      "learning_rate": 2.7996922644817126e-07,
+      "loss": 0.5011,
+      "step": 11403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9758733729187016,
+      "learning_rate": 2.7935101614019354e-07,
+      "loss": 0.5011,
+      "step": 11404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0575231778547123,
+      "learning_rate": 2.7873347946456684e-07,
+      "loss": 0.4679,
+      "step": 11405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9270474171243843,
+      "learning_rate": 2.7811661646408915e-07,
+      "loss": 0.4469,
+      "step": 11406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9410766556350038,
+      "learning_rate": 2.7750042718150514e-07,
+      "loss": 0.4999,
+      "step": 11407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8715469726023198,
+      "learning_rate": 2.768849116595185e-07,
+      "loss": 0.434,
+      "step": 11408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9514148647505117,
+      "learning_rate": 2.762700699407828e-07,
+      "loss": 0.4589,
+      "step": 11409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9448864220490569,
+      "learning_rate": 2.7565590206790613e-07,
+      "loss": 0.4982,
+      "step": 11410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9840067038967264,
+      "learning_rate": 2.7504240808344906e-07,
+      "loss": 0.497,
+      "step": 11411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9206231070024818,
+      "learning_rate": 2.744295880299297e-07,
+      "loss": 0.451,
+      "step": 11412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.995856159062488,
+      "learning_rate": 2.7381744194980963e-07,
+      "loss": 0.5632,
+      "step": 11413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9625196276014706,
+      "learning_rate": 2.732059698855172e-07,
+      "loss": 0.5134,
+      "step": 11414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.948921370108036,
+      "learning_rate": 2.7259517187942174e-07,
+      "loss": 0.4777,
+      "step": 11415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9561701392693762,
+      "learning_rate": 2.7198504797385286e-07,
+      "loss": 0.4564,
+      "step": 11416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.013002742298461,
+      "learning_rate": 2.7137559821109104e-07,
+      "loss": 0.5106,
+      "step": 11417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9886559129055978,
+      "learning_rate": 2.7076682263337264e-07,
+      "loss": 0.5182,
+      "step": 11418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9297366408530876,
+      "learning_rate": 2.701587212828816e-07,
+      "loss": 0.4637,
+      "step": 11419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8628281199103401,
+      "learning_rate": 2.6955129420176193e-07,
+      "loss": 0.462,
+      "step": 11420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8915919614811436,
+      "learning_rate": 2.689445414321057e-07,
+      "loss": 0.4719,
+      "step": 11421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9077066766545857,
+      "learning_rate": 2.6833846301596246e-07,
+      "loss": 0.4627,
+      "step": 11422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9649476009742295,
+      "learning_rate": 2.6773305899533084e-07,
+      "loss": 0.5364,
+      "step": 11423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9958264945406542,
+      "learning_rate": 2.6712832941216735e-07,
+      "loss": 0.494,
+      "step": 11424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9259154281760127,
+      "learning_rate": 2.6652427430837513e-07,
+      "loss": 0.4736,
+      "step": 11425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9181811509348105,
+      "learning_rate": 2.659208937258195e-07,
+      "loss": 0.4899,
+      "step": 11426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9557135320962066,
+      "learning_rate": 2.653181877063105e-07,
+      "loss": 0.4539,
+      "step": 11427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9424434836374508,
+      "learning_rate": 2.6471615629161564e-07,
+      "loss": 0.4835,
+      "step": 11428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.006653551445918,
+      "learning_rate": 2.641147995234572e-07,
+      "loss": 0.451,
+      "step": 11429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.972800610895364,
+      "learning_rate": 2.6351411744350853e-07,
+      "loss": 0.4451,
+      "step": 11430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9801622688087769,
+      "learning_rate": 2.6291411009339184e-07,
+      "loss": 0.4533,
+      "step": 11431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9436460199587007,
+      "learning_rate": 2.623147775146939e-07,
+      "loss": 0.5323,
+      "step": 11432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9656819963627655,
+      "learning_rate": 2.617161197489426e-07,
+      "loss": 0.4455,
+      "step": 11433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8882390107869854,
+      "learning_rate": 2.61118136837627e-07,
+      "loss": 0.4535,
+      "step": 11434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9408173864700979,
+      "learning_rate": 2.605208288221861e-07,
+      "loss": 0.4936,
+      "step": 11435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9464672305317218,
+      "learning_rate": 2.599241957440135e-07,
+      "loss": 0.5004,
+      "step": 11436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9689635678352339,
+      "learning_rate": 2.593282376444539e-07,
+      "loss": 0.5032,
+      "step": 11437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9571939695204263,
+      "learning_rate": 2.587329545648076e-07,
+      "loss": 0.4825,
+      "step": 11438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0129788746581043,
+      "learning_rate": 2.581383465463272e-07,
+      "loss": 0.4993,
+      "step": 11439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9450247269430384,
+      "learning_rate": 2.5754441363021854e-07,
+      "loss": 0.4955,
+      "step": 11440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9431474803716476,
+      "learning_rate": 2.5695115585763985e-07,
+      "loss": 0.4658,
+      "step": 11441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9351517255823307,
+      "learning_rate": 2.5635857326970494e-07,
+      "loss": 0.4728,
+      "step": 11442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0523809194784852,
+      "learning_rate": 2.5576666590747647e-07,
+      "loss": 0.6033,
+      "step": 11443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9171374814882678,
+      "learning_rate": 2.5517543381197715e-07,
+      "loss": 0.4813,
+      "step": 11444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9422827978562496,
+      "learning_rate": 2.5458487702417544e-07,
+      "loss": 0.4849,
+      "step": 11445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9136321288515592,
+      "learning_rate": 2.539949955849985e-07,
+      "loss": 0.4322,
+      "step": 11446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9397003802292838,
+      "learning_rate": 2.5340578953532256e-07,
+      "loss": 0.4441,
+      "step": 11447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0308380015724954,
+      "learning_rate": 2.5281725891598166e-07,
+      "loss": 0.5122,
+      "step": 11448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8825546477677132,
+      "learning_rate": 2.522294037677564e-07,
+      "loss": 0.4516,
+      "step": 11449
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8923664008410982,
+      "learning_rate": 2.516422241313898e-07,
+      "loss": 0.4327,
+      "step": 11450
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9709155618170434,
+      "learning_rate": 2.5105572004756827e-07,
+      "loss": 0.4749,
+      "step": 11451
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9445659754416634,
+      "learning_rate": 2.504698915569392e-07,
+      "loss": 0.4851,
+      "step": 11452
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0358932911976195,
+      "learning_rate": 2.49884738700098e-07,
+      "loss": 0.4879,
+      "step": 11453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9421156716446982,
+      "learning_rate": 2.493002615175977e-07,
+      "loss": 0.5585,
+      "step": 11454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9830053855246939,
+      "learning_rate": 2.487164600499381e-07,
+      "loss": 0.5162,
+      "step": 11455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.852493323525767,
+      "learning_rate": 2.481333343375802e-07,
+      "loss": 0.466,
+      "step": 11456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9097985826103769,
+      "learning_rate": 2.475508844209318e-07,
+      "loss": 0.5315,
+      "step": 11457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.977593827825391,
+      "learning_rate": 2.469691103403571e-07,
+      "loss": 0.5219,
+      "step": 11458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8948840283251344,
+      "learning_rate": 2.463880121361717e-07,
+      "loss": 0.4955,
+      "step": 11459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9137935694279657,
+      "learning_rate": 2.4580758984864675e-07,
+      "loss": 0.4488,
+      "step": 11460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9623904350660415,
+      "learning_rate": 2.452278435180011e-07,
+      "loss": 0.4811,
+      "step": 11461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9768541913562684,
+      "learning_rate": 2.44648773184416e-07,
+      "loss": 0.5311,
+      "step": 11462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9811267247772674,
+      "learning_rate": 2.440703788880172e-07,
+      "loss": 0.5502,
+      "step": 11463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9739306300978986,
+      "learning_rate": 2.4349266066888697e-07,
+      "loss": 0.4836,
+      "step": 11464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.885085228057854,
+      "learning_rate": 2.4291561856706224e-07,
+      "loss": 0.4172,
+      "step": 11465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0019125523263028,
+      "learning_rate": 2.42339252622531e-07,
+      "loss": 0.4975,
+      "step": 11466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0379549781503792,
+      "learning_rate": 2.417635628752324e-07,
+      "loss": 0.5312,
+      "step": 11467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9748225775151601,
+      "learning_rate": 2.411885493650656e-07,
+      "loss": 0.5303,
+      "step": 11468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9013719123124777,
+      "learning_rate": 2.4061421213187553e-07,
+      "loss": 0.4759,
+      "step": 11469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9363216769145684,
+      "learning_rate": 2.4004055121546354e-07,
+      "loss": 0.4897,
+      "step": 11470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9190009329597287,
+      "learning_rate": 2.3946756665558457e-07,
+      "loss": 0.5444,
+      "step": 11471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9129701848377056,
+      "learning_rate": 2.3889525849194573e-07,
+      "loss": 0.4854,
+      "step": 11472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9861834302459811,
+      "learning_rate": 2.383236267642064e-07,
+      "loss": 0.5198,
+      "step": 11473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9487541856662814,
+      "learning_rate": 2.3775267151198268e-07,
+      "loss": 0.4662,
+      "step": 11474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9313717143200725,
+      "learning_rate": 2.3718239277483957e-07,
+      "loss": 0.4983,
+      "step": 11475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9914572028788192,
+      "learning_rate": 2.3661279059229547e-07,
+      "loss": 0.5113,
+      "step": 11476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9384206526654694,
+      "learning_rate": 2.3604386500382658e-07,
+      "loss": 0.4799,
+      "step": 11477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.91476640221212,
+      "learning_rate": 2.3547561604885693e-07,
+      "loss": 0.4808,
+      "step": 11478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.959476425363997,
+      "learning_rate": 2.34908043766765e-07,
+      "loss": 0.5028,
+      "step": 11479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8586439006338363,
+      "learning_rate": 2.3434114819688492e-07,
+      "loss": 0.4727,
+      "step": 11480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9686850870111136,
+      "learning_rate": 2.3377492937850077e-07,
+      "loss": 0.4618,
+      "step": 11481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9513867379474965,
+      "learning_rate": 2.332093873508512e-07,
+      "loss": 0.4792,
+      "step": 11482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9789415893310047,
+      "learning_rate": 2.3264452215312817e-07,
+      "loss": 0.4585,
+      "step": 11483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8857061877304372,
+      "learning_rate": 2.3208033382447703e-07,
+      "loss": 0.4137,
+      "step": 11484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1075940052741962,
+      "learning_rate": 2.315168224039932e-07,
+      "loss": 0.524,
+      "step": 11485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0648445676642262,
+      "learning_rate": 2.3095398793072988e-07,
+      "loss": 0.4903,
+      "step": 11486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9380650984956824,
+      "learning_rate": 2.3039183044368918e-07,
+      "loss": 0.4808,
+      "step": 11487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9195255391855791,
+      "learning_rate": 2.2983034998182997e-07,
+      "loss": 0.5148,
+      "step": 11488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9767623140075157,
+      "learning_rate": 2.292695465840611e-07,
+      "loss": 0.509,
+      "step": 11489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9511340960467927,
+      "learning_rate": 2.2870942028924592e-07,
+      "loss": 0.4871,
+      "step": 11490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9257448355190214,
+      "learning_rate": 2.2814997113620008e-07,
+      "loss": 0.4549,
+      "step": 11491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9765459552883505,
+      "learning_rate": 2.2759119916369475e-07,
+      "loss": 0.4934,
+      "step": 11492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9485530275142726,
+      "learning_rate": 2.2703310441045012e-07,
+      "loss": 0.5296,
+      "step": 11493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9788948524867847,
+      "learning_rate": 2.264756869151441e-07,
+      "loss": 0.4941,
+      "step": 11494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.976896256270598,
+      "learning_rate": 2.2591894671640246e-07,
+      "loss": 0.5197,
+      "step": 11495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9282784520344509,
+      "learning_rate": 2.253628838528088e-07,
+      "loss": 0.462,
+      "step": 11496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.978036939703617,
+      "learning_rate": 2.2480749836289672e-07,
+      "loss": 0.5081,
+      "step": 11497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9645700321451685,
+      "learning_rate": 2.2425279028515658e-07,
+      "loss": 0.4979,
+      "step": 11498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.904580208924153,
+      "learning_rate": 2.2369875965802424e-07,
+      "loss": 0.5012,
+      "step": 11499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0060872487309571,
+      "learning_rate": 2.231454065198979e-07,
+      "loss": 0.518,
+      "step": 11500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0372439651101641,
+      "learning_rate": 2.225927309091225e-07,
+      "loss": 0.545,
+      "step": 11501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7751415828819416,
+      "learning_rate": 2.220407328639973e-07,
+      "loss": 0.4071,
+      "step": 11502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9450030394120906,
+      "learning_rate": 2.2148941242277732e-07,
+      "loss": 0.45,
+      "step": 11503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9381623403458342,
+      "learning_rate": 2.2093876962366755e-07,
+      "loss": 0.4971,
+      "step": 11504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0936081725087992,
+      "learning_rate": 2.2038880450482635e-07,
+      "loss": 0.5285,
+      "step": 11505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9464753191272031,
+      "learning_rate": 2.1983951710436768e-07,
+      "loss": 0.5045,
+      "step": 11506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9439441443903467,
+      "learning_rate": 2.1929090746035442e-07,
+      "loss": 0.4977,
+      "step": 11507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9693561399893567,
+      "learning_rate": 2.1874297561080616e-07,
+      "loss": 0.4464,
+      "step": 11508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9700127600386009,
+      "learning_rate": 2.1819572159369362e-07,
+      "loss": 0.5189,
+      "step": 11509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9969850062131683,
+      "learning_rate": 2.1764914544694203e-07,
+      "loss": 0.5442,
+      "step": 11510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9328818708633978,
+      "learning_rate": 2.1710324720842556e-07,
+      "loss": 0.5314,
+      "step": 11511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9966265518299623,
+      "learning_rate": 2.1655802691597837e-07,
+      "loss": 0.4564,
+      "step": 11512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1030087220932174,
+      "learning_rate": 2.1601348460738136e-07,
+      "loss": 0.5491,
+      "step": 11513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0014104948023606,
+      "learning_rate": 2.1546962032037211e-07,
+      "loss": 0.4876,
+      "step": 11514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0001591152367717,
+      "learning_rate": 2.1492643409263826e-07,
+      "loss": 0.469,
+      "step": 11515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9063598650995034,
+      "learning_rate": 2.1438392596182522e-07,
+      "loss": 0.5024,
+      "step": 11516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.969634253237918,
+      "learning_rate": 2.1384209596552297e-07,
+      "loss": 0.5086,
+      "step": 11517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9050830053343426,
+      "learning_rate": 2.1330094414128588e-07,
+      "loss": 0.4827,
+      "step": 11518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9616622729265908,
+      "learning_rate": 2.1276047052661176e-07,
+      "loss": 0.4942,
+      "step": 11519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9292719969220156,
+      "learning_rate": 2.1222067515895618e-07,
+      "loss": 0.5031,
+      "step": 11520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8583985298289181,
+      "learning_rate": 2.1168155807572476e-07,
+      "loss": 0.4731,
+      "step": 11521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0368097064485176,
+      "learning_rate": 2.11143119314281e-07,
+      "loss": 0.4492,
+      "step": 11522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9604901260255988,
+      "learning_rate": 2.1060535891193502e-07,
+      "loss": 0.5228,
+      "step": 11523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9001466531715809,
+      "learning_rate": 2.1006827690595478e-07,
+      "loss": 0.4525,
+      "step": 11524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9317278582374193,
+      "learning_rate": 2.095318733335594e-07,
+      "loss": 0.4114,
+      "step": 11525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.933645590445262,
+      "learning_rate": 2.089961482319214e-07,
+      "loss": 0.4291,
+      "step": 11526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9756302117349039,
+      "learning_rate": 2.0846110163816547e-07,
+      "loss": 0.49,
+      "step": 11527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0351340753968052,
+      "learning_rate": 2.0792673358936978e-07,
+      "loss": 0.5204,
+      "step": 11528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.893173614049143,
+      "learning_rate": 2.0739304412256578e-07,
+      "loss": 0.4897,
+      "step": 11529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.93363362407905,
+      "learning_rate": 2.0686003327473837e-07,
+      "loss": 0.4362,
+      "step": 11530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.94204371654142,
+      "learning_rate": 2.0632770108282462e-07,
+      "loss": 0.4555,
+      "step": 11531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9595420828187516,
+      "learning_rate": 2.0579604758371286e-07,
+      "loss": 0.496,
+      "step": 11532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9189135525666059,
+      "learning_rate": 2.05265072814248e-07,
+      "loss": 0.4318,
+      "step": 11533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8494707099918426,
+      "learning_rate": 2.047347768112262e-07,
+      "loss": 0.4793,
+      "step": 11534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0353601474519885,
+      "learning_rate": 2.0420515961139475e-07,
+      "loss": 0.4841,
+      "step": 11535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8312928039244486,
+      "learning_rate": 2.0367622125145868e-07,
+      "loss": 0.5352,
+      "step": 11536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9498648164663277,
+      "learning_rate": 2.0314796176806984e-07,
+      "loss": 0.4937,
+      "step": 11537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9609294146325073,
+      "learning_rate": 2.0262038119783778e-07,
+      "loss": 0.5138,
+      "step": 11538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9108235617292304,
+      "learning_rate": 2.0209347957732328e-07,
+      "loss": 0.5079,
+      "step": 11539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8957003143894978,
+      "learning_rate": 2.0156725694303936e-07,
+      "loss": 0.4517,
+      "step": 11540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9433740502699994,
+      "learning_rate": 2.010417133314535e-07,
+      "loss": 0.4947,
+      "step": 11541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0002718530207013,
+      "learning_rate": 2.0051684877898547e-07,
+      "loss": 0.4633,
+      "step": 11542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9221570895181908,
+      "learning_rate": 1.999926633220084e-07,
+      "loss": 0.4623,
+      "step": 11543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9640475017965786,
+      "learning_rate": 1.9946915699684653e-07,
+      "loss": 0.4682,
+      "step": 11544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9504884884382505,
+      "learning_rate": 1.9894632983977868e-07,
+      "loss": 0.4907,
+      "step": 11545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9108308409096453,
+      "learning_rate": 1.9842418188703694e-07,
+      "loss": 0.4612,
+      "step": 11546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.967929844407366,
+      "learning_rate": 1.9790271317480458e-07,
+      "loss": 0.5103,
+      "step": 11547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9606433023604526,
+      "learning_rate": 1.973819237392205e-07,
+      "loss": 0.5287,
+      "step": 11548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8816364193756345,
+      "learning_rate": 1.968618136163747e-07,
+      "loss": 0.4454,
+      "step": 11549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.977224015587263,
+      "learning_rate": 1.9634238284230945e-07,
+      "loss": 0.4818,
+      "step": 11550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1635547112143767,
+      "learning_rate": 1.9582363145302152e-07,
+      "loss": 0.4756,
+      "step": 11551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.029802951948784,
+      "learning_rate": 1.9530555948445883e-07,
+      "loss": 0.5323,
+      "step": 11552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0420832717806676,
+      "learning_rate": 1.947881669725249e-07,
+      "loss": 0.4768,
+      "step": 11553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8901052369947996,
+      "learning_rate": 1.9427145395307322e-07,
+      "loss": 0.4326,
+      "step": 11554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0468425019933045,
+      "learning_rate": 1.9375542046191297e-07,
+      "loss": 0.5095,
+      "step": 11555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8997505720146292,
+      "learning_rate": 1.9324006653480332e-07,
+      "loss": 0.4905,
+      "step": 11556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9338220987399636,
+      "learning_rate": 1.927253922074579e-07,
+      "loss": 0.4623,
+      "step": 11557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8516812243067089,
+      "learning_rate": 1.9221139751554373e-07,
+      "loss": 0.4504,
+      "step": 11558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9289614121327456,
+      "learning_rate": 1.9169808249468125e-07,
+      "loss": 0.5018,
+      "step": 11559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9779634493064184,
+      "learning_rate": 1.9118544718044084e-07,
+      "loss": 0.5542,
+      "step": 11560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8994294029334284,
+      "learning_rate": 1.906734916083497e-07,
+      "loss": 0.4226,
+      "step": 11561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9682669746822895,
+      "learning_rate": 1.9016221581388272e-07,
+      "loss": 0.5273,
+      "step": 11562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9761134799360173,
+      "learning_rate": 1.8965161983247494e-07,
+      "loss": 0.5418,
+      "step": 11563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0555396355770037,
+      "learning_rate": 1.891417036995069e-07,
+      "loss": 0.5418,
+      "step": 11564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9106038553461743,
+      "learning_rate": 1.8863246745031704e-07,
+      "loss": 0.5275,
+      "step": 11565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9721292591901912,
+      "learning_rate": 1.881239111201949e-07,
+      "loss": 0.4808,
+      "step": 11566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2443193603745144,
+      "learning_rate": 1.876160347443823e-07,
+      "loss": 0.4499,
+      "step": 11567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9113977420502493,
+      "learning_rate": 1.8710883835807437e-07,
+      "loss": 0.4509,
+      "step": 11568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9157774007045484,
+      "learning_rate": 1.866023219964208e-07,
+      "loss": 0.4312,
+      "step": 11569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9780163504554356,
+      "learning_rate": 1.8609648569452132e-07,
+      "loss": 0.4702,
+      "step": 11570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9718901158746666,
+      "learning_rate": 1.8559132948743007e-07,
+      "loss": 0.4922,
+      "step": 11571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0308700763315264,
+      "learning_rate": 1.8508685341015465e-07,
+      "loss": 0.4986,
+      "step": 11572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8754438065720972,
+      "learning_rate": 1.845830574976548e-07,
+      "loss": 0.4633,
+      "step": 11573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9917172090708526,
+      "learning_rate": 1.8407994178484155e-07,
+      "loss": 0.4981,
+      "step": 11574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.897217849065791,
+      "learning_rate": 1.8357750630658367e-07,
+      "loss": 0.4333,
+      "step": 11575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8707656123011874,
+      "learning_rate": 1.830757510976966e-07,
+      "loss": 0.4689,
+      "step": 11576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9677192816664028,
+      "learning_rate": 1.8257467619295143e-07,
+      "loss": 0.5139,
+      "step": 11577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0527355591460235,
+      "learning_rate": 1.8207428162707374e-07,
+      "loss": 0.497,
+      "step": 11578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9315945499644502,
+      "learning_rate": 1.8157456743474133e-07,
+      "loss": 0.5082,
+      "step": 11579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8556914362013863,
+      "learning_rate": 1.8107553365057983e-07,
+      "loss": 0.4383,
+      "step": 11580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9714958630042003,
+      "learning_rate": 1.8057718030917714e-07,
+      "loss": 0.4837,
+      "step": 11581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7953942637319843,
+      "learning_rate": 1.8007950744506454e-07,
+      "loss": 0.4541,
+      "step": 11582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0624559488720686,
+      "learning_rate": 1.7958251509273106e-07,
+      "loss": 0.5234,
+      "step": 11583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0426144132894177,
+      "learning_rate": 1.790862032866203e-07,
+      "loss": 0.4869,
+      "step": 11584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8442765483827533,
+      "learning_rate": 1.7859057206112361e-07,
+      "loss": 0.3807,
+      "step": 11585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1444035809631279,
+      "learning_rate": 1.78095621450588e-07,
+      "loss": 0.5447,
+      "step": 11586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9206041439662114,
+      "learning_rate": 1.776013514893149e-07,
+      "loss": 0.4671,
+      "step": 11587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0077606054336101,
+      "learning_rate": 1.7710776221155578e-07,
+      "loss": 0.4941,
+      "step": 11588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8720283736829387,
+      "learning_rate": 1.7661485365151553e-07,
+      "loss": 0.4178,
+      "step": 11589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0005935424660755,
+      "learning_rate": 1.761226258433524e-07,
+      "loss": 0.5458,
+      "step": 11590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9880090308093082,
+      "learning_rate": 1.756310788211779e-07,
+      "loss": 0.4767,
+      "step": 11591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9555884247413868,
+      "learning_rate": 1.751402126190549e-07,
+      "loss": 0.5261,
+      "step": 11592
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9378522243429254,
+      "learning_rate": 1.7465002727100055e-07,
+      "loss": 0.4732,
+      "step": 11593
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0058916977963035,
+      "learning_rate": 1.741605228109844e-07,
+      "loss": 0.5062,
+      "step": 11594
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9147081497526133,
+      "learning_rate": 1.7367169927292925e-07,
+      "loss": 0.4483,
+      "step": 11595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8592595462485942,
+      "learning_rate": 1.7318355669070807e-07,
+      "loss": 0.4526,
+      "step": 11596
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.831385833810018,
+      "learning_rate": 1.7269609509815156e-07,
+      "loss": 0.43,
+      "step": 11597
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.924732035163592,
+      "learning_rate": 1.7220931452903712e-07,
+      "loss": 0.4925,
+      "step": 11598
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9140520056972847,
+      "learning_rate": 1.7172321501710109e-07,
+      "loss": 0.4782,
+      "step": 11599
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.975889917886747,
+      "learning_rate": 1.712377965960288e-07,
+      "loss": 0.5251,
+      "step": 11600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9777399777316796,
+      "learning_rate": 1.7075305929945775e-07,
+      "loss": 0.4418,
+      "step": 11601
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9104619502187835,
+      "learning_rate": 1.7026900316098217e-07,
+      "loss": 0.4639,
+      "step": 11602
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0389219870998976,
+      "learning_rate": 1.6978562821414635e-07,
+      "loss": 0.5252,
+      "step": 11603
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8893812909710748,
+      "learning_rate": 1.6930293449244573e-07,
+      "loss": 0.4562,
+      "step": 11604
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0139611356319098,
+      "learning_rate": 1.6882092202933242e-07,
+      "loss": 0.5476,
+      "step": 11605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9920808416502068,
+      "learning_rate": 1.683395908582097e-07,
+      "loss": 0.5021,
+      "step": 11606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9401684008955127,
+      "learning_rate": 1.6785894101243205e-07,
+      "loss": 0.4315,
+      "step": 11607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0326745169755602,
+      "learning_rate": 1.6737897252530832e-07,
+      "loss": 0.4319,
+      "step": 11608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4288513899528752,
+      "learning_rate": 1.6689968543010082e-07,
+      "loss": 0.4886,
+      "step": 11609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8585485958823579,
+      "learning_rate": 1.6642107976002186e-07,
+      "loss": 0.4083,
+      "step": 11610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9069449809793708,
+      "learning_rate": 1.659431555482416e-07,
+      "loss": 0.4786,
+      "step": 11611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1631691820206738,
+      "learning_rate": 1.6546591282787683e-07,
+      "loss": 0.5776,
+      "step": 11612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0616439313560415,
+      "learning_rate": 1.6498935163200114e-07,
+      "loss": 0.5342,
+      "step": 11613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8923630631274349,
+      "learning_rate": 1.6451347199364032e-07,
+      "loss": 0.517,
+      "step": 11614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9362423351584785,
+      "learning_rate": 1.6403827394577244e-07,
+      "loss": 0.4897,
+      "step": 11615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9151508211734752,
+      "learning_rate": 1.6356375752132558e-07,
+      "loss": 0.4951,
+      "step": 11616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8389818007257338,
+      "learning_rate": 1.630899227531868e-07,
+      "loss": 0.4103,
+      "step": 11617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9323848955910646,
+      "learning_rate": 1.6261676967419094e-07,
+      "loss": 0.468,
+      "step": 11618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8786132211991765,
+      "learning_rate": 1.621442983171262e-07,
+      "loss": 0.4523,
+      "step": 11619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8676920931406709,
+      "learning_rate": 1.616725087147364e-07,
+      "loss": 0.49,
+      "step": 11620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9795059798461183,
+      "learning_rate": 1.6120140089971536e-07,
+      "loss": 0.5262,
+      "step": 11621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0434011389872648,
+      "learning_rate": 1.6073097490470924e-07,
+      "loss": 0.5017,
+      "step": 11622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9551902553910484,
+      "learning_rate": 1.602612307623208e-07,
+      "loss": 0.4933,
+      "step": 11623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9575849488802687,
+      "learning_rate": 1.5979216850509848e-07,
+      "loss": 0.4581,
+      "step": 11624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.903798519803686,
+      "learning_rate": 1.5932378816555405e-07,
+      "loss": 0.4304,
+      "step": 11625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.903238053594976,
+      "learning_rate": 1.5885608977614043e-07,
+      "loss": 0.4708,
+      "step": 11626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0195498094816038,
+      "learning_rate": 1.5838907336927055e-07,
+      "loss": 0.493,
+      "step": 11627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.948691606472617,
+      "learning_rate": 1.5792273897730858e-07,
+      "loss": 0.5152,
+      "step": 11628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9576141491811475,
+      "learning_rate": 1.5745708663257199e-07,
+      "loss": 0.4619,
+      "step": 11629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9671330738151036,
+      "learning_rate": 1.5699211636732714e-07,
+      "loss": 0.4614,
+      "step": 11630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9491379972972037,
+      "learning_rate": 1.5652782821379942e-07,
+      "loss": 0.4687,
+      "step": 11631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8516708278279836,
+      "learning_rate": 1.5606422220416196e-07,
+      "loss": 0.4502,
+      "step": 11632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.922946174629565,
+      "learning_rate": 1.5560129837054127e-07,
+      "loss": 0.4468,
+      "step": 11633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.893755085344053,
+      "learning_rate": 1.551390567450195e-07,
+      "loss": 0.4817,
+      "step": 11634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0082700354791259,
+      "learning_rate": 1.5467749735962878e-07,
+      "loss": 0.4937,
+      "step": 11635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9353409160951671,
+      "learning_rate": 1.5421662024635353e-07,
+      "loss": 0.4265,
+      "step": 11636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9418824879588475,
+      "learning_rate": 1.5375642543713488e-07,
+      "loss": 0.4844,
+      "step": 11637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9828534306732699,
+      "learning_rate": 1.5329691296386174e-07,
+      "loss": 0.4977,
+      "step": 11638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.044063249037734,
+      "learning_rate": 1.5283808285837754e-07,
+      "loss": 0.5204,
+      "step": 11639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.027630071859994,
+      "learning_rate": 1.5237993515248017e-07,
+      "loss": 0.4924,
+      "step": 11640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9350323418549664,
+      "learning_rate": 1.519224698779198e-07,
+      "loss": 0.5401,
+      "step": 11641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9073285796594803,
+      "learning_rate": 1.5146568706639552e-07,
+      "loss": 0.4147,
+      "step": 11642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9303432068459472,
+      "learning_rate": 1.5100958674956424e-07,
+      "loss": 0.4919,
+      "step": 11643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8080326063767643,
+      "learning_rate": 1.5055416895903284e-07,
+      "loss": 0.413,
+      "step": 11644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8638585497713773,
+      "learning_rate": 1.500994337263606e-07,
+      "loss": 0.4604,
+      "step": 11645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0138512478569417,
+      "learning_rate": 1.4964538108306004e-07,
+      "loss": 0.4739,
+      "step": 11646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0237844054054825,
+      "learning_rate": 1.4919201106059932e-07,
+      "loss": 0.5367,
+      "step": 11647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9469968312258004,
+      "learning_rate": 1.4873932369039223e-07,
+      "loss": 0.4471,
+      "step": 11648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9348968855520209,
+      "learning_rate": 1.4828731900381366e-07,
+      "loss": 0.4464,
+      "step": 11649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9454783504849241,
+      "learning_rate": 1.478359970321852e-07,
+      "loss": 0.4805,
+      "step": 11650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9696086869071582,
+      "learning_rate": 1.47385357806783e-07,
+      "loss": 0.4803,
+      "step": 11651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9258787941079295,
+      "learning_rate": 1.4693540135883533e-07,
+      "loss": 0.4343,
+      "step": 11652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.859426434883943,
+      "learning_rate": 1.4648612771952618e-07,
+      "loss": 0.4454,
+      "step": 11653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9625867462394057,
+      "learning_rate": 1.4603753691998735e-07,
+      "loss": 0.4783,
+      "step": 11654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9656030188852532,
+      "learning_rate": 1.4558962899130724e-07,
+      "loss": 0.5484,
+      "step": 11655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9501134464292218,
+      "learning_rate": 1.4514240396452438e-07,
+      "loss": 0.4818,
+      "step": 11656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0097149440429334,
+      "learning_rate": 1.4469586187063289e-07,
+      "loss": 0.524,
+      "step": 11657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0225104022893932,
+      "learning_rate": 1.4425000274057577e-07,
+      "loss": 0.5142,
+      "step": 11658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8977241993375531,
+      "learning_rate": 1.4380482660525164e-07,
+      "loss": 0.4279,
+      "step": 11659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9465529505683918,
+      "learning_rate": 1.4336033349550916e-07,
+      "loss": 0.4816,
+      "step": 11660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9588676788041502,
+      "learning_rate": 1.429165234421548e-07,
+      "loss": 0.4907,
+      "step": 11661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9247383698200462,
+      "learning_rate": 1.424733964759406e-07,
+      "loss": 0.4896,
+      "step": 11662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9732645453172724,
+      "learning_rate": 1.420309526275776e-07,
+      "loss": 0.4748,
+      "step": 11663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8902318164715665,
+      "learning_rate": 1.4158919192772458e-07,
+      "loss": 0.4787,
+      "step": 11664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9836209649754226,
+      "learning_rate": 1.4114811440699706e-07,
+      "loss": 0.4465,
+      "step": 11665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9865694640029874,
+      "learning_rate": 1.4070772009595944e-07,
+      "loss": 0.446,
+      "step": 11666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1353592274720519,
+      "learning_rate": 1.4026800902513293e-07,
+      "loss": 0.5372,
+      "step": 11667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9894621895910385,
+      "learning_rate": 1.3982898122498755e-07,
+      "loss": 0.4594,
+      "step": 11668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8808762541092449,
+      "learning_rate": 1.3939063672594677e-07,
+      "loss": 0.415,
+      "step": 11669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9543222001061445,
+      "learning_rate": 1.389529755583885e-07,
+      "loss": 0.5108,
+      "step": 11670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9031544153763041,
+      "learning_rate": 1.3851599775264403e-07,
+      "loss": 0.4892,
+      "step": 11671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.051975022845536,
+      "learning_rate": 1.3807970333899133e-07,
+      "loss": 0.5505,
+      "step": 11672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9494252597569944,
+      "learning_rate": 1.3764409234766962e-07,
+      "loss": 0.4922,
+      "step": 11673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9844827083958534,
+      "learning_rate": 1.3720916480886359e-07,
+      "loss": 0.5427,
+      "step": 11674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0213755760482455,
+      "learning_rate": 1.367749207527147e-07,
+      "loss": 0.4659,
+      "step": 11675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1219645031442136,
+      "learning_rate": 1.3634136020931444e-07,
+      "loss": 0.5685,
+      "step": 11676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9198137919526792,
+      "learning_rate": 1.3590848320870874e-07,
+      "loss": 0.4975,
+      "step": 11677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9814407636972431,
+      "learning_rate": 1.35476289780897e-07,
+      "loss": 0.4589,
+      "step": 11678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9265380091276773,
+      "learning_rate": 1.3504477995582744e-07,
+      "loss": 0.4739,
+      "step": 11679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9235688603751667,
+      "learning_rate": 1.3461395376340502e-07,
+      "loss": 0.5288,
+      "step": 11680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9593894678972065,
+      "learning_rate": 1.3418381123348477e-07,
+      "loss": 0.4638,
+      "step": 11681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9690623318955943,
+      "learning_rate": 1.337543523958751e-07,
+      "loss": 0.5005,
+      "step": 11682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.846608089060443,
+      "learning_rate": 1.333255772803377e-07,
+      "loss": 0.408,
+      "step": 11683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.969082169503564,
+      "learning_rate": 1.3289748591658546e-07,
+      "loss": 0.4923,
+      "step": 11684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8407132879000558,
+      "learning_rate": 1.3247007833428694e-07,
+      "loss": 0.4794,
+      "step": 11685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0111315851398217,
+      "learning_rate": 1.320433545630584e-07,
+      "loss": 0.532,
+      "step": 11686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0221013452148755,
+      "learning_rate": 1.3161731463247284e-07,
+      "loss": 0.4515,
+      "step": 11687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8748986501368735,
+      "learning_rate": 1.3119195857205337e-07,
+      "loss": 0.4978,
+      "step": 11688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9497318881951801,
+      "learning_rate": 1.3076728641127857e-07,
+      "loss": 0.478,
+      "step": 11689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.915779655254726,
+      "learning_rate": 1.3034329817957603e-07,
+      "loss": 0.424,
+      "step": 11690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9553825667674619,
+      "learning_rate": 1.2991999390632892e-07,
+      "loss": 0.4347,
+      "step": 11691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9041932919429256,
+      "learning_rate": 1.2949737362087156e-07,
+      "loss": 0.4739,
+      "step": 11692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8968792184451714,
+      "learning_rate": 1.2907543735249163e-07,
+      "loss": 0.4658,
+      "step": 11693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0653933580624912,
+      "learning_rate": 1.286541851304268e-07,
+      "loss": 0.5547,
+      "step": 11694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9609039869721445,
+      "learning_rate": 1.282336169838727e-07,
+      "loss": 0.4894,
+      "step": 11695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.020385805816101,
+      "learning_rate": 1.278137329419715e-07,
+      "loss": 0.5107,
+      "step": 11696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8729444977310155,
+      "learning_rate": 1.2739453303382222e-07,
+      "loss": 0.4691,
+      "step": 11697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9736436538183029,
+      "learning_rate": 1.2697601728847596e-07,
+      "loss": 0.4704,
+      "step": 11698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.009484260023586,
+      "learning_rate": 1.2655818573493295e-07,
+      "loss": 0.4842,
+      "step": 11699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9396092829306968,
+      "learning_rate": 1.261410384021511e-07,
+      "loss": 0.5107,
+      "step": 11700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.872628805875551,
+      "learning_rate": 1.2572457531903614e-07,
+      "loss": 0.4342,
+      "step": 11701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9414232683288218,
+      "learning_rate": 1.2530879651444949e-07,
+      "loss": 0.4543,
+      "step": 11702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.864248669810779,
+      "learning_rate": 1.2489370201720473e-07,
+      "loss": 0.4684,
+      "step": 11703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9694194096349622,
+      "learning_rate": 1.2447929185606778e-07,
+      "loss": 0.4468,
+      "step": 11704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0015075266884892,
+      "learning_rate": 1.2406556605975673e-07,
+      "loss": 0.5852,
+      "step": 11705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9396979059242895,
+      "learning_rate": 1.2365252465694088e-07,
+      "loss": 0.4968,
+      "step": 11706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9359777283532162,
+      "learning_rate": 1.2324016767624515e-07,
+      "loss": 0.4768,
+      "step": 11707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0363612056953866,
+      "learning_rate": 1.228284951462444e-07,
+      "loss": 0.5069,
+      "step": 11708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9450225369152239,
+      "learning_rate": 1.2241750709546918e-07,
+      "loss": 0.4555,
+      "step": 11709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.945398340483411,
+      "learning_rate": 1.2200720355239893e-07,
+      "loss": 0.4209,
+      "step": 11710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9641827946918541,
+      "learning_rate": 1.2159758454546643e-07,
+      "loss": 0.465,
+      "step": 11711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0174498372544294,
+      "learning_rate": 1.2118865010306124e-07,
+      "loss": 0.5257,
+      "step": 11712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8686355378031086,
+      "learning_rate": 1.2078040025351844e-07,
+      "loss": 0.4278,
+      "step": 11713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1133168236049928,
+      "learning_rate": 1.2037283502513208e-07,
+      "loss": 0.4626,
+      "step": 11714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9425915566814234,
+      "learning_rate": 1.1996595444614511e-07,
+      "loss": 0.4682,
+      "step": 11715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8035223312922553,
+      "learning_rate": 1.1955975854475388e-07,
+      "loss": 0.479,
+      "step": 11716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8922305833970967,
+      "learning_rate": 1.1915424734910585e-07,
+      "loss": 0.4497,
+      "step": 11717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9559911600741078,
+      "learning_rate": 1.1874942088730635e-07,
+      "loss": 0.4816,
+      "step": 11718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9758579168957457,
+      "learning_rate": 1.1834527918740624e-07,
+      "loss": 0.5539,
+      "step": 11719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0506259051311835,
+      "learning_rate": 1.1794182227741314e-07,
+      "loss": 0.4521,
+      "step": 11720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9705461383571655,
+      "learning_rate": 1.1753905018528688e-07,
+      "loss": 0.5648,
+      "step": 11721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0157017913842958,
+      "learning_rate": 1.171369629389385e-07,
+      "loss": 0.5057,
+      "step": 11722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9483830944456155,
+      "learning_rate": 1.1673556056623237e-07,
+      "loss": 0.4671,
+      "step": 11723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9072532108846191,
+      "learning_rate": 1.1633484309498511e-07,
+      "loss": 0.4809,
+      "step": 11724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9568152372299135,
+      "learning_rate": 1.1593481055296673e-07,
+      "loss": 0.4797,
+      "step": 11725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8821466617527249,
+      "learning_rate": 1.1553546296789952e-07,
+      "loss": 0.4821,
+      "step": 11726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9175803795045503,
+      "learning_rate": 1.1513680036745578e-07,
+      "loss": 0.4646,
+      "step": 11727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0322661343458974,
+      "learning_rate": 1.1473882277926562e-07,
+      "loss": 0.4446,
+      "step": 11728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0408277460676767,
+      "learning_rate": 1.1434153023090589e-07,
+      "loss": 0.5255,
+      "step": 11729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.027227376134949,
+      "learning_rate": 1.1394492274991009e-07,
+      "loss": 0.5529,
+      "step": 11730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9732458654486648,
+      "learning_rate": 1.1354900036376181e-07,
+      "loss": 0.5195,
+      "step": 11731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0535313265351196,
+      "learning_rate": 1.13153763099898e-07,
+      "loss": 0.5504,
+      "step": 11732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9198569397489443,
+      "learning_rate": 1.1275921098570896e-07,
+      "loss": 0.4748,
+      "step": 11733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.917712254839116,
+      "learning_rate": 1.1236534404853727e-07,
+      "loss": 0.4632,
+      "step": 11734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.856909983032463,
+      "learning_rate": 1.1197216231567664e-07,
+      "loss": 0.4514,
+      "step": 11735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9469785652163093,
+      "learning_rate": 1.1157966581437419e-07,
+      "loss": 0.4618,
+      "step": 11736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9599933656874255,
+      "learning_rate": 1.1118785457183034e-07,
+      "loss": 0.4708,
+      "step": 11737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.094933152970052,
+      "learning_rate": 1.1079672861519675e-07,
+      "loss": 0.56,
+      "step": 11738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.073555695579005,
+      "learning_rate": 1.1040628797157727e-07,
+      "loss": 0.4836,
+      "step": 11739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9336131282070538,
+      "learning_rate": 1.1001653266803136e-07,
+      "loss": 0.4616,
+      "step": 11740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9517994592150936,
+      "learning_rate": 1.0962746273156633e-07,
+      "loss": 0.5279,
+      "step": 11741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9092708185894586,
+      "learning_rate": 1.0923907818914614e-07,
+      "loss": 0.4783,
+      "step": 11742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9137401074747392,
+      "learning_rate": 1.0885137906768373e-07,
+      "loss": 0.4915,
+      "step": 11743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.886062036636107,
+      "learning_rate": 1.084643653940487e-07,
+      "loss": 0.41,
+      "step": 11744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0168251352297248,
+      "learning_rate": 1.0807803719505849e-07,
+      "loss": 0.4606,
+      "step": 11745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8764053557786847,
+      "learning_rate": 1.0769239449748614e-07,
+      "loss": 0.4161,
+      "step": 11746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9543498410492781,
+      "learning_rate": 1.0730743732805581e-07,
+      "loss": 0.4976,
+      "step": 11747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8988669209801309,
+      "learning_rate": 1.0692316571344619e-07,
+      "loss": 0.4746,
+      "step": 11748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9705097304261154,
+      "learning_rate": 1.0653957968028594e-07,
+      "loss": 0.531,
+      "step": 11749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8087770163850684,
+      "learning_rate": 1.0615667925515716e-07,
+      "loss": 0.4503,
+      "step": 11750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8585770296241498,
+      "learning_rate": 1.0577446446459416e-07,
+      "loss": 0.394,
+      "step": 11751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8991131732136636,
+      "learning_rate": 1.0539293533508577e-07,
+      "loss": 0.4604,
+      "step": 11752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.043978687288304,
+      "learning_rate": 1.0501209189306972e-07,
+      "loss": 0.5689,
+      "step": 11753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.012373658397403,
+      "learning_rate": 1.0463193416493933e-07,
+      "loss": 0.5616,
+      "step": 11754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9632718766958349,
+      "learning_rate": 1.04252462177038e-07,
+      "loss": 0.4913,
+      "step": 11755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.952867772462053,
+      "learning_rate": 1.0387367595566355e-07,
+      "loss": 0.4671,
+      "step": 11756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9312195502524956,
+      "learning_rate": 1.0349557552706613e-07,
+      "loss": 0.5035,
+      "step": 11757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9671971360550594,
+      "learning_rate": 1.0311816091744698e-07,
+      "loss": 0.4186,
+      "step": 11758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9494946903358517,
+      "learning_rate": 1.0274143215296073e-07,
+      "loss": 0.4798,
+      "step": 11759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8579853360947254,
+      "learning_rate": 1.0236538925971429e-07,
+      "loss": 0.4214,
+      "step": 11760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9821881441558292,
+      "learning_rate": 1.019900322637668e-07,
+      "loss": 0.4483,
+      "step": 11761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.919312383287609,
+      "learning_rate": 1.016153611911308e-07,
+      "loss": 0.5016,
+      "step": 11762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9690368869195982,
+      "learning_rate": 1.0124137606777107e-07,
+      "loss": 0.4951,
+      "step": 11763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9219625016430492,
+      "learning_rate": 1.0086807691960243e-07,
+      "loss": 0.4489,
+      "step": 11764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9490039649323291,
+      "learning_rate": 1.0049546377249642e-07,
+      "loss": 0.5141,
+      "step": 11765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0328827891759012,
+      "learning_rate": 1.0012353665227458e-07,
+      "loss": 0.4956,
+      "step": 11766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8300609810503681,
+      "learning_rate": 9.975229558470967e-08,
+      "loss": 0.4694,
+      "step": 11767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9583215834991651,
+      "learning_rate": 9.938174059552885e-08,
+      "loss": 0.4489,
+      "step": 11768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1407787654736952,
+      "learning_rate": 9.901187171041271e-08,
+      "loss": 0.5123,
+      "step": 11769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9188810440416877,
+      "learning_rate": 9.864268895499074e-08,
+      "loss": 0.475,
+      "step": 11770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9768542214641207,
+      "learning_rate": 9.827419235484803e-08,
+      "loss": 0.5083,
+      "step": 11771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.971834748912786,
+      "learning_rate": 9.790638193552082e-08,
+      "loss": 0.4618,
+      "step": 11772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9809466241068123,
+      "learning_rate": 9.753925772249873e-08,
+      "loss": 0.5157,
+      "step": 11773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.046921175293461,
+      "learning_rate": 9.71728197412225e-08,
+      "loss": 0.5045,
+      "step": 11774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9619930134152861,
+      "learning_rate": 9.680706801708517e-08,
+      "loss": 0.4513,
+      "step": 11775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0247706813665163,
+      "learning_rate": 9.644200257543534e-08,
+      "loss": 0.5578,
+      "step": 11776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9856133551445193,
+      "learning_rate": 9.607762344156946e-08,
+      "loss": 0.5058,
+      "step": 11777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9145922251023627,
+      "learning_rate": 9.571393064073953e-08,
+      "loss": 0.4999,
+      "step": 11778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9405462474920527,
+      "learning_rate": 9.535092419814873e-08,
+      "loss": 0.4495,
+      "step": 11779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0159098070267714,
+      "learning_rate": 9.498860413895472e-08,
+      "loss": 0.5144,
+      "step": 11780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8960909471248009,
+      "learning_rate": 9.462697048826408e-08,
+      "loss": 0.4778,
+      "step": 11781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9080420987831723,
+      "learning_rate": 9.426602327113788e-08,
+      "loss": 0.4559,
+      "step": 11782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9544893878383572,
+      "learning_rate": 9.390576251258943e-08,
+      "loss": 0.4975,
+      "step": 11783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9637330708653101,
+      "learning_rate": 9.354618823758654e-08,
+      "loss": 0.4813,
+      "step": 11784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9727251888984069,
+      "learning_rate": 9.318730047104484e-08,
+      "loss": 0.4482,
+      "step": 11785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0301592271271238,
+      "learning_rate": 9.282909923783557e-08,
+      "loss": 0.4473,
+      "step": 11786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9666339201017452,
+      "learning_rate": 9.247158456278327e-08,
+      "loss": 0.4611,
+      "step": 11787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9457148871860079,
+      "learning_rate": 9.211475647066148e-08,
+      "loss": 0.4772,
+      "step": 11788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9110564036623252,
+      "learning_rate": 9.175861498619821e-08,
+      "loss": 0.4615,
+      "step": 11789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9682975620629547,
+      "learning_rate": 9.140316013407479e-08,
+      "loss": 0.4591,
+      "step": 11790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0789231927916287,
+      "learning_rate": 9.104839193892379e-08,
+      "loss": 0.4509,
+      "step": 11791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.07751499957963,
+      "learning_rate": 9.069431042532995e-08,
+      "loss": 0.5149,
+      "step": 11792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9618200013702595,
+      "learning_rate": 9.034091561783032e-08,
+      "loss": 0.541,
+      "step": 11793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.033258536988785,
+      "learning_rate": 8.99882075409153e-08,
+      "loss": 0.5121,
+      "step": 11794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7940648088676413,
+      "learning_rate": 8.963618621902759e-08,
+      "loss": 0.4257,
+      "step": 11795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9676953589118892,
+      "learning_rate": 8.928485167656208e-08,
+      "loss": 0.482,
+      "step": 11796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9316837620156098,
+      "learning_rate": 8.89342039378649e-08,
+      "loss": 0.4366,
+      "step": 11797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9919385490875576,
+      "learning_rate": 8.858424302723767e-08,
+      "loss": 0.4911,
+      "step": 11798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8620054709348294,
+      "learning_rate": 8.823496896892991e-08,
+      "loss": 0.4312,
+      "step": 11799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9563793794175025,
+      "learning_rate": 8.78863817871467e-08,
+      "loss": 0.4966,
+      "step": 11800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9019147172069846,
+      "learning_rate": 8.753848150604538e-08,
+      "loss": 0.457,
+      "step": 11801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.007078630021055,
+      "learning_rate": 8.719126814973556e-08,
+      "loss": 0.491,
+      "step": 11802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9815358823107124,
+      "learning_rate": 8.684474174227797e-08,
+      "loss": 0.5185,
+      "step": 11803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9684981077551171,
+      "learning_rate": 8.649890230768676e-08,
+      "loss": 0.4662,
+      "step": 11804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8355804604535023,
+      "learning_rate": 8.615374986992831e-08,
+      "loss": 0.4454,
+      "step": 11805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9829781903898468,
+      "learning_rate": 8.580928445292124e-08,
+      "loss": 0.5163,
+      "step": 11806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9820450700426611,
+      "learning_rate": 8.54655060805365e-08,
+      "loss": 0.5061,
+      "step": 11807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9170387812859044,
+      "learning_rate": 8.512241477659944e-08,
+      "loss": 0.4526,
+      "step": 11808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8588684291904626,
+      "learning_rate": 8.478001056488327e-08,
+      "loss": 0.4278,
+      "step": 11809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.004476225320024,
+      "learning_rate": 8.443829346911792e-08,
+      "loss": 0.4339,
+      "step": 11810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9955030591553591,
+      "learning_rate": 8.409726351298441e-08,
+      "loss": 0.4798,
+      "step": 11811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0319475619112393,
+      "learning_rate": 8.375692072011388e-08,
+      "loss": 0.5098,
+      "step": 11812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9512626411668296,
+      "learning_rate": 8.341726511409409e-08,
+      "loss": 0.4742,
+      "step": 11813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8647860949043975,
+      "learning_rate": 8.307829671846179e-08,
+      "loss": 0.4773,
+      "step": 11814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.966937294863239,
+      "learning_rate": 8.274001555670597e-08,
+      "loss": 0.5045,
+      "step": 11815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0511562530909382,
+      "learning_rate": 8.24024216522723e-08,
+      "loss": 0.5244,
+      "step": 11816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0529033296786536,
+      "learning_rate": 8.20655150285521e-08,
+      "loss": 0.4635,
+      "step": 11817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9419210827460939,
+      "learning_rate": 8.172929570889553e-08,
+      "loss": 0.4529,
+      "step": 11818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.057298308091194,
+      "learning_rate": 8.139376371660179e-08,
+      "loss": 0.5259,
+      "step": 11819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9911938418309101,
+      "learning_rate": 8.105891907492224e-08,
+      "loss": 0.5373,
+      "step": 11820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9799078980237341,
+      "learning_rate": 8.072476180705946e-08,
+      "loss": 0.5207,
+      "step": 11821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9169456394142715,
+      "learning_rate": 8.03912919361749e-08,
+      "loss": 0.4771,
+      "step": 11822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9701500291611556,
+      "learning_rate": 8.005850948537453e-08,
+      "loss": 0.4812,
+      "step": 11823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9376443184836865,
+      "learning_rate": 7.972641447771989e-08,
+      "loss": 0.4849,
+      "step": 11824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8836556164802101,
+      "learning_rate": 7.939500693622481e-08,
+      "loss": 0.4832,
+      "step": 11825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0238583224039954,
+      "learning_rate": 7.906428688385759e-08,
+      "loss": 0.5001,
+      "step": 11826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9368949132542735,
+      "learning_rate": 7.873425434353432e-08,
+      "loss": 0.4871,
+      "step": 11827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9951882378946979,
+      "learning_rate": 7.840490933812783e-08,
+      "loss": 0.4845,
+      "step": 11828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8502504258314992,
+      "learning_rate": 7.807625189046098e-08,
+      "loss": 0.4095,
+      "step": 11829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9569555972363435,
+      "learning_rate": 7.774828202330776e-08,
+      "loss": 0.4657,
+      "step": 11830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9069133566490915,
+      "learning_rate": 7.742099975939888e-08,
+      "loss": 0.4634,
+      "step": 11831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9656614654654957,
+      "learning_rate": 7.709440512141286e-08,
+      "loss": 0.5422,
+      "step": 11832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9992615048473477,
+      "learning_rate": 7.676849813198272e-08,
+      "loss": 0.5122,
+      "step": 11833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.928341283376421,
+      "learning_rate": 7.644327881369485e-08,
+      "loss": 0.4389,
+      "step": 11834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0356863433018246,
+      "learning_rate": 7.611874718908452e-08,
+      "loss": 0.5778,
+      "step": 11835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9279200145760254,
+      "learning_rate": 7.579490328064265e-08,
+      "loss": 0.4886,
+      "step": 11836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0535959759120828,
+      "learning_rate": 7.547174711081128e-08,
+      "loss": 0.4395,
+      "step": 11837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9987245906603993,
+      "learning_rate": 7.514927870198475e-08,
+      "loss": 0.5428,
+      "step": 11838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9292302350613081,
+      "learning_rate": 7.482749807650958e-08,
+      "loss": 0.4683,
+      "step": 11839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.999598281395877,
+      "learning_rate": 7.450640525668573e-08,
+      "loss": 0.5147,
+      "step": 11840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9016518616157955,
+      "learning_rate": 7.41860002647643e-08,
+      "loss": 0.4539,
+      "step": 11841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9069015870968691,
+      "learning_rate": 7.386628312294863e-08,
+      "loss": 0.3802,
+      "step": 11842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9050331418584707,
+      "learning_rate": 7.354725385339546e-08,
+      "loss": 0.4511,
+      "step": 11843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.083380607724216,
+      "learning_rate": 7.322891247821151e-08,
+      "loss": 0.5082,
+      "step": 11844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8287870334091962,
+      "learning_rate": 7.291125901946027e-08,
+      "loss": 0.4287,
+      "step": 11845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9519750070137744,
+      "learning_rate": 7.259429349915303e-08,
+      "loss": 0.5054,
+      "step": 11846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9186236888951284,
+      "learning_rate": 7.227801593925555e-08,
+      "loss": 0.4396,
+      "step": 11847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0156722316213873,
+      "learning_rate": 7.196242636168582e-08,
+      "loss": 0.5054,
+      "step": 11848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9044042392774271,
+      "learning_rate": 7.164752478831305e-08,
+      "loss": 0.442,
+      "step": 11849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.071712246415441,
+      "learning_rate": 7.133331124096087e-08,
+      "loss": 0.5517,
+      "step": 11850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9086881899095651,
+      "learning_rate": 7.101978574140411e-08,
+      "loss": 0.5043,
+      "step": 11851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9903068859691339,
+      "learning_rate": 7.070694831136871e-08,
+      "loss": 0.5096,
+      "step": 11852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7962250008333884,
+      "learning_rate": 7.039479897253509e-08,
+      "loss": 0.4403,
+      "step": 11853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9390602910887271,
+      "learning_rate": 7.008333774653376e-08,
+      "loss": 0.4721,
+      "step": 11854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9426156451013604,
+      "learning_rate": 6.977256465494853e-08,
+      "loss": 0.4614,
+      "step": 11855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9799328098160812,
+      "learning_rate": 6.946247971931774e-08,
+      "loss": 0.474,
+      "step": 11856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9250217103653412,
+      "learning_rate": 6.915308296112755e-08,
+      "loss": 0.4106,
+      "step": 11857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9025833591773156,
+      "learning_rate": 6.88443744018208e-08,
+      "loss": 0.4388,
+      "step": 11858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9340544233362317,
+      "learning_rate": 6.853635406279035e-08,
+      "loss": 0.4985,
+      "step": 11859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8905377200903977,
+      "learning_rate": 6.822902196538028e-08,
+      "loss": 0.4445,
+      "step": 11860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9160671303363982,
+      "learning_rate": 6.792237813089131e-08,
+      "loss": 0.4647,
+      "step": 11861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0165528326770312,
+      "learning_rate": 6.761642258056977e-08,
+      "loss": 0.4915,
+      "step": 11862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9790731286567422,
+      "learning_rate": 6.731115533562094e-08,
+      "loss": 0.4872,
+      "step": 11863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.940150772056775,
+      "learning_rate": 6.7006576417199e-08,
+      "loss": 0.5599,
+      "step": 11864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8891355445636352,
+      "learning_rate": 6.670268584641148e-08,
+      "loss": 0.5077,
+      "step": 11865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0427807053065226,
+      "learning_rate": 6.639948364431492e-08,
+      "loss": 0.4946,
+      "step": 11866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9213346316427382,
+      "learning_rate": 6.60969698319247e-08,
+      "loss": 0.4955,
+      "step": 11867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9091893947814887,
+      "learning_rate": 6.579514443020296e-08,
+      "loss": 0.4964,
+      "step": 11868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0501205485179521,
+      "learning_rate": 6.549400746006629e-08,
+      "loss": 0.4752,
+      "step": 11869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.883953055819093,
+      "learning_rate": 6.519355894238245e-08,
+      "loss": 0.4733,
+      "step": 11870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9817174202768156,
+      "learning_rate": 6.489379889797254e-08,
+      "loss": 0.4572,
+      "step": 11871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9823410939883479,
+      "learning_rate": 6.459472734760997e-08,
+      "loss": 0.521,
+      "step": 11872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0098939609336977,
+      "learning_rate": 6.429634431202036e-08,
+      "loss": 0.5193,
+      "step": 11873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0662717994985875,
+      "learning_rate": 6.399864981188164e-08,
+      "loss": 0.5038,
+      "step": 11874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9839336837767094,
+      "learning_rate": 6.370164386782285e-08,
+      "loss": 0.4969,
+      "step": 11875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9035783762848933,
+      "learning_rate": 6.340532650042641e-08,
+      "loss": 0.4771,
+      "step": 11876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.969875008374207,
+      "learning_rate": 6.310969773022701e-08,
+      "loss": 0.5079,
+      "step": 11877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9436460184683367,
+      "learning_rate": 6.281475757771161e-08,
+      "loss": 0.4733,
+      "step": 11878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.972534101604993,
+      "learning_rate": 6.252050606332049e-08,
+      "loss": 0.501,
+      "step": 11879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9578199633119227,
+      "learning_rate": 6.222694320744182e-08,
+      "loss": 0.4724,
+      "step": 11880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8301656923384986,
+      "learning_rate": 6.193406903042265e-08,
+      "loss": 0.475,
+      "step": 11881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9039558845681417,
+      "learning_rate": 6.164188355255673e-08,
+      "loss": 0.4935,
+      "step": 11882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9081915049960225,
+      "learning_rate": 6.135038679409344e-08,
+      "loss": 0.5057,
+      "step": 11883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0382548299120893,
+      "learning_rate": 6.105957877523216e-08,
+      "loss": 0.4951,
+      "step": 11884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.905766580597293,
+      "learning_rate": 6.076945951612678e-08,
+      "loss": 0.4418,
+      "step": 11885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8931595039530028,
+      "learning_rate": 6.048002903688121e-08,
+      "loss": 0.4915,
+      "step": 11886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0160363862096509,
+      "learning_rate": 6.019128735755386e-08,
+      "loss": 0.5485,
+      "step": 11887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8877215619065384,
+      "learning_rate": 5.990323449815316e-08,
+      "loss": 0.4455,
+      "step": 11888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9249415508767169,
+      "learning_rate": 5.961587047864204e-08,
+      "loss": 0.4407,
+      "step": 11889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8467704348053258,
+      "learning_rate": 5.932919531893344e-08,
+      "loss": 0.4518,
+      "step": 11890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.928828790733546,
+      "learning_rate": 5.9043209038894825e-08,
+      "loss": 0.4528,
+      "step": 11891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9258296897330218,
+      "learning_rate": 5.8757911658343657e-08,
+      "loss": 0.4908,
+      "step": 11892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.993635545931022,
+      "learning_rate": 5.847330319705191e-08,
+      "loss": 0.5191,
+      "step": 11893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0503873247954665,
+      "learning_rate": 5.818938367474159e-08,
+      "loss": 0.4728,
+      "step": 11894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.909336100675651,
+      "learning_rate": 5.790615311108805e-08,
+      "loss": 0.4634,
+      "step": 11895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0197521172435708,
+      "learning_rate": 5.7623611525721155e-08,
+      "loss": 0.5245,
+      "step": 11896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8289684625646265,
+      "learning_rate": 5.7341758938217474e-08,
+      "loss": 0.404,
+      "step": 11897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0007797587248377,
+      "learning_rate": 5.706059536811137e-08,
+      "loss": 0.4432,
+      "step": 11898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8384808151151713,
+      "learning_rate": 5.6780120834887264e-08,
+      "loss": 0.3979,
+      "step": 11899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8634048374650863,
+      "learning_rate": 5.650033535798072e-08,
+      "loss": 0.4548,
+      "step": 11900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9543880848177454,
+      "learning_rate": 5.6221238956780664e-08,
+      "loss": 0.4442,
+      "step": 11901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9048198696439131,
+      "learning_rate": 5.5942831650628303e-08,
+      "loss": 0.4611,
+      "step": 11902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9538814286699375,
+      "learning_rate": 5.566511345881931e-08,
+      "loss": 0.5113,
+      "step": 11903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9043391175027943,
+      "learning_rate": 5.5388084400594954e-08,
+      "loss": 0.4949,
+      "step": 11904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9293652419355056,
+      "learning_rate": 5.511174449515655e-08,
+      "loss": 0.46,
+      "step": 11905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0324711413947731,
+      "learning_rate": 5.483609376165322e-08,
+      "loss": 0.5127,
+      "step": 11906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8941503163390319,
+      "learning_rate": 5.456113221918746e-08,
+      "loss": 0.4751,
+      "step": 11907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.919163316555529,
+      "learning_rate": 5.428685988681292e-08,
+      "loss": 0.4502,
+      "step": 11908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9342751986032585,
+      "learning_rate": 5.401327678353774e-08,
+      "loss": 0.4755,
+      "step": 11909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9684954869778296,
+      "learning_rate": 5.3740382928320065e-08,
+      "loss": 0.4781,
+      "step": 11910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0860550064505223,
+      "learning_rate": 5.346817834007145e-08,
+      "loss": 0.4832,
+      "step": 11911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.915553037036275,
+      "learning_rate": 5.3196663037655695e-08,
+      "loss": 0.5023,
+      "step": 11912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9504761418845167,
+      "learning_rate": 5.292583703988885e-08,
+      "loss": 0.4172,
+      "step": 11913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9750402476226417,
+      "learning_rate": 5.265570036553813e-08,
+      "loss": 0.5118,
+      "step": 11914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.950881631306124,
+      "learning_rate": 5.238625303332412e-08,
+      "loss": 0.4694,
+      "step": 11915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9596144934035122,
+      "learning_rate": 5.2117495061918544e-08,
+      "loss": 0.4668,
+      "step": 11916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9621390065847045,
+      "learning_rate": 5.184942646994762e-08,
+      "loss": 0.5706,
+      "step": 11917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9387416367581379,
+      "learning_rate": 5.158204727598759e-08,
+      "loss": 0.4659,
+      "step": 11918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0693341512877415,
+      "learning_rate": 5.131535749856698e-08,
+      "loss": 0.4836,
+      "step": 11919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0015756147477828,
+      "learning_rate": 5.104935715616766e-08,
+      "loss": 0.5097,
+      "step": 11920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.272574570392914,
+      "learning_rate": 5.0784046267223775e-08,
+      "loss": 0.4799,
+      "step": 11921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8719092888254251,
+      "learning_rate": 5.0519424850119516e-08,
+      "loss": 0.4908,
+      "step": 11922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9723855937950857,
+      "learning_rate": 5.025549292319465e-08,
+      "loss": 0.4476,
+      "step": 11923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8891720838974949,
+      "learning_rate": 4.999225050473788e-08,
+      "loss": 0.4679,
+      "step": 11924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.00799732731766,
+      "learning_rate": 4.972969761299351e-08,
+      "loss": 0.4652,
+      "step": 11925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9705146089469759,
+      "learning_rate": 4.9467834266154756e-08,
+      "loss": 0.4753,
+      "step": 11926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8757512034841,
+      "learning_rate": 4.920666048236933e-08,
+      "loss": 0.438,
+      "step": 11927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9998789701294412,
+      "learning_rate": 4.894617627973497e-08,
+      "loss": 0.4694,
+      "step": 11928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9252960647533439,
+      "learning_rate": 4.8686381676305015e-08,
+      "loss": 0.4889,
+      "step": 11929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9785224715228302,
+      "learning_rate": 4.8427276690081735e-08,
+      "loss": 0.4981,
+      "step": 11930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9302323682272238,
+      "learning_rate": 4.8168861339020766e-08,
+      "loss": 0.4768,
+      "step": 11931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9128889123665305,
+      "learning_rate": 4.791113564103111e-08,
+      "loss": 0.4583,
+      "step": 11932
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9159577759229364,
+      "learning_rate": 4.7654099613971825e-08,
+      "loss": 0.4738,
+      "step": 11933
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9667416898072759,
+      "learning_rate": 4.739775327565532e-08,
+      "loss": 0.5195,
+      "step": 11934
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9583053889816117,
+      "learning_rate": 4.714209664384739e-08,
+      "loss": 0.4987,
+      "step": 11935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0593700920519042,
+      "learning_rate": 4.688712973626386e-08,
+      "loss": 0.4761,
+      "step": 11936
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9518152352507855,
+      "learning_rate": 4.663285257057393e-08,
+      "loss": 0.4734,
+      "step": 11937
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9369693183511989,
+      "learning_rate": 4.637926516439795e-08,
+      "loss": 0.4472,
+      "step": 11938
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9575936036237775,
+      "learning_rate": 4.612636753531075e-08,
+      "loss": 0.4948,
+      "step": 11939
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0008413774296308,
+      "learning_rate": 4.58741597008372e-08,
+      "loss": 0.4879,
+      "step": 11940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9777885044908675,
+      "learning_rate": 4.5622641678454424e-08,
+      "loss": 0.4558,
+      "step": 11941
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9945116465130329,
+      "learning_rate": 4.537181348559405e-08,
+      "loss": 0.4254,
+      "step": 11942
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9461087646435123,
+      "learning_rate": 4.512167513963661e-08,
+      "loss": 0.5021,
+      "step": 11943
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9954162897161014,
+      "learning_rate": 4.487222665791713e-08,
+      "loss": 0.5154,
+      "step": 11944
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9237004610309977,
+      "learning_rate": 4.4623468057722886e-08,
+      "loss": 0.466,
+      "step": 11945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9637871569395646,
+      "learning_rate": 4.437539935629009e-08,
+      "loss": 0.4733,
+      "step": 11946
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8388679921392534,
+      "learning_rate": 4.412802057081278e-08,
+      "loss": 0.4363,
+      "step": 11947
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9179034699969338,
+      "learning_rate": 4.388133171843278e-08,
+      "loss": 0.5191,
+      "step": 11948
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8819156108312363,
+      "learning_rate": 4.3635332816245324e-08,
+      "loss": 0.4188,
+      "step": 11949
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9320418097418713,
+      "learning_rate": 4.339002388129787e-08,
+      "loss": 0.4578,
+      "step": 11950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.972402913307539,
+      "learning_rate": 4.3145404930591275e-08,
+      "loss": 0.4979,
+      "step": 11951
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9727729975849237,
+      "learning_rate": 4.2901475981074195e-08,
+      "loss": 0.5532,
+      "step": 11952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.035867758045349,
+      "learning_rate": 4.2658237049655325e-08,
+      "loss": 0.4867,
+      "step": 11953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0089929989437811,
+      "learning_rate": 4.241568815318675e-08,
+      "loss": 0.4247,
+      "step": 11954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0064757121513352,
+      "learning_rate": 4.2173829308479466e-08,
+      "loss": 0.4637,
+      "step": 11955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9584510660983068,
+      "learning_rate": 4.193266053229339e-08,
+      "loss": 0.486,
+      "step": 11956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0334196207407,
+      "learning_rate": 4.1692181841340716e-08,
+      "loss": 0.4957,
+      "step": 11957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.971222793955482,
+      "learning_rate": 4.1452393252285894e-08,
+      "loss": 0.4645,
+      "step": 11958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0089128046903748,
+      "learning_rate": 4.1213294781748956e-08,
+      "loss": 0.4739,
+      "step": 11959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.916419587613008,
+      "learning_rate": 4.097488644629555e-08,
+      "loss": 0.4691,
+      "step": 11960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9609627617532643,
+      "learning_rate": 4.0737168262450224e-08,
+      "loss": 0.4733,
+      "step": 11961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9136837067317344,
+      "learning_rate": 4.050014024668425e-08,
+      "loss": 0.4675,
+      "step": 11962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8788214931690304,
+      "learning_rate": 4.02638024154256e-08,
+      "loss": 0.4911,
+      "step": 11963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9283939728597378,
+      "learning_rate": 4.002815478505007e-08,
+      "loss": 0.4347,
+      "step": 11964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9048435075657899,
+      "learning_rate": 3.9793197371889026e-08,
+      "loss": 0.4434,
+      "step": 11965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0281323566018774,
+      "learning_rate": 3.955893019222501e-08,
+      "loss": 0.4806,
+      "step": 11966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9526276177157981,
+      "learning_rate": 3.93253532622917e-08,
+      "loss": 0.4953,
+      "step": 11967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9364679716036685,
+      "learning_rate": 3.909246659827726e-08,
+      "loss": 0.4233,
+      "step": 11968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9841615329388482,
+      "learning_rate": 3.8860270216319885e-08,
+      "loss": 0.5168,
+      "step": 11969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.047709700000648,
+      "learning_rate": 3.862876413250893e-08,
+      "loss": 0.5589,
+      "step": 11970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.863210692248217,
+      "learning_rate": 3.839794836288935e-08,
+      "loss": 0.4264,
+      "step": 11971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8951007919327821,
+      "learning_rate": 3.816782292345611e-08,
+      "loss": 0.442,
+      "step": 11972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0667982781868577,
+      "learning_rate": 3.7938387830156464e-08,
+      "loss": 0.4819,
+      "step": 11973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8919483517785504,
+      "learning_rate": 3.7709643098891024e-08,
+      "loss": 0.4757,
+      "step": 11974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0189103671289825,
+      "learning_rate": 3.748158874550934e-08,
+      "loss": 0.5292,
+      "step": 11975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8422196036000987,
+      "learning_rate": 3.725422478581764e-08,
+      "loss": 0.468,
+      "step": 11976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9799552575419327,
+      "learning_rate": 3.702755123557111e-08,
+      "loss": 0.4893,
+      "step": 11977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0308217161205298,
+      "learning_rate": 3.6801568110478304e-08,
+      "loss": 0.5068,
+      "step": 11978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8779731833204971,
+      "learning_rate": 3.6576275426200014e-08,
+      "loss": 0.4277,
+      "step": 11979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9053805546833162,
+      "learning_rate": 3.635167319834709e-08,
+      "loss": 0.465,
+      "step": 11980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8666755276899345,
+      "learning_rate": 3.612776144248597e-08,
+      "loss": 0.4476,
+      "step": 11981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9849607754708499,
+      "learning_rate": 3.590454017413314e-08,
+      "loss": 0.5167,
+      "step": 11982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.857166337628939,
+      "learning_rate": 3.568200940875732e-08,
+      "loss": 0.442,
+      "step": 11983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9616493391315466,
+      "learning_rate": 3.546016916178063e-08,
+      "loss": 0.5472,
+      "step": 11984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9561335160933164,
+      "learning_rate": 3.523901944857522e-08,
+      "loss": 0.4597,
+      "step": 11985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0152458863216052,
+      "learning_rate": 3.5018560284466595e-08,
+      "loss": 0.5377,
+      "step": 11986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9116349271646171,
+      "learning_rate": 3.4798791684733655e-08,
+      "loss": 0.4812,
+      "step": 11987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9578005502526648,
+      "learning_rate": 3.457971366460422e-08,
+      "loss": 0.4809,
+      "step": 11988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9950944460120628,
+      "learning_rate": 3.436132623926169e-08,
+      "loss": 0.5159,
+      "step": 11989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0323696694058275,
+      "learning_rate": 3.414362942384064e-08,
+      "loss": 0.473,
+      "step": 11990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8910655305456949,
+      "learning_rate": 3.392662323342566e-08,
+      "loss": 0.5181,
+      "step": 11991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0142482538193889,
+      "learning_rate": 3.371030768305583e-08,
+      "loss": 0.4883,
+      "step": 11992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8941388462250626,
+      "learning_rate": 3.349468278772139e-08,
+      "loss": 0.4608,
+      "step": 11993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9806960085628176,
+      "learning_rate": 3.3279748562364824e-08,
+      "loss": 0.4906,
+      "step": 11994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9284249097360684,
+      "learning_rate": 3.3065505021881995e-08,
+      "loss": 0.5173,
+      "step": 11995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.962827598357366,
+      "learning_rate": 3.2851952181118805e-08,
+      "loss": 0.4712,
+      "step": 11996
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9785884863012381,
+      "learning_rate": 3.2639090054874534e-08,
+      "loss": 0.4511,
+      "step": 11997
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9825408225097995,
+      "learning_rate": 3.242691865790071e-08,
+      "loss": 0.5059,
+      "step": 11998
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0175010878886457,
+      "learning_rate": 3.22154380048989e-08,
+      "loss": 0.551,
+      "step": 11999
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8098232963770842,
+      "learning_rate": 3.200464811052628e-08,
+      "loss": 0.4347,
+      "step": 12000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.943884107957791,
+      "learning_rate": 3.1794548989391163e-08,
+      "loss": 0.4533,
+      "step": 12001
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9370564155146761,
+      "learning_rate": 3.158514065605078e-08,
+      "loss": 0.4661,
+      "step": 12002
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8380704807977243,
+      "learning_rate": 3.1376423125019093e-08,
+      "loss": 0.4335,
+      "step": 12003
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9139190205140328,
+      "learning_rate": 3.116839641075786e-08,
+      "loss": 0.4962,
+      "step": 12004
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.018881269398722,
+      "learning_rate": 3.0961060527685546e-08,
+      "loss": 0.5046,
+      "step": 12005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9367199554553238,
+      "learning_rate": 3.0754415490168446e-08,
+      "loss": 0.4828,
+      "step": 12006
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8418975025483183,
+      "learning_rate": 3.054846131252731e-08,
+      "loss": 0.4576,
+      "step": 12007
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9456790269529668,
+      "learning_rate": 3.034319800903629e-08,
+      "loss": 0.4752,
+      "step": 12008
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9483389987156057,
+      "learning_rate": 3.013862559391734e-08,
+      "loss": 0.4741,
+      "step": 12009
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9600136688183358,
+      "learning_rate": 2.993474408134911e-08,
+      "loss": 0.4849,
+      "step": 12010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9480643663449927,
+      "learning_rate": 2.9731553485459197e-08,
+      "loss": 0.4928,
+      "step": 12011
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8911982744195088,
+      "learning_rate": 2.9529053820329667e-08,
+      "loss": 0.4711,
+      "step": 12012
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.97406248259707,
+      "learning_rate": 2.932724509999263e-08,
+      "loss": 0.4866,
+      "step": 12013
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9693131231472446,
+      "learning_rate": 2.9126127338432454e-08,
+      "loss": 0.5288,
+      "step": 12014
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8760148975933071,
+      "learning_rate": 2.8925700549589096e-08,
+      "loss": 0.4298,
+      "step": 12015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9383849892653244,
+      "learning_rate": 2.8725964747350342e-08,
+      "loss": 0.4584,
+      "step": 12016
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9725559701311824,
+      "learning_rate": 2.852691994555623e-08,
+      "loss": 0.4955,
+      "step": 12017
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0068941623928347,
+      "learning_rate": 2.8328566158002392e-08,
+      "loss": 0.4362,
+      "step": 12018
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9606479146924113,
+      "learning_rate": 2.8130903398434496e-08,
+      "loss": 0.4623,
+      "step": 12019
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9535023751808042,
+      "learning_rate": 2.7933931680550476e-08,
+      "loss": 0.5133,
+      "step": 12020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8856677096306654,
+      "learning_rate": 2.7737651017998303e-08,
+      "loss": 0.4567,
+      "step": 12021
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8725884384204944,
+      "learning_rate": 2.754206142438265e-08,
+      "loss": 0.4795,
+      "step": 12022
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0567535458362625,
+      "learning_rate": 2.734716291325712e-08,
+      "loss": 0.5445,
+      "step": 12023
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9246990984926732,
+      "learning_rate": 2.7152955498126465e-08,
+      "loss": 0.4927,
+      "step": 12024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9147568261561264,
+      "learning_rate": 2.695943919244992e-08,
+      "loss": 0.4427,
+      "step": 12025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8809789552613674,
+      "learning_rate": 2.676661400963898e-08,
+      "loss": 0.496,
+      "step": 12026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9264478746652443,
+      "learning_rate": 2.6574479963054068e-08,
+      "loss": 0.4888,
+      "step": 12027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0732755625344113,
+      "learning_rate": 2.6383037066013417e-08,
+      "loss": 0.4947,
+      "step": 12028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9045542732560331,
+      "learning_rate": 2.6192285331779754e-08,
+      "loss": 0.4218,
+      "step": 12029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.926681609482207,
+      "learning_rate": 2.6002224773574725e-08,
+      "loss": 0.4675,
+      "step": 12030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9571897718833516,
+      "learning_rate": 2.5812855404568903e-08,
+      "loss": 0.5279,
+      "step": 12031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8379559114592033,
+      "learning_rate": 2.5624177237884017e-08,
+      "loss": 0.4349,
+      "step": 12032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9385323324140011,
+      "learning_rate": 2.5436190286597384e-08,
+      "loss": 0.5041,
+      "step": 12033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9581379073190894,
+      "learning_rate": 2.524889456373525e-08,
+      "loss": 0.5283,
+      "step": 12034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9226299394723871,
+      "learning_rate": 2.506229008227723e-08,
+      "loss": 0.4763,
+      "step": 12035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0014369490594117,
+      "learning_rate": 2.4876376855154095e-08,
+      "loss": 0.5233,
+      "step": 12036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9786997303191647,
+      "learning_rate": 2.469115489525109e-08,
+      "loss": 0.4833,
+      "step": 12037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0225673016820687,
+      "learning_rate": 2.4506624215402396e-08,
+      "loss": 0.4294,
+      "step": 12038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0178203672748507,
+      "learning_rate": 2.4322784828395562e-08,
+      "loss": 0.5122,
+      "step": 12039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.960735625727089,
+      "learning_rate": 2.4139636746972617e-08,
+      "loss": 0.4957,
+      "step": 12040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9003834052306827,
+      "learning_rate": 2.395717998382341e-08,
+      "loss": 0.4717,
+      "step": 12041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0236375174607777,
+      "learning_rate": 2.377541455159338e-08,
+      "loss": 0.4855,
+      "step": 12042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9208499654286599,
+      "learning_rate": 2.3594340462878007e-08,
+      "loss": 0.4828,
+      "step": 12043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9747632253006461,
+      "learning_rate": 2.3413957730226144e-08,
+      "loss": 0.5228,
+      "step": 12044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8512407956116504,
+      "learning_rate": 2.3234266366137794e-08,
+      "loss": 0.486,
+      "step": 12045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9528617655477253,
+      "learning_rate": 2.305526638306521e-08,
+      "loss": 0.4579,
+      "step": 12046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9290253967263423,
+      "learning_rate": 2.2876957793412923e-08,
+      "loss": 0.4823,
+      "step": 12047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7917551407534277,
+      "learning_rate": 2.2699340609537713e-08,
+      "loss": 0.3943,
+      "step": 12048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.969076220647739,
+      "learning_rate": 2.2522414843748618e-08,
+      "loss": 0.4756,
+      "step": 12049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9774628839866263,
+      "learning_rate": 2.2346180508305836e-08,
+      "loss": 0.4676,
+      "step": 12050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9611488238609713,
+      "learning_rate": 2.217063761542293e-08,
+      "loss": 0.4263,
+      "step": 12051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0969373743056006,
+      "learning_rate": 2.1995786177264612e-08,
+      "loss": 0.4796,
+      "step": 12052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9629598594205779,
+      "learning_rate": 2.1821626205947854e-08,
+      "loss": 0.5476,
+      "step": 12053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9234136061403723,
+      "learning_rate": 2.1648157713540786e-08,
+      "loss": 0.4845,
+      "step": 12054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0456502401405645,
+      "learning_rate": 2.147538071206712e-08,
+      "loss": 0.4983,
+      "step": 12055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9620817812734163,
+      "learning_rate": 2.130329521349728e-08,
+      "loss": 0.5425,
+      "step": 12056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9936463735647684,
+      "learning_rate": 2.113190122975839e-08,
+      "loss": 0.4874,
+      "step": 12057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9191229093038062,
+      "learning_rate": 2.096119877272873e-08,
+      "loss": 0.4658,
+      "step": 12058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0217672518961693,
+      "learning_rate": 2.0791187854234396e-08,
+      "loss": 0.5153,
+      "step": 12059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0039471938425795,
+      "learning_rate": 2.0621868486060402e-08,
+      "loss": 0.5135,
+      "step": 12060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9101474014009246,
+      "learning_rate": 2.045324067993959e-08,
+      "loss": 0.4847,
+      "step": 12061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9451141667951678,
+      "learning_rate": 2.0285304447557052e-08,
+      "loss": 0.4758,
+      "step": 12062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1071361661435415,
+      "learning_rate": 2.011805980055015e-08,
+      "loss": 0.5314,
+      "step": 12063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9478821974145435,
+      "learning_rate": 1.9951506750510718e-08,
+      "loss": 0.5055,
+      "step": 12064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9473836420133486,
+      "learning_rate": 1.9785645308978417e-08,
+      "loss": 0.4738,
+      "step": 12065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.047067744675001,
+      "learning_rate": 1.962047548744961e-08,
+      "loss": 0.4963,
+      "step": 12066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9487194859396089,
+      "learning_rate": 1.9455997297368467e-08,
+      "loss": 0.4543,
+      "step": 12067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9020631622345036,
+      "learning_rate": 1.9292210750134766e-08,
+      "loss": 0.4766,
+      "step": 12068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9240780684461876,
+      "learning_rate": 1.9129115857097203e-08,
+      "loss": 0.4414,
+      "step": 12069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0183690019040301,
+      "learning_rate": 1.896671262955896e-08,
+      "loss": 0.4414,
+      "step": 12070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9719576260291167,
+      "learning_rate": 1.8805001078774364e-08,
+      "loss": 0.4942,
+      "step": 12071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8229267596156057,
+      "learning_rate": 1.8643981215951125e-08,
+      "loss": 0.4737,
+      "step": 12072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.926850956701825,
+      "learning_rate": 1.8483653052244754e-08,
+      "loss": 0.4606,
+      "step": 12073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9709028631688023,
+      "learning_rate": 1.832401659876859e-08,
+      "loss": 0.4355,
+      "step": 12074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0114948920634896,
+      "learning_rate": 1.8165071866583785e-08,
+      "loss": 0.4659,
+      "step": 12075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.996860587805961,
+      "learning_rate": 1.8006818866705968e-08,
+      "loss": 0.5256,
+      "step": 12076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8907455895141424,
+      "learning_rate": 1.7849257610101923e-08,
+      "loss": 0.431,
+      "step": 12077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9866829023109346,
+      "learning_rate": 1.7692388107689584e-08,
+      "loss": 0.4946,
+      "step": 12078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9840949466758064,
+      "learning_rate": 1.7536210370341366e-08,
+      "loss": 0.4973,
+      "step": 12079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.934940510323555,
+      "learning_rate": 1.7380724408878613e-08,
+      "loss": 0.4776,
+      "step": 12080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9316063795807219,
+      "learning_rate": 1.7225930234077147e-08,
+      "loss": 0.448,
+      "step": 12081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9043191059114086,
+      "learning_rate": 1.7071827856663947e-08,
+      "loss": 0.4002,
+      "step": 12082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0040959642431642,
+      "learning_rate": 1.6918417287318245e-08,
+      "loss": 0.4702,
+      "step": 12083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0118369320866807,
+      "learning_rate": 1.6765698536671538e-08,
+      "loss": 0.5233,
+      "step": 12084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9343418206195916,
+      "learning_rate": 1.661367161530647e-08,
+      "loss": 0.4925,
+      "step": 12085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9724808816874202,
+      "learning_rate": 1.646233653375795e-08,
+      "loss": 0.5342,
+      "step": 12086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9234714441172094,
+      "learning_rate": 1.6311693302515364e-08,
+      "loss": 0.4269,
+      "step": 12087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9908340940280683,
+      "learning_rate": 1.6161741932017026e-08,
+      "loss": 0.5122,
+      "step": 12088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0136407090289303,
+      "learning_rate": 1.601248243265352e-08,
+      "loss": 0.555,
+      "step": 12089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8792657315259043,
+      "learning_rate": 1.586391481476879e-08,
+      "loss": 0.497,
+      "step": 12090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9361415339097172,
+      "learning_rate": 1.5716039088660152e-08,
+      "loss": 0.4718,
+      "step": 12091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.051550955427514,
+      "learning_rate": 1.5568855264572745e-08,
+      "loss": 0.4684,
+      "step": 12092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0455735739038166,
+      "learning_rate": 1.5422363352708414e-08,
+      "loss": 0.4771,
+      "step": 12093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9894632078153739,
+      "learning_rate": 1.5276563363217923e-08,
+      "loss": 0.5236,
+      "step": 12094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9306248195163555,
+      "learning_rate": 1.51314553062043e-08,
+      "loss": 0.5081,
+      "step": 12095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.030265257408779,
+      "learning_rate": 1.498703919172506e-08,
+      "loss": 0.5067,
+      "step": 12096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9951251961523201,
+      "learning_rate": 1.4843315029786642e-08,
+      "loss": 0.4647,
+      "step": 12097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9202627425313655,
+      "learning_rate": 1.4700282830351077e-08,
+      "loss": 0.4568,
+      "step": 12098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0646728167615809,
+      "learning_rate": 1.4557942603327103e-08,
+      "loss": 0.5333,
+      "step": 12099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9165375441309248,
+      "learning_rate": 1.4416294358582383e-08,
+      "loss": 0.474,
+      "step": 12100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9989449749096093,
+      "learning_rate": 1.4275338105930181e-08,
+      "loss": 0.5541,
+      "step": 12101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8842241995287015,
+      "learning_rate": 1.4135073855139348e-08,
+      "loss": 0.4333,
+      "step": 12102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0020455888538515,
+      "learning_rate": 1.3995501615930996e-08,
+      "loss": 0.5113,
+      "step": 12103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9392848855909458,
+      "learning_rate": 1.3856621397977388e-08,
+      "loss": 0.4491,
+      "step": 12104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8645236971420108,
+      "learning_rate": 1.3718433210901938e-08,
+      "loss": 0.4684,
+      "step": 12105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9476125835258605,
+      "learning_rate": 1.358093706428032e-08,
+      "loss": 0.4455,
+      "step": 12106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9120247324159362,
+      "learning_rate": 1.3444132967642687e-08,
+      "loss": 0.4567,
+      "step": 12107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9439904162223841,
+      "learning_rate": 1.3308020930468123e-08,
+      "loss": 0.4762,
+      "step": 12108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8476734435812402,
+      "learning_rate": 1.3172600962190196e-08,
+      "loss": 0.3775,
+      "step": 12109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0085984962827712,
+      "learning_rate": 1.3037873072192509e-08,
+      "loss": 0.5417,
+      "step": 12110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9458556391105006,
+      "learning_rate": 1.2903837269810926e-08,
+      "loss": 0.4422,
+      "step": 12111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.927148332515927,
+      "learning_rate": 1.2770493564335795e-08,
+      "loss": 0.4492,
+      "step": 12112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8648138309308088,
+      "learning_rate": 1.2637841965006392e-08,
+      "loss": 0.4853,
+      "step": 12113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9085857938849952,
+      "learning_rate": 1.2505882481016473e-08,
+      "loss": 0.4589,
+      "step": 12114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9850194006674354,
+      "learning_rate": 1.2374615121508726e-08,
+      "loss": 0.5077,
+      "step": 12115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8683619860038134,
+      "learning_rate": 1.2244039895582538e-08,
+      "loss": 0.4737,
+      "step": 12116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.986180421412581,
+      "learning_rate": 1.2114156812284006e-08,
+      "loss": 0.4238,
+      "step": 12117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8630904059814098,
+      "learning_rate": 1.1984965880615929e-08,
+      "loss": 0.4534,
+      "step": 12118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9567877247689517,
+      "learning_rate": 1.1856467109530034e-08,
+      "loss": 0.4496,
+      "step": 12119
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9244238624328863,
+      "learning_rate": 1.1728660507931423e-08,
+      "loss": 0.4996,
+      "step": 12120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9355603986596689,
+      "learning_rate": 1.1601546084677451e-08,
+      "loss": 0.4538,
+      "step": 12121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0998083959992173,
+      "learning_rate": 1.147512384857663e-08,
+      "loss": 0.5296,
+      "step": 12122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0029466284947572,
+      "learning_rate": 1.134939380838973e-08,
+      "loss": 0.5109,
+      "step": 12123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.018903049451067,
+      "learning_rate": 1.1224355972829782e-08,
+      "loss": 0.53,
+      "step": 12124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9814802302583996,
+      "learning_rate": 1.1100010350562073e-08,
+      "loss": 0.4953,
+      "step": 12125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9378454731447133,
+      "learning_rate": 1.0976356950203049e-08,
+      "loss": 0.4791,
+      "step": 12126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8459287341705666,
+      "learning_rate": 1.0853395780322518e-08,
+      "loss": 0.4792,
+      "step": 12127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9874219066355951,
+      "learning_rate": 1.0731126849441442e-08,
+      "loss": 0.5082,
+      "step": 12128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9790531177343795,
+      "learning_rate": 1.0609550166033045e-08,
+      "loss": 0.4642,
+      "step": 12129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8633723815748132,
+      "learning_rate": 1.0488665738521697e-08,
+      "loss": 0.4535,
+      "step": 12130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9307003256715729,
+      "learning_rate": 1.0368473575285143e-08,
+      "loss": 0.491,
+      "step": 12131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9499840617739655,
+      "learning_rate": 1.0248973684653385e-08,
+      "loss": 0.4717,
+      "step": 12132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9656425326858633,
+      "learning_rate": 1.0130166074906467e-08,
+      "loss": 0.4833,
+      "step": 12133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.969248297391733,
+      "learning_rate": 1.0012050754277802e-08,
+      "loss": 0.4377,
+      "step": 12134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0964434122566502,
+      "learning_rate": 9.894627730953066e-09,
+      "loss": 0.4803,
+      "step": 12135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9802224309989128,
+      "learning_rate": 9.777897013069082e-09,
+      "loss": 0.5238,
+      "step": 12136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8730273662548644,
+      "learning_rate": 9.661858608716045e-09,
+      "loss": 0.4808,
+      "step": 12137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9342879322938532,
+      "learning_rate": 9.546512525934193e-09,
+      "loss": 0.4506,
+      "step": 12138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8701452487731054,
+      "learning_rate": 9.43185877271824e-09,
+      "loss": 0.4778,
+      "step": 12139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0066801314510307,
+      "learning_rate": 9.317897357011829e-09,
+      "loss": 0.4871,
+      "step": 12140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9434565414798493,
+      "learning_rate": 9.204628286714202e-09,
+      "loss": 0.5085,
+      "step": 12141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.960720693115298,
+      "learning_rate": 9.092051569674632e-09,
+      "loss": 0.5068,
+      "step": 12142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0266753458833535,
+      "learning_rate": 8.980167213692437e-09,
+      "loss": 0.4912,
+      "step": 12143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8768551645097515,
+      "learning_rate": 8.868975226523634e-09,
+      "loss": 0.4614,
+      "step": 12144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9258059202902077,
+      "learning_rate": 8.758475615872065e-09,
+      "loss": 0.4769,
+      "step": 12145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8688950692394761,
+      "learning_rate": 8.648668389397153e-09,
+      "loss": 0.4356,
+      "step": 12146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9116665627644456,
+      "learning_rate": 8.539553554706148e-09,
+      "loss": 0.479,
+      "step": 12147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9170045332553114,
+      "learning_rate": 8.431131119361891e-09,
+      "loss": 0.4745,
+      "step": 12148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9571233551136623,
+      "learning_rate": 8.323401090877258e-09,
+      "loss": 0.4227,
+      "step": 12149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9635271757833536,
+      "learning_rate": 8.216363476718503e-09,
+      "loss": 0.4495,
+      "step": 12150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.972196287676129,
+      "learning_rate": 8.110018284304132e-09,
+      "loss": 0.5253,
+      "step": 12151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9668650890442178,
+      "learning_rate": 8.004365521001589e-09,
+      "loss": 0.468,
+      "step": 12152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9220578685108054,
+      "learning_rate": 7.899405194133902e-09,
+      "loss": 0.4399,
+      "step": 12153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.109967136629721,
+      "learning_rate": 7.795137310974143e-09,
+      "loss": 0.5623,
+      "step": 12154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9759862425878394,
+      "learning_rate": 7.691561878748755e-09,
+      "loss": 0.4836,
+      "step": 12155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9096238591337812,
+      "learning_rate": 7.588678904635328e-09,
+      "loss": 0.4849,
+      "step": 12156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0074088319601273,
+      "learning_rate": 7.486488395762604e-09,
+      "loss": 0.556,
+      "step": 12157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9484713726196886,
+      "learning_rate": 7.384990359212696e-09,
+      "loss": 0.4814,
+      "step": 12158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9423528640935501,
+      "learning_rate": 7.284184802019978e-09,
+      "loss": 0.4615,
+      "step": 12159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0101292494710938,
+      "learning_rate": 7.1840717311688625e-09,
+      "loss": 0.5216,
+      "step": 12160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9788949263707301,
+      "learning_rate": 7.084651153599353e-09,
+      "loss": 0.4846,
+      "step": 12161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8823697123892938,
+      "learning_rate": 6.985923076199275e-09,
+      "loss": 0.4516,
+      "step": 12162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9332411928489063,
+      "learning_rate": 6.88788750580982e-09,
+      "loss": 0.4917,
+      "step": 12163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8939196205465503,
+      "learning_rate": 6.790544449227776e-09,
+      "loss": 0.5001,
+      "step": 12164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.915649796606784,
+      "learning_rate": 6.693893913195526e-09,
+      "loss": 0.4448,
+      "step": 12165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8985794749672081,
+      "learning_rate": 6.597935904413267e-09,
+      "loss": 0.4557,
+      "step": 12166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8898060428434278,
+      "learning_rate": 6.502670429529012e-09,
+      "loss": 0.4438,
+      "step": 12167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.911361729983376,
+      "learning_rate": 6.40809749514637e-09,
+      "loss": 0.4731,
+      "step": 12168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9853357071007716,
+      "learning_rate": 6.314217107817877e-09,
+      "loss": 0.5042,
+      "step": 12169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9595718090401706,
+      "learning_rate": 6.221029274049439e-09,
+      "loss": 0.4843,
+      "step": 12170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9264152524313263,
+      "learning_rate": 6.1285340003003346e-09,
+      "loss": 0.4648,
+      "step": 12171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0823810828783662,
+      "learning_rate": 6.03673129297877e-09,
+      "loss": 0.5243,
+      "step": 12172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0166043314385933,
+      "learning_rate": 5.945621158446324e-09,
+      "loss": 0.4803,
+      "step": 12173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0495585663202873,
+      "learning_rate": 5.855203603017945e-09,
+      "loss": 0.4966,
+      "step": 12174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9321136127621876,
+      "learning_rate": 5.765478632959731e-09,
+      "loss": 0.4823,
+      "step": 12175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.89851925044116,
+      "learning_rate": 5.676446254488932e-09,
+      "loss": 0.4719,
+      "step": 12176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0085946206564493,
+      "learning_rate": 5.588106473775057e-09,
+      "loss": 0.5553,
+      "step": 12177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8649896612244649,
+      "learning_rate": 5.500459296939875e-09,
+      "loss": 0.4842,
+      "step": 12178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9366458571588505,
+      "learning_rate": 5.413504730058527e-09,
+      "loss": 0.4386,
+      "step": 12179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8616063953962477,
+      "learning_rate": 5.327242779156194e-09,
+      "loss": 0.485,
+      "step": 12180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.947643360116589,
+      "learning_rate": 5.2416734502103165e-09,
+      "loss": 0.4673,
+      "step": 12181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5494313108108435,
+      "learning_rate": 5.156796749150595e-09,
+      "loss": 0.458,
+      "step": 12182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9505065847530575,
+      "learning_rate": 5.0726126818601e-09,
+      "loss": 0.456,
+      "step": 12183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9026091368438374,
+      "learning_rate": 4.989121254171947e-09,
+      "loss": 0.4619,
+      "step": 12184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9139503054391478,
+      "learning_rate": 4.9063224718726154e-09,
+      "loss": 0.4758,
+      "step": 12185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0371267800898043,
+      "learning_rate": 4.824216340698629e-09,
+      "loss": 0.5083,
+      "step": 12186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9140285052535965,
+      "learning_rate": 4.74280286634099e-09,
+      "loss": 0.4906,
+      "step": 12187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0079408181761793,
+      "learning_rate": 4.662082054441852e-09,
+      "loss": 0.5653,
+      "step": 12188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9830808916635849,
+      "learning_rate": 4.582053910594519e-09,
+      "loss": 0.4774,
+      "step": 12189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9077023704770761,
+      "learning_rate": 4.502718440344556e-09,
+      "loss": 0.4558,
+      "step": 12190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7935871680265658,
+      "learning_rate": 4.4240756491897854e-09,
+      "loss": 0.4481,
+      "step": 12191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8963596426210041,
+      "learning_rate": 4.346125542581403e-09,
+      "loss": 0.4413,
+      "step": 12192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9176413396869422,
+      "learning_rate": 4.268868125919534e-09,
+      "loss": 0.4847,
+      "step": 12193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8930937176633225,
+      "learning_rate": 4.192303404559894e-09,
+      "loss": 0.4818,
+      "step": 12194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9774095396399091,
+      "learning_rate": 4.1164313838060184e-09,
+      "loss": 0.4618,
+      "step": 12195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9304736319360605,
+      "learning_rate": 4.041252068918145e-09,
+      "loss": 0.469,
+      "step": 12196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0330209884154204,
+      "learning_rate": 3.966765465105438e-09,
+      "loss": 0.4876,
+      "step": 12197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8581916788578621,
+      "learning_rate": 3.892971577528215e-09,
+      "loss": 0.4076,
+      "step": 12198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.018750336762057,
+      "learning_rate": 3.819870411302385e-09,
+      "loss": 0.5133,
+      "step": 12199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9301758200353073,
+      "learning_rate": 3.747461971492783e-09,
+      "loss": 0.4803,
+      "step": 12200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9524198887721386,
+      "learning_rate": 3.6757462631176186e-09,
+      "loss": 0.5102,
+      "step": 12201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9998380514117441,
+      "learning_rate": 3.6047232911462506e-09,
+      "loss": 0.4627,
+      "step": 12202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9303895789749116,
+      "learning_rate": 3.5343930605002964e-09,
+      "loss": 0.4983,
+      "step": 12203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9814676213506741,
+      "learning_rate": 3.4647555760547456e-09,
+      "loss": 0.5019,
+      "step": 12204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8899742237254751,
+      "learning_rate": 3.3958108426346283e-09,
+      "loss": 0.4679,
+      "step": 12205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9854215413848316,
+      "learning_rate": 3.327558865017233e-09,
+      "loss": 0.4697,
+      "step": 12206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9949671671252079,
+      "learning_rate": 3.259999647933221e-09,
+      "loss": 0.4586,
+      "step": 12207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9767718153567564,
+      "learning_rate": 3.193133196064402e-09,
+      "loss": 0.5054,
+      "step": 12208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0785275131904468,
+      "learning_rate": 3.126959514043737e-09,
+      "loss": 0.5084,
+      "step": 12209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8674149722017763,
+      "learning_rate": 3.0614786064586676e-09,
+      "loss": 0.4421,
+      "step": 12210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9315561827978559,
+      "learning_rate": 2.996690477844455e-09,
+      "loss": 0.4992,
+      "step": 12211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9269373342663652,
+      "learning_rate": 2.9325951326930614e-09,
+      "loss": 0.5106,
+      "step": 12212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0034662424592324,
+      "learning_rate": 2.8691925754453785e-09,
+      "loss": 0.5164,
+      "step": 12213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9529358562010914,
+      "learning_rate": 2.8064828104956697e-09,
+      "loss": 0.4973,
+      "step": 12214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0100316056822873,
+      "learning_rate": 2.7444658421882375e-09,
+      "loss": 0.516,
+      "step": 12215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.947131493385448,
+      "learning_rate": 2.6831416748229755e-09,
+      "loss": 0.4978,
+      "step": 12216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.013465954899412,
+      "learning_rate": 2.622510312647597e-09,
+      "loss": 0.4745,
+      "step": 12217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0572863965813828,
+      "learning_rate": 2.5625717598642962e-09,
+      "loss": 0.5839,
+      "step": 12218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8450663723277408,
+      "learning_rate": 2.5033260206275277e-09,
+      "loss": 0.4414,
+      "step": 12219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9349043179917564,
+      "learning_rate": 2.4447730990428964e-09,
+      "loss": 0.4565,
+      "step": 12220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8921858303213847,
+      "learning_rate": 2.386912999167157e-09,
+      "loss": 0.4308,
+      "step": 12221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1106808138466342,
+      "learning_rate": 2.329745725010435e-09,
+      "loss": 0.5236,
+      "step": 12222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9107858977787934,
+      "learning_rate": 2.273271280534006e-09,
+      "loss": 0.4344,
+      "step": 12223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0104185102825205,
+      "learning_rate": 2.217489669652517e-09,
+      "loss": 0.5316,
+      "step": 12224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9485027710249507,
+      "learning_rate": 2.1624008962306543e-09,
+      "loss": 0.4754,
+      "step": 12225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.886480688086442,
+      "learning_rate": 2.108004964086474e-09,
+      "loss": 0.4888,
+      "step": 12226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.937058241545652,
+      "learning_rate": 2.0543018769902946e-09,
+      "loss": 0.4611,
+      "step": 12227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9598546147109381,
+      "learning_rate": 2.0012916386613625e-09,
+      "loss": 0.4744,
+      "step": 12228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.974531542912316,
+      "learning_rate": 1.9489742527756263e-09,
+      "loss": 0.5349,
+      "step": 12229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9519705243013669,
+      "learning_rate": 1.8973497229568537e-09,
+      "loss": 0.4761,
+      "step": 12230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9238988769473946,
+      "learning_rate": 1.8464180527844044e-09,
+      "loss": 0.4833,
+      "step": 12231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0111873319625468,
+      "learning_rate": 1.7961792457865668e-09,
+      "loss": 0.5243,
+      "step": 12232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.967548228009322,
+      "learning_rate": 1.7466333054450001e-09,
+      "loss": 0.5089,
+      "step": 12233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.918909362153841,
+      "learning_rate": 1.6977802351936246e-09,
+      "loss": 0.4543,
+      "step": 12234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7785482161984433,
+      "learning_rate": 1.6496200384163996e-09,
+      "loss": 0.3861,
+      "step": 12235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9384527289121802,
+      "learning_rate": 1.6021527184528761e-09,
+      "loss": 0.4751,
+      "step": 12236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9505856087988792,
+      "learning_rate": 1.555378278591535e-09,
+      "loss": 0.4957,
+      "step": 12237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9896863520416174,
+      "learning_rate": 1.5092967220742272e-09,
+      "loss": 0.4905,
+      "step": 12238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0768453855263935,
+      "learning_rate": 1.4639080520939541e-09,
+      "loss": 0.5181,
+      "step": 12239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0133044427198161,
+      "learning_rate": 1.4192122717959777e-09,
+      "loss": 0.5047,
+      "step": 12240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9404999214974434,
+      "learning_rate": 1.3752093842778204e-09,
+      "loss": 0.4799,
+      "step": 12241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0552931054831358,
+      "learning_rate": 1.3318993925881541e-09,
+      "loss": 0.5162,
+      "step": 12242
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0552380665620096,
+      "learning_rate": 1.2892822997301324e-09,
+      "loss": 0.5035,
+      "step": 12243
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9391633731274371,
+      "learning_rate": 1.2473581086558383e-09,
+      "loss": 0.4921,
+      "step": 12244
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0052164661643428,
+      "learning_rate": 1.2061268222707257e-09,
+      "loss": 0.4663,
+      "step": 12245
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9722915792156718,
+      "learning_rate": 1.165588443431398e-09,
+      "loss": 0.4943,
+      "step": 12246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9102927236377634,
+      "learning_rate": 1.12574297494783e-09,
+      "loss": 0.4471,
+      "step": 12247
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9398310361694413,
+      "learning_rate": 1.0865904195822563e-09,
+      "loss": 0.4707,
+      "step": 12248
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.808454573445147,
+      "learning_rate": 1.048130780046952e-09,
+      "loss": 0.398,
+      "step": 12249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9746782424320476,
+      "learning_rate": 1.0103640590064524e-09,
+      "loss": 0.5442,
+      "step": 12250
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9649576296411937,
+      "learning_rate": 9.73290259078663e-10,
+      "loss": 0.5274,
+      "step": 12251
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8107375447971296,
+      "learning_rate": 9.369093828326403e-10,
+      "loss": 0.431,
+      "step": 12252
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0699241121892324,
+      "learning_rate": 9.012214327897006e-10,
+      "loss": 0.5015,
+      "step": 12253
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.905378821625332,
+      "learning_rate": 8.662264114234209e-10,
+      "loss": 0.4383,
+      "step": 12254
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0107933303670742,
+      "learning_rate": 8.319243211585281e-10,
+      "loss": 0.4994,
+      "step": 12255
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8628225975921563,
+      "learning_rate": 7.983151643708997e-10,
+      "loss": 0.4456,
+      "step": 12256
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0056042550071769,
+      "learning_rate": 7.653989433920039e-10,
+      "loss": 0.5566,
+      "step": 12257
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0487775395495176,
+      "learning_rate": 7.331756605011286e-10,
+      "loss": 0.511,
+      "step": 12258
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9034474611671817,
+      "learning_rate": 7.016453179320426e-10,
+      "loss": 0.4756,
+      "step": 12259
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.860531555843401,
+      "learning_rate": 6.708079178685545e-10,
+      "loss": 0.4613,
+      "step": 12260
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.894604087497659,
+      "learning_rate": 6.40663462450064e-10,
+      "loss": 0.4818,
+      "step": 12261
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.956638263621099,
+      "learning_rate": 6.112119537637906e-10,
+      "loss": 0.4995,
+      "step": 12262
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.950141944738411,
+      "learning_rate": 5.82453393850324e-10,
+      "loss": 0.4644,
+      "step": 12263
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9197142556995642,
+      "learning_rate": 5.54387784703625e-10,
+      "loss": 0.4667,
+      "step": 12264
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9409690029323907,
+      "learning_rate": 5.270151282688041e-10,
+      "loss": 0.4442,
+      "step": 12265
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9224085642149458,
+      "learning_rate": 5.003354264421223e-10,
+      "loss": 0.4995,
+      "step": 12266
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9039475339596145,
+      "learning_rate": 4.743486810732111e-10,
+      "loss": 0.4782,
+      "step": 12267
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9500554852385881,
+      "learning_rate": 4.490548939617423e-10,
+      "loss": 0.5374,
+      "step": 12268
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.021632409812602,
+      "learning_rate": 4.2445406686075776e-10,
+      "loss": 0.5037,
+      "step": 12269
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9019301685248371,
+      "learning_rate": 4.005462014766703e-10,
+      "loss": 0.4847,
+      "step": 12270
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0448459387319498,
+      "learning_rate": 3.7733129946371237e-10,
+      "loss": 0.5018,
+      "step": 12271
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9567017288240108,
+      "learning_rate": 3.548093624328175e-10,
+      "loss": 0.5002,
+      "step": 12272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9393370125389402,
+      "learning_rate": 3.32980391943849e-10,
+      "loss": 0.5022,
+      "step": 12273
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0039457861354868,
+      "learning_rate": 3.118443895100409e-10,
+      "loss": 0.4833,
+      "step": 12274
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.876708231374574,
+      "learning_rate": 2.914013565957774e-10,
+      "loss": 0.4747,
+      "step": 12275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0483753373478573,
+      "learning_rate": 2.716512946165928e-10,
+      "loss": 0.5458,
+      "step": 12276
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.854264871286635,
+      "learning_rate": 2.525942049436125e-10,
+      "loss": 0.4194,
+      "step": 12277
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9330203416041252,
+      "learning_rate": 2.3423008889467134e-10,
+      "loss": 0.4675,
+      "step": 12278
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.928285603265314,
+      "learning_rate": 2.165589477443053e-10,
+      "loss": 0.4496,
+      "step": 12279
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.896315861070505,
+      "learning_rate": 1.9958078271709037e-10,
+      "loss": 0.5164,
+      "step": 12280
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9970984789464427,
+      "learning_rate": 1.8329559498875272e-10,
+      "loss": 0.4903,
+      "step": 12281
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9862019097831741,
+      "learning_rate": 1.6770338568838918e-10,
+      "loss": 0.4294,
+      "step": 12282
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8625424437572193,
+      "learning_rate": 1.5280415589624676e-10,
+      "loss": 0.4841,
+      "step": 12283
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6996509177567276,
+      "learning_rate": 1.3859790664483287e-10,
+      "loss": 0.5431,
+      "step": 12284
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.983653006981926,
+      "learning_rate": 1.250846389189153e-10,
+      "loss": 0.4933,
+      "step": 12285
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9691698910333484,
+      "learning_rate": 1.1226435365441212e-10,
+      "loss": 0.478,
+      "step": 12286
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8999462139403965,
+      "learning_rate": 1.0013705174061195e-10,
+      "loss": 0.4665,
+      "step": 12287
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9701615848783587,
+      "learning_rate": 8.870273401684338e-11,
+      "loss": 0.469,
+      "step": 12288
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9788648850371204,
+      "learning_rate": 7.796140127691587e-11,
+      "loss": 0.4688,
+      "step": 12289
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9041305602197512,
+      "learning_rate": 6.791305426356865e-11,
+      "loss": 0.4562,
+      "step": 12290
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9908625759964073,
+      "learning_rate": 5.855769367402176e-11,
+      "loss": 0.5208,
+      "step": 12291
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9195748995284373,
+      "learning_rate": 4.9895320156645445e-11,
+      "loss": 0.407,
+      "step": 12292
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0408210431502338,
+      "learning_rate": 4.192593431096015e-11,
+      "loss": 0.5351,
+      "step": 12293
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9092443947587706,
+      "learning_rate": 3.4649536690967154e-11,
+      "loss": 0.4844,
+      "step": 12294
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9725729468080113,
+      "learning_rate": 2.8066127798487274e-11,
+      "loss": 0.5036,
+      "step": 12295
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9842735837618908,
+      "learning_rate": 2.2175708092042615e-11,
+      "loss": 0.5061,
+      "step": 12296
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9175628272236795,
+      "learning_rate": 1.6978277979085023e-11,
+      "loss": 0.4502,
+      "step": 12297
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0374820658516282,
+      "learning_rate": 1.2473837819326762e-11,
+      "loss": 0.4847,
+      "step": 12298
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8909161564068082,
+      "learning_rate": 8.662387924740501e-12,
+      "loss": 0.4672,
+      "step": 12299
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9415518726489468,
+      "learning_rate": 5.54392855955932e-12,
+      "loss": 0.4625,
+      "step": 12300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9500390955938162,
+      "learning_rate": 3.118459941386931e-12,
+      "loss": 0.4868,
+      "step": 12301
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.922814982524896,
+      "learning_rate": 1.3859822356465657e-12,
+      "loss": 0.5076,
+      "step": 12302
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0370158785381784,
+      "learning_rate": 3.4649556446275656e-13,
+      "loss": 0.5309,
+      "step": 12303
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.844434207744529,
+      "learning_rate": 0.0,
+      "loss": 0.4437,
+      "step": 12304
+    },
+    {
+      "epoch": 1.0,
+      "step": 12304,
+      "total_flos": 1.5084962346172416e+16,
+      "train_loss": 0.5391508782323648,
+      "train_runtime": 261842.5175,
+      "train_samples_per_second": 6.015,
+      "train_steps_per_second": 0.047
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12304,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1846,
+  "total_flos": 1.5084962346172416e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}