diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,26294 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998001465591899,
+  "eval_steps": 500,
+  "global_step": 3752,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.090684425124426,
+      "learning_rate": 1.7699115044247788e-07,
+      "loss": 1.5788,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.519008330686292,
+      "learning_rate": 3.5398230088495575e-07,
+      "loss": 1.3421,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.082133725597737,
+      "learning_rate": 5.309734513274336e-07,
+      "loss": 1.3304,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.329335589436722,
+      "learning_rate": 7.079646017699115e-07,
+      "loss": 1.2285,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.700651089621816,
+      "learning_rate": 8.849557522123895e-07,
+      "loss": 1.6597,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.311559104193037,
+      "learning_rate": 1.0619469026548673e-06,
+      "loss": 1.1015,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.592171733731611,
+      "learning_rate": 1.2389380530973452e-06,
+      "loss": 1.2313,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.956089783443849,
+      "learning_rate": 1.415929203539823e-06,
+      "loss": 1.3023,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.873877151046992,
+      "learning_rate": 1.592920353982301e-06,
+      "loss": 1.5633,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.858226116216994,
+      "learning_rate": 1.769911504424779e-06,
+      "loss": 1.1939,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.19890955198723,
+      "learning_rate": 1.9469026548672567e-06,
+      "loss": 1.1438,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.011000786159999,
+      "learning_rate": 2.1238938053097345e-06,
+      "loss": 1.0597,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.316340634019374,
+      "learning_rate": 2.3008849557522127e-06,
+      "loss": 1.5034,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.760806439558829,
+      "learning_rate": 2.4778761061946905e-06,
+      "loss": 0.9258,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.93399681137179,
+      "learning_rate": 2.6548672566371687e-06,
+      "loss": 1.0598,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.2056010378090525,
+      "learning_rate": 2.831858407079646e-06,
+      "loss": 1.0163,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.047896205155173,
+      "learning_rate": 3.0088495575221242e-06,
+      "loss": 1.0555,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.02944744657154,
+      "learning_rate": 3.185840707964602e-06,
+      "loss": 1.1692,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.787216642150343,
+      "learning_rate": 3.36283185840708e-06,
+      "loss": 1.16,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.613549048879829,
+      "learning_rate": 3.539823008849558e-06,
+      "loss": 1.5257,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.819009275475223,
+      "learning_rate": 3.7168141592920357e-06,
+      "loss": 0.919,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.073357632808168,
+      "learning_rate": 3.8938053097345135e-06,
+      "loss": 1.3794,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.467838749387505,
+      "learning_rate": 4.070796460176992e-06,
+      "loss": 1.3949,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2879864012511066,
+      "learning_rate": 4.247787610619469e-06,
+      "loss": 0.9467,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.99812501637886,
+      "learning_rate": 4.424778761061948e-06,
+      "loss": 0.8098,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6491762139893322,
+      "learning_rate": 4.6017699115044254e-06,
+      "loss": 0.8966,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4919306450116325,
+      "learning_rate": 4.778761061946903e-06,
+      "loss": 0.9472,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.842555496626711,
+      "learning_rate": 4.955752212389381e-06,
+      "loss": 0.9868,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.719456725156765,
+      "learning_rate": 5.132743362831859e-06,
+      "loss": 0.8978,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5236981643505656,
+      "learning_rate": 5.309734513274337e-06,
+      "loss": 0.7909,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.234164845636012,
+      "learning_rate": 5.486725663716814e-06,
+      "loss": 0.826,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3649919359963656,
+      "learning_rate": 5.663716814159292e-06,
+      "loss": 0.9121,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8490673030682476,
+      "learning_rate": 5.840707964601771e-06,
+      "loss": 1.3237,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.866859739886757,
+      "learning_rate": 6.0176991150442484e-06,
+      "loss": 0.8036,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1432681263541755,
+      "learning_rate": 6.194690265486726e-06,
+      "loss": 1.3656,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.550388537934869,
+      "learning_rate": 6.371681415929204e-06,
+      "loss": 0.726,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2828179523063805,
+      "learning_rate": 6.548672566371682e-06,
+      "loss": 1.2553,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.477229174035435,
+      "learning_rate": 6.72566371681416e-06,
+      "loss": 0.8555,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4373650656964028,
+      "learning_rate": 6.902654867256637e-06,
+      "loss": 0.8159,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.021075995406555,
+      "learning_rate": 7.079646017699116e-06,
+      "loss": 1.3141,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.840892133630191,
+      "learning_rate": 7.256637168141594e-06,
+      "loss": 1.3171,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.00972490683521,
+      "learning_rate": 7.4336283185840714e-06,
+      "loss": 1.3237,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7549480309140906,
+      "learning_rate": 7.610619469026549e-06,
+      "loss": 0.7743,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.520450468793066,
+      "learning_rate": 7.787610619469027e-06,
+      "loss": 0.7806,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4800923637162815,
+      "learning_rate": 7.964601769911505e-06,
+      "loss": 0.8205,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.733597088379438,
+      "learning_rate": 8.141592920353984e-06,
+      "loss": 1.288,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7232924484407044,
+      "learning_rate": 8.31858407079646e-06,
+      "loss": 1.3401,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.755340728171708,
+      "learning_rate": 8.495575221238938e-06,
+      "loss": 1.2202,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7346258380970356,
+      "learning_rate": 8.672566371681418e-06,
+      "loss": 0.9501,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8373080457609823,
+      "learning_rate": 8.849557522123895e-06,
+      "loss": 1.2661,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.373404076494007,
+      "learning_rate": 9.026548672566371e-06,
+      "loss": 0.8717,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3290693761388876,
+      "learning_rate": 9.203539823008851e-06,
+      "loss": 0.7425,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3752898978022,
+      "learning_rate": 9.380530973451329e-06,
+      "loss": 0.7494,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9078460465682654,
+      "learning_rate": 9.557522123893806e-06,
+      "loss": 1.2372,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5251420794591337,
+      "learning_rate": 9.734513274336284e-06,
+      "loss": 0.766,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.088283449527855,
+      "learning_rate": 9.911504424778762e-06,
+      "loss": 0.76,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.172417460190972,
+      "learning_rate": 1.008849557522124e-05,
+      "loss": 0.7267,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3659473269657325,
+      "learning_rate": 1.0265486725663717e-05,
+      "loss": 0.7175,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3063020295016146,
+      "learning_rate": 1.0442477876106197e-05,
+      "loss": 0.6973,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6651970794756377,
+      "learning_rate": 1.0619469026548675e-05,
+      "loss": 1.2723,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7712869744345063,
+      "learning_rate": 1.079646017699115e-05,
+      "loss": 1.2098,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0642094029919353,
+      "learning_rate": 1.0973451327433629e-05,
+      "loss": 0.7182,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2966021317620684,
+      "learning_rate": 1.1150442477876106e-05,
+      "loss": 0.755,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5265064376553625,
+      "learning_rate": 1.1327433628318584e-05,
+      "loss": 0.716,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.221027915793282,
+      "learning_rate": 1.1504424778761064e-05,
+      "loss": 0.7051,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.42779764165399,
+      "learning_rate": 1.1681415929203541e-05,
+      "loss": 0.7877,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0674499431141053,
+      "learning_rate": 1.1858407079646019e-05,
+      "loss": 0.7062,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0740102139377186,
+      "learning_rate": 1.2035398230088497e-05,
+      "loss": 0.7336,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.458129159634638,
+      "learning_rate": 1.2212389380530973e-05,
+      "loss": 0.7166,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.213058833976562,
+      "learning_rate": 1.2389380530973452e-05,
+      "loss": 0.7598,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.245878495803006,
+      "learning_rate": 1.256637168141593e-05,
+      "loss": 0.6564,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1430202186828584,
+      "learning_rate": 1.2743362831858408e-05,
+      "loss": 1.2399,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0230875541553797,
+      "learning_rate": 1.2920353982300886e-05,
+      "loss": 0.6177,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7913042254300353,
+      "learning_rate": 1.3097345132743363e-05,
+      "loss": 1.2794,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.387638899530596,
+      "learning_rate": 1.3274336283185843e-05,
+      "loss": 0.8285,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2389570990155407,
+      "learning_rate": 1.345132743362832e-05,
+      "loss": 0.7135,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.298907720453503,
+      "learning_rate": 1.3628318584070797e-05,
+      "loss": 0.7084,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.775617121283823,
+      "learning_rate": 1.3805309734513275e-05,
+      "loss": 1.2448,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.386858949995019,
+      "learning_rate": 1.3982300884955752e-05,
+      "loss": 0.7238,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6819320136133937,
+      "learning_rate": 1.4159292035398232e-05,
+      "loss": 1.2809,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.722767255846862,
+      "learning_rate": 1.433628318584071e-05,
+      "loss": 1.2865,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1263862266632625,
+      "learning_rate": 1.4513274336283187e-05,
+      "loss": 0.7299,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3073897673747523,
+      "learning_rate": 1.4690265486725665e-05,
+      "loss": 0.7914,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2455784294248526,
+      "learning_rate": 1.4867256637168143e-05,
+      "loss": 0.7518,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1986343064864595,
+      "learning_rate": 1.5044247787610619e-05,
+      "loss": 1.2888,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5654560084230345,
+      "learning_rate": 1.5221238938053098e-05,
+      "loss": 1.2587,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2108272146313706,
+      "learning_rate": 1.5398230088495576e-05,
+      "loss": 0.7222,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1678946159049555,
+      "learning_rate": 1.5575221238938054e-05,
+      "loss": 0.7344,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4092633072030702,
+      "learning_rate": 1.5752212389380532e-05,
+      "loss": 0.7574,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.31894860108094,
+      "learning_rate": 1.592920353982301e-05,
+      "loss": 0.7265,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8736828797596337,
+      "learning_rate": 1.6106194690265487e-05,
+      "loss": 1.2067,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1095272815850348,
+      "learning_rate": 1.628318584070797e-05,
+      "loss": 0.7018,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.146911840129064,
+      "learning_rate": 1.6460176991150443e-05,
+      "loss": 0.7166,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.122250886568982,
+      "learning_rate": 1.663716814159292e-05,
+      "loss": 0.7019,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.165161196042364,
+      "learning_rate": 1.68141592920354e-05,
+      "loss": 0.664,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4247079849773665,
+      "learning_rate": 1.6991150442477876e-05,
+      "loss": 0.6416,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.146501410615504,
+      "learning_rate": 1.7168141592920354e-05,
+      "loss": 1.242,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.121645523304816,
+      "learning_rate": 1.7345132743362835e-05,
+      "loss": 0.6851,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6784314031223393,
+      "learning_rate": 1.7522123893805313e-05,
+      "loss": 1.2868,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1852033489573217,
+      "learning_rate": 1.769911504424779e-05,
+      "loss": 0.7263,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2385482265764503,
+      "learning_rate": 1.7876106194690265e-05,
+      "loss": 0.6177,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1101619284530115,
+      "learning_rate": 1.8053097345132743e-05,
+      "loss": 0.639,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1864743212955813,
+      "learning_rate": 1.823008849557522e-05,
+      "loss": 0.7705,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8128167917607776,
+      "learning_rate": 1.8407079646017702e-05,
+      "loss": 1.1644,
+      "step": 104
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.053354296358307,
+      "learning_rate": 1.858407079646018e-05,
+      "loss": 0.6187,
+      "step": 105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.009914047800546,
+      "learning_rate": 1.8761061946902657e-05,
+      "loss": 0.6491,
+      "step": 106
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.334637476008098,
+      "learning_rate": 1.8938053097345135e-05,
+      "loss": 0.663,
+      "step": 107
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0300286583401546,
+      "learning_rate": 1.9115044247787613e-05,
+      "loss": 1.1815,
+      "step": 108
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7117980652320455,
+      "learning_rate": 1.929203539823009e-05,
+      "loss": 1.2481,
+      "step": 109
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6197607608675293,
+      "learning_rate": 1.946902654867257e-05,
+      "loss": 0.655,
+      "step": 110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.33951445362461,
+      "learning_rate": 1.9646017699115046e-05,
+      "loss": 0.6469,
+      "step": 111
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6397153166632887,
+      "learning_rate": 1.9823008849557524e-05,
+      "loss": 0.7485,
+      "step": 112
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1787658844045916,
+      "learning_rate": 2e-05,
+      "loss": 1.2212,
+      "step": 113
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2056199463151525,
+      "learning_rate": 1.9999996273461508e-05,
+      "loss": 0.6702,
+      "step": 114
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6290815147884268,
+      "learning_rate": 1.9999985093848808e-05,
+      "loss": 0.687,
+      "step": 115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4272512376334077,
+      "learning_rate": 1.999996646117023e-05,
+      "loss": 0.7505,
+      "step": 116
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.451698468323556,
+      "learning_rate": 1.9999940375439664e-05,
+      "loss": 0.7743,
+      "step": 117
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6577338690665875,
+      "learning_rate": 1.999990683667655e-05,
+      "loss": 0.7358,
+      "step": 118
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2207677477743517,
+      "learning_rate": 1.9999865844905887e-05,
+      "loss": 0.7142,
+      "step": 119
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.191648154451648,
+      "learning_rate": 1.9999817400158227e-05,
+      "loss": 0.5986,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2741067429827093,
+      "learning_rate": 1.9999761502469673e-05,
+      "loss": 0.7231,
+      "step": 121
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.616409257123111,
+      "learning_rate": 1.9999698151881887e-05,
+      "loss": 1.2142,
+      "step": 122
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.799937397997997,
+      "learning_rate": 1.9999627348442085e-05,
+      "loss": 0.7338,
+      "step": 123
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3435523176529767,
+      "learning_rate": 1.9999549092203038e-05,
+      "loss": 0.6699,
+      "step": 124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3038962159544485,
+      "learning_rate": 1.999946338322307e-05,
+      "loss": 0.7273,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2959289108294563,
+      "learning_rate": 1.999937022156606e-05,
+      "loss": 0.7111,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6383846224794882,
+      "learning_rate": 1.999926960730145e-05,
+      "loss": 1.2518,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3960622651931254,
+      "learning_rate": 1.9999161540504212e-05,
+      "loss": 0.7589,
+      "step": 128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.309785004519786,
+      "learning_rate": 1.9999046021254905e-05,
+      "loss": 0.7951,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0226636639960933,
+      "learning_rate": 1.999892304963962e-05,
+      "loss": 0.6905,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.745394750824896,
+      "learning_rate": 1.999879262575e-05,
+      "loss": 1.2213,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5757579963876127,
+      "learning_rate": 1.9998654749683267e-05,
+      "loss": 1.2486,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3551916278967977,
+      "learning_rate": 1.9998509421542175e-05,
+      "loss": 0.7208,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1289671218852155,
+      "learning_rate": 1.9998356641435037e-05,
+      "loss": 0.7199,
+      "step": 134
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.258247463088468,
+      "learning_rate": 1.999819640947572e-05,
+      "loss": 0.6936,
+      "step": 135
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2151704648128305,
+      "learning_rate": 1.9998028725783647e-05,
+      "loss": 0.7338,
+      "step": 136
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.518657123568622,
+      "learning_rate": 1.9997853590483793e-05,
+      "loss": 1.2591,
+      "step": 137
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.426324887924248,
+      "learning_rate": 1.9997671003706688e-05,
+      "loss": 0.6664,
+      "step": 138
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6959116362224944,
+      "learning_rate": 1.9997480965588417e-05,
+      "loss": 1.2568,
+      "step": 139
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2475359514136746,
+      "learning_rate": 1.999728347627062e-05,
+      "loss": 0.6852,
+      "step": 140
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.220180353249594,
+      "learning_rate": 1.999707853590048e-05,
+      "loss": 0.7349,
+      "step": 141
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.33726125302252,
+      "learning_rate": 1.9996866144630746e-05,
+      "loss": 0.7314,
+      "step": 142
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.068211138506116,
+      "learning_rate": 1.9996646302619712e-05,
+      "loss": 0.6716,
+      "step": 143
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.22871312738648,
+      "learning_rate": 1.999641901003123e-05,
+      "loss": 0.7317,
+      "step": 144
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.133315679224587,
+      "learning_rate": 1.99961842670347e-05,
+      "loss": 0.7495,
+      "step": 145
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9394792828857943,
+      "learning_rate": 1.999594207380508e-05,
+      "loss": 0.7151,
+      "step": 146
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1152095767872066,
+      "learning_rate": 1.9995692430522877e-05,
+      "loss": 0.6717,
+      "step": 147
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1536231121422023,
+      "learning_rate": 1.9995435337374155e-05,
+      "loss": 0.7221,
+      "step": 148
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.127017130743878,
+      "learning_rate": 1.9995170794550524e-05,
+      "loss": 0.6635,
+      "step": 149
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.547272216136701,
+      "learning_rate": 1.999489880224915e-05,
+      "loss": 1.349,
+      "step": 150
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3396407325034247,
+      "learning_rate": 1.9994619360672757e-05,
+      "loss": 0.6719,
+      "step": 151
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.710583236435864,
+      "learning_rate": 1.9994332470029608e-05,
+      "loss": 1.2202,
+      "step": 152
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.229667493699025,
+      "learning_rate": 1.999403813053353e-05,
+      "loss": 0.7675,
+      "step": 153
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3161212261168678,
+      "learning_rate": 1.9993736342403887e-05,
+      "loss": 0.7561,
+      "step": 154
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.443435197234292,
+      "learning_rate": 1.9993427105865615e-05,
+      "loss": 0.7703,
+      "step": 155
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1441941914202824,
+      "learning_rate": 1.9993110421149183e-05,
+      "loss": 1.2963,
+      "step": 156
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.071069731651454,
+      "learning_rate": 1.9992786288490627e-05,
+      "loss": 1.2168,
+      "step": 157
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8850220508713527,
+      "learning_rate": 1.9992454708131514e-05,
+      "loss": 1.2489,
+      "step": 158
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2696582600540127,
+      "learning_rate": 1.9992115680318983e-05,
+      "loss": 0.6358,
+      "step": 159
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.205713954317766,
+      "learning_rate": 1.9991769205305712e-05,
+      "loss": 0.6934,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.663213865356555,
+      "learning_rate": 1.9991415283349926e-05,
+      "loss": 1.2546,
+      "step": 161
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.296014874504943,
+      "learning_rate": 1.9991053914715414e-05,
+      "loss": 0.7461,
+      "step": 162
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.208399516687059,
+      "learning_rate": 1.99906850996715e-05,
+      "loss": 0.7306,
+      "step": 163
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.105556334667362,
+      "learning_rate": 1.9990308838493065e-05,
+      "loss": 1.2286,
+      "step": 164
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.419739046342952,
+      "learning_rate": 1.9989925131460545e-05,
+      "loss": 0.7208,
+      "step": 165
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.369637827271588,
+      "learning_rate": 1.998953397885992e-05,
+      "loss": 0.7155,
+      "step": 166
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2839169892981923,
+      "learning_rate": 1.9989135380982708e-05,
+      "loss": 0.712,
+      "step": 167
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0406847579091973,
+      "learning_rate": 1.9988729338126e-05,
+      "loss": 0.6768,
+      "step": 168
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2808094507854153,
+      "learning_rate": 1.9988315850592418e-05,
+      "loss": 0.6867,
+      "step": 169
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.765056568535897,
+      "learning_rate": 1.998789491869013e-05,
+      "loss": 1.263,
+      "step": 170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0349350091853027,
+      "learning_rate": 1.9987466542732873e-05,
+      "loss": 1.2412,
+      "step": 171
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.177726438160662,
+      "learning_rate": 1.998703072303991e-05,
+      "loss": 0.7192,
+      "step": 172
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.231513421709446,
+      "learning_rate": 1.9986587459936065e-05,
+      "loss": 0.6808,
+      "step": 173
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1670966881752527,
+      "learning_rate": 1.9986136753751703e-05,
+      "loss": 0.7051,
+      "step": 174
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1177400532013744,
+      "learning_rate": 1.998567860482274e-05,
+      "loss": 0.7382,
+      "step": 175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0840207355179454,
+      "learning_rate": 1.9985213013490638e-05,
+      "loss": 1.2849,
+      "step": 176
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6489271310175537,
+      "learning_rate": 1.9984739980102405e-05,
+      "loss": 1.1926,
+      "step": 177
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.201850698975346,
+      "learning_rate": 1.9984259505010593e-05,
+      "loss": 1.2724,
+      "step": 178
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4991007502026,
+      "learning_rate": 1.9983771588573312e-05,
+      "loss": 1.1777,
+      "step": 179
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.338773742178541,
+      "learning_rate": 1.9983276231154204e-05,
+      "loss": 0.6437,
+      "step": 180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8480634578422444,
+      "learning_rate": 1.998277343312246e-05,
+      "loss": 0.7413,
+      "step": 181
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3490038225330725,
+      "learning_rate": 1.998226319485283e-05,
+      "loss": 0.6929,
+      "step": 182
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.663196311693036,
+      "learning_rate": 1.998174551672559e-05,
+      "loss": 1.1433,
+      "step": 183
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0977144664963165,
+      "learning_rate": 1.9981220399126572e-05,
+      "loss": 1.2586,
+      "step": 184
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8802943728628394,
+      "learning_rate": 1.9980687842447146e-05,
+      "loss": 1.1828,
+      "step": 185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.454971582475382,
+      "learning_rate": 1.9980147847084237e-05,
+      "loss": 0.7028,
+      "step": 186
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7663598609940565,
+      "learning_rate": 1.9979600413440303e-05,
+      "loss": 1.1869,
+      "step": 187
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2630438871933363,
+      "learning_rate": 1.9979045541923354e-05,
+      "loss": 0.7055,
+      "step": 188
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.453232344922628,
+      "learning_rate": 1.9978483232946937e-05,
+      "loss": 0.7125,
+      "step": 189
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9387594157592865,
+      "learning_rate": 1.9977913486930147e-05,
+      "loss": 0.6035,
+      "step": 190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0723711172682653,
+      "learning_rate": 1.997733630429762e-05,
+      "loss": 0.6841,
+      "step": 191
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2178760578891206,
+      "learning_rate": 1.9976751685479533e-05,
+      "loss": 0.6814,
+      "step": 192
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.080581079118226,
+      "learning_rate": 1.9976159630911608e-05,
+      "loss": 0.6439,
+      "step": 193
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7002235608147327,
+      "learning_rate": 1.9975560141035112e-05,
+      "loss": 1.2119,
+      "step": 194
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.020665589407722,
+      "learning_rate": 1.9974953216296842e-05,
+      "loss": 0.7255,
+      "step": 195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.075437181645661,
+      "learning_rate": 1.9974338857149147e-05,
+      "loss": 0.7038,
+      "step": 196
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.776977708092158,
+      "learning_rate": 1.9973717064049912e-05,
+      "loss": 1.2277,
+      "step": 197
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.283411570734202,
+      "learning_rate": 1.9973087837462574e-05,
+      "loss": 0.7043,
+      "step": 198
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1898084419915174,
+      "learning_rate": 1.9972451177856084e-05,
+      "loss": 0.5967,
+      "step": 199
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2093423562134475,
+      "learning_rate": 1.997180708570496e-05,
+      "loss": 0.7362,
+      "step": 200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2370025557197146,
+      "learning_rate": 1.9971155561489247e-05,
+      "loss": 0.6981,
+      "step": 201
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9372364253216714,
+      "learning_rate": 1.997049660569453e-05,
+      "loss": 0.6481,
+      "step": 202
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.306123335704072,
+      "learning_rate": 1.9969830218811933e-05,
+      "loss": 0.6456,
+      "step": 203
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.316640297611442,
+      "learning_rate": 1.9969156401338124e-05,
+      "loss": 0.7137,
+      "step": 204
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.160078265361317,
+      "learning_rate": 1.99684751537753e-05,
+      "loss": 0.6877,
+      "step": 205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8261997839301087,
+      "learning_rate": 1.99677864766312e-05,
+      "loss": 1.2168,
+      "step": 206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1244118185532463,
+      "learning_rate": 1.99670903704191e-05,
+      "loss": 0.6884,
+      "step": 207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.147758820649942,
+      "learning_rate": 1.9966386835657817e-05,
+      "loss": 0.678,
+      "step": 208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.044086494228525,
+      "learning_rate": 1.9965675872871697e-05,
+      "loss": 1.2469,
+      "step": 209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.250913776584532,
+      "learning_rate": 1.9964957482590627e-05,
+      "loss": 0.6511,
+      "step": 210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3058410434164887,
+      "learning_rate": 1.9964231665350034e-05,
+      "loss": 0.6533,
+      "step": 211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9721873723857783,
+      "learning_rate": 1.9963498421690864e-05,
+      "loss": 0.6709,
+      "step": 212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.995308366220785,
+      "learning_rate": 1.996275775215962e-05,
+      "loss": 0.6753,
+      "step": 213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0978553065191385,
+      "learning_rate": 1.9962009657308322e-05,
+      "loss": 0.6909,
+      "step": 214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2298195434402093,
+      "learning_rate": 1.9961254137694535e-05,
+      "loss": 0.6879,
+      "step": 215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6682640134316795,
+      "learning_rate": 1.9960491193881354e-05,
+      "loss": 1.267,
+      "step": 216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.608730358746501,
+      "learning_rate": 1.99597208264374e-05,
+      "loss": 1.221,
+      "step": 217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.491591588462461,
+      "learning_rate": 1.9958943035936844e-05,
+      "loss": 1.1927,
+      "step": 218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3179338089885015,
+      "learning_rate": 1.995815782295937e-05,
+      "loss": 0.6897,
+      "step": 219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3526556561022534,
+      "learning_rate": 1.995736518809021e-05,
+      "loss": 0.7026,
+      "step": 220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3386070236552734,
+      "learning_rate": 1.9956565131920112e-05,
+      "loss": 0.6504,
+      "step": 221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.136702321261758,
+      "learning_rate": 1.9955757655045373e-05,
+      "loss": 0.5965,
+      "step": 222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.96501338603252,
+      "learning_rate": 1.995494275806781e-05,
+      "loss": 0.6816,
+      "step": 223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1288786062302156,
+      "learning_rate": 1.995412044159477e-05,
+      "loss": 1.1976,
+      "step": 224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.411255713715843,
+      "learning_rate": 1.9953290706239134e-05,
+      "loss": 0.7001,
+      "step": 225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.860788059358995,
+      "learning_rate": 1.9952453552619305e-05,
+      "loss": 1.227,
+      "step": 226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0557894056391555,
+      "learning_rate": 1.9951608981359226e-05,
+      "loss": 0.7276,
+      "step": 227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.655284476369075,
+      "learning_rate": 1.995075699308836e-05,
+      "loss": 1.2122,
+      "step": 228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.588038908772318,
+      "learning_rate": 1.9949897588441702e-05,
+      "loss": 1.214,
+      "step": 229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.040413397983155,
+      "learning_rate": 1.9949030768059773e-05,
+      "loss": 1.2439,
+      "step": 230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.577550999098623,
+      "learning_rate": 1.9948156532588614e-05,
+      "loss": 1.1861,
+      "step": 231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7251839203510295,
+      "learning_rate": 1.9947274882679807e-05,
+      "loss": 1.2042,
+      "step": 232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.780816076317283,
+      "learning_rate": 1.994638581899045e-05,
+      "loss": 0.6652,
+      "step": 233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.340900834611378,
+      "learning_rate": 1.994548934218317e-05,
+      "loss": 0.6818,
+      "step": 234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4479131235437976,
+      "learning_rate": 1.9944585452926116e-05,
+      "loss": 0.7477,
+      "step": 235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.990595739742056,
+      "learning_rate": 1.994367415189297e-05,
+      "loss": 0.6992,
+      "step": 236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.379514931570371,
+      "learning_rate": 1.994275543976292e-05,
+      "loss": 1.2266,
+      "step": 237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9879375514757074,
+      "learning_rate": 1.99418293172207e-05,
+      "loss": 1.2189,
+      "step": 238
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0949454730162613,
+      "learning_rate": 1.994089578495655e-05,
+      "loss": 0.7298,
+      "step": 239
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2676273754308216,
+      "learning_rate": 1.993995484366624e-05,
+      "loss": 0.5995,
+      "step": 240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1433144550274075,
+      "learning_rate": 1.9939006494051062e-05,
+      "loss": 0.719,
+      "step": 241
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1790930889368942,
+      "learning_rate": 1.993805073681783e-05,
+      "loss": 0.6662,
+      "step": 242
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3340106297542857,
+      "learning_rate": 1.9937087572678874e-05,
+      "loss": 1.2642,
+      "step": 243
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.281125082682405,
+      "learning_rate": 1.993611700235205e-05,
+      "loss": 0.7453,
+      "step": 244
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0532922971849006,
+      "learning_rate": 1.993513902656073e-05,
+      "loss": 1.2535,
+      "step": 245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0397686196852587,
+      "learning_rate": 1.9934153646033805e-05,
+      "loss": 0.6284,
+      "step": 246
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.277239524456707,
+      "learning_rate": 1.9933160861505692e-05,
+      "loss": 0.7232,
+      "step": 247
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.098477929234546,
+      "learning_rate": 1.9932160673716317e-05,
+      "loss": 0.6328,
+      "step": 248
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.217851880422591,
+      "learning_rate": 1.9931153083411127e-05,
+      "loss": 1.2891,
+      "step": 249
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7293821970622445,
+      "learning_rate": 1.993013809134109e-05,
+      "loss": 0.674,
+      "step": 250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.079731982736342,
+      "learning_rate": 1.9929115698262686e-05,
+      "loss": 0.6798,
+      "step": 251
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.969783845205896,
+      "learning_rate": 1.992808590493791e-05,
+      "loss": 0.5693,
+      "step": 252
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.930763650689499,
+      "learning_rate": 1.992704871213428e-05,
+      "loss": 1.2723,
+      "step": 253
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8464903271337376,
+      "learning_rate": 1.9926004120624816e-05,
+      "loss": 0.5974,
+      "step": 254
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.94684785985451,
+      "learning_rate": 1.9924952131188072e-05,
+      "loss": 1.2187,
+      "step": 255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.578928852037161,
+      "learning_rate": 1.9923892744608092e-05,
+      "loss": 1.1745,
+      "step": 256
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5185863636731955,
+      "learning_rate": 1.992282596167445e-05,
+      "loss": 0.7475,
+      "step": 257
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.169773416848961,
+      "learning_rate": 1.9921751783182227e-05,
+      "loss": 0.6511,
+      "step": 258
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.123231477210999,
+      "learning_rate": 1.9920670209932022e-05,
+      "loss": 0.7966,
+      "step": 259
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9225939252441557,
+      "learning_rate": 1.9919581242729928e-05,
+      "loss": 0.6701,
+      "step": 260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.976712412075055,
+      "learning_rate": 1.991848488238757e-05,
+      "loss": 1.2416,
+      "step": 261
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9355046165141876,
+      "learning_rate": 1.991738112972207e-05,
+      "loss": 1.2125,
+      "step": 262
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9517646911303252,
+      "learning_rate": 1.9916269985556065e-05,
+      "loss": 0.6596,
+      "step": 263
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1044745298622543,
+      "learning_rate": 1.99151514507177e-05,
+      "loss": 0.6194,
+      "step": 264
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.731161564237486,
+      "learning_rate": 1.9914025526040624e-05,
+      "loss": 1.2182,
+      "step": 265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.517366099196497,
+      "learning_rate": 1.9912892212364003e-05,
+      "loss": 1.1296,
+      "step": 266
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.188157171006201,
+      "learning_rate": 1.99117515105325e-05,
+      "loss": 0.7657,
+      "step": 267
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.054663110007371,
+      "learning_rate": 1.991060342139629e-05,
+      "loss": 1.1918,
+      "step": 268
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.626914101659628,
+      "learning_rate": 1.990944794581105e-05,
+      "loss": 1.2117,
+      "step": 269
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1823683185110467,
+      "learning_rate": 1.990828508463797e-05,
+      "loss": 0.7365,
+      "step": 270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0579149853823595,
+      "learning_rate": 1.9907114838743737e-05,
+      "loss": 1.1788,
+      "step": 271
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9031766852121685,
+      "learning_rate": 1.9905937209000545e-05,
+      "loss": 0.6641,
+      "step": 272
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.958400980282488,
+      "learning_rate": 1.9904752196286087e-05,
+      "loss": 0.629,
+      "step": 273
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7692778002972753,
+      "learning_rate": 1.9903559801483566e-05,
+      "loss": 1.2626,
+      "step": 274
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0649723804225344,
+      "learning_rate": 1.9902360025481683e-05,
+      "loss": 0.6564,
+      "step": 275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2580767370448243,
+      "learning_rate": 1.990115286917464e-05,
+      "loss": 0.7461,
+      "step": 276
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.099818702821693,
+      "learning_rate": 1.9899938333462136e-05,
+      "loss": 1.2548,
+      "step": 277
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8698829348324169,
+      "learning_rate": 1.9898716419249376e-05,
+      "loss": 0.6504,
+      "step": 278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9075731669391893,
+      "learning_rate": 1.9897487127447066e-05,
+      "loss": 0.7159,
+      "step": 279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.098717477431955,
+      "learning_rate": 1.9896250458971404e-05,
+      "loss": 0.6993,
+      "step": 280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9327977765857778,
+      "learning_rate": 1.9895006414744084e-05,
+      "loss": 0.605,
+      "step": 281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7875078658343053,
+      "learning_rate": 1.989375499569231e-05,
+      "loss": 1.1416,
+      "step": 282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.924418138928877,
+      "learning_rate": 1.9892496202748766e-05,
+      "loss": 1.2154,
+      "step": 283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.020642775447524,
+      "learning_rate": 1.9891230036851648e-05,
+      "loss": 0.6229,
+      "step": 284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0327506572732905,
+      "learning_rate": 1.9889956498944633e-05,
+      "loss": 0.6006,
+      "step": 285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1318944858241666,
+      "learning_rate": 1.98886755899769e-05,
+      "loss": 0.6346,
+      "step": 286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8818148103927754,
+      "learning_rate": 1.9887387310903125e-05,
+      "loss": 1.2418,
+      "step": 287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2084314829549,
+      "learning_rate": 1.9886091662683468e-05,
+      "loss": 0.7261,
+      "step": 288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8476115831060833,
+      "learning_rate": 1.988478864628358e-05,
+      "loss": 0.6923,
+      "step": 289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0735746883862882,
+      "learning_rate": 1.9883478262674624e-05,
+      "loss": 0.7149,
+      "step": 290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0646018628217804,
+      "learning_rate": 1.9882160512833223e-05,
+      "loss": 0.6719,
+      "step": 291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0543313556291043,
+      "learning_rate": 1.988083539774152e-05,
+      "loss": 1.241,
+      "step": 292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.923731279712935,
+      "learning_rate": 1.9879502918387122e-05,
+      "loss": 1.22,
+      "step": 293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9178601241765318,
+      "learning_rate": 1.987816307576314e-05,
+      "loss": 0.6551,
+      "step": 294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0463930542058613,
+      "learning_rate": 1.9876815870868173e-05,
+      "loss": 1.1982,
+      "step": 295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9410521710992616,
+      "learning_rate": 1.98754613047063e-05,
+      "loss": 0.6795,
+      "step": 296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.110068739987925,
+      "learning_rate": 1.987409937828709e-05,
+      "loss": 0.6795,
+      "step": 297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.846803702741638,
+      "learning_rate": 1.9872730092625593e-05,
+      "loss": 1.2599,
+      "step": 298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0862711775875407,
+      "learning_rate": 1.9871353448742357e-05,
+      "loss": 0.6491,
+      "step": 299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6296922602597292,
+      "learning_rate": 1.98699694476634e-05,
+      "loss": 1.2445,
+      "step": 300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.223697056840505,
+      "learning_rate": 1.9868578090420226e-05,
+      "loss": 0.6903,
+      "step": 301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9826697042463006,
+      "learning_rate": 1.986717937804983e-05,
+      "loss": 0.6885,
+      "step": 302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.508668604918639,
+      "learning_rate": 1.986577331159468e-05,
+      "loss": 1.2393,
+      "step": 303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1370521993990277,
+      "learning_rate": 1.9864359892102727e-05,
+      "loss": 1.2195,
+      "step": 304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2397082869393796,
+      "learning_rate": 1.9862939120627408e-05,
+      "loss": 0.7599,
+      "step": 305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2561426870402244,
+      "learning_rate": 1.986151099822763e-05,
+      "loss": 0.6832,
+      "step": 306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.45155727781265,
+      "learning_rate": 1.9860075525967787e-05,
+      "loss": 0.6974,
+      "step": 307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0926867769543076,
+      "learning_rate": 1.9858632704917743e-05,
+      "loss": 0.6889,
+      "step": 308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9584870784195447,
+      "learning_rate": 1.985718253615285e-05,
+      "loss": 0.6949,
+      "step": 309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3252329431026855,
+      "learning_rate": 1.9855725020753925e-05,
+      "loss": 0.6656,
+      "step": 310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.12875106338558,
+      "learning_rate": 1.9854260159807268e-05,
+      "loss": 0.6962,
+      "step": 311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9334529406557572,
+      "learning_rate": 1.985278795440465e-05,
+      "loss": 0.673,
+      "step": 312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.268842228166189,
+      "learning_rate": 1.985130840564332e-05,
+      "loss": 0.6407,
+      "step": 313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.715536790411519,
+      "learning_rate": 1.9849821514625992e-05,
+      "loss": 1.2503,
+      "step": 314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8756806669895925,
+      "learning_rate": 1.9848327282460858e-05,
+      "loss": 1.2489,
+      "step": 315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.729344971393248,
+      "learning_rate": 1.9846825710261585e-05,
+      "loss": 0.7393,
+      "step": 316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7500757608318382,
+      "learning_rate": 1.98453167991473e-05,
+      "loss": 0.6575,
+      "step": 317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.656825190508439,
+      "learning_rate": 1.9843800550242612e-05,
+      "loss": 0.6995,
+      "step": 318
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2688217275918907,
+      "learning_rate": 1.9842276964677593e-05,
+      "loss": 0.7121,
+      "step": 319
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.835903993486976,
+      "learning_rate": 1.984074604358778e-05,
+      "loss": 0.6163,
+      "step": 320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.181388063704869,
+      "learning_rate": 1.983920778811418e-05,
+      "loss": 0.7307,
+      "step": 321
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1917418543921707,
+      "learning_rate": 1.9837662199403266e-05,
+      "loss": 0.6792,
+      "step": 322
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.127493147598884,
+      "learning_rate": 1.9836109278606984e-05,
+      "loss": 1.2115,
+      "step": 323
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1561608545150857,
+      "learning_rate": 1.9834549026882727e-05,
+      "loss": 0.6427,
+      "step": 324
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.821941049009498,
+      "learning_rate": 1.9832981445393372e-05,
+      "loss": 1.3108,
+      "step": 325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8486264404859063,
+      "learning_rate": 1.9831406535307246e-05,
+      "loss": 0.6264,
+      "step": 326
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.574510154889229,
+      "learning_rate": 1.982982429779814e-05,
+      "loss": 0.5852,
+      "step": 327
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7164398078830767,
+      "learning_rate": 1.9828234734045312e-05,
+      "loss": 0.7216,
+      "step": 328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1844875614452723,
+      "learning_rate": 1.982663784523347e-05,
+      "loss": 0.6728,
+      "step": 329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.058262089878628,
+      "learning_rate": 1.982503363255279e-05,
+      "loss": 1.2417,
+      "step": 330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1124435071570837,
+      "learning_rate": 1.982342209719891e-05,
+      "loss": 1.2151,
+      "step": 331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.881228966296008,
+      "learning_rate": 1.9821803240372913e-05,
+      "loss": 1.2396,
+      "step": 332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1774952253682853,
+      "learning_rate": 1.9820177063281348e-05,
+      "loss": 0.5981,
+      "step": 333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.565864690950024,
+      "learning_rate": 1.9818543567136213e-05,
+      "loss": 0.7017,
+      "step": 334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.509389961779903,
+      "learning_rate": 1.9816902753154973e-05,
+      "loss": 0.6055,
+      "step": 335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1034443112184875,
+      "learning_rate": 1.981525462256053e-05,
+      "loss": 1.2484,
+      "step": 336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4722331455939113,
+      "learning_rate": 1.9813599176581257e-05,
+      "loss": 1.2953,
+      "step": 337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5431951633479892,
+      "learning_rate": 1.9811936416450966e-05,
+      "loss": 0.7286,
+      "step": 338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.486443304594545,
+      "learning_rate": 1.981026634340893e-05,
+      "loss": 0.7105,
+      "step": 339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2971421865759956,
+      "learning_rate": 1.9808588958699858e-05,
+      "loss": 0.6536,
+      "step": 340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4178731222086887,
+      "learning_rate": 1.9806904263573925e-05,
+      "loss": 0.6603,
+      "step": 341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1946743788401375,
+      "learning_rate": 1.9805212259286747e-05,
+      "loss": 0.6266,
+      "step": 342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.916535306781055,
+      "learning_rate": 1.980351294709938e-05,
+      "loss": 1.185,
+      "step": 343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.94382065021091,
+      "learning_rate": 1.9801806328278346e-05,
+      "loss": 1.2623,
+      "step": 344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0576317932455184,
+      "learning_rate": 1.9800092404095596e-05,
+      "loss": 0.7184,
+      "step": 345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4287845863923176,
+      "learning_rate": 1.9798371175828526e-05,
+      "loss": 0.7465,
+      "step": 346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.625094016677003,
+      "learning_rate": 1.979664264475999e-05,
+      "loss": 0.7327,
+      "step": 347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7810065675825264,
+      "learning_rate": 1.9794906812178272e-05,
+      "loss": 0.7042,
+      "step": 348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.249352644987833,
+      "learning_rate": 1.9793163679377098e-05,
+      "loss": 0.6487,
+      "step": 349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.178938572784818,
+      "learning_rate": 1.979141324765564e-05,
+      "loss": 0.7902,
+      "step": 350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9900767323045385,
+      "learning_rate": 1.978965551831851e-05,
+      "loss": 0.6412,
+      "step": 351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8514220842982105,
+      "learning_rate": 1.9787890492675755e-05,
+      "loss": 1.2576,
+      "step": 352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8756367490976267,
+      "learning_rate": 1.978611817204286e-05,
+      "loss": 1.2364,
+      "step": 353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.76048505235624,
+      "learning_rate": 1.9784338557740757e-05,
+      "loss": 0.7091,
+      "step": 354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.244260152881739,
+      "learning_rate": 1.97825516510958e-05,
+      "loss": 0.6753,
+      "step": 355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.787778385462312,
+      "learning_rate": 1.9780757453439787e-05,
+      "loss": 0.7098,
+      "step": 356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2270813451269094,
+      "learning_rate": 1.9778955966109945e-05,
+      "loss": 1.2534,
+      "step": 357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7800777632916316,
+      "learning_rate": 1.9777147190448937e-05,
+      "loss": 1.2736,
+      "step": 358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.229617876873017,
+      "learning_rate": 1.977533112780486e-05,
+      "loss": 0.6493,
+      "step": 359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1337607292042486,
+      "learning_rate": 1.9773507779531237e-05,
+      "loss": 0.6667,
+      "step": 360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4417542039695577,
+      "learning_rate": 1.9771677146987023e-05,
+      "loss": 0.7329,
+      "step": 361
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3478324970727242,
+      "learning_rate": 1.9769839231536604e-05,
+      "loss": 0.6503,
+      "step": 362
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1411219620856303,
+      "learning_rate": 1.976799403454979e-05,
+      "loss": 0.5762,
+      "step": 363
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.9254555894688155,
+      "learning_rate": 1.9766141557401824e-05,
+      "loss": 1.2482,
+      "step": 364
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.187527689580912,
+      "learning_rate": 1.976428180147337e-05,
+      "loss": 0.652,
+      "step": 365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1257691227790123,
+      "learning_rate": 1.976241476815052e-05,
+      "loss": 0.6816,
+      "step": 366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3616416161477796,
+      "learning_rate": 1.9760540458824786e-05,
+      "loss": 1.2294,
+      "step": 367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.10482168289612,
+      "learning_rate": 1.9758658874893104e-05,
+      "loss": 0.7324,
+      "step": 368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.215501523277715,
+      "learning_rate": 1.9756770017757837e-05,
+      "loss": 0.6565,
+      "step": 369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.206366830461685,
+      "learning_rate": 1.975487388882676e-05,
+      "loss": 0.7189,
+      "step": 370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4447495815656968,
+      "learning_rate": 1.975297048951308e-05,
+      "loss": 0.7027,
+      "step": 371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1319037151983764,
+      "learning_rate": 1.9751059821235406e-05,
+      "loss": 1.1568,
+      "step": 372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4043815727916757,
+      "learning_rate": 1.974914188541778e-05,
+      "loss": 0.7768,
+      "step": 373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8749935146306491,
+      "learning_rate": 1.9747216683489655e-05,
+      "loss": 0.6318,
+      "step": 374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.108048541220435,
+      "learning_rate": 1.974528421688589e-05,
+      "loss": 0.7198,
+      "step": 375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1367967793638125,
+      "learning_rate": 1.974334448704678e-05,
+      "loss": 0.5665,
+      "step": 376
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0211383765632167,
+      "learning_rate": 1.974139749541801e-05,
+      "loss": 0.6768,
+      "step": 377
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7673959840475923,
+      "learning_rate": 1.9739443243450693e-05,
+      "loss": 1.182,
+      "step": 378
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.108744811868046,
+      "learning_rate": 1.9737481732601347e-05,
+      "loss": 0.6343,
+      "step": 379
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0486477595627766,
+      "learning_rate": 1.9735512964331903e-05,
+      "loss": 0.7387,
+      "step": 380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0750860227885277,
+      "learning_rate": 1.9733536940109693e-05,
+      "loss": 1.2112,
+      "step": 381
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.089080877158942,
+      "learning_rate": 1.973155366140747e-05,
+      "loss": 0.6545,
+      "step": 382
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.629420620367027,
+      "learning_rate": 1.9729563129703384e-05,
+      "loss": 1.1572,
+      "step": 383
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.061447534595425,
+      "learning_rate": 1.9727565346480994e-05,
+      "loss": 0.7121,
+      "step": 384
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.150597503076926,
+      "learning_rate": 1.972556031322926e-05,
+      "loss": 0.6766,
+      "step": 385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1796437593166456,
+      "learning_rate": 1.9723548031442557e-05,
+      "loss": 0.6915,
+      "step": 386
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.083374655836838,
+      "learning_rate": 1.9721528502620648e-05,
+      "loss": 0.7152,
+      "step": 387
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.057622256216699,
+      "learning_rate": 1.97195017282687e-05,
+      "loss": 0.6453,
+      "step": 388
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9458348555877742,
+      "learning_rate": 1.9717467709897292e-05,
+      "loss": 0.7072,
+      "step": 389
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.015412698978649,
+      "learning_rate": 1.971542644902239e-05,
+      "loss": 0.708,
+      "step": 390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0280258950036627,
+      "learning_rate": 1.9713377947165354e-05,
+      "loss": 0.6832,
+      "step": 391
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1491694164268,
+      "learning_rate": 1.971132220585296e-05,
+      "loss": 0.6935,
+      "step": 392
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.698038323269799,
+      "learning_rate": 1.9709259226617362e-05,
+      "loss": 1.2015,
+      "step": 393
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0895233798248105,
+      "learning_rate": 1.9707189010996117e-05,
+      "loss": 0.6507,
+      "step": 394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2498340632496285,
+      "learning_rate": 1.970511156053217e-05,
+      "loss": 0.7153,
+      "step": 395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2095769968488885,
+      "learning_rate": 1.970302687677386e-05,
+      "loss": 0.7262,
+      "step": 396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.825717478173587,
+      "learning_rate": 1.970093496127492e-05,
+      "loss": 1.2229,
+      "step": 397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.122888403460392,
+      "learning_rate": 1.9698835815594467e-05,
+      "loss": 0.6662,
+      "step": 398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1465230530353145,
+      "learning_rate": 1.9696729441297017e-05,
+      "loss": 0.6701,
+      "step": 399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0405005945608314,
+      "learning_rate": 1.9694615839952464e-05,
+      "loss": 0.6192,
+      "step": 400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0595173470453756,
+      "learning_rate": 1.969249501313609e-05,
+      "loss": 0.6858,
+      "step": 401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.025218144940984,
+      "learning_rate": 1.9690366962428563e-05,
+      "loss": 0.7164,
+      "step": 402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9876692815235164,
+      "learning_rate": 1.968823168941594e-05,
+      "loss": 0.7241,
+      "step": 403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9391081876740204,
+      "learning_rate": 1.9686089195689654e-05,
+      "loss": 0.6455,
+      "step": 404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8858878262361143,
+      "learning_rate": 1.9683939482846517e-05,
+      "loss": 0.6257,
+      "step": 405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0040066012627595,
+      "learning_rate": 1.968178255248873e-05,
+      "loss": 0.654,
+      "step": 406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7542548027345592,
+      "learning_rate": 1.9679618406223876e-05,
+      "loss": 1.1785,
+      "step": 407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0983927508688747,
+      "learning_rate": 1.9677447045664906e-05,
+      "loss": 0.6834,
+      "step": 408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.018077915279923,
+      "learning_rate": 1.967526847243015e-05,
+      "loss": 0.6718,
+      "step": 409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5833649661276703,
+      "learning_rate": 1.967308268814331e-05,
+      "loss": 1.1388,
+      "step": 410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.193612554776183,
+      "learning_rate": 1.967088969443348e-05,
+      "loss": 0.6578,
+      "step": 411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1709946424799504,
+      "learning_rate": 1.9668689492935103e-05,
+      "loss": 0.6287,
+      "step": 412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8875120024994114,
+      "learning_rate": 1.9666482085288018e-05,
+      "loss": 0.6878,
+      "step": 413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9100784708750018,
+      "learning_rate": 1.966426747313741e-05,
+      "loss": 0.6587,
+      "step": 414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.916292685545292,
+      "learning_rate": 1.966204565813386e-05,
+      "loss": 1.1978,
+      "step": 415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.067712890098875,
+      "learning_rate": 1.9659816641933292e-05,
+      "loss": 0.7069,
+      "step": 416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.894610440566612,
+      "learning_rate": 1.965758042619702e-05,
+      "loss": 0.5764,
+      "step": 417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5527488185399365,
+      "learning_rate": 1.9655337012591703e-05,
+      "loss": 1.3222,
+      "step": 418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1679704482256144,
+      "learning_rate": 1.9653086402789376e-05,
+      "loss": 0.7197,
+      "step": 419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5684565270475934,
+      "learning_rate": 1.965082859846744e-05,
+      "loss": 0.6641,
+      "step": 420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3428757398492546,
+      "learning_rate": 1.9648563601308654e-05,
+      "loss": 0.7002,
+      "step": 421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.001760376999293,
+      "learning_rate": 1.9646291413001136e-05,
+      "loss": 0.683,
+      "step": 422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9458859161564048,
+      "learning_rate": 1.9644012035238363e-05,
+      "loss": 0.6653,
+      "step": 423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.044947114702168,
+      "learning_rate": 1.9641725469719177e-05,
+      "loss": 0.6912,
+      "step": 424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9018811763670114,
+      "learning_rate": 1.963943171814777e-05,
+      "loss": 0.5959,
+      "step": 425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0993884494817387,
+      "learning_rate": 1.9637130782233694e-05,
+      "loss": 0.6293,
+      "step": 426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.033178824145469,
+      "learning_rate": 1.9634822663691853e-05,
+      "loss": 1.2368,
+      "step": 427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.091635659299788,
+      "learning_rate": 1.9632507364242505e-05,
+      "loss": 0.634,
+      "step": 428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4547067846914223,
+      "learning_rate": 1.9630184885611266e-05,
+      "loss": 1.217,
+      "step": 429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1251027306775123,
+      "learning_rate": 1.9627855229529092e-05,
+      "loss": 0.6397,
+      "step": 430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.994465768399648,
+      "learning_rate": 1.9625518397732296e-05,
+      "loss": 0.6532,
+      "step": 431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3031138181315907,
+      "learning_rate": 1.9623174391962536e-05,
+      "loss": 0.7007,
+      "step": 432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0028457052386477,
+      "learning_rate": 1.962082321396681e-05,
+      "loss": 0.6315,
+      "step": 433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.138876984418124,
+      "learning_rate": 1.9618464865497483e-05,
+      "loss": 0.6724,
+      "step": 434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9410692124158262,
+      "learning_rate": 1.9616099348312244e-05,
+      "loss": 0.655,
+      "step": 435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.205160910803586,
+      "learning_rate": 1.9613726664174126e-05,
+      "loss": 0.7434,
+      "step": 436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9671785647563802,
+      "learning_rate": 1.9611346814851515e-05,
+      "loss": 0.6605,
+      "step": 437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0859646303942294,
+      "learning_rate": 1.960895980211813e-05,
+      "loss": 0.6144,
+      "step": 438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9817341090466987,
+      "learning_rate": 1.9606565627753028e-05,
+      "loss": 0.6595,
+      "step": 439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9441700826954547,
+      "learning_rate": 1.960416429354061e-05,
+      "loss": 1.148,
+      "step": 440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7939341316275925,
+      "learning_rate": 1.9601755801270603e-05,
+      "loss": 0.6524,
+      "step": 441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.649622698064225,
+      "learning_rate": 1.959934015273808e-05,
+      "loss": 1.1848,
+      "step": 442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.925283284608547,
+      "learning_rate": 1.9596917349743436e-05,
+      "loss": 0.6398,
+      "step": 443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0010256211681625,
+      "learning_rate": 1.959448739409241e-05,
+      "loss": 0.6353,
+      "step": 444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.200029223290801,
+      "learning_rate": 1.9592050287596073e-05,
+      "loss": 0.6186,
+      "step": 445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.319317301678621,
+      "learning_rate": 1.95896060320708e-05,
+      "loss": 1.2893,
+      "step": 446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8665393539242374,
+      "learning_rate": 1.9587154629338333e-05,
+      "loss": 0.6589,
+      "step": 447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3004644297224552,
+      "learning_rate": 1.958469608122571e-05,
+      "loss": 1.315,
+      "step": 448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8951530360951447,
+      "learning_rate": 1.9582230389565312e-05,
+      "loss": 0.6512,
+      "step": 449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.130030275971379,
+      "learning_rate": 1.9579757556194834e-05,
+      "loss": 1.1867,
+      "step": 450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0413211880054085,
+      "learning_rate": 1.9577277582957297e-05,
+      "loss": 0.63,
+      "step": 451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.041615260487738,
+      "learning_rate": 1.957479047170105e-05,
+      "loss": 0.6502,
+      "step": 452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7225642405689428,
+      "learning_rate": 1.957229622427975e-05,
+      "loss": 1.2234,
+      "step": 453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1589137933334515,
+      "learning_rate": 1.956979484255238e-05,
+      "loss": 0.686,
+      "step": 454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8264853506844143,
+      "learning_rate": 1.9567286328383242e-05,
+      "loss": 0.6277,
+      "step": 455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2404083020995875,
+      "learning_rate": 1.956477068364195e-05,
+      "loss": 1.2675,
+      "step": 456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.08334172841068,
+      "learning_rate": 1.956224791020343e-05,
+      "loss": 0.6535,
+      "step": 457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8832525918932286,
+      "learning_rate": 1.955971800994793e-05,
+      "loss": 0.5945,
+      "step": 458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7144869380726067,
+      "learning_rate": 1.9557180984761e-05,
+      "loss": 1.1588,
+      "step": 459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1231470136238983,
+      "learning_rate": 1.9554636836533506e-05,
+      "loss": 0.7288,
+      "step": 460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1306943752129284,
+      "learning_rate": 1.9552085567161618e-05,
+      "loss": 0.732,
+      "step": 461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9430000253813267,
+      "learning_rate": 1.954952717854682e-05,
+      "loss": 0.6282,
+      "step": 462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9219361044658392,
+      "learning_rate": 1.95469616725959e-05,
+      "loss": 0.6168,
+      "step": 463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7694983861797522,
+      "learning_rate": 1.954438905122095e-05,
+      "loss": 0.6782,
+      "step": 464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.851800509654484,
+      "learning_rate": 1.9541809316339356e-05,
+      "loss": 0.7074,
+      "step": 465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.01666771902627,
+      "learning_rate": 1.9539222469873826e-05,
+      "loss": 0.6503,
+      "step": 466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.023243088551264,
+      "learning_rate": 1.9536628513752346e-05,
+      "loss": 0.7474,
+      "step": 467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.878920235293422,
+      "learning_rate": 1.9534027449908218e-05,
+      "loss": 1.1971,
+      "step": 468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7976128614739566,
+      "learning_rate": 1.9531419280280034e-05,
+      "loss": 1.2176,
+      "step": 469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1102822978017812,
+      "learning_rate": 1.9528804006811682e-05,
+      "loss": 0.6935,
+      "step": 470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.713577278039233,
+      "learning_rate": 1.9526181631452346e-05,
+      "loss": 1.2323,
+      "step": 471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.868969983468065,
+      "learning_rate": 1.9523552156156503e-05,
+      "loss": 1.2219,
+      "step": 472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6688105187211355,
+      "learning_rate": 1.952091558288392e-05,
+      "loss": 1.2236,
+      "step": 473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.039745254612324,
+      "learning_rate": 1.9518271913599657e-05,
+      "loss": 0.7126,
+      "step": 474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.908064988511034,
+      "learning_rate": 1.9515621150274054e-05,
+      "loss": 0.7004,
+      "step": 475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7855965012774995,
+      "learning_rate": 1.951296329488276e-05,
+      "loss": 1.2177,
+      "step": 476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8037721470640817,
+      "learning_rate": 1.951029834940668e-05,
+      "loss": 0.611,
+      "step": 477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2649054973919496,
+      "learning_rate": 1.9507626315832024e-05,
+      "loss": 0.5837,
+      "step": 478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.864073653825091,
+      "learning_rate": 1.9504947196150285e-05,
+      "loss": 1.2374,
+      "step": 479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0160916660967882,
+      "learning_rate": 1.9502260992358223e-05,
+      "loss": 0.6564,
+      "step": 480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9372439218993895,
+      "learning_rate": 1.949956770645789e-05,
+      "loss": 0.6349,
+      "step": 481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5276106604302537,
+      "learning_rate": 1.9496867340456612e-05,
+      "loss": 1.1938,
+      "step": 482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.311350210741097,
+      "learning_rate": 1.949415989636699e-05,
+      "loss": 0.7305,
+      "step": 483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0409466891762094,
+      "learning_rate": 1.9491445376206913e-05,
+      "loss": 0.6029,
+      "step": 484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1536669439655767,
+      "learning_rate": 1.948872378199952e-05,
+      "loss": 0.7142,
+      "step": 485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9652998174261256,
+      "learning_rate": 1.9485995115773245e-05,
+      "loss": 0.7397,
+      "step": 486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.677517573036276,
+      "learning_rate": 1.9483259379561784e-05,
+      "loss": 1.1619,
+      "step": 487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0667795235866784,
+      "learning_rate": 1.94805165754041e-05,
+      "loss": 0.666,
+      "step": 488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1036498175777973,
+      "learning_rate": 1.9477766705344422e-05,
+      "loss": 0.6313,
+      "step": 489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7158991097601333,
+      "learning_rate": 1.9475009771432255e-05,
+      "loss": 1.136,
+      "step": 490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1377113322563663,
+      "learning_rate": 1.947224577572236e-05,
+      "loss": 0.6163,
+      "step": 491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9950680832996472,
+      "learning_rate": 1.946947472027477e-05,
+      "loss": 0.6399,
+      "step": 492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5599235390139716,
+      "learning_rate": 1.946669660715477e-05,
+      "loss": 1.2204,
+      "step": 493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0213449942914523,
+      "learning_rate": 1.9463911438432902e-05,
+      "loss": 0.6903,
+      "step": 494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6857086224679976,
+      "learning_rate": 1.9461119216184988e-05,
+      "loss": 1.2912,
+      "step": 495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1095081040348953,
+      "learning_rate": 1.9458319942492078e-05,
+      "loss": 0.6399,
+      "step": 496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.83848150248232,
+      "learning_rate": 1.9455513619440506e-05,
+      "loss": 0.6515,
+      "step": 497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.192983760478116,
+      "learning_rate": 1.9452700249121833e-05,
+      "loss": 0.6575,
+      "step": 498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.207382265982575,
+      "learning_rate": 1.9449879833632897e-05,
+      "loss": 1.2392,
+      "step": 499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7964070367292722,
+      "learning_rate": 1.944705237507577e-05,
+      "loss": 0.6474,
+      "step": 500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0050551213349843,
+      "learning_rate": 1.9444217875557777e-05,
+      "loss": 0.6841,
+      "step": 501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9459464894123344,
+      "learning_rate": 1.9441376337191498e-05,
+      "loss": 1.1691,
+      "step": 502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8933630534508665,
+      "learning_rate": 1.9438527762094747e-05,
+      "loss": 0.6528,
+      "step": 503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0800574807335335,
+      "learning_rate": 1.9435672152390593e-05,
+      "loss": 0.7041,
+      "step": 504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9915260322492405,
+      "learning_rate": 1.9432809510207344e-05,
+      "loss": 0.6724,
+      "step": 505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9490876275248794,
+      "learning_rate": 1.9429939837678544e-05,
+      "loss": 0.6487,
+      "step": 506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8033015495701576,
+      "learning_rate": 1.942706313694299e-05,
+      "loss": 1.2014,
+      "step": 507
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0637143758746217,
+      "learning_rate": 1.9424179410144702e-05,
+      "loss": 0.7069,
+      "step": 508
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8703915057483456,
+      "learning_rate": 1.942128865943295e-05,
+      "loss": 0.6883,
+      "step": 509
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8598843503782678,
+      "learning_rate": 1.941839088696223e-05,
+      "loss": 0.7022,
+      "step": 510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.73865494084379,
+      "learning_rate": 1.9415486094892273e-05,
+      "loss": 0.6214,
+      "step": 511
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1509792756566735,
+      "learning_rate": 1.9412574285388043e-05,
+      "loss": 0.7163,
+      "step": 512
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.059814012239148,
+      "learning_rate": 1.9409655460619734e-05,
+      "loss": 1.1975,
+      "step": 513
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9152476891609755,
+      "learning_rate": 1.940672962276277e-05,
+      "loss": 0.6884,
+      "step": 514
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9481110936156611,
+      "learning_rate": 1.94037967739978e-05,
+      "loss": 0.6403,
+      "step": 515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.029669531717513,
+      "learning_rate": 1.9400856916510698e-05,
+      "loss": 0.7177,
+      "step": 516
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9694369596889705,
+      "learning_rate": 1.9397910052492566e-05,
+      "loss": 0.6533,
+      "step": 517
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0653472446870444,
+      "learning_rate": 1.939495618413972e-05,
+      "loss": 0.6621,
+      "step": 518
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.989963369598015,
+      "learning_rate": 1.93919953136537e-05,
+      "loss": 0.6333,
+      "step": 519
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.216127473713011,
+      "learning_rate": 1.938902744324127e-05,
+      "loss": 0.7,
+      "step": 520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8819443595826812,
+      "learning_rate": 1.9386052575114408e-05,
+      "loss": 1.2238,
+      "step": 521
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1311411269509066,
+      "learning_rate": 1.9383070711490293e-05,
+      "loss": 0.7255,
+      "step": 522
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4209533622443278,
+      "learning_rate": 1.938008185459135e-05,
+      "loss": 0.7234,
+      "step": 523
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2857788974791418,
+      "learning_rate": 1.9377086006645186e-05,
+      "loss": 0.7207,
+      "step": 524
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3845929058850546,
+      "learning_rate": 1.937408316988463e-05,
+      "loss": 0.7134,
+      "step": 525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8870517148396622,
+      "learning_rate": 1.937107334654772e-05,
+      "loss": 0.7008,
+      "step": 526
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.804655577688891,
+      "learning_rate": 1.93680565388777e-05,
+      "loss": 1.2047,
+      "step": 527
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8532655809362506,
+      "learning_rate": 1.936503274912302e-05,
+      "loss": 1.2595,
+      "step": 528
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0574789171691963,
+      "learning_rate": 1.9362001979537338e-05,
+      "loss": 0.6499,
+      "step": 529
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.186081985697203,
+      "learning_rate": 1.935896423237951e-05,
+      "loss": 1.279,
+      "step": 530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.095509745558331,
+      "learning_rate": 1.935591950991358e-05,
+      "loss": 0.6155,
+      "step": 531
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0420953095648353,
+      "learning_rate": 1.9352867814408818e-05,
+      "loss": 0.5815,
+      "step": 532
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8625072731682644,
+      "learning_rate": 1.934980914813967e-05,
+      "loss": 0.7373,
+      "step": 533
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.97802269595636,
+      "learning_rate": 1.934674351338578e-05,
+      "loss": 1.1866,
+      "step": 534
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9693280310946042,
+      "learning_rate": 1.9343670912432e-05,
+      "loss": 0.64,
+      "step": 535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0945409305069047,
+      "learning_rate": 1.934059134756835e-05,
+      "loss": 0.5874,
+      "step": 536
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9273671452533003,
+      "learning_rate": 1.933750482109006e-05,
+      "loss": 1.2762,
+      "step": 537
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1035186546317712,
+      "learning_rate": 1.9334411335297544e-05,
+      "loss": 0.6803,
+      "step": 538
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1212914763419435,
+      "learning_rate": 1.9331310892496397e-05,
+      "loss": 0.608,
+      "step": 539
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.706989642076053,
+      "learning_rate": 1.9328203494997405e-05,
+      "loss": 1.2353,
+      "step": 540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9547056302026546,
+      "learning_rate": 1.9325089145116536e-05,
+      "loss": 0.5953,
+      "step": 541
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1843380677463515,
+      "learning_rate": 1.9321967845174935e-05,
+      "loss": 0.6679,
+      "step": 542
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3760511519661103,
+      "learning_rate": 1.9318839597498934e-05,
+      "loss": 0.6401,
+      "step": 543
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1404195708699425,
+      "learning_rate": 1.9315704404420038e-05,
+      "loss": 0.7336,
+      "step": 544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9735127995590795,
+      "learning_rate": 1.9312562268274933e-05,
+      "loss": 0.6574,
+      "step": 545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8274812192699827,
+      "learning_rate": 1.9309413191405476e-05,
+      "loss": 1.2175,
+      "step": 546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.973341968485768,
+      "learning_rate": 1.93062571761587e-05,
+      "loss": 0.6658,
+      "step": 547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.894518794662943,
+      "learning_rate": 1.9303094224886805e-05,
+      "loss": 0.68,
+      "step": 548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8570387655002212,
+      "learning_rate": 1.9299924339947163e-05,
+      "loss": 0.6572,
+      "step": 549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2646241101259412,
+      "learning_rate": 1.9296747523702314e-05,
+      "loss": 0.695,
+      "step": 550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5462890065963184,
+      "learning_rate": 1.9293563778519962e-05,
+      "loss": 1.2989,
+      "step": 551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8310478927534684,
+      "learning_rate": 1.9290373106772982e-05,
+      "loss": 1.2571,
+      "step": 552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.160838213489556,
+      "learning_rate": 1.9287175510839403e-05,
+      "loss": 0.6741,
+      "step": 553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.095426749577576,
+      "learning_rate": 1.928397099310242e-05,
+      "loss": 0.676,
+      "step": 554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2335570631741337,
+      "learning_rate": 1.9280759555950378e-05,
+      "loss": 0.6939,
+      "step": 555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9679772685894368,
+      "learning_rate": 1.9277541201776792e-05,
+      "loss": 0.5742,
+      "step": 556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9174437074535453,
+      "learning_rate": 1.9274315932980324e-05,
+      "loss": 0.6718,
+      "step": 557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.264455670236973,
+      "learning_rate": 1.9271083751964794e-05,
+      "loss": 0.6364,
+      "step": 558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.639852005470425,
+      "learning_rate": 1.926784466113917e-05,
+      "loss": 1.2111,
+      "step": 559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8302757328569292,
+      "learning_rate": 1.926459866291757e-05,
+      "loss": 1.2855,
+      "step": 560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2259104432047585,
+      "learning_rate": 1.926134575971926e-05,
+      "loss": 0.6848,
+      "step": 561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6152966603351633,
+      "learning_rate": 1.925808595396866e-05,
+      "loss": 1.1797,
+      "step": 562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1172072220183042,
+      "learning_rate": 1.9254819248095323e-05,
+      "loss": 0.6195,
+      "step": 563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.33393003222303,
+      "learning_rate": 1.9251545644533953e-05,
+      "loss": 0.6926,
+      "step": 564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9563893035610516,
+      "learning_rate": 1.924826514572439e-05,
+      "loss": 1.2215,
+      "step": 565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.339900932116602,
+      "learning_rate": 1.9244977754111613e-05,
+      "loss": 1.2396,
+      "step": 566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0651081889771286,
+      "learning_rate": 1.9241683472145746e-05,
+      "loss": 0.6824,
+      "step": 567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9708605279928744,
+      "learning_rate": 1.9238382302282036e-05,
+      "loss": 0.6392,
+      "step": 568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.235753201096843,
+      "learning_rate": 1.923507424698088e-05,
+      "loss": 0.7309,
+      "step": 569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0197153991587506,
+      "learning_rate": 1.9231759308707784e-05,
+      "loss": 0.6932,
+      "step": 570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3508758949899677,
+      "learning_rate": 1.922843748993341e-05,
+      "loss": 1.1676,
+      "step": 571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8739171719580598,
+      "learning_rate": 1.9225108793133525e-05,
+      "loss": 0.7045,
+      "step": 572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.049870693520711,
+      "learning_rate": 1.9221773220789037e-05,
+      "loss": 0.7213,
+      "step": 573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9382101291570715,
+      "learning_rate": 1.9218430775385975e-05,
+      "loss": 0.6993,
+      "step": 574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.778764473812285,
+      "learning_rate": 1.9215081459415485e-05,
+      "loss": 1.197,
+      "step": 575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.615216629777846,
+      "learning_rate": 1.9211725275373846e-05,
+      "loss": 1.1948,
+      "step": 576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.600104250432078,
+      "learning_rate": 1.9208362225762436e-05,
+      "loss": 1.2549,
+      "step": 577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7688361372147683,
+      "learning_rate": 1.920499231308777e-05,
+      "loss": 1.2397,
+      "step": 578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.240866995031444,
+      "learning_rate": 1.9201615539861467e-05,
+      "loss": 0.6727,
+      "step": 579
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0963782758876657,
+      "learning_rate": 1.9198231908600266e-05,
+      "loss": 0.7054,
+      "step": 580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8278286219887474,
+      "learning_rate": 1.9194841421826006e-05,
+      "loss": 1.2934,
+      "step": 581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1349100730089203,
+      "learning_rate": 1.919144408206565e-05,
+      "loss": 0.7202,
+      "step": 582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9129768302186048,
+      "learning_rate": 1.918803989185126e-05,
+      "loss": 0.6185,
+      "step": 583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0729584943765618,
+      "learning_rate": 1.918462885372e-05,
+      "loss": 0.636,
+      "step": 584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0913604213633437,
+      "learning_rate": 1.918121097021415e-05,
+      "loss": 0.7018,
+      "step": 585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.212673988345927,
+      "learning_rate": 1.917778624388108e-05,
+      "loss": 1.1505,
+      "step": 586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8871175633527069,
+      "learning_rate": 1.917435467727327e-05,
+      "loss": 0.6027,
+      "step": 587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8743465320201798,
+      "learning_rate": 1.9170916272948287e-05,
+      "loss": 0.6907,
+      "step": 588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8360244011580673,
+      "learning_rate": 1.9167471033468802e-05,
+      "loss": 0.6662,
+      "step": 589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7136132826297445,
+      "learning_rate": 1.9164018961402582e-05,
+      "loss": 1.188,
+      "step": 590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8556108270412877,
+      "learning_rate": 1.9160560059322476e-05,
+      "loss": 0.6602,
+      "step": 591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8413587236183688,
+      "learning_rate": 1.915709432980644e-05,
+      "loss": 0.7225,
+      "step": 592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.637512181325404,
+      "learning_rate": 1.91536217754375e-05,
+      "loss": 1.23,
+      "step": 593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8601725685263515,
+      "learning_rate": 1.9150142398803782e-05,
+      "loss": 0.6562,
+      "step": 594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7976978116432663,
+      "learning_rate": 1.9146656202498493e-05,
+      "loss": 1.2422,
+      "step": 595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8930604477598327,
+      "learning_rate": 1.9143163189119918e-05,
+      "loss": 0.5996,
+      "step": 596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.981246852579915,
+      "learning_rate": 1.913966336127143e-05,
+      "loss": 0.6713,
+      "step": 597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.027439751200127,
+      "learning_rate": 1.9136156721561477e-05,
+      "loss": 0.69,
+      "step": 598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7954174701902212,
+      "learning_rate": 1.9132643272603583e-05,
+      "loss": 0.6316,
+      "step": 599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8915415432938985,
+      "learning_rate": 1.912912301701635e-05,
+      "loss": 0.6211,
+      "step": 600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1789381859511803,
+      "learning_rate": 1.9125595957423453e-05,
+      "loss": 0.6669,
+      "step": 601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0589039961482456,
+      "learning_rate": 1.9122062096453634e-05,
+      "loss": 1.2468,
+      "step": 602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8039203264282446,
+      "learning_rate": 1.9118521436740705e-05,
+      "loss": 0.6412,
+      "step": 603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.001587779810438,
+      "learning_rate": 1.9114973980923552e-05,
+      "loss": 0.77,
+      "step": 604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8290857720077316,
+      "learning_rate": 1.9111419731646118e-05,
+      "loss": 0.6209,
+      "step": 605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2915632917456894,
+      "learning_rate": 1.9107858691557416e-05,
+      "loss": 0.6766,
+      "step": 606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.33655945625957,
+      "learning_rate": 1.910429086331151e-05,
+      "loss": 0.7004,
+      "step": 607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.989823168102007,
+      "learning_rate": 1.9100716249567537e-05,
+      "loss": 0.6497,
+      "step": 608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8509597679800056,
+      "learning_rate": 1.909713485298968e-05,
+      "loss": 0.6321,
+      "step": 609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0241658789320147,
+      "learning_rate": 1.909354667624718e-05,
+      "loss": 1.2423,
+      "step": 610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.680191298989103,
+      "learning_rate": 1.9089951722014338e-05,
+      "loss": 1.2496,
+      "step": 611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1611556944089947,
+      "learning_rate": 1.9086349992970495e-05,
+      "loss": 0.6981,
+      "step": 612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.311232364801531,
+      "learning_rate": 1.9082741491800053e-05,
+      "loss": 0.6649,
+      "step": 613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.560349502040165,
+      "learning_rate": 1.9079126221192452e-05,
+      "loss": 1.2111,
+      "step": 614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.165142750447097,
+      "learning_rate": 1.907550418384218e-05,
+      "loss": 1.168,
+      "step": 615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.153999712108513,
+      "learning_rate": 1.9071875382448777e-05,
+      "loss": 0.6206,
+      "step": 616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.294867162803576,
+      "learning_rate": 1.9068239819716807e-05,
+      "loss": 0.6285,
+      "step": 617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4259902906537167,
+      "learning_rate": 1.906459749835589e-05,
+      "loss": 1.218,
+      "step": 618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1954568563107686,
+      "learning_rate": 1.9060948421080668e-05,
+      "loss": 0.7145,
+      "step": 619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.187457784111085,
+      "learning_rate": 1.9057292590610835e-05,
+      "loss": 0.675,
+      "step": 620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0562880899731906,
+      "learning_rate": 1.9053630009671106e-05,
+      "loss": 0.5803,
+      "step": 621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0965645785178655,
+      "learning_rate": 1.904996068099123e-05,
+      "loss": 0.5778,
+      "step": 622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0613910685833985,
+      "learning_rate": 1.9046284607305988e-05,
+      "loss": 0.6873,
+      "step": 623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0563862298971998,
+      "learning_rate": 1.9042601791355184e-05,
+      "loss": 0.6786,
+      "step": 624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1110130343255067,
+      "learning_rate": 1.903891223588365e-05,
+      "loss": 1.2401,
+      "step": 625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6893272969369337,
+      "learning_rate": 1.9035215943641237e-05,
+      "loss": 1.2171,
+      "step": 626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0804480286207694,
+      "learning_rate": 1.9031512917382827e-05,
+      "loss": 0.6134,
+      "step": 627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.818758141448102,
+      "learning_rate": 1.902780315986831e-05,
+      "loss": 1.1499,
+      "step": 628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.502374777907341,
+      "learning_rate": 1.9024086673862595e-05,
+      "loss": 0.726,
+      "step": 629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.252765559093462,
+      "learning_rate": 1.902036346213561e-05,
+      "loss": 0.5967,
+      "step": 630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.342411533633952,
+      "learning_rate": 1.9016633527462294e-05,
+      "loss": 1.1763,
+      "step": 631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.042476623768581,
+      "learning_rate": 1.9012896872622597e-05,
+      "loss": 0.6258,
+      "step": 632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5979764221735038,
+      "learning_rate": 1.900915350040147e-05,
+      "loss": 0.6144,
+      "step": 633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2285957092935997,
+      "learning_rate": 1.9005403413588882e-05,
+      "loss": 1.249,
+      "step": 634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.97693609087175,
+      "learning_rate": 1.90016466149798e-05,
+      "loss": 1.2089,
+      "step": 635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0857643925438865,
+      "learning_rate": 1.8997883107374203e-05,
+      "loss": 0.7395,
+      "step": 636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.058407034001628,
+      "learning_rate": 1.899411289357705e-05,
+      "loss": 0.6426,
+      "step": 637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9640972051729597,
+      "learning_rate": 1.8990335976398317e-05,
+      "loss": 0.7006,
+      "step": 638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0984603025157553,
+      "learning_rate": 1.8986552358652968e-05,
+      "loss": 0.746,
+      "step": 639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0830285487975546,
+      "learning_rate": 1.8982762043160964e-05,
+      "loss": 0.7088,
+      "step": 640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8553865547617725,
+      "learning_rate": 1.8978965032747253e-05,
+      "loss": 0.6243,
+      "step": 641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8107378884072258,
+      "learning_rate": 1.8975161330241776e-05,
+      "loss": 0.6012,
+      "step": 642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0724150019820886,
+      "learning_rate": 1.8971350938479467e-05,
+      "loss": 0.7202,
+      "step": 643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.044723222051349,
+      "learning_rate": 1.8967533860300236e-05,
+      "loss": 0.611,
+      "step": 644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.89832714079423,
+      "learning_rate": 1.8963710098548983e-05,
+      "loss": 0.7283,
+      "step": 645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.001175383440918,
+      "learning_rate": 1.8959879656075584e-05,
+      "loss": 0.6694,
+      "step": 646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9294931108673203,
+      "learning_rate": 1.89560425357349e-05,
+      "loss": 0.6513,
+      "step": 647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9771789849348942,
+      "learning_rate": 1.895219874038677e-05,
+      "loss": 0.6434,
+      "step": 648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.664219285399324,
+      "learning_rate": 1.894834827289599e-05,
+      "loss": 0.6394,
+      "step": 649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.210224988891224,
+      "learning_rate": 1.894449113613236e-05,
+      "loss": 1.1809,
+      "step": 650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9023173421489323,
+      "learning_rate": 1.8940627332970624e-05,
+      "loss": 1.1825,
+      "step": 651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0814740604454696,
+      "learning_rate": 1.8936756866290507e-05,
+      "loss": 0.6408,
+      "step": 652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9653513073415607,
+      "learning_rate": 1.8932879738976695e-05,
+      "loss": 0.6346,
+      "step": 653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.261224872931458,
+      "learning_rate": 1.8928995953918847e-05,
+      "loss": 0.6796,
+      "step": 654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0235278557324077,
+      "learning_rate": 1.8925105514011576e-05,
+      "loss": 0.5885,
+      "step": 655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.154204151079518,
+      "learning_rate": 1.892120842215445e-05,
+      "loss": 0.7238,
+      "step": 656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.971724718415186,
+      "learning_rate": 1.8917304681252008e-05,
+      "loss": 0.7395,
+      "step": 657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.362874022000197,
+      "learning_rate": 1.891339429421374e-05,
+      "loss": 0.6369,
+      "step": 658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9447910846454137,
+      "learning_rate": 1.890947726395408e-05,
+      "loss": 1.2231,
+      "step": 659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9294948378910846,
+      "learning_rate": 1.8905553593392427e-05,
+      "loss": 0.5906,
+      "step": 660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3005112682666646,
+      "learning_rate": 1.890162328545312e-05,
+      "loss": 0.6364,
+      "step": 661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1046708787553956,
+      "learning_rate": 1.889768634306545e-05,
+      "loss": 0.622,
+      "step": 662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4839758663607143,
+      "learning_rate": 1.8893742769163646e-05,
+      "loss": 0.6565,
+      "step": 663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1974289314907938,
+      "learning_rate": 1.888979256668689e-05,
+      "loss": 0.6743,
+      "step": 664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8277251554314549,
+      "learning_rate": 1.8885835738579296e-05,
+      "loss": 0.6674,
+      "step": 665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.983243892590935,
+      "learning_rate": 1.8881872287789916e-05,
+      "loss": 0.6793,
+      "step": 666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.204621432685124,
+      "learning_rate": 1.8877902217272743e-05,
+      "loss": 0.6658,
+      "step": 667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9670357986669702,
+      "learning_rate": 1.8873925529986703e-05,
+      "loss": 0.6238,
+      "step": 668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.295350747263723,
+      "learning_rate": 1.8869942228895647e-05,
+      "loss": 0.6692,
+      "step": 669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1980264121122604,
+      "learning_rate": 1.886595231696836e-05,
+      "loss": 0.6274,
+      "step": 670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.118363511465668,
+      "learning_rate": 1.886195579717856e-05,
+      "loss": 0.6206,
+      "step": 671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9733303792132153,
+      "learning_rate": 1.8857952672504875e-05,
+      "loss": 0.6936,
+      "step": 672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1142800458560487,
+      "learning_rate": 1.885394294593087e-05,
+      "loss": 0.7058,
+      "step": 673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7378267553718203,
+      "learning_rate": 1.884992662044503e-05,
+      "loss": 1.1915,
+      "step": 674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.170030330818317,
+      "learning_rate": 1.8845903699040743e-05,
+      "loss": 0.6523,
+      "step": 675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.406205268731773,
+      "learning_rate": 1.8841874184716333e-05,
+      "loss": 1.1674,
+      "step": 676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.330904890663938,
+      "learning_rate": 1.8837838080475023e-05,
+      "loss": 0.689,
+      "step": 677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9575949755669502,
+      "learning_rate": 1.883379538932495e-05,
+      "loss": 0.6455,
+      "step": 678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4112770972798785,
+      "learning_rate": 1.8829746114279167e-05,
+      "loss": 1.19,
+      "step": 679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.917288052732672,
+      "learning_rate": 1.8825690258355628e-05,
+      "loss": 0.6402,
+      "step": 680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0540480649745922,
+      "learning_rate": 1.8821627824577196e-05,
+      "loss": 0.6038,
+      "step": 681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0935037461573818,
+      "learning_rate": 1.8817558815971632e-05,
+      "loss": 0.7395,
+      "step": 682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0979268026740425,
+      "learning_rate": 1.8813483235571598e-05,
+      "loss": 0.6655,
+      "step": 683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8460256848420724,
+      "learning_rate": 1.880940108641466e-05,
+      "loss": 1.2,
+      "step": 684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7307841975222433,
+      "learning_rate": 1.880531237154327e-05,
+      "loss": 1.2225,
+      "step": 685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9077872925403645,
+      "learning_rate": 1.8801217094004786e-05,
+      "loss": 0.6177,
+      "step": 686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6900144259871497,
+      "learning_rate": 1.879711525685144e-05,
+      "loss": 1.186,
+      "step": 687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.007744273031034,
+      "learning_rate": 1.8793006863140367e-05,
+      "loss": 0.6654,
+      "step": 688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2309131520501837,
+      "learning_rate": 1.878889191593359e-05,
+      "loss": 0.7328,
+      "step": 689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.054823463692116,
+      "learning_rate": 1.8784770418298003e-05,
+      "loss": 1.1995,
+      "step": 690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1770890604320536,
+      "learning_rate": 1.8780642373305397e-05,
+      "loss": 0.6955,
+      "step": 691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9150995844805367,
+      "learning_rate": 1.877650778403243e-05,
+      "loss": 0.6717,
+      "step": 692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1826130193041093,
+      "learning_rate": 1.8772366653560644e-05,
+      "loss": 0.7111,
+      "step": 693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0712115785545864,
+      "learning_rate": 1.8768218984976457e-05,
+      "loss": 0.6587,
+      "step": 694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1846759773007824,
+      "learning_rate": 1.876406478137116e-05,
+      "loss": 0.7116,
+      "step": 695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8845845455981913,
+      "learning_rate": 1.8759904045840905e-05,
+      "loss": 0.6566,
+      "step": 696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1492578867076815,
+      "learning_rate": 1.875573678148673e-05,
+      "loss": 0.6834,
+      "step": 697
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9686582939240287,
+      "learning_rate": 1.8751562991414526e-05,
+      "loss": 0.6992,
+      "step": 698
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9304733919239723,
+      "learning_rate": 1.874738267873505e-05,
+      "loss": 0.6167,
+      "step": 699
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9252396983332665,
+      "learning_rate": 1.8743195846563918e-05,
+      "loss": 0.6501,
+      "step": 700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.16116225195509,
+      "learning_rate": 1.8739002498021617e-05,
+      "loss": 0.7276,
+      "step": 701
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8997109738846731,
+      "learning_rate": 1.873480263623347e-05,
+      "loss": 0.6383,
+      "step": 702
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0244967958304994,
+      "learning_rate": 1.8730596264329678e-05,
+      "loss": 0.6373,
+      "step": 703
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3343158927093564,
+      "learning_rate": 1.8726383385445274e-05,
+      "loss": 0.6796,
+      "step": 704
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4618905838447422,
+      "learning_rate": 1.8722164002720155e-05,
+      "loss": 1.2263,
+      "step": 705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7406158775598481,
+      "learning_rate": 1.8717938119299054e-05,
+      "loss": 0.5799,
+      "step": 706
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.67981813589424,
+      "learning_rate": 1.871370573833156e-05,
+      "loss": 1.2158,
+      "step": 707
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.98369369075784,
+      "learning_rate": 1.8709466862972093e-05,
+      "loss": 0.6764,
+      "step": 708
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.459607740238956,
+      "learning_rate": 1.8705221496379927e-05,
+      "loss": 1.212,
+      "step": 709
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0023188504385345,
+      "learning_rate": 1.870096964171916e-05,
+      "loss": 0.6897,
+      "step": 710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1596209740947834,
+      "learning_rate": 1.8696711302158733e-05,
+      "loss": 0.7247,
+      "step": 711
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9993506024072325,
+      "learning_rate": 1.869244648087242e-05,
+      "loss": 0.6321,
+      "step": 712
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0297423205484213,
+      "learning_rate": 1.8688175181038825e-05,
+      "loss": 1.2655,
+      "step": 713
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.000775549755625,
+      "learning_rate": 1.8683897405841384e-05,
+      "loss": 0.6475,
+      "step": 714
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.582856150100595,
+      "learning_rate": 1.867961315846835e-05,
+      "loss": 1.1727,
+      "step": 715
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0216927188931693,
+      "learning_rate": 1.867532244211281e-05,
+      "loss": 0.6694,
+      "step": 716
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0132362817134255,
+      "learning_rate": 1.8671025259972666e-05,
+      "loss": 0.6866,
+      "step": 717
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1045133450099502,
+      "learning_rate": 1.866672161525064e-05,
+      "loss": 0.642,
+      "step": 718
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.175756529499933,
+      "learning_rate": 1.8662411511154276e-05,
+      "loss": 0.626,
+      "step": 719
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1200744369891367,
+      "learning_rate": 1.865809495089592e-05,
+      "loss": 0.6317,
+      "step": 720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1344746480416137,
+      "learning_rate": 1.865377193769275e-05,
+      "loss": 1.2806,
+      "step": 721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.729714144376913,
+      "learning_rate": 1.8649442474766727e-05,
+      "loss": 1.1733,
+      "step": 722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.60404845662394,
+      "learning_rate": 1.864510656534464e-05,
+      "loss": 1.1818,
+      "step": 723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0361120888632978,
+      "learning_rate": 1.8640764212658075e-05,
+      "loss": 0.6554,
+      "step": 724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.381964543197577,
+      "learning_rate": 1.8636415419943422e-05,
+      "loss": 0.7016,
+      "step": 725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1765696083983452,
+      "learning_rate": 1.8632060190441868e-05,
+      "loss": 0.6796,
+      "step": 726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1300264583260504,
+      "learning_rate": 1.8627698527399402e-05,
+      "loss": 0.6676,
+      "step": 727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0443018331419043,
+      "learning_rate": 1.86233304340668e-05,
+      "loss": 0.6228,
+      "step": 728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.021560183199297,
+      "learning_rate": 1.861895591369964e-05,
+      "loss": 1.182,
+      "step": 729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.982971505986617,
+      "learning_rate": 1.8614574969558284e-05,
+      "loss": 0.6223,
+      "step": 730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0501783651099266,
+      "learning_rate": 1.8610187604907883e-05,
+      "loss": 0.6798,
+      "step": 731
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0131272987607356,
+      "learning_rate": 1.860579382301837e-05,
+      "loss": 0.5975,
+      "step": 732
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8561454902533434,
+      "learning_rate": 1.8601393627164474e-05,
+      "loss": 0.5889,
+      "step": 733
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0543185230875296,
+      "learning_rate": 1.8596987020625685e-05,
+      "loss": 1.2089,
+      "step": 734
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8163071711437944,
+      "learning_rate": 1.859257400668629e-05,
+      "loss": 1.1751,
+      "step": 735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9377907719992158,
+      "learning_rate": 1.858815458863533e-05,
+      "loss": 0.6837,
+      "step": 736
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.462768641072179,
+      "learning_rate": 1.8583728769766645e-05,
+      "loss": 1.1785,
+      "step": 737
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1291126472883235,
+      "learning_rate": 1.8579296553378824e-05,
+      "loss": 0.6919,
+      "step": 738
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9432120790318403,
+      "learning_rate": 1.8574857942775236e-05,
+      "loss": 0.6244,
+      "step": 739
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0481002427081405,
+      "learning_rate": 1.8570412941264004e-05,
+      "loss": 0.6649,
+      "step": 740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9422200743255444,
+      "learning_rate": 1.8565961552158032e-05,
+      "loss": 0.6591,
+      "step": 741
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9881114627639755,
+      "learning_rate": 1.8561503778774967e-05,
+      "loss": 0.6388,
+      "step": 742
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9586870845483026,
+      "learning_rate": 1.8557039624437224e-05,
+      "loss": 0.6187,
+      "step": 743
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0761893015284696,
+      "learning_rate": 1.855256909247197e-05,
+      "loss": 0.6568,
+      "step": 744
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.936834142311765,
+      "learning_rate": 1.8548092186211135e-05,
+      "loss": 0.6967,
+      "step": 745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.116283059567498,
+      "learning_rate": 1.854360890899138e-05,
+      "loss": 0.6887,
+      "step": 746
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.070644026044796,
+      "learning_rate": 1.8539119264154135e-05,
+      "loss": 0.6312,
+      "step": 747
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.021957234015503,
+      "learning_rate": 1.853462325504556e-05,
+      "loss": 0.6973,
+      "step": 748
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0218388721756395,
+      "learning_rate": 1.8530120885016573e-05,
+      "loss": 1.2181,
+      "step": 749
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.027457108662091,
+      "learning_rate": 1.8525612157422815e-05,
+      "loss": 0.6897,
+      "step": 750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0589823818348516,
+      "learning_rate": 1.8521097075624687e-05,
+      "loss": 0.5994,
+      "step": 751
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0045028140167473,
+      "learning_rate": 1.8516575642987308e-05,
+      "loss": 0.6709,
+      "step": 752
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.858896101553582,
+      "learning_rate": 1.8512047862880534e-05,
+      "loss": 0.5947,
+      "step": 753
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.043779463459385,
+      "learning_rate": 1.850751373867896e-05,
+      "loss": 0.7113,
+      "step": 754
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0262739545943353,
+      "learning_rate": 1.8502973273761897e-05,
+      "loss": 0.6551,
+      "step": 755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6800199717105415,
+      "learning_rate": 1.8498426471513396e-05,
+      "loss": 1.2422,
+      "step": 756
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2101674481447775,
+      "learning_rate": 1.849387333532222e-05,
+      "loss": 0.6801,
+      "step": 757
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0883493351632194,
+      "learning_rate": 1.8489313868581857e-05,
+      "loss": 0.6771,
+      "step": 758
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0264454115494406,
+      "learning_rate": 1.848474807469051e-05,
+      "loss": 0.5979,
+      "step": 759
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.950267584959136,
+      "learning_rate": 1.8480175957051104e-05,
+      "loss": 0.7164,
+      "step": 760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.958238519113366,
+      "learning_rate": 1.8475597519071268e-05,
+      "loss": 0.691,
+      "step": 761
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.058120846682873,
+      "learning_rate": 1.8471012764163356e-05,
+      "loss": 0.6592,
+      "step": 762
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.894317867265161,
+      "learning_rate": 1.8466421695744414e-05,
+      "loss": 1.1515,
+      "step": 763
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.483338406984078,
+      "learning_rate": 1.8461824317236203e-05,
+      "loss": 1.1099,
+      "step": 764
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8953102056812232,
+      "learning_rate": 1.8457220632065185e-05,
+      "loss": 0.6025,
+      "step": 765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0917429759389727,
+      "learning_rate": 1.8452610643662523e-05,
+      "loss": 0.667,
+      "step": 766
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0347355049134124,
+      "learning_rate": 1.8447994355464075e-05,
+      "loss": 0.6826,
+      "step": 767
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.838906507155251,
+      "learning_rate": 1.8443371770910395e-05,
+      "loss": 0.7052,
+      "step": 768
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9095886258743964,
+      "learning_rate": 1.8438742893446733e-05,
+      "loss": 0.6991,
+      "step": 769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0311849382362217,
+      "learning_rate": 1.8434107726523028e-05,
+      "loss": 1.2286,
+      "step": 770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8465479670273028,
+      "learning_rate": 1.8429466273593903e-05,
+      "loss": 0.6682,
+      "step": 771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9179598514677971,
+      "learning_rate": 1.8424818538118665e-05,
+      "loss": 0.6786,
+      "step": 772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8410232774028654,
+      "learning_rate": 1.8420164523561316e-05,
+      "loss": 1.2296,
+      "step": 773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0038908548645824,
+      "learning_rate": 1.8415504233390524e-05,
+      "loss": 0.6368,
+      "step": 774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7781664292359403,
+      "learning_rate": 1.841083767107964e-05,
+      "loss": 0.5873,
+      "step": 775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.167922195328591,
+      "learning_rate": 1.840616484010669e-05,
+      "loss": 0.6502,
+      "step": 776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8379377959840093,
+      "learning_rate": 1.8401485743954366e-05,
+      "loss": 0.6251,
+      "step": 777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.078601259023026,
+      "learning_rate": 1.8396800386110038e-05,
+      "loss": 0.6669,
+      "step": 778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7724251186456894,
+      "learning_rate": 1.8392108770065738e-05,
+      "loss": 1.1637,
+      "step": 779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.092945252867908,
+      "learning_rate": 1.838741089931817e-05,
+      "loss": 0.6274,
+      "step": 780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8807106777686224,
+      "learning_rate": 1.838270677736868e-05,
+      "loss": 0.5588,
+      "step": 781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8567491018259852,
+      "learning_rate": 1.83779964077233e-05,
+      "loss": 0.5999,
+      "step": 782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9609833848697777,
+      "learning_rate": 1.8373279793892697e-05,
+      "loss": 0.6628,
+      "step": 783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9857413782168318,
+      "learning_rate": 1.8368556939392202e-05,
+      "loss": 0.6088,
+      "step": 784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9809665169347828,
+      "learning_rate": 1.8363827847741794e-05,
+      "loss": 0.6519,
+      "step": 785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.718880968780902,
+      "learning_rate": 1.83590925224661e-05,
+      "loss": 1.1305,
+      "step": 786
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9367529285527594,
+      "learning_rate": 1.8354350967094397e-05,
+      "loss": 0.6948,
+      "step": 787
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6022876144711717,
+      "learning_rate": 1.8349603185160606e-05,
+      "loss": 1.1896,
+      "step": 788
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9840077238678295,
+      "learning_rate": 1.8344849180203276e-05,
+      "loss": 0.6741,
+      "step": 789
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0103570071140595,
+      "learning_rate": 1.834008895576561e-05,
+      "loss": 0.6925,
+      "step": 790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8904140373751885,
+      "learning_rate": 1.8335322515395436e-05,
+      "loss": 0.6454,
+      "step": 791
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8683773583600187,
+      "learning_rate": 1.8330549862645222e-05,
+      "loss": 0.6838,
+      "step": 792
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9855882845423123,
+      "learning_rate": 1.8325771001072067e-05,
+      "loss": 0.6593,
+      "step": 793
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8243990057251345,
+      "learning_rate": 1.8320985934237685e-05,
+      "loss": 0.6323,
+      "step": 794
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8907290795422513,
+      "learning_rate": 1.8316194665708428e-05,
+      "loss": 0.735,
+      "step": 795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5840867412231354,
+      "learning_rate": 1.8311397199055264e-05,
+      "loss": 1.1855,
+      "step": 796
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7699297307095805,
+      "learning_rate": 1.8306593537853778e-05,
+      "loss": 0.6407,
+      "step": 797
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.983665859053877,
+      "learning_rate": 1.8301783685684184e-05,
+      "loss": 0.7068,
+      "step": 798
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.703443309032314,
+      "learning_rate": 1.8296967646131295e-05,
+      "loss": 1.2181,
+      "step": 799
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9240221520592944,
+      "learning_rate": 1.8292145422784543e-05,
+      "loss": 0.6554,
+      "step": 800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0675363763306174,
+      "learning_rate": 1.8287317019237975e-05,
+      "loss": 0.7264,
+      "step": 801
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.103845842200734,
+      "learning_rate": 1.828248243909023e-05,
+      "loss": 0.7064,
+      "step": 802
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.11567518434236,
+      "learning_rate": 1.8277641685944555e-05,
+      "loss": 1.1993,
+      "step": 803
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1609342055050313,
+      "learning_rate": 1.827279476340881e-05,
+      "loss": 0.7816,
+      "step": 804
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.021614775418659,
+      "learning_rate": 1.826794167509544e-05,
+      "loss": 0.6832,
+      "step": 805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8053397937846425,
+      "learning_rate": 1.8263082424621488e-05,
+      "loss": 0.6039,
+      "step": 806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8927681449782678,
+      "learning_rate": 1.825821701560859e-05,
+      "loss": 0.6776,
+      "step": 807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.101975949503301,
+      "learning_rate": 1.8253345451682972e-05,
+      "loss": 0.5989,
+      "step": 808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9877080250653585,
+      "learning_rate": 1.8248467736475452e-05,
+      "loss": 0.6852,
+      "step": 809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2203666735853504,
+      "learning_rate": 1.8243583873621426e-05,
+      "loss": 0.7024,
+      "step": 810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8981835612016067,
+      "learning_rate": 1.8238693866760876e-05,
+      "loss": 0.7037,
+      "step": 811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.17019339001328,
+      "learning_rate": 1.823379771953836e-05,
+      "loss": 1.1844,
+      "step": 812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.99306199987272,
+      "learning_rate": 1.8228895435603018e-05,
+      "loss": 0.5865,
+      "step": 813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4890361119647233,
+      "learning_rate": 1.8223987018608556e-05,
+      "loss": 1.265,
+      "step": 814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.457747309281537,
+      "learning_rate": 1.8219072472213258e-05,
+      "loss": 0.6701,
+      "step": 815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1845104730612337,
+      "learning_rate": 1.8214151800079968e-05,
+      "loss": 0.7037,
+      "step": 816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3237160259210676,
+      "learning_rate": 1.8209225005876105e-05,
+      "loss": 0.6491,
+      "step": 817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.48525133950815,
+      "learning_rate": 1.8204292093273645e-05,
+      "loss": 1.2428,
+      "step": 818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9468735182096222,
+      "learning_rate": 1.8199353065949127e-05,
+      "loss": 0.6793,
+      "step": 819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8247593731841958,
+      "learning_rate": 1.819440792758365e-05,
+      "loss": 0.6193,
+      "step": 820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.79783650834899,
+      "learning_rate": 1.8189456681862853e-05,
+      "loss": 0.6455,
+      "step": 821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.945575928192751,
+      "learning_rate": 1.818449933247695e-05,
+      "loss": 0.6309,
+      "step": 822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9342174254451685,
+      "learning_rate": 1.817953588312068e-05,
+      "loss": 0.7318,
+      "step": 823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.085508074291521,
+      "learning_rate": 1.817456633749335e-05,
+      "loss": 0.606,
+      "step": 824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8854139150118179,
+      "learning_rate": 1.8169590699298792e-05,
+      "loss": 0.6335,
+      "step": 825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9259671563756717,
+      "learning_rate": 1.8164608972245397e-05,
+      "loss": 0.6783,
+      "step": 826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.598045116702108,
+      "learning_rate": 1.8159621160046078e-05,
+      "loss": 1.2446,
+      "step": 827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5289129824118253,
+      "learning_rate": 1.8154627266418287e-05,
+      "loss": 1.2003,
+      "step": 828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9734147023233162,
+      "learning_rate": 1.814962729508402e-05,
+      "loss": 0.6929,
+      "step": 829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4519348034553716,
+      "learning_rate": 1.8144621249769785e-05,
+      "loss": 0.7091,
+      "step": 830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2513169788679948,
+      "learning_rate": 1.8139609134206634e-05,
+      "loss": 0.6144,
+      "step": 831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.061488760187201,
+      "learning_rate": 1.8134590952130132e-05,
+      "loss": 0.6374,
+      "step": 832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9933636900425624,
+      "learning_rate": 1.812956670728037e-05,
+      "loss": 0.6882,
+      "step": 833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1576964943112036,
+      "learning_rate": 1.8124536403401953e-05,
+      "loss": 0.7256,
+      "step": 834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8248369097823713,
+      "learning_rate": 1.811950004424401e-05,
+      "loss": 0.6134,
+      "step": 835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1297954868314046,
+      "learning_rate": 1.8114457633560173e-05,
+      "loss": 1.2051,
+      "step": 836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9611625745301182,
+      "learning_rate": 1.810940917510859e-05,
+      "loss": 0.6138,
+      "step": 837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.691195820470262,
+      "learning_rate": 1.810435467265192e-05,
+      "loss": 1.1081,
+      "step": 838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9488804164944507,
+      "learning_rate": 1.8099294129957322e-05,
+      "loss": 0.5946,
+      "step": 839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8830297710348294,
+      "learning_rate": 1.8094227550796453e-05,
+      "loss": 0.6656,
+      "step": 840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.951282910123267,
+      "learning_rate": 1.808915493894548e-05,
+      "loss": 0.623,
+      "step": 841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8359261065202859,
+      "learning_rate": 1.8084076298185053e-05,
+      "loss": 0.7367,
+      "step": 842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6806685297040014,
+      "learning_rate": 1.8078991632300327e-05,
+      "loss": 1.2517,
+      "step": 843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.548770350089678,
+      "learning_rate": 1.807390094508094e-05,
+      "loss": 1.1736,
+      "step": 844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8174607529596745,
+      "learning_rate": 1.8068804240321024e-05,
+      "loss": 0.6332,
+      "step": 845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2682887702628323,
+      "learning_rate": 1.8063701521819184e-05,
+      "loss": 1.1649,
+      "step": 846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.957251960856155,
+      "learning_rate": 1.8058592793378528e-05,
+      "loss": 0.6093,
+      "step": 847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.118857021960424,
+      "learning_rate": 1.805347805880662e-05,
+      "loss": 1.2081,
+      "step": 848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9707340042394248,
+      "learning_rate": 1.8048357321915513e-05,
+      "loss": 0.6366,
+      "step": 849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.664030386164068,
+      "learning_rate": 1.8043230586521736e-05,
+      "loss": 1.1808,
+      "step": 850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.556161788399397,
+      "learning_rate": 1.803809785644628e-05,
+      "loss": 1.1915,
+      "step": 851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.399108242899878,
+      "learning_rate": 1.803295913551461e-05,
+      "loss": 1.1497,
+      "step": 852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.580341004354619,
+      "learning_rate": 1.8027814427556653e-05,
+      "loss": 1.2407,
+      "step": 853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2575919118726357,
+      "learning_rate": 1.8022663736406804e-05,
+      "loss": 0.6711,
+      "step": 854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.678337547167258,
+      "learning_rate": 1.8017507065903904e-05,
+      "loss": 1.1862,
+      "step": 855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.015569930065479,
+      "learning_rate": 1.8012344419891268e-05,
+      "loss": 0.6777,
+      "step": 856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8140612408226313,
+      "learning_rate": 1.8007175802216652e-05,
+      "loss": 0.7471,
+      "step": 857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.989693685680955,
+      "learning_rate": 1.800200121673227e-05,
+      "loss": 0.5935,
+      "step": 858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8267639082271683,
+      "learning_rate": 1.799682066729477e-05,
+      "loss": 0.6048,
+      "step": 859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.942125912005556,
+      "learning_rate": 1.799163415776527e-05,
+      "loss": 1.1617,
+      "step": 860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9862339727933747,
+      "learning_rate": 1.7986441692009302e-05,
+      "loss": 0.5998,
+      "step": 861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.880108092954622,
+      "learning_rate": 1.798124327389686e-05,
+      "loss": 0.6712,
+      "step": 862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.816549956213632,
+      "learning_rate": 1.7976038907302363e-05,
+      "loss": 0.6731,
+      "step": 863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.013597835205249,
+      "learning_rate": 1.7970828596104662e-05,
+      "loss": 0.6343,
+      "step": 864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7688711832845936,
+      "learning_rate": 1.7965612344187046e-05,
+      "loss": 1.211,
+      "step": 865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.683972918340153,
+      "learning_rate": 1.7960390155437225e-05,
+      "loss": 1.2585,
+      "step": 866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8978919070189568,
+      "learning_rate": 1.7955162033747337e-05,
+      "loss": 0.6638,
+      "step": 867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.569570563125772,
+      "learning_rate": 1.7949927983013946e-05,
+      "loss": 1.1311,
+      "step": 868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7580315348210944,
+      "learning_rate": 1.7944688007138023e-05,
+      "loss": 1.2258,
+      "step": 869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.939186251233678,
+      "learning_rate": 1.7939442110024967e-05,
+      "loss": 0.6027,
+      "step": 870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.97564979909395,
+      "learning_rate": 1.7934190295584583e-05,
+      "loss": 0.63,
+      "step": 871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.959479903549918,
+      "learning_rate": 1.7928932567731088e-05,
+      "loss": 1.2079,
+      "step": 872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8057221668883088,
+      "learning_rate": 1.792366893038311e-05,
+      "loss": 1.2116,
+      "step": 873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0092861109047555,
+      "learning_rate": 1.791839938746368e-05,
+      "loss": 0.6864,
+      "step": 874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8911521092989612,
+      "learning_rate": 1.7913123942900224e-05,
+      "loss": 0.6231,
+      "step": 875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.19140994514885,
+      "learning_rate": 1.7907842600624574e-05,
+      "loss": 0.7239,
+      "step": 876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.993536637737408,
+      "learning_rate": 1.7902555364572952e-05,
+      "loss": 1.2058,
+      "step": 877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.796343018630878,
+      "learning_rate": 1.7897262238685985e-05,
+      "loss": 0.6125,
+      "step": 878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.878027003915466,
+      "learning_rate": 1.7891963226908667e-05,
+      "loss": 0.6627,
+      "step": 879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7083236386301612,
+      "learning_rate": 1.7886658333190406e-05,
+      "loss": 1.2089,
+      "step": 880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0307887792443777,
+      "learning_rate": 1.7881347561484967e-05,
+      "loss": 0.6615,
+      "step": 881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8423123739514617,
+      "learning_rate": 1.787603091575052e-05,
+      "loss": 0.5939,
+      "step": 882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9350853645466533,
+      "learning_rate": 1.7870708399949595e-05,
+      "loss": 0.6173,
+      "step": 883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.055749953310832,
+      "learning_rate": 1.786538001804911e-05,
+      "loss": 0.6177,
+      "step": 884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0580601754997927,
+      "learning_rate": 1.7860045774020342e-05,
+      "loss": 0.6864,
+      "step": 885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2391615559394307,
+      "learning_rate": 1.7854705671838948e-05,
+      "loss": 1.2023,
+      "step": 886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6340602785188865,
+      "learning_rate": 1.7849359715484946e-05,
+      "loss": 1.1758,
+      "step": 887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.850127953842047,
+      "learning_rate": 1.784400790894272e-05,
+      "loss": 0.626,
+      "step": 888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.089135675003734,
+      "learning_rate": 1.7838650256201014e-05,
+      "loss": 0.6989,
+      "step": 889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.636446359439468,
+      "learning_rate": 1.7833286761252924e-05,
+      "loss": 1.1933,
+      "step": 890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1650353051142868,
+      "learning_rate": 1.7827917428095905e-05,
+      "loss": 1.1759,
+      "step": 891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.919915365978683,
+      "learning_rate": 1.7822542260731763e-05,
+      "loss": 1.2546,
+      "step": 892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.469017455014408,
+      "learning_rate": 1.7817161263166655e-05,
+      "loss": 1.2153,
+      "step": 893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8620429338185995,
+      "learning_rate": 1.781177443941107e-05,
+      "loss": 0.615,
+      "step": 894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9764683700304042,
+      "learning_rate": 1.7806381793479863e-05,
+      "loss": 0.7046,
+      "step": 895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0273509664787164,
+      "learning_rate": 1.7800983329392205e-05,
+      "loss": 0.725,
+      "step": 896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.824819623776793,
+      "learning_rate": 1.7795579051171614e-05,
+      "loss": 0.6274,
+      "step": 897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8875980650850082,
+      "learning_rate": 1.7790168962845943e-05,
+      "loss": 0.6208,
+      "step": 898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0451509924103153,
+      "learning_rate": 1.778475306844737e-05,
+      "loss": 0.6627,
+      "step": 899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.061649630702216,
+      "learning_rate": 1.7779331372012404e-05,
+      "loss": 0.6759,
+      "step": 900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.836727441403032,
+      "learning_rate": 1.7773903877581876e-05,
+      "loss": 0.6841,
+      "step": 901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.055359500265629,
+      "learning_rate": 1.7768470589200945e-05,
+      "loss": 0.6615,
+      "step": 902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.811670866691933,
+      "learning_rate": 1.776303151091907e-05,
+      "loss": 1.2194,
+      "step": 903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1349061535089664,
+      "learning_rate": 1.775758664679005e-05,
+      "loss": 0.7116,
+      "step": 904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1776682684410233,
+      "learning_rate": 1.775213600087198e-05,
+      "loss": 0.6338,
+      "step": 905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.024267186616694,
+      "learning_rate": 1.7746679577227268e-05,
+      "loss": 0.63,
+      "step": 906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.197220751023389,
+      "learning_rate": 1.7741217379922627e-05,
+      "loss": 0.7485,
+      "step": 907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9051367141307787,
+      "learning_rate": 1.773574941302908e-05,
+      "loss": 1.1796,
+      "step": 908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.550855139457756,
+      "learning_rate": 1.7730275680621938e-05,
+      "loss": 1.224,
+      "step": 909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1510695446057047,
+      "learning_rate": 1.772479618678082e-05,
+      "loss": 0.5677,
+      "step": 910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.200019032394908,
+      "learning_rate": 1.7719310935589638e-05,
+      "loss": 1.2036,
+      "step": 911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0916097981673643,
+      "learning_rate": 1.7713819931136585e-05,
+      "loss": 0.6819,
+      "step": 912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7605826732828034,
+      "learning_rate": 1.770832317751415e-05,
+      "loss": 0.5997,
+      "step": 913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6035905137038498,
+      "learning_rate": 1.770282067881911e-05,
+      "loss": 1.1757,
+      "step": 914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.685035957106679,
+      "learning_rate": 1.769731243915252e-05,
+      "loss": 1.107,
+      "step": 915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8000179772747096,
+      "learning_rate": 1.7691798462619706e-05,
+      "loss": 0.6739,
+      "step": 916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0215502567902126,
+      "learning_rate": 1.7686278753330283e-05,
+      "loss": 0.643,
+      "step": 917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4720752109104915,
+      "learning_rate": 1.7680753315398135e-05,
+      "loss": 1.1783,
+      "step": 918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.001642366383445,
+      "learning_rate": 1.7675222152941407e-05,
+      "loss": 0.6961,
+      "step": 919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9781563577008903,
+      "learning_rate": 1.766968527008252e-05,
+      "loss": 0.6774,
+      "step": 920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.859333058042282,
+      "learning_rate": 1.7664142670948158e-05,
+      "loss": 1.2075,
+      "step": 921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9058147941654577,
+      "learning_rate": 1.7658594359669262e-05,
+      "loss": 0.6006,
+      "step": 922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8796943678044213,
+      "learning_rate": 1.7653040340381028e-05,
+      "loss": 0.6975,
+      "step": 923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0945835174271616,
+      "learning_rate": 1.764748061722291e-05,
+      "loss": 0.6944,
+      "step": 924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0620664357507947,
+      "learning_rate": 1.7641915194338614e-05,
+      "loss": 0.6457,
+      "step": 925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0925883593704526,
+      "learning_rate": 1.763634407587609e-05,
+      "loss": 0.6987,
+      "step": 926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8057701618871036,
+      "learning_rate": 1.763076726598754e-05,
+      "loss": 0.6242,
+      "step": 927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9881651521477746,
+      "learning_rate": 1.76251847688294e-05,
+      "loss": 0.665,
+      "step": 928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6009677562647493,
+      "learning_rate": 1.7619596588562347e-05,
+      "loss": 1.238,
+      "step": 929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.552500222754272,
+      "learning_rate": 1.76140027293513e-05,
+      "loss": 1.1922,
+      "step": 930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0022180646656085,
+      "learning_rate": 1.76084031953654e-05,
+      "loss": 0.7079,
+      "step": 931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5685033035932143,
+      "learning_rate": 1.7602797990778025e-05,
+      "loss": 1.2062,
+      "step": 932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.073796566163718,
+      "learning_rate": 1.7597187119766775e-05,
+      "loss": 0.6391,
+      "step": 933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8661388732286794,
+      "learning_rate": 1.759157058651348e-05,
+      "loss": 0.623,
+      "step": 934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9458485624929227,
+      "learning_rate": 1.7585948395204182e-05,
+      "loss": 0.6923,
+      "step": 935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9569437538933494,
+      "learning_rate": 1.7580320550029145e-05,
+      "loss": 1.2125,
+      "step": 936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7911861683583434,
+      "learning_rate": 1.7574687055182842e-05,
+      "loss": 0.6487,
+      "step": 937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8292823973844334,
+      "learning_rate": 1.7569047914863965e-05,
+      "loss": 1.1635,
+      "step": 938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7874772617423247,
+      "learning_rate": 1.7563403133275406e-05,
+      "loss": 0.7048,
+      "step": 939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9773479540151229,
+      "learning_rate": 1.7557752714624264e-05,
+      "loss": 0.6357,
+      "step": 940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9054547174417413,
+      "learning_rate": 1.7552096663121837e-05,
+      "loss": 0.7137,
+      "step": 941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7934851808800225,
+      "learning_rate": 1.754643498298363e-05,
+      "loss": 0.6529,
+      "step": 942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.012050663928789,
+      "learning_rate": 1.7540767678429334e-05,
+      "loss": 0.6545,
+      "step": 943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.042739406760189,
+      "learning_rate": 1.7535094753682833e-05,
+      "loss": 1.1628,
+      "step": 944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0102046440395434,
+      "learning_rate": 1.7529416212972203e-05,
+      "loss": 0.6611,
+      "step": 945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.819694722773333,
+      "learning_rate": 1.7523732060529703e-05,
+      "loss": 0.5547,
+      "step": 946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.561652703889409,
+      "learning_rate": 1.7518042300591778e-05,
+      "loss": 1.1758,
+      "step": 947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9189160786129655,
+      "learning_rate": 1.7512346937399048e-05,
+      "loss": 0.6618,
+      "step": 948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8853410744661896,
+      "learning_rate": 1.750664597519631e-05,
+      "loss": 1.1665,
+      "step": 949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.817251605732727,
+      "learning_rate": 1.7500939418232537e-05,
+      "loss": 0.6151,
+      "step": 950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8111600772902596,
+      "learning_rate": 1.749522727076087e-05,
+      "loss": 1.2121,
+      "step": 951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.030887725385215,
+      "learning_rate": 1.7489509537038615e-05,
+      "loss": 0.6415,
+      "step": 952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0107410977317723,
+      "learning_rate": 1.7483786221327245e-05,
+      "loss": 0.6845,
+      "step": 953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7584597684782424,
+      "learning_rate": 1.7478057327892387e-05,
+      "loss": 0.6377,
+      "step": 954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9308103640572758,
+      "learning_rate": 1.7472322861003833e-05,
+      "loss": 0.6164,
+      "step": 955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.43793301372731,
+      "learning_rate": 1.7466582824935525e-05,
+      "loss": 1.2244,
+      "step": 956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7774578682513285,
+      "learning_rate": 1.7460837223965556e-05,
+      "loss": 0.6498,
+      "step": 957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0636573629059716,
+      "learning_rate": 1.7455086062376167e-05,
+      "loss": 0.6154,
+      "step": 958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.199992542810336,
+      "learning_rate": 1.744932934445374e-05,
+      "loss": 1.1945,
+      "step": 959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6275723947409624,
+      "learning_rate": 1.7443567074488802e-05,
+      "loss": 1.2128,
+      "step": 960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6219802308160767,
+      "learning_rate": 1.7437799256776024e-05,
+      "loss": 1.1159,
+      "step": 961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1456135549487256,
+      "learning_rate": 1.7432025895614197e-05,
+      "loss": 0.6447,
+      "step": 962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7504903239767118,
+      "learning_rate": 1.7426246995306253e-05,
+      "loss": 0.6826,
+      "step": 963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9900591271746633,
+      "learning_rate": 1.7420462560159256e-05,
+      "loss": 0.6824,
+      "step": 964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.951788737649214,
+      "learning_rate": 1.741467259448438e-05,
+      "loss": 1.1847,
+      "step": 965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.966417489416672,
+      "learning_rate": 1.7408877102596943e-05,
+      "loss": 0.6153,
+      "step": 966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0995925295349895,
+      "learning_rate": 1.740307608881636e-05,
+      "loss": 0.6727,
+      "step": 967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.063883471072763,
+      "learning_rate": 1.7397269557466178e-05,
+      "loss": 0.6377,
+      "step": 968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.083382564422627,
+      "learning_rate": 1.7391457512874042e-05,
+      "loss": 0.6685,
+      "step": 969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.967367438812025,
+      "learning_rate": 1.738563995937172e-05,
+      "loss": 0.6523,
+      "step": 970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0644185797703876,
+      "learning_rate": 1.737981690129508e-05,
+      "loss": 1.1742,
+      "step": 971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0308361025247623,
+      "learning_rate": 1.7373988342984085e-05,
+      "loss": 1.1734,
+      "step": 972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1055995425977687,
+      "learning_rate": 1.7368154288782814e-05,
+      "loss": 0.6194,
+      "step": 973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0148580471393585,
+      "learning_rate": 1.7362314743039422e-05,
+      "loss": 0.6666,
+      "step": 974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3445875189519803,
+      "learning_rate": 1.735646971010618e-05,
+      "loss": 0.7257,
+      "step": 975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.002326384127028,
+      "learning_rate": 1.7350619194339423e-05,
+      "loss": 0.6241,
+      "step": 976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.06334198344689,
+      "learning_rate": 1.734476320009959e-05,
+      "loss": 1.22,
+      "step": 977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.680840081465969,
+      "learning_rate": 1.7338901731751204e-05,
+      "loss": 1.2235,
+      "step": 978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8754677465934946,
+      "learning_rate": 1.7333034793662856e-05,
+      "loss": 0.6494,
+      "step": 979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8507940327486843,
+      "learning_rate": 1.7327162390207223e-05,
+      "loss": 0.6948,
+      "step": 980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0036398856073485,
+      "learning_rate": 1.7321284525761054e-05,
+      "loss": 0.6972,
+      "step": 981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.163525141259841,
+      "learning_rate": 1.7315401204705163e-05,
+      "loss": 0.6633,
+      "step": 982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8423238270577258,
+      "learning_rate": 1.7309512431424435e-05,
+      "loss": 0.6504,
+      "step": 983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9478883762084243,
+      "learning_rate": 1.730361821030782e-05,
+      "loss": 1.1653,
+      "step": 984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.719632456040685,
+      "learning_rate": 1.7297718545748327e-05,
+      "loss": 1.2974,
+      "step": 985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.859892905699994,
+      "learning_rate": 1.7291813442143018e-05,
+      "loss": 0.5662,
+      "step": 986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1671749777584313,
+      "learning_rate": 1.7285902903893014e-05,
+      "loss": 0.6251,
+      "step": 987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9627098154745108,
+      "learning_rate": 1.7279986935403487e-05,
+      "loss": 0.6613,
+      "step": 988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0887404667467613,
+      "learning_rate": 1.727406554108365e-05,
+      "loss": 0.6319,
+      "step": 989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9994196998695055,
+      "learning_rate": 1.7268138725346766e-05,
+      "loss": 0.6949,
+      "step": 990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.511789565494871,
+      "learning_rate": 1.7262206492610136e-05,
+      "loss": 1.2134,
+      "step": 991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.638721433542888,
+      "learning_rate": 1.7256268847295096e-05,
+      "loss": 1.1765,
+      "step": 992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.780857153205876,
+      "learning_rate": 1.725032579382702e-05,
+      "loss": 1.1873,
+      "step": 993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.347064681695982,
+      "learning_rate": 1.724437733663532e-05,
+      "loss": 1.1572,
+      "step": 994
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3979840978050686,
+      "learning_rate": 1.7238423480153413e-05,
+      "loss": 1.1786,
+      "step": 995
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0556862346212394,
+      "learning_rate": 1.7232464228818767e-05,
+      "loss": 0.6447,
+      "step": 996
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.283728265706108,
+      "learning_rate": 1.7226499587072846e-05,
+      "loss": 0.651,
+      "step": 997
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.180860370577453,
+      "learning_rate": 1.7220529559361153e-05,
+      "loss": 0.6534,
+      "step": 998
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8809302375101968,
+      "learning_rate": 1.721455415013319e-05,
+      "loss": 0.6025,
+      "step": 999
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5486719464405625,
+      "learning_rate": 1.7208573363842477e-05,
+      "loss": 1.175,
+      "step": 1000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8210473668442617,
+      "learning_rate": 1.720258720494654e-05,
+      "loss": 0.6323,
+      "step": 1001
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9801881166649589,
+      "learning_rate": 1.7196595677906912e-05,
+      "loss": 0.6486,
+      "step": 1002
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9302794570813897,
+      "learning_rate": 1.719059878718912e-05,
+      "loss": 0.5839,
+      "step": 1003
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7530740453712697,
+      "learning_rate": 1.7184596537262696e-05,
+      "loss": 1.2771,
+      "step": 1004
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9325830619917133,
+      "learning_rate": 1.717858893260116e-05,
+      "loss": 0.6657,
+      "step": 1005
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.943134550410303,
+      "learning_rate": 1.7172575977682027e-05,
+      "loss": 0.6485,
+      "step": 1006
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7999518292999508,
+      "learning_rate": 1.71665576769868e-05,
+      "loss": 0.6033,
+      "step": 1007
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.070030044749354,
+      "learning_rate": 1.7160534035000966e-05,
+      "loss": 0.6438,
+      "step": 1008
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9561246292467036,
+      "learning_rate": 1.7154505056213985e-05,
+      "loss": 0.6491,
+      "step": 1009
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9905483573084506,
+      "learning_rate": 1.7148470745119313e-05,
+      "loss": 0.627,
+      "step": 1010
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1870497895074092,
+      "learning_rate": 1.7142431106214356e-05,
+      "loss": 1.1966,
+      "step": 1011
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.809600009636495,
+      "learning_rate": 1.7136386144000513e-05,
+      "loss": 1.1923,
+      "step": 1012
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.369814887363469,
+      "learning_rate": 1.7130335862983133e-05,
+      "loss": 1.1669,
+      "step": 1013
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2876108634692636,
+      "learning_rate": 1.7124280267671546e-05,
+      "loss": 0.6842,
+      "step": 1014
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8581929739963554,
+      "learning_rate": 1.7118219362579026e-05,
+      "loss": 1.2059,
+      "step": 1015
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.459508840194685,
+      "learning_rate": 1.7112153152222815e-05,
+      "loss": 0.6271,
+      "step": 1016
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.26576007505901,
+      "learning_rate": 1.7106081641124107e-05,
+      "loss": 0.6585,
+      "step": 1017
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9184061873913074,
+      "learning_rate": 1.7100004833808046e-05,
+      "loss": 1.1806,
+      "step": 1018
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1625354125554033,
+      "learning_rate": 1.709392273480372e-05,
+      "loss": 0.6521,
+      "step": 1019
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8503035705408712,
+      "learning_rate": 1.7087835348644167e-05,
+      "loss": 1.2087,
+      "step": 1020
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9915673426148066,
+      "learning_rate": 1.7081742679866368e-05,
+      "loss": 1.2091,
+      "step": 1021
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0456928504550715,
+      "learning_rate": 1.7075644733011227e-05,
+      "loss": 0.6636,
+      "step": 1022
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0722951611126597,
+      "learning_rate": 1.7069541512623593e-05,
+      "loss": 0.6433,
+      "step": 1023
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8694091884723605,
+      "learning_rate": 1.7063433023252244e-05,
+      "loss": 0.6144,
+      "step": 1024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.552689152070877,
+      "learning_rate": 1.7057319269449887e-05,
+      "loss": 1.229,
+      "step": 1025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9423266128341283,
+      "learning_rate": 1.7051200255773142e-05,
+      "loss": 0.6791,
+      "step": 1026
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8535851895971733,
+      "learning_rate": 1.7045075986782567e-05,
+      "loss": 1.1561,
+      "step": 1027
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7829560898430055,
+      "learning_rate": 1.7038946467042623e-05,
+      "loss": 1.2558,
+      "step": 1028
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8900696709716174,
+      "learning_rate": 1.7032811701121684e-05,
+      "loss": 0.6371,
+      "step": 1029
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.089534925058146,
+      "learning_rate": 1.7026671693592043e-05,
+      "loss": 0.6328,
+      "step": 1030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.861141627716313,
+      "learning_rate": 1.7020526449029894e-05,
+      "loss": 1.2595,
+      "step": 1031
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.735044697949092,
+      "learning_rate": 1.7014375972015334e-05,
+      "loss": 1.1802,
+      "step": 1032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7465022979296163,
+      "learning_rate": 1.7008220267132364e-05,
+      "loss": 0.6039,
+      "step": 1033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8294272006827355,
+      "learning_rate": 1.700205933896887e-05,
+      "loss": 0.6337,
+      "step": 1034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.231485498036314,
+      "learning_rate": 1.699589319211665e-05,
+      "loss": 0.643,
+      "step": 1035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3320553914771613,
+      "learning_rate": 1.6989721831171376e-05,
+      "loss": 0.654,
+      "step": 1036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0968201273198668,
+      "learning_rate": 1.698354526073261e-05,
+      "loss": 0.6497,
+      "step": 1037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.0926783511344285,
+      "learning_rate": 1.6977363485403794e-05,
+      "loss": 1.1545,
+      "step": 1038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9281762078878044,
+      "learning_rate": 1.6971176509792257e-05,
+      "loss": 0.591,
+      "step": 1039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9782166058391975,
+      "learning_rate": 1.6964984338509203e-05,
+      "loss": 0.6235,
+      "step": 1040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3917914734885013,
+      "learning_rate": 1.6958786976169696e-05,
+      "loss": 0.6864,
+      "step": 1041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.420131368179053,
+      "learning_rate": 1.695258442739269e-05,
+      "loss": 0.6581,
+      "step": 1042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3499516796802404,
+      "learning_rate": 1.694637669680098e-05,
+      "loss": 0.7004,
+      "step": 1043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0154821388242494,
+      "learning_rate": 1.6940163789021243e-05,
+      "loss": 0.6588,
+      "step": 1044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9282093297296898,
+      "learning_rate": 1.6933945708684004e-05,
+      "loss": 0.6883,
+      "step": 1045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.091514615409396,
+      "learning_rate": 1.6927722460423645e-05,
+      "loss": 0.6086,
+      "step": 1046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0063003671465958,
+      "learning_rate": 1.692149404887841e-05,
+      "loss": 0.5683,
+      "step": 1047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2305355564593308,
+      "learning_rate": 1.6915260478690368e-05,
+      "loss": 0.6158,
+      "step": 1048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9952185625922674,
+      "learning_rate": 1.6909021754505454e-05,
+      "loss": 1.1789,
+      "step": 1049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6048079064370517,
+      "learning_rate": 1.690277788097344e-05,
+      "loss": 0.635,
+      "step": 1050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6616693974764627,
+      "learning_rate": 1.689652886274793e-05,
+      "loss": 1.2018,
+      "step": 1051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8898153246931821,
+      "learning_rate": 1.689027470448636e-05,
+      "loss": 0.6215,
+      "step": 1052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2395477050723698,
+      "learning_rate": 1.688401541085001e-05,
+      "loss": 1.1632,
+      "step": 1053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8694243508413169,
+      "learning_rate": 1.687775098650398e-05,
+      "loss": 0.6823,
+      "step": 1054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2209823202421104,
+      "learning_rate": 1.687148143611719e-05,
+      "loss": 1.1752,
+      "step": 1055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2154485407027042,
+      "learning_rate": 1.6865206764362385e-05,
+      "loss": 0.6875,
+      "step": 1056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9254624045102509,
+      "learning_rate": 1.6858926975916125e-05,
+      "loss": 0.6211,
+      "step": 1057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9953243788377517,
+      "learning_rate": 1.6852642075458784e-05,
+      "loss": 0.6622,
+      "step": 1058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1648202269573167,
+      "learning_rate": 1.6846352067674555e-05,
+      "loss": 0.6225,
+      "step": 1059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8717118394731238,
+      "learning_rate": 1.6840056957251417e-05,
+      "loss": 0.6335,
+      "step": 1060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.57763639830572,
+      "learning_rate": 1.6833756748881172e-05,
+      "loss": 1.2557,
+      "step": 1061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.504031651213194,
+      "learning_rate": 1.6827451447259414e-05,
+      "loss": 1.1852,
+      "step": 1062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8601238185655697,
+      "learning_rate": 1.682114105708553e-05,
+      "loss": 1.1705,
+      "step": 1063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.519677407242201,
+      "learning_rate": 1.68148255830627e-05,
+      "loss": 0.6433,
+      "step": 1064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3625901203067445,
+      "learning_rate": 1.6808505029897902e-05,
+      "loss": 0.6509,
+      "step": 1065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.543109901275445,
+      "learning_rate": 1.680217940230189e-05,
+      "loss": 1.1943,
+      "step": 1066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9126912143646655,
+      "learning_rate": 1.6795848704989202e-05,
+      "loss": 0.6802,
+      "step": 1067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.519575625608034,
+      "learning_rate": 1.6789512942678155e-05,
+      "loss": 1.2109,
+      "step": 1068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8434588125664109,
+      "learning_rate": 1.6783172120090843e-05,
+      "loss": 0.5282,
+      "step": 1069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2408393958850574,
+      "learning_rate": 1.6776826241953128e-05,
+      "loss": 0.665,
+      "step": 1070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9518183038120234,
+      "learning_rate": 1.6770475312994642e-05,
+      "loss": 1.1887,
+      "step": 1071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.136963517445899,
+      "learning_rate": 1.6764119337948788e-05,
+      "loss": 1.1714,
+      "step": 1072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2442307826716092,
+      "learning_rate": 1.6757758321552713e-05,
+      "loss": 0.6882,
+      "step": 1073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7828405062390644,
+      "learning_rate": 1.6751392268547335e-05,
+      "loss": 0.6078,
+      "step": 1074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7375181155023445,
+      "learning_rate": 1.6745021183677327e-05,
+      "loss": 0.684,
+      "step": 1075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.000815630262636,
+      "learning_rate": 1.67386450716911e-05,
+      "loss": 0.6273,
+      "step": 1076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8330755177160274,
+      "learning_rate": 1.6732263937340823e-05,
+      "loss": 0.6346,
+      "step": 1077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9133661539472575,
+      "learning_rate": 1.672587778538241e-05,
+      "loss": 0.613,
+      "step": 1078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9686565775325944,
+      "learning_rate": 1.6719486620575497e-05,
+      "loss": 1.1576,
+      "step": 1079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.154339948897071,
+      "learning_rate": 1.671309044768348e-05,
+      "loss": 1.1919,
+      "step": 1080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3361722491492407,
+      "learning_rate": 1.670668927147347e-05,
+      "loss": 0.6342,
+      "step": 1081
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1351719682139554,
+      "learning_rate": 1.6700283096716316e-05,
+      "loss": 0.5893,
+      "step": 1082
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6159099493470497,
+      "learning_rate": 1.6693871928186585e-05,
+      "loss": 0.6917,
+      "step": 1083
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.368078679416766,
+      "learning_rate": 1.6687455770662572e-05,
+      "loss": 0.656,
+      "step": 1084
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.074154708432402,
+      "learning_rate": 1.668103462892629e-05,
+      "loss": 0.6267,
+      "step": 1085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1046396532893885,
+      "learning_rate": 1.6674608507763465e-05,
+      "loss": 0.7503,
+      "step": 1086
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.149002287340963,
+      "learning_rate": 1.666817741196354e-05,
+      "loss": 1.2217,
+      "step": 1087
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.016587970180482,
+      "learning_rate": 1.6661741346319648e-05,
+      "loss": 0.6061,
+      "step": 1088
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.629667831046643,
+      "learning_rate": 1.6655300315628643e-05,
+      "loss": 0.6863,
+      "step": 1089
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0079527573851,
+      "learning_rate": 1.664885432469108e-05,
+      "loss": 1.1998,
+      "step": 1090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.671357840934744,
+      "learning_rate": 1.66424033783112e-05,
+      "loss": 1.1526,
+      "step": 1091
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5473586867776277,
+      "learning_rate": 1.663594748129695e-05,
+      "loss": 1.1455,
+      "step": 1092
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0982839653077536,
+      "learning_rate": 1.6629486638459953e-05,
+      "loss": 0.6833,
+      "step": 1093
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1560444555254197,
+      "learning_rate": 1.6623020854615523e-05,
+      "loss": 0.628,
+      "step": 1094
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.017329020369992,
+      "learning_rate": 1.661655013458267e-05,
+      "loss": 0.6241,
+      "step": 1095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0318819905139587,
+      "learning_rate": 1.6610074483184054e-05,
+      "loss": 0.7067,
+      "step": 1096
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9781218903524462,
+      "learning_rate": 1.6603593905246044e-05,
+      "loss": 0.6692,
+      "step": 1097
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6591676123354213,
+      "learning_rate": 1.6597108405598653e-05,
+      "loss": 0.6437,
+      "step": 1098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1172881475566565,
+      "learning_rate": 1.6590617989075585e-05,
+      "loss": 0.6881,
+      "step": 1099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0885551254110264,
+      "learning_rate": 1.658412266051419e-05,
+      "loss": 0.6623,
+      "step": 1100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.242888255398427,
+      "learning_rate": 1.657762242475549e-05,
+      "loss": 0.7433,
+      "step": 1101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9409368139136638,
+      "learning_rate": 1.6571117286644153e-05,
+      "loss": 0.6269,
+      "step": 1102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.103775610661748,
+      "learning_rate": 1.6564607251028517e-05,
+      "loss": 0.6345,
+      "step": 1103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.005556062964028,
+      "learning_rate": 1.655809232276056e-05,
+      "loss": 0.5522,
+      "step": 1104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.717409551479153,
+      "learning_rate": 1.6551572506695905e-05,
+      "loss": 1.2321,
+      "step": 1105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.817322883750143,
+      "learning_rate": 1.6545047807693824e-05,
+      "loss": 1.1431,
+      "step": 1106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1142127478038555,
+      "learning_rate": 1.6538518230617226e-05,
+      "loss": 0.681,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1657954474690824,
+      "learning_rate": 1.653198378033265e-05,
+      "loss": 0.6993,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2299784649121572,
+      "learning_rate": 1.652544446171028e-05,
+      "loss": 0.7146,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.971949390707741,
+      "learning_rate": 1.6518900279623915e-05,
+      "loss": 1.2066,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0601288762182226,
+      "learning_rate": 1.651235123895098e-05,
+      "loss": 0.6776,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8418053057004355,
+      "learning_rate": 1.6505797344572538e-05,
+      "loss": 1.2117,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7177784333481765,
+      "learning_rate": 1.6499238601373243e-05,
+      "loss": 0.6426,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8627343128348612,
+      "learning_rate": 1.649267501424139e-05,
+      "loss": 1.1806,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.069456720892328,
+      "learning_rate": 1.648610658806886e-05,
+      "loss": 0.6222,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6728521110913572,
+      "learning_rate": 1.647953332775116e-05,
+      "loss": 1.161,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0666790914146227,
+      "learning_rate": 1.6472955238187386e-05,
+      "loss": 0.6751,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8613508647040278,
+      "learning_rate": 1.646637232428024e-05,
+      "loss": 0.5787,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9100831679088544,
+      "learning_rate": 1.645978459093602e-05,
+      "loss": 0.6464,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.238480180282339,
+      "learning_rate": 1.6453192043064614e-05,
+      "loss": 1.2102,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.049159488965949,
+      "learning_rate": 1.64465946855795e-05,
+      "loss": 0.6901,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8903731978515212,
+      "learning_rate": 1.643999252339773e-05,
+      "loss": 1.1363,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.202969383242651,
+      "learning_rate": 1.643338556143996e-05,
+      "loss": 1.2093,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0724165470071365,
+      "learning_rate": 1.64267738046304e-05,
+      "loss": 0.6816,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8188185760040259,
+      "learning_rate": 1.6420157257896848e-05,
+      "loss": 0.6183,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9409397396121741,
+      "learning_rate": 1.6413535926170667e-05,
+      "loss": 0.656,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9566225871520173,
+      "learning_rate": 1.640690981438678e-05,
+      "loss": 0.6446,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9987295823099203,
+      "learning_rate": 1.6400278927483692e-05,
+      "loss": 1.2336,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7510641777483715,
+      "learning_rate": 1.639364327040344e-05,
+      "loss": 0.6011,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.000739384275506,
+      "learning_rate": 1.638700284809164e-05,
+      "loss": 0.6532,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.863312296320136,
+      "learning_rate": 1.6380357665497445e-05,
+      "loss": 0.6135,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8730531455226875,
+      "learning_rate": 1.637370772757356e-05,
+      "loss": 0.5503,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.963759305580031,
+      "learning_rate": 1.6367053039276243e-05,
+      "loss": 0.695,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1357700033377798,
+      "learning_rate": 1.6360393605565278e-05,
+      "loss": 1.1773,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.781178719841549,
+      "learning_rate": 1.635372943140399e-05,
+      "loss": 0.6632,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6706508517575114,
+      "learning_rate": 1.6347060521759243e-05,
+      "loss": 1.2242,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4901576469148536,
+      "learning_rate": 1.6340386881601428e-05,
+      "loss": 1.1659,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0700369387681135,
+      "learning_rate": 1.6333708515904455e-05,
+      "loss": 0.7268,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9112591081336887,
+      "learning_rate": 1.6327025429645763e-05,
+      "loss": 0.5918,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.680552053805452,
+      "learning_rate": 1.6320337627806314e-05,
+      "loss": 1.1892,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.909132163895421,
+      "learning_rate": 1.6313645115370572e-05,
+      "loss": 0.716,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.112333287280215,
+      "learning_rate": 1.6306947897326518e-05,
+      "loss": 0.7027,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8756120587760499,
+      "learning_rate": 1.6300245978665645e-05,
+      "loss": 0.6469,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9713331253247521,
+      "learning_rate": 1.6293539364382938e-05,
+      "loss": 0.6806,
+      "step": 1144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.837138342817252,
+      "learning_rate": 1.6286828059476896e-05,
+      "loss": 0.6349,
+      "step": 1145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.678227938668459,
+      "learning_rate": 1.6280112068949495e-05,
+      "loss": 1.2255,
+      "step": 1146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.419934993921298,
+      "learning_rate": 1.6273391397806226e-05,
+      "loss": 1.0955,
+      "step": 1147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9717595482786918,
+      "learning_rate": 1.626666605105605e-05,
+      "loss": 0.7075,
+      "step": 1148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.891061269296996,
+      "learning_rate": 1.6259936033711422e-05,
+      "loss": 1.2259,
+      "step": 1149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6344571245452544,
+      "learning_rate": 1.6253201350788278e-05,
+      "loss": 1.2204,
+      "step": 1150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0874322800972793,
+      "learning_rate": 1.6246462007306026e-05,
+      "loss": 0.6651,
+      "step": 1151
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.088445137202298,
+      "learning_rate": 1.623971800828755e-05,
+      "loss": 0.6984,
+      "step": 1152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8718631954030598,
+      "learning_rate": 1.6232969358759206e-05,
+      "loss": 0.67,
+      "step": 1153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.124708728954157,
+      "learning_rate": 1.6226216063750814e-05,
+      "loss": 1.2391,
+      "step": 1154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5565617221103576,
+      "learning_rate": 1.6219458128295657e-05,
+      "loss": 1.2105,
+      "step": 1155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7734062722539194,
+      "learning_rate": 1.6212695557430476e-05,
+      "loss": 0.5881,
+      "step": 1156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.145179106679771,
+      "learning_rate": 1.6205928356195467e-05,
+      "loss": 0.8138,
+      "step": 1157
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.845450829675184,
+      "learning_rate": 1.6199156529634278e-05,
+      "loss": 0.6347,
+      "step": 1158
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7016949249362898,
+      "learning_rate": 1.6192380082794e-05,
+      "loss": 0.6174,
+      "step": 1159
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7845508055712902,
+      "learning_rate": 1.6185599020725175e-05,
+      "loss": 0.6537,
+      "step": 1160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.800357160774649,
+      "learning_rate": 1.6178813348481783e-05,
+      "loss": 0.6614,
+      "step": 1161
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.970183961673574,
+      "learning_rate": 1.6172023071121233e-05,
+      "loss": 0.673,
+      "step": 1162
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9565529408002205,
+      "learning_rate": 1.616522819370437e-05,
+      "loss": 0.6733,
+      "step": 1163
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.00830529598555,
+      "learning_rate": 1.615842872129547e-05,
+      "loss": 0.7176,
+      "step": 1164
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.645499926849591,
+      "learning_rate": 1.615162465896224e-05,
+      "loss": 0.6191,
+      "step": 1165
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5680540950065116,
+      "learning_rate": 1.6144816011775786e-05,
+      "loss": 1.1684,
+      "step": 1166
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.344883131870658,
+      "learning_rate": 1.6138002784810655e-05,
+      "loss": 1.1467,
+      "step": 1167
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9059197732886963,
+      "learning_rate": 1.6131184983144798e-05,
+      "loss": 0.6224,
+      "step": 1168
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.601230836075243,
+      "learning_rate": 1.6124362611859566e-05,
+      "loss": 1.1938,
+      "step": 1169
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0774340686280155,
+      "learning_rate": 1.611753567603973e-05,
+      "loss": 0.6884,
+      "step": 1170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8264938927999985,
+      "learning_rate": 1.6110704180773466e-05,
+      "loss": 0.5765,
+      "step": 1171
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.050377463165344,
+      "learning_rate": 1.6103868131152324e-05,
+      "loss": 0.5939,
+      "step": 1172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7382572030630676,
+      "learning_rate": 1.6097027532271278e-05,
+      "loss": 1.2611,
+      "step": 1173
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9036382908005873,
+      "learning_rate": 1.6090182389228673e-05,
+      "loss": 0.7228,
+      "step": 1174
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8196463604231352,
+      "learning_rate": 1.6083332707126245e-05,
+      "loss": 0.6265,
+      "step": 1175
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.370236022316346,
+      "learning_rate": 1.607647849106912e-05,
+      "loss": 1.1887,
+      "step": 1176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3363480256478253,
+      "learning_rate": 1.6069619746165795e-05,
+      "loss": 1.1902,
+      "step": 1177
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0332572723413547,
+      "learning_rate": 1.6062756477528147e-05,
+      "loss": 0.6761,
+      "step": 1178
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3248294511505483,
+      "learning_rate": 1.6055888690271418e-05,
+      "loss": 1.1626,
+      "step": 1179
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.887605765119441,
+      "learning_rate": 1.6049016389514232e-05,
+      "loss": 0.6687,
+      "step": 1180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8643753600855917,
+      "learning_rate": 1.604213958037856e-05,
+      "loss": 0.6497,
+      "step": 1181
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.688040395217157,
+      "learning_rate": 1.6035258267989743e-05,
+      "loss": 1.2083,
+      "step": 1182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.03314390801392,
+      "learning_rate": 1.6028372457476475e-05,
+      "loss": 0.5959,
+      "step": 1183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8530678012888397,
+      "learning_rate": 1.6021482153970803e-05,
+      "loss": 0.6835,
+      "step": 1184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.008336956104643,
+      "learning_rate": 1.601458736260813e-05,
+      "loss": 0.5996,
+      "step": 1185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.860766197228706,
+      "learning_rate": 1.6007688088527187e-05,
+      "loss": 1.2132,
+      "step": 1186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6997289772619304,
+      "learning_rate": 1.6000784336870062e-05,
+      "loss": 0.5432,
+      "step": 1187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.059139384488414,
+      "learning_rate": 1.5993876112782177e-05,
+      "loss": 0.6717,
+      "step": 1188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.402952892367057,
+      "learning_rate": 1.5986963421412278e-05,
+      "loss": 1.1823,
+      "step": 1189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.785756082000142,
+      "learning_rate": 1.598004626791245e-05,
+      "loss": 0.6287,
+      "step": 1190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8159177204948074,
+      "learning_rate": 1.59731246574381e-05,
+      "loss": 0.6535,
+      "step": 1191
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.846670814929946,
+      "learning_rate": 1.596619859514796e-05,
+      "loss": 1.1983,
+      "step": 1192
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.086128984941042,
+      "learning_rate": 1.5959268086204075e-05,
+      "loss": 0.7655,
+      "step": 1193
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1162516267545146,
+      "learning_rate": 1.5952333135771806e-05,
+      "loss": 0.7376,
+      "step": 1194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.033754224635934,
+      "learning_rate": 1.594539374901983e-05,
+      "loss": 0.622,
+      "step": 1195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6388735407002195,
+      "learning_rate": 1.593844993112012e-05,
+      "loss": 1.1866,
+      "step": 1196
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0381981741781625,
+      "learning_rate": 1.5931501687247962e-05,
+      "loss": 0.705,
+      "step": 1197
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4474736924833627,
+      "learning_rate": 1.592454902258193e-05,
+      "loss": 1.2161,
+      "step": 1198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9735451543866953,
+      "learning_rate": 1.59175919423039e-05,
+      "loss": 0.6457,
+      "step": 1199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1834527543973277,
+      "learning_rate": 1.5910630451599036e-05,
+      "loss": 0.7546,
+      "step": 1200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.568892035072757,
+      "learning_rate": 1.59036645556558e-05,
+      "loss": 1.1617,
+      "step": 1201
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7880363098661445,
+      "learning_rate": 1.5896694259665916e-05,
+      "loss": 0.6249,
+      "step": 1202
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.28940131083805,
+      "learning_rate": 1.5889719568824403e-05,
+      "loss": 0.6063,
+      "step": 1203
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9355227204688379,
+      "learning_rate": 1.5882740488329557e-05,
+      "loss": 0.6019,
+      "step": 1204
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1224528989246982,
+      "learning_rate": 1.5875757023382934e-05,
+      "loss": 0.6243,
+      "step": 1205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7846385679112098,
+      "learning_rate": 1.5868769179189367e-05,
+      "loss": 0.695,
+      "step": 1206
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.979546749894294,
+      "learning_rate": 1.586177696095695e-05,
+      "loss": 0.6175,
+      "step": 1207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8513969215210788,
+      "learning_rate": 1.5854780373897038e-05,
+      "loss": 0.5597,
+      "step": 1208
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8677995711243287,
+      "learning_rate": 1.5847779423224237e-05,
+      "loss": 0.719,
+      "step": 1209
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8951233057505164,
+      "learning_rate": 1.5840774114156417e-05,
+      "loss": 0.6346,
+      "step": 1210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.191708896053898,
+      "learning_rate": 1.583376445191468e-05,
+      "loss": 1.1709,
+      "step": 1211
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.718273195442717,
+      "learning_rate": 1.5826750441723385e-05,
+      "loss": 1.127,
+      "step": 1212
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8113693598928455,
+      "learning_rate": 1.5819732088810132e-05,
+      "loss": 0.589,
+      "step": 1213
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2544197303021454,
+      "learning_rate": 1.5812709398405748e-05,
+      "loss": 1.1514,
+      "step": 1214
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.982190315403083,
+      "learning_rate": 1.5805682375744297e-05,
+      "loss": 0.6445,
+      "step": 1215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0810518572070222,
+      "learning_rate": 1.579865102606308e-05,
+      "loss": 0.7437,
+      "step": 1216
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.654167148996302,
+      "learning_rate": 1.5791615354602612e-05,
+      "loss": 1.1178,
+      "step": 1217
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8263125169442196,
+      "learning_rate": 1.578457536660663e-05,
+      "loss": 0.6952,
+      "step": 1218
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.758121502826959,
+      "learning_rate": 1.5777531067322095e-05,
+      "loss": 1.2041,
+      "step": 1219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9757029544344091,
+      "learning_rate": 1.5770482461999178e-05,
+      "loss": 0.6482,
+      "step": 1220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.925715935799873,
+      "learning_rate": 1.576342955589126e-05,
+      "loss": 0.6292,
+      "step": 1221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9007620080184906,
+      "learning_rate": 1.5756372354254922e-05,
+      "loss": 0.7221,
+      "step": 1222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.039632483012965,
+      "learning_rate": 1.5749310862349948e-05,
+      "loss": 0.7159,
+      "step": 1223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6146691087362335,
+      "learning_rate": 1.5742245085439332e-05,
+      "loss": 1.1614,
+      "step": 1224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8422271476579581,
+      "learning_rate": 1.5735175028789247e-05,
+      "loss": 0.5604,
+      "step": 1225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0450759729483643,
+      "learning_rate": 1.5728100697669057e-05,
+      "loss": 0.6197,
+      "step": 1226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9631982805636135,
+      "learning_rate": 1.572102209735132e-05,
+      "loss": 0.5814,
+      "step": 1227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.023702357840359,
+      "learning_rate": 1.5713939233111774e-05,
+      "loss": 0.6834,
+      "step": 1228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.904582688744744,
+      "learning_rate": 1.5706852110229326e-05,
+      "loss": 0.6581,
+      "step": 1229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2355878536687976,
+      "learning_rate": 1.5699760733986066e-05,
+      "loss": 0.6744,
+      "step": 1230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.619476506854909,
+      "learning_rate": 1.569266510966725e-05,
+      "loss": 1.1869,
+      "step": 1231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8016620772311476,
+      "learning_rate": 1.5685565242561303e-05,
+      "loss": 0.6594,
+      "step": 1232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4808877955153283,
+      "learning_rate": 1.567846113795981e-05,
+      "loss": 1.2212,
+      "step": 1233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.880630611539389,
+      "learning_rate": 1.5671352801157517e-05,
+      "loss": 0.6532,
+      "step": 1234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5617260906443255,
+      "learning_rate": 1.566424023745231e-05,
+      "loss": 1.1775,
+      "step": 1235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.107446223790652,
+      "learning_rate": 1.565712345214526e-05,
+      "loss": 0.6008,
+      "step": 1236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8964996815224755,
+      "learning_rate": 1.5650002450540545e-05,
+      "loss": 0.6366,
+      "step": 1237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9378433306303562,
+      "learning_rate": 1.564287723794551e-05,
+      "loss": 0.67,
+      "step": 1238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.638464147174468,
+      "learning_rate": 1.563574781967062e-05,
+      "loss": 0.5701,
+      "step": 1239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.458871180387442,
+      "learning_rate": 1.5628614201029497e-05,
+      "loss": 1.2215,
+      "step": 1240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.374224084811959,
+      "learning_rate": 1.5621476387338877e-05,
+      "loss": 1.2011,
+      "step": 1241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2829497360494964,
+      "learning_rate": 1.561433438391863e-05,
+      "loss": 1.2279,
+      "step": 1242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3182470408276594,
+      "learning_rate": 1.560718819609174e-05,
+      "loss": 1.1623,
+      "step": 1243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.930832235454234,
+      "learning_rate": 1.5600037829184324e-05,
+      "loss": 0.6276,
+      "step": 1244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9579697573456105,
+      "learning_rate": 1.5592883288525603e-05,
+      "loss": 0.6545,
+      "step": 1245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.87462508304546,
+      "learning_rate": 1.558572457944791e-05,
+      "loss": 0.6985,
+      "step": 1246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8827558199610885,
+      "learning_rate": 1.5578561707286684e-05,
+      "loss": 0.6712,
+      "step": 1247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.580148281369055,
+      "learning_rate": 1.557139467738047e-05,
+      "loss": 1.0853,
+      "step": 1248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8158560957891259,
+      "learning_rate": 1.556422349507091e-05,
+      "loss": 0.6605,
+      "step": 1249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.516657911012937,
+      "learning_rate": 1.5557048165702746e-05,
+      "loss": 1.2341,
+      "step": 1250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4263504264270916,
+      "learning_rate": 1.55498686946238e-05,
+      "loss": 1.1882,
+      "step": 1251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0015447616226574,
+      "learning_rate": 1.554268508718499e-05,
+      "loss": 0.7101,
+      "step": 1252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8439286390799616,
+      "learning_rate": 1.553549734874032e-05,
+      "loss": 0.6252,
+      "step": 1253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9380723570956302,
+      "learning_rate": 1.5528305484646854e-05,
+      "loss": 0.6497,
+      "step": 1254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.989744733159473,
+      "learning_rate": 1.552110950026475e-05,
+      "loss": 0.7421,
+      "step": 1255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1393024551387025,
+      "learning_rate": 1.5513909400957234e-05,
+      "loss": 0.6917,
+      "step": 1256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9686544813864546,
+      "learning_rate": 1.550670519209059e-05,
+      "loss": 0.6674,
+      "step": 1257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0183408185147016,
+      "learning_rate": 1.5499496879034176e-05,
+      "loss": 0.6579,
+      "step": 1258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0528023089721907,
+      "learning_rate": 1.5492284467160395e-05,
+      "loss": 0.7542,
+      "step": 1259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8716396685775325,
+      "learning_rate": 1.548506796184472e-05,
+      "loss": 0.7038,
+      "step": 1260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6765505626353323,
+      "learning_rate": 1.547784736846567e-05,
+      "loss": 1.1757,
+      "step": 1261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6535075488731765,
+      "learning_rate": 1.54706226924048e-05,
+      "loss": 1.2175,
+      "step": 1262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6480517729967104,
+      "learning_rate": 1.546339393904672e-05,
+      "loss": 1.1166,
+      "step": 1263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.114904794832528,
+      "learning_rate": 1.5456161113779082e-05,
+      "loss": 0.5733,
+      "step": 1264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3978104430168217,
+      "learning_rate": 1.544892422199256e-05,
+      "loss": 0.7066,
+      "step": 1265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1020883452192276,
+      "learning_rate": 1.5441683269080865e-05,
+      "loss": 0.6376,
+      "step": 1266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4756263775148066,
+      "learning_rate": 1.5434438260440737e-05,
+      "loss": 0.6639,
+      "step": 1267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.235162932109842,
+      "learning_rate": 1.5427189201471933e-05,
+      "loss": 0.6727,
+      "step": 1268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.119054169065409,
+      "learning_rate": 1.5419936097577243e-05,
+      "loss": 0.7017,
+      "step": 1269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9020244750041975,
+      "learning_rate": 1.5412678954162448e-05,
+      "loss": 0.6889,
+      "step": 1270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1721419506484723,
+      "learning_rate": 1.540541777663636e-05,
+      "loss": 0.6591,
+      "step": 1271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.929531344891643,
+      "learning_rate": 1.5398152570410787e-05,
+      "loss": 0.5773,
+      "step": 1272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8261422926307462,
+      "learning_rate": 1.5390883340900548e-05,
+      "loss": 1.198,
+      "step": 1273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1089071063389073,
+      "learning_rate": 1.5383610093523452e-05,
+      "loss": 0.7111,
+      "step": 1274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8325543451686934,
+      "learning_rate": 1.5376332833700308e-05,
+      "loss": 0.6013,
+      "step": 1275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9455139932539496,
+      "learning_rate": 1.5369051566854914e-05,
+      "loss": 0.6271,
+      "step": 1276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.933206113490888,
+      "learning_rate": 1.536176629841405e-05,
+      "loss": 0.6107,
+      "step": 1277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9954133182861893,
+      "learning_rate": 1.535447703380749e-05,
+      "loss": 0.6492,
+      "step": 1278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.124869187807971,
+      "learning_rate": 1.5347183778467973e-05,
+      "loss": 1.1509,
+      "step": 1279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.716587581750191,
+      "learning_rate": 1.533988653783122e-05,
+      "loss": 0.6306,
+      "step": 1280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8531000433786495,
+      "learning_rate": 1.533258531733592e-05,
+      "loss": 0.6477,
+      "step": 1281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9880912301875615,
+      "learning_rate": 1.532528012242373e-05,
+      "loss": 0.7322,
+      "step": 1282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.691633006365506,
+      "learning_rate": 1.5317970958539267e-05,
+      "loss": 1.1768,
+      "step": 1283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.932913532231692,
+      "learning_rate": 1.5310657831130112e-05,
+      "loss": 0.6446,
+      "step": 1284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5615145570504625,
+      "learning_rate": 1.530334074564679e-05,
+      "loss": 1.1911,
+      "step": 1285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2753481549195285,
+      "learning_rate": 1.5296019707542778e-05,
+      "loss": 1.1336,
+      "step": 1286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0630931422221237,
+      "learning_rate": 1.528869472227451e-05,
+      "loss": 0.6568,
+      "step": 1287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1206915796826173,
+      "learning_rate": 1.5281365795301352e-05,
+      "loss": 0.6349,
+      "step": 1288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8803882555464457,
+      "learning_rate": 1.5274032932085603e-05,
+      "loss": 0.6641,
+      "step": 1289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1910458952715075,
+      "learning_rate": 1.526669613809251e-05,
+      "loss": 1.1961,
+      "step": 1290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.851523406673763,
+      "learning_rate": 1.5259355418790235e-05,
+      "loss": 0.6825,
+      "step": 1291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8233400628016356,
+      "learning_rate": 1.5252010779649884e-05,
+      "loss": 0.6614,
+      "step": 1292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.104839759285013,
+      "learning_rate": 1.5244662226145463e-05,
+      "loss": 0.6992,
+      "step": 1293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8692133361515693,
+      "learning_rate": 1.5237309763753906e-05,
+      "loss": 0.6267,
+      "step": 1294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8215158942945078,
+      "learning_rate": 1.5229953397955064e-05,
+      "loss": 0.632,
+      "step": 1295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.546471829616648,
+      "learning_rate": 1.5222593134231693e-05,
+      "loss": 1.0791,
+      "step": 1296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9057555282854737,
+      "learning_rate": 1.5215228978069452e-05,
+      "loss": 0.6525,
+      "step": 1297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0397036489509333,
+      "learning_rate": 1.5207860934956901e-05,
+      "loss": 0.7068,
+      "step": 1298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4872307026159532,
+      "learning_rate": 1.5200489010385506e-05,
+      "loss": 1.1706,
+      "step": 1299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.286915488237803,
+      "learning_rate": 1.519311320984961e-05,
+      "loss": 0.6915,
+      "step": 1300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6606973973742494,
+      "learning_rate": 1.5185733538846465e-05,
+      "loss": 1.1965,
+      "step": 1301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9582453229286043,
+      "learning_rate": 1.5178350002876187e-05,
+      "loss": 0.6917,
+      "step": 1302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7949431153879902,
+      "learning_rate": 1.5170962607441787e-05,
+      "loss": 0.6043,
+      "step": 1303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7077528332112657,
+      "learning_rate": 1.5163571358049149e-05,
+      "loss": 0.6141,
+      "step": 1304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7884161466462332,
+      "learning_rate": 1.5156176260207026e-05,
+      "loss": 0.6035,
+      "step": 1305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9162638809662482,
+      "learning_rate": 1.514877731942704e-05,
+      "loss": 0.605,
+      "step": 1306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9206252183026107,
+      "learning_rate": 1.5141374541223676e-05,
+      "loss": 0.6381,
+      "step": 1307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9761748956750305,
+      "learning_rate": 1.5133967931114289e-05,
+      "loss": 0.6047,
+      "step": 1308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9346873844263712,
+      "learning_rate": 1.5126557494619078e-05,
+      "loss": 0.7078,
+      "step": 1309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7603419623469505,
+      "learning_rate": 1.51191432372611e-05,
+      "loss": 1.1521,
+      "step": 1310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.920021672965111,
+      "learning_rate": 1.5111725164566255e-05,
+      "loss": 0.6446,
+      "step": 1311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.587807282364832,
+      "learning_rate": 1.5104303282063295e-05,
+      "loss": 1.2093,
+      "step": 1312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0427094586151666,
+      "learning_rate": 1.5096877595283803e-05,
+      "loss": 0.6841,
+      "step": 1313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2167943421820575,
+      "learning_rate": 1.5089448109762196e-05,
+      "loss": 0.6903,
+      "step": 1314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6193512852618555,
+      "learning_rate": 1.5082014831035735e-05,
+      "loss": 1.142,
+      "step": 1315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5457454844646685,
+      "learning_rate": 1.5074577764644499e-05,
+      "loss": 1.1873,
+      "step": 1316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9361423241569942,
+      "learning_rate": 1.506713691613138e-05,
+      "loss": 0.7106,
+      "step": 1317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8832148541649985,
+      "learning_rate": 1.5059692291042114e-05,
+      "loss": 0.5799,
+      "step": 1318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0235582484641896,
+      "learning_rate": 1.505224389492523e-05,
+      "loss": 0.6176,
+      "step": 1319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9618478064802403,
+      "learning_rate": 1.5044791733332073e-05,
+      "loss": 0.5925,
+      "step": 1320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.11401070264941,
+      "learning_rate": 1.5037335811816805e-05,
+      "loss": 0.6538,
+      "step": 1321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0898651285261707,
+      "learning_rate": 1.5029876135936375e-05,
+      "loss": 0.7329,
+      "step": 1322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9203765340232692,
+      "learning_rate": 1.5022412711250533e-05,
+      "loss": 0.6984,
+      "step": 1323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.060582055811385,
+      "learning_rate": 1.5014945543321834e-05,
+      "loss": 0.6293,
+      "step": 1324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7218869501322795,
+      "learning_rate": 1.5007474637715618e-05,
+      "loss": 0.5758,
+      "step": 1325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6705765878949124,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1883,
+      "step": 1326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0755166369181124,
+      "learning_rate": 1.4992521635745891e-05,
+      "loss": 0.6549,
+      "step": 1327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8383250097780433,
+      "learning_rate": 1.4985039550526972e-05,
+      "loss": 0.5629,
+      "step": 1328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0908633979092404,
+      "learning_rate": 1.4977553749919698e-05,
+      "loss": 0.6328,
+      "step": 1329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0132790418869906,
+      "learning_rate": 1.4970064239503295e-05,
+      "loss": 0.6294,
+      "step": 1330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.890961927497465,
+      "learning_rate": 1.4962571024859748e-05,
+      "loss": 0.6302,
+      "step": 1331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8729468372145726,
+      "learning_rate": 1.4955074111573815e-05,
+      "loss": 1.1847,
+      "step": 1332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0319366802944363,
+      "learning_rate": 1.4947573505233002e-05,
+      "loss": 0.6469,
+      "step": 1333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8887467613384779,
+      "learning_rate": 1.4940069211427562e-05,
+      "loss": 0.6585,
+      "step": 1334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.783442714137494,
+      "learning_rate": 1.493256123575051e-05,
+      "loss": 1.253,
+      "step": 1335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9900674094662578,
+      "learning_rate": 1.492504958379759e-05,
+      "loss": 0.661,
+      "step": 1336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8420811777321198,
+      "learning_rate": 1.4917534261167306e-05,
+      "loss": 0.597,
+      "step": 1337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1954636226411752,
+      "learning_rate": 1.4910015273460876e-05,
+      "loss": 0.6796,
+      "step": 1338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6971947390219468,
+      "learning_rate": 1.4902492626282261e-05,
+      "loss": 0.591,
+      "step": 1339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1141169458117037,
+      "learning_rate": 1.4894966325238154e-05,
+      "loss": 0.6183,
+      "step": 1340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8219739821646626,
+      "learning_rate": 1.4887436375937958e-05,
+      "loss": 0.6352,
+      "step": 1341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.93880565752181,
+      "learning_rate": 1.4879902783993804e-05,
+      "loss": 0.6219,
+      "step": 1342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8924608839532704,
+      "learning_rate": 1.4872365555020539e-05,
+      "loss": 0.6961,
+      "step": 1343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2049454255193632,
+      "learning_rate": 1.4864824694635714e-05,
+      "loss": 1.1786,
+      "step": 1344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.876430021272057,
+      "learning_rate": 1.4857280208459595e-05,
+      "loss": 0.5812,
+      "step": 1345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2637435339068532,
+      "learning_rate": 1.484973210211514e-05,
+      "loss": 0.6822,
+      "step": 1346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.856433743367476,
+      "learning_rate": 1.4842180381228013e-05,
+      "loss": 0.7104,
+      "step": 1347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.558256108271051,
+      "learning_rate": 1.483462505142657e-05,
+      "loss": 1.1779,
+      "step": 1348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8655810738823522,
+      "learning_rate": 1.4827066118341857e-05,
+      "loss": 0.6209,
+      "step": 1349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.819893019878148,
+      "learning_rate": 1.4819503587607607e-05,
+      "loss": 0.7046,
+      "step": 1350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9544936688630556,
+      "learning_rate": 1.4811937464860227e-05,
+      "loss": 0.5755,
+      "step": 1351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8258283034185385,
+      "learning_rate": 1.480436775573881e-05,
+      "loss": 0.6198,
+      "step": 1352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.800033384439128,
+      "learning_rate": 1.4796794465885116e-05,
+      "loss": 0.6012,
+      "step": 1353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.582444809685742,
+      "learning_rate": 1.478921760094358e-05,
+      "loss": 1.18,
+      "step": 1354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5751335220675404,
+      "learning_rate": 1.4781637166561296e-05,
+      "loss": 1.1699,
+      "step": 1355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4654162750661253,
+      "learning_rate": 1.477405316838802e-05,
+      "loss": 1.1521,
+      "step": 1356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9209861865826452,
+      "learning_rate": 1.4766465612076166e-05,
+      "loss": 0.6171,
+      "step": 1357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9602923004131219,
+      "learning_rate": 1.4758874503280796e-05,
+      "loss": 0.6385,
+      "step": 1358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.907324718267176,
+      "learning_rate": 1.475127984765962e-05,
+      "loss": 0.5967,
+      "step": 1359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8587414634118458,
+      "learning_rate": 1.4743681650872995e-05,
+      "loss": 0.6687,
+      "step": 1360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9965673490436981,
+      "learning_rate": 1.473607991858392e-05,
+      "loss": 0.6909,
+      "step": 1361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.866631922983096,
+      "learning_rate": 1.4728474656458019e-05,
+      "loss": 0.6078,
+      "step": 1362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8260841578447402,
+      "learning_rate": 1.472086587016355e-05,
+      "loss": 0.6349,
+      "step": 1363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7402225575984753,
+      "learning_rate": 1.4713253565371409e-05,
+      "loss": 0.6019,
+      "step": 1364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.453537330636166,
+      "learning_rate": 1.4705637747755099e-05,
+      "loss": 1.1757,
+      "step": 1365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7951817559604568,
+      "learning_rate": 1.4698018422990749e-05,
+      "loss": 0.5959,
+      "step": 1366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.928874994176511,
+      "learning_rate": 1.4690395596757099e-05,
+      "loss": 0.6019,
+      "step": 1367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8961025480071394,
+      "learning_rate": 1.4682769274735504e-05,
+      "loss": 0.6915,
+      "step": 1368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8081336510199282,
+      "learning_rate": 1.4675139462609917e-05,
+      "loss": 0.6524,
+      "step": 1369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8431513727798639,
+      "learning_rate": 1.4667506166066896e-05,
+      "loss": 0.5924,
+      "step": 1370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0294257297382168,
+      "learning_rate": 1.4659869390795596e-05,
+      "loss": 1.1794,
+      "step": 1371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.052928418472855,
+      "learning_rate": 1.4652229142487767e-05,
+      "loss": 0.6708,
+      "step": 1372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9371510394515525,
+      "learning_rate": 1.464458542683774e-05,
+      "loss": 0.595,
+      "step": 1373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.906980283538782,
+      "learning_rate": 1.4636938249542438e-05,
+      "loss": 0.6279,
+      "step": 1374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.361536851220504,
+      "learning_rate": 1.4629287616301362e-05,
+      "loss": 1.1726,
+      "step": 1375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.796082876417057,
+      "learning_rate": 1.4621633532816586e-05,
+      "loss": 0.6309,
+      "step": 1376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9025123583921233,
+      "learning_rate": 1.461397600479276e-05,
+      "loss": 0.6467,
+      "step": 1377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.779431020048955,
+      "learning_rate": 1.4606315037937095e-05,
+      "loss": 0.6831,
+      "step": 1378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.817792272591361,
+      "learning_rate": 1.459865063795937e-05,
+      "loss": 1.1953,
+      "step": 1379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.587914891711827,
+      "learning_rate": 1.4590982810571923e-05,
+      "loss": 1.1562,
+      "step": 1380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7026070807952196,
+      "learning_rate": 1.4583311561489643e-05,
+      "loss": 0.5734,
+      "step": 1381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6610593942167617,
+      "learning_rate": 1.457563689642997e-05,
+      "loss": 0.6349,
+      "step": 1382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.503920814682198,
+      "learning_rate": 1.4567958821112892e-05,
+      "loss": 1.1697,
+      "step": 1383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.386901193274087,
+      "learning_rate": 1.456027734126094e-05,
+      "loss": 1.2069,
+      "step": 1384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.47655068403206,
+      "learning_rate": 1.4552592462599177e-05,
+      "loss": 1.1305,
+      "step": 1385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9315506389645816,
+      "learning_rate": 1.4544904190855205e-05,
+      "loss": 0.6343,
+      "step": 1386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9085744332910997,
+      "learning_rate": 1.4537212531759146e-05,
+      "loss": 0.6063,
+      "step": 1387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4605558372556757,
+      "learning_rate": 1.4529517491043658e-05,
+      "loss": 1.205,
+      "step": 1388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8082718221448293,
+      "learning_rate": 1.4521819074443917e-05,
+      "loss": 0.57,
+      "step": 1389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4789066583377584,
+      "learning_rate": 1.4514117287697604e-05,
+      "loss": 1.1238,
+      "step": 1390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9399398700004153,
+      "learning_rate": 1.4506412136544926e-05,
+      "loss": 0.7239,
+      "step": 1391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6640126569867677,
+      "learning_rate": 1.4498703626728589e-05,
+      "loss": 0.5885,
+      "step": 1392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.687053523457553,
+      "learning_rate": 1.4490991763993809e-05,
+      "loss": 1.2181,
+      "step": 1393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9898429803713182,
+      "learning_rate": 1.448327655408829e-05,
+      "loss": 0.6213,
+      "step": 1394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.166208641220349,
+      "learning_rate": 1.447555800276224e-05,
+      "loss": 0.7036,
+      "step": 1395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7838767118887477,
+      "learning_rate": 1.4467836115768356e-05,
+      "loss": 0.7174,
+      "step": 1396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9099185150073636,
+      "learning_rate": 1.446011089886182e-05,
+      "loss": 0.6147,
+      "step": 1397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.29201599194328,
+      "learning_rate": 1.4452382357800296e-05,
+      "loss": 1.2303,
+      "step": 1398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.956476985514293,
+      "learning_rate": 1.4444650498343919e-05,
+      "loss": 0.6851,
+      "step": 1399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7660553403640515,
+      "learning_rate": 1.4436915326255312e-05,
+      "loss": 0.6087,
+      "step": 1400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5965847194033203,
+      "learning_rate": 1.4429176847299556e-05,
+      "loss": 0.6253,
+      "step": 1401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6360470289327567,
+      "learning_rate": 1.4421435067244194e-05,
+      "loss": 0.6167,
+      "step": 1402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7421350968722986,
+      "learning_rate": 1.4413689991859237e-05,
+      "loss": 0.6641,
+      "step": 1403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7298787194142147,
+      "learning_rate": 1.4405941626917152e-05,
+      "loss": 1.2374,
+      "step": 1404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.927545280045091,
+      "learning_rate": 1.4398189978192854e-05,
+      "loss": 0.6697,
+      "step": 1405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.51592970594387,
+      "learning_rate": 1.4390435051463705e-05,
+      "loss": 1.1551,
+      "step": 1406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9359005609138384,
+      "learning_rate": 1.438267685250951e-05,
+      "loss": 0.6873,
+      "step": 1407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0273509142125334,
+      "learning_rate": 1.4374915387112518e-05,
+      "loss": 0.6953,
+      "step": 1408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0502334813962473,
+      "learning_rate": 1.4367150661057406e-05,
+      "loss": 0.6767,
+      "step": 1409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5504043309649314,
+      "learning_rate": 1.4359382680131286e-05,
+      "loss": 1.2114,
+      "step": 1410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.258186654820411,
+      "learning_rate": 1.4351611450123693e-05,
+      "loss": 1.1247,
+      "step": 1411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.929667115071505,
+      "learning_rate": 1.4343836976826589e-05,
+      "loss": 0.6491,
+      "step": 1412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8630538754397956,
+      "learning_rate": 1.433605926603434e-05,
+      "loss": 0.5808,
+      "step": 1413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.998636724992996,
+      "learning_rate": 1.4328278323543743e-05,
+      "loss": 0.6989,
+      "step": 1414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7977707134796688,
+      "learning_rate": 1.4320494155153986e-05,
+      "loss": 0.5567,
+      "step": 1415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9868508821190372,
+      "learning_rate": 1.4312706766666675e-05,
+      "loss": 0.6884,
+      "step": 1416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9061632794071048,
+      "learning_rate": 1.4304916163885811e-05,
+      "loss": 0.6921,
+      "step": 1417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9828395289561913,
+      "learning_rate": 1.4297122352617787e-05,
+      "loss": 0.6523,
+      "step": 1418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9834594859046215,
+      "learning_rate": 1.4289325338671391e-05,
+      "loss": 0.6209,
+      "step": 1419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7468719224774443,
+      "learning_rate": 1.4281525127857801e-05,
+      "loss": 0.5604,
+      "step": 1420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0720334528690048,
+      "learning_rate": 1.427372172599057e-05,
+      "loss": 1.1596,
+      "step": 1421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7736062992467874,
+      "learning_rate": 1.4265915138885635e-05,
+      "loss": 1.1834,
+      "step": 1422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9397206152735156,
+      "learning_rate": 1.4258105372361306e-05,
+      "loss": 0.7043,
+      "step": 1423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.848253194138749,
+      "learning_rate": 1.425029243223826e-05,
+      "loss": 0.6524,
+      "step": 1424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.597407187475036,
+      "learning_rate": 1.4242476324339545e-05,
+      "loss": 1.1421,
+      "step": 1425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.083999787971743,
+      "learning_rate": 1.4234657054490563e-05,
+      "loss": 0.7158,
+      "step": 1426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8009995835612698,
+      "learning_rate": 1.422683462851908e-05,
+      "loss": 0.55,
+      "step": 1427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.810061110930546,
+      "learning_rate": 1.4219009052255204e-05,
+      "loss": 0.5738,
+      "step": 1428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9491462361107377,
+      "learning_rate": 1.4211180331531405e-05,
+      "loss": 0.6739,
+      "step": 1429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.892625001326539,
+      "learning_rate": 1.4203348472182484e-05,
+      "loss": 0.5706,
+      "step": 1430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7968720245548977,
+      "learning_rate": 1.4195513480045584e-05,
+      "loss": 0.6212,
+      "step": 1431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.902734374402517,
+      "learning_rate": 1.4187675360960192e-05,
+      "loss": 0.6709,
+      "step": 1432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.073565574047457,
+      "learning_rate": 1.4179834120768112e-05,
+      "loss": 0.6099,
+      "step": 1433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8149691298821087,
+      "learning_rate": 1.4171989765313486e-05,
+      "loss": 0.6137,
+      "step": 1434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7931428831307896,
+      "learning_rate": 1.4164142300442765e-05,
+      "loss": 0.5942,
+      "step": 1435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7747494792224645,
+      "learning_rate": 1.4156291732004732e-05,
+      "loss": 0.6293,
+      "step": 1436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.821101551737612,
+      "learning_rate": 1.4148438065850475e-05,
+      "loss": 0.6382,
+      "step": 1437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9222744593414138,
+      "learning_rate": 1.414058130783339e-05,
+      "loss": 0.6402,
+      "step": 1438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8570319917499765,
+      "learning_rate": 1.4132721463809179e-05,
+      "loss": 0.6876,
+      "step": 1439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9352745668715186,
+      "learning_rate": 1.4124858539635842e-05,
+      "loss": 0.6184,
+      "step": 1440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8152625488148155,
+      "learning_rate": 1.4116992541173684e-05,
+      "loss": 0.67,
+      "step": 1441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9235605119426884,
+      "learning_rate": 1.4109123474285289e-05,
+      "loss": 0.5618,
+      "step": 1442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5937245846199763,
+      "learning_rate": 1.4101251344835533e-05,
+      "loss": 1.1563,
+      "step": 1443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7613306143687948,
+      "learning_rate": 1.409337615869158e-05,
+      "loss": 0.6034,
+      "step": 1444
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9597113567659927,
+      "learning_rate": 1.4085497921722857e-05,
+      "loss": 0.696,
+      "step": 1445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7760372212239337,
+      "learning_rate": 1.4077616639801084e-05,
+      "loss": 0.6153,
+      "step": 1446
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8907760409838286,
+      "learning_rate": 1.4069732318800238e-05,
+      "loss": 0.6874,
+      "step": 1447
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3513434358135097,
+      "learning_rate": 1.4061844964596563e-05,
+      "loss": 1.1632,
+      "step": 1448
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.49004359041212,
+      "learning_rate": 1.4053954583068565e-05,
+      "loss": 1.1289,
+      "step": 1449
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.694390333129094,
+      "learning_rate": 1.4046061180097007e-05,
+      "loss": 0.6078,
+      "step": 1450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0323538767121594,
+      "learning_rate": 1.4038164761564902e-05,
+      "loss": 0.6817,
+      "step": 1451
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2669929059757674,
+      "learning_rate": 1.4030265333357514e-05,
+      "loss": 1.1475,
+      "step": 1452
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.417243285174133,
+      "learning_rate": 1.4022362901362343e-05,
+      "loss": 1.2096,
+      "step": 1453
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.490147835581923,
+      "learning_rate": 1.4014457471469138e-05,
+      "loss": 1.1318,
+      "step": 1454
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8498462486019858,
+      "learning_rate": 1.4006549049569874e-05,
+      "loss": 0.6239,
+      "step": 1455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9334868319152607,
+      "learning_rate": 1.3998637641558754e-05,
+      "loss": 0.6264,
+      "step": 1456
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8038775235670377,
+      "learning_rate": 1.399072325333222e-05,
+      "loss": 0.7209,
+      "step": 1457
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9846387383005195,
+      "learning_rate": 1.3982805890788924e-05,
+      "loss": 0.6315,
+      "step": 1458
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0504866556201544,
+      "learning_rate": 1.3974885559829734e-05,
+      "loss": 0.6858,
+      "step": 1459
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.579709778733863,
+      "learning_rate": 1.3966962266357733e-05,
+      "loss": 1.196,
+      "step": 1460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.796428439356534,
+      "learning_rate": 1.3959036016278218e-05,
+      "loss": 1.1284,
+      "step": 1461
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7619814787356847,
+      "learning_rate": 1.3951106815498679e-05,
+      "loss": 0.5762,
+      "step": 1462
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.46929772227998,
+      "learning_rate": 1.3943174669928813e-05,
+      "loss": 1.1866,
+      "step": 1463
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3068008350470026,
+      "learning_rate": 1.3935239585480508e-05,
+      "loss": 1.0907,
+      "step": 1464
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.360120898650647,
+      "learning_rate": 1.3927301568067848e-05,
+      "loss": 1.1813,
+      "step": 1465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9011669811498662,
+      "learning_rate": 1.3919360623607094e-05,
+      "loss": 0.6542,
+      "step": 1466
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8359976327039766,
+      "learning_rate": 1.3911416758016692e-05,
+      "loss": 0.6234,
+      "step": 1467
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9137645572386277,
+      "learning_rate": 1.3903469977217268e-05,
+      "loss": 0.6421,
+      "step": 1468
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8653255499243389,
+      "learning_rate": 1.3895520287131619e-05,
+      "loss": 0.6727,
+      "step": 1469
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0139097855275376,
+      "learning_rate": 1.3887567693684712e-05,
+      "loss": 0.6487,
+      "step": 1470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9288860076894696,
+      "learning_rate": 1.3879612202803672e-05,
+      "loss": 0.7302,
+      "step": 1471
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6624526240153183,
+      "learning_rate": 1.3871653820417787e-05,
+      "loss": 0.5752,
+      "step": 1472
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8979899710492734,
+      "learning_rate": 1.3863692552458509e-05,
+      "loss": 0.747,
+      "step": 1473
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7800189005217426,
+      "learning_rate": 1.3855728404859426e-05,
+      "loss": 1.1693,
+      "step": 1474
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5999525620993533,
+      "learning_rate": 1.3847761383556278e-05,
+      "loss": 1.2059,
+      "step": 1475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3868146819247076,
+      "learning_rate": 1.3839791494486946e-05,
+      "loss": 1.2095,
+      "step": 1476
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9308712637622167,
+      "learning_rate": 1.3831818743591457e-05,
+      "loss": 0.6545,
+      "step": 1477
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.041927955167395,
+      "learning_rate": 1.3823843136811956e-05,
+      "loss": 0.6739,
+      "step": 1478
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9988046051643718,
+      "learning_rate": 1.3815864680092727e-05,
+      "loss": 0.5681,
+      "step": 1479
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7697973002652976,
+      "learning_rate": 1.3807883379380176e-05,
+      "loss": 1.1845,
+      "step": 1480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.557628629781112,
+      "learning_rate": 1.3799899240622828e-05,
+      "loss": 1.1909,
+      "step": 1481
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8509901506686324,
+      "learning_rate": 1.3791912269771319e-05,
+      "loss": 0.6811,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9103153963517405,
+      "learning_rate": 1.3783922472778408e-05,
+      "loss": 0.649,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.981335643873522,
+      "learning_rate": 1.3775929855598941e-05,
+      "loss": 0.6396,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8229701581106694,
+      "learning_rate": 1.376793442418989e-05,
+      "loss": 0.5977,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.82570484471643,
+      "learning_rate": 1.3759936184510304e-05,
+      "loss": 0.611,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8726350099900737,
+      "learning_rate": 1.375193514252133e-05,
+      "loss": 0.5984,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9414846107735775,
+      "learning_rate": 1.3743931304186212e-05,
+      "loss": 0.6905,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.868775683696041,
+      "learning_rate": 1.3735924675470268e-05,
+      "loss": 0.609,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.724690697985275,
+      "learning_rate": 1.3727915262340905e-05,
+      "loss": 1.1879,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5175558656611647,
+      "learning_rate": 1.3719903070767597e-05,
+      "loss": 1.0933,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7170509974283932,
+      "learning_rate": 1.3711888106721891e-05,
+      "loss": 0.6172,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9961843591525896,
+      "learning_rate": 1.3703870376177404e-05,
+      "loss": 0.6134,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0033782667908953,
+      "learning_rate": 1.3695849885109812e-05,
+      "loss": 0.6526,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1685874910397076,
+      "learning_rate": 1.3687826639496847e-05,
+      "loss": 0.6289,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.999392841606343,
+      "learning_rate": 1.3679800645318293e-05,
+      "loss": 0.613,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7480273401614363,
+      "learning_rate": 1.3671771908555994e-05,
+      "loss": 1.1184,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.766935644200237,
+      "learning_rate": 1.366374043519382e-05,
+      "loss": 0.6114,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0502444390381553,
+      "learning_rate": 1.3655706231217697e-05,
+      "loss": 0.699,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9739904983775336,
+      "learning_rate": 1.3647669302615573e-05,
+      "loss": 0.6326,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9171164722209637,
+      "learning_rate": 1.363962965537744e-05,
+      "loss": 0.6043,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.753608971537807,
+      "learning_rate": 1.3631587295495305e-05,
+      "loss": 1.1351,
+      "step": 1502
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8672076529707509,
+      "learning_rate": 1.36235422289632e-05,
+      "loss": 0.5426,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.850331667726579,
+      "learning_rate": 1.3615494461777173e-05,
+      "loss": 0.5433,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6899775378009805,
+      "learning_rate": 1.3607443999935293e-05,
+      "loss": 0.5903,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.426964835209477,
+      "learning_rate": 1.3599390849437626e-05,
+      "loss": 1.148,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8607226786154634,
+      "learning_rate": 1.3591335016286249e-05,
+      "loss": 0.6065,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.103950520521851,
+      "learning_rate": 1.3583276506485235e-05,
+      "loss": 0.5924,
+      "step": 1508
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7955877167503913,
+      "learning_rate": 1.3575215326040657e-05,
+      "loss": 0.6126,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.062590475276482,
+      "learning_rate": 1.3567151480960571e-05,
+      "loss": 1.222,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.834797277102248,
+      "learning_rate": 1.3559084977255027e-05,
+      "loss": 0.6364,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.258117594481597,
+      "learning_rate": 1.3551015820936044e-05,
+      "loss": 0.6325,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4891185305296326,
+      "learning_rate": 1.3542944018017637e-05,
+      "loss": 1.1688,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8445913713406838,
+      "learning_rate": 1.3534869574515777e-05,
+      "loss": 0.6118,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6219636301028175,
+      "learning_rate": 1.3526792496448406e-05,
+      "loss": 1.1394,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5346359089441757,
+      "learning_rate": 1.3518712789835439e-05,
+      "loss": 1.1616,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.933300776872874,
+      "learning_rate": 1.351063046069874e-05,
+      "loss": 0.5749,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6353217316212827,
+      "learning_rate": 1.3502545515062132e-05,
+      "loss": 0.5796,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0051690887384197,
+      "learning_rate": 1.349445795895139e-05,
+      "loss": 0.6674,
+      "step": 1519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7420215236701546,
+      "learning_rate": 1.3486367798394225e-05,
+      "loss": 0.6078,
+      "step": 1520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.62143565430264,
+      "learning_rate": 1.3478275039420301e-05,
+      "loss": 1.0722,
+      "step": 1521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5535273620349455,
+      "learning_rate": 1.3470179688061214e-05,
+      "loss": 1.1821,
+      "step": 1522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.525056607782132,
+      "learning_rate": 1.3462081750350488e-05,
+      "loss": 1.1518,
+      "step": 1523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9234615392293188,
+      "learning_rate": 1.345398123232358e-05,
+      "loss": 0.6216,
+      "step": 1524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0874538940122678,
+      "learning_rate": 1.3445878140017871e-05,
+      "loss": 0.5699,
+      "step": 1525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6358991455637624,
+      "learning_rate": 1.3437772479472655e-05,
+      "loss": 0.6315,
+      "step": 1526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.047471141467564,
+      "learning_rate": 1.3429664256729145e-05,
+      "loss": 0.5498,
+      "step": 1527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7110999897748047,
+      "learning_rate": 1.342155347783046e-05,
+      "loss": 0.6087,
+      "step": 1528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8188029787375344,
+      "learning_rate": 1.3413440148821626e-05,
+      "loss": 0.6014,
+      "step": 1529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8199084207945702,
+      "learning_rate": 1.3405324275749574e-05,
+      "loss": 0.588,
+      "step": 1530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0591876054129044,
+      "learning_rate": 1.3397205864663121e-05,
+      "loss": 0.6355,
+      "step": 1531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.413584674558013,
+      "learning_rate": 1.338908492161298e-05,
+      "loss": 1.1252,
+      "step": 1532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8147538036283768,
+      "learning_rate": 1.3380961452651762e-05,
+      "loss": 0.5918,
+      "step": 1533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6336679417249296,
+      "learning_rate": 1.337283546383394e-05,
+      "loss": 0.6039,
+      "step": 1534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2020984090122577,
+      "learning_rate": 1.3364706961215883e-05,
+      "loss": 1.1208,
+      "step": 1535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.864707652248465,
+      "learning_rate": 1.335657595085582e-05,
+      "loss": 0.6018,
+      "step": 1536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9245880286604424,
+      "learning_rate": 1.3348442438813867e-05,
+      "loss": 0.6747,
+      "step": 1537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4162470145806334,
+      "learning_rate": 1.3340306431151984e-05,
+      "loss": 1.1506,
+      "step": 1538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.729400023727809,
+      "learning_rate": 1.3332167933934002e-05,
+      "loss": 0.7008,
+      "step": 1539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8378616332088342,
+      "learning_rate": 1.3324026953225604e-05,
+      "loss": 0.5951,
+      "step": 1540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7730261860869387,
+      "learning_rate": 1.3315883495094331e-05,
+      "loss": 0.647,
+      "step": 1541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9021731273667999,
+      "learning_rate": 1.330773756560956e-05,
+      "loss": 0.6321,
+      "step": 1542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.838724375271201,
+      "learning_rate": 1.3299589170842517e-05,
+      "loss": 0.639,
+      "step": 1543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0970206773343225,
+      "learning_rate": 1.329143831686626e-05,
+      "loss": 0.719,
+      "step": 1544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.135766768444061,
+      "learning_rate": 1.328328500975569e-05,
+      "loss": 0.6338,
+      "step": 1545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8201847464965415,
+      "learning_rate": 1.3275129255587525e-05,
+      "loss": 0.5983,
+      "step": 1546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7723485505272496,
+      "learning_rate": 1.326697106044031e-05,
+      "loss": 0.5538,
+      "step": 1547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8784750713299492,
+      "learning_rate": 1.3258810430394413e-05,
+      "loss": 0.6316,
+      "step": 1548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7940501457135525,
+      "learning_rate": 1.3250647371532015e-05,
+      "loss": 0.5723,
+      "step": 1549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7563699104153767,
+      "learning_rate": 1.3242481889937106e-05,
+      "loss": 1.1628,
+      "step": 1550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.903990782726359,
+      "learning_rate": 1.323431399169548e-05,
+      "loss": 0.5957,
+      "step": 1551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9242787194004414,
+      "learning_rate": 1.3226143682894738e-05,
+      "loss": 0.6642,
+      "step": 1552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6272621821606381,
+      "learning_rate": 1.3217970969624272e-05,
+      "loss": 0.5392,
+      "step": 1553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.859701783887628,
+      "learning_rate": 1.320979585797527e-05,
+      "loss": 0.6009,
+      "step": 1554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4557137081259377,
+      "learning_rate": 1.3201618354040702e-05,
+      "loss": 1.1871,
+      "step": 1555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9820761218559604,
+      "learning_rate": 1.3193438463915328e-05,
+      "loss": 0.6768,
+      "step": 1556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8040598599212574,
+      "learning_rate": 1.3185256193695681e-05,
+      "loss": 1.1224,
+      "step": 1557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3064230344057584,
+      "learning_rate": 1.3177071549480073e-05,
+      "loss": 0.6697,
+      "step": 1558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3186366328423134,
+      "learning_rate": 1.316888453736858e-05,
+      "loss": 1.1409,
+      "step": 1559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4859642696303967,
+      "learning_rate": 1.3160695163463045e-05,
+      "loss": 1.1992,
+      "step": 1560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.554305239058523,
+      "learning_rate": 1.315250343386707e-05,
+      "loss": 1.1631,
+      "step": 1561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.013652140403078,
+      "learning_rate": 1.314430935468602e-05,
+      "loss": 0.6041,
+      "step": 1562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.881118097801502,
+      "learning_rate": 1.3136112932026999e-05,
+      "loss": 1.1794,
+      "step": 1563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.049559810324001,
+      "learning_rate": 1.3127914171998863e-05,
+      "loss": 0.649,
+      "step": 1564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.957054671828374,
+      "learning_rate": 1.3119713080712219e-05,
+      "loss": 0.686,
+      "step": 1565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.024136970682176,
+      "learning_rate": 1.3111509664279398e-05,
+      "loss": 0.6389,
+      "step": 1566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8758264479932514,
+      "learning_rate": 1.310330392881447e-05,
+      "loss": 0.6889,
+      "step": 1567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9030095673543577,
+      "learning_rate": 1.3095095880433229e-05,
+      "loss": 0.686,
+      "step": 1568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3112823565499876,
+      "learning_rate": 1.3086885525253203e-05,
+      "loss": 1.2274,
+      "step": 1569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.784945282332013,
+      "learning_rate": 1.3078672869393632e-05,
+      "loss": 0.6318,
+      "step": 1570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.479996292039343,
+      "learning_rate": 1.3070457918975467e-05,
+      "loss": 1.268,
+      "step": 1571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.331841273673135,
+      "learning_rate": 1.306224068012138e-05,
+      "loss": 1.1668,
+      "step": 1572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.41832217541647,
+      "learning_rate": 1.3054021158955736e-05,
+      "loss": 1.1369,
+      "step": 1573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9032270089151628,
+      "learning_rate": 1.3045799361604617e-05,
+      "loss": 0.6639,
+      "step": 1574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5944008582044407,
+      "learning_rate": 1.3037575294195781e-05,
+      "loss": 1.1422,
+      "step": 1575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3726159332186483,
+      "learning_rate": 1.3029348962858693e-05,
+      "loss": 1.1266,
+      "step": 1576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9426413738202335,
+      "learning_rate": 1.30211203737245e-05,
+      "loss": 0.6761,
+      "step": 1577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0584709225934295,
+      "learning_rate": 1.3012889532926036e-05,
+      "loss": 0.6569,
+      "step": 1578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.537849059777212,
+      "learning_rate": 1.300465644659781e-05,
+      "loss": 1.1943,
+      "step": 1579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1087185043453633,
+      "learning_rate": 1.2996421120876e-05,
+      "loss": 0.7276,
+      "step": 1580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9541501466064641,
+      "learning_rate": 1.2988183561898463e-05,
+      "loss": 0.7099,
+      "step": 1581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7912137776551615,
+      "learning_rate": 1.297994377580471e-05,
+      "loss": 0.6512,
+      "step": 1582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.846136521095365,
+      "learning_rate": 1.2971701768735923e-05,
+      "loss": 0.6173,
+      "step": 1583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8272705634654332,
+      "learning_rate": 1.2963457546834926e-05,
+      "loss": 0.647,
+      "step": 1584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8925399658071054,
+      "learning_rate": 1.2955211116246206e-05,
+      "loss": 0.6949,
+      "step": 1585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.023948129241329,
+      "learning_rate": 1.2946962483115893e-05,
+      "loss": 1.1996,
+      "step": 1586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9941007069834866,
+      "learning_rate": 1.2938711653591753e-05,
+      "loss": 0.6358,
+      "step": 1587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9430140849595272,
+      "learning_rate": 1.293045863382319e-05,
+      "loss": 0.6729,
+      "step": 1588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7645060454613626,
+      "learning_rate": 1.2922203429961249e-05,
+      "loss": 0.6411,
+      "step": 1589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4692152265293563,
+      "learning_rate": 1.2913946048158596e-05,
+      "loss": 1.2133,
+      "step": 1590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.031664809885615,
+      "learning_rate": 1.290568649456952e-05,
+      "loss": 0.6767,
+      "step": 1591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.608609038955653,
+      "learning_rate": 1.289742477534993e-05,
+      "loss": 1.1951,
+      "step": 1592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.746724632738632,
+      "learning_rate": 1.2889160896657349e-05,
+      "loss": 0.5809,
+      "step": 1593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8408131236170326,
+      "learning_rate": 1.2880894864650908e-05,
+      "loss": 0.6787,
+      "step": 1594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8261962953378579,
+      "learning_rate": 1.2872626685491347e-05,
+      "loss": 0.5901,
+      "step": 1595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.851632605199923,
+      "learning_rate": 1.2864356365341e-05,
+      "loss": 0.5728,
+      "step": 1596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9503971374058946,
+      "learning_rate": 1.2856083910363806e-05,
+      "loss": 0.6491,
+      "step": 1597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.829554823472215,
+      "learning_rate": 1.2847809326725284e-05,
+      "loss": 1.176,
+      "step": 1598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.728531008345621,
+      "learning_rate": 1.2839532620592547e-05,
+      "loss": 0.5604,
+      "step": 1599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3471341528710177,
+      "learning_rate": 1.2831253798134286e-05,
+      "loss": 1.172,
+      "step": 1600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8115538096279706,
+      "learning_rate": 1.2822972865520775e-05,
+      "loss": 0.6633,
+      "step": 1601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9619767998118975,
+      "learning_rate": 1.2814689828923855e-05,
+      "loss": 0.6508,
+      "step": 1602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3996745670556683,
+      "learning_rate": 1.2806404694516936e-05,
+      "loss": 1.1536,
+      "step": 1603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9520205798183483,
+      "learning_rate": 1.2798117468474989e-05,
+      "loss": 0.5855,
+      "step": 1604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0913887007459513,
+      "learning_rate": 1.2789828156974555e-05,
+      "loss": 0.6763,
+      "step": 1605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6544582168573942,
+      "learning_rate": 1.2781536766193718e-05,
+      "loss": 1.1921,
+      "step": 1606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8011200155189413,
+      "learning_rate": 1.2773243302312113e-05,
+      "loss": 1.1352,
+      "step": 1607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7554741072851794,
+      "learning_rate": 1.2764947771510928e-05,
+      "loss": 0.6679,
+      "step": 1608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9538574489188216,
+      "learning_rate": 1.2756650179972877e-05,
+      "loss": 0.6268,
+      "step": 1609
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9040650299121478,
+      "learning_rate": 1.2748350533882232e-05,
+      "loss": 0.6598,
+      "step": 1610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.886611554906667,
+      "learning_rate": 1.2740048839424772e-05,
+      "loss": 1.1528,
+      "step": 1611
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9600882572127996,
+      "learning_rate": 1.2731745102787816e-05,
+      "loss": 0.6251,
+      "step": 1612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.970659149769396,
+      "learning_rate": 1.2723439330160208e-05,
+      "loss": 0.6214,
+      "step": 1613
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8893455452622767,
+      "learning_rate": 1.2715131527732298e-05,
+      "loss": 0.6665,
+      "step": 1614
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.708752938096864,
+      "learning_rate": 1.2706821701695962e-05,
+      "loss": 0.6721,
+      "step": 1615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7753707173474664,
+      "learning_rate": 1.2698509858244567e-05,
+      "loss": 1.1786,
+      "step": 1616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7458734221308834,
+      "learning_rate": 1.2690196003573003e-05,
+      "loss": 0.6634,
+      "step": 1617
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.466619844591997,
+      "learning_rate": 1.2681880143877647e-05,
+      "loss": 1.1955,
+      "step": 1618
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0062617059931944,
+      "learning_rate": 1.2673562285356375e-05,
+      "loss": 0.6479,
+      "step": 1619
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.667934297591169,
+      "learning_rate": 1.2665242434208547e-05,
+      "loss": 1.1792,
+      "step": 1620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0732428890134758,
+      "learning_rate": 1.2656920596635014e-05,
+      "loss": 0.6862,
+      "step": 1621
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9028241539605588,
+      "learning_rate": 1.2648596778838106e-05,
+      "loss": 0.6238,
+      "step": 1622
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.088600759364443,
+      "learning_rate": 1.264027098702163e-05,
+      "loss": 0.7113,
+      "step": 1623
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.676910809952784,
+      "learning_rate": 1.2631943227390858e-05,
+      "loss": 0.5699,
+      "step": 1624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.859154480685328,
+      "learning_rate": 1.2623613506152538e-05,
+      "loss": 1.1793,
+      "step": 1625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.508819864403052,
+      "learning_rate": 1.2615281829514873e-05,
+      "loss": 1.1576,
+      "step": 1626
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.740052510202756,
+      "learning_rate": 1.2606948203687528e-05,
+      "loss": 1.2087,
+      "step": 1627
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9916425759647851,
+      "learning_rate": 1.2598612634881615e-05,
+      "loss": 0.6274,
+      "step": 1628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3739204120869526,
+      "learning_rate": 1.2590275129309703e-05,
+      "loss": 1.1442,
+      "step": 1629
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9533825182591817,
+      "learning_rate": 1.2581935693185794e-05,
+      "loss": 0.6265,
+      "step": 1630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9957717210402437,
+      "learning_rate": 1.2573594332725336e-05,
+      "loss": 0.6487,
+      "step": 1631
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9997483641663352,
+      "learning_rate": 1.2565251054145206e-05,
+      "loss": 0.5839,
+      "step": 1632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.739366564788738,
+      "learning_rate": 1.255690586366372e-05,
+      "loss": 0.6556,
+      "step": 1633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8182419222180166,
+      "learning_rate": 1.254855876750061e-05,
+      "loss": 1.1391,
+      "step": 1634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8683407512935102,
+      "learning_rate": 1.254020977187703e-05,
+      "loss": 0.6472,
+      "step": 1635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9086571426010497,
+      "learning_rate": 1.2531858883015548e-05,
+      "loss": 0.577,
+      "step": 1636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9106046403991304,
+      "learning_rate": 1.2523506107140152e-05,
+      "loss": 0.5737,
+      "step": 1637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9251799763311153,
+      "learning_rate": 1.2515151450476226e-05,
+      "loss": 0.673,
+      "step": 1638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.131798190692535,
+      "learning_rate": 1.250679491925056e-05,
+      "loss": 0.6714,
+      "step": 1639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.514994217327796,
+      "learning_rate": 1.2498436519691341e-05,
+      "loss": 1.1216,
+      "step": 1640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.776389351990468,
+      "learning_rate": 1.2490076258028152e-05,
+      "loss": 0.5527,
+      "step": 1641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.095953902384477,
+      "learning_rate": 1.2481714140491959e-05,
+      "loss": 0.6411,
+      "step": 1642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8988091063381238,
+      "learning_rate": 1.2473350173315106e-05,
+      "loss": 0.612,
+      "step": 1643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9876666289350198,
+      "learning_rate": 1.2464984362731332e-05,
+      "loss": 0.6444,
+      "step": 1644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8919250673785102,
+      "learning_rate": 1.2456616714975733e-05,
+      "loss": 0.6715,
+      "step": 1645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.351117113567901,
+      "learning_rate": 1.2448247236284788e-05,
+      "loss": 1.16,
+      "step": 1646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.838999229668501,
+      "learning_rate": 1.2439875932896326e-05,
+      "loss": 1.1536,
+      "step": 1647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.669239475141962,
+      "learning_rate": 1.243150281104955e-05,
+      "loss": 0.6374,
+      "step": 1648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7072329630565954,
+      "learning_rate": 1.2423127876985008e-05,
+      "loss": 0.5709,
+      "step": 1649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.888540210170697,
+      "learning_rate": 1.2414751136944605e-05,
+      "loss": 0.6067,
+      "step": 1650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9866835036634312,
+      "learning_rate": 1.2406372597171593e-05,
+      "loss": 0.6588,
+      "step": 1651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.585361328623722,
+      "learning_rate": 1.2397992263910554e-05,
+      "loss": 1.1472,
+      "step": 1652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6865584211858657,
+      "learning_rate": 1.2389610143407423e-05,
+      "loss": 0.5729,
+      "step": 1653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7740291601168838,
+      "learning_rate": 1.2381226241909454e-05,
+      "loss": 0.6301,
+      "step": 1654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7744582611705262,
+      "learning_rate": 1.2372840565665233e-05,
+      "loss": 0.6411,
+      "step": 1655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.013038375132442,
+      "learning_rate": 1.2364453120924673e-05,
+      "loss": 0.6393,
+      "step": 1656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7847174653510816,
+      "learning_rate": 1.2356063913939e-05,
+      "loss": 0.6503,
+      "step": 1657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0070191312350527,
+      "learning_rate": 1.2347672950960755e-05,
+      "loss": 0.652,
+      "step": 1658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7975250150507125,
+      "learning_rate": 1.2339280238243782e-05,
+      "loss": 0.5467,
+      "step": 1659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8093134743969737,
+      "learning_rate": 1.2330885782043238e-05,
+      "loss": 1.1722,
+      "step": 1660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8004679387975182,
+      "learning_rate": 1.232248958861558e-05,
+      "loss": 0.6455,
+      "step": 1661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.368779015814652,
+      "learning_rate": 1.2314091664218548e-05,
+      "loss": 1.1676,
+      "step": 1662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7291004219706663,
+      "learning_rate": 1.2305692015111186e-05,
+      "loss": 0.6316,
+      "step": 1663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8354634053616126,
+      "learning_rate": 1.2297290647553813e-05,
+      "loss": 0.6587,
+      "step": 1664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7334229460458164,
+      "learning_rate": 1.2288887567808034e-05,
+      "loss": 0.6416,
+      "step": 1665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9807403736285305,
+      "learning_rate": 1.2280482782136733e-05,
+      "loss": 0.6064,
+      "step": 1666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.774471504213978,
+      "learning_rate": 1.2272076296804056e-05,
+      "loss": 0.5636,
+      "step": 1667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.625258018401385,
+      "learning_rate": 1.2263668118075422e-05,
+      "loss": 1.1887,
+      "step": 1668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.803721365103422,
+      "learning_rate": 1.2255258252217514e-05,
+      "loss": 0.559,
+      "step": 1669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.822030443060436,
+      "learning_rate": 1.224684670549827e-05,
+      "loss": 0.6457,
+      "step": 1670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6313523834301344,
+      "learning_rate": 1.2238433484186879e-05,
+      "loss": 1.1707,
+      "step": 1671
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0159265248241622,
+      "learning_rate": 1.2230018594553777e-05,
+      "loss": 0.5845,
+      "step": 1672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.619144155907245,
+      "learning_rate": 1.2221602042870649e-05,
+      "loss": 0.5612,
+      "step": 1673
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.766366461901434,
+      "learning_rate": 1.2213183835410416e-05,
+      "loss": 0.6245,
+      "step": 1674
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.33092479093917,
+      "learning_rate": 1.2204763978447232e-05,
+      "loss": 1.0845,
+      "step": 1675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7616707456838314,
+      "learning_rate": 1.219634247825648e-05,
+      "loss": 0.6099,
+      "step": 1676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.553125947851537,
+      "learning_rate": 1.2187919341114773e-05,
+      "loss": 1.1729,
+      "step": 1677
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.853891908367791,
+      "learning_rate": 1.2179494573299937e-05,
+      "loss": 0.6178,
+      "step": 1678
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.971317972432188,
+      "learning_rate": 1.2171068181091013e-05,
+      "loss": 0.6707,
+      "step": 1679
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9251157986864689,
+      "learning_rate": 1.2162640170768263e-05,
+      "loss": 0.6344,
+      "step": 1680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.429242498051686,
+      "learning_rate": 1.2154210548613142e-05,
+      "loss": 1.2071,
+      "step": 1681
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7840438412049557,
+      "learning_rate": 1.2145779320908316e-05,
+      "loss": 0.5556,
+      "step": 1682
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8142348586088706,
+      "learning_rate": 1.2137346493937643e-05,
+      "loss": 0.6284,
+      "step": 1683
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0023497528985152,
+      "learning_rate": 1.2128912073986173e-05,
+      "loss": 0.7353,
+      "step": 1684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4573744923617875,
+      "learning_rate": 1.2120476067340145e-05,
+      "loss": 1.145,
+      "step": 1685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8762124497993675,
+      "learning_rate": 1.211203848028698e-05,
+      "loss": 0.5921,
+      "step": 1686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4585898949968636,
+      "learning_rate": 1.2103599319115276e-05,
+      "loss": 1.1851,
+      "step": 1687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8896586011458214,
+      "learning_rate": 1.2095158590114804e-05,
+      "loss": 1.151,
+      "step": 1688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9030913003425178,
+      "learning_rate": 1.2086716299576505e-05,
+      "loss": 0.615,
+      "step": 1689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7956338657135726,
+      "learning_rate": 1.2078272453792485e-05,
+      "loss": 0.6579,
+      "step": 1690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8224780448451787,
+      "learning_rate": 1.2069827059056006e-05,
+      "loss": 0.6054,
+      "step": 1691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.79313038219523,
+      "learning_rate": 1.2061380121661478e-05,
+      "loss": 1.1467,
+      "step": 1692
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.087099734147116,
+      "learning_rate": 1.2052931647904484e-05,
+      "loss": 0.673,
+      "step": 1693
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0309119860213682,
+      "learning_rate": 1.2044481644081725e-05,
+      "loss": 0.6362,
+      "step": 1694
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5994745911876516,
+      "learning_rate": 1.2036030116491057e-05,
+      "loss": 1.155,
+      "step": 1695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.847319656050317,
+      "learning_rate": 1.202757707143147e-05,
+      "loss": 0.6085,
+      "step": 1696
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9089018180599278,
+      "learning_rate": 1.201912251520308e-05,
+      "loss": 0.6418,
+      "step": 1697
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.41408364211794,
+      "learning_rate": 1.2010666454107137e-05,
+      "loss": 1.1187,
+      "step": 1698
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8726579732558073,
+      "learning_rate": 1.2002208894446006e-05,
+      "loss": 0.68,
+      "step": 1699
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0564885167900386,
+      "learning_rate": 1.1993749842523168e-05,
+      "loss": 0.7013,
+      "step": 1700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9463744868345294,
+      "learning_rate": 1.1985289304643228e-05,
+      "loss": 0.5935,
+      "step": 1701
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9778719615467748,
+      "learning_rate": 1.1976827287111884e-05,
+      "loss": 0.6989,
+      "step": 1702
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.857923999660307,
+      "learning_rate": 1.1968363796235947e-05,
+      "loss": 0.6321,
+      "step": 1703
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0028917437585343,
+      "learning_rate": 1.1959898838323314e-05,
+      "loss": 0.6123,
+      "step": 1704
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9366619002439738,
+      "learning_rate": 1.1951432419682994e-05,
+      "loss": 0.6717,
+      "step": 1705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9883614203073214,
+      "learning_rate": 1.1942964546625064e-05,
+      "loss": 0.641,
+      "step": 1706
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9602520281320193,
+      "learning_rate": 1.1934495225460703e-05,
+      "loss": 0.6432,
+      "step": 1707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0440424053092316,
+      "learning_rate": 1.1926024462502152e-05,
+      "loss": 0.684,
+      "step": 1708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9771970226763906,
+      "learning_rate": 1.1917552264062746e-05,
+      "loss": 0.5685,
+      "step": 1709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.450918992587182,
+      "learning_rate": 1.1909078636456874e-05,
+      "loss": 1.133,
+      "step": 1710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9189165603738354,
+      "learning_rate": 1.1900603586e-05,
+      "loss": 0.6719,
+      "step": 1711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7912193587051983,
+      "learning_rate": 1.1892127119008637e-05,
+      "loss": 0.5933,
+      "step": 1712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9927130127238222,
+      "learning_rate": 1.1883649241800368e-05,
+      "loss": 0.6955,
+      "step": 1713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6672204542875475,
+      "learning_rate": 1.1875169960693818e-05,
+      "loss": 1.1701,
+      "step": 1714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.472030851944355,
+      "learning_rate": 1.1866689282008657e-05,
+      "loss": 1.1295,
+      "step": 1715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.960811176526034,
+      "learning_rate": 1.1858207212065605e-05,
+      "loss": 0.5665,
+      "step": 1716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8610089101074307,
+      "learning_rate": 1.1849723757186412e-05,
+      "loss": 0.5763,
+      "step": 1717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.948731318508705,
+      "learning_rate": 1.1841238923693862e-05,
+      "loss": 0.6122,
+      "step": 1718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3910151785974456,
+      "learning_rate": 1.183275271791177e-05,
+      "loss": 1.2122,
+      "step": 1719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.762149956434838,
+      "learning_rate": 1.1824265146164966e-05,
+      "loss": 1.249,
+      "step": 1720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8827661724704179,
+      "learning_rate": 1.1815776214779304e-05,
+      "loss": 0.6711,
+      "step": 1721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4088640812693343,
+      "learning_rate": 1.1807285930081653e-05,
+      "loss": 1.1788,
+      "step": 1722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3861833014419314,
+      "learning_rate": 1.1798794298399883e-05,
+      "loss": 1.1652,
+      "step": 1723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.597925262465035,
+      "learning_rate": 1.1790301326062872e-05,
+      "loss": 1.1181,
+      "step": 1724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6856041957132752,
+      "learning_rate": 1.1781807019400504e-05,
+      "loss": 0.5827,
+      "step": 1725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9041653311239035,
+      "learning_rate": 1.1773311384743647e-05,
+      "loss": 0.5506,
+      "step": 1726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2771871347277592,
+      "learning_rate": 1.176481442842416e-05,
+      "loss": 0.7362,
+      "step": 1727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9189142929211116,
+      "learning_rate": 1.1756316156774892e-05,
+      "loss": 0.5772,
+      "step": 1728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8304441628168986,
+      "learning_rate": 1.1747816576129674e-05,
+      "loss": 1.2137,
+      "step": 1729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6277493907100937,
+      "learning_rate": 1.1739315692823307e-05,
+      "loss": 1.1666,
+      "step": 1730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8483195113886208,
+      "learning_rate": 1.1730813513191564e-05,
+      "loss": 0.6439,
+      "step": 1731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.065838828936817,
+      "learning_rate": 1.1722310043571182e-05,
+      "loss": 0.6977,
+      "step": 1732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8433942931148855,
+      "learning_rate": 1.1713805290299867e-05,
+      "loss": 0.6292,
+      "step": 1733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8449601130233448,
+      "learning_rate": 1.1705299259716272e-05,
+      "loss": 0.639,
+      "step": 1734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.06912012368247,
+      "learning_rate": 1.1696791958160014e-05,
+      "loss": 0.6058,
+      "step": 1735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9665697210372246,
+      "learning_rate": 1.1688283391971645e-05,
+      "loss": 0.5632,
+      "step": 1736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7617987783250963,
+      "learning_rate": 1.1679773567492666e-05,
+      "loss": 0.6069,
+      "step": 1737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9061837057256774,
+      "learning_rate": 1.1671262491065519e-05,
+      "loss": 0.5513,
+      "step": 1738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.834361419018219,
+      "learning_rate": 1.1662750169033569e-05,
+      "loss": 0.6628,
+      "step": 1739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8186268531142604,
+      "learning_rate": 1.1654236607741116e-05,
+      "loss": 0.6522,
+      "step": 1740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.439145731020705,
+      "learning_rate": 1.1645721813533386e-05,
+      "loss": 1.0965,
+      "step": 1741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3134347686659864,
+      "learning_rate": 1.1637205792756519e-05,
+      "loss": 1.1553,
+      "step": 1742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.813880048985399,
+      "learning_rate": 1.162868855175757e-05,
+      "loss": 0.5834,
+      "step": 1743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0828709641452345,
+      "learning_rate": 1.1620170096884502e-05,
+      "loss": 0.6428,
+      "step": 1744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.123178778611842,
+      "learning_rate": 1.161165043448619e-05,
+      "loss": 0.5723,
+      "step": 1745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2846674590376814,
+      "learning_rate": 1.1603129570912404e-05,
+      "loss": 0.6553,
+      "step": 1746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.757627149007369,
+      "learning_rate": 1.1594607512513805e-05,
+      "loss": 1.1372,
+      "step": 1747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.989161707552885,
+      "learning_rate": 1.1586084265641948e-05,
+      "loss": 1.1264,
+      "step": 1748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9107361241597716,
+      "learning_rate": 1.157755983664928e-05,
+      "loss": 0.6005,
+      "step": 1749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7681397521096258,
+      "learning_rate": 1.1569034231889122e-05,
+      "loss": 0.6636,
+      "step": 1750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0609837433118834,
+      "learning_rate": 1.1560507457715669e-05,
+      "loss": 0.6522,
+      "step": 1751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8641999243304446,
+      "learning_rate": 1.1551979520483995e-05,
+      "loss": 0.6434,
+      "step": 1752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0852655431040743,
+      "learning_rate": 1.1543450426550036e-05,
+      "loss": 0.6382,
+      "step": 1753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0703356958784065,
+      "learning_rate": 1.1534920182270592e-05,
+      "loss": 0.6157,
+      "step": 1754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0273533739563887,
+      "learning_rate": 1.1526388794003322e-05,
+      "loss": 0.5693,
+      "step": 1755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7900284452670414,
+      "learning_rate": 1.151785626810673e-05,
+      "loss": 1.1395,
+      "step": 1756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9910849068800596,
+      "learning_rate": 1.1509322610940177e-05,
+      "loss": 0.6586,
+      "step": 1757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6232150002158487,
+      "learning_rate": 1.1500787828863861e-05,
+      "loss": 1.0826,
+      "step": 1758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6512914645838106,
+      "learning_rate": 1.1492251928238827e-05,
+      "loss": 0.5946,
+      "step": 1759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0165198148666477,
+      "learning_rate": 1.1483714915426936e-05,
+      "loss": 0.6443,
+      "step": 1760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9893050331079973,
+      "learning_rate": 1.1475176796790901e-05,
+      "loss": 0.6406,
+      "step": 1761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.008718292644802,
+      "learning_rate": 1.1466637578694242e-05,
+      "loss": 0.6267,
+      "step": 1762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0994108437735823,
+      "learning_rate": 1.1458097267501302e-05,
+      "loss": 0.6005,
+      "step": 1763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0246982524135206,
+      "learning_rate": 1.1449555869577244e-05,
+      "loss": 0.6663,
+      "step": 1764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.927242690794933,
+      "learning_rate": 1.1441013391288039e-05,
+      "loss": 0.6015,
+      "step": 1765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9074063866460509,
+      "learning_rate": 1.1432469839000459e-05,
+      "loss": 0.6851,
+      "step": 1766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.785652954695507,
+      "learning_rate": 1.142392521908208e-05,
+      "loss": 0.587,
+      "step": 1767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7010984213530818,
+      "learning_rate": 1.1415379537901272e-05,
+      "loss": 1.1656,
+      "step": 1768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9741741345468118,
+      "learning_rate": 1.1406832801827198e-05,
+      "loss": 0.6777,
+      "step": 1769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6479627710553724,
+      "learning_rate": 1.1398285017229807e-05,
+      "loss": 0.5745,
+      "step": 1770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.85571644320314,
+      "learning_rate": 1.1389736190479828e-05,
+      "loss": 0.5965,
+      "step": 1771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9186350771486755,
+      "learning_rate": 1.1381186327948767e-05,
+      "loss": 0.5881,
+      "step": 1772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.14545447714804,
+      "learning_rate": 1.1372635436008901e-05,
+      "loss": 0.6798,
+      "step": 1773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7819565140032962,
+      "learning_rate": 1.136408352103328e-05,
+      "loss": 0.6019,
+      "step": 1774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9824478665471483,
+      "learning_rate": 1.1355530589395708e-05,
+      "loss": 0.6346,
+      "step": 1775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7527365030103483,
+      "learning_rate": 1.1346976647470753e-05,
+      "loss": 0.625,
+      "step": 1776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8820013147421624,
+      "learning_rate": 1.133842170163373e-05,
+      "loss": 1.1531,
+      "step": 1777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4241915218471046,
+      "learning_rate": 1.1329865758260712e-05,
+      "loss": 1.1173,
+      "step": 1778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.401209332969751,
+      "learning_rate": 1.1321308823728506e-05,
+      "loss": 1.161,
+      "step": 1779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7005734399092067,
+      "learning_rate": 1.1312750904414662e-05,
+      "loss": 0.5624,
+      "step": 1780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7533788971929252,
+      "learning_rate": 1.1304192006697458e-05,
+      "loss": 0.5729,
+      "step": 1781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8640883933136803,
+      "learning_rate": 1.1295632136955917e-05,
+      "loss": 0.6538,
+      "step": 1782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8836352968822085,
+      "learning_rate": 1.1287071301569767e-05,
+      "loss": 0.6461,
+      "step": 1783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8103508735554794,
+      "learning_rate": 1.1278509506919468e-05,
+      "loss": 0.6283,
+      "step": 1784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8410146910172294,
+      "learning_rate": 1.126994675938619e-05,
+      "loss": 0.6149,
+      "step": 1785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.087195463156637,
+      "learning_rate": 1.1261383065351816e-05,
+      "loss": 1.0903,
+      "step": 1786
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9864788427582944,
+      "learning_rate": 1.1252818431198931e-05,
+      "loss": 0.6447,
+      "step": 1787
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6881588320349605,
+      "learning_rate": 1.1244252863310825e-05,
+      "loss": 0.5992,
+      "step": 1788
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.108042312420301,
+      "learning_rate": 1.123568636807148e-05,
+      "loss": 0.6695,
+      "step": 1789
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.569503001825289,
+      "learning_rate": 1.1227118951865577e-05,
+      "loss": 1.1672,
+      "step": 1790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7622664890653512,
+      "learning_rate": 1.1218550621078468e-05,
+      "loss": 0.5821,
+      "step": 1791
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8616484611824524,
+      "learning_rate": 1.1209981382096202e-05,
+      "loss": 0.6452,
+      "step": 1792
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.958574669034025,
+      "learning_rate": 1.1201411241305493e-05,
+      "loss": 0.5937,
+      "step": 1793
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7571968017770339,
+      "learning_rate": 1.1192840205093741e-05,
+      "loss": 0.6398,
+      "step": 1794
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.807174952076636,
+      "learning_rate": 1.1184268279849e-05,
+      "loss": 0.6616,
+      "step": 1795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.563360434837779,
+      "learning_rate": 1.1175695471959994e-05,
+      "loss": 1.1579,
+      "step": 1796
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9494872877497222,
+      "learning_rate": 1.1167121787816096e-05,
+      "loss": 1.113,
+      "step": 1797
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3399890158122014,
+      "learning_rate": 1.1158547233807348e-05,
+      "loss": 1.1439,
+      "step": 1798
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.517012151109439,
+      "learning_rate": 1.1149971816324429e-05,
+      "loss": 1.0977,
+      "step": 1799
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9665695611989944,
+      "learning_rate": 1.1141395541758662e-05,
+      "loss": 0.5859,
+      "step": 1800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8929587264878054,
+      "learning_rate": 1.1132818416502006e-05,
+      "loss": 0.6635,
+      "step": 1801
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9998158498508443,
+      "learning_rate": 1.112424044694707e-05,
+      "loss": 0.5751,
+      "step": 1802
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.042954835857277,
+      "learning_rate": 1.1115661639487069e-05,
+      "loss": 0.6312,
+      "step": 1803
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8349815212757152,
+      "learning_rate": 1.1107082000515862e-05,
+      "loss": 0.646,
+      "step": 1804
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8105178878055783,
+      "learning_rate": 1.1098501536427915e-05,
+      "loss": 0.6191,
+      "step": 1805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9108072978078328,
+      "learning_rate": 1.1089920253618314e-05,
+      "loss": 1.2156,
+      "step": 1806
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9838160987381008,
+      "learning_rate": 1.1081338158482762e-05,
+      "loss": 0.6689,
+      "step": 1807
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.69028209136999,
+      "learning_rate": 1.1072755257417553e-05,
+      "loss": 0.5702,
+      "step": 1808
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7742935397977044,
+      "learning_rate": 1.106417155681959e-05,
+      "loss": 1.11,
+      "step": 1809
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9133479917027056,
+      "learning_rate": 1.1055587063086378e-05,
+      "loss": 0.6161,
+      "step": 1810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9483721266943659,
+      "learning_rate": 1.1047001782615997e-05,
+      "loss": 0.6238,
+      "step": 1811
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2918067137219973,
+      "learning_rate": 1.1038415721807132e-05,
+      "loss": 1.1571,
+      "step": 1812
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.970561832009966,
+      "learning_rate": 1.102982888705903e-05,
+      "loss": 1.1789,
+      "step": 1813
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8955941029439147,
+      "learning_rate": 1.1021241284771533e-05,
+      "loss": 0.5672,
+      "step": 1814
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.021998400362541,
+      "learning_rate": 1.1012652921345044e-05,
+      "loss": 0.6625,
+      "step": 1815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6412143354284163,
+      "learning_rate": 1.1004063803180538e-05,
+      "loss": 1.2106,
+      "step": 1816
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9188144078874638,
+      "learning_rate": 1.0995473936679546e-05,
+      "loss": 0.5565,
+      "step": 1817
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.840044009954613,
+      "learning_rate": 1.098688332824417e-05,
+      "loss": 0.6334,
+      "step": 1818
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7524288743839442,
+      "learning_rate": 1.097829198427705e-05,
+      "loss": 0.5945,
+      "step": 1819
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.471185458499901,
+      "learning_rate": 1.0969699911181381e-05,
+      "loss": 1.1275,
+      "step": 1820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8252594935891124,
+      "learning_rate": 1.0961107115360903e-05,
+      "loss": 0.6123,
+      "step": 1821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8127783523299497,
+      "learning_rate": 1.0952513603219895e-05,
+      "loss": 0.6756,
+      "step": 1822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9053702194925959,
+      "learning_rate": 1.0943919381163163e-05,
+      "loss": 0.6186,
+      "step": 1823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.002270079234928,
+      "learning_rate": 1.093532445559605e-05,
+      "loss": 0.6744,
+      "step": 1824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9225979264961166,
+      "learning_rate": 1.0926728832924422e-05,
+      "loss": 0.6545,
+      "step": 1825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.631872422508765,
+      "learning_rate": 1.0918132519554655e-05,
+      "loss": 1.2056,
+      "step": 1826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8792070214598422,
+      "learning_rate": 1.0909535521893659e-05,
+      "loss": 0.6373,
+      "step": 1827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8305376173708705,
+      "learning_rate": 1.0900937846348833e-05,
+      "loss": 0.5936,
+      "step": 1828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.536946388042619,
+      "learning_rate": 1.0892339499328095e-05,
+      "loss": 1.1806,
+      "step": 1829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.452718963134155,
+      "learning_rate": 1.088374048723986e-05,
+      "loss": 1.0976,
+      "step": 1830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.725064360836767,
+      "learning_rate": 1.0875140816493035e-05,
+      "loss": 0.5882,
+      "step": 1831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7943737019604582,
+      "learning_rate": 1.0866540493497022e-05,
+      "loss": 0.6056,
+      "step": 1832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8263695325511564,
+      "learning_rate": 1.0857939524661708e-05,
+      "loss": 0.6163,
+      "step": 1833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8175411483135,
+      "learning_rate": 1.0849337916397462e-05,
+      "loss": 0.6659,
+      "step": 1834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4414721310899465,
+      "learning_rate": 1.0840735675115129e-05,
+      "loss": 1.1007,
+      "step": 1835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6482863927481712,
+      "learning_rate": 1.0832132807226024e-05,
+      "loss": 0.5901,
+      "step": 1836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7958591553713423,
+      "learning_rate": 1.0823529319141928e-05,
+      "loss": 0.6306,
+      "step": 1837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8772810502231112,
+      "learning_rate": 1.0814925217275093e-05,
+      "loss": 0.6475,
+      "step": 1838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8632836367621572,
+      "learning_rate": 1.0806320508038224e-05,
+      "loss": 0.5762,
+      "step": 1839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9262914580229447,
+      "learning_rate": 1.079771519784447e-05,
+      "loss": 0.6178,
+      "step": 1840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9086643196219282,
+      "learning_rate": 1.0789109293107433e-05,
+      "loss": 0.6407,
+      "step": 1841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9724235522118474,
+      "learning_rate": 1.078050280024117e-05,
+      "loss": 0.5834,
+      "step": 1842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.794748671118757,
+      "learning_rate": 1.077189572566016e-05,
+      "loss": 0.5614,
+      "step": 1843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.781354246445819,
+      "learning_rate": 1.0763288075779323e-05,
+      "loss": 0.6446,
+      "step": 1844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6322382862653055,
+      "learning_rate": 1.0754679857014005e-05,
+      "loss": 0.542,
+      "step": 1845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5213535751671623,
+      "learning_rate": 1.0746071075779985e-05,
+      "loss": 1.1438,
+      "step": 1846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7683671698578476,
+      "learning_rate": 1.0737461738493444e-05,
+      "loss": 0.6682,
+      "step": 1847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5485991262738152,
+      "learning_rate": 1.0728851851570993e-05,
+      "loss": 1.1291,
+      "step": 1848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.776743287486541,
+      "learning_rate": 1.0720241421429647e-05,
+      "loss": 0.5676,
+      "step": 1849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.449066808923292,
+      "learning_rate": 1.0711630454486824e-05,
+      "loss": 1.1802,
+      "step": 1850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0005179910222943,
+      "learning_rate": 1.0703018957160344e-05,
+      "loss": 0.6215,
+      "step": 1851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.82774330701236,
+      "learning_rate": 1.0694406935868426e-05,
+      "loss": 0.5602,
+      "step": 1852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3530956523757687,
+      "learning_rate": 1.068579439702967e-05,
+      "loss": 1.1247,
+      "step": 1853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.324593260340001,
+      "learning_rate": 1.067718134706307e-05,
+      "loss": 1.1392,
+      "step": 1854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9306860786191324,
+      "learning_rate": 1.0668567792388e-05,
+      "loss": 0.7015,
+      "step": 1855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7342369101382993,
+      "learning_rate": 1.0659953739424207e-05,
+      "loss": 0.5886,
+      "step": 1856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.77371707208596,
+      "learning_rate": 1.065133919459181e-05,
+      "loss": 0.621,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.762508672851476,
+      "learning_rate": 1.06427241643113e-05,
+      "loss": 0.6043,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8040591461036937,
+      "learning_rate": 1.0634108655003521e-05,
+      "loss": 0.6405,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6028205785239729,
+      "learning_rate": 1.0625492673089676e-05,
+      "loss": 0.6077,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7153934275745097,
+      "learning_rate": 1.061687622499133e-05,
+      "loss": 0.6061,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9715828969486557,
+      "learning_rate": 1.0608259317130383e-05,
+      "loss": 0.7078,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7167252767609895,
+      "learning_rate": 1.0599641955929085e-05,
+      "loss": 1.1462,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5514667396661137,
+      "learning_rate": 1.0591024147810022e-05,
+      "loss": 1.1518,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8537929645790412,
+      "learning_rate": 1.0582405899196108e-05,
+      "loss": 0.6418,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2173645602594148,
+      "learning_rate": 1.0573787216510598e-05,
+      "loss": 1.1147,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7876894071843628,
+      "learning_rate": 1.0565168106177058e-05,
+      "loss": 0.6166,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6428766760641444,
+      "learning_rate": 1.0556548574619377e-05,
+      "loss": 1.1308,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9083072609428877,
+      "learning_rate": 1.0547928628261755e-05,
+      "loss": 0.6428,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.373567347196092,
+      "learning_rate": 1.053930827352871e-05,
+      "loss": 1.1196,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.467527638817866,
+      "learning_rate": 1.053068751684506e-05,
+      "loss": 1.0966,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6460401325588756,
+      "learning_rate": 1.0522066364635914e-05,
+      "loss": 1.196,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8579642521827913,
+      "learning_rate": 1.0513444823326688e-05,
+      "loss": 0.6544,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3625258436391823,
+      "learning_rate": 1.0504822899343083e-05,
+      "loss": 1.1557,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8910679748265833,
+      "learning_rate": 1.0496200599111086e-05,
+      "loss": 0.587,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2312061357614104,
+      "learning_rate": 1.0487577929056963e-05,
+      "loss": 1.1413,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7652784347187034,
+      "learning_rate": 1.0478954895607254e-05,
+      "loss": 0.6745,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6386681969440318,
+      "learning_rate": 1.0470331505188773e-05,
+      "loss": 1.1215,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9879113725728934,
+      "learning_rate": 1.0461707764228601e-05,
+      "loss": 0.6688,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8817139095245008,
+      "learning_rate": 1.0453083679154078e-05,
+      "loss": 0.6744,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1297605543815843,
+      "learning_rate": 1.04444592563928e-05,
+      "loss": 0.6187,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8467371513502864,
+      "learning_rate": 1.043583450237262e-05,
+      "loss": 0.6571,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7119574028853606,
+      "learning_rate": 1.0427209423521627e-05,
+      "loss": 0.5696,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.657989800427006,
+      "learning_rate": 1.0418584026268164e-05,
+      "loss": 0.6191,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.503860984797557,
+      "learning_rate": 1.04099583170408e-05,
+      "loss": 1.2385,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7672037907312779,
+      "learning_rate": 1.040133230226835e-05,
+      "loss": 0.554,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6905819733594365,
+      "learning_rate": 1.0392705988379846e-05,
+      "loss": 0.5154,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.444324297907412,
+      "learning_rate": 1.0384079381804543e-05,
+      "loss": 1.1818,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0119619239866924,
+      "learning_rate": 1.037545248897192e-05,
+      "loss": 0.5803,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9095239030690188,
+      "learning_rate": 1.0366825316311668e-05,
+      "loss": 0.5996,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.746356742279472,
+      "learning_rate": 1.0358197870253683e-05,
+      "loss": 0.5871,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7737797857452293,
+      "learning_rate": 1.0349570157228068e-05,
+      "loss": 0.5909,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.846198936666428,
+      "learning_rate": 1.0340942183665121e-05,
+      "loss": 0.6119,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.821784225926696,
+      "learning_rate": 1.0332313955995343e-05,
+      "loss": 0.6321,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6756928844768322,
+      "learning_rate": 1.0323685480649415e-05,
+      "loss": 0.5901,
+      "step": 1895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7400592700881967,
+      "learning_rate": 1.0315056764058205e-05,
+      "loss": 1.1618,
+      "step": 1896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8248240303314816,
+      "learning_rate": 1.0306427812652763e-05,
+      "loss": 0.5982,
+      "step": 1897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6951152656624737,
+      "learning_rate": 1.0297798632864313e-05,
+      "loss": 0.6234,
+      "step": 1898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3122789867075233,
+      "learning_rate": 1.0289169231124249e-05,
+      "loss": 1.1442,
+      "step": 1899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7767339056774207,
+      "learning_rate": 1.0280539613864133e-05,
+      "loss": 0.6346,
+      "step": 1900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7717431658265885,
+      "learning_rate": 1.0271909787515681e-05,
+      "loss": 0.5412,
+      "step": 1901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.723414485336943,
+      "learning_rate": 1.0263279758510774e-05,
+      "loss": 1.1378,
+      "step": 1902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.759259978676682,
+      "learning_rate": 1.0254649533281436e-05,
+      "loss": 1.1646,
+      "step": 1903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.697173039206291,
+      "learning_rate": 1.0246019118259838e-05,
+      "loss": 1.2473,
+      "step": 1904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3876893822717156,
+      "learning_rate": 1.0237388519878294e-05,
+      "loss": 1.217,
+      "step": 1905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6946668227335957,
+      "learning_rate": 1.0228757744569264e-05,
+      "loss": 0.5616,
+      "step": 1906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8672612736465053,
+      "learning_rate": 1.0220126798765327e-05,
+      "loss": 0.6162,
+      "step": 1907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8156565107738485,
+      "learning_rate": 1.0211495688899189e-05,
+      "loss": 0.6204,
+      "step": 1908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8985306095707122,
+      "learning_rate": 1.0202864421403684e-05,
+      "loss": 0.6152,
+      "step": 1909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.922255338696469,
+      "learning_rate": 1.019423300271177e-05,
+      "loss": 1.1258,
+      "step": 1910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6737812714367801,
+      "learning_rate": 1.0185601439256498e-05,
+      "loss": 0.5524,
+      "step": 1911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8377404884614075,
+      "learning_rate": 1.0176969737471048e-05,
+      "loss": 0.6327,
+      "step": 1912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7931016848765027,
+      "learning_rate": 1.0168337903788688e-05,
+      "loss": 0.5845,
+      "step": 1913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7278805984467775,
+      "learning_rate": 1.0159705944642792e-05,
+      "loss": 0.5969,
+      "step": 1914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9650728663617696,
+      "learning_rate": 1.0151073866466827e-05,
+      "loss": 0.5878,
+      "step": 1915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7214569789257586,
+      "learning_rate": 1.0142441675694347e-05,
+      "loss": 0.5901,
+      "step": 1916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9672142391724856,
+      "learning_rate": 1.0133809378758987e-05,
+      "loss": 0.6794,
+      "step": 1917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.441206708711904,
+      "learning_rate": 1.0125176982094469e-05,
+      "loss": 1.1082,
+      "step": 1918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8729019407552479,
+      "learning_rate": 1.0116544492134583e-05,
+      "loss": 0.7106,
+      "step": 1919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9053659875799096,
+      "learning_rate": 1.0107911915313192e-05,
+      "loss": 0.6613,
+      "step": 1920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4090050096123785,
+      "learning_rate": 1.0099279258064214e-05,
+      "loss": 1.1317,
+      "step": 1921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.566615513641779,
+      "learning_rate": 1.0090646526821644e-05,
+      "loss": 1.1158,
+      "step": 1922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8143497551217935,
+      "learning_rate": 1.0082013728019519e-05,
+      "loss": 0.5823,
+      "step": 1923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8534497749549026,
+      "learning_rate": 1.0073380868091932e-05,
+      "loss": 0.6666,
+      "step": 1924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9760465865015864,
+      "learning_rate": 1.0064747953473017e-05,
+      "loss": 0.6223,
+      "step": 1925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3849307319557282,
+      "learning_rate": 1.0056114990596956e-05,
+      "loss": 1.15,
+      "step": 1926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4462850847606084,
+      "learning_rate": 1.0047481985897963e-05,
+      "loss": 1.0897,
+      "step": 1927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7933055333517778,
+      "learning_rate": 1.0038848945810276e-05,
+      "loss": 0.5754,
+      "step": 1928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1334115858895197,
+      "learning_rate": 1.003021587676817e-05,
+      "loss": 0.6752,
+      "step": 1929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8194652354655159,
+      "learning_rate": 1.002158278520594e-05,
+      "loss": 0.6262,
+      "step": 1930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9212745761871044,
+      "learning_rate": 1.0012949677557895e-05,
+      "loss": 0.5601,
+      "step": 1931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7975088021828374,
+      "learning_rate": 1.0004316560258354e-05,
+      "loss": 0.5701,
+      "step": 1932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.824211287080967,
+      "learning_rate": 9.995683439741649e-06,
+      "loss": 0.6531,
+      "step": 1933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9498519375024046,
+      "learning_rate": 9.987050322442108e-06,
+      "loss": 0.6036,
+      "step": 1934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8827652790873695,
+      "learning_rate": 9.978417214794061e-06,
+      "loss": 0.5984,
+      "step": 1935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9552114042004636,
+      "learning_rate": 9.96978412323183e-06,
+      "loss": 0.5986,
+      "step": 1936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7652194205924419,
+      "learning_rate": 9.961151054189729e-06,
+      "loss": 0.636,
+      "step": 1937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.88406518475265,
+      "learning_rate": 9.952518014102042e-06,
+      "loss": 0.6316,
+      "step": 1938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7463504695182037,
+      "learning_rate": 9.943885009403044e-06,
+      "loss": 0.6167,
+      "step": 1939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5700091264542753,
+      "learning_rate": 9.935252046526986e-06,
+      "loss": 1.0903,
+      "step": 1940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6912559051804574,
+      "learning_rate": 9.92661913190807e-06,
+      "loss": 0.5738,
+      "step": 1941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5483483217531804,
+      "learning_rate": 9.917986271980486e-06,
+      "loss": 1.146,
+      "step": 1942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.80885204506323,
+      "learning_rate": 9.90935347317836e-06,
+      "loss": 0.6146,
+      "step": 1943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9644938099204485,
+      "learning_rate": 9.900720741935787e-06,
+      "loss": 0.6228,
+      "step": 1944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.714622259568418,
+      "learning_rate": 9.892088084686815e-06,
+      "loss": 0.623,
+      "step": 1945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8511429209106622,
+      "learning_rate": 9.883455507865419e-06,
+      "loss": 0.6274,
+      "step": 1946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.500967760358848,
+      "learning_rate": 9.87482301790553e-06,
+      "loss": 1.116,
+      "step": 1947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7532153994067556,
+      "learning_rate": 9.866190621241016e-06,
+      "loss": 0.6287,
+      "step": 1948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9823633522908615,
+      "learning_rate": 9.857558324305656e-06,
+      "loss": 0.6235,
+      "step": 1949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6773541258633904,
+      "learning_rate": 9.848926133533176e-06,
+      "loss": 0.6282,
+      "step": 1950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6075821308639273,
+      "learning_rate": 9.84029405535721e-06,
+      "loss": 0.5332,
+      "step": 1951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7087485646116487,
+      "learning_rate": 9.831662096211315e-06,
+      "loss": 0.5712,
+      "step": 1952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7581078424769545,
+      "learning_rate": 9.823030262528956e-06,
+      "loss": 0.6336,
+      "step": 1953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.423782374044007,
+      "learning_rate": 9.814398560743504e-06,
+      "loss": 1.1302,
+      "step": 1954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.835619664955005,
+      "learning_rate": 9.805766997288232e-06,
+      "loss": 0.6131,
+      "step": 1955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7500128474998455,
+      "learning_rate": 9.797135578596318e-06,
+      "loss": 0.6637,
+      "step": 1956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8879136082144106,
+      "learning_rate": 9.788504311100814e-06,
+      "loss": 0.5684,
+      "step": 1957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8550751454369325,
+      "learning_rate": 9.779873201234678e-06,
+      "loss": 0.6683,
+      "step": 1958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3640815496599252,
+      "learning_rate": 9.771242255430738e-06,
+      "loss": 1.2033,
+      "step": 1959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.594295341590083,
+      "learning_rate": 9.762611480121704e-06,
+      "loss": 1.209,
+      "step": 1960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.80166984225749,
+      "learning_rate": 9.753980881740167e-06,
+      "loss": 0.6372,
+      "step": 1961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7058330600077276,
+      "learning_rate": 9.745350466718569e-06,
+      "loss": 0.5547,
+      "step": 1962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6659051952244046,
+      "learning_rate": 9.736720241489226e-06,
+      "loss": 0.5961,
+      "step": 1963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6749755752480078,
+      "learning_rate": 9.72809021248432e-06,
+      "loss": 0.5788,
+      "step": 1964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7985754321845642,
+      "learning_rate": 9.719460386135869e-06,
+      "loss": 0.5792,
+      "step": 1965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.880322128856686,
+      "learning_rate": 9.710830768875753e-06,
+      "loss": 0.6086,
+      "step": 1966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7989707609839947,
+      "learning_rate": 9.70220136713569e-06,
+      "loss": 0.5871,
+      "step": 1967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.911298767885244,
+      "learning_rate": 9.69357218734724e-06,
+      "loss": 0.5998,
+      "step": 1968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5423677232596207,
+      "learning_rate": 9.684943235941799e-06,
+      "loss": 1.1491,
+      "step": 1969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3086325596937507,
+      "learning_rate": 9.676314519350588e-06,
+      "loss": 1.1268,
+      "step": 1970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7808064637985217,
+      "learning_rate": 9.66768604400466e-06,
+      "loss": 0.5383,
+      "step": 1971
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.244952428484954,
+      "learning_rate": 9.659057816334882e-06,
+      "loss": 1.1616,
+      "step": 1972
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7167401345002566,
+      "learning_rate": 9.650429842771934e-06,
+      "loss": 0.574,
+      "step": 1973
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7773180506789275,
+      "learning_rate": 9.641802129746323e-06,
+      "loss": 0.6513,
+      "step": 1974
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5381006965602952,
+      "learning_rate": 9.633174683688336e-06,
+      "loss": 1.1812,
+      "step": 1975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.781136012155204,
+      "learning_rate": 9.62454751102808e-06,
+      "loss": 0.5632,
+      "step": 1976
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.009792569402177,
+      "learning_rate": 9.615920618195462e-06,
+      "loss": 0.6169,
+      "step": 1977
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.381812804677831,
+      "learning_rate": 9.607294011620158e-06,
+      "loss": 1.1219,
+      "step": 1978
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.556050236115859,
+      "learning_rate": 9.59866769773165e-06,
+      "loss": 1.1578,
+      "step": 1979
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8677881657063815,
+      "learning_rate": 9.590041682959201e-06,
+      "loss": 0.5876,
+      "step": 1980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.435834140058692,
+      "learning_rate": 9.581415973731838e-06,
+      "loss": 1.184,
+      "step": 1981
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0448512128089917,
+      "learning_rate": 9.572790576478376e-06,
+      "loss": 0.6266,
+      "step": 1982
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8439887436110682,
+      "learning_rate": 9.564165497627384e-06,
+      "loss": 0.681,
+      "step": 1983
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8176125381219,
+      "learning_rate": 9.5555407436072e-06,
+      "loss": 0.5611,
+      "step": 1984
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0412215656362678,
+      "learning_rate": 9.546916320845923e-06,
+      "loss": 0.65,
+      "step": 1985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7934670871817142,
+      "learning_rate": 9.538292235771402e-06,
+      "loss": 0.619,
+      "step": 1986
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.840901569489654,
+      "learning_rate": 9.529668494811229e-06,
+      "loss": 1.2151,
+      "step": 1987
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4168076015568625,
+      "learning_rate": 9.52104510439275e-06,
+      "loss": 1.1569,
+      "step": 1988
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.789590586639511,
+      "learning_rate": 9.512422070943039e-06,
+      "loss": 0.6307,
+      "step": 1989
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8580997171921836,
+      "learning_rate": 9.503799400888918e-06,
+      "loss": 0.6393,
+      "step": 1990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9001639669820676,
+      "learning_rate": 9.495177100656919e-06,
+      "loss": 0.6572,
+      "step": 1991
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8608103148295192,
+      "learning_rate": 9.486555176673312e-06,
+      "loss": 0.5257,
+      "step": 1992
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8152180147028958,
+      "learning_rate": 9.47793363536409e-06,
+      "loss": 0.6402,
+      "step": 1993
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.831011254377759,
+      "learning_rate": 9.469312483154944e-06,
+      "loss": 0.588,
+      "step": 1994
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9091628278461878,
+      "learning_rate": 9.46069172647129e-06,
+      "loss": 0.6216,
+      "step": 1995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2366612795019924,
+      "learning_rate": 9.452071371738249e-06,
+      "loss": 0.6484,
+      "step": 1996
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.75724978560826,
+      "learning_rate": 9.443451425380627e-06,
+      "loss": 0.5865,
+      "step": 1997
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7640434022841258,
+      "learning_rate": 9.434831893822949e-06,
+      "loss": 0.6121,
+      "step": 1998
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8889823044149647,
+      "learning_rate": 9.426212783489405e-06,
+      "loss": 0.7199,
+      "step": 1999
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8700607404888914,
+      "learning_rate": 9.417594100803891e-06,
+      "loss": 0.6144,
+      "step": 2000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8648322767604566,
+      "learning_rate": 9.408975852189983e-06,
+      "loss": 0.6226,
+      "step": 2001
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7910284863923676,
+      "learning_rate": 9.400358044070917e-06,
+      "loss": 0.6179,
+      "step": 2002
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7546251624469442,
+      "learning_rate": 9.391740682869618e-06,
+      "loss": 0.6541,
+      "step": 2003
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.798215366339828,
+      "learning_rate": 9.383123775008673e-06,
+      "loss": 0.7028,
+      "step": 2004
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8392619495195595,
+      "learning_rate": 9.374507326910327e-06,
+      "loss": 0.6095,
+      "step": 2005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.748995529227339,
+      "learning_rate": 9.365891344996485e-06,
+      "loss": 0.6065,
+      "step": 2006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8848583505589176,
+      "learning_rate": 9.357275835688704e-06,
+      "loss": 0.6069,
+      "step": 2007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1586203047099866,
+      "learning_rate": 9.348660805408191e-06,
+      "loss": 0.6885,
+      "step": 2008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6193065009850096,
+      "learning_rate": 9.340046260575796e-06,
+      "loss": 0.6278,
+      "step": 2009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7584698750614745,
+      "learning_rate": 9.331432207612003e-06,
+      "loss": 0.6168,
+      "step": 2010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5667192098614886,
+      "learning_rate": 9.32281865293693e-06,
+      "loss": 1.0666,
+      "step": 2011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7450614932364974,
+      "learning_rate": 9.314205602970334e-06,
+      "loss": 0.6124,
+      "step": 2012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.321555029368742,
+      "learning_rate": 9.305593064131577e-06,
+      "loss": 1.1409,
+      "step": 2013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8889456382295258,
+      "learning_rate": 9.29698104283966e-06,
+      "loss": 0.6717,
+      "step": 2014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.386314406278732,
+      "learning_rate": 9.288369545513178e-06,
+      "loss": 1.1049,
+      "step": 2015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.876168775389138,
+      "learning_rate": 9.279758578570355e-06,
+      "loss": 0.6363,
+      "step": 2016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8941991185780191,
+      "learning_rate": 9.27114814842901e-06,
+      "loss": 0.5911,
+      "step": 2017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7597065493931139,
+      "learning_rate": 9.26253826150656e-06,
+      "loss": 0.5914,
+      "step": 2018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.83181892121832,
+      "learning_rate": 9.253928924220017e-06,
+      "loss": 0.6127,
+      "step": 2019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5807256882672656,
+      "learning_rate": 9.245320142985996e-06,
+      "loss": 1.1494,
+      "step": 2020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0257098121634427,
+      "learning_rate": 9.236711924220679e-06,
+      "loss": 0.6039,
+      "step": 2021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.843531136190665,
+      "learning_rate": 9.228104274339843e-06,
+      "loss": 0.5933,
+      "step": 2022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7584455178898328,
+      "learning_rate": 9.219497199758832e-06,
+      "loss": 0.6056,
+      "step": 2023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8344422937496483,
+      "learning_rate": 9.210890706892569e-06,
+      "loss": 0.675,
+      "step": 2024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8125697900511522,
+      "learning_rate": 9.202284802155537e-06,
+      "loss": 0.6143,
+      "step": 2025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8363675452859654,
+      "learning_rate": 9.19367949196178e-06,
+      "loss": 0.7355,
+      "step": 2026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3586377963796825,
+      "learning_rate": 9.185074782724905e-06,
+      "loss": 1.1367,
+      "step": 2027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7093168814786592,
+      "learning_rate": 9.176470680858074e-06,
+      "loss": 0.562,
+      "step": 2028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6649729207164472,
+      "learning_rate": 9.167867192773981e-06,
+      "loss": 0.6352,
+      "step": 2029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.941330429191624,
+      "learning_rate": 9.159264324884878e-06,
+      "loss": 0.6415,
+      "step": 2030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.796625827895185,
+      "learning_rate": 9.150662083602541e-06,
+      "loss": 0.5844,
+      "step": 2031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5663932840671084,
+      "learning_rate": 9.142060475338294e-06,
+      "loss": 0.6389,
+      "step": 2032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8965199212864743,
+      "learning_rate": 9.133459506502983e-06,
+      "loss": 0.6008,
+      "step": 2033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6342760415851505,
+      "learning_rate": 9.124859183506968e-06,
+      "loss": 0.5172,
+      "step": 2034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8946638834933396,
+      "learning_rate": 9.116259512760142e-06,
+      "loss": 0.563,
+      "step": 2035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8376831726066185,
+      "learning_rate": 9.107660500671908e-06,
+      "loss": 0.6345,
+      "step": 2036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8534148377234405,
+      "learning_rate": 9.099062153651169e-06,
+      "loss": 0.6039,
+      "step": 2037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7335418702427638,
+      "learning_rate": 9.090464478106346e-06,
+      "loss": 0.5862,
+      "step": 2038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8956397799809757,
+      "learning_rate": 9.081867480445346e-06,
+      "loss": 0.5775,
+      "step": 2039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.823739774307718,
+      "learning_rate": 9.073271167075581e-06,
+      "loss": 0.5525,
+      "step": 2040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5458842758521567,
+      "learning_rate": 9.064675544403952e-06,
+      "loss": 1.132,
+      "step": 2041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7156993641046512,
+      "learning_rate": 9.056080618836838e-06,
+      "loss": 0.6599,
+      "step": 2042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7172052627553989,
+      "learning_rate": 9.047486396780108e-06,
+      "loss": 0.601,
+      "step": 2043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8527680968884686,
+      "learning_rate": 9.0388928846391e-06,
+      "loss": 0.6037,
+      "step": 2044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4285076828845664,
+      "learning_rate": 9.03030008881862e-06,
+      "loss": 1.1329,
+      "step": 2045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.816906654709464,
+      "learning_rate": 9.021708015722957e-06,
+      "loss": 0.5966,
+      "step": 2046
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.544629498616538,
+      "learning_rate": 9.013116671755834e-06,
+      "loss": 1.1239,
+      "step": 2047
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7580606462975186,
+      "learning_rate": 9.004526063320454e-06,
+      "loss": 0.4903,
+      "step": 2048
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3931536211478335,
+      "learning_rate": 8.995936196819467e-06,
+      "loss": 1.1835,
+      "step": 2049
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4050458503906396,
+      "learning_rate": 8.987347078654957e-06,
+      "loss": 1.131,
+      "step": 2050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.291491354796924,
+      "learning_rate": 8.978758715228467e-06,
+      "loss": 1.1145,
+      "step": 2051
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3289873134625876,
+      "learning_rate": 8.970171112940973e-06,
+      "loss": 1.1055,
+      "step": 2052
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.397002763552894,
+      "learning_rate": 8.961584278192873e-06,
+      "loss": 1.1469,
+      "step": 2053
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.309344732623122,
+      "learning_rate": 8.952998217384006e-06,
+      "loss": 1.1034,
+      "step": 2054
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3947705528705194,
+      "learning_rate": 8.944412936913626e-06,
+      "loss": 1.1573,
+      "step": 2055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7764666714501627,
+      "learning_rate": 8.93582844318041e-06,
+      "loss": 0.5932,
+      "step": 2056
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.855876588025043,
+      "learning_rate": 8.927244742582453e-06,
+      "loss": 0.5928,
+      "step": 2057
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8069365458690205,
+      "learning_rate": 8.918661841517241e-06,
+      "loss": 0.5766,
+      "step": 2058
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.020480063020796,
+      "learning_rate": 8.910079746381686e-06,
+      "loss": 0.5698,
+      "step": 2059
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7650360174759103,
+      "learning_rate": 8.90149846357209e-06,
+      "loss": 0.563,
+      "step": 2060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9449408770965084,
+      "learning_rate": 8.892917999484143e-06,
+      "loss": 0.5967,
+      "step": 2061
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2776302682592027,
+      "learning_rate": 8.884338360512934e-06,
+      "loss": 1.1331,
+      "step": 2062
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7190895359493752,
+      "learning_rate": 8.875759553052934e-06,
+      "loss": 0.6434,
+      "step": 2063
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9443614652555672,
+      "learning_rate": 8.867181583497992e-06,
+      "loss": 0.6256,
+      "step": 2064
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8778073768153496,
+      "learning_rate": 8.858604458241343e-06,
+      "loss": 0.5859,
+      "step": 2065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6142720588007924,
+      "learning_rate": 8.850028183675573e-06,
+      "loss": 0.6247,
+      "step": 2066
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7521876936295473,
+      "learning_rate": 8.841452766192652e-06,
+      "loss": 0.6122,
+      "step": 2067
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6651960399129564,
+      "learning_rate": 8.832878212183906e-06,
+      "loss": 0.5836,
+      "step": 2068
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.568956078698448,
+      "learning_rate": 8.824304528040011e-06,
+      "loss": 0.5883,
+      "step": 2069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8991698806928856,
+      "learning_rate": 8.815731720151006e-06,
+      "loss": 0.6021,
+      "step": 2070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.934728564757415,
+      "learning_rate": 8.807159794906262e-06,
+      "loss": 0.679,
+      "step": 2071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.752828412709649,
+      "learning_rate": 8.798588758694507e-06,
+      "loss": 0.6336,
+      "step": 2072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.745849659233598,
+      "learning_rate": 8.790018617903804e-06,
+      "loss": 1.0797,
+      "step": 2073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4266802812589243,
+      "learning_rate": 8.781449378921535e-06,
+      "loss": 1.1789,
+      "step": 2074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.837293608344429,
+      "learning_rate": 8.772881048134424e-06,
+      "loss": 0.6257,
+      "step": 2075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.800666737721762,
+      "learning_rate": 8.764313631928521e-06,
+      "loss": 0.6017,
+      "step": 2076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.759724380391782,
+      "learning_rate": 8.755747136689176e-06,
+      "loss": 0.6278,
+      "step": 2077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8560643212833168,
+      "learning_rate": 8.747181568801072e-06,
+      "loss": 0.5669,
+      "step": 2078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.678676549037832,
+      "learning_rate": 8.738616934648187e-06,
+      "loss": 1.0943,
+      "step": 2079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4440815678351884,
+      "learning_rate": 8.730053240613815e-06,
+      "loss": 1.1804,
+      "step": 2080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8161336432746455,
+      "learning_rate": 8.721490493080537e-06,
+      "loss": 0.7024,
+      "step": 2081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8523402685271444,
+      "learning_rate": 8.712928698430237e-06,
+      "loss": 0.651,
+      "step": 2082
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7982280311795757,
+      "learning_rate": 8.704367863044085e-06,
+      "loss": 0.5854,
+      "step": 2083
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7093153729625599,
+      "learning_rate": 8.695807993302544e-06,
+      "loss": 0.6059,
+      "step": 2084
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4897994388224673,
+      "learning_rate": 8.687249095585342e-06,
+      "loss": 1.1328,
+      "step": 2085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7413507789291667,
+      "learning_rate": 8.6786911762715e-06,
+      "loss": 0.6261,
+      "step": 2086
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.689011550438518,
+      "learning_rate": 8.670134241739291e-06,
+      "loss": 0.5357,
+      "step": 2087
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6670343086615944,
+      "learning_rate": 8.66157829836627e-06,
+      "loss": 0.5875,
+      "step": 2088
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8360328461950144,
+      "learning_rate": 8.653023352529252e-06,
+      "loss": 0.6164,
+      "step": 2089
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.939440075214224,
+      "learning_rate": 8.644469410604295e-06,
+      "loss": 0.5647,
+      "step": 2090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7606533088671827,
+      "learning_rate": 8.635916478966722e-06,
+      "loss": 0.6536,
+      "step": 2091
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.008007129099646,
+      "learning_rate": 8.627364563991102e-06,
+      "loss": 0.6095,
+      "step": 2092
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6251530871232815,
+      "learning_rate": 8.618813672051236e-06,
+      "loss": 1.1578,
+      "step": 2093
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6834542182561723,
+      "learning_rate": 8.610263809520176e-06,
+      "loss": 0.5752,
+      "step": 2094
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.974601647204923,
+      "learning_rate": 8.601714982770197e-06,
+      "loss": 0.6024,
+      "step": 2095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.788846535375357,
+      "learning_rate": 8.593167198172805e-06,
+      "loss": 0.6577,
+      "step": 2096
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7283167241055744,
+      "learning_rate": 8.584620462098732e-06,
+      "loss": 0.6113,
+      "step": 2097
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7675552487669461,
+      "learning_rate": 8.576074780917922e-06,
+      "loss": 0.5964,
+      "step": 2098
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6365759784528606,
+      "learning_rate": 8.567530160999546e-06,
+      "loss": 0.5633,
+      "step": 2099
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.547342887604147,
+      "learning_rate": 8.558986608711963e-06,
+      "loss": 1.1091,
+      "step": 2100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7996406114391894,
+      "learning_rate": 8.550444130422756e-06,
+      "loss": 0.6092,
+      "step": 2101
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.004540694503004,
+      "learning_rate": 8.541902732498701e-06,
+      "loss": 0.5761,
+      "step": 2102
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6362078106390736,
+      "learning_rate": 8.533362421305761e-06,
+      "loss": 0.6075,
+      "step": 2103
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8165199774132534,
+      "learning_rate": 8.5248232032091e-06,
+      "loss": 0.6041,
+      "step": 2104
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.661259385895624,
+      "learning_rate": 8.516285084573066e-06,
+      "loss": 1.146,
+      "step": 2105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7000047123227475,
+      "learning_rate": 8.507748071761176e-06,
+      "loss": 0.5845,
+      "step": 2106
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8642333148290615,
+      "learning_rate": 8.499212171136142e-06,
+      "loss": 0.5589,
+      "step": 2107
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8277576271140263,
+      "learning_rate": 8.490677389059828e-06,
+      "loss": 0.5859,
+      "step": 2108
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.557440996377917,
+      "learning_rate": 8.482143731893272e-06,
+      "loss": 0.5382,
+      "step": 2109
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6818445342977548,
+      "learning_rate": 8.473611205996683e-06,
+      "loss": 0.5148,
+      "step": 2110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8156482915577254,
+      "learning_rate": 8.465079817729411e-06,
+      "loss": 0.6352,
+      "step": 2111
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7452190266145973,
+      "learning_rate": 8.456549573449965e-06,
+      "loss": 0.7112,
+      "step": 2112
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.628818143003967,
+      "learning_rate": 8.448020479516006e-06,
+      "loss": 1.1221,
+      "step": 2113
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7875782945618128,
+      "learning_rate": 8.439492542284334e-06,
+      "loss": 0.6119,
+      "step": 2114
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5425859768823758,
+      "learning_rate": 8.430965768110883e-06,
+      "loss": 1.1532,
+      "step": 2115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8382872907806593,
+      "learning_rate": 8.422440163350723e-06,
+      "loss": 0.5929,
+      "step": 2116
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.536519110139197,
+      "learning_rate": 8.413915734358053e-06,
+      "loss": 1.1214,
+      "step": 2117
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6209681654659565,
+      "learning_rate": 8.405392487486202e-06,
+      "loss": 1.1503,
+      "step": 2118
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7257224290420687,
+      "learning_rate": 8.3968704290876e-06,
+      "loss": 0.5813,
+      "step": 2119
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6665132320922416,
+      "learning_rate": 8.38834956551381e-06,
+      "loss": 0.6051,
+      "step": 2120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6819530164721495,
+      "learning_rate": 8.379829903115502e-06,
+      "loss": 0.5526,
+      "step": 2121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7134817338764106,
+      "learning_rate": 8.371311448242433e-06,
+      "loss": 1.163,
+      "step": 2122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.791753195819067,
+      "learning_rate": 8.362794207243488e-06,
+      "loss": 1.1506,
+      "step": 2123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7990430909257784,
+      "learning_rate": 8.354278186466618e-06,
+      "loss": 0.6248,
+      "step": 2124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.037115513299112,
+      "learning_rate": 8.345763392258885e-06,
+      "loss": 0.6521,
+      "step": 2125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7259855738214258,
+      "learning_rate": 8.337249830966436e-06,
+      "loss": 0.6096,
+      "step": 2126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9481229697910147,
+      "learning_rate": 8.328737508934485e-06,
+      "loss": 0.6375,
+      "step": 2127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7969303018212477,
+      "learning_rate": 8.320226432507333e-06,
+      "loss": 0.5636,
+      "step": 2128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7240062137327592,
+      "learning_rate": 8.31171660802836e-06,
+      "loss": 0.5866,
+      "step": 2129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6600772703049085,
+      "learning_rate": 8.30320804183999e-06,
+      "loss": 1.0738,
+      "step": 2130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8038569258872532,
+      "learning_rate": 8.294700740283732e-06,
+      "loss": 0.6142,
+      "step": 2131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6550073372606744,
+      "learning_rate": 8.286194709700137e-06,
+      "loss": 0.5081,
+      "step": 2132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9067257415576322,
+      "learning_rate": 8.277689956428823e-06,
+      "loss": 0.5638,
+      "step": 2133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.961369898677521,
+      "learning_rate": 8.269186486808443e-06,
+      "loss": 0.6093,
+      "step": 2134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7515143331703416,
+      "learning_rate": 8.260684307176696e-06,
+      "loss": 0.6108,
+      "step": 2135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.445962418087565,
+      "learning_rate": 8.252183423870326e-06,
+      "loss": 1.139,
+      "step": 2136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7300917822163486,
+      "learning_rate": 8.24368384322511e-06,
+      "loss": 0.661,
+      "step": 2137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.515257009673268,
+      "learning_rate": 8.235185571575843e-06,
+      "loss": 1.1558,
+      "step": 2138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7476405613737025,
+      "learning_rate": 8.22668861525636e-06,
+      "loss": 0.6425,
+      "step": 2139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9459186359136498,
+      "learning_rate": 8.2181929805995e-06,
+      "loss": 0.615,
+      "step": 2140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7828304655307363,
+      "learning_rate": 8.209698673937128e-06,
+      "loss": 0.5117,
+      "step": 2141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.939510580739434,
+      "learning_rate": 8.201205701600122e-06,
+      "loss": 0.6106,
+      "step": 2142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4706733447164946,
+      "learning_rate": 8.19271406991835e-06,
+      "loss": 1.0825,
+      "step": 2143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4079072569429987,
+      "learning_rate": 8.184223785220696e-06,
+      "loss": 1.0906,
+      "step": 2144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0048292614900975,
+      "learning_rate": 8.175734853835037e-06,
+      "loss": 0.6318,
+      "step": 2145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8215263791316576,
+      "learning_rate": 8.167247282088232e-06,
+      "loss": 0.5969,
+      "step": 2146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5287024067073984,
+      "learning_rate": 8.15876107630614e-06,
+      "loss": 1.0993,
+      "step": 2147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8616035085789022,
+      "learning_rate": 8.15027624281359e-06,
+      "loss": 0.5954,
+      "step": 2148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.552086912553073,
+      "learning_rate": 8.141792787934397e-06,
+      "loss": 1.1592,
+      "step": 2149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3481987651104244,
+      "learning_rate": 8.133310717991347e-06,
+      "loss": 1.0885,
+      "step": 2150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.587164426634463,
+      "learning_rate": 8.124830039306187e-06,
+      "loss": 1.1175,
+      "step": 2151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9235359927371043,
+      "learning_rate": 8.116350758199636e-06,
+      "loss": 0.6491,
+      "step": 2152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6177100605847428,
+      "learning_rate": 8.107872880991366e-06,
+      "loss": 0.5816,
+      "step": 2153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8095740453736688,
+      "learning_rate": 8.099396414000004e-06,
+      "loss": 0.6047,
+      "step": 2154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.472518336019492,
+      "learning_rate": 8.090921363543129e-06,
+      "loss": 1.1027,
+      "step": 2155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6640082337465616,
+      "learning_rate": 8.082447735937257e-06,
+      "loss": 0.5407,
+      "step": 2156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7897408150515852,
+      "learning_rate": 8.073975537497848e-06,
+      "loss": 0.5772,
+      "step": 2157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9968692250280182,
+      "learning_rate": 8.065504774539304e-06,
+      "loss": 1.1392,
+      "step": 2158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7842492425533896,
+      "learning_rate": 8.057035453374939e-06,
+      "loss": 0.637,
+      "step": 2159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.951753526287782,
+      "learning_rate": 8.048567580317008e-06,
+      "loss": 0.6523,
+      "step": 2160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8931744070415193,
+      "learning_rate": 8.040101161676688e-06,
+      "loss": 0.5991,
+      "step": 2161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5509935386190545,
+      "learning_rate": 8.031636203764056e-06,
+      "loss": 0.6129,
+      "step": 2162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7800096696108236,
+      "learning_rate": 8.02317271288812e-06,
+      "loss": 0.6313,
+      "step": 2163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8999253489078967,
+      "learning_rate": 8.014710695356774e-06,
+      "loss": 0.6128,
+      "step": 2164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7848200632229434,
+      "learning_rate": 8.006250157476832e-06,
+      "loss": 0.6359,
+      "step": 2165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.84956444141318,
+      "learning_rate": 7.997791105553999e-06,
+      "loss": 0.6282,
+      "step": 2166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7731753041484724,
+      "learning_rate": 7.989333545892868e-06,
+      "loss": 0.5759,
+      "step": 2167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8473081129322866,
+      "learning_rate": 7.980877484796923e-06,
+      "loss": 0.5672,
+      "step": 2168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3291734760819414,
+      "learning_rate": 7.972422928568534e-06,
+      "loss": 1.1508,
+      "step": 2169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7724962474011405,
+      "learning_rate": 7.963969883508944e-06,
+      "loss": 0.598,
+      "step": 2170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.524289172313138,
+      "learning_rate": 7.955518355918279e-06,
+      "loss": 1.1141,
+      "step": 2171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8904618782156717,
+      "learning_rate": 7.947068352095519e-06,
+      "loss": 0.6681,
+      "step": 2172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7871168274930251,
+      "learning_rate": 7.93861987833852e-06,
+      "loss": 0.5669,
+      "step": 2173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8588336879317788,
+      "learning_rate": 7.930172940944001e-06,
+      "loss": 0.5755,
+      "step": 2174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6628278344869312,
+      "learning_rate": 7.921727546207518e-06,
+      "loss": 0.605,
+      "step": 2175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8907621555276413,
+      "learning_rate": 7.913283700423495e-06,
+      "loss": 0.5953,
+      "step": 2176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.813287145055464,
+      "learning_rate": 7.9048414098852e-06,
+      "loss": 0.5786,
+      "step": 2177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4186520394640927,
+      "learning_rate": 7.896400680884725e-06,
+      "loss": 1.1969,
+      "step": 2178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8780492172322085,
+      "learning_rate": 7.887961519713024e-06,
+      "loss": 0.5992,
+      "step": 2179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3554154645294845,
+      "learning_rate": 7.879523932659858e-06,
+      "loss": 1.1356,
+      "step": 2180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.839045100893246,
+      "learning_rate": 7.871087926013829e-06,
+      "loss": 0.5998,
+      "step": 2181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.983582340297044,
+      "learning_rate": 7.862653506062362e-06,
+      "loss": 0.6086,
+      "step": 2182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7031763796116854,
+      "learning_rate": 7.854220679091686e-06,
+      "loss": 0.6147,
+      "step": 2183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.462987946462516,
+      "learning_rate": 7.84578945138686e-06,
+      "loss": 1.1302,
+      "step": 2184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9398012070800355,
+      "learning_rate": 7.83735982923174e-06,
+      "loss": 0.6962,
+      "step": 2185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1847339723503545,
+      "learning_rate": 7.82893181890899e-06,
+      "loss": 1.1385,
+      "step": 2186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8314453370119708,
+      "learning_rate": 7.820505426700068e-06,
+      "loss": 0.5981,
+      "step": 2187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8992425671037352,
+      "learning_rate": 7.81208065888523e-06,
+      "loss": 0.6413,
+      "step": 2188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.863661228867366,
+      "learning_rate": 7.80365752174352e-06,
+      "loss": 0.5936,
+      "step": 2189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.378105052303617,
+      "learning_rate": 7.795236021552772e-06,
+      "loss": 1.0954,
+      "step": 2190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7912027271124558,
+      "learning_rate": 7.786816164589586e-06,
+      "loss": 0.5552,
+      "step": 2191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.642845788179634,
+      "learning_rate": 7.778397957129351e-06,
+      "loss": 0.5663,
+      "step": 2192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9262457492764087,
+      "learning_rate": 7.769981405446228e-06,
+      "loss": 0.6781,
+      "step": 2193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5312900761791233,
+      "learning_rate": 7.761566515813123e-06,
+      "loss": 1.1137,
+      "step": 2194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8336508310706685,
+      "learning_rate": 7.753153294501735e-06,
+      "loss": 0.6341,
+      "step": 2195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.120786469087602,
+      "learning_rate": 7.744741747782488e-06,
+      "loss": 0.6428,
+      "step": 2196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2507450337749235,
+      "learning_rate": 7.736331881924578e-06,
+      "loss": 1.1537,
+      "step": 2197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7911409618248262,
+      "learning_rate": 7.727923703195947e-06,
+      "loss": 0.6044,
+      "step": 2198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.357192112240095,
+      "learning_rate": 7.71951721786327e-06,
+      "loss": 1.1469,
+      "step": 2199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8472146615122986,
+      "learning_rate": 7.711112432191966e-06,
+      "loss": 0.554,
+      "step": 2200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7728580204673947,
+      "learning_rate": 7.70270935244619e-06,
+      "loss": 0.5949,
+      "step": 2201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7791584184227713,
+      "learning_rate": 7.694307984888817e-06,
+      "loss": 0.568,
+      "step": 2202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.798932702458932,
+      "learning_rate": 7.685908335781455e-06,
+      "loss": 0.6683,
+      "step": 2203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.453521692534765,
+      "learning_rate": 7.677510411384424e-06,
+      "loss": 1.1528,
+      "step": 2204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4709423649540896,
+      "learning_rate": 7.669114217956765e-06,
+      "loss": 1.1202,
+      "step": 2205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6318595776693012,
+      "learning_rate": 7.660719761756223e-06,
+      "loss": 1.1322,
+      "step": 2206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5678305263659746,
+      "learning_rate": 7.65232704903925e-06,
+      "loss": 1.1179,
+      "step": 2207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.861177402299145,
+      "learning_rate": 7.643936086061002e-06,
+      "loss": 0.5899,
+      "step": 2208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7868323958162922,
+      "learning_rate": 7.63554687907533e-06,
+      "loss": 0.604,
+      "step": 2209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4381210046336594,
+      "learning_rate": 7.627159434334767e-06,
+      "loss": 1.0787,
+      "step": 2210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9227181130105875,
+      "learning_rate": 7.618773758090552e-06,
+      "loss": 0.6188,
+      "step": 2211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7673821442175226,
+      "learning_rate": 7.610389856592581e-06,
+      "loss": 0.5856,
+      "step": 2212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0266838091036714,
+      "learning_rate": 7.602007736089447e-06,
+      "loss": 0.6845,
+      "step": 2213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6679363262477103,
+      "learning_rate": 7.593627402828413e-06,
+      "loss": 1.1167,
+      "step": 2214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.844875783439187,
+      "learning_rate": 7.5852488630553965e-06,
+      "loss": 0.6197,
+      "step": 2215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7851893888900778,
+      "learning_rate": 7.576872123014993e-06,
+      "loss": 0.6005,
+      "step": 2216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7110686562295105,
+      "learning_rate": 7.568497188950454e-06,
+      "loss": 0.5645,
+      "step": 2217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7295728174388647,
+      "learning_rate": 7.560124067103676e-06,
+      "loss": 0.5969,
+      "step": 2218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9610218935481618,
+      "learning_rate": 7.551752763715217e-06,
+      "loss": 0.6843,
+      "step": 2219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3978593163113997,
+      "learning_rate": 7.543383285024269e-06,
+      "loss": 1.122,
+      "step": 2220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4111201189645994,
+      "learning_rate": 7.53501563726867e-06,
+      "loss": 1.1503,
+      "step": 2221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8564174464153234,
+      "learning_rate": 7.526649826684895e-06,
+      "loss": 0.5867,
+      "step": 2222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7563702475005545,
+      "learning_rate": 7.518285859508045e-06,
+      "loss": 0.6555,
+      "step": 2223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.023489403512432,
+      "learning_rate": 7.50992374197185e-06,
+      "loss": 0.5593,
+      "step": 2224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7381694693195038,
+      "learning_rate": 7.501563480308661e-06,
+      "loss": 0.6011,
+      "step": 2225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9090338415319312,
+      "learning_rate": 7.4932050807494415e-06,
+      "loss": 0.6288,
+      "step": 2226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8667879352889556,
+      "learning_rate": 7.48484854952378e-06,
+      "loss": 0.6518,
+      "step": 2227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8927216526269777,
+      "learning_rate": 7.476493892859852e-06,
+      "loss": 0.6279,
+      "step": 2228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.684421533886023,
+      "learning_rate": 7.468141116984453e-06,
+      "loss": 1.1517,
+      "step": 2229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8437496243863833,
+      "learning_rate": 7.459790228122976e-06,
+      "loss": 0.6418,
+      "step": 2230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7732922551222223,
+      "learning_rate": 7.451441232499392e-06,
+      "loss": 0.5275,
+      "step": 2231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8477746539611233,
+      "learning_rate": 7.4430941363362795e-06,
+      "loss": 0.5446,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.372745568842616,
+      "learning_rate": 7.434748945854795e-06,
+      "loss": 1.11,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6234400174731047,
+      "learning_rate": 7.426405667274667e-06,
+      "loss": 0.5351,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4897519226792593,
+      "learning_rate": 7.418064306814211e-06,
+      "loss": 1.1303,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5553649521839747,
+      "learning_rate": 7.4097248706903e-06,
+      "loss": 1.1576,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6205809335081012,
+      "learning_rate": 7.401387365118385e-06,
+      "loss": 0.564,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7156933780445842,
+      "learning_rate": 7.3930517963124735e-06,
+      "loss": 0.5493,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.834441547270687,
+      "learning_rate": 7.38471817048513e-06,
+      "loss": 0.6098,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7458716357172877,
+      "learning_rate": 7.376386493847465e-06,
+      "loss": 0.6089,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.940228875249776,
+      "learning_rate": 7.368056772609145e-06,
+      "loss": 0.6377,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4969761070071574,
+      "learning_rate": 7.359729012978374e-06,
+      "loss": 1.1706,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8842357001884122,
+      "learning_rate": 7.351403221161899e-06,
+      "loss": 0.6756,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.736278049796794,
+      "learning_rate": 7.34307940336499e-06,
+      "loss": 1.1642,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6901818151773778,
+      "learning_rate": 7.334757565791455e-06,
+      "loss": 0.6529,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9536179906954239,
+      "learning_rate": 7.326437714643631e-06,
+      "loss": 0.6171,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6527013483472,
+      "learning_rate": 7.318119856122354e-06,
+      "loss": 1.0783,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8283452537828746,
+      "learning_rate": 7.3098039964269965e-06,
+      "loss": 0.5762,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7536215623419598,
+      "learning_rate": 7.3014901417554365e-06,
+      "loss": 0.6169,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8316446582076948,
+      "learning_rate": 7.293178298304042e-06,
+      "loss": 0.5678,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.388776957725871,
+      "learning_rate": 7.284868472267705e-06,
+      "loss": 1.1595,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5135544967907848,
+      "learning_rate": 7.276560669839796e-06,
+      "loss": 0.5659,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6929996843355697,
+      "learning_rate": 7.268254897212184e-06,
+      "loss": 0.491,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9213877573871232,
+      "learning_rate": 7.2599511605752334e-06,
+      "loss": 0.6099,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9294581618926097,
+      "learning_rate": 7.251649466117771e-06,
+      "loss": 0.6976,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.732393281593926,
+      "learning_rate": 7.2433498200271225e-06,
+      "loss": 0.5942,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7437150984517644,
+      "learning_rate": 7.2350522284890755e-06,
+      "loss": 0.5607,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.947021718051942,
+      "learning_rate": 7.22675669768789e-06,
+      "loss": 0.5319,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7982556251334474,
+      "learning_rate": 7.218463233806288e-06,
+      "loss": 0.5954,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8386069823198772,
+      "learning_rate": 7.210171843025448e-06,
+      "loss": 0.6144,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8532960175744184,
+      "learning_rate": 7.20188253152501e-06,
+      "loss": 0.6396,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5427395150281846,
+      "learning_rate": 7.19359530548307e-06,
+      "loss": 1.0546,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7324965336952067,
+      "learning_rate": 7.1853101710761485e-06,
+      "loss": 0.5749,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6367311267739568,
+      "learning_rate": 7.177027134479226e-06,
+      "loss": 0.5416,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8378778022899835,
+      "learning_rate": 7.1687462018657175e-06,
+      "loss": 0.6015,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.989479130758528,
+      "learning_rate": 7.160467379407456e-06,
+      "loss": 0.5823,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.818967456253556,
+      "learning_rate": 7.152190673274722e-06,
+      "loss": 0.6892,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.351523546150868,
+      "learning_rate": 7.143916089636198e-06,
+      "loss": 1.0886,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.731936109511886,
+      "learning_rate": 7.135643634659e-06,
+      "loss": 0.6487,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9251145319732264,
+      "learning_rate": 7.127373314508658e-06,
+      "loss": 0.5851,
+      "step": 2270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.552045303609423,
+      "learning_rate": 7.119105135349095e-06,
+      "loss": 1.1038,
+      "step": 2271
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8967050443723872,
+      "learning_rate": 7.110839103342653e-06,
+      "loss": 0.5963,
+      "step": 2272
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4839285668857825,
+      "learning_rate": 7.102575224650074e-06,
+      "loss": 1.1093,
+      "step": 2273
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8019340926084781,
+      "learning_rate": 7.094313505430483e-06,
+      "loss": 0.6078,
+      "step": 2274
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3839049112234356,
+      "learning_rate": 7.086053951841407e-06,
+      "loss": 1.1355,
+      "step": 2275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8383537728553792,
+      "learning_rate": 7.077796570038754e-06,
+      "loss": 0.6292,
+      "step": 2276
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3644864145871,
+      "learning_rate": 7.069541366176812e-06,
+      "loss": 1.1451,
+      "step": 2277
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2488929770320776,
+      "learning_rate": 7.0612883464082524e-06,
+      "loss": 1.1386,
+      "step": 2278
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7902396018350628,
+      "learning_rate": 7.0530375168841095e-06,
+      "loss": 0.5773,
+      "step": 2279
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7548095011535527,
+      "learning_rate": 7.044788883753793e-06,
+      "loss": 0.6064,
+      "step": 2280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7630337780575949,
+      "learning_rate": 7.036542453165076e-06,
+      "loss": 0.6171,
+      "step": 2281
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.667623852058585,
+      "learning_rate": 7.0282982312640805e-06,
+      "loss": 0.5848,
+      "step": 2282
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.405738213611042,
+      "learning_rate": 7.020056224195293e-06,
+      "loss": 1.0611,
+      "step": 2283
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7984490193575282,
+      "learning_rate": 7.01181643810154e-06,
+      "loss": 0.5843,
+      "step": 2284
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.657153822239189,
+      "learning_rate": 7.003578879124e-06,
+      "loss": 0.6565,
+      "step": 2285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9623812362573492,
+      "learning_rate": 6.995343553402194e-06,
+      "loss": 0.6778,
+      "step": 2286
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7944128100551853,
+      "learning_rate": 6.987110467073965e-06,
+      "loss": 0.6238,
+      "step": 2287
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7996486828880256,
+      "learning_rate": 6.9788796262755e-06,
+      "loss": 0.5309,
+      "step": 2288
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6181172537354604,
+      "learning_rate": 6.9706510371413115e-06,
+      "loss": 1.1403,
+      "step": 2289
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9035847170399451,
+      "learning_rate": 6.962424705804223e-06,
+      "loss": 0.679,
+      "step": 2290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9776726329613932,
+      "learning_rate": 6.954200638395389e-06,
+      "loss": 0.685,
+      "step": 2291
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8072764984449203,
+      "learning_rate": 6.945978841044265e-06,
+      "loss": 0.6021,
+      "step": 2292
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8279390732590965,
+      "learning_rate": 6.9377593198786216e-06,
+      "loss": 0.6094,
+      "step": 2293
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.617079946126471,
+      "learning_rate": 6.929542081024534e-06,
+      "loss": 1.1464,
+      "step": 2294
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8837871973744156,
+      "learning_rate": 6.921327130606372e-06,
+      "loss": 0.5795,
+      "step": 2295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8395848639384613,
+      "learning_rate": 6.913114474746798e-06,
+      "loss": 0.607,
+      "step": 2296
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9784964120185762,
+      "learning_rate": 6.9049041195667734e-06,
+      "loss": 0.6477,
+      "step": 2297
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8270354999812362,
+      "learning_rate": 6.8966960711855335e-06,
+      "loss": 0.5622,
+      "step": 2298
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.366584937735813,
+      "learning_rate": 6.888490335720608e-06,
+      "loss": 1.1221,
+      "step": 2299
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2262033529564564,
+      "learning_rate": 6.880286919287784e-06,
+      "loss": 1.1269,
+      "step": 2300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.491157031733425,
+      "learning_rate": 6.872085828001136e-06,
+      "loss": 1.1323,
+      "step": 2301
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.916406131214398,
+      "learning_rate": 6.8638870679730065e-06,
+      "loss": 0.5744,
+      "step": 2302
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.855099014731594,
+      "learning_rate": 6.855690645313983e-06,
+      "loss": 0.6022,
+      "step": 2303
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.839662908218264,
+      "learning_rate": 6.84749656613293e-06,
+      "loss": 0.6476,
+      "step": 2304
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3161348322529407,
+      "learning_rate": 6.83930483653696e-06,
+      "loss": 1.1076,
+      "step": 2305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8301417070173862,
+      "learning_rate": 6.8311154626314225e-06,
+      "loss": 0.5876,
+      "step": 2306
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8022241722715298,
+      "learning_rate": 6.8229284505199325e-06,
+      "loss": 0.5947,
+      "step": 2307
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8939727044602062,
+      "learning_rate": 6.81474380630432e-06,
+      "loss": 0.5904,
+      "step": 2308
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.743149148730459,
+      "learning_rate": 6.806561536084674e-06,
+      "loss": 0.5382,
+      "step": 2309
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7888535215802457,
+      "learning_rate": 6.798381645959302e-06,
+      "loss": 0.6174,
+      "step": 2310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7908736622134325,
+      "learning_rate": 6.790204142024735e-06,
+      "loss": 0.6061,
+      "step": 2311
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5506714774932644,
+      "learning_rate": 6.7820290303757295e-06,
+      "loss": 1.106,
+      "step": 2312
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9415705585315508,
+      "learning_rate": 6.773856317105265e-06,
+      "loss": 0.65,
+      "step": 2313
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9233731381253454,
+      "learning_rate": 6.765686008304523e-06,
+      "loss": 0.6108,
+      "step": 2314
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8426498656051185,
+      "learning_rate": 6.7575181100629e-06,
+      "loss": 0.6337,
+      "step": 2315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.30002081256518,
+      "learning_rate": 6.749352628467988e-06,
+      "loss": 1.0537,
+      "step": 2316
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5034042503103713,
+      "learning_rate": 6.7411895696055884e-06,
+      "loss": 1.0693,
+      "step": 2317
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6310191682131954,
+      "learning_rate": 6.733028939559694e-06,
+      "loss": 0.6024,
+      "step": 2318
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8033527878271336,
+      "learning_rate": 6.724870744412479e-06,
+      "loss": 0.6063,
+      "step": 2319
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7835583225736475,
+      "learning_rate": 6.7167149902443105e-06,
+      "loss": 0.6263,
+      "step": 2320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7451214194689653,
+      "learning_rate": 6.708561683133743e-06,
+      "loss": 0.5174,
+      "step": 2321
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8006140641432888,
+      "learning_rate": 6.700410829157487e-06,
+      "loss": 0.5889,
+      "step": 2322
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5147411462169584,
+      "learning_rate": 6.692262434390445e-06,
+      "loss": 1.096,
+      "step": 2323
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7662734383406347,
+      "learning_rate": 6.684116504905672e-06,
+      "loss": 0.6322,
+      "step": 2324
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4270976877131205,
+      "learning_rate": 6.675973046774396e-06,
+      "loss": 1.1058,
+      "step": 2325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6816095992414606,
+      "learning_rate": 6.6678320660660025e-06,
+      "loss": 0.5908,
+      "step": 2326
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6095372651507427,
+      "learning_rate": 6.659693568848018e-06,
+      "loss": 1.1415,
+      "step": 2327
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8315940630501573,
+      "learning_rate": 6.651557561186133e-06,
+      "loss": 0.6019,
+      "step": 2328
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9613947050108682,
+      "learning_rate": 6.64342404914418e-06,
+      "loss": 0.5779,
+      "step": 2329
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.41543575815019,
+      "learning_rate": 6.635293038784121e-06,
+      "loss": 1.1397,
+      "step": 2330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.110397996634484,
+      "learning_rate": 6.627164536166064e-06,
+      "loss": 0.7091,
+      "step": 2331
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3629339266531932,
+      "learning_rate": 6.619038547348242e-06,
+      "loss": 1.1395,
+      "step": 2332
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8017728293752582,
+      "learning_rate": 6.610915078387022e-06,
+      "loss": 0.6302,
+      "step": 2333
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8188783759015221,
+      "learning_rate": 6.6027941353368855e-06,
+      "loss": 0.6394,
+      "step": 2334
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7360375210960912,
+      "learning_rate": 6.59467572425043e-06,
+      "loss": 0.58,
+      "step": 2335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9920262648366844,
+      "learning_rate": 6.586559851178373e-06,
+      "loss": 0.5915,
+      "step": 2336
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.579331705043958,
+      "learning_rate": 6.578446522169543e-06,
+      "loss": 1.111,
+      "step": 2337
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4053936697507896,
+      "learning_rate": 6.570335743270857e-06,
+      "loss": 1.0477,
+      "step": 2338
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7131896355405367,
+      "learning_rate": 6.562227520527349e-06,
+      "loss": 0.6359,
+      "step": 2339
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.779119352933402,
+      "learning_rate": 6.5541218599821324e-06,
+      "loss": 0.533,
+      "step": 2340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7311722132370893,
+      "learning_rate": 6.546018767676421e-06,
+      "loss": 0.5968,
+      "step": 2341
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5548220625509517,
+      "learning_rate": 6.537918249649516e-06,
+      "loss": 1.0907,
+      "step": 2342
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0021615244038604,
+      "learning_rate": 6.5298203119387906e-06,
+      "loss": 0.5986,
+      "step": 2343
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7795883387483589,
+      "learning_rate": 6.5217249605797005e-06,
+      "loss": 0.6179,
+      "step": 2344
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8411468201319037,
+      "learning_rate": 6.513632201605779e-06,
+      "loss": 0.5825,
+      "step": 2345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8585516386749714,
+      "learning_rate": 6.505542041048614e-06,
+      "loss": 0.5952,
+      "step": 2346
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.834923982638499,
+      "learning_rate": 6.497454484937871e-06,
+      "loss": 0.628,
+      "step": 2347
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.384924183471111,
+      "learning_rate": 6.489369539301262e-06,
+      "loss": 1.177,
+      "step": 2348
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9108114102552785,
+      "learning_rate": 6.481287210164563e-06,
+      "loss": 0.6342,
+      "step": 2349
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.195385890330883,
+      "learning_rate": 6.473207503551597e-06,
+      "loss": 1.0977,
+      "step": 2350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7431930602304475,
+      "learning_rate": 6.4651304254842274e-06,
+      "loss": 0.6359,
+      "step": 2351
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9826474838932533,
+      "learning_rate": 6.457055981982367e-06,
+      "loss": 0.5963,
+      "step": 2352
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8999370511792415,
+      "learning_rate": 6.4489841790639585e-06,
+      "loss": 0.591,
+      "step": 2353
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7771968237962976,
+      "learning_rate": 6.440915022744978e-06,
+      "loss": 0.5818,
+      "step": 2354
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2907272651829484,
+      "learning_rate": 6.4328485190394344e-06,
+      "loss": 1.1203,
+      "step": 2355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2784246896332174,
+      "learning_rate": 6.424784673959347e-06,
+      "loss": 1.1138,
+      "step": 2356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.349214763347442,
+      "learning_rate": 6.416723493514766e-06,
+      "loss": 1.1006,
+      "step": 2357
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.85326770217161,
+      "learning_rate": 6.408664983713755e-06,
+      "loss": 0.5988,
+      "step": 2358
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.86874777949982,
+      "learning_rate": 6.400609150562378e-06,
+      "loss": 0.6414,
+      "step": 2359
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8133645796296518,
+      "learning_rate": 6.392556000064709e-06,
+      "loss": 0.5931,
+      "step": 2360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6756255555300315,
+      "learning_rate": 6.384505538222829e-06,
+      "loss": 0.6216,
+      "step": 2361
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5892566635492407,
+      "learning_rate": 6.376457771036804e-06,
+      "loss": 1.0699,
+      "step": 2362
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4339736364244424,
+      "learning_rate": 6.3684127045047006e-06,
+      "loss": 1.1411,
+      "step": 2363
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8340991932402722,
+      "learning_rate": 6.360370344622562e-06,
+      "loss": 0.6213,
+      "step": 2364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9347629315599073,
+      "learning_rate": 6.352330697384426e-06,
+      "loss": 0.6572,
+      "step": 2365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.709259336670832,
+      "learning_rate": 6.344293768782306e-06,
+      "loss": 0.5883,
+      "step": 2366
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0205550096133162,
+      "learning_rate": 6.336259564806182e-06,
+      "loss": 0.5888,
+      "step": 2367
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.846167192165536,
+      "learning_rate": 6.328228091444009e-06,
+      "loss": 0.5741,
+      "step": 2368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9993954572853085,
+      "learning_rate": 6.320199354681708e-06,
+      "loss": 0.6115,
+      "step": 2369
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1172344801871885,
+      "learning_rate": 6.3121733605031564e-06,
+      "loss": 0.5543,
+      "step": 2370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5407530871569834,
+      "learning_rate": 6.304150114890194e-06,
+      "loss": 1.069,
+      "step": 2371
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.215250346258818,
+      "learning_rate": 6.2961296238226e-06,
+      "loss": 1.0469,
+      "step": 2372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8591140492670646,
+      "learning_rate": 6.28811189327811e-06,
+      "loss": 0.5956,
+      "step": 2373
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9351338591033673,
+      "learning_rate": 6.280096929232408e-06,
+      "loss": 0.7392,
+      "step": 2374
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3274327437983002,
+      "learning_rate": 6.272084737659096e-06,
+      "loss": 1.1362,
+      "step": 2375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3049046070148695,
+      "learning_rate": 6.264075324529732e-06,
+      "loss": 1.0876,
+      "step": 2376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.350714751884907,
+      "learning_rate": 6.256068695813792e-06,
+      "loss": 1.1688,
+      "step": 2377
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9303721949727148,
+      "learning_rate": 6.248064857478672e-06,
+      "loss": 0.6319,
+      "step": 2378
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9435195182675793,
+      "learning_rate": 6.240063815489703e-06,
+      "loss": 0.646,
+      "step": 2379
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7890708214354112,
+      "learning_rate": 6.2320655758101135e-06,
+      "loss": 0.4969,
+      "step": 2380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5991642537892952,
+      "learning_rate": 6.224070144401058e-06,
+      "loss": 0.4902,
+      "step": 2381
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.890904037626853,
+      "learning_rate": 6.216077527221598e-06,
+      "loss": 0.6349,
+      "step": 2382
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.736867160605311,
+      "learning_rate": 6.208087730228682e-06,
+      "loss": 0.5626,
+      "step": 2383
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6925559466972953,
+      "learning_rate": 6.2001007593771745e-06,
+      "loss": 0.5844,
+      "step": 2384
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.965643491795176,
+      "learning_rate": 6.192116620619826e-06,
+      "loss": 0.6809,
+      "step": 2385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8210143015090352,
+      "learning_rate": 6.184135319907276e-06,
+      "loss": 0.6158,
+      "step": 2386
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.677679088291029,
+      "learning_rate": 6.176156863188047e-06,
+      "loss": 0.6119,
+      "step": 2387
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8069590784957104,
+      "learning_rate": 6.168181256408546e-06,
+      "loss": 0.6218,
+      "step": 2388
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.676886477090964,
+      "learning_rate": 6.160208505513054e-06,
+      "loss": 1.1619,
+      "step": 2389
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.074037390131822,
+      "learning_rate": 6.152238616443726e-06,
+      "loss": 0.5708,
+      "step": 2390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.800952356433981,
+      "learning_rate": 6.144271595140577e-06,
+      "loss": 0.6433,
+      "step": 2391
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.408891509068325,
+      "learning_rate": 6.136307447541492e-06,
+      "loss": 1.1,
+      "step": 2392
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8090502767361105,
+      "learning_rate": 6.128346179582214e-06,
+      "loss": 0.6178,
+      "step": 2393
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7914457464813918,
+      "learning_rate": 6.120387797196331e-06,
+      "loss": 0.5364,
+      "step": 2394
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.802065110780784,
+      "learning_rate": 6.112432306315295e-06,
+      "loss": 0.582,
+      "step": 2395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8600750946097377,
+      "learning_rate": 6.104479712868384e-06,
+      "loss": 0.6544,
+      "step": 2396
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7935833448041167,
+      "learning_rate": 6.096530022782734e-06,
+      "loss": 0.6056,
+      "step": 2397
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2904963324145746,
+      "learning_rate": 6.088583241983312e-06,
+      "loss": 1.1625,
+      "step": 2398
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0002789776711283,
+      "learning_rate": 6.080639376392909e-06,
+      "loss": 0.6493,
+      "step": 2399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6789994373684831,
+      "learning_rate": 6.072698431932152e-06,
+      "loss": 0.6102,
+      "step": 2400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.601117163308175,
+      "learning_rate": 6.064760414519493e-06,
+      "loss": 0.6175,
+      "step": 2401
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7903341304045624,
+      "learning_rate": 6.0568253300711876e-06,
+      "loss": 0.6012,
+      "step": 2402
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9314038635155848,
+      "learning_rate": 6.048893184501323e-06,
+      "loss": 0.5532,
+      "step": 2403
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4149477299026474,
+      "learning_rate": 6.040963983721784e-06,
+      "loss": 1.091,
+      "step": 2404
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.980180576359981,
+      "learning_rate": 6.033037733642269e-06,
+      "loss": 0.6298,
+      "step": 2405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8235994652946788,
+      "learning_rate": 6.02511444017027e-06,
+      "loss": 0.6323,
+      "step": 2406
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6864308381841817,
+      "learning_rate": 6.017194109211078e-06,
+      "loss": 0.583,
+      "step": 2407
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6462438129181525,
+      "learning_rate": 6.009276746667779e-06,
+      "loss": 0.5369,
+      "step": 2408
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.802266825190394,
+      "learning_rate": 6.001362358441247e-06,
+      "loss": 0.6119,
+      "step": 2409
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9989048660858322,
+      "learning_rate": 5.993450950430129e-06,
+      "loss": 0.5865,
+      "step": 2410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8181581779203466,
+      "learning_rate": 5.985542528530867e-06,
+      "loss": 0.6018,
+      "step": 2411
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7027542765524308,
+      "learning_rate": 5.977637098637659e-06,
+      "loss": 0.5963,
+      "step": 2412
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.928467268491428,
+      "learning_rate": 5.969734666642488e-06,
+      "loss": 0.5748,
+      "step": 2413
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8545229820293025,
+      "learning_rate": 5.961835238435102e-06,
+      "loss": 0.5649,
+      "step": 2414
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.724889361189641,
+      "learning_rate": 5.953938819902996e-06,
+      "loss": 0.5583,
+      "step": 2415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8105938657537044,
+      "learning_rate": 5.946045416931436e-06,
+      "loss": 0.5821,
+      "step": 2416
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6720252924976635,
+      "learning_rate": 5.938155035403442e-06,
+      "loss": 0.5775,
+      "step": 2417
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7623080578654915,
+      "learning_rate": 5.930267681199765e-06,
+      "loss": 0.5962,
+      "step": 2418
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3825718255384425,
+      "learning_rate": 5.922383360198918e-06,
+      "loss": 1.0977,
+      "step": 2419
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3812509307592364,
+      "learning_rate": 5.914502078277146e-06,
+      "loss": 1.0707,
+      "step": 2420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8024373826087057,
+      "learning_rate": 5.906623841308424e-06,
+      "loss": 0.6285,
+      "step": 2421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7889857132277929,
+      "learning_rate": 5.898748655164469e-06,
+      "loss": 0.5873,
+      "step": 2422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.029952117675518,
+      "learning_rate": 5.890876525714713e-06,
+      "loss": 0.6218,
+      "step": 2423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.530282280396807,
+      "learning_rate": 5.883007458826319e-06,
+      "loss": 1.1639,
+      "step": 2424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6756372970300255,
+      "learning_rate": 5.87514146036416e-06,
+      "loss": 1.1156,
+      "step": 2425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4103372864033425,
+      "learning_rate": 5.867278536190824e-06,
+      "loss": 1.175,
+      "step": 2426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9048288236077997,
+      "learning_rate": 5.859418692166616e-06,
+      "loss": 0.6401,
+      "step": 2427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7306820510981153,
+      "learning_rate": 5.851561934149528e-06,
+      "loss": 0.6381,
+      "step": 2428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7367169132141493,
+      "learning_rate": 5.843708267995268e-06,
+      "loss": 0.5526,
+      "step": 2429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.831259886873753,
+      "learning_rate": 5.835857699557236e-06,
+      "loss": 0.4929,
+      "step": 2430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6280588157800824,
+      "learning_rate": 5.828010234686519e-06,
+      "loss": 1.1335,
+      "step": 2431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7806201835748667,
+      "learning_rate": 5.820165879231891e-06,
+      "loss": 0.6457,
+      "step": 2432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7837629985765706,
+      "learning_rate": 5.812324639039812e-06,
+      "loss": 0.6271,
+      "step": 2433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8294897966618997,
+      "learning_rate": 5.804486519954414e-06,
+      "loss": 0.6,
+      "step": 2434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9533351220836903,
+      "learning_rate": 5.7966515278175206e-06,
+      "loss": 0.6126,
+      "step": 2435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4068679012731464,
+      "learning_rate": 5.788819668468596e-06,
+      "loss": 1.1147,
+      "step": 2436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.865115127102477,
+      "learning_rate": 5.780990947744796e-06,
+      "loss": 0.6372,
+      "step": 2437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9702394777986993,
+      "learning_rate": 5.773165371480923e-06,
+      "loss": 0.6434,
+      "step": 2438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7836089674423594,
+      "learning_rate": 5.76534294550944e-06,
+      "loss": 0.6374,
+      "step": 2439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6659350903949586,
+      "learning_rate": 5.7575236756604595e-06,
+      "loss": 0.604,
+      "step": 2440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.676102130564393,
+      "learning_rate": 5.7497075677617446e-06,
+      "loss": 0.5505,
+      "step": 2441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.57475518058731,
+      "learning_rate": 5.741894627638695e-06,
+      "loss": 1.1945,
+      "step": 2442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8247057072703297,
+      "learning_rate": 5.7340848611143705e-06,
+      "loss": 0.597,
+      "step": 2443
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.928083002991912,
+      "learning_rate": 5.726278274009434e-06,
+      "loss": 0.6765,
+      "step": 2444
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6344941066914493,
+      "learning_rate": 5.718474872142203e-06,
+      "loss": 0.5642,
+      "step": 2445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7866430556523605,
+      "learning_rate": 5.710674661328612e-06,
+      "loss": 0.5197,
+      "step": 2446
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3825562211062845,
+      "learning_rate": 5.702877647382217e-06,
+      "loss": 1.1382,
+      "step": 2447
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6932649693378043,
+      "learning_rate": 5.695083836114194e-06,
+      "loss": 0.6006,
+      "step": 2448
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9398387825149253,
+      "learning_rate": 5.6872932333333245e-06,
+      "loss": 1.0492,
+      "step": 2449
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7269216055028542,
+      "learning_rate": 5.679505844846015e-06,
+      "loss": 0.5921,
+      "step": 2450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4496147148847482,
+      "learning_rate": 5.67172167645626e-06,
+      "loss": 1.1545,
+      "step": 2451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7307400922794323,
+      "learning_rate": 5.663940733965662e-06,
+      "loss": 0.6015,
+      "step": 2452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4637056943547258,
+      "learning_rate": 5.656163023173411e-06,
+      "loss": 1.1348,
+      "step": 2453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3738849431840485,
+      "learning_rate": 5.648388549876309e-06,
+      "loss": 1.1122,
+      "step": 2454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3978651186934545,
+      "learning_rate": 5.640617319868714e-06,
+      "loss": 1.1152,
+      "step": 2455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.861737459074332,
+      "learning_rate": 5.6328493389426e-06,
+      "loss": 0.5064,
+      "step": 2456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.359610890420577,
+      "learning_rate": 5.625084612887486e-06,
+      "loss": 1.1185,
+      "step": 2457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9002159301223593,
+      "learning_rate": 5.6173231474904945e-06,
+      "loss": 0.591,
+      "step": 2458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7958954977519812,
+      "learning_rate": 5.609564948536301e-06,
+      "loss": 0.6352,
+      "step": 2459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.867373760121339,
+      "learning_rate": 5.601810021807151e-06,
+      "loss": 0.614,
+      "step": 2460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.823198969733031,
+      "learning_rate": 5.594058373082849e-06,
+      "loss": 0.5953,
+      "step": 2461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9302441987359507,
+      "learning_rate": 5.586310008140768e-06,
+      "loss": 0.6518,
+      "step": 2462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8902069413221336,
+      "learning_rate": 5.578564932755809e-06,
+      "loss": 0.6905,
+      "step": 2463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7866704039924681,
+      "learning_rate": 5.570823152700448e-06,
+      "loss": 0.6248,
+      "step": 2464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.753085184602148,
+      "learning_rate": 5.56308467374469e-06,
+      "loss": 0.5814,
+      "step": 2465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3349230475902147,
+      "learning_rate": 5.555349501656082e-06,
+      "loss": 1.1231,
+      "step": 2466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7974184624937808,
+      "learning_rate": 5.5476176421997095e-06,
+      "loss": 0.5746,
+      "step": 2467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7449778257763575,
+      "learning_rate": 5.539889101138179e-06,
+      "loss": 0.6594,
+      "step": 2468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.799867043930739,
+      "learning_rate": 5.532163884231642e-06,
+      "loss": 0.6023,
+      "step": 2469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8502337230386139,
+      "learning_rate": 5.52444199723776e-06,
+      "loss": 0.6815,
+      "step": 2470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.654941240060386,
+      "learning_rate": 5.516723445911711e-06,
+      "loss": 0.5838,
+      "step": 2471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.987845287573239,
+      "learning_rate": 5.509008236006195e-06,
+      "loss": 0.6492,
+      "step": 2472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3370102020774985,
+      "learning_rate": 5.501296373271412e-06,
+      "loss": 1.125,
+      "step": 2473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9046113614546163,
+      "learning_rate": 5.4935878634550735e-06,
+      "loss": 0.6033,
+      "step": 2474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7963894294639995,
+      "learning_rate": 5.4858827123023995e-06,
+      "loss": 0.6353,
+      "step": 2475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.887678158143403,
+      "learning_rate": 5.478180925556084e-06,
+      "loss": 0.6851,
+      "step": 2476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.717941981276722,
+      "learning_rate": 5.470482508956343e-06,
+      "loss": 0.4844,
+      "step": 2477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7960829134626952,
+      "learning_rate": 5.462787468240857e-06,
+      "loss": 0.6424,
+      "step": 2478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.264089359299886,
+      "learning_rate": 5.455095809144801e-06,
+      "loss": 1.1161,
+      "step": 2479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9305216148421065,
+      "learning_rate": 5.447407537400827e-06,
+      "loss": 0.6393,
+      "step": 2480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9359728633443083,
+      "learning_rate": 5.439722658739065e-06,
+      "loss": 0.6646,
+      "step": 2481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4558976846408913,
+      "learning_rate": 5.4320411788871085e-06,
+      "loss": 1.0635,
+      "step": 2482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.702649624310225,
+      "learning_rate": 5.424363103570032e-06,
+      "loss": 0.57,
+      "step": 2483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7948965475698355,
+      "learning_rate": 5.41668843851036e-06,
+      "loss": 0.5475,
+      "step": 2484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7725173587036955,
+      "learning_rate": 5.409017189428081e-06,
+      "loss": 0.6187,
+      "step": 2485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.72768805330029,
+      "learning_rate": 5.401349362040633e-06,
+      "loss": 0.5888,
+      "step": 2486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9070435895513915,
+      "learning_rate": 5.393684962062905e-06,
+      "loss": 0.6403,
+      "step": 2487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0513585985995246,
+      "learning_rate": 5.386023995207245e-06,
+      "loss": 0.5697,
+      "step": 2488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8206332723682503,
+      "learning_rate": 5.378366467183414e-06,
+      "loss": 0.647,
+      "step": 2489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.413964556040643,
+      "learning_rate": 5.370712383698639e-06,
+      "loss": 1.1078,
+      "step": 2490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9788346273935655,
+      "learning_rate": 5.363061750457563e-06,
+      "loss": 0.5719,
+      "step": 2491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7278346783927918,
+      "learning_rate": 5.355414573162263e-06,
+      "loss": 0.5972,
+      "step": 2492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8327857653711228,
+      "learning_rate": 5.347770857512233e-06,
+      "loss": 0.616,
+      "step": 2493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8757710012525297,
+      "learning_rate": 5.340130609204408e-06,
+      "loss": 0.601,
+      "step": 2494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2612140486578975,
+      "learning_rate": 5.332493833933105e-06,
+      "loss": 1.0905,
+      "step": 2495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7811295316907185,
+      "learning_rate": 5.324860537390088e-06,
+      "loss": 0.6056,
+      "step": 2496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7461291094872091,
+      "learning_rate": 5.3172307252645e-06,
+      "loss": 0.5829,
+      "step": 2497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6336051335841906,
+      "learning_rate": 5.309604403242903e-06,
+      "loss": 0.5159,
+      "step": 2498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.292750264543701,
+      "learning_rate": 5.301981577009255e-06,
+      "loss": 1.0423,
+      "step": 2499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7735602745714012,
+      "learning_rate": 5.2943622522449055e-06,
+      "loss": 0.5237,
+      "step": 2500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.657731078799844,
+      "learning_rate": 5.286746434628593e-06,
+      "loss": 0.6376,
+      "step": 2501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3989744114663516,
+      "learning_rate": 5.2791341298364506e-06,
+      "loss": 1.1106,
+      "step": 2502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4393014537637683,
+      "learning_rate": 5.2715253435419854e-06,
+      "loss": 1.1288,
+      "step": 2503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2576425106188394,
+      "learning_rate": 5.2639200814160855e-06,
+      "loss": 1.1143,
+      "step": 2504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.612078167782831,
+      "learning_rate": 5.256318349127008e-06,
+      "loss": 1.1427,
+      "step": 2505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7686187685433437,
+      "learning_rate": 5.248720152340381e-06,
+      "loss": 0.6236,
+      "step": 2506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8385530260860057,
+      "learning_rate": 5.2411254967192104e-06,
+      "loss": 0.5536,
+      "step": 2507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7597389397281133,
+      "learning_rate": 5.233534387923835e-06,
+      "loss": 0.5694,
+      "step": 2508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.790133759321074,
+      "learning_rate": 5.225946831611981e-06,
+      "loss": 0.509,
+      "step": 2509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4359210426789533,
+      "learning_rate": 5.218362833438706e-06,
+      "loss": 1.1469,
+      "step": 2510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8929090797423571,
+      "learning_rate": 5.210782399056422e-06,
+      "loss": 0.5496,
+      "step": 2511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9248564271751312,
+      "learning_rate": 5.203205534114887e-06,
+      "loss": 0.6718,
+      "step": 2512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9262296366979768,
+      "learning_rate": 5.195632244261195e-06,
+      "loss": 0.6038,
+      "step": 2513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6613783330584833,
+      "learning_rate": 5.188062535139775e-06,
+      "loss": 0.5637,
+      "step": 2514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8328372413098666,
+      "learning_rate": 5.180496412392401e-06,
+      "loss": 0.6228,
+      "step": 2515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9547328875327457,
+      "learning_rate": 5.172933881658144e-06,
+      "loss": 0.617,
+      "step": 2516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7086613017152368,
+      "learning_rate": 5.165374948573433e-06,
+      "loss": 0.504,
+      "step": 2517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6674661242407156,
+      "learning_rate": 5.157819618771991e-06,
+      "loss": 0.6144,
+      "step": 2518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8836367224587398,
+      "learning_rate": 5.150267897884864e-06,
+      "loss": 0.599,
+      "step": 2519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8549259118012436,
+      "learning_rate": 5.1427197915404115e-06,
+      "loss": 0.6147,
+      "step": 2520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5449938720843512,
+      "learning_rate": 5.135175305364286e-06,
+      "loss": 1.078,
+      "step": 2521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.728922012181704,
+      "learning_rate": 5.127634444979462e-06,
+      "loss": 0.5847,
+      "step": 2522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.924622688749933,
+      "learning_rate": 5.1200972160061964e-06,
+      "loss": 0.5724,
+      "step": 2523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6395122039548986,
+      "learning_rate": 5.1125636240620445e-06,
+      "loss": 0.5183,
+      "step": 2524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.767398193795481,
+      "learning_rate": 5.105033674761845e-06,
+      "loss": 0.5909,
+      "step": 2525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.753198053296005,
+      "learning_rate": 5.0975073737177404e-06,
+      "loss": 0.6051,
+      "step": 2526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7394524594284821,
+      "learning_rate": 5.089984726539126e-06,
+      "loss": 0.6192,
+      "step": 2527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8631820396529502,
+      "learning_rate": 5.0824657388327e-06,
+      "loss": 0.6294,
+      "step": 2528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.322279120515422,
+      "learning_rate": 5.074950416202411e-06,
+      "loss": 1.0957,
+      "step": 2529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2997907374729323,
+      "learning_rate": 5.067438764249495e-06,
+      "loss": 1.1026,
+      "step": 2530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8754982051057407,
+      "learning_rate": 5.0599307885724426e-06,
+      "loss": 0.6026,
+      "step": 2531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9258501065621914,
+      "learning_rate": 5.052426494767004e-06,
+      "loss": 0.7184,
+      "step": 2532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7393804430820072,
+      "learning_rate": 5.044925888426184e-06,
+      "loss": 0.594,
+      "step": 2533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4361955881414197,
+      "learning_rate": 5.037428975140255e-06,
+      "loss": 1.0741,
+      "step": 2534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4681671006248576,
+      "learning_rate": 5.029935760496708e-06,
+      "loss": 1.1112,
+      "step": 2535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.021119525157526,
+      "learning_rate": 5.022446250080304e-06,
+      "loss": 0.5845,
+      "step": 2536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5908174343687909,
+      "learning_rate": 5.01496044947303e-06,
+      "loss": 0.5446,
+      "step": 2537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9973372680698627,
+      "learning_rate": 5.007478364254111e-06,
+      "loss": 0.5992,
+      "step": 2538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7701555347971099,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 0.5745,
+      "step": 2539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7332087763987098,
+      "learning_rate": 4.992525362284382e-06,
+      "loss": 0.5829,
+      "step": 2540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8277076022216938,
+      "learning_rate": 4.985054456678164e-06,
+      "loss": 0.651,
+      "step": 2541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.821699675338382,
+      "learning_rate": 4.977587288749467e-06,
+      "loss": 0.6059,
+      "step": 2542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.55529997367405,
+      "learning_rate": 4.970123864063629e-06,
+      "loss": 1.0521,
+      "step": 2543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.575903195596301,
+      "learning_rate": 4.962664188183198e-06,
+      "loss": 0.498,
+      "step": 2544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.840905384096372,
+      "learning_rate": 4.955208266667928e-06,
+      "loss": 0.5563,
+      "step": 2545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7834218373147455,
+      "learning_rate": 4.94775610507477e-06,
+      "loss": 0.5671,
+      "step": 2546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.425188347709505,
+      "learning_rate": 4.940307708957891e-06,
+      "loss": 1.1176,
+      "step": 2547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7431737409241015,
+      "learning_rate": 4.932863083868619e-06,
+      "loss": 0.5941,
+      "step": 2548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.455664854191056,
+      "learning_rate": 4.925422235355506e-06,
+      "loss": 1.1413,
+      "step": 2549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8458448528037361,
+      "learning_rate": 4.917985168964268e-06,
+      "loss": 0.5783,
+      "step": 2550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.452247950299204,
+      "learning_rate": 4.910551890237807e-06,
+      "loss": 1.1574,
+      "step": 2551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8228473938937901,
+      "learning_rate": 4.903122404716203e-06,
+      "loss": 0.5969,
+      "step": 2552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9765264167128838,
+      "learning_rate": 4.89569671793671e-06,
+      "loss": 0.5525,
+      "step": 2553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7599237669906385,
+      "learning_rate": 4.888274835433745e-06,
+      "loss": 0.5953,
+      "step": 2554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6817771264966355,
+      "learning_rate": 4.880856762738901e-06,
+      "loss": 0.5474,
+      "step": 2555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8875981302232705,
+      "learning_rate": 4.873442505380923e-06,
+      "loss": 0.5911,
+      "step": 2556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.421053611541478,
+      "learning_rate": 4.866032068885713e-06,
+      "loss": 1.1585,
+      "step": 2557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.484589765628312,
+      "learning_rate": 4.858625458776326e-06,
+      "loss": 1.1291,
+      "step": 2558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9514447555128704,
+      "learning_rate": 4.851222680572961e-06,
+      "loss": 0.6367,
+      "step": 2559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.283303255791724,
+      "learning_rate": 4.84382373979298e-06,
+      "loss": 1.0949,
+      "step": 2560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.310233465540795,
+      "learning_rate": 4.836428641950851e-06,
+      "loss": 1.0626,
+      "step": 2561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6604687168190666,
+      "learning_rate": 4.8290373925582125e-06,
+      "loss": 0.5841,
+      "step": 2562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7422747479094571,
+      "learning_rate": 4.821649997123814e-06,
+      "loss": 0.6536,
+      "step": 2563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9684547802671897,
+      "learning_rate": 4.8142664611535374e-06,
+      "loss": 0.6709,
+      "step": 2564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.610928624906475,
+      "learning_rate": 4.806886790150387e-06,
+      "loss": 1.1627,
+      "step": 2565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.904882786672151,
+      "learning_rate": 4.7995109896145e-06,
+      "loss": 0.6557,
+      "step": 2566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7209946414433785,
+      "learning_rate": 4.7921390650430995e-06,
+      "loss": 0.5816,
+      "step": 2567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.907345424819328,
+      "learning_rate": 4.784771021930556e-06,
+      "loss": 0.6759,
+      "step": 2568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.427624786563428,
+      "learning_rate": 4.7774068657683106e-06,
+      "loss": 1.1312,
+      "step": 2569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7579650488161356,
+      "learning_rate": 4.770046602044939e-06,
+      "loss": 0.6573,
+      "step": 2570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0363176710106687,
+      "learning_rate": 4.762690236246097e-06,
+      "loss": 0.6216,
+      "step": 2571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4953307657952,
+      "learning_rate": 4.7553377738545425e-06,
+      "loss": 1.0933,
+      "step": 2572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7494516598029586,
+      "learning_rate": 4.747989220350118e-06,
+      "loss": 0.5721,
+      "step": 2573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.080103255313883,
+      "learning_rate": 4.740644581209763e-06,
+      "loss": 0.6031,
+      "step": 2574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.640955417415073,
+      "learning_rate": 4.733303861907494e-06,
+      "loss": 0.5475,
+      "step": 2575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9005307657193489,
+      "learning_rate": 4.7259670679144e-06,
+      "loss": 0.5726,
+      "step": 2576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7589268913456417,
+      "learning_rate": 4.718634204698653e-06,
+      "loss": 0.5958,
+      "step": 2577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.791279801866013,
+      "learning_rate": 4.7113052777254885e-06,
+      "loss": 0.6159,
+      "step": 2578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7950313777156375,
+      "learning_rate": 4.7039802924572244e-06,
+      "loss": 0.6341,
+      "step": 2579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.611683231056148,
+      "learning_rate": 4.696659254353212e-06,
+      "loss": 1.1275,
+      "step": 2580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8710732167041417,
+      "learning_rate": 4.689342168869889e-06,
+      "loss": 0.5935,
+      "step": 2581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.195551683781475,
+      "learning_rate": 4.682029041460732e-06,
+      "loss": 1.0847,
+      "step": 2582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8387069692645033,
+      "learning_rate": 4.6747198775762716e-06,
+      "loss": 0.5607,
+      "step": 2583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.573160000906917,
+      "learning_rate": 4.667414682664083e-06,
+      "loss": 1.1141,
+      "step": 2584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7059608205624777,
+      "learning_rate": 4.660113462168784e-06,
+      "loss": 0.6146,
+      "step": 2585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3992180563863137,
+      "learning_rate": 4.652816221532028e-06,
+      "loss": 1.1049,
+      "step": 2586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6260454102431363,
+      "learning_rate": 4.645522966192515e-06,
+      "loss": 0.5208,
+      "step": 2587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.311252616949436,
+      "learning_rate": 4.63823370158595e-06,
+      "loss": 1.0722,
+      "step": 2588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9537430524080834,
+      "learning_rate": 4.630948433145088e-06,
+      "loss": 0.5854,
+      "step": 2589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0257957185257176,
+      "learning_rate": 4.623667166299694e-06,
+      "loss": 0.6212,
+      "step": 2590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4314631585020754,
+      "learning_rate": 4.616389906476551e-06,
+      "loss": 1.0779,
+      "step": 2591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.819700395713027,
+      "learning_rate": 4.609116659099455e-06,
+      "loss": 0.572,
+      "step": 2592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7683873205870135,
+      "learning_rate": 4.601847429589213e-06,
+      "loss": 0.682,
+      "step": 2593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.311158756123246,
+      "learning_rate": 4.594582223363642e-06,
+      "loss": 1.1368,
+      "step": 2594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9245303353726053,
+      "learning_rate": 4.587321045837555e-06,
+      "loss": 0.6613,
+      "step": 2595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7538056656259287,
+      "learning_rate": 4.580063902422762e-06,
+      "loss": 0.5858,
+      "step": 2596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9143510711285112,
+      "learning_rate": 4.572810798528064e-06,
+      "loss": 0.6874,
+      "step": 2597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.786628038053231,
+      "learning_rate": 4.565561739559268e-06,
+      "loss": 0.5957,
+      "step": 2598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8828010469794128,
+      "learning_rate": 4.558316730919137e-06,
+      "loss": 0.6196,
+      "step": 2599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6322157714157208,
+      "learning_rate": 4.551075778007447e-06,
+      "loss": 0.5697,
+      "step": 2600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.347106285259032,
+      "learning_rate": 4.54383888622092e-06,
+      "loss": 1.099,
+      "step": 2601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.961622826409772,
+      "learning_rate": 4.53660606095328e-06,
+      "loss": 0.5875,
+      "step": 2602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.404973991556621,
+      "learning_rate": 4.529377307595203e-06,
+      "loss": 1.109,
+      "step": 2603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6811164344501117,
+      "learning_rate": 4.522152631534336e-06,
+      "loss": 0.5734,
+      "step": 2604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9039637902700774,
+      "learning_rate": 4.514932038155278e-06,
+      "loss": 0.5753,
+      "step": 2605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8870559801408475,
+      "learning_rate": 4.507715532839608e-06,
+      "loss": 0.6035,
+      "step": 2606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.276699260449193,
+      "learning_rate": 4.500503120965827e-06,
+      "loss": 1.0995,
+      "step": 2607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8090820530579703,
+      "learning_rate": 4.493294807909415e-06,
+      "loss": 0.6529,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4312530515676576,
+      "learning_rate": 4.48609059904277e-06,
+      "loss": 1.0872,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8152382354936218,
+      "learning_rate": 4.4788904997352525e-06,
+      "loss": 0.6482,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2778127401563704,
+      "learning_rate": 4.471694515353151e-06,
+      "loss": 1.1619,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5652534542709615,
+      "learning_rate": 4.464502651259683e-06,
+      "loss": 1.06,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9679352982225002,
+      "learning_rate": 4.4573149128150096e-06,
+      "loss": 0.6268,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9696583312054512,
+      "learning_rate": 4.450131305376201e-06,
+      "loss": 0.6007,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9743241586578608,
+      "learning_rate": 4.442951834297257e-06,
+      "loss": 0.564,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7174297145990522,
+      "learning_rate": 4.435776504929092e-06,
+      "loss": 0.5969,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7952145802921384,
+      "learning_rate": 4.428605322619535e-06,
+      "loss": 0.6477,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6575553695182346,
+      "learning_rate": 4.421438292713318e-06,
+      "loss": 1.1746,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8161178400344211,
+      "learning_rate": 4.4142754205520965e-06,
+      "loss": 0.5839,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.593651830700048,
+      "learning_rate": 4.4071167114743985e-06,
+      "loss": 1.1276,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6508293863327008,
+      "learning_rate": 4.399962170815676e-06,
+      "loss": 0.5692,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.453574003050369,
+      "learning_rate": 4.39281180390826e-06,
+      "loss": 1.0795,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4709687790093806,
+      "learning_rate": 4.385665616081374e-06,
+      "loss": 1.0932,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3394545797364215,
+      "learning_rate": 4.378523612661126e-06,
+      "loss": 1.1055,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7804634539496988,
+      "learning_rate": 4.371385798970508e-06,
+      "loss": 0.6454,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.564845641556418,
+      "learning_rate": 4.364252180329381e-06,
+      "loss": 1.1445,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.485174941335601,
+      "learning_rate": 4.3571227620545e-06,
+      "loss": 1.0914,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9429748183717805,
+      "learning_rate": 4.349997549459458e-06,
+      "loss": 0.6244,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8701232678522057,
+      "learning_rate": 4.342876547854743e-06,
+      "loss": 0.5735,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5957618668195486,
+      "learning_rate": 4.3357597625476885e-06,
+      "loss": 1.1293,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9390560353435482,
+      "learning_rate": 4.32864719884249e-06,
+      "loss": 0.5752,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8504719637104867,
+      "learning_rate": 4.321538862040196e-06,
+      "loss": 0.6701,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2967543747607775,
+      "learning_rate": 4.3144347574387e-06,
+      "loss": 1.0984,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.795475887786526,
+      "learning_rate": 4.307334890332753e-06,
+      "loss": 0.5651,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8464119591941437,
+      "learning_rate": 4.300239266013938e-06,
+      "loss": 0.6107,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6224618929680403,
+      "learning_rate": 4.293147889770678e-06,
+      "loss": 0.5345,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7815795255950113,
+      "learning_rate": 4.286060766888227e-06,
+      "loss": 0.5441,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7503786058676882,
+      "learning_rate": 4.278977902648682e-06,
+      "loss": 0.6297,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8623220644833192,
+      "learning_rate": 4.271899302330943e-06,
+      "loss": 0.6098,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3323188373785384,
+      "learning_rate": 4.264824971210759e-06,
+      "loss": 1.1204,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7432637709368966,
+      "learning_rate": 4.2577549145606704e-06,
+      "loss": 0.608,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.832921879019371,
+      "learning_rate": 4.2506891376500534e-06,
+      "loss": 0.5703,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7153951103699063,
+      "learning_rate": 4.243627645745083e-06,
+      "loss": 0.5499,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.436379644780466,
+      "learning_rate": 4.236570444108745e-06,
+      "loss": 1.074,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9893579375506003,
+      "learning_rate": 4.229517538000822e-06,
+      "loss": 0.6619,
+      "step": 2645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6239162651385035,
+      "learning_rate": 4.222468932677909e-06,
+      "loss": 1.147,
+      "step": 2646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8113626389102604,
+      "learning_rate": 4.215424633393372e-06,
+      "loss": 0.5808,
+      "step": 2647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9315015336092443,
+      "learning_rate": 4.208384645397392e-06,
+      "loss": 0.6363,
+      "step": 2648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.877492395642846,
+      "learning_rate": 4.201348973936923e-06,
+      "loss": 0.6084,
+      "step": 2649
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.693044859802817,
+      "learning_rate": 4.194317624255705e-06,
+      "loss": 0.5523,
+      "step": 2650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.066896513361684,
+      "learning_rate": 4.187290601594257e-06,
+      "loss": 0.6286,
+      "step": 2651
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7614496308195848,
+      "learning_rate": 4.180267911189869e-06,
+      "loss": 0.6265,
+      "step": 2652
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3949227278214953,
+      "learning_rate": 4.173249558276615e-06,
+      "loss": 1.1271,
+      "step": 2653
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7943376882141326,
+      "learning_rate": 4.166235548085322e-06,
+      "loss": 0.5578,
+      "step": 2654
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.299619268603521,
+      "learning_rate": 4.159225885843588e-06,
+      "loss": 1.0982,
+      "step": 2655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8094973785568653,
+      "learning_rate": 4.152220576775765e-06,
+      "loss": 0.5978,
+      "step": 2656
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7703920341410784,
+      "learning_rate": 4.1452196261029666e-06,
+      "loss": 0.5975,
+      "step": 2657
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0480981093538575,
+      "learning_rate": 4.13822303904305e-06,
+      "loss": 0.5989,
+      "step": 2658
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8431271370868787,
+      "learning_rate": 4.1312308208106376e-06,
+      "loss": 0.6433,
+      "step": 2659
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.898314457677467,
+      "learning_rate": 4.124242976617068e-06,
+      "loss": 0.5656,
+      "step": 2660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7410366998027054,
+      "learning_rate": 4.117259511670446e-06,
+      "loss": 1.1111,
+      "step": 2661
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6425248477501226,
+      "learning_rate": 4.1102804311755985e-06,
+      "loss": 0.545,
+      "step": 2662
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.38343828092289,
+      "learning_rate": 4.1033057403340895e-06,
+      "loss": 1.1134,
+      "step": 2663
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8866982728225772,
+      "learning_rate": 4.096335444344206e-06,
+      "loss": 0.6128,
+      "step": 2664
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.39048664138113,
+      "learning_rate": 4.089369548400968e-06,
+      "loss": 1.1503,
+      "step": 2665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9177474957417966,
+      "learning_rate": 4.0824080576961035e-06,
+      "loss": 0.5343,
+      "step": 2666
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4360504460265173,
+      "learning_rate": 4.075450977418074e-06,
+      "loss": 1.1344,
+      "step": 2667
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.845207267237328,
+      "learning_rate": 4.068498312752043e-06,
+      "loss": 0.5987,
+      "step": 2668
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.851533625185319,
+      "learning_rate": 4.0615500688798825e-06,
+      "loss": 0.56,
+      "step": 2669
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.519266237555693,
+      "learning_rate": 4.0546062509801735e-06,
+      "loss": 1.1508,
+      "step": 2670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9498372695221438,
+      "learning_rate": 4.047666864228193e-06,
+      "loss": 0.6591,
+      "step": 2671
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7041588927955353,
+      "learning_rate": 4.04073191379593e-06,
+      "loss": 0.577,
+      "step": 2672
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6749686272917435,
+      "learning_rate": 4.033801404852042e-06,
+      "loss": 0.5709,
+      "step": 2673
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9673683043047556,
+      "learning_rate": 4.026875342561902e-06,
+      "loss": 0.6624,
+      "step": 2674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4731877569324343,
+      "learning_rate": 4.0199537320875535e-06,
+      "loss": 1.0878,
+      "step": 2675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1177788233370736,
+      "learning_rate": 4.013036578587726e-06,
+      "loss": 0.5726,
+      "step": 2676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6952053754554064,
+      "learning_rate": 4.006123887217825e-06,
+      "loss": 1.131,
+      "step": 2677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.863964773439302,
+      "learning_rate": 3.999215663129942e-06,
+      "loss": 0.5415,
+      "step": 2678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2529129302086512,
+      "learning_rate": 3.992311911472815e-06,
+      "loss": 1.113,
+      "step": 2679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4576758670699244,
+      "learning_rate": 3.985412637391878e-06,
+      "loss": 1.0916,
+      "step": 2680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.068396625014615,
+      "learning_rate": 3.978517846029199e-06,
+      "loss": 0.5954,
+      "step": 2681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8397337609982258,
+      "learning_rate": 3.9716275425235294e-06,
+      "loss": 0.6244,
+      "step": 2682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.728321955095718,
+      "learning_rate": 3.964741732010262e-06,
+      "loss": 0.5798,
+      "step": 2683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9480336679879975,
+      "learning_rate": 3.9578604196214456e-06,
+      "loss": 0.6265,
+      "step": 2684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5939398304188264,
+      "learning_rate": 3.950983610485769e-06,
+      "loss": 1.0448,
+      "step": 2685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8489067689038938,
+      "learning_rate": 3.944111309728581e-06,
+      "loss": 0.6656,
+      "step": 2686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7612698994459264,
+      "learning_rate": 3.937243522471855e-06,
+      "loss": 0.6203,
+      "step": 2687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8398309061507865,
+      "learning_rate": 3.930380253834207e-06,
+      "loss": 0.6225,
+      "step": 2688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9494758500777634,
+      "learning_rate": 3.923521508930882e-06,
+      "loss": 0.6375,
+      "step": 2689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7824827159128669,
+      "learning_rate": 3.916667292873754e-06,
+      "loss": 0.6093,
+      "step": 2690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3584799725182455,
+      "learning_rate": 3.909817610771332e-06,
+      "loss": 1.0709,
+      "step": 2691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4236647844252155,
+      "learning_rate": 3.902972467728724e-06,
+      "loss": 1.0851,
+      "step": 2692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3956985176717165,
+      "learning_rate": 3.8961318688476765e-06,
+      "loss": 1.0922,
+      "step": 2693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.453636027056736,
+      "learning_rate": 3.8892958192265385e-06,
+      "loss": 1.1189,
+      "step": 2694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7697770939012538,
+      "learning_rate": 3.882464323960271e-06,
+      "loss": 0.6205,
+      "step": 2695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8519138891010922,
+      "learning_rate": 3.8756373881404375e-06,
+      "loss": 0.6457,
+      "step": 2696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.915863069307979,
+      "learning_rate": 3.868815016855209e-06,
+      "loss": 0.6235,
+      "step": 2697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8978887417103962,
+      "learning_rate": 3.861997215189345e-06,
+      "loss": 0.6075,
+      "step": 2698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9348735746211472,
+      "learning_rate": 3.855183988224218e-06,
+      "loss": 0.6836,
+      "step": 2699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8857580871359263,
+      "learning_rate": 3.848375341037763e-06,
+      "loss": 0.5962,
+      "step": 2700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.936606840967372,
+      "learning_rate": 3.841571278704529e-06,
+      "loss": 0.6798,
+      "step": 2701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7769011752380144,
+      "learning_rate": 3.834771806295633e-06,
+      "loss": 0.6499,
+      "step": 2702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8463807264052698,
+      "learning_rate": 3.827976928878772e-06,
+      "loss": 0.5895,
+      "step": 2703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6316687665367793,
+      "learning_rate": 3.8211866515182214e-06,
+      "loss": 0.6077,
+      "step": 2704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6641656881592017,
+      "learning_rate": 3.8144009792748238e-06,
+      "loss": 0.5872,
+      "step": 2705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9641482358101514,
+      "learning_rate": 3.8076199172060013e-06,
+      "loss": 0.6201,
+      "step": 2706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.742561285413826,
+      "learning_rate": 3.800843470365726e-06,
+      "loss": 0.5557,
+      "step": 2707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8400527167203014,
+      "learning_rate": 3.7940716438045364e-06,
+      "loss": 0.5938,
+      "step": 2708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0888393070157427,
+      "learning_rate": 3.787304442569525e-06,
+      "loss": 0.6157,
+      "step": 2709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8586211624324283,
+      "learning_rate": 3.7805418717043473e-06,
+      "loss": 0.5978,
+      "step": 2710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.461550478705761,
+      "learning_rate": 3.773783936249187e-06,
+      "loss": 1.1652,
+      "step": 2711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6254031327317973,
+      "learning_rate": 3.7670306412407985e-06,
+      "loss": 0.5118,
+      "step": 2712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8622554967111187,
+      "learning_rate": 3.7602819917124524e-06,
+      "loss": 0.6388,
+      "step": 2713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8706870312526513,
+      "learning_rate": 3.7535379926939765e-06,
+      "loss": 0.5511,
+      "step": 2714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0061640842164996,
+      "learning_rate": 3.7467986492117247e-06,
+      "loss": 0.5995,
+      "step": 2715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8611473846797666,
+      "learning_rate": 3.7400639662885794e-06,
+      "loss": 0.4743,
+      "step": 2716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.840284457771445,
+      "learning_rate": 3.733333948943949e-06,
+      "loss": 0.6189,
+      "step": 2717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7810755831077363,
+      "learning_rate": 3.726608602193779e-06,
+      "loss": 0.572,
+      "step": 2718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8323081600757825,
+      "learning_rate": 3.7198879310505053e-06,
+      "loss": 0.599,
+      "step": 2719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7135126929246003,
+      "learning_rate": 3.7131719405231082e-06,
+      "loss": 0.5389,
+      "step": 2720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.393150715244422,
+      "learning_rate": 3.706460635617064e-06,
+      "loss": 1.0013,
+      "step": 2721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8139472215968264,
+      "learning_rate": 3.6997540213343587e-06,
+      "loss": 0.6007,
+      "step": 2722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7583933507660874,
+      "learning_rate": 3.6930521026734843e-06,
+      "loss": 0.6829,
+      "step": 2723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8595206020639397,
+      "learning_rate": 3.6863548846294283e-06,
+      "loss": 0.5994,
+      "step": 2724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.866637834871972,
+      "learning_rate": 3.6796623721936876e-06,
+      "loss": 0.5432,
+      "step": 2725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7514854687543908,
+      "learning_rate": 3.6729745703542373e-06,
+      "loss": 0.5971,
+      "step": 2726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7741821771617148,
+      "learning_rate": 3.6662914840955487e-06,
+      "loss": 0.6062,
+      "step": 2727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7496132777815114,
+      "learning_rate": 3.6596131183985773e-06,
+      "loss": 0.5907,
+      "step": 2728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7024581553346463,
+      "learning_rate": 3.652939478240761e-06,
+      "loss": 0.5713,
+      "step": 2729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7191913911181966,
+      "learning_rate": 3.646270568596011e-06,
+      "loss": 0.4777,
+      "step": 2730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7881308609013975,
+      "learning_rate": 3.639606394434728e-06,
+      "loss": 0.5889,
+      "step": 2731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8424851959324764,
+      "learning_rate": 3.632946960723759e-06,
+      "loss": 0.6166,
+      "step": 2732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9536006504763606,
+      "learning_rate": 3.6262922724264394e-06,
+      "loss": 0.6222,
+      "step": 2733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3062369701677157,
+      "learning_rate": 3.619642334502558e-06,
+      "loss": 1.0692,
+      "step": 2734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.359061887454257,
+      "learning_rate": 3.612997151908364e-06,
+      "loss": 1.0958,
+      "step": 2735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7791795904960397,
+      "learning_rate": 3.606356729596564e-06,
+      "loss": 0.5738,
+      "step": 2736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7535085457308508,
+      "learning_rate": 3.5997210725163147e-06,
+      "loss": 0.5733,
+      "step": 2737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8676195977579348,
+      "learning_rate": 3.5930901856132203e-06,
+      "loss": 0.5796,
+      "step": 2738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0186199839614214,
+      "learning_rate": 3.586464073829338e-06,
+      "loss": 0.5808,
+      "step": 2739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6943613509282098,
+      "learning_rate": 3.5798427421031557e-06,
+      "loss": 0.5083,
+      "step": 2740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.917314273565661,
+      "learning_rate": 3.5732261953696035e-06,
+      "loss": 0.6365,
+      "step": 2741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8839250056945323,
+      "learning_rate": 3.5666144385600454e-06,
+      "loss": 0.6197,
+      "step": 2742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.821386560734248,
+      "learning_rate": 3.5600074766022707e-06,
+      "loss": 0.5773,
+      "step": 2743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7426933502589312,
+      "learning_rate": 3.5534053144205084e-06,
+      "loss": 0.5444,
+      "step": 2744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.713390467983297,
+      "learning_rate": 3.5468079569353888e-06,
+      "loss": 0.6321,
+      "step": 2745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6335942077874668,
+      "learning_rate": 3.540215409063982e-06,
+      "loss": 0.4938,
+      "step": 2746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.773531817698764,
+      "learning_rate": 3.5336276757197616e-06,
+      "loss": 0.5978,
+      "step": 2747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4575917755466246,
+      "learning_rate": 3.527044761812617e-06,
+      "loss": 1.1379,
+      "step": 2748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7578502096677995,
+      "learning_rate": 3.52046667224884e-06,
+      "loss": 0.6032,
+      "step": 2749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9132154022744636,
+      "learning_rate": 3.513893411931143e-06,
+      "loss": 0.5826,
+      "step": 2750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3800835051208495,
+      "learning_rate": 3.507324985758612e-06,
+      "loss": 1.0664,
+      "step": 2751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2952828894829516,
+      "learning_rate": 3.5007613986267607e-06,
+      "loss": 1.1103,
+      "step": 2752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8120035865490667,
+      "learning_rate": 3.4942026554274665e-06,
+      "loss": 0.57,
+      "step": 2753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8695115561855469,
+      "learning_rate": 3.4876487610490216e-06,
+      "loss": 0.5913,
+      "step": 2754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.580157631452974,
+      "learning_rate": 3.481099720376091e-06,
+      "loss": 1.1149,
+      "step": 2755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.573808425944145,
+      "learning_rate": 3.4745555382897255e-06,
+      "loss": 0.5923,
+      "step": 2756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6677626780974073,
+      "learning_rate": 3.4680162196673507e-06,
+      "loss": 0.5484,
+      "step": 2757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7602726040944825,
+      "learning_rate": 3.4614817693827764e-06,
+      "loss": 0.6319,
+      "step": 2758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.876861103916033,
+      "learning_rate": 3.454952192306178e-06,
+      "loss": 0.6098,
+      "step": 2759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7514239251852999,
+      "learning_rate": 3.4484274933040975e-06,
+      "loss": 0.5561,
+      "step": 2760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9188905225661352,
+      "learning_rate": 3.4419076772394433e-06,
+      "loss": 0.6119,
+      "step": 2761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8040516444257304,
+      "learning_rate": 3.435392748971482e-06,
+      "loss": 0.5594,
+      "step": 2762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6122583944514146,
+      "learning_rate": 3.42888271335585e-06,
+      "loss": 0.5415,
+      "step": 2763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9093720105235437,
+      "learning_rate": 3.422377575244513e-06,
+      "loss": 0.6279,
+      "step": 2764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5879258310722069,
+      "learning_rate": 3.4158773394858135e-06,
+      "loss": 0.5686,
+      "step": 2765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7924195519847972,
+      "learning_rate": 3.409382010924416e-06,
+      "loss": 0.5672,
+      "step": 2766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.383915834906072,
+      "learning_rate": 3.4028915944013465e-06,
+      "loss": 1.0965,
+      "step": 2767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8436891175729626,
+      "learning_rate": 3.396406094753959e-06,
+      "loss": 0.5811,
+      "step": 2768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.344567904458836,
+      "learning_rate": 3.3899255168159483e-06,
+      "loss": 1.0867,
+      "step": 2769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.390324992908514,
+      "learning_rate": 3.3834498654173344e-06,
+      "loss": 1.0951,
+      "step": 2770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8515580084326948,
+      "learning_rate": 3.37697914538448e-06,
+      "loss": 0.6132,
+      "step": 2771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8977026309426726,
+      "learning_rate": 3.370513361540051e-06,
+      "loss": 0.6343,
+      "step": 2772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9461142947575865,
+      "learning_rate": 3.3640525187030526e-06,
+      "loss": 0.6669,
+      "step": 2773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.225767435881883,
+      "learning_rate": 3.3575966216888e-06,
+      "loss": 1.1003,
+      "step": 2774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.597345077099322,
+      "learning_rate": 3.3511456753089233e-06,
+      "loss": 0.6165,
+      "step": 2775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0062352597107536,
+      "learning_rate": 3.3446996843713596e-06,
+      "loss": 0.5395,
+      "step": 2776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9775343221731325,
+      "learning_rate": 3.338258653680355e-06,
+      "loss": 0.6771,
+      "step": 2777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6852356297539786,
+      "learning_rate": 3.331822588036465e-06,
+      "loss": 0.6027,
+      "step": 2778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.220859992660356,
+      "learning_rate": 3.325391492236535e-06,
+      "loss": 1.1531,
+      "step": 2779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9702354470403425,
+      "learning_rate": 3.318965371073711e-06,
+      "loss": 0.6143,
+      "step": 2780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0692341577372035,
+      "learning_rate": 3.312544229337431e-06,
+      "loss": 0.6142,
+      "step": 2781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7166595260959039,
+      "learning_rate": 3.306128071813419e-06,
+      "loss": 0.5609,
+      "step": 2782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6597620799915105,
+      "learning_rate": 3.299716903283686e-06,
+      "loss": 0.4984,
+      "step": 2783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7420589460744833,
+      "learning_rate": 3.293310728526534e-06,
+      "loss": 0.6197,
+      "step": 2784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9587527715608264,
+      "learning_rate": 3.2869095523165216e-06,
+      "loss": 0.6292,
+      "step": 2785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.768020116957326,
+      "learning_rate": 3.280513379424504e-06,
+      "loss": 0.6581,
+      "step": 2786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7752880383785714,
+      "learning_rate": 3.2741222146175955e-06,
+      "loss": 0.5746,
+      "step": 2787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4088619511998477,
+      "learning_rate": 3.267736062659179e-06,
+      "loss": 1.1761,
+      "step": 2788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.633314147544877,
+      "learning_rate": 3.2613549283089042e-06,
+      "loss": 0.5487,
+      "step": 2789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.156074500305198,
+      "learning_rate": 3.2549788163226793e-06,
+      "loss": 1.0863,
+      "step": 2790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.568458141955435,
+      "learning_rate": 3.2486077314526666e-06,
+      "loss": 0.5581,
+      "step": 2791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.786617085293935,
+      "learning_rate": 3.242241678447291e-06,
+      "loss": 0.6089,
+      "step": 2792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.838555278766746,
+      "learning_rate": 3.235880662051216e-06,
+      "loss": 0.5849,
+      "step": 2793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.256193466840037,
+      "learning_rate": 3.229524687005359e-06,
+      "loss": 1.0796,
+      "step": 2794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8017137032080455,
+      "learning_rate": 3.223173758046876e-06,
+      "loss": 0.5722,
+      "step": 2795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.665877873540797,
+      "learning_rate": 3.216827879909158e-06,
+      "loss": 0.6255,
+      "step": 2796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7915718086558428,
+      "learning_rate": 3.210487057321849e-06,
+      "loss": 0.5358,
+      "step": 2797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7884376565226787,
+      "learning_rate": 3.2041512950108e-06,
+      "loss": 0.6238,
+      "step": 2798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7923763297457942,
+      "learning_rate": 3.197820597698111e-06,
+      "loss": 0.5746,
+      "step": 2799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7089359900592112,
+      "learning_rate": 3.1914949701020993e-06,
+      "loss": 0.5616,
+      "step": 2800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8498678190283544,
+      "learning_rate": 3.1851744169373e-06,
+      "loss": 0.5421,
+      "step": 2801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7262494877382744,
+      "learning_rate": 3.17885894291447e-06,
+      "loss": 0.5835,
+      "step": 2802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.646195311870674,
+      "learning_rate": 3.1725485527405898e-06,
+      "loss": 1.062,
+      "step": 2803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8680137830836394,
+      "learning_rate": 3.1662432511188266e-06,
+      "loss": 0.5376,
+      "step": 2804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.281667124184148,
+      "learning_rate": 3.1599430427485866e-06,
+      "loss": 1.115,
+      "step": 2805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9910926144584613,
+      "learning_rate": 3.153647932325448e-06,
+      "loss": 0.5957,
+      "step": 2806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.765002924346476,
+      "learning_rate": 3.1473579245412155e-06,
+      "loss": 0.5585,
+      "step": 2807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0267451527521962,
+      "learning_rate": 3.141073024083878e-06,
+      "loss": 0.6066,
+      "step": 2808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.734854539530825,
+      "learning_rate": 3.13479323563762e-06,
+      "loss": 0.5472,
+      "step": 2809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7211770358166791,
+      "learning_rate": 3.128518563882812e-06,
+      "loss": 0.5655,
+      "step": 2810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9769940815097842,
+      "learning_rate": 3.1222490134960214e-06,
+      "loss": 0.6217,
+      "step": 2811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8273422332433105,
+      "learning_rate": 3.1159845891499894e-06,
+      "loss": 0.6124,
+      "step": 2812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.851098990670245,
+      "learning_rate": 3.1097252955136416e-06,
+      "loss": 0.6228,
+      "step": 2813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9331933434008877,
+      "learning_rate": 3.1034711372520756e-06,
+      "loss": 0.6245,
+      "step": 2814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7935243243395707,
+      "learning_rate": 3.0972221190265617e-06,
+      "loss": 0.6341,
+      "step": 2815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8047998236393514,
+      "learning_rate": 3.090978245494549e-06,
+      "loss": 0.5804,
+      "step": 2816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.64243120408363,
+      "learning_rate": 3.084739521309634e-06,
+      "loss": 0.614,
+      "step": 2817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7767610967197034,
+      "learning_rate": 3.078505951121594e-06,
+      "loss": 0.5806,
+      "step": 2818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7192266761098405,
+      "learning_rate": 3.072277539576355e-06,
+      "loss": 0.5802,
+      "step": 2819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.741794632591887,
+      "learning_rate": 3.0660542913159984e-06,
+      "loss": 0.5767,
+      "step": 2820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.799220458009852,
+      "learning_rate": 3.0598362109787603e-06,
+      "loss": 0.575,
+      "step": 2821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6692182925228516,
+      "learning_rate": 3.0536233031990237e-06,
+      "loss": 0.6327,
+      "step": 2822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0301797459762296,
+      "learning_rate": 3.047415572607313e-06,
+      "loss": 0.6702,
+      "step": 2823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6661988249415893,
+      "learning_rate": 3.041213023830307e-06,
+      "loss": 0.5791,
+      "step": 2824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.887410050293082,
+      "learning_rate": 3.0350156614908e-06,
+      "loss": 0.6183,
+      "step": 2825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.340743715208589,
+      "learning_rate": 3.028823490207745e-06,
+      "loss": 1.1121,
+      "step": 2826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9877173968584356,
+      "learning_rate": 3.022636514596211e-06,
+      "loss": 0.5684,
+      "step": 2827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4860588860175827,
+      "learning_rate": 3.016454739267397e-06,
+      "loss": 1.0577,
+      "step": 2828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.781115827336361,
+      "learning_rate": 3.0102781688286297e-06,
+      "loss": 0.5633,
+      "step": 2829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7514283289474086,
+      "learning_rate": 3.004106807883351e-06,
+      "loss": 0.6488,
+      "step": 2830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8024289266591609,
+      "learning_rate": 2.9979406610311303e-06,
+      "loss": 0.6073,
+      "step": 2831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8660848511671735,
+      "learning_rate": 2.99177973286764e-06,
+      "loss": 0.5962,
+      "step": 2832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9130658728642416,
+      "learning_rate": 2.985624027984668e-06,
+      "loss": 0.6375,
+      "step": 2833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6086567973299841,
+      "learning_rate": 2.9794735509701056e-06,
+      "loss": 0.5589,
+      "step": 2834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9653373749816363,
+      "learning_rate": 2.97332830640796e-06,
+      "loss": 0.5787,
+      "step": 2835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8473243733584372,
+      "learning_rate": 2.9671882988783163e-06,
+      "loss": 0.6304,
+      "step": 2836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7070159951667636,
+      "learning_rate": 2.961053532957383e-06,
+      "loss": 0.5491,
+      "step": 2837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.820699441021506,
+      "learning_rate": 2.954924013217434e-06,
+      "loss": 0.5746,
+      "step": 2838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4029113546131344,
+      "learning_rate": 2.948799744226859e-06,
+      "loss": 1.1374,
+      "step": 2839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8586227718775137,
+      "learning_rate": 2.9426807305501182e-06,
+      "loss": 0.6171,
+      "step": 2840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9088095527917228,
+      "learning_rate": 2.93656697674776e-06,
+      "loss": 0.6634,
+      "step": 2841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7087962902655085,
+      "learning_rate": 2.9304584873764096e-06,
+      "loss": 1.139,
+      "step": 2842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8658161020606712,
+      "learning_rate": 2.9243552669887797e-06,
+      "loss": 0.5979,
+      "step": 2843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2087204951845867,
+      "learning_rate": 2.918257320133635e-06,
+      "loss": 1.0997,
+      "step": 2844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9085698977937386,
+      "learning_rate": 2.9121646513558322e-06,
+      "loss": 0.5149,
+      "step": 2845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296973522795832,
+      "learning_rate": 2.906077265196281e-06,
+      "loss": 1.0948,
+      "step": 2846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.482917116839101,
+      "learning_rate": 2.899995166191958e-06,
+      "loss": 1.1781,
+      "step": 2847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.438550718408801,
+      "learning_rate": 2.893918358875896e-06,
+      "loss": 1.1033,
+      "step": 2848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3700969139535535,
+      "learning_rate": 2.887846847777185e-06,
+      "loss": 1.1267,
+      "step": 2849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2072829074697196,
+      "learning_rate": 2.8817806374209754e-06,
+      "loss": 1.0816,
+      "step": 2850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8798936704056972,
+      "learning_rate": 2.875719732328456e-06,
+      "loss": 0.6378,
+      "step": 2851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.423446271787683,
+      "learning_rate": 2.8696641370168675e-06,
+      "loss": 1.0158,
+      "step": 2852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.344876375998864,
+      "learning_rate": 2.863613855999491e-06,
+      "loss": 1.066,
+      "step": 2853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8372365240197521,
+      "learning_rate": 2.857568893785646e-06,
+      "loss": 0.574,
+      "step": 2854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3922967284479815,
+      "learning_rate": 2.8515292548806895e-06,
+      "loss": 1.0607,
+      "step": 2855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6912402888978564,
+      "learning_rate": 2.845494943786018e-06,
+      "loss": 0.562,
+      "step": 2856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.774904386470753,
+      "learning_rate": 2.8394659649990375e-06,
+      "loss": 0.6495,
+      "step": 2857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.002165992961411,
+      "learning_rate": 2.8334423230132024e-06,
+      "loss": 0.5942,
+      "step": 2858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9275370484296643,
+      "learning_rate": 2.8274240223179774e-06,
+      "loss": 0.6813,
+      "step": 2859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8003719876177147,
+      "learning_rate": 2.8214110673988447e-06,
+      "loss": 0.6358,
+      "step": 2860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9461004605143812,
+      "learning_rate": 2.8154034627373107e-06,
+      "loss": 0.6511,
+      "step": 2861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6309476425053053,
+      "learning_rate": 2.809401212810885e-06,
+      "loss": 1.1186,
+      "step": 2862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8625480363167783,
+      "learning_rate": 2.8034043220930905e-06,
+      "loss": 0.5898,
+      "step": 2863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.635503552813505,
+      "learning_rate": 2.7974127950534614e-06,
+      "loss": 0.5856,
+      "step": 2864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5549013590805654,
+      "learning_rate": 2.791426636157526e-06,
+      "loss": 1.128,
+      "step": 2865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.36317504573532,
+      "learning_rate": 2.7854458498668135e-06,
+      "loss": 1.0843,
+      "step": 2866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6147230834452757,
+      "learning_rate": 2.7794704406388506e-06,
+      "loss": 0.5739,
+      "step": 2867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7721433640444948,
+      "learning_rate": 2.773500412927154e-06,
+      "loss": 0.5557,
+      "step": 2868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8581064860346566,
+      "learning_rate": 2.767535771181239e-06,
+      "loss": 0.6025,
+      "step": 2869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7754473576098022,
+      "learning_rate": 2.761576519846586e-06,
+      "loss": 0.579,
+      "step": 2870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9638581007101463,
+      "learning_rate": 2.755622663364682e-06,
+      "loss": 0.6319,
+      "step": 2871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7788467161047918,
+      "learning_rate": 2.7496742061729785e-06,
+      "loss": 0.5984,
+      "step": 2872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4139202206115664,
+      "learning_rate": 2.743731152704907e-06,
+      "loss": 1.0636,
+      "step": 2873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8762415661738803,
+      "learning_rate": 2.7377935073898665e-06,
+      "loss": 0.6129,
+      "step": 2874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1089124549246545,
+      "learning_rate": 2.731861274653239e-06,
+      "loss": 0.5911,
+      "step": 2875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.386505881398116,
+      "learning_rate": 2.725934458916352e-06,
+      "loss": 1.0846,
+      "step": 2876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7450923051479923,
+      "learning_rate": 2.7200130645965185e-06,
+      "loss": 0.6231,
+      "step": 2877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.521152219188254,
+      "learning_rate": 2.714097096106988e-06,
+      "loss": 1.1061,
+      "step": 2878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5802517796056448,
+      "learning_rate": 2.7081865578569843e-06,
+      "loss": 0.6118,
+      "step": 2879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.45338341595798,
+      "learning_rate": 2.7022814542516774e-06,
+      "loss": 1.0965,
+      "step": 2880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6620009009759025,
+      "learning_rate": 2.696381789692183e-06,
+      "loss": 0.5441,
+      "step": 2881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7856322860780314,
+      "learning_rate": 2.6904875685755662e-06,
+      "loss": 0.5528,
+      "step": 2882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7620978734390684,
+      "learning_rate": 2.6845987952948395e-06,
+      "loss": 0.5983,
+      "step": 2883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.781497722984175,
+      "learning_rate": 2.6787154742389487e-06,
+      "loss": 0.5806,
+      "step": 2884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2677580114512503,
+      "learning_rate": 2.672837609792779e-06,
+      "loss": 1.0899,
+      "step": 2885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.279529876615507,
+      "learning_rate": 2.6669652063371465e-06,
+      "loss": 1.0976,
+      "step": 2886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8363740069675294,
+      "learning_rate": 2.6610982682487963e-06,
+      "loss": 0.5573,
+      "step": 2887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8398030602155777,
+      "learning_rate": 2.6552367999004123e-06,
+      "loss": 0.5584,
+      "step": 2888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8361366732272342,
+      "learning_rate": 2.6493808056605795e-06,
+      "loss": 0.5593,
+      "step": 2889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8448524026965982,
+      "learning_rate": 2.6435302898938254e-06,
+      "loss": 0.6238,
+      "step": 2890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.775722822831256,
+      "learning_rate": 2.6376852569605795e-06,
+      "loss": 0.6274,
+      "step": 2891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7766271772214153,
+      "learning_rate": 2.63184571121719e-06,
+      "loss": 0.5463,
+      "step": 2892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7777646276719976,
+      "learning_rate": 2.6260116570159168e-06,
+      "loss": 0.5826,
+      "step": 2893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7664849763825623,
+      "learning_rate": 2.6201830987049248e-06,
+      "loss": 0.5449,
+      "step": 2894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.790787977298713,
+      "learning_rate": 2.6143600406282797e-06,
+      "loss": 0.6067,
+      "step": 2895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.312976041947536,
+      "learning_rate": 2.6085424871259624e-06,
+      "loss": 1.1384,
+      "step": 2896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4608390615584605,
+      "learning_rate": 2.602730442533826e-06,
+      "loss": 1.1481,
+      "step": 2897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9614906097750013,
+      "learning_rate": 2.596923911183643e-06,
+      "loss": 0.6294,
+      "step": 2898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5791210962586841,
+      "learning_rate": 2.591122897403061e-06,
+      "loss": 0.6357,
+      "step": 2899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8674384116705152,
+      "learning_rate": 2.585327405515622e-06,
+      "loss": 0.5726,
+      "step": 2900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7540278020355662,
+      "learning_rate": 2.5795374398407492e-06,
+      "loss": 0.5903,
+      "step": 2901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4994522976867874,
+      "learning_rate": 2.5737530046937463e-06,
+      "loss": 1.1288,
+      "step": 2902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.719502375282584,
+      "learning_rate": 2.5679741043858043e-06,
+      "loss": 0.6034,
+      "step": 2903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.677260655510472,
+      "learning_rate": 2.562200743223978e-06,
+      "loss": 0.548,
+      "step": 2904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5030467459661017,
+      "learning_rate": 2.5564329255111976e-06,
+      "loss": 0.5224,
+      "step": 2905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.66602858746986,
+      "learning_rate": 2.55067065554626e-06,
+      "loss": 0.6026,
+      "step": 2906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2758905478193534,
+      "learning_rate": 2.5449139376238385e-06,
+      "loss": 1.07,
+      "step": 2907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8506457636217855,
+      "learning_rate": 2.539162776034445e-06,
+      "loss": 0.5585,
+      "step": 2908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8904670427571484,
+      "learning_rate": 2.53341717506448e-06,
+      "loss": 0.5553,
+      "step": 2909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8981827712011516,
+      "learning_rate": 2.5276771389961694e-06,
+      "loss": 0.6204,
+      "step": 2910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.478127359791622,
+      "learning_rate": 2.521942672107617e-06,
+      "loss": 1.1712,
+      "step": 2911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7051588469428005,
+      "learning_rate": 2.51621377867276e-06,
+      "loss": 0.5939,
+      "step": 2912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7291645943214875,
+      "learning_rate": 2.510490462961389e-06,
+      "loss": 0.5868,
+      "step": 2913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.43580549511695,
+      "learning_rate": 2.504772729239131e-06,
+      "loss": 1.0587,
+      "step": 2914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3713193385215208,
+      "learning_rate": 2.4990605817674673e-06,
+      "loss": 1.1118,
+      "step": 2915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7909881540975139,
+      "learning_rate": 2.4933540248036913e-06,
+      "loss": 0.6113,
+      "step": 2916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3166983465114948,
+      "learning_rate": 2.4876530626009543e-06,
+      "loss": 1.1462,
+      "step": 2917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6644968075768025,
+      "learning_rate": 2.481957699408225e-06,
+      "loss": 0.5449,
+      "step": 2918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8834155267044685,
+      "learning_rate": 2.476267939470299e-06,
+      "loss": 0.6299,
+      "step": 2919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7834558420289597,
+      "learning_rate": 2.4705837870278003e-06,
+      "loss": 0.5467,
+      "step": 2920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7877399578787871,
+      "learning_rate": 2.4649052463171673e-06,
+      "loss": 0.5826,
+      "step": 2921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9252734014876924,
+      "learning_rate": 2.459232321570667e-06,
+      "loss": 0.6437,
+      "step": 2922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6727442977178313,
+      "learning_rate": 2.4535650170163703e-06,
+      "loss": 0.525,
+      "step": 2923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8448651635192566,
+      "learning_rate": 2.4479033368781637e-06,
+      "loss": 0.6115,
+      "step": 2924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.493659598386063,
+      "learning_rate": 2.4422472853757396e-06,
+      "loss": 1.1311,
+      "step": 2925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7151012049508032,
+      "learning_rate": 2.4365968667245985e-06,
+      "loss": 0.5274,
+      "step": 2926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.345859380768141,
+      "learning_rate": 2.430952085136036e-06,
+      "loss": 1.0972,
+      "step": 2927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7444475772891292,
+      "learning_rate": 2.425312944817162e-06,
+      "loss": 0.5683,
+      "step": 2928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7535717271438334,
+      "learning_rate": 2.4196794499708575e-06,
+      "loss": 0.5606,
+      "step": 2929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6315400957285926,
+      "learning_rate": 2.4140516047958197e-06,
+      "loss": 0.612,
+      "step": 2930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8357418038291957,
+      "learning_rate": 2.4084294134865226e-06,
+      "loss": 0.6032,
+      "step": 2931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9472620700365733,
+      "learning_rate": 2.4028128802332274e-06,
+      "loss": 0.5636,
+      "step": 2932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7687995839273878,
+      "learning_rate": 2.3972020092219795e-06,
+      "loss": 0.6178,
+      "step": 2933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7095384993516147,
+      "learning_rate": 2.391596804634605e-06,
+      "loss": 0.5269,
+      "step": 2934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8286514966152418,
+      "learning_rate": 2.385997270648702e-06,
+      "loss": 0.5603,
+      "step": 2935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7716370153178929,
+      "learning_rate": 2.3804034114376538e-06,
+      "loss": 0.5571,
+      "step": 2936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8018598140758066,
+      "learning_rate": 2.374815231170603e-06,
+      "loss": 0.6338,
+      "step": 2937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5594143574939239,
+      "learning_rate": 2.3692327340124633e-06,
+      "loss": 0.5395,
+      "step": 2938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.770759276710845,
+      "learning_rate": 2.363655924123913e-06,
+      "loss": 0.6214,
+      "step": 2939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.446192576417244,
+      "learning_rate": 2.358084805661388e-06,
+      "loss": 1.1249,
+      "step": 2940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.561501573091589,
+      "learning_rate": 2.3525193827770953e-06,
+      "loss": 1.095,
+      "step": 2941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7665064768540153,
+      "learning_rate": 2.346959659618975e-06,
+      "loss": 0.6293,
+      "step": 2942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.410070198671885,
+      "learning_rate": 2.3414056403307404e-06,
+      "loss": 1.071,
+      "step": 2943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8317490760409225,
+      "learning_rate": 2.3358573290518427e-06,
+      "loss": 0.5971,
+      "step": 2944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4861718673333035,
+      "learning_rate": 2.330314729917481e-06,
+      "loss": 1.1437,
+      "step": 2945
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.315633629693504,
+      "learning_rate": 2.3247778470585926e-06,
+      "loss": 1.1518,
+      "step": 2946
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8409435267998016,
+      "learning_rate": 2.319246684601869e-06,
+      "loss": 0.51,
+      "step": 2947
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6193259100110908,
+      "learning_rate": 2.313721246669717e-06,
+      "loss": 0.5809,
+      "step": 2948
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7435116073554082,
+      "learning_rate": 2.308201537380299e-06,
+      "loss": 0.6211,
+      "step": 2949
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7716282859806811,
+      "learning_rate": 2.302687560847484e-06,
+      "loss": 0.5783,
+      "step": 2950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7283824015676006,
+      "learning_rate": 2.2971793211808924e-06,
+      "loss": 0.5959,
+      "step": 2951
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.734535274533178,
+      "learning_rate": 2.291676822485852e-06,
+      "loss": 0.591,
+      "step": 2952
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3385546900635066,
+      "learning_rate": 2.2861800688634196e-06,
+      "loss": 1.0522,
+      "step": 2953
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8331830221776728,
+      "learning_rate": 2.280689064410364e-06,
+      "loss": 0.5601,
+      "step": 2954
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2809331935009327,
+      "learning_rate": 2.27520381321918e-06,
+      "loss": 1.0807,
+      "step": 2955
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6765167964072207,
+      "learning_rate": 2.269724319378063e-06,
+      "loss": 0.5473,
+      "step": 2956
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.780579738796459,
+      "learning_rate": 2.264250586970923e-06,
+      "loss": 0.5138,
+      "step": 2957
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7750948112718976,
+      "learning_rate": 2.2587826200773743e-06,
+      "loss": 0.6005,
+      "step": 2958
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8201927135307612,
+      "learning_rate": 2.2533204227727323e-06,
+      "loss": 0.6235,
+      "step": 2959
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7537577009040635,
+      "learning_rate": 2.2478639991280236e-06,
+      "loss": 0.6079,
+      "step": 2960
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7375542926131273,
+      "learning_rate": 2.24241335320995e-06,
+      "loss": 0.6035,
+      "step": 2961
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8383607850342114,
+      "learning_rate": 2.2369684890809305e-06,
+      "loss": 0.6377,
+      "step": 2962
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3053274390680563,
+      "learning_rate": 2.2315294107990603e-06,
+      "loss": 1.1225,
+      "step": 2963
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.588914544303894,
+      "learning_rate": 2.2260961224181254e-06,
+      "loss": 0.5678,
+      "step": 2964
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3026573675193935,
+      "learning_rate": 2.220668627987599e-06,
+      "loss": 1.1123,
+      "step": 2965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8221615206719717,
+      "learning_rate": 2.2152469315526327e-06,
+      "loss": 0.6002,
+      "step": 2966
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5571921869231735,
+      "learning_rate": 2.2098310371540578e-06,
+      "loss": 1.1234,
+      "step": 2967
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8869162761779597,
+      "learning_rate": 2.2044209488283895e-06,
+      "loss": 0.6311,
+      "step": 2968
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9585251489193756,
+      "learning_rate": 2.1990166706077975e-06,
+      "loss": 0.5824,
+      "step": 2969
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7283173016517213,
+      "learning_rate": 2.1936182065201395e-06,
+      "loss": 0.6105,
+      "step": 2970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.317394380857667,
+      "learning_rate": 2.18822556058893e-06,
+      "loss": 1.112,
+      "step": 2971
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6610557523975433,
+      "learning_rate": 2.1828387368333504e-06,
+      "loss": 0.6078,
+      "step": 2972
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7257063701437079,
+      "learning_rate": 2.1774577392682405e-06,
+      "loss": 0.5728,
+      "step": 2973
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5071563032005924,
+      "learning_rate": 2.172082571904097e-06,
+      "loss": 1.1318,
+      "step": 2974
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0057059846879506,
+      "learning_rate": 2.166713238747079e-06,
+      "loss": 0.6219,
+      "step": 2975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8427368286896226,
+      "learning_rate": 2.1613497437989893e-06,
+      "loss": 0.5764,
+      "step": 2976
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7017933241242422,
+      "learning_rate": 2.155992091057282e-06,
+      "loss": 0.6016,
+      "step": 2977
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7839972490317615,
+      "learning_rate": 2.1506402845150543e-06,
+      "loss": 0.6226,
+      "step": 2978
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7462710242465418,
+      "learning_rate": 2.145294328161056e-06,
+      "loss": 0.601,
+      "step": 2979
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8599816225601826,
+      "learning_rate": 2.1399542259796602e-06,
+      "loss": 0.552,
+      "step": 2980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9299461681211998,
+      "learning_rate": 2.134619981950896e-06,
+      "loss": 0.5567,
+      "step": 2981
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.721314154031287,
+      "learning_rate": 2.129291600050406e-06,
+      "loss": 0.5505,
+      "step": 2982
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5796784991286177,
+      "learning_rate": 2.123969084249482e-06,
+      "loss": 0.6287,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7617760137597294,
+      "learning_rate": 2.1186524385150343e-06,
+      "loss": 0.601,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7931263724998454,
+      "learning_rate": 2.1133416668095986e-06,
+      "loss": 0.6177,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8840744324196315,
+      "learning_rate": 2.108036773091332e-06,
+      "loss": 0.6814,
+      "step": 2986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7654741808494179,
+      "learning_rate": 2.10273776131402e-06,
+      "loss": 0.5899,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.50668007828263,
+      "learning_rate": 2.097444635427047e-06,
+      "loss": 1.1029,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9880505890123819,
+      "learning_rate": 2.0921573993754285e-06,
+      "loss": 0.5431,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8886144032438454,
+      "learning_rate": 2.0868760570997794e-06,
+      "loss": 0.5724,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.213624477847324,
+      "learning_rate": 2.081600612536324e-06,
+      "loss": 1.0395,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7322360022743888,
+      "learning_rate": 2.0763310696168927e-06,
+      "loss": 0.559,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.26641368102322,
+      "learning_rate": 2.071067432268913e-06,
+      "loss": 1.0602,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2937479767762547,
+      "learning_rate": 2.065809704415421e-06,
+      "loss": 1.1158,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.662805879069354,
+      "learning_rate": 2.0605578899750377e-06,
+      "loss": 0.5602,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8502110470740916,
+      "learning_rate": 2.05531199286198e-06,
+      "loss": 0.6019,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7279317541520518,
+      "learning_rate": 2.0500720169860588e-06,
+      "loss": 0.5504,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7316730246101273,
+      "learning_rate": 2.044837966252665e-06,
+      "loss": 0.5244,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7380087025703876,
+      "learning_rate": 2.0396098445627755e-06,
+      "loss": 0.5637,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6953320525167654,
+      "learning_rate": 2.0343876558129583e-06,
+      "loss": 0.5833,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.034673959559532,
+      "learning_rate": 2.0291714038953394e-06,
+      "loss": 0.5646,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.836367396001169,
+      "learning_rate": 2.0239610926976395e-06,
+      "loss": 0.6281,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7803265725235728,
+      "learning_rate": 2.0187567261031415e-06,
+      "loss": 0.5844,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8220865909174082,
+      "learning_rate": 2.0135583079906996e-06,
+      "loss": 0.5747,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.234339983795349,
+      "learning_rate": 2.0083658422347354e-06,
+      "loss": 1.1151,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.911248312458243,
+      "learning_rate": 2.0031793327052317e-06,
+      "loss": 0.5814,
+      "step": 3006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8857402844076365,
+      "learning_rate": 1.997998783267734e-06,
+      "loss": 0.649,
+      "step": 3007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.262694490999627,
+      "learning_rate": 1.9928241977833494e-06,
+      "loss": 1.1123,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.653428098845633,
+      "learning_rate": 1.987655580108734e-06,
+      "loss": 0.6182,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8470842622408854,
+      "learning_rate": 1.9824929340960976e-06,
+      "loss": 0.6738,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7968147598830921,
+      "learning_rate": 1.9773362635932e-06,
+      "loss": 0.6036,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.789044001319078,
+      "learning_rate": 1.972185572443347e-06,
+      "loss": 0.5845,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8220794156142592,
+      "learning_rate": 1.9670408644853943e-06,
+      "loss": 0.6463,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6389511544064992,
+      "learning_rate": 1.961902143553722e-06,
+      "loss": 0.5573,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.73809009069911,
+      "learning_rate": 1.9567694134782665e-06,
+      "loss": 0.5626,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5232171430812396,
+      "learning_rate": 1.9516426780844887e-06,
+      "loss": 1.054,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8148429387934688,
+      "learning_rate": 1.9465219411933833e-06,
+      "loss": 0.5189,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.948196088269232,
+      "learning_rate": 1.941407206621473e-06,
+      "loss": 0.551,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3680969451748504,
+      "learning_rate": 1.936298478180817e-06,
+      "loss": 1.1099,
+      "step": 3019
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8547514176809856,
+      "learning_rate": 1.9311957596789776e-06,
+      "loss": 0.5657,
+      "step": 3020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.618629320891213,
+      "learning_rate": 1.9260990549190628e-06,
+      "loss": 1.1796,
+      "step": 3021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7098095311797739,
+      "learning_rate": 1.921008367699675e-06,
+      "loss": 0.5327,
+      "step": 3022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9139953507748049,
+      "learning_rate": 1.915923701814949e-06,
+      "loss": 0.6985,
+      "step": 3023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.893080610289198,
+      "learning_rate": 1.910845061054524e-06,
+      "loss": 0.5819,
+      "step": 3024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5862420636969996,
+      "learning_rate": 1.905772449203549e-06,
+      "loss": 1.0922,
+      "step": 3025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.042458084659804,
+      "learning_rate": 1.9007058700426805e-06,
+      "loss": 0.5864,
+      "step": 3026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7422361758118847,
+      "learning_rate": 1.8956453273480813e-06,
+      "loss": 0.5765,
+      "step": 3027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4716610496133944,
+      "learning_rate": 1.8905908248914117e-06,
+      "loss": 1.1268,
+      "step": 3028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7154865799756873,
+      "learning_rate": 1.885542366439832e-06,
+      "loss": 0.5575,
+      "step": 3029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2991643367967454,
+      "learning_rate": 1.8804999557559955e-06,
+      "loss": 1.0896,
+      "step": 3030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7074679757775262,
+      "learning_rate": 1.8754635965980484e-06,
+      "loss": 0.5595,
+      "step": 3031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7781520488203728,
+      "learning_rate": 1.8704332927196345e-06,
+      "loss": 0.5782,
+      "step": 3032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.822873662034147,
+      "learning_rate": 1.8654090478698684e-06,
+      "loss": 0.5933,
+      "step": 3033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8305936164327308,
+      "learning_rate": 1.8603908657933666e-06,
+      "loss": 0.5843,
+      "step": 3034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6498307001469696,
+      "learning_rate": 1.8553787502302156e-06,
+      "loss": 0.584,
+      "step": 3035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.837647296176136,
+      "learning_rate": 1.8503727049159837e-06,
+      "loss": 0.5779,
+      "step": 3036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8714339932028659,
+      "learning_rate": 1.8453727335817152e-06,
+      "loss": 0.5951,
+      "step": 3037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3415894994914765,
+      "learning_rate": 1.8403788399539268e-06,
+      "loss": 1.1138,
+      "step": 3038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.609645291124782,
+      "learning_rate": 1.835391027754605e-06,
+      "loss": 0.621,
+      "step": 3039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7731497525200055,
+      "learning_rate": 1.8304093007012102e-06,
+      "loss": 0.6163,
+      "step": 3040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.461678985021767,
+      "learning_rate": 1.8254336625066526e-06,
+      "loss": 1.0666,
+      "step": 3041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7781617832765866,
+      "learning_rate": 1.8204641168793212e-06,
+      "loss": 0.6331,
+      "step": 3042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.916278852331029,
+      "learning_rate": 1.8155006675230547e-06,
+      "loss": 0.5732,
+      "step": 3043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7243941470499613,
+      "learning_rate": 1.810543318137149e-06,
+      "loss": 0.6212,
+      "step": 3044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7948301959575241,
+      "learning_rate": 1.8055920724163556e-06,
+      "loss": 0.5158,
+      "step": 3045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.823405644843727,
+      "learning_rate": 1.8006469340508725e-06,
+      "loss": 0.6502,
+      "step": 3046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2756741690810736,
+      "learning_rate": 1.795707906726356e-06,
+      "loss": 1.1053,
+      "step": 3047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.866169467734974,
+      "learning_rate": 1.7907749941238972e-06,
+      "loss": 0.5749,
+      "step": 3048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.767118494816128,
+      "learning_rate": 1.785848199920035e-06,
+      "loss": 0.606,
+      "step": 3049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4154710825144914,
+      "learning_rate": 1.7809275277867443e-06,
+      "loss": 1.1306,
+      "step": 3050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.118810839657473,
+      "learning_rate": 1.7760129813914474e-06,
+      "loss": 0.6183,
+      "step": 3051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8298195466955687,
+      "learning_rate": 1.7711045643969816e-06,
+      "loss": 0.5859,
+      "step": 3052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7164406730356239,
+      "learning_rate": 1.7662022804616407e-06,
+      "loss": 0.5844,
+      "step": 3053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.766655873998742,
+      "learning_rate": 1.7613061332391245e-06,
+      "loss": 0.6151,
+      "step": 3054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7184654476875225,
+      "learning_rate": 1.756416126378575e-06,
+      "loss": 0.5905,
+      "step": 3055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.84434425578982,
+      "learning_rate": 1.751532263524549e-06,
+      "loss": 0.6033,
+      "step": 3056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6431967580872746,
+      "learning_rate": 1.7466545483170305e-06,
+      "loss": 0.5517,
+      "step": 3057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.334484205967325,
+      "learning_rate": 1.7417829843914125e-06,
+      "loss": 1.1136,
+      "step": 3058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9265175471028286,
+      "learning_rate": 1.7369175753785172e-06,
+      "loss": 0.5836,
+      "step": 3059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.936759726087052,
+      "learning_rate": 1.7320583249045619e-06,
+      "loss": 0.5562,
+      "step": 3060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8064766464142143,
+      "learning_rate": 1.727205236591194e-06,
+      "loss": 0.5861,
+      "step": 3061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8250452397847081,
+      "learning_rate": 1.7223583140554457e-06,
+      "loss": 0.6048,
+      "step": 3062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8794387961773418,
+      "learning_rate": 1.717517560909775e-06,
+      "loss": 0.6176,
+      "step": 3063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3177730439000994,
+      "learning_rate": 1.7126829807620294e-06,
+      "loss": 1.0601,
+      "step": 3064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4466626999112484,
+      "learning_rate": 1.7078545772154565e-06,
+      "loss": 1.1487,
+      "step": 3065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.377566801764708,
+      "learning_rate": 1.7030323538687066e-06,
+      "loss": 1.0968,
+      "step": 3066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.381263492867772,
+      "learning_rate": 1.6982163143158192e-06,
+      "loss": 1.1248,
+      "step": 3067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8881110698706238,
+      "learning_rate": 1.6934064621462232e-06,
+      "loss": 0.6151,
+      "step": 3068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8714094941999915,
+      "learning_rate": 1.6886028009447408e-06,
+      "loss": 0.6316,
+      "step": 3069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8564353064251196,
+      "learning_rate": 1.683805334291576e-06,
+      "loss": 0.5705,
+      "step": 3070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.880107880857629,
+      "learning_rate": 1.6790140657623155e-06,
+      "loss": 0.6171,
+      "step": 3071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7686324549743557,
+      "learning_rate": 1.674228998927937e-06,
+      "loss": 0.5903,
+      "step": 3072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.804296973465962,
+      "learning_rate": 1.6694501373547767e-06,
+      "loss": 0.5725,
+      "step": 3073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.011816509565535,
+      "learning_rate": 1.6646774846045645e-06,
+      "loss": 0.6265,
+      "step": 3074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6594162727423165,
+      "learning_rate": 1.6599110442343935e-06,
+      "loss": 0.597,
+      "step": 3075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7490843999519685,
+      "learning_rate": 1.6551508197967281e-06,
+      "loss": 0.6198,
+      "step": 3076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.640373761697442,
+      "learning_rate": 1.6503968148394001e-06,
+      "loss": 0.5777,
+      "step": 3077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.354830591491219,
+      "learning_rate": 1.6456490329056052e-06,
+      "loss": 1.093,
+      "step": 3078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.418862164817516,
+      "learning_rate": 1.6409074775339007e-06,
+      "loss": 1.1794,
+      "step": 3079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7947222123711744,
+      "learning_rate": 1.6361721522582109e-06,
+      "loss": 0.5943,
+      "step": 3080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3616167602139466,
+      "learning_rate": 1.6314430606078002e-06,
+      "loss": 1.1184,
+      "step": 3081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9878703463792025,
+      "learning_rate": 1.6267202061073052e-06,
+      "loss": 0.7325,
+      "step": 3082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7920607591752713,
+      "learning_rate": 1.622003592276702e-06,
+      "loss": 0.6143,
+      "step": 3083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8524980439943268,
+      "learning_rate": 1.617293222631321e-06,
+      "loss": 0.6633,
+      "step": 3084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7707147075686305,
+      "learning_rate": 1.6125891006818351e-06,
+      "loss": 0.5865,
+      "step": 3085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8807244206442562,
+      "learning_rate": 1.6078912299342608e-06,
+      "loss": 0.6064,
+      "step": 3086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1815786643393427,
+      "learning_rate": 1.603199613889963e-06,
+      "loss": 1.0935,
+      "step": 3087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7503244689126138,
+      "learning_rate": 1.5985142560456367e-06,
+      "loss": 0.5293,
+      "step": 3088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6991876815544118,
+      "learning_rate": 1.5938351598933145e-06,
+      "loss": 0.5314,
+      "step": 3089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.202441919403861,
+      "learning_rate": 1.5891623289203594e-06,
+      "loss": 1.0283,
+      "step": 3090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3347677353488856,
+      "learning_rate": 1.5844957666094785e-06,
+      "loss": 1.0639,
+      "step": 3091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8284237320018657,
+      "learning_rate": 1.5798354764386837e-06,
+      "loss": 0.6095,
+      "step": 3092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8865974940601398,
+      "learning_rate": 1.5751814618813377e-06,
+      "loss": 0.565,
+      "step": 3093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.839372531155159,
+      "learning_rate": 1.5705337264061004e-06,
+      "loss": 0.5799,
+      "step": 3094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9212429997332807,
+      "learning_rate": 1.5658922734769754e-06,
+      "loss": 0.6437,
+      "step": 3095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5960999976397678,
+      "learning_rate": 1.5612571065532688e-06,
+      "loss": 0.5934,
+      "step": 3096
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7982352819407252,
+      "learning_rate": 1.5566282290896073e-06,
+      "loss": 0.5566,
+      "step": 3097
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8637229917338636,
+      "learning_rate": 1.5520056445359288e-06,
+      "loss": 0.5763,
+      "step": 3098
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7283553944864043,
+      "learning_rate": 1.5473893563374799e-06,
+      "loss": 0.5989,
+      "step": 3099
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.65627891863996,
+      "learning_rate": 1.5427793679348147e-06,
+      "loss": 0.6172,
+      "step": 3100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7350831598008507,
+      "learning_rate": 1.5381756827637972e-06,
+      "loss": 0.6394,
+      "step": 3101
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6836764074574069,
+      "learning_rate": 1.5335783042555874e-06,
+      "loss": 0.5826,
+      "step": 3102
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9643906716615849,
+      "learning_rate": 1.5289872358366475e-06,
+      "loss": 0.5832,
+      "step": 3103
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2386902117352996,
+      "learning_rate": 1.5244024809287338e-06,
+      "loss": 1.0706,
+      "step": 3104
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5709572895978545,
+      "learning_rate": 1.5198240429488997e-06,
+      "loss": 1.1059,
+      "step": 3105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.776341819144069,
+      "learning_rate": 1.5152519253094956e-06,
+      "loss": 0.5975,
+      "step": 3106
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7878805059518166,
+      "learning_rate": 1.5106861314181475e-06,
+      "loss": 0.5543,
+      "step": 3107
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2847685547411567,
+      "learning_rate": 1.506126664677784e-06,
+      "loss": 1.0838,
+      "step": 3108
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.845651156012885,
+      "learning_rate": 1.5015735284866074e-06,
+      "loss": 0.6776,
+      "step": 3109
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7431185366173938,
+      "learning_rate": 1.4970267262381056e-06,
+      "loss": 0.515,
+      "step": 3110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9087723371824719,
+      "learning_rate": 1.4924862613210421e-06,
+      "loss": 0.5658,
+      "step": 3111
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.322841071093742,
+      "learning_rate": 1.4879521371194694e-06,
+      "loss": 1.0999,
+      "step": 3112
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3408014030905515,
+      "learning_rate": 1.483424357012695e-06,
+      "loss": 1.1008,
+      "step": 3113
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2571064494569444,
+      "learning_rate": 1.4789029243753173e-06,
+      "loss": 1.0887,
+      "step": 3114
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8050410457218686,
+      "learning_rate": 1.4743878425771852e-06,
+      "loss": 0.5721,
+      "step": 3115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5977256866858536,
+      "learning_rate": 1.4698791149834312e-06,
+      "loss": 1.1009,
+      "step": 3116
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.606645214633525,
+      "learning_rate": 1.4653767449544421e-06,
+      "loss": 0.5454,
+      "step": 3117
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.406769793468292,
+      "learning_rate": 1.46088073584587e-06,
+      "loss": 1.032,
+      "step": 3118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.781635373538149,
+      "learning_rate": 1.456391091008622e-06,
+      "loss": 0.5708,
+      "step": 3119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7642119891666477,
+      "learning_rate": 1.4519078137888688e-06,
+      "loss": 0.6396,
+      "step": 3120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.359502251607647,
+      "learning_rate": 1.4474309075280313e-06,
+      "loss": 1.0983,
+      "step": 3121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7522820560891281,
+      "learning_rate": 1.4429603755627796e-06,
+      "loss": 0.5982,
+      "step": 3122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7539746908141838,
+      "learning_rate": 1.4384962212250374e-06,
+      "loss": 0.6244,
+      "step": 3123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3333058188387765,
+      "learning_rate": 1.434038447841971e-06,
+      "loss": 1.0622,
+      "step": 3124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4462917402897313,
+      "learning_rate": 1.4295870587359995e-06,
+      "loss": 1.047,
+      "step": 3125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3872510578116923,
+      "learning_rate": 1.4251420572247677e-06,
+      "loss": 1.1028,
+      "step": 3126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.687103592291555,
+      "learning_rate": 1.4207034466211778e-06,
+      "loss": 1.1143,
+      "step": 3127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8770220292795308,
+      "learning_rate": 1.416271230233357e-06,
+      "loss": 0.6073,
+      "step": 3128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8616459583771268,
+      "learning_rate": 1.4118454113646706e-06,
+      "loss": 0.6388,
+      "step": 3129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3331602358007633,
+      "learning_rate": 1.4074259933137146e-06,
+      "loss": 1.0928,
+      "step": 3130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.818128815869297,
+      "learning_rate": 1.4030129793743175e-06,
+      "loss": 0.6165,
+      "step": 3131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9580093092157218,
+      "learning_rate": 1.3986063728355293e-06,
+      "loss": 0.6507,
+      "step": 3132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.632336168196171,
+      "learning_rate": 1.394206176981633e-06,
+      "loss": 0.5551,
+      "step": 3133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8665144808631446,
+      "learning_rate": 1.3898123950921216e-06,
+      "loss": 0.6144,
+      "step": 3134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7566420948484325,
+      "learning_rate": 1.3854250304417205e-06,
+      "loss": 0.5988,
+      "step": 3135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8616277495175593,
+      "learning_rate": 1.3810440863003628e-06,
+      "loss": 0.6573,
+      "step": 3136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5509065897162935,
+      "learning_rate": 1.376669565933202e-06,
+      "loss": 0.5971,
+      "step": 3137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.766281241257663,
+      "learning_rate": 1.3723014726006012e-06,
+      "loss": 0.6084,
+      "step": 3138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.27678491155514,
+      "learning_rate": 1.367939809558131e-06,
+      "loss": 1.1396,
+      "step": 3139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8711218236816245,
+      "learning_rate": 1.3635845800565783e-06,
+      "loss": 0.5826,
+      "step": 3140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.933672653502336,
+      "learning_rate": 1.359235787341925e-06,
+      "loss": 0.6012,
+      "step": 3141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6984753780727335,
+      "learning_rate": 1.354893434655361e-06,
+      "loss": 0.6032,
+      "step": 3142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.362570875408876,
+      "learning_rate": 1.3505575252332737e-06,
+      "loss": 1.1071,
+      "step": 3143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3054313946725955,
+      "learning_rate": 1.3462280623072543e-06,
+      "loss": 1.1484,
+      "step": 3144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2574104762983933,
+      "learning_rate": 1.3419050491040775e-06,
+      "loss": 1.0807,
+      "step": 3145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4665136561258296,
+      "learning_rate": 1.337588488845728e-06,
+      "loss": 1.0984,
+      "step": 3146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.964660848065622,
+      "learning_rate": 1.3332783847493603e-06,
+      "loss": 0.6083,
+      "step": 3147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8134605046184349,
+      "learning_rate": 1.3289747400273356e-06,
+      "loss": 0.4979,
+      "step": 3148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7520769768523727,
+      "learning_rate": 1.324677557887193e-06,
+      "loss": 0.5471,
+      "step": 3149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.778782372771125,
+      "learning_rate": 1.3203868415316535e-06,
+      "loss": 0.6373,
+      "step": 3150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.554564835301154,
+      "learning_rate": 1.316102594158618e-06,
+      "loss": 0.5292,
+      "step": 3151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.402778863892697,
+      "learning_rate": 1.3118248189611782e-06,
+      "loss": 1.1467,
+      "step": 3152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3209187196133465,
+      "learning_rate": 1.3075535191275823e-06,
+      "loss": 1.1628,
+      "step": 3153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8205846742173852,
+      "learning_rate": 1.3032886978412706e-06,
+      "loss": 0.6098,
+      "step": 3154
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2506267956131003,
+      "learning_rate": 1.299030358280845e-06,
+      "loss": 1.1075,
+      "step": 3155
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.724681879916815,
+      "learning_rate": 1.2947785036200778e-06,
+      "loss": 0.582,
+      "step": 3156
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9479406582956347,
+      "learning_rate": 1.2905331370279095e-06,
+      "loss": 0.6155,
+      "step": 3157
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.390878226785235,
+      "learning_rate": 1.2862942616684415e-06,
+      "loss": 1.1147,
+      "step": 3158
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9736043934944503,
+      "learning_rate": 1.2820618807009466e-06,
+      "loss": 0.6208,
+      "step": 3159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2331548159835624,
+      "learning_rate": 1.277835997279847e-06,
+      "loss": 1.0603,
+      "step": 3160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8020390896040024,
+      "learning_rate": 1.2736166145547268e-06,
+      "loss": 0.5556,
+      "step": 3161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.836897987166998,
+      "learning_rate": 1.269403735670325e-06,
+      "loss": 0.6,
+      "step": 3162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8200421700911464,
+      "learning_rate": 1.2651973637665316e-06,
+      "loss": 0.52,
+      "step": 3163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2814371547472305,
+      "learning_rate": 1.260997501978386e-06,
+      "loss": 1.1249,
+      "step": 3164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7612222548663345,
+      "learning_rate": 1.2568041534360842e-06,
+      "loss": 0.5641,
+      "step": 3165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8114423907055708,
+      "learning_rate": 1.2526173212649529e-06,
+      "loss": 0.6034,
+      "step": 3166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7007102999864012,
+      "learning_rate": 1.2484370085854757e-06,
+      "loss": 0.5652,
+      "step": 3167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7473401763250422,
+      "learning_rate": 1.2442632185132707e-06,
+      "loss": 0.6437,
+      "step": 3168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2739138600155933,
+      "learning_rate": 1.2400959541590962e-06,
+      "loss": 1.1164,
+      "step": 3169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3506585950766423,
+      "learning_rate": 1.2359352186288453e-06,
+      "loss": 1.1576,
+      "step": 3170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.864071844788345,
+      "learning_rate": 1.2317810150235465e-06,
+      "loss": 0.5793,
+      "step": 3171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.875270039418249,
+      "learning_rate": 1.2276333464393574e-06,
+      "loss": 0.6014,
+      "step": 3172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5759102408108892,
+      "learning_rate": 1.2234922159675732e-06,
+      "loss": 0.5645,
+      "step": 3173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7030546218652305,
+      "learning_rate": 1.2193576266946061e-06,
+      "loss": 0.5697,
+      "step": 3174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8107232547777703,
+      "learning_rate": 1.2152295817019977e-06,
+      "loss": 0.5591,
+      "step": 3175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6787137222998436,
+      "learning_rate": 1.2111080840664125e-06,
+      "loss": 0.5755,
+      "step": 3176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7619715162895824,
+      "learning_rate": 1.206993136859632e-06,
+      "loss": 0.6174,
+      "step": 3177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7867349368024605,
+      "learning_rate": 1.2028847431485647e-06,
+      "loss": 0.6431,
+      "step": 3178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7598370857623646,
+      "learning_rate": 1.1987829059952184e-06,
+      "loss": 0.5621,
+      "step": 3179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.827691985026925,
+      "learning_rate": 1.194687628456731e-06,
+      "loss": 0.5944,
+      "step": 3180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7888844409847762,
+      "learning_rate": 1.1905989135853434e-06,
+      "loss": 0.5974,
+      "step": 3181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6314440086420892,
+      "learning_rate": 1.1865167644284047e-06,
+      "loss": 0.505,
+      "step": 3182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.757986114804363,
+      "learning_rate": 1.18244118402837e-06,
+      "loss": 0.5432,
+      "step": 3183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.29350941110979,
+      "learning_rate": 1.1783721754228083e-06,
+      "loss": 1.091,
+      "step": 3184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.378753053074132,
+      "learning_rate": 1.1743097416443737e-06,
+      "loss": 1.0878,
+      "step": 3185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6827916896623272,
+      "learning_rate": 1.170253885720838e-06,
+      "loss": 0.5688,
+      "step": 3186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4363013706234944,
+      "learning_rate": 1.1662046106750524e-06,
+      "loss": 1.1148,
+      "step": 3187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2773574902342415,
+      "learning_rate": 1.1621619195249812e-06,
+      "loss": 1.1213,
+      "step": 3188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7951006340915578,
+      "learning_rate": 1.1581258152836693e-06,
+      "loss": 0.6041,
+      "step": 3189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6390743593673023,
+      "learning_rate": 1.1540963009592576e-06,
+      "loss": 0.5672,
+      "step": 3190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.832005152214475,
+      "learning_rate": 1.1500733795549712e-06,
+      "loss": 0.646,
+      "step": 3191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8616819606296537,
+      "learning_rate": 1.1460570540691285e-06,
+      "loss": 0.6082,
+      "step": 3192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6883846398440385,
+      "learning_rate": 1.142047327495126e-06,
+      "loss": 0.5452,
+      "step": 3193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6663891933749346,
+      "learning_rate": 1.1380442028214445e-06,
+      "loss": 0.5174,
+      "step": 3194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8564957335011218,
+      "learning_rate": 1.1340476830316427e-06,
+      "loss": 0.5819,
+      "step": 3195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7796658329216846,
+      "learning_rate": 1.130057771104356e-06,
+      "loss": 0.5636,
+      "step": 3196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.751307888172845,
+      "learning_rate": 1.1260744700133008e-06,
+      "loss": 0.567,
+      "step": 3197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.839649928578178,
+      "learning_rate": 1.1220977827272572e-06,
+      "loss": 0.6593,
+      "step": 3198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.24344582319898,
+      "learning_rate": 1.1181277122100842e-06,
+      "loss": 1.0962,
+      "step": 3199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6443166336746982,
+      "learning_rate": 1.114164261420706e-06,
+      "loss": 0.5274,
+      "step": 3200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7822697150119169,
+      "learning_rate": 1.1102074333131108e-06,
+      "loss": 0.5936,
+      "step": 3201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9557385147046924,
+      "learning_rate": 1.106257230836354e-06,
+      "loss": 0.5825,
+      "step": 3202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8504783867840346,
+      "learning_rate": 1.1023136569345526e-06,
+      "loss": 0.5934,
+      "step": 3203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9090215826717731,
+      "learning_rate": 1.0983767145468804e-06,
+      "loss": 0.6189,
+      "step": 3204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.270098453212827,
+      "learning_rate": 1.0944464066075766e-06,
+      "loss": 1.0902,
+      "step": 3205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.442714185705128,
+      "learning_rate": 1.0905227360459213e-06,
+      "loss": 1.0878,
+      "step": 3206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5840922185933135,
+      "learning_rate": 1.0866057057862634e-06,
+      "loss": 0.5541,
+      "step": 3207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2978439862335853,
+      "learning_rate": 1.0826953187479928e-06,
+      "loss": 1.0993,
+      "step": 3208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.052442227104053,
+      "learning_rate": 1.0787915778455515e-06,
+      "loss": 0.6766,
+      "step": 3209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.299565733898223,
+      "learning_rate": 1.074894485988428e-06,
+      "loss": 1.0799,
+      "step": 3210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.286085365029018,
+      "learning_rate": 1.0710040460811522e-06,
+      "loss": 1.082,
+      "step": 3211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3469072989853674,
+      "learning_rate": 1.0671202610233045e-06,
+      "loss": 1.0719,
+      "step": 3212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.280953698238122,
+      "learning_rate": 1.063243133709495e-06,
+      "loss": 1.138,
+      "step": 3213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.921280475222433,
+      "learning_rate": 1.0593726670293792e-06,
+      "loss": 0.6119,
+      "step": 3214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8627947092250372,
+      "learning_rate": 1.0555088638676414e-06,
+      "loss": 0.6006,
+      "step": 3215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8076948504298163,
+      "learning_rate": 1.0516517271040128e-06,
+      "loss": 0.5788,
+      "step": 3216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9542898192324485,
+      "learning_rate": 1.047801259613236e-06,
+      "loss": 0.6702,
+      "step": 3217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8081206275599249,
+      "learning_rate": 1.043957464265104e-06,
+      "loss": 0.6419,
+      "step": 3218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8874079761717049,
+      "learning_rate": 1.0401203439244179e-06,
+      "loss": 0.5748,
+      "step": 3219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4232275577296023,
+      "learning_rate": 1.0362899014510197e-06,
+      "loss": 1.0496,
+      "step": 3220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8876533151221857,
+      "learning_rate": 1.0324661396997671e-06,
+      "loss": 0.6195,
+      "step": 3221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8852976568123214,
+      "learning_rate": 1.0286490615205358e-06,
+      "loss": 0.5842,
+      "step": 3222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.295258187758776,
+      "learning_rate": 1.0248386697582246e-06,
+      "loss": 1.0688,
+      "step": 3223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6380199262599144,
+      "learning_rate": 1.021034967252752e-06,
+      "loss": 0.5311,
+      "step": 3224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.751491380147118,
+      "learning_rate": 1.01723795683904e-06,
+      "loss": 0.653,
+      "step": 3225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.880459812111415,
+      "learning_rate": 1.013447641347034e-06,
+      "loss": 0.5603,
+      "step": 3226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8766410921778889,
+      "learning_rate": 1.0096640236016853e-06,
+      "loss": 0.5299,
+      "step": 3227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9453416633921747,
+      "learning_rate": 1.0058871064229526e-06,
+      "loss": 0.5935,
+      "step": 3228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7417758682849873,
+      "learning_rate": 1.0021168926258006e-06,
+      "loss": 0.5939,
+      "step": 3229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2527687401250147,
+      "learning_rate": 9.983533850201975e-07,
+      "loss": 1.1572,
+      "step": 3230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3733509649951907,
+      "learning_rate": 9.945965864111185e-07,
+      "loss": 1.0862,
+      "step": 3231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9283786182351077,
+      "learning_rate": 9.908464995985322e-07,
+      "loss": 0.6104,
+      "step": 3232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8067908426399286,
+      "learning_rate": 9.871031273774079e-07,
+      "loss": 0.6101,
+      "step": 3233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7564830847936836,
+      "learning_rate": 9.833664725377079e-07,
+      "loss": 0.5816,
+      "step": 3234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8576006944446188,
+      "learning_rate": 9.796365378643912e-07,
+      "loss": 0.5654,
+      "step": 3235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4085582649921733,
+      "learning_rate": 9.759133261374054e-07,
+      "loss": 1.0819,
+      "step": 3236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.78990959601853,
+      "learning_rate": 9.721968401316928e-07,
+      "loss": 0.6161,
+      "step": 3237
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7359513232370707,
+      "learning_rate": 9.684870826171733e-07,
+      "loss": 0.5972,
+      "step": 3238
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8382796908878418,
+      "learning_rate": 9.64784056358763e-07,
+      "loss": 0.6327,
+      "step": 3239
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.966689721077219,
+      "learning_rate": 9.61087764116353e-07,
+      "loss": 0.6067,
+      "step": 3240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9165726647843087,
+      "learning_rate": 9.573982086448197e-07,
+      "loss": 0.6438,
+      "step": 3241
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8992322285049388,
+      "learning_rate": 9.537153926940156e-07,
+      "loss": 0.6563,
+      "step": 3242
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7489404794311154,
+      "learning_rate": 9.500393190087731e-07,
+      "loss": 0.5737,
+      "step": 3243
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.797416232855419,
+      "learning_rate": 9.463699903288958e-07,
+      "loss": 0.5426,
+      "step": 3244
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3214430596183653,
+      "learning_rate": 9.427074093891675e-07,
+      "loss": 1.0272,
+      "step": 3245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7259190791565961,
+      "learning_rate": 9.390515789193333e-07,
+      "loss": 0.4841,
+      "step": 3246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8452172330785535,
+      "learning_rate": 9.354025016441148e-07,
+      "loss": 0.5414,
+      "step": 3247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0209967100760933,
+      "learning_rate": 9.317601802831966e-07,
+      "loss": 0.6054,
+      "step": 3248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.572329844015877,
+      "learning_rate": 9.281246175512249e-07,
+      "loss": 1.0365,
+      "step": 3249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5660860014871345,
+      "learning_rate": 9.24495816157821e-07,
+      "loss": 0.5396,
+      "step": 3250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8335256182589998,
+      "learning_rate": 9.208737788075495e-07,
+      "loss": 0.5164,
+      "step": 3251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8437754551425447,
+      "learning_rate": 9.172585081999497e-07,
+      "loss": 0.5634,
+      "step": 3252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3550879939471585,
+      "learning_rate": 9.136500070295062e-07,
+      "loss": 1.0469,
+      "step": 3253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8162030253523151,
+      "learning_rate": 9.100482779856656e-07,
+      "loss": 0.6038,
+      "step": 3254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8605322391265808,
+      "learning_rate": 9.064533237528205e-07,
+      "loss": 0.5946,
+      "step": 3255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9217793815950845,
+      "learning_rate": 9.028651470103244e-07,
+      "loss": 0.6391,
+      "step": 3256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3188942383325553,
+      "learning_rate": 8.992837504324648e-07,
+      "loss": 1.0483,
+      "step": 3257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7192629766333123,
+      "learning_rate": 8.957091366884929e-07,
+      "loss": 0.5824,
+      "step": 3258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.857636200533355,
+      "learning_rate": 8.921413084425867e-07,
+      "loss": 0.5981,
+      "step": 3259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6238891181945279,
+      "learning_rate": 8.885802683538835e-07,
+      "loss": 0.5504,
+      "step": 3260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8758037727955825,
+      "learning_rate": 8.850260190764504e-07,
+      "loss": 0.6609,
+      "step": 3261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6624022120579753,
+      "learning_rate": 8.81478563259297e-07,
+      "loss": 0.4897,
+      "step": 3262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.770470481747253,
+      "learning_rate": 8.779379035463698e-07,
+      "loss": 0.6188,
+      "step": 3263
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8808659069975169,
+      "learning_rate": 8.744040425765499e-07,
+      "loss": 0.5809,
+      "step": 3264
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8253277897817797,
+      "learning_rate": 8.708769829836517e-07,
+      "loss": 0.6341,
+      "step": 3265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9239184838278018,
+      "learning_rate": 8.673567273964189e-07,
+      "loss": 0.5677,
+      "step": 3266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7740461202295261,
+      "learning_rate": 8.638432784385265e-07,
+      "loss": 0.6251,
+      "step": 3267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8113390114344685,
+      "learning_rate": 8.603366387285705e-07,
+      "loss": 0.6114,
+      "step": 3268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8700710530656983,
+      "learning_rate": 8.56836810880085e-07,
+      "loss": 0.5891,
+      "step": 3269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7350303387894166,
+      "learning_rate": 8.53343797501509e-07,
+      "loss": 0.6582,
+      "step": 3270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8043587193348551,
+      "learning_rate": 8.498576011962189e-07,
+      "loss": 0.6668,
+      "step": 3271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.948617488880242,
+      "learning_rate": 8.463782245625018e-07,
+      "loss": 0.6404,
+      "step": 3272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.371500834867399,
+      "learning_rate": 8.429056701935634e-07,
+      "loss": 1.0867,
+      "step": 3273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8177135681635956,
+      "learning_rate": 8.394399406775244e-07,
+      "loss": 0.6167,
+      "step": 3274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7755308786469344,
+      "learning_rate": 8.35981038597422e-07,
+      "loss": 0.6618,
+      "step": 3275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2113803162949925,
+      "learning_rate": 8.325289665311998e-07,
+      "loss": 1.116,
+      "step": 3276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7860122207722593,
+      "learning_rate": 8.290837270517182e-07,
+      "loss": 0.5635,
+      "step": 3277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3321537912439996,
+      "learning_rate": 8.256453227267336e-07,
+      "loss": 1.0656,
+      "step": 3278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0020432529427628,
+      "learning_rate": 8.222137561189214e-07,
+      "loss": 0.6103,
+      "step": 3279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8000909404875478,
+      "learning_rate": 8.187890297858525e-07,
+      "loss": 0.5181,
+      "step": 3280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7109083361375566,
+      "learning_rate": 8.153711462800018e-07,
+      "loss": 0.5735,
+      "step": 3281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.72964396045948,
+      "learning_rate": 8.11960108148745e-07,
+      "loss": 0.5211,
+      "step": 3282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8493837224834853,
+      "learning_rate": 8.085559179343516e-07,
+      "loss": 0.6093,
+      "step": 3283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2909672221991264,
+      "learning_rate": 8.051585781739956e-07,
+      "loss": 1.1253,
+      "step": 3284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7286124385322845,
+      "learning_rate": 8.017680913997373e-07,
+      "loss": 0.6209,
+      "step": 3285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7808026905457137,
+      "learning_rate": 7.983844601385349e-07,
+      "loss": 0.5512,
+      "step": 3286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7276758200552245,
+      "learning_rate": 7.950076869122314e-07,
+      "loss": 0.5292,
+      "step": 3287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8675204897398874,
+      "learning_rate": 7.916377742375669e-07,
+      "loss": 0.6391,
+      "step": 3288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8590838708257842,
+      "learning_rate": 7.882747246261579e-07,
+      "loss": 0.6344,
+      "step": 3289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2544108042887854,
+      "learning_rate": 7.849185405845172e-07,
+      "loss": 1.0819,
+      "step": 3290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9269979952774714,
+      "learning_rate": 7.815692246140272e-07,
+      "loss": 0.6422,
+      "step": 3291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.88265260183878,
+      "learning_rate": 7.782267792109633e-07,
+      "loss": 0.5822,
+      "step": 3292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7085584832876033,
+      "learning_rate": 7.748912068664771e-07,
+      "loss": 0.5936,
+      "step": 3293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.340619634379051,
+      "learning_rate": 7.715625100665935e-07,
+      "loss": 1.0873,
+      "step": 3294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.678076879506429,
+      "learning_rate": 7.682406912922146e-07,
+      "loss": 0.6026,
+      "step": 3295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8436000738579224,
+      "learning_rate": 7.649257530191245e-07,
+      "loss": 0.6185,
+      "step": 3296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6419420717518722,
+      "learning_rate": 7.616176977179634e-07,
+      "loss": 0.5901,
+      "step": 3297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8207697553756945,
+      "learning_rate": 7.583165278542559e-07,
+      "loss": 0.559,
+      "step": 3298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.383743743036724,
+      "learning_rate": 7.550222458883883e-07,
+      "loss": 1.1283,
+      "step": 3299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4954362781556854,
+      "learning_rate": 7.517348542756131e-07,
+      "loss": 1.0315,
+      "step": 3300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7973010025558183,
+      "learning_rate": 7.484543554660495e-07,
+      "loss": 0.5466,
+      "step": 3301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.084690470909893,
+      "learning_rate": 7.451807519046772e-07,
+      "loss": 0.6333,
+      "step": 3302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7506224534284986,
+      "learning_rate": 7.419140460313412e-07,
+      "loss": 0.5628,
+      "step": 3303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.841010479839977,
+      "learning_rate": 7.386542402807395e-07,
+      "loss": 0.5058,
+      "step": 3304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3016102550740776,
+      "learning_rate": 7.35401337082432e-07,
+      "loss": 1.0214,
+      "step": 3305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7001083794193017,
+      "learning_rate": 7.321553388608326e-07,
+      "loss": 0.5938,
+      "step": 3306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.224423035226227,
+      "learning_rate": 7.289162480352074e-07,
+      "loss": 1.1492,
+      "step": 3307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.810258514673398,
+      "learning_rate": 7.256840670196752e-07,
+      "loss": 0.6556,
+      "step": 3308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2783469622217893,
+      "learning_rate": 7.224587982232101e-07,
+      "loss": 1.0887,
+      "step": 3309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6593871703044105,
+      "learning_rate": 7.192404440496236e-07,
+      "loss": 0.5129,
+      "step": 3310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8505584924985423,
+      "learning_rate": 7.16029006897584e-07,
+      "loss": 0.5674,
+      "step": 3311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8212680654268258,
+      "learning_rate": 7.128244891605985e-07,
+      "loss": 0.6108,
+      "step": 3312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7832877056922984,
+      "learning_rate": 7.096268932270189e-07,
+      "loss": 0.5908,
+      "step": 3313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.832347038939212,
+      "learning_rate": 7.064362214800391e-07,
+      "loss": 0.6181,
+      "step": 3314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.317070803872008,
+      "learning_rate": 7.0325247629769e-07,
+      "loss": 1.1391,
+      "step": 3315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2394319173912645,
+      "learning_rate": 7.000756600528391e-07,
+      "loss": 1.075,
+      "step": 3316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8390711285920271,
+      "learning_rate": 6.96905775113198e-07,
+      "loss": 0.6924,
+      "step": 3317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6714360886158786,
+      "learning_rate": 6.937428238413024e-07,
+      "loss": 0.5352,
+      "step": 3318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7720704989613387,
+      "learning_rate": 6.905868085945256e-07,
+      "loss": 0.5847,
+      "step": 3319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5847932267105427,
+      "learning_rate": 6.874377317250691e-07,
+      "loss": 0.5696,
+      "step": 3320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9393031037067547,
+      "learning_rate": 6.842955955799624e-07,
+      "loss": 0.5215,
+      "step": 3321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8301054771630827,
+      "learning_rate": 6.811604025010698e-07,
+      "loss": 0.5955,
+      "step": 3322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9430431870853755,
+      "learning_rate": 6.780321548250668e-07,
+      "loss": 0.5893,
+      "step": 3323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.454005960278684,
+      "learning_rate": 6.749108548834671e-07,
+      "loss": 1.0292,
+      "step": 3324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8335106125705682,
+      "learning_rate": 6.717965050025966e-07,
+      "loss": 0.6521,
+      "step": 3325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.784095203401173,
+      "learning_rate": 6.686891075036039e-07,
+      "loss": 0.7222,
+      "step": 3326
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8526817354674805,
+      "learning_rate": 6.655886647024556e-07,
+      "loss": 0.6218,
+      "step": 3327
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8738215124002948,
+      "learning_rate": 6.62495178909941e-07,
+      "loss": 0.5855,
+      "step": 3328
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.729685431328782,
+      "learning_rate": 6.594086524316512e-07,
+      "loss": 0.6177,
+      "step": 3329
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7116906946701964,
+      "learning_rate": 6.563290875680051e-07,
+      "loss": 0.5823,
+      "step": 3330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.958140669435682,
+      "learning_rate": 6.532564866142199e-07,
+      "loss": 0.5572,
+      "step": 3331
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.267815592831765,
+      "learning_rate": 6.501908518603339e-07,
+      "loss": 1.1395,
+      "step": 3332
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.375114794455744,
+      "learning_rate": 6.471321855911849e-07,
+      "loss": 1.1516,
+      "step": 3333
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3646280041109784,
+      "learning_rate": 6.440804900864217e-07,
+      "loss": 1.109,
+      "step": 3334
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7313733426376665,
+      "learning_rate": 6.410357676204959e-07,
+      "loss": 0.6106,
+      "step": 3335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7318176435724353,
+      "learning_rate": 6.379980204626635e-07,
+      "loss": 0.5395,
+      "step": 3336
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8252502869787957,
+      "learning_rate": 6.349672508769811e-07,
+      "loss": 0.5829,
+      "step": 3337
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8100354402479137,
+      "learning_rate": 6.31943461122303e-07,
+      "loss": 0.5423,
+      "step": 3338
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7614367376446503,
+      "learning_rate": 6.289266534522842e-07,
+      "loss": 0.6154,
+      "step": 3339
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7035515168398718,
+      "learning_rate": 6.259168301153728e-07,
+      "loss": 0.5876,
+      "step": 3340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.277661834215904,
+      "learning_rate": 6.229139933548178e-07,
+      "loss": 1.0597,
+      "step": 3341
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9473565095284235,
+      "learning_rate": 6.199181454086511e-07,
+      "loss": 0.5671,
+      "step": 3342
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7707931115629296,
+      "learning_rate": 6.169292885097056e-07,
+      "loss": 0.6693,
+      "step": 3343
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.355247322518236,
+      "learning_rate": 6.139474248855959e-07,
+      "loss": 1.096,
+      "step": 3344
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7739864190128127,
+      "learning_rate": 6.109725567587311e-07,
+      "loss": 0.656,
+      "step": 3345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8713637102092961,
+      "learning_rate": 6.080046863463007e-07,
+      "loss": 0.5847,
+      "step": 3346
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7859591002687312,
+      "learning_rate": 6.050438158602833e-07,
+      "loss": 0.5601,
+      "step": 3347
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7608278006854619,
+      "learning_rate": 6.02089947507436e-07,
+      "loss": 0.6041,
+      "step": 3348
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6808510341340335,
+      "learning_rate": 5.991430834893031e-07,
+      "loss": 0.6437,
+      "step": 3349
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7786887900465755,
+      "learning_rate": 5.962032260022011e-07,
+      "loss": 0.5833,
+      "step": 3350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8032804206072512,
+      "learning_rate": 5.932703772372317e-07,
+      "loss": 0.5882,
+      "step": 3351
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8577189138008445,
+      "learning_rate": 5.903445393802687e-07,
+      "loss": 0.5654,
+      "step": 3352
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6697163907705923,
+      "learning_rate": 5.874257146119611e-07,
+      "loss": 0.5422,
+      "step": 3353
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8672299937099628,
+      "learning_rate": 5.845139051077309e-07,
+      "loss": 0.6585,
+      "step": 3354
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.385841118752315,
+      "learning_rate": 5.816091130377722e-07,
+      "loss": 1.1141,
+      "step": 3355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8324274216774308,
+      "learning_rate": 5.787113405670508e-07,
+      "loss": 0.601,
+      "step": 3356
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4236454359530635,
+      "learning_rate": 5.758205898552982e-07,
+      "loss": 1.1068,
+      "step": 3357
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3738148156461865,
+      "learning_rate": 5.72936863057012e-07,
+      "loss": 1.1427,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8149698512598818,
+      "learning_rate": 5.700601623214552e-07,
+      "loss": 0.5888,
+      "step": 3359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6757725004473483,
+      "learning_rate": 5.671904897926594e-07,
+      "loss": 0.5257,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6889959888154218,
+      "learning_rate": 5.643278476094072e-07,
+      "loss": 0.5688,
+      "step": 3361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.363198698590651,
+      "learning_rate": 5.614722379052551e-07,
+      "loss": 1.0547,
+      "step": 3362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8130716317822793,
+      "learning_rate": 5.58623662808504e-07,
+      "loss": 0.5946,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2044693210539408,
+      "learning_rate": 5.557821244422235e-07,
+      "loss": 1.0805,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2836725771422963,
+      "learning_rate": 5.529476249242316e-07,
+      "loss": 1.0574,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5216288608396713,
+      "learning_rate": 5.501201663671041e-07,
+      "loss": 0.5647,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3377488795921035,
+      "learning_rate": 5.472997508781664e-07,
+      "loss": 1.0465,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9197301854045645,
+      "learning_rate": 5.44486380559498e-07,
+      "loss": 0.6225,
+      "step": 3368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.709547919139198,
+      "learning_rate": 5.416800575079229e-07,
+      "loss": 0.5764,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.948443477335575,
+      "learning_rate": 5.388807838150167e-07,
+      "loss": 0.6135,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.840699515744075,
+      "learning_rate": 5.360885615670986e-07,
+      "loss": 0.5564,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7623602392501825,
+      "learning_rate": 5.333033928452358e-07,
+      "loss": 0.587,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.675883303259724,
+      "learning_rate": 5.305252797252325e-07,
+      "loss": 0.6014,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6939706339359708,
+      "learning_rate": 5.27754224277639e-07,
+      "loss": 0.5241,
+      "step": 3374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8167884203044389,
+      "learning_rate": 5.249902285677466e-07,
+      "loss": 0.5408,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.416469868676564,
+      "learning_rate": 5.222332946555797e-07,
+      "loss": 1.0868,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8116268542125942,
+      "learning_rate": 5.194834245959046e-07,
+      "loss": 0.6256,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8396718272026078,
+      "learning_rate": 5.167406204382186e-07,
+      "loss": 0.6445,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9251048620667572,
+      "learning_rate": 5.140048842267564e-07,
+      "loss": 0.6545,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.865291804788193,
+      "learning_rate": 5.112762180004805e-07,
+      "loss": 0.6108,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8768053101762654,
+      "learning_rate": 5.085546237930916e-07,
+      "loss": 0.5265,
+      "step": 3381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8169262215182276,
+      "learning_rate": 5.058401036330096e-07,
+      "loss": 0.5872,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.867265066422815,
+      "learning_rate": 5.031326595433905e-07,
+      "loss": 0.5872,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2873854143294348,
+      "learning_rate": 5.004322935421135e-07,
+      "loss": 1.082,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2421124572321203,
+      "learning_rate": 4.977390076417809e-07,
+      "loss": 1.0874,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9734571277420538,
+      "learning_rate": 4.950528038497182e-07,
+      "loss": 0.5701,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0003660806477725,
+      "learning_rate": 4.923736841679772e-07,
+      "loss": 0.5962,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3183987767439445,
+      "learning_rate": 4.897016505933216e-07,
+      "loss": 1.115,
+      "step": 3388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7845506356580976,
+      "learning_rate": 4.870367051172431e-07,
+      "loss": 0.5913,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2788872553457082,
+      "learning_rate": 4.843788497259449e-07,
+      "loss": 1.0754,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.719316825423068,
+      "learning_rate": 4.81728086400346e-07,
+      "loss": 0.5178,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.756571601081864,
+      "learning_rate": 4.790844171160813e-07,
+      "loss": 0.6115,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8790651283413569,
+      "learning_rate": 4.764478438434972e-07,
+      "loss": 0.5744,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.58658621758503,
+      "learning_rate": 4.73818368547655e-07,
+      "loss": 1.1157,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8156093173296666,
+      "learning_rate": 4.71195993188317e-07,
+      "loss": 0.6071,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2286010846750783,
+      "learning_rate": 4.6858071971996656e-07,
+      "loss": 1.0565,
+      "step": 3396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8069538112032169,
+      "learning_rate": 4.6597255009178224e-07,
+      "loss": 0.5405,
+      "step": 3397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8930597776209952,
+      "learning_rate": 4.633714862476557e-07,
+      "loss": 0.5673,
+      "step": 3398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.877035486932075,
+      "learning_rate": 4.607775301261763e-07,
+      "loss": 0.5798,
+      "step": 3399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8884072662139821,
+      "learning_rate": 4.5819068366064535e-07,
+      "loss": 0.6435,
+      "step": 3400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.764929768643127,
+      "learning_rate": 4.556109487790539e-07,
+      "loss": 0.5845,
+      "step": 3401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7990155364546472,
+      "learning_rate": 4.530383274041028e-07,
+      "loss": 0.5242,
+      "step": 3402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2202887048244917,
+      "learning_rate": 4.504728214531806e-07,
+      "loss": 1.0907,
+      "step": 3403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.749351510759308,
+      "learning_rate": 4.4791443283838445e-07,
+      "loss": 0.6242,
+      "step": 3404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8621748836082812,
+      "learning_rate": 4.453631634664979e-07,
+      "loss": 0.5338,
+      "step": 3405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8221244856218366,
+      "learning_rate": 4.428190152390033e-07,
+      "loss": 0.5402,
+      "step": 3406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7313188664955037,
+      "learning_rate": 4.402819900520716e-07,
+      "loss": 0.5221,
+      "step": 3407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2539071927686747,
+      "learning_rate": 4.3775208979657034e-07,
+      "loss": 1.1584,
+      "step": 3408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2595675840074634,
+      "learning_rate": 4.352293163580512e-07,
+      "loss": 1.1019,
+      "step": 3409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5459092524748717,
+      "learning_rate": 4.3271367161675904e-07,
+      "loss": 0.5384,
+      "step": 3410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2968083342865135,
+      "learning_rate": 4.302051574476207e-07,
+      "loss": 1.104,
+      "step": 3411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0386166126387315,
+      "learning_rate": 4.277037757202507e-07,
+      "loss": 0.6023,
+      "step": 3412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.333899289919978,
+      "learning_rate": 4.2520952829895325e-07,
+      "loss": 1.0709,
+      "step": 3413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6607426962913743,
+      "learning_rate": 4.2272241704270247e-07,
+      "loss": 0.4709,
+      "step": 3414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6654080251507737,
+      "learning_rate": 4.202424438051678e-07,
+      "loss": 0.5213,
+      "step": 3415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6594899718425982,
+      "learning_rate": 4.1776961043468956e-07,
+      "loss": 0.5872,
+      "step": 3416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8307622363594183,
+      "learning_rate": 4.153039187742902e-07,
+      "loss": 0.5604,
+      "step": 3417
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2160773570797496,
+      "learning_rate": 4.128453706616686e-07,
+      "loss": 1.0171,
+      "step": 3418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.262206400705667,
+      "learning_rate": 4.103939679292002e-07,
+      "loss": 1.1055,
+      "step": 3419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.717504824121182,
+      "learning_rate": 4.079497124039311e-07,
+      "loss": 0.5826,
+      "step": 3420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.710870951548572,
+      "learning_rate": 4.055126059075898e-07,
+      "loss": 0.6039,
+      "step": 3421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.26288623297175,
+      "learning_rate": 4.0308265025656433e-07,
+      "loss": 1.0899,
+      "step": 3422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.293730759305058,
+      "learning_rate": 4.006598472619228e-07,
+      "loss": 1.1575,
+      "step": 3423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8274438285162757,
+      "learning_rate": 3.9824419872939837e-07,
+      "loss": 0.6412,
+      "step": 3424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7556630811885714,
+      "learning_rate": 3.9583570645939315e-07,
+      "loss": 0.6164,
+      "step": 3425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.698688600106655,
+      "learning_rate": 3.9343437224697244e-07,
+      "loss": 0.5584,
+      "step": 3426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.616809766628362,
+      "learning_rate": 3.910401978818712e-07,
+      "loss": 0.533,
+      "step": 3427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7505789900795958,
+      "learning_rate": 3.886531851484865e-07,
+      "loss": 0.5938,
+      "step": 3428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9142827864101077,
+      "learning_rate": 3.8627333582587635e-07,
+      "loss": 0.6338,
+      "step": 3429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8104830102549652,
+      "learning_rate": 3.839006516877608e-07,
+      "loss": 0.5523,
+      "step": 3430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7069947466317827,
+      "learning_rate": 3.8153513450251865e-07,
+      "loss": 0.6168,
+      "step": 3431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2431937107321356,
+      "learning_rate": 3.7917678603319075e-07,
+      "loss": 1.0904,
+      "step": 3432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.046455062078342,
+      "learning_rate": 3.768256080374688e-07,
+      "loss": 0.6222,
+      "step": 3433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8134915457287732,
+      "learning_rate": 3.744816022677078e-07,
+      "loss": 0.6588,
+      "step": 3434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.749186133909876,
+      "learning_rate": 3.721447704709091e-07,
+      "loss": 0.6521,
+      "step": 3435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9024465931686398,
+      "learning_rate": 3.698151143887352e-07,
+      "loss": 0.5709,
+      "step": 3436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7521744860962252,
+      "learning_rate": 3.6749263575749483e-07,
+      "loss": 0.5431,
+      "step": 3437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8815117136339834,
+      "learning_rate": 3.65177336308149e-07,
+      "loss": 0.6361,
+      "step": 3438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3387484598166637,
+      "learning_rate": 3.628692177663096e-07,
+      "loss": 1.008,
+      "step": 3439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.761274497761487,
+      "learning_rate": 3.605682818522338e-07,
+      "loss": 0.6204,
+      "step": 3440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.222893738164028,
+      "learning_rate": 3.582745302808255e-07,
+      "loss": 1.1513,
+      "step": 3441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3439742740855496,
+      "learning_rate": 3.5598796476163935e-07,
+      "loss": 1.1114,
+      "step": 3442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.729548303420874,
+      "learning_rate": 3.537085869988677e-07,
+      "loss": 0.5941,
+      "step": 3443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9496981250066308,
+      "learning_rate": 3.514363986913483e-07,
+      "loss": 0.6354,
+      "step": 3444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.192208881022722,
+      "learning_rate": 3.491714015325609e-07,
+      "loss": 1.0034,
+      "step": 3445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8087129246977465,
+      "learning_rate": 3.469135972106252e-07,
+      "loss": 0.5997,
+      "step": 3446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4333362788819612,
+      "learning_rate": 3.4466298740830186e-07,
+      "loss": 1.0951,
+      "step": 3447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7846074965047325,
+      "learning_rate": 3.4241957380298454e-07,
+      "loss": 0.5735,
+      "step": 3448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6715303606084775,
+      "learning_rate": 3.401833580667091e-07,
+      "loss": 0.5014,
+      "step": 3449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8674548273056688,
+      "learning_rate": 3.379543418661435e-07,
+      "loss": 0.6059,
+      "step": 3450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7161579374231173,
+      "learning_rate": 3.3573252686258993e-07,
+      "loss": 0.5883,
+      "step": 3451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.658707209806315,
+      "learning_rate": 3.335179147119849e-07,
+      "loss": 0.5727,
+      "step": 3452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.348645799787643,
+      "learning_rate": 3.313105070648981e-07,
+      "loss": 1.1169,
+      "step": 3453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7502668439894806,
+      "learning_rate": 3.2911030556652344e-07,
+      "loss": 0.6383,
+      "step": 3454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7309010336258321,
+      "learning_rate": 3.269173118566915e-07,
+      "loss": 0.6151,
+      "step": 3455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7705383632944698,
+      "learning_rate": 3.247315275698548e-07,
+      "loss": 0.5702,
+      "step": 3456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.018976539779313,
+      "learning_rate": 3.225529543350958e-07,
+      "loss": 0.5767,
+      "step": 3457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6605641044943722,
+      "learning_rate": 3.2038159377612345e-07,
+      "loss": 0.6354,
+      "step": 3458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7336112221343103,
+      "learning_rate": 3.182174475112676e-07,
+      "loss": 0.4975,
+      "step": 3459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9154340889539836,
+      "learning_rate": 3.1606051715348365e-07,
+      "loss": 0.5631,
+      "step": 3460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8556410215771924,
+      "learning_rate": 3.1391080431034894e-07,
+      "loss": 0.6191,
+      "step": 3461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1800829246078934,
+      "learning_rate": 3.117683105840619e-07,
+      "loss": 0.5613,
+      "step": 3462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2774797512311475,
+      "learning_rate": 3.096330375714374e-07,
+      "loss": 1.0382,
+      "step": 3463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8832414535205066,
+      "learning_rate": 3.075049868639124e-07,
+      "loss": 0.603,
+      "step": 3464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7621113607277141,
+      "learning_rate": 3.0538416004753713e-07,
+      "loss": 0.6344,
+      "step": 3465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8901338231952112,
+      "learning_rate": 3.0327055870298494e-07,
+      "loss": 0.5563,
+      "step": 3466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.779979598497243,
+      "learning_rate": 3.011641844055335e-07,
+      "loss": 0.6335,
+      "step": 3467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.304191240489385,
+      "learning_rate": 2.9906503872508375e-07,
+      "loss": 1.1062,
+      "step": 3468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7073476817561533,
+      "learning_rate": 2.96973123226143e-07,
+      "loss": 0.5146,
+      "step": 3469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7659762552482934,
+      "learning_rate": 2.948884394678342e-07,
+      "loss": 0.5739,
+      "step": 3470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0147430820090735,
+      "learning_rate": 2.928109890038866e-07,
+      "loss": 0.589,
+      "step": 3471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7233775373996565,
+      "learning_rate": 2.9074077338263953e-07,
+      "loss": 0.5125,
+      "step": 3472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7326438030899298,
+      "learning_rate": 2.886777941470409e-07,
+      "loss": 0.5887,
+      "step": 3473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8937933980871389,
+      "learning_rate": 2.866220528346475e-07,
+      "loss": 0.6034,
+      "step": 3474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.302761293556923,
+      "learning_rate": 2.8457355097761485e-07,
+      "loss": 1.0822,
+      "step": 3475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9814200248193292,
+      "learning_rate": 2.8253229010271055e-07,
+      "loss": 0.6186,
+      "step": 3476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8538760874544167,
+      "learning_rate": 2.8049827173130096e-07,
+      "loss": 0.6085,
+      "step": 3477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2649115070459866,
+      "learning_rate": 2.784714973793556e-07,
+      "loss": 1.0911,
+      "step": 3478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.820349004788762,
+      "learning_rate": 2.7645196855744515e-07,
+      "loss": 0.5745,
+      "step": 3479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8015489179758621,
+      "learning_rate": 2.744396867707388e-07,
+      "loss": 0.5259,
+      "step": 3480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.839622375112756,
+      "learning_rate": 2.724346535190081e-07,
+      "loss": 0.6156,
+      "step": 3481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2944865762047133,
+      "learning_rate": 2.704368702966176e-07,
+      "loss": 1.0971,
+      "step": 3482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.711559313037306,
+      "learning_rate": 2.684463385925318e-07,
+      "loss": 0.5525,
+      "step": 3483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.403114663632663,
+      "learning_rate": 2.664630598903073e-07,
+      "loss": 1.0899,
+      "step": 3484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5971932702033325,
+      "learning_rate": 2.644870356681006e-07,
+      "loss": 0.4482,
+      "step": 3485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.711013907156889,
+      "learning_rate": 2.6251826739865485e-07,
+      "loss": 0.5874,
+      "step": 3486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.863032960770031,
+      "learning_rate": 2.605567565493095e-07,
+      "loss": 0.6402,
+      "step": 3487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3698714213437073,
+      "learning_rate": 2.5860250458199176e-07,
+      "loss": 1.0798,
+      "step": 3488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8247789636794138,
+      "learning_rate": 2.5665551295322335e-07,
+      "loss": 0.6495,
+      "step": 3489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7168192024473423,
+      "learning_rate": 2.5471578311411005e-07,
+      "loss": 0.5503,
+      "step": 3490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.514706196772943,
+      "learning_rate": 2.527833165103488e-07,
+      "loss": 0.4735,
+      "step": 3491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8632242747476144,
+      "learning_rate": 2.5085811458221974e-07,
+      "loss": 0.5364,
+      "step": 3492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8760884975090946,
+      "learning_rate": 2.489401787645951e-07,
+      "loss": 0.5362,
+      "step": 3493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.416699616928953,
+      "learning_rate": 2.4702951048692137e-07,
+      "loss": 1.0789,
+      "step": 3494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8160035756579116,
+      "learning_rate": 2.4512611117323947e-07,
+      "loss": 0.5842,
+      "step": 3495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8188870714034564,
+      "learning_rate": 2.4322998224216463e-07,
+      "loss": 0.5442,
+      "step": 3496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9096694323492849,
+      "learning_rate": 2.4134112510689754e-07,
+      "loss": 0.6228,
+      "step": 3497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.387996752613773,
+      "learning_rate": 2.394595411752165e-07,
+      "loss": 1.1238,
+      "step": 3498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.795165227400552,
+      "learning_rate": 2.3758523184948202e-07,
+      "loss": 0.5549,
+      "step": 3499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6195261546400714,
+      "learning_rate": 2.3571819852663103e-07,
+      "loss": 1.1111,
+      "step": 3500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.860236358883597,
+      "learning_rate": 2.3385844259817715e-07,
+      "loss": 0.5936,
+      "step": 3501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7191841180385146,
+      "learning_rate": 2.320059654502116e-07,
+      "loss": 0.5595,
+      "step": 3502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7833690748326816,
+      "learning_rate": 2.3016076846339997e-07,
+      "loss": 0.4977,
+      "step": 3503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.801504402389232,
+      "learning_rate": 2.2832285301297995e-07,
+      "loss": 0.5702,
+      "step": 3504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8689276527082341,
+      "learning_rate": 2.2649222046876584e-07,
+      "loss": 0.5596,
+      "step": 3505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4540160563868647,
+      "learning_rate": 2.246688721951429e-07,
+      "loss": 1.047,
+      "step": 3506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.76230846816657,
+      "learning_rate": 2.2285280955106403e-07,
+      "loss": 0.5421,
+      "step": 3507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.903756735649877,
+      "learning_rate": 2.2104403389005657e-07,
+      "loss": 0.6411,
+      "step": 3508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7499747201830957,
+      "learning_rate": 2.192425465602155e-07,
+      "loss": 0.5683,
+      "step": 3509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8138825737608417,
+      "learning_rate": 2.1744834890420118e-07,
+      "loss": 0.6141,
+      "step": 3510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.283062752857871,
+      "learning_rate": 2.1566144225924513e-07,
+      "loss": 1.1064,
+      "step": 3511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.282112380282223,
+      "learning_rate": 2.1388182795714084e-07,
+      "loss": 1.0938,
+      "step": 3512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6837576897008901,
+      "learning_rate": 2.121095073242474e-07,
+      "loss": 0.5108,
+      "step": 3513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9483354509287703,
+      "learning_rate": 2.103444816814937e-07,
+      "loss": 0.6355,
+      "step": 3514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7133596183483144,
+      "learning_rate": 2.0858675234436187e-07,
+      "loss": 0.6272,
+      "step": 3515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.500440267988904,
+      "learning_rate": 2.0683632062290404e-07,
+      "loss": 1.1064,
+      "step": 3516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2452964584647117,
+      "learning_rate": 2.0509318782172994e-07,
+      "loss": 1.0816,
+      "step": 3517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3519545645458875,
+      "learning_rate": 2.0335735524000922e-07,
+      "loss": 1.11,
+      "step": 3518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.424869496702636,
+      "learning_rate": 2.0162882417147256e-07,
+      "loss": 1.0514,
+      "step": 3519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7305712058485327,
+      "learning_rate": 1.99907595904405e-07,
+      "loss": 0.5599,
+      "step": 3520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8116467548717037,
+      "learning_rate": 1.9819367172165482e-07,
+      "loss": 0.6026,
+      "step": 3521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.728782874704094,
+      "learning_rate": 1.964870529006191e-07,
+      "loss": 0.5527,
+      "step": 3522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4141514793615646,
+      "learning_rate": 1.94787740713257e-07,
+      "loss": 1.1451,
+      "step": 3523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.516772191061238,
+      "learning_rate": 1.930957364260766e-07,
+      "loss": 1.0979,
+      "step": 3524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8887611966477984,
+      "learning_rate": 1.9141104130014465e-07,
+      "loss": 0.5427,
+      "step": 3525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3532242087442383,
+      "learning_rate": 1.8973365659107345e-07,
+      "loss": 1.0903,
+      "step": 3526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7613434910102304,
+      "learning_rate": 1.880635835490352e-07,
+      "loss": 0.6133,
+      "step": 3527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7088592497401829,
+      "learning_rate": 1.864008234187442e-07,
+      "loss": 0.5793,
+      "step": 3528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8773709221492245,
+      "learning_rate": 1.8474537743947142e-07,
+      "loss": 0.6571,
+      "step": 3529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8332870427577226,
+      "learning_rate": 1.8309724684503094e-07,
+      "loss": 0.5974,
+      "step": 3530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.838210715526418,
+      "learning_rate": 1.8145643286378912e-07,
+      "loss": 0.6685,
+      "step": 3531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6553793738898417,
+      "learning_rate": 1.7982293671865548e-07,
+      "loss": 0.5374,
+      "step": 3532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.063518282669557,
+      "learning_rate": 1.7819675962708948e-07,
+      "loss": 0.6042,
+      "step": 3533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6070010488318742,
+      "learning_rate": 1.7657790280109165e-07,
+      "loss": 0.5285,
+      "step": 3534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.785222395513093,
+      "learning_rate": 1.7496636744720908e-07,
+      "loss": 0.6293,
+      "step": 3535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2962213699290315,
+      "learning_rate": 1.7336215476653206e-07,
+      "loss": 1.1572,
+      "step": 3536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4026221452582828,
+      "learning_rate": 1.71765265954692e-07,
+      "loss": 1.1222,
+      "step": 3537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.737583992182857,
+      "learning_rate": 1.701757022018624e-07,
+      "loss": 0.5491,
+      "step": 3538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8007117423017156,
+      "learning_rate": 1.685934646927556e-07,
+      "loss": 0.6057,
+      "step": 3539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7813602114752645,
+      "learning_rate": 1.6701855460662942e-07,
+      "loss": 0.5934,
+      "step": 3540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8377741052058447,
+      "learning_rate": 1.6545097311727375e-07,
+      "loss": 0.5951,
+      "step": 3541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.889937641469589,
+      "learning_rate": 1.6389072139301965e-07,
+      "loss": 0.56,
+      "step": 3542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7342622992673589,
+      "learning_rate": 1.6233780059673466e-07,
+      "loss": 0.5305,
+      "step": 3543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8853915490017445,
+      "learning_rate": 1.6079221188582295e-07,
+      "loss": 0.6242,
+      "step": 3544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8016815055395974,
+      "learning_rate": 1.5925395641222307e-07,
+      "loss": 0.5913,
+      "step": 3545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9575479454600313,
+      "learning_rate": 1.5772303532241018e-07,
+      "loss": 0.5852,
+      "step": 3546
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.831600139575866,
+      "learning_rate": 1.5619944975738822e-07,
+      "loss": 0.6114,
+      "step": 3547
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8414145160652091,
+      "learning_rate": 1.5468320085270105e-07,
+      "loss": 0.6066,
+      "step": 3548
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4245679274315894,
+      "learning_rate": 1.531742897384192e-07,
+      "loss": 1.1152,
+      "step": 3549
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8441737085861218,
+      "learning_rate": 1.5167271753914525e-07,
+      "loss": 0.606,
+      "step": 3550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7937113980057595,
+      "learning_rate": 1.5017848537401293e-07,
+      "loss": 0.6152,
+      "step": 3551
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.912424048493025,
+      "learning_rate": 1.4869159435668468e-07,
+      "loss": 0.5869,
+      "step": 3552
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.390710844508197,
+      "learning_rate": 1.4721204559535075e-07,
+      "loss": 1.1513,
+      "step": 3553
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8241781846605258,
+      "learning_rate": 1.4573984019273346e-07,
+      "loss": 0.5713,
+      "step": 3554
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3319058068467284,
+      "learning_rate": 1.4427497924607624e-07,
+      "loss": 1.113,
+      "step": 3555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8109029426435046,
+      "learning_rate": 1.4281746384715133e-07,
+      "loss": 0.5673,
+      "step": 3556
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.394841396161931,
+      "learning_rate": 1.4136729508225755e-07,
+      "loss": 1.1385,
+      "step": 3557
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.214323627841144,
+      "learning_rate": 1.3992447403221476e-07,
+      "loss": 1.0528,
+      "step": 3558
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3246189269019766,
+      "learning_rate": 1.3848900177237167e-07,
+      "loss": 1.0824,
+      "step": 3559
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.839780905666832,
+      "learning_rate": 1.370608793725936e-07,
+      "loss": 0.6326,
+      "step": 3560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8524688278517876,
+      "learning_rate": 1.3564010789727355e-07,
+      "loss": 0.5838,
+      "step": 3561
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7419845086680572,
+      "learning_rate": 1.3422668840532227e-07,
+      "loss": 0.5171,
+      "step": 3562
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.409253826109116,
+      "learning_rate": 1.3282062195017154e-07,
+      "loss": 1.0927,
+      "step": 3563
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6224686178627348,
+      "learning_rate": 1.3142190957977418e-07,
+      "loss": 0.5505,
+      "step": 3564
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.806188299354381,
+      "learning_rate": 1.30030552336603e-07,
+      "loss": 0.593,
+      "step": 3565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6156309504626998,
+      "learning_rate": 1.2864655125764404e-07,
+      "loss": 0.5322,
+      "step": 3566
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4405061270540784,
+      "learning_rate": 1.272699073744077e-07,
+      "loss": 1.1192,
+      "step": 3567
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.387809694864741,
+      "learning_rate": 1.2590062171291329e-07,
+      "loss": 1.0887,
+      "step": 3568
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6748309518323958,
+      "learning_rate": 1.2453869529370222e-07,
+      "loss": 0.568,
+      "step": 3569
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7115834721611307,
+      "learning_rate": 1.231841291318292e-07,
+      "loss": 0.571,
+      "step": 3570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.426207098147732,
+      "learning_rate": 1.2183692423686111e-07,
+      "loss": 1.0327,
+      "step": 3571
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7800748611171358,
+      "learning_rate": 1.2049708161288144e-07,
+      "loss": 0.5573,
+      "step": 3572
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7704828253224607,
+      "learning_rate": 1.191646022584847e-07,
+      "loss": 0.6339,
+      "step": 3573
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8914829416829535,
+      "learning_rate": 1.1783948716677762e-07,
+      "loss": 0.6318,
+      "step": 3574
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8465772035441532,
+      "learning_rate": 1.1652173732538019e-07,
+      "loss": 0.6134,
+      "step": 3575
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.311777799440466,
+      "learning_rate": 1.1521135371642012e-07,
+      "loss": 1.1174,
+      "step": 3576
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7732872349278777,
+      "learning_rate": 1.1390833731653617e-07,
+      "loss": 0.5795,
+      "step": 3577
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2897770485017133,
+      "learning_rate": 1.1261268909687816e-07,
+      "loss": 1.0779,
+      "step": 3578
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7597689076584333,
+      "learning_rate": 1.1132441002310035e-07,
+      "loss": 0.6104,
+      "step": 3579
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7974077029107196,
+      "learning_rate": 1.100435010553691e-07,
+      "loss": 0.5474,
+      "step": 3580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.072828536579476,
+      "learning_rate": 1.0876996314835409e-07,
+      "loss": 0.5861,
+      "step": 3581
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.76509177457468,
+      "learning_rate": 1.0750379725123495e-07,
+      "loss": 0.5983,
+      "step": 3582
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.913447847910714,
+      "learning_rate": 1.0624500430769346e-07,
+      "loss": 0.615,
+      "step": 3583
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6372946814815552,
+      "learning_rate": 1.0499358525591696e-07,
+      "loss": 0.5879,
+      "step": 3584
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6930455676105478,
+      "learning_rate": 1.0374954102859935e-07,
+      "loss": 0.5822,
+      "step": 3585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.069099854500893,
+      "learning_rate": 1.0251287255293674e-07,
+      "loss": 0.6335,
+      "step": 3586
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.737742464423501,
+      "learning_rate": 1.0128358075062406e-07,
+      "loss": 0.5232,
+      "step": 3587
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2170796959118717,
+      "learning_rate": 1.0006166653786619e-07,
+      "loss": 1.0192,
+      "step": 3588
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8328174091642553,
+      "learning_rate": 9.884713082536357e-08,
+      "loss": 0.6474,
+      "step": 3589
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.667534483195104,
+      "learning_rate": 9.763997451831875e-08,
+      "loss": 0.5582,
+      "step": 3590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8154100044236576,
+      "learning_rate": 9.644019851643538e-08,
+      "loss": 0.5147,
+      "step": 3591
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7793025041608561,
+      "learning_rate": 9.524780371391373e-08,
+      "loss": 0.5303,
+      "step": 3592
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3344070862824,
+      "learning_rate": 9.406279099945737e-08,
+      "loss": 1.1025,
+      "step": 3593
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9814736572850975,
+      "learning_rate": 9.288516125626424e-08,
+      "loss": 0.5681,
+      "step": 3594
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.191428715196389,
+      "learning_rate": 9.171491536203114e-08,
+      "loss": 1.0803,
+      "step": 3595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.880830608644516,
+      "learning_rate": 9.055205418894931e-08,
+      "loss": 0.6279,
+      "step": 3596
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6912606002943122,
+      "learning_rate": 8.939657860371209e-08,
+      "loss": 0.4815,
+      "step": 3597
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6767060967822742,
+      "learning_rate": 8.824848946750175e-08,
+      "loss": 0.5909,
+      "step": 3598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8127215144613906,
+      "learning_rate": 8.710778763599826e-08,
+      "loss": 0.6936,
+      "step": 3599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2030601599149087,
+      "learning_rate": 8.597447395937597e-08,
+      "loss": 1.108,
+      "step": 3600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7151163350121579,
+      "learning_rate": 8.484854928230146e-08,
+      "loss": 0.4979,
+      "step": 3601
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7503155088812268,
+      "learning_rate": 8.373001444393569e-08,
+      "loss": 0.5222,
+      "step": 3602
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3090840010969083,
+      "learning_rate": 8.261887027793181e-08,
+      "loss": 1.0585,
+      "step": 3603
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6757724961823004,
+      "learning_rate": 8.151511761243181e-08,
+      "loss": 0.5971,
+      "step": 3604
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8125273827825135,
+      "learning_rate": 8.041875727007432e-08,
+      "loss": 0.6497,
+      "step": 3605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.388135485441079,
+      "learning_rate": 7.932979006798236e-08,
+      "loss": 1.106,
+      "step": 3606
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9422229676048892,
+      "learning_rate": 7.824821681777339e-08,
+      "loss": 0.5971,
+      "step": 3607
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7492191913310922,
+      "learning_rate": 7.717403832555148e-08,
+      "loss": 0.5479,
+      "step": 3608
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7159852958863315,
+      "learning_rate": 7.610725539191066e-08,
+      "loss": 0.4582,
+      "step": 3609
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6911143473656796,
+      "learning_rate": 7.50478688119316e-08,
+      "loss": 0.6071,
+      "step": 3610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3250942590800303,
+      "learning_rate": 7.399587937518381e-08,
+      "loss": 1.0692,
+      "step": 3611
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6508215054184938,
+      "learning_rate": 7.295128786572347e-08,
+      "loss": 0.5263,
+      "step": 3612
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8889559416996213,
+      "learning_rate": 7.191409506209113e-08,
+      "loss": 0.5673,
+      "step": 3613
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2763280494760845,
+      "learning_rate": 7.088430173731731e-08,
+      "loss": 1.1278,
+      "step": 3614
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7730794691146725,
+      "learning_rate": 6.986190865891252e-08,
+      "loss": 0.5721,
+      "step": 3615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3372206279607592,
+      "learning_rate": 6.8846916588875e-08,
+      "loss": 1.0903,
+      "step": 3616
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2373925073431367,
+      "learning_rate": 6.783932628368517e-08,
+      "loss": 1.0822,
+      "step": 3617
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7207641302336536,
+      "learning_rate": 6.683913849431012e-08,
+      "loss": 0.5053,
+      "step": 3618
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7067805748011784,
+      "learning_rate": 6.584635396619577e-08,
+      "loss": 0.5731,
+      "step": 3619
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8320315879918405,
+      "learning_rate": 6.486097343927245e-08,
+      "loss": 0.6644,
+      "step": 3620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7334611030287475,
+      "learning_rate": 6.388299764795158e-08,
+      "loss": 0.673,
+      "step": 3621
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8998317724980465,
+      "learning_rate": 6.291242732112679e-08,
+      "loss": 0.5884,
+      "step": 3622
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.670644318871945,
+      "learning_rate": 6.194926318217165e-08,
+      "loss": 0.5849,
+      "step": 3623
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9141924037887332,
+      "learning_rate": 6.099350594893861e-08,
+      "loss": 0.5934,
+      "step": 3624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.355444993483092,
+      "learning_rate": 6.004515633376117e-08,
+      "loss": 1.1148,
+      "step": 3625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7161723811962217,
+      "learning_rate": 5.910421504345287e-08,
+      "loss": 0.5445,
+      "step": 3626
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.840367184224101,
+      "learning_rate": 5.8170682779303825e-08,
+      "loss": 0.5769,
+      "step": 3627
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7790739707976704,
+      "learning_rate": 5.7244560237081914e-08,
+      "loss": 0.5522,
+      "step": 3628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8401272624913956,
+      "learning_rate": 5.632584810703501e-08,
+      "loss": 0.6041,
+      "step": 3629
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.682063836709373,
+      "learning_rate": 5.541454707388427e-08,
+      "loss": 0.5486,
+      "step": 3630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7277953583240995,
+      "learning_rate": 5.451065781683196e-08,
+      "loss": 0.5107,
+      "step": 3631
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8404127051869161,
+      "learning_rate": 5.36141810095514e-08,
+      "loss": 0.5616,
+      "step": 3632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8748138847300784,
+      "learning_rate": 5.272511732019481e-08,
+      "loss": 0.5974,
+      "step": 3633
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.911606225645968,
+      "learning_rate": 5.184346741138768e-08,
+      "loss": 0.5989,
+      "step": 3634
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7323093914328742,
+      "learning_rate": 5.0969231940231065e-08,
+      "loss": 0.6422,
+      "step": 3635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.889679066007662,
+      "learning_rate": 5.010241155829931e-08,
+      "loss": 0.582,
+      "step": 3636
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3645668350697067,
+      "learning_rate": 4.9243006911640076e-08,
+      "loss": 1.0872,
+      "step": 3637
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4741682200086843,
+      "learning_rate": 4.8391018640774335e-08,
+      "loss": 1.0879,
+      "step": 3638
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6716651152513982,
+      "learning_rate": 4.754644738069525e-08,
+      "loss": 0.5401,
+      "step": 3639
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8568343023938525,
+      "learning_rate": 4.670929376086819e-08,
+      "loss": 0.5562,
+      "step": 3640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1869363660822283,
+      "learning_rate": 4.587955840523073e-08,
+      "loss": 1.0715,
+      "step": 3641
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7116126209177442,
+      "learning_rate": 4.505724193219041e-08,
+      "loss": 0.501,
+      "step": 3642
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.491018221036229,
+      "learning_rate": 4.424234495462698e-08,
+      "loss": 1.1512,
+      "step": 3643
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8189556089282464,
+      "learning_rate": 4.343486807988906e-08,
+      "loss": 0.5782,
+      "step": 3644
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4084083788315573,
+      "learning_rate": 4.2634811909794125e-08,
+      "loss": 1.1069,
+      "step": 3645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9270845035935034,
+      "learning_rate": 4.184217704063187e-08,
+      "loss": 0.6002,
+      "step": 3646
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.388227058501542,
+      "learning_rate": 4.1056964063158625e-08,
+      "loss": 1.1139,
+      "step": 3647
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6716132090553195,
+      "learning_rate": 4.027917356259958e-08,
+      "loss": 0.5583,
+      "step": 3648
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.81433451269193,
+      "learning_rate": 3.9508806118647714e-08,
+      "loss": 0.6441,
+      "step": 3649
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7556620202008963,
+      "learning_rate": 3.874586230546484e-08,
+      "loss": 0.562,
+      "step": 3650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8351624174718617,
+      "learning_rate": 3.7990342691677226e-08,
+      "loss": 0.6408,
+      "step": 3651
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.839411845273074,
+      "learning_rate": 3.7242247840381104e-08,
+      "loss": 0.615,
+      "step": 3652
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7266433100915792,
+      "learning_rate": 3.650157830913603e-08,
+      "loss": 0.49,
+      "step": 3653
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4942405801869323,
+      "learning_rate": 3.5768334649969316e-08,
+      "loss": 1.062,
+      "step": 3654
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8610013165590016,
+      "learning_rate": 3.504251740937381e-08,
+      "loss": 0.5919,
+      "step": 3655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9709856924004758,
+      "learning_rate": 3.4324127128304575e-08,
+      "loss": 0.6323,
+      "step": 3656
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2296936477232054,
+      "learning_rate": 3.361316434218553e-08,
+      "loss": 1.1721,
+      "step": 3657
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8934440556345535,
+      "learning_rate": 3.290962958090171e-08,
+      "loss": 0.6635,
+      "step": 3658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6525049922905322,
+      "learning_rate": 3.221352336880257e-08,
+      "loss": 0.5231,
+      "step": 3659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8400236696770342,
+      "learning_rate": 3.152484622470309e-08,
+      "loss": 0.5869,
+      "step": 3660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6542600194319685,
+      "learning_rate": 3.084359866187825e-08,
+      "loss": 0.5586,
+      "step": 3661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.876271541447307,
+      "learning_rate": 3.016978118806746e-08,
+      "loss": 0.5676,
+      "step": 3662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7333896266758784,
+      "learning_rate": 2.9503394305472334e-08,
+      "loss": 0.5702,
+      "step": 3663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2463341124216964,
+      "learning_rate": 2.884443851075447e-08,
+      "loss": 1.0339,
+      "step": 3664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.463001564561556,
+      "learning_rate": 2.8192914295042118e-08,
+      "loss": 1.075,
+      "step": 3665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6632748969495939,
+      "learning_rate": 2.7548822143917962e-08,
+      "loss": 0.6151,
+      "step": 3666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.954322071187969,
+      "learning_rate": 2.6912162537431342e-08,
+      "loss": 0.6241,
+      "step": 3667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7761826673424792,
+      "learning_rate": 2.628293595008713e-08,
+      "loss": 0.5656,
+      "step": 3668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.55442676296194,
+      "learning_rate": 2.5661142850854637e-08,
+      "loss": 1.0452,
+      "step": 3669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2479263680949897,
+      "learning_rate": 2.5046783703159828e-08,
+      "loss": 1.0689,
+      "step": 3670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.312602232467346,
+      "learning_rate": 2.4439858964890872e-08,
+      "loss": 1.0674,
+      "step": 3671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.630631945214499,
+      "learning_rate": 2.384036908839149e-08,
+      "loss": 0.5845,
+      "step": 3672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8543997196891204,
+      "learning_rate": 2.3248314520468717e-08,
+      "loss": 0.7084,
+      "step": 3673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6724977483830956,
+      "learning_rate": 2.2663695702381805e-08,
+      "loss": 0.5678,
+      "step": 3674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7956577125491486,
+      "learning_rate": 2.208651306985443e-08,
+      "loss": 0.6186,
+      "step": 3675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7529302324970237,
+      "learning_rate": 2.1516767053064713e-08,
+      "loss": 0.5642,
+      "step": 3676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6052283427283696,
+      "learning_rate": 2.095445807664853e-08,
+      "loss": 0.5338,
+      "step": 3677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6835256101046834,
+      "learning_rate": 2.0399586559698425e-08,
+      "loss": 0.5787,
+      "step": 3678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2025243552172578,
+      "learning_rate": 1.9852152915764698e-08,
+      "loss": 1.116,
+      "step": 3679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2615150479405424,
+      "learning_rate": 1.9312157552855425e-08,
+      "loss": 1.0834,
+      "step": 3680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7340802993894628,
+      "learning_rate": 1.8779600873430893e-08,
+      "loss": 0.5491,
+      "step": 3681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8390043193790213,
+      "learning_rate": 1.8254483274411372e-08,
+      "loss": 0.5967,
+      "step": 3682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.612622436971519,
+      "learning_rate": 1.7736805147170465e-08,
+      "loss": 0.5566,
+      "step": 3683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6537984464750755,
+      "learning_rate": 1.7226566877538432e-08,
+      "loss": 0.5409,
+      "step": 3684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0315158385721954,
+      "learning_rate": 1.672376884579774e-08,
+      "loss": 0.6512,
+      "step": 3685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2361226926921343,
+      "learning_rate": 1.6228411426689737e-08,
+      "loss": 1.0892,
+      "step": 3686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9869381934250787,
+      "learning_rate": 1.5740494989406883e-08,
+      "loss": 0.5377,
+      "step": 3687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8831991803162942,
+      "learning_rate": 1.5260019897598288e-08,
+      "loss": 0.6032,
+      "step": 3688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9008287150396115,
+      "learning_rate": 1.4786986509364165e-08,
+      "loss": 0.6012,
+      "step": 3689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7118184583845768,
+      "learning_rate": 1.4321395177261387e-08,
+      "loss": 0.6057,
+      "step": 3690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6429634864747291,
+      "learning_rate": 1.3863246248297935e-08,
+      "loss": 0.5227,
+      "step": 3691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.862307473263137,
+      "learning_rate": 1.3412540063936219e-08,
+      "loss": 0.6265,
+      "step": 3692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6266976546867304,
+      "learning_rate": 1.296927696009087e-08,
+      "loss": 0.5499,
+      "step": 3693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.422153083642856,
+      "learning_rate": 1.2533457267128735e-08,
+      "loss": 1.0275,
+      "step": 3694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8240738801755116,
+      "learning_rate": 1.2105081309869981e-08,
+      "loss": 0.5904,
+      "step": 3695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9845666309011114,
+      "learning_rate": 1.1684149407585887e-08,
+      "loss": 0.5694,
+      "step": 3696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3346460914023557,
+      "learning_rate": 1.1270661874001053e-08,
+      "loss": 1.0683,
+      "step": 3697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6877386641806877,
+      "learning_rate": 1.08646190172923e-08,
+      "loss": 0.5658,
+      "step": 3698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.719098709252683,
+      "learning_rate": 1.0466021140083104e-08,
+      "loss": 0.5287,
+      "step": 3699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6971890025734067,
+      "learning_rate": 1.0074868539454719e-08,
+      "loss": 0.5592,
+      "step": 3700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3101457789341437,
+      "learning_rate": 9.691161506933943e-09,
+      "loss": 1.1157,
+      "step": 3701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7666218790322252,
+      "learning_rate": 9.314900328502019e-09,
+      "loss": 0.5665,
+      "step": 3702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8212057414120424,
+      "learning_rate": 8.946085284587958e-09,
+      "loss": 0.5951,
+      "step": 3703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7537666512187748,
+      "learning_rate": 8.584716650074099e-09,
+      "loss": 0.5315,
+      "step": 3704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9201067266688419,
+      "learning_rate": 8.230794694290556e-09,
+      "loss": 0.6724,
+      "step": 3705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8549943990652629,
+      "learning_rate": 7.884319681017438e-09,
+      "loss": 0.6009,
+      "step": 3706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8436871759027404,
+      "learning_rate": 7.54529186848596e-09,
+      "loss": 0.551,
+      "step": 3707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9775642699700016,
+      "learning_rate": 7.21371150937622e-09,
+      "loss": 0.6363,
+      "step": 3708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.427559395051086,
+      "learning_rate": 6.889578850817202e-09,
+      "loss": 1.1139,
+      "step": 3709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.735297838913253,
+      "learning_rate": 6.572894134386776e-09,
+      "loss": 0.5643,
+      "step": 3710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.147471983431588,
+      "learning_rate": 6.2636575961139145e-09,
+      "loss": 1.0857,
+      "step": 3711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5078171616195704,
+      "learning_rate": 5.961869466474257e-09,
+      "loss": 1.1284,
+      "step": 3712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7837543770407494,
+      "learning_rate": 5.667529970393437e-09,
+      "loss": 0.5898,
+      "step": 3713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.843553764657596,
+      "learning_rate": 5.380639327244863e-09,
+      "loss": 0.6325,
+      "step": 3714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.647492511557017,
+      "learning_rate": 5.101197750849718e-09,
+      "loss": 0.535,
+      "step": 3715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7121122022018234,
+      "learning_rate": 4.829205449476959e-09,
+      "loss": 0.4786,
+      "step": 3716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8710718551412433,
+      "learning_rate": 4.56466262584776e-09,
+      "loss": 0.5662,
+      "step": 3717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7263154257647637,
+      "learning_rate": 4.307569477124407e-09,
+      "loss": 0.5946,
+      "step": 3718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.829162228482798,
+      "learning_rate": 4.057926194922513e-09,
+      "loss": 0.5515,
+      "step": 3719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2284992961899004,
+      "learning_rate": 3.815732965303243e-09,
+      "loss": 1.0703,
+      "step": 3720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.79532256605668,
+      "learning_rate": 3.5809899687733186e-09,
+      "loss": 0.5024,
+      "step": 3721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0022319435874185,
+      "learning_rate": 3.353697380290566e-09,
+      "loss": 0.6535,
+      "step": 3722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.738628337247635,
+      "learning_rate": 3.1338553692561446e-09,
+      "loss": 0.521,
+      "step": 3723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7523663786951718,
+      "learning_rate": 2.9214640995212097e-09,
+      "loss": 0.5941,
+      "step": 3724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8775860384390548,
+      "learning_rate": 2.7165237293813594e-09,
+      "loss": 0.5457,
+      "step": 3725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.817830020542788,
+      "learning_rate": 2.5190344115821886e-09,
+      "loss": 0.5209,
+      "step": 3726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8446953932202192,
+      "learning_rate": 2.328996293312624e-09,
+      "loss": 0.6131,
+      "step": 3727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.320226453689705,
+      "learning_rate": 2.146409516209369e-09,
+      "loss": 1.0434,
+      "step": 3728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.858032336257363,
+      "learning_rate": 1.9712742163557898e-09,
+      "loss": 0.6052,
+      "step": 3729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8048378673455525,
+      "learning_rate": 1.8035905242830277e-09,
+      "loss": 0.6146,
+      "step": 3730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.692061247036353,
+      "learning_rate": 1.6433585649655582e-09,
+      "loss": 0.6163,
+      "step": 3731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0713343502275814,
+      "learning_rate": 1.490578457825631e-09,
+      "loss": 0.6257,
+      "step": 3732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.348117040982282,
+      "learning_rate": 1.3452503167321606e-09,
+      "loss": 1.0606,
+      "step": 3733
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.324386324198412,
+      "learning_rate": 1.2073742499985053e-09,
+      "loss": 1.1281,
+      "step": 3734
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8179952569333864,
+      "learning_rate": 1.0769503603846876e-09,
+      "loss": 0.5811,
+      "step": 3735
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.337861517262373,
+      "learning_rate": 9.53978745098505e-10,
+      "loss": 1.1589,
+      "step": 3736
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8020411311878894,
+      "learning_rate": 8.384594957888681e-10,
+      "loss": 0.6156,
+      "step": 3737
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8330075153553205,
+      "learning_rate": 7.303926985546828e-10,
+      "loss": 0.5674,
+      "step": 3738
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.387405764045576,
+      "learning_rate": 6.297784339392987e-10,
+      "loss": 1.1183,
+      "step": 3739
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7189534103455781,
+      "learning_rate": 5.366167769305097e-10,
+      "loss": 0.547,
+      "step": 3740
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2998765033597413,
+      "learning_rate": 4.5090779696277396e-10,
+      "loss": 1.1374,
+      "step": 3741
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7484397654502877,
+      "learning_rate": 3.7265155791610386e-10,
+      "loss": 0.5808,
+      "step": 3742
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.293957663321462,
+      "learning_rate": 3.018481181138455e-10,
+      "loss": 1.093,
+      "step": 3743
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3452351182183495,
+      "learning_rate": 2.3849753032823e-10,
+      "loss": 1.1405,
+      "step": 3744
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7609443643472347,
+      "learning_rate": 1.8259984177371181e-10,
+      "loss": 0.5945,
+      "step": 3745
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9016534408767585,
+      "learning_rate": 1.3415509411252024e-10,
+      "loss": 0.6386,
+      "step": 3746
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.914101359329578,
+      "learning_rate": 9.316332344910806e-11,
+      "loss": 0.6205,
+      "step": 3747
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8872682934857807,
+      "learning_rate": 5.962456033681285e-11,
+      "loss": 0.5699,
+      "step": 3748
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4149117273565577,
+      "learning_rate": 3.3538829771195824e-11,
+      "loss": 1.0771,
+      "step": 3749
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.291107127828432,
+      "learning_rate": 1.4906151193372352e-11,
+      "loss": 1.131,
+      "step": 3750
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.219930924501786,
+      "learning_rate": 3.726538493342702e-12,
+      "loss": 1.1079,
+      "step": 3751
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2524600844649556,
+      "learning_rate": 0.0,
+      "loss": 0.9535,
+      "step": 3752
+    },
+    {
+      "epoch": 1.0,
+      "step": 3752,
+      "total_flos": 7826242895216640.0,
+      "train_loss": 0.7736163951043509,
+      "train_runtime": 87529.3125,
+      "train_samples_per_second": 10.975,
+      "train_steps_per_second": 0.043
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3752,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 7826242895216640.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}