diff --git "a/final/trainer_state.json" "b/final/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/final/trainer_state.json"
@@ -0,0 +1,11396 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996918335901387,
+  "eval_steps": 500,
+  "global_step": 1622,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006163328197226503,
+      "grad_norm": 7.7007341384887695,
+      "learning_rate": 2.45398773006135e-07,
+      "loss": 1.9014,
+      "step": 1
+    },
+    {
+      "epoch": 0.0012326656394453005,
+      "grad_norm": 7.831578254699707,
+      "learning_rate": 4.9079754601227e-07,
+      "loss": 1.9231,
+      "step": 2
+    },
+    {
+      "epoch": 0.0018489984591679508,
+      "grad_norm": 7.626069068908691,
+      "learning_rate": 7.36196319018405e-07,
+      "loss": 1.9178,
+      "step": 3
+    },
+    {
+      "epoch": 0.002465331278890601,
+      "grad_norm": 7.49867057800293,
+      "learning_rate": 9.8159509202454e-07,
+      "loss": 1.9453,
+      "step": 4
+    },
+    {
+      "epoch": 0.0030816640986132513,
+      "grad_norm": 7.060399532318115,
+      "learning_rate": 1.226993865030675e-06,
+      "loss": 1.8867,
+      "step": 5
+    },
+    {
+      "epoch": 0.0036979969183359015,
+      "grad_norm": 6.3593010902404785,
+      "learning_rate": 1.47239263803681e-06,
+      "loss": 1.9149,
+      "step": 6
+    },
+    {
+      "epoch": 0.004314329738058551,
+      "grad_norm": 5.817101955413818,
+      "learning_rate": 1.717791411042945e-06,
+      "loss": 1.8836,
+      "step": 7
+    },
+    {
+      "epoch": 0.004930662557781202,
+      "grad_norm": 4.121643543243408,
+      "learning_rate": 1.96319018404908e-06,
+      "loss": 1.8643,
+      "step": 8
+    },
+    {
+      "epoch": 0.005546995377503852,
+      "grad_norm": 3.0739834308624268,
+      "learning_rate": 2.208588957055215e-06,
+      "loss": 1.7948,
+      "step": 9
+    },
+    {
+      "epoch": 0.0061633281972265025,
+      "grad_norm": 2.5905909538269043,
+      "learning_rate": 2.45398773006135e-06,
+      "loss": 1.7899,
+      "step": 10
+    },
+    {
+      "epoch": 0.006779661016949152,
+      "grad_norm": 4.185056209564209,
+      "learning_rate": 2.699386503067485e-06,
+      "loss": 1.8681,
+      "step": 11
+    },
+    {
+      "epoch": 0.007395993836671803,
+      "grad_norm": 4.769819259643555,
+      "learning_rate": 2.94478527607362e-06,
+      "loss": 1.8309,
+      "step": 12
+    },
+    {
+      "epoch": 0.008012326656394453,
+      "grad_norm": 4.520302772521973,
+      "learning_rate": 3.190184049079755e-06,
+      "loss": 1.7631,
+      "step": 13
+    },
+    {
+      "epoch": 0.008628659476117103,
+      "grad_norm": 4.179266452789307,
+      "learning_rate": 3.43558282208589e-06,
+      "loss": 1.878,
+      "step": 14
+    },
+    {
+      "epoch": 0.009244992295839754,
+      "grad_norm": 3.0450217723846436,
+      "learning_rate": 3.680981595092025e-06,
+      "loss": 1.7691,
+      "step": 15
+    },
+    {
+      "epoch": 0.009861325115562404,
+      "grad_norm": 2.297912359237671,
+      "learning_rate": 3.92638036809816e-06,
+      "loss": 1.778,
+      "step": 16
+    },
+    {
+      "epoch": 0.010477657935285054,
+      "grad_norm": 1.6684463024139404,
+      "learning_rate": 4.171779141104294e-06,
+      "loss": 1.7915,
+      "step": 17
+    },
+    {
+      "epoch": 0.011093990755007704,
+      "grad_norm": 1.660719871520996,
+      "learning_rate": 4.41717791411043e-06,
+      "loss": 1.7559,
+      "step": 18
+    },
+    {
+      "epoch": 0.011710323574730355,
+      "grad_norm": 1.5748722553253174,
+      "learning_rate": 4.662576687116564e-06,
+      "loss": 1.7705,
+      "step": 19
+    },
+    {
+      "epoch": 0.012326656394453005,
+      "grad_norm": 1.7511894702911377,
+      "learning_rate": 4.9079754601227e-06,
+      "loss": 1.7254,
+      "step": 20
+    },
+    {
+      "epoch": 0.012942989214175655,
+      "grad_norm": 1.5858523845672607,
+      "learning_rate": 5.153374233128835e-06,
+      "loss": 1.7507,
+      "step": 21
+    },
+    {
+      "epoch": 0.013559322033898305,
+      "grad_norm": 1.4151811599731445,
+      "learning_rate": 5.39877300613497e-06,
+      "loss": 1.6743,
+      "step": 22
+    },
+    {
+      "epoch": 0.014175654853620955,
+      "grad_norm": 1.2708033323287964,
+      "learning_rate": 5.644171779141104e-06,
+      "loss": 1.6913,
+      "step": 23
+    },
+    {
+      "epoch": 0.014791987673343606,
+      "grad_norm": 1.2384111881256104,
+      "learning_rate": 5.88957055214724e-06,
+      "loss": 1.7353,
+      "step": 24
+    },
+    {
+      "epoch": 0.015408320493066256,
+      "grad_norm": 1.2881749868392944,
+      "learning_rate": 6.134969325153375e-06,
+      "loss": 1.767,
+      "step": 25
+    },
+    {
+      "epoch": 0.016024653312788906,
+      "grad_norm": 1.2464679479599,
+      "learning_rate": 6.38036809815951e-06,
+      "loss": 1.7554,
+      "step": 26
+    },
+    {
+      "epoch": 0.016640986132511557,
+      "grad_norm": 1.24666166305542,
+      "learning_rate": 6.625766871165644e-06,
+      "loss": 1.6588,
+      "step": 27
+    },
+    {
+      "epoch": 0.017257318952234205,
+      "grad_norm": 1.20462167263031,
+      "learning_rate": 6.87116564417178e-06,
+      "loss": 1.7058,
+      "step": 28
+    },
+    {
+      "epoch": 0.017873651771956857,
+      "grad_norm": 1.073009967803955,
+      "learning_rate": 7.116564417177915e-06,
+      "loss": 1.7216,
+      "step": 29
+    },
+    {
+      "epoch": 0.01848998459167951,
+      "grad_norm": 1.090959906578064,
+      "learning_rate": 7.36196319018405e-06,
+      "loss": 1.6912,
+      "step": 30
+    },
+    {
+      "epoch": 0.019106317411402157,
+      "grad_norm": 1.118828535079956,
+      "learning_rate": 7.6073619631901856e-06,
+      "loss": 1.7231,
+      "step": 31
+    },
+    {
+      "epoch": 0.019722650231124808,
+      "grad_norm": 1.1499671936035156,
+      "learning_rate": 7.85276073619632e-06,
+      "loss": 1.658,
+      "step": 32
+    },
+    {
+      "epoch": 0.020338983050847456,
+      "grad_norm": 1.1840991973876953,
+      "learning_rate": 8.098159509202455e-06,
+      "loss": 1.679,
+      "step": 33
+    },
+    {
+      "epoch": 0.020955315870570108,
+      "grad_norm": 1.049609661102295,
+      "learning_rate": 8.343558282208589e-06,
+      "loss": 1.7687,
+      "step": 34
+    },
+    {
+      "epoch": 0.02157164869029276,
+      "grad_norm": 1.0060491561889648,
+      "learning_rate": 8.588957055214725e-06,
+      "loss": 1.685,
+      "step": 35
+    },
+    {
+      "epoch": 0.022187981510015407,
+      "grad_norm": 1.0304157733917236,
+      "learning_rate": 8.83435582822086e-06,
+      "loss": 1.6884,
+      "step": 36
+    },
+    {
+      "epoch": 0.02280431432973806,
+      "grad_norm": 0.9977303743362427,
+      "learning_rate": 9.079754601226994e-06,
+      "loss": 1.6903,
+      "step": 37
+    },
+    {
+      "epoch": 0.02342064714946071,
+      "grad_norm": 1.008644938468933,
+      "learning_rate": 9.325153374233129e-06,
+      "loss": 1.6521,
+      "step": 38
+    },
+    {
+      "epoch": 0.02403697996918336,
+      "grad_norm": 1.046408772468567,
+      "learning_rate": 9.570552147239264e-06,
+      "loss": 1.7299,
+      "step": 39
+    },
+    {
+      "epoch": 0.02465331278890601,
+      "grad_norm": 1.0336506366729736,
+      "learning_rate": 9.8159509202454e-06,
+      "loss": 1.6146,
+      "step": 40
+    },
+    {
+      "epoch": 0.025269645608628658,
+      "grad_norm": 1.0193053483963013,
+      "learning_rate": 1.0061349693251534e-05,
+      "loss": 1.6733,
+      "step": 41
+    },
+    {
+      "epoch": 0.02588597842835131,
+      "grad_norm": 1.0004470348358154,
+      "learning_rate": 1.030674846625767e-05,
+      "loss": 1.7239,
+      "step": 42
+    },
+    {
+      "epoch": 0.02650231124807396,
+      "grad_norm": 1.0116422176361084,
+      "learning_rate": 1.0552147239263804e-05,
+      "loss": 1.7232,
+      "step": 43
+    },
+    {
+      "epoch": 0.02711864406779661,
+      "grad_norm": 0.9728295207023621,
+      "learning_rate": 1.079754601226994e-05,
+      "loss": 1.6871,
+      "step": 44
+    },
+    {
+      "epoch": 0.02773497688751926,
+      "grad_norm": 0.9763592481613159,
+      "learning_rate": 1.1042944785276076e-05,
+      "loss": 1.6709,
+      "step": 45
+    },
+    {
+      "epoch": 0.02835130970724191,
+      "grad_norm": 0.9567804932594299,
+      "learning_rate": 1.1288343558282208e-05,
+      "loss": 1.6082,
+      "step": 46
+    },
+    {
+      "epoch": 0.02896764252696456,
+      "grad_norm": 0.9812591671943665,
+      "learning_rate": 1.1533742331288344e-05,
+      "loss": 1.6607,
+      "step": 47
+    },
+    {
+      "epoch": 0.029583975346687212,
+      "grad_norm": 1.1011967658996582,
+      "learning_rate": 1.177914110429448e-05,
+      "loss": 1.7133,
+      "step": 48
+    },
+    {
+      "epoch": 0.03020030816640986,
+      "grad_norm": 0.997948944568634,
+      "learning_rate": 1.2024539877300614e-05,
+      "loss": 1.6553,
+      "step": 49
+    },
+    {
+      "epoch": 0.030816640986132512,
+      "grad_norm": 0.9882333874702454,
+      "learning_rate": 1.226993865030675e-05,
+      "loss": 1.6809,
+      "step": 50
+    },
+    {
+      "epoch": 0.03143297380585516,
+      "grad_norm": 1.0237276554107666,
+      "learning_rate": 1.2515337423312886e-05,
+      "loss": 1.7517,
+      "step": 51
+    },
+    {
+      "epoch": 0.03204930662557781,
+      "grad_norm": 0.9705806374549866,
+      "learning_rate": 1.276073619631902e-05,
+      "loss": 1.7056,
+      "step": 52
+    },
+    {
+      "epoch": 0.03266563944530046,
+      "grad_norm": 0.9457777738571167,
+      "learning_rate": 1.3006134969325156e-05,
+      "loss": 1.6469,
+      "step": 53
+    },
+    {
+      "epoch": 0.033281972265023115,
+      "grad_norm": 1.01027250289917,
+      "learning_rate": 1.3251533742331288e-05,
+      "loss": 1.6744,
+      "step": 54
+    },
+    {
+      "epoch": 0.03389830508474576,
+      "grad_norm": 0.9895347356796265,
+      "learning_rate": 1.3496932515337424e-05,
+      "loss": 1.6506,
+      "step": 55
+    },
+    {
+      "epoch": 0.03451463790446841,
+      "grad_norm": 1.055320143699646,
+      "learning_rate": 1.374233128834356e-05,
+      "loss": 1.6655,
+      "step": 56
+    },
+    {
+      "epoch": 0.035130970724191066,
+      "grad_norm": 1.0253647565841675,
+      "learning_rate": 1.3987730061349694e-05,
+      "loss": 1.6899,
+      "step": 57
+    },
+    {
+      "epoch": 0.035747303543913714,
+      "grad_norm": 0.9785537123680115,
+      "learning_rate": 1.423312883435583e-05,
+      "loss": 1.6384,
+      "step": 58
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 1.0246118307113647,
+      "learning_rate": 1.4478527607361965e-05,
+      "loss": 1.7097,
+      "step": 59
+    },
+    {
+      "epoch": 0.03697996918335902,
+      "grad_norm": 1.0297993421554565,
+      "learning_rate": 1.47239263803681e-05,
+      "loss": 1.7113,
+      "step": 60
+    },
+    {
+      "epoch": 0.037596302003081665,
+      "grad_norm": 1.0864713191986084,
+      "learning_rate": 1.4969325153374235e-05,
+      "loss": 1.689,
+      "step": 61
+    },
+    {
+      "epoch": 0.03821263482280431,
+      "grad_norm": 0.8881139159202576,
+      "learning_rate": 1.5214723926380371e-05,
+      "loss": 1.6708,
+      "step": 62
+    },
+    {
+      "epoch": 0.03882896764252696,
+      "grad_norm": 1.0513116121292114,
+      "learning_rate": 1.5460122699386504e-05,
+      "loss": 1.6636,
+      "step": 63
+    },
+    {
+      "epoch": 0.039445300462249616,
+      "grad_norm": 0.9132838249206543,
+      "learning_rate": 1.570552147239264e-05,
+      "loss": 1.6633,
+      "step": 64
+    },
+    {
+      "epoch": 0.040061633281972264,
+      "grad_norm": 0.9439066648483276,
+      "learning_rate": 1.5950920245398772e-05,
+      "loss": 1.708,
+      "step": 65
+    },
+    {
+      "epoch": 0.04067796610169491,
+      "grad_norm": 0.8892993927001953,
+      "learning_rate": 1.619631901840491e-05,
+      "loss": 1.6697,
+      "step": 66
+    },
+    {
+      "epoch": 0.04129429892141757,
+      "grad_norm": 0.8928653597831726,
+      "learning_rate": 1.6441717791411043e-05,
+      "loss": 1.6138,
+      "step": 67
+    },
+    {
+      "epoch": 0.041910631741140215,
+      "grad_norm": 0.9380534887313843,
+      "learning_rate": 1.6687116564417178e-05,
+      "loss": 1.6451,
+      "step": 68
+    },
+    {
+      "epoch": 0.042526964560862864,
+      "grad_norm": 0.9998738169670105,
+      "learning_rate": 1.6932515337423315e-05,
+      "loss": 1.7299,
+      "step": 69
+    },
+    {
+      "epoch": 0.04314329738058552,
+      "grad_norm": 0.9376276731491089,
+      "learning_rate": 1.717791411042945e-05,
+      "loss": 1.6444,
+      "step": 70
+    },
+    {
+      "epoch": 0.04375963020030817,
+      "grad_norm": 0.9520479440689087,
+      "learning_rate": 1.7423312883435583e-05,
+      "loss": 1.667,
+      "step": 71
+    },
+    {
+      "epoch": 0.044375963020030815,
+      "grad_norm": 0.8968117237091064,
+      "learning_rate": 1.766871165644172e-05,
+      "loss": 1.6785,
+      "step": 72
+    },
+    {
+      "epoch": 0.04499229583975347,
+      "grad_norm": 0.9188838601112366,
+      "learning_rate": 1.7914110429447855e-05,
+      "loss": 1.6208,
+      "step": 73
+    },
+    {
+      "epoch": 0.04560862865947612,
+      "grad_norm": 0.8976795077323914,
+      "learning_rate": 1.815950920245399e-05,
+      "loss": 1.5895,
+      "step": 74
+    },
+    {
+      "epoch": 0.046224961479198766,
+      "grad_norm": 0.8823075294494629,
+      "learning_rate": 1.8404907975460123e-05,
+      "loss": 1.646,
+      "step": 75
+    },
+    {
+      "epoch": 0.04684129429892142,
+      "grad_norm": 0.8592455387115479,
+      "learning_rate": 1.8650306748466257e-05,
+      "loss": 1.5647,
+      "step": 76
+    },
+    {
+      "epoch": 0.04745762711864407,
+      "grad_norm": 0.8513899445533752,
+      "learning_rate": 1.8895705521472395e-05,
+      "loss": 1.6174,
+      "step": 77
+    },
+    {
+      "epoch": 0.04807395993836672,
+      "grad_norm": 0.9042425155639648,
+      "learning_rate": 1.914110429447853e-05,
+      "loss": 1.634,
+      "step": 78
+    },
+    {
+      "epoch": 0.048690292758089365,
+      "grad_norm": 0.8949453234672546,
+      "learning_rate": 1.9386503067484663e-05,
+      "loss": 1.5943,
+      "step": 79
+    },
+    {
+      "epoch": 0.04930662557781202,
+      "grad_norm": 0.8401027321815491,
+      "learning_rate": 1.96319018404908e-05,
+      "loss": 1.5745,
+      "step": 80
+    },
+    {
+      "epoch": 0.04992295839753467,
+      "grad_norm": 0.8911848664283752,
+      "learning_rate": 1.9877300613496935e-05,
+      "loss": 1.6336,
+      "step": 81
+    },
+    {
+      "epoch": 0.050539291217257316,
+      "grad_norm": 0.8529688119888306,
+      "learning_rate": 2.012269938650307e-05,
+      "loss": 1.6164,
+      "step": 82
+    },
+    {
+      "epoch": 0.05115562403697997,
+      "grad_norm": 0.9154197573661804,
+      "learning_rate": 2.0368098159509203e-05,
+      "loss": 1.6758,
+      "step": 83
+    },
+    {
+      "epoch": 0.05177195685670262,
+      "grad_norm": 0.866750955581665,
+      "learning_rate": 2.061349693251534e-05,
+      "loss": 1.5921,
+      "step": 84
+    },
+    {
+      "epoch": 0.05238828967642527,
+      "grad_norm": 0.8651569485664368,
+      "learning_rate": 2.0858895705521474e-05,
+      "loss": 1.6166,
+      "step": 85
+    },
+    {
+      "epoch": 0.05300462249614792,
+      "grad_norm": 0.9068999290466309,
+      "learning_rate": 2.110429447852761e-05,
+      "loss": 1.6648,
+      "step": 86
+    },
+    {
+      "epoch": 0.05362095531587057,
+      "grad_norm": 0.8784379959106445,
+      "learning_rate": 2.1349693251533746e-05,
+      "loss": 1.6612,
+      "step": 87
+    },
+    {
+      "epoch": 0.05423728813559322,
+      "grad_norm": 0.8599914908409119,
+      "learning_rate": 2.159509202453988e-05,
+      "loss": 1.6372,
+      "step": 88
+    },
+    {
+      "epoch": 0.054853620955315874,
+      "grad_norm": 0.8836574554443359,
+      "learning_rate": 2.184049079754601e-05,
+      "loss": 1.6742,
+      "step": 89
+    },
+    {
+      "epoch": 0.05546995377503852,
+      "grad_norm": 0.9129705429077148,
+      "learning_rate": 2.2085889570552152e-05,
+      "loss": 1.6252,
+      "step": 90
+    },
+    {
+      "epoch": 0.05608628659476117,
+      "grad_norm": 0.8584738373756409,
+      "learning_rate": 2.2331288343558283e-05,
+      "loss": 1.6668,
+      "step": 91
+    },
+    {
+      "epoch": 0.05670261941448382,
+      "grad_norm": 0.8560867309570312,
+      "learning_rate": 2.2576687116564417e-05,
+      "loss": 1.5992,
+      "step": 92
+    },
+    {
+      "epoch": 0.05731895223420647,
+      "grad_norm": 0.8884056806564331,
+      "learning_rate": 2.2822085889570554e-05,
+      "loss": 1.6916,
+      "step": 93
+    },
+    {
+      "epoch": 0.05793528505392912,
+      "grad_norm": 0.8534591197967529,
+      "learning_rate": 2.3067484662576688e-05,
+      "loss": 1.627,
+      "step": 94
+    },
+    {
+      "epoch": 0.05855161787365177,
+      "grad_norm": 0.8151088356971741,
+      "learning_rate": 2.3312883435582822e-05,
+      "loss": 1.5843,
+      "step": 95
+    },
+    {
+      "epoch": 0.059167950693374424,
+      "grad_norm": 0.8539857864379883,
+      "learning_rate": 2.355828220858896e-05,
+      "loss": 1.624,
+      "step": 96
+    },
+    {
+      "epoch": 0.05978428351309707,
+      "grad_norm": 0.8533106446266174,
+      "learning_rate": 2.3803680981595094e-05,
+      "loss": 1.7183,
+      "step": 97
+    },
+    {
+      "epoch": 0.06040061633281972,
+      "grad_norm": 0.9716623425483704,
+      "learning_rate": 2.4049079754601228e-05,
+      "loss": 1.7222,
+      "step": 98
+    },
+    {
+      "epoch": 0.061016949152542375,
+      "grad_norm": 0.8098641037940979,
+      "learning_rate": 2.4294478527607366e-05,
+      "loss": 1.6397,
+      "step": 99
+    },
+    {
+      "epoch": 0.061633281972265024,
+      "grad_norm": 0.8411961793899536,
+      "learning_rate": 2.45398773006135e-05,
+      "loss": 1.6258,
+      "step": 100
+    },
+    {
+      "epoch": 0.06224961479198767,
+      "grad_norm": 0.911243736743927,
+      "learning_rate": 2.4785276073619634e-05,
+      "loss": 1.6559,
+      "step": 101
+    },
+    {
+      "epoch": 0.06286594761171033,
+      "grad_norm": 0.9482967257499695,
+      "learning_rate": 2.503067484662577e-05,
+      "loss": 1.7146,
+      "step": 102
+    },
+    {
+      "epoch": 0.06348228043143297,
+      "grad_norm": 0.8271360397338867,
+      "learning_rate": 2.5276073619631905e-05,
+      "loss": 1.5776,
+      "step": 103
+    },
+    {
+      "epoch": 0.06409861325115562,
+      "grad_norm": 0.8761653900146484,
+      "learning_rate": 2.552147239263804e-05,
+      "loss": 1.6654,
+      "step": 104
+    },
+    {
+      "epoch": 0.06471494607087827,
+      "grad_norm": 0.8657371401786804,
+      "learning_rate": 2.576687116564417e-05,
+      "loss": 1.7236,
+      "step": 105
+    },
+    {
+      "epoch": 0.06533127889060092,
+      "grad_norm": 0.810516357421875,
+      "learning_rate": 2.601226993865031e-05,
+      "loss": 1.6357,
+      "step": 106
+    },
+    {
+      "epoch": 0.06594761171032358,
+      "grad_norm": 0.8823238015174866,
+      "learning_rate": 2.6257668711656442e-05,
+      "loss": 1.6358,
+      "step": 107
+    },
+    {
+      "epoch": 0.06656394453004623,
+      "grad_norm": 1.0349003076553345,
+      "learning_rate": 2.6503067484662576e-05,
+      "loss": 1.6785,
+      "step": 108
+    },
+    {
+      "epoch": 0.06718027734976888,
+      "grad_norm": 0.7891526222229004,
+      "learning_rate": 2.6748466257668714e-05,
+      "loss": 1.6366,
+      "step": 109
+    },
+    {
+      "epoch": 0.06779661016949153,
+      "grad_norm": 0.8495930433273315,
+      "learning_rate": 2.6993865030674848e-05,
+      "loss": 1.6392,
+      "step": 110
+    },
+    {
+      "epoch": 0.06841294298921417,
+      "grad_norm": 0.8164300322532654,
+      "learning_rate": 2.7239263803680982e-05,
+      "loss": 1.6739,
+      "step": 111
+    },
+    {
+      "epoch": 0.06902927580893682,
+      "grad_norm": 0.8112594485282898,
+      "learning_rate": 2.748466257668712e-05,
+      "loss": 1.6826,
+      "step": 112
+    },
+    {
+      "epoch": 0.06964560862865947,
+      "grad_norm": 0.8023330569267273,
+      "learning_rate": 2.7730061349693253e-05,
+      "loss": 1.6632,
+      "step": 113
+    },
+    {
+      "epoch": 0.07026194144838213,
+      "grad_norm": 0.788388729095459,
+      "learning_rate": 2.7975460122699388e-05,
+      "loss": 1.6048,
+      "step": 114
+    },
+    {
+      "epoch": 0.07087827426810478,
+      "grad_norm": 0.7653685808181763,
+      "learning_rate": 2.8220858895705525e-05,
+      "loss": 1.5936,
+      "step": 115
+    },
+    {
+      "epoch": 0.07149460708782743,
+      "grad_norm": 0.7973049283027649,
+      "learning_rate": 2.846625766871166e-05,
+      "loss": 1.6586,
+      "step": 116
+    },
+    {
+      "epoch": 0.07211093990755008,
+      "grad_norm": 0.7768449783325195,
+      "learning_rate": 2.8711656441717793e-05,
+      "loss": 1.5832,
+      "step": 117
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 0.7748964428901672,
+      "learning_rate": 2.895705521472393e-05,
+      "loss": 1.6792,
+      "step": 118
+    },
+    {
+      "epoch": 0.07334360554699537,
+      "grad_norm": 0.8378171920776367,
+      "learning_rate": 2.9202453987730065e-05,
+      "loss": 1.6247,
+      "step": 119
+    },
+    {
+      "epoch": 0.07395993836671803,
+      "grad_norm": 0.7740994691848755,
+      "learning_rate": 2.94478527607362e-05,
+      "loss": 1.6401,
+      "step": 120
+    },
+    {
+      "epoch": 0.07457627118644068,
+      "grad_norm": 0.7645397186279297,
+      "learning_rate": 2.9693251533742337e-05,
+      "loss": 1.6607,
+      "step": 121
+    },
+    {
+      "epoch": 0.07519260400616333,
+      "grad_norm": 0.7948278188705444,
+      "learning_rate": 2.993865030674847e-05,
+      "loss": 1.6642,
+      "step": 122
+    },
+    {
+      "epoch": 0.07580893682588598,
+      "grad_norm": 0.7775226831436157,
+      "learning_rate": 3.01840490797546e-05,
+      "loss": 1.6617,
+      "step": 123
+    },
+    {
+      "epoch": 0.07642526964560863,
+      "grad_norm": 0.8228120803833008,
+      "learning_rate": 3.0429447852760742e-05,
+      "loss": 1.5793,
+      "step": 124
+    },
+    {
+      "epoch": 0.07704160246533127,
+      "grad_norm": 0.8086126446723938,
+      "learning_rate": 3.067484662576687e-05,
+      "loss": 1.6518,
+      "step": 125
+    },
+    {
+      "epoch": 0.07765793528505392,
+      "grad_norm": 0.7439844608306885,
+      "learning_rate": 3.092024539877301e-05,
+      "loss": 1.6284,
+      "step": 126
+    },
+    {
+      "epoch": 0.07827426810477658,
+      "grad_norm": 0.8310662508010864,
+      "learning_rate": 3.116564417177915e-05,
+      "loss": 1.6417,
+      "step": 127
+    },
+    {
+      "epoch": 0.07889060092449923,
+      "grad_norm": 0.8119212985038757,
+      "learning_rate": 3.141104294478528e-05,
+      "loss": 1.6033,
+      "step": 128
+    },
+    {
+      "epoch": 0.07950693374422188,
+      "grad_norm": 0.7821022272109985,
+      "learning_rate": 3.1656441717791416e-05,
+      "loss": 1.6896,
+      "step": 129
+    },
+    {
+      "epoch": 0.08012326656394453,
+      "grad_norm": 0.7965499758720398,
+      "learning_rate": 3.1901840490797544e-05,
+      "loss": 1.6492,
+      "step": 130
+    },
+    {
+      "epoch": 0.08073959938366718,
+      "grad_norm": 0.7747324109077454,
+      "learning_rate": 3.2147239263803684e-05,
+      "loss": 1.6301,
+      "step": 131
+    },
+    {
+      "epoch": 0.08135593220338982,
+      "grad_norm": 0.7778410911560059,
+      "learning_rate": 3.239263803680982e-05,
+      "loss": 1.5949,
+      "step": 132
+    },
+    {
+      "epoch": 0.08197226502311249,
+      "grad_norm": 0.8136535286903381,
+      "learning_rate": 3.263803680981595e-05,
+      "loss": 1.5787,
+      "step": 133
+    },
+    {
+      "epoch": 0.08258859784283513,
+      "grad_norm": 0.7390981316566467,
+      "learning_rate": 3.288343558282209e-05,
+      "loss": 1.6162,
+      "step": 134
+    },
+    {
+      "epoch": 0.08320493066255778,
+      "grad_norm": 0.8133728504180908,
+      "learning_rate": 3.312883435582822e-05,
+      "loss": 1.6204,
+      "step": 135
+    },
+    {
+      "epoch": 0.08382126348228043,
+      "grad_norm": 0.7785617113113403,
+      "learning_rate": 3.3374233128834355e-05,
+      "loss": 1.664,
+      "step": 136
+    },
+    {
+      "epoch": 0.08443759630200308,
+      "grad_norm": 0.8097333908081055,
+      "learning_rate": 3.3619631901840496e-05,
+      "loss": 1.6337,
+      "step": 137
+    },
+    {
+      "epoch": 0.08505392912172573,
+      "grad_norm": 0.7560635209083557,
+      "learning_rate": 3.386503067484663e-05,
+      "loss": 1.6624,
+      "step": 138
+    },
+    {
+      "epoch": 0.08567026194144839,
+      "grad_norm": 0.735256016254425,
+      "learning_rate": 3.4110429447852764e-05,
+      "loss": 1.6295,
+      "step": 139
+    },
+    {
+      "epoch": 0.08628659476117104,
+      "grad_norm": 0.8018369674682617,
+      "learning_rate": 3.43558282208589e-05,
+      "loss": 1.6347,
+      "step": 140
+    },
+    {
+      "epoch": 0.08690292758089369,
+      "grad_norm": 0.7610778212547302,
+      "learning_rate": 3.460122699386503e-05,
+      "loss": 1.6675,
+      "step": 141
+    },
+    {
+      "epoch": 0.08751926040061633,
+      "grad_norm": 0.7676482200622559,
+      "learning_rate": 3.4846625766871166e-05,
+      "loss": 1.7137,
+      "step": 142
+    },
+    {
+      "epoch": 0.08813559322033898,
+      "grad_norm": 0.7479026913642883,
+      "learning_rate": 3.509202453987731e-05,
+      "loss": 1.7188,
+      "step": 143
+    },
+    {
+      "epoch": 0.08875192604006163,
+      "grad_norm": 0.7722792625427246,
+      "learning_rate": 3.533742331288344e-05,
+      "loss": 1.6932,
+      "step": 144
+    },
+    {
+      "epoch": 0.08936825885978428,
+      "grad_norm": 0.7511256337165833,
+      "learning_rate": 3.5582822085889576e-05,
+      "loss": 1.6144,
+      "step": 145
+    },
+    {
+      "epoch": 0.08998459167950694,
+      "grad_norm": 0.7605886459350586,
+      "learning_rate": 3.582822085889571e-05,
+      "loss": 1.635,
+      "step": 146
+    },
+    {
+      "epoch": 0.09060092449922959,
+      "grad_norm": 0.7359566688537598,
+      "learning_rate": 3.6073619631901844e-05,
+      "loss": 1.6192,
+      "step": 147
+    },
+    {
+      "epoch": 0.09121725731895224,
+      "grad_norm": 0.7442858815193176,
+      "learning_rate": 3.631901840490798e-05,
+      "loss": 1.5986,
+      "step": 148
+    },
+    {
+      "epoch": 0.09183359013867488,
+      "grad_norm": 0.7447007298469543,
+      "learning_rate": 3.656441717791412e-05,
+      "loss": 1.5948,
+      "step": 149
+    },
+    {
+      "epoch": 0.09244992295839753,
+      "grad_norm": 0.7689185738563538,
+      "learning_rate": 3.6809815950920246e-05,
+      "loss": 1.6534,
+      "step": 150
+    },
+    {
+      "epoch": 0.09306625577812018,
+      "grad_norm": 0.781383752822876,
+      "learning_rate": 3.705521472392638e-05,
+      "loss": 1.6269,
+      "step": 151
+    },
+    {
+      "epoch": 0.09368258859784284,
+      "grad_norm": 0.7619852423667908,
+      "learning_rate": 3.7300613496932514e-05,
+      "loss": 1.5797,
+      "step": 152
+    },
+    {
+      "epoch": 0.09429892141756549,
+      "grad_norm": 0.7769779562950134,
+      "learning_rate": 3.7546012269938655e-05,
+      "loss": 1.5976,
+      "step": 153
+    },
+    {
+      "epoch": 0.09491525423728814,
+      "grad_norm": 0.7349963188171387,
+      "learning_rate": 3.779141104294479e-05,
+      "loss": 1.6322,
+      "step": 154
+    },
+    {
+      "epoch": 0.09553158705701079,
+      "grad_norm": 0.7396400570869446,
+      "learning_rate": 3.8036809815950924e-05,
+      "loss": 1.6398,
+      "step": 155
+    },
+    {
+      "epoch": 0.09614791987673343,
+      "grad_norm": 0.7363472580909729,
+      "learning_rate": 3.828220858895706e-05,
+      "loss": 1.6313,
+      "step": 156
+    },
+    {
+      "epoch": 0.09676425269645608,
+      "grad_norm": 0.7341015338897705,
+      "learning_rate": 3.852760736196319e-05,
+      "loss": 1.7016,
+      "step": 157
+    },
+    {
+      "epoch": 0.09738058551617873,
+      "grad_norm": 0.6993916034698486,
+      "learning_rate": 3.8773006134969326e-05,
+      "loss": 1.5703,
+      "step": 158
+    },
+    {
+      "epoch": 0.09799691833590139,
+      "grad_norm": 0.7287641167640686,
+      "learning_rate": 3.901840490797547e-05,
+      "loss": 1.6438,
+      "step": 159
+    },
+    {
+      "epoch": 0.09861325115562404,
+      "grad_norm": 0.7177438139915466,
+      "learning_rate": 3.92638036809816e-05,
+      "loss": 1.6235,
+      "step": 160
+    },
+    {
+      "epoch": 0.09922958397534669,
+      "grad_norm": 0.6875810623168945,
+      "learning_rate": 3.9509202453987735e-05,
+      "loss": 1.593,
+      "step": 161
+    },
+    {
+      "epoch": 0.09984591679506934,
+      "grad_norm": 0.7038248777389526,
+      "learning_rate": 3.975460122699387e-05,
+      "loss": 1.6023,
+      "step": 162
+    },
+    {
+      "epoch": 0.10046224961479198,
+      "grad_norm": 0.687907338142395,
+      "learning_rate": 4e-05,
+      "loss": 1.608,
+      "step": 163
+    },
+    {
+      "epoch": 0.10107858243451463,
+      "grad_norm": 0.7126633524894714,
+      "learning_rate": 3.999995363513562e-05,
+      "loss": 1.6547,
+      "step": 164
+    },
+    {
+      "epoch": 0.1016949152542373,
+      "grad_norm": 0.7042025923728943,
+      "learning_rate": 3.999981454075744e-05,
+      "loss": 1.667,
+      "step": 165
+    },
+    {
+      "epoch": 0.10231124807395994,
+      "grad_norm": 0.6862210631370544,
+      "learning_rate": 3.999958271751038e-05,
+      "loss": 1.6081,
+      "step": 166
+    },
+    {
+      "epoch": 0.10292758089368259,
+      "grad_norm": 0.6989213824272156,
+      "learning_rate": 3.999925816646928e-05,
+      "loss": 1.6115,
+      "step": 167
+    },
+    {
+      "epoch": 0.10354391371340524,
+      "grad_norm": 0.690365195274353,
+      "learning_rate": 3.9998840889138925e-05,
+      "loss": 1.5594,
+      "step": 168
+    },
+    {
+      "epoch": 0.10416024653312789,
+      "grad_norm": 0.6726807951927185,
+      "learning_rate": 3.9998330887454e-05,
+      "loss": 1.6238,
+      "step": 169
+    },
+    {
+      "epoch": 0.10477657935285054,
+      "grad_norm": 0.6911548972129822,
+      "learning_rate": 3.9997728163779134e-05,
+      "loss": 1.6105,
+      "step": 170
+    },
+    {
+      "epoch": 0.10539291217257318,
+      "grad_norm": 0.6829053163528442,
+      "learning_rate": 3.999703272090884e-05,
+      "loss": 1.6448,
+      "step": 171
+    },
+    {
+      "epoch": 0.10600924499229585,
+      "grad_norm": 0.6991954445838928,
+      "learning_rate": 3.9996244562067525e-05,
+      "loss": 1.5788,
+      "step": 172
+    },
+    {
+      "epoch": 0.1066255778120185,
+      "grad_norm": 0.6998047828674316,
+      "learning_rate": 3.9995363690909487e-05,
+      "loss": 1.7358,
+      "step": 173
+    },
+    {
+      "epoch": 0.10724191063174114,
+      "grad_norm": 0.7109401822090149,
+      "learning_rate": 3.999439011151887e-05,
+      "loss": 1.5783,
+      "step": 174
+    },
+    {
+      "epoch": 0.10785824345146379,
+      "grad_norm": 0.7161952257156372,
+      "learning_rate": 3.9993323828409666e-05,
+      "loss": 1.6098,
+      "step": 175
+    },
+    {
+      "epoch": 0.10847457627118644,
+      "grad_norm": 0.7027195692062378,
+      "learning_rate": 3.999216484652567e-05,
+      "loss": 1.6,
+      "step": 176
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 0.6891278028488159,
+      "learning_rate": 3.999091317124049e-05,
+      "loss": 1.6253,
+      "step": 177
+    },
+    {
+      "epoch": 0.10970724191063175,
+      "grad_norm": 0.6901974081993103,
+      "learning_rate": 3.998956880835751e-05,
+      "loss": 1.6493,
+      "step": 178
+    },
+    {
+      "epoch": 0.1103235747303544,
+      "grad_norm": 0.7213894724845886,
+      "learning_rate": 3.998813176410985e-05,
+      "loss": 1.6286,
+      "step": 179
+    },
+    {
+      "epoch": 0.11093990755007704,
+      "grad_norm": 0.6912412643432617,
+      "learning_rate": 3.998660204516033e-05,
+      "loss": 1.5866,
+      "step": 180
+    },
+    {
+      "epoch": 0.11155624036979969,
+      "grad_norm": 0.6759384274482727,
+      "learning_rate": 3.998497965860149e-05,
+      "loss": 1.6171,
+      "step": 181
+    },
+    {
+      "epoch": 0.11217257318952234,
+      "grad_norm": 0.7188528180122375,
+      "learning_rate": 3.9983264611955496e-05,
+      "loss": 1.7641,
+      "step": 182
+    },
+    {
+      "epoch": 0.11278890600924499,
+      "grad_norm": 0.6806765198707581,
+      "learning_rate": 3.9981456913174135e-05,
+      "loss": 1.6582,
+      "step": 183
+    },
+    {
+      "epoch": 0.11340523882896764,
+      "grad_norm": 0.6784571409225464,
+      "learning_rate": 3.997955657063878e-05,
+      "loss": 1.5771,
+      "step": 184
+    },
+    {
+      "epoch": 0.1140215716486903,
+      "grad_norm": 0.7074373364448547,
+      "learning_rate": 3.997756359316035e-05,
+      "loss": 1.5754,
+      "step": 185
+    },
+    {
+      "epoch": 0.11463790446841295,
+      "grad_norm": 0.6583380699157715,
+      "learning_rate": 3.997547798997924e-05,
+      "loss": 1.6057,
+      "step": 186
+    },
+    {
+      "epoch": 0.1152542372881356,
+      "grad_norm": 0.6994054913520813,
+      "learning_rate": 3.9973299770765345e-05,
+      "loss": 1.5706,
+      "step": 187
+    },
+    {
+      "epoch": 0.11587057010785824,
+      "grad_norm": 0.6755475401878357,
+      "learning_rate": 3.997102894561793e-05,
+      "loss": 1.6144,
+      "step": 188
+    },
+    {
+      "epoch": 0.11648690292758089,
+      "grad_norm": 0.6866325736045837,
+      "learning_rate": 3.9968665525065654e-05,
+      "loss": 1.6255,
+      "step": 189
+    },
+    {
+      "epoch": 0.11710323574730354,
+      "grad_norm": 0.6652048230171204,
+      "learning_rate": 3.996620952006648e-05,
+      "loss": 1.605,
+      "step": 190
+    },
+    {
+      "epoch": 0.1177195685670262,
+      "grad_norm": 0.6833721399307251,
+      "learning_rate": 3.996366094200765e-05,
+      "loss": 1.6686,
+      "step": 191
+    },
+    {
+      "epoch": 0.11833590138674885,
+      "grad_norm": 0.6793074607849121,
+      "learning_rate": 3.9961019802705603e-05,
+      "loss": 1.637,
+      "step": 192
+    },
+    {
+      "epoch": 0.1189522342064715,
+      "grad_norm": 0.6983605623245239,
+      "learning_rate": 3.995828611440595e-05,
+      "loss": 1.6507,
+      "step": 193
+    },
+    {
+      "epoch": 0.11956856702619414,
+      "grad_norm": 0.6783537864685059,
+      "learning_rate": 3.995545988978339e-05,
+      "loss": 1.538,
+      "step": 194
+    },
+    {
+      "epoch": 0.12018489984591679,
+      "grad_norm": 0.6404557824134827,
+      "learning_rate": 3.9952541141941695e-05,
+      "loss": 1.5394,
+      "step": 195
+    },
+    {
+      "epoch": 0.12080123266563944,
+      "grad_norm": 0.6974167823791504,
+      "learning_rate": 3.994952988441358e-05,
+      "loss": 1.6152,
+      "step": 196
+    },
+    {
+      "epoch": 0.12141756548536209,
+      "grad_norm": 0.7140214443206787,
+      "learning_rate": 3.9946426131160706e-05,
+      "loss": 1.6507,
+      "step": 197
+    },
+    {
+      "epoch": 0.12203389830508475,
+      "grad_norm": 0.6546657681465149,
+      "learning_rate": 3.9943229896573594e-05,
+      "loss": 1.6298,
+      "step": 198
+    },
+    {
+      "epoch": 0.1226502311248074,
+      "grad_norm": 0.6407018303871155,
+      "learning_rate": 3.993994119547153e-05,
+      "loss": 1.5801,
+      "step": 199
+    },
+    {
+      "epoch": 0.12326656394453005,
+      "grad_norm": 0.6628043055534363,
+      "learning_rate": 3.993656004310253e-05,
+      "loss": 1.5879,
+      "step": 200
+    },
+    {
+      "epoch": 0.1238828967642527,
+      "grad_norm": 0.6629613637924194,
+      "learning_rate": 3.993308645514327e-05,
+      "loss": 1.5517,
+      "step": 201
+    },
+    {
+      "epoch": 0.12449922958397534,
+      "grad_norm": 0.6973757743835449,
+      "learning_rate": 3.992952044769898e-05,
+      "loss": 1.6574,
+      "step": 202
+    },
+    {
+      "epoch": 0.125115562403698,
+      "grad_norm": 0.677291750907898,
+      "learning_rate": 3.9925862037303425e-05,
+      "loss": 1.5528,
+      "step": 203
+    },
+    {
+      "epoch": 0.12573189522342065,
+      "grad_norm": 0.6733261942863464,
+      "learning_rate": 3.992211124091876e-05,
+      "loss": 1.6105,
+      "step": 204
+    },
+    {
+      "epoch": 0.1263482280431433,
+      "grad_norm": 0.6975075602531433,
+      "learning_rate": 3.9918268075935514e-05,
+      "loss": 1.6032,
+      "step": 205
+    },
+    {
+      "epoch": 0.12696456086286595,
+      "grad_norm": 0.6801272034645081,
+      "learning_rate": 3.991433256017246e-05,
+      "loss": 1.6238,
+      "step": 206
+    },
+    {
+      "epoch": 0.1275808936825886,
+      "grad_norm": 0.6633975505828857,
+      "learning_rate": 3.9910304711876556e-05,
+      "loss": 1.5586,
+      "step": 207
+    },
+    {
+      "epoch": 0.12819722650231125,
+      "grad_norm": 0.6919019818305969,
+      "learning_rate": 3.990618454972288e-05,
+      "loss": 1.5929,
+      "step": 208
+    },
+    {
+      "epoch": 0.1288135593220339,
+      "grad_norm": 0.6731187105178833,
+      "learning_rate": 3.9901972092814504e-05,
+      "loss": 1.6332,
+      "step": 209
+    },
+    {
+      "epoch": 0.12942989214175654,
+      "grad_norm": 0.6671748757362366,
+      "learning_rate": 3.989766736068242e-05,
+      "loss": 1.6109,
+      "step": 210
+    },
+    {
+      "epoch": 0.1300462249614792,
+      "grad_norm": 0.6681280136108398,
+      "learning_rate": 3.989327037328548e-05,
+      "loss": 1.5844,
+      "step": 211
+    },
+    {
+      "epoch": 0.13066255778120184,
+      "grad_norm": 0.7066627740859985,
+      "learning_rate": 3.988878115101024e-05,
+      "loss": 1.6193,
+      "step": 212
+    },
+    {
+      "epoch": 0.13127889060092449,
+      "grad_norm": 0.6514896750450134,
+      "learning_rate": 3.9884199714670914e-05,
+      "loss": 1.6017,
+      "step": 213
+    },
+    {
+      "epoch": 0.13189522342064716,
+      "grad_norm": 0.689854621887207,
+      "learning_rate": 3.987952608550928e-05,
+      "loss": 1.6162,
+      "step": 214
+    },
+    {
+      "epoch": 0.1325115562403698,
+      "grad_norm": 0.6585028171539307,
+      "learning_rate": 3.9874760285194554e-05,
+      "loss": 1.5866,
+      "step": 215
+    },
+    {
+      "epoch": 0.13312788906009246,
+      "grad_norm": 0.6556795239448547,
+      "learning_rate": 3.9869902335823296e-05,
+      "loss": 1.6076,
+      "step": 216
+    },
+    {
+      "epoch": 0.1337442218798151,
+      "grad_norm": 0.6920654773712158,
+      "learning_rate": 3.986495225991934e-05,
+      "loss": 1.593,
+      "step": 217
+    },
+    {
+      "epoch": 0.13436055469953775,
+      "grad_norm": 0.678615152835846,
+      "learning_rate": 3.985991008043363e-05,
+      "loss": 1.6143,
+      "step": 218
+    },
+    {
+      "epoch": 0.1349768875192604,
+      "grad_norm": 0.6753882169723511,
+      "learning_rate": 3.985477582074417e-05,
+      "loss": 1.6432,
+      "step": 219
+    },
+    {
+      "epoch": 0.13559322033898305,
+      "grad_norm": 0.6829013228416443,
+      "learning_rate": 3.984954950465589e-05,
+      "loss": 1.5916,
+      "step": 220
+    },
+    {
+      "epoch": 0.1362095531587057,
+      "grad_norm": 0.6789626479148865,
+      "learning_rate": 3.984423115640051e-05,
+      "loss": 1.5781,
+      "step": 221
+    },
+    {
+      "epoch": 0.13682588597842835,
+      "grad_norm": 0.7018295526504517,
+      "learning_rate": 3.983882080063652e-05,
+      "loss": 1.6833,
+      "step": 222
+    },
+    {
+      "epoch": 0.137442218798151,
+      "grad_norm": 0.6346052289009094,
+      "learning_rate": 3.983331846244892e-05,
+      "loss": 1.5468,
+      "step": 223
+    },
+    {
+      "epoch": 0.13805855161787364,
+      "grad_norm": 0.6694582104682922,
+      "learning_rate": 3.982772416734925e-05,
+      "loss": 1.5935,
+      "step": 224
+    },
+    {
+      "epoch": 0.1386748844375963,
+      "grad_norm": 0.6665736436843872,
+      "learning_rate": 3.982203794127538e-05,
+      "loss": 1.5285,
+      "step": 225
+    },
+    {
+      "epoch": 0.13929121725731894,
+      "grad_norm": 0.6625133156776428,
+      "learning_rate": 3.9816259810591416e-05,
+      "loss": 1.6179,
+      "step": 226
+    },
+    {
+      "epoch": 0.13990755007704161,
+      "grad_norm": 0.6552127599716187,
+      "learning_rate": 3.981038980208759e-05,
+      "loss": 1.6083,
+      "step": 227
+    },
+    {
+      "epoch": 0.14052388289676426,
+      "grad_norm": 0.6420478224754333,
+      "learning_rate": 3.98044279429801e-05,
+      "loss": 1.5898,
+      "step": 228
+    },
+    {
+      "epoch": 0.1411402157164869,
+      "grad_norm": 0.6572580337524414,
+      "learning_rate": 3.979837426091105e-05,
+      "loss": 1.6002,
+      "step": 229
+    },
+    {
+      "epoch": 0.14175654853620956,
+      "grad_norm": 0.6583701968193054,
+      "learning_rate": 3.979222878394823e-05,
+      "loss": 1.5608,
+      "step": 230
+    },
+    {
+      "epoch": 0.1423728813559322,
+      "grad_norm": 0.670501708984375,
+      "learning_rate": 3.978599154058508e-05,
+      "loss": 1.6096,
+      "step": 231
+    },
+    {
+      "epoch": 0.14298921417565486,
+      "grad_norm": 0.6815168857574463,
+      "learning_rate": 3.977966255974048e-05,
+      "loss": 1.6108,
+      "step": 232
+    },
+    {
+      "epoch": 0.1436055469953775,
+      "grad_norm": 0.6337635517120361,
+      "learning_rate": 3.9773241870758665e-05,
+      "loss": 1.6007,
+      "step": 233
+    },
+    {
+      "epoch": 0.14422187981510015,
+      "grad_norm": 0.692473292350769,
+      "learning_rate": 3.976672950340909e-05,
+      "loss": 1.633,
+      "step": 234
+    },
+    {
+      "epoch": 0.1448382126348228,
+      "grad_norm": 0.6315524578094482,
+      "learning_rate": 3.9760125487886235e-05,
+      "loss": 1.5258,
+      "step": 235
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 0.6339032053947449,
+      "learning_rate": 3.975342985480955e-05,
+      "loss": 1.561,
+      "step": 236
+    },
+    {
+      "epoch": 0.1460708782742681,
+      "grad_norm": 0.6732640862464905,
+      "learning_rate": 3.974664263522322e-05,
+      "loss": 1.6279,
+      "step": 237
+    },
+    {
+      "epoch": 0.14668721109399074,
+      "grad_norm": 0.6514915227890015,
+      "learning_rate": 3.973976386059612e-05,
+      "loss": 1.5986,
+      "step": 238
+    },
+    {
+      "epoch": 0.1473035439137134,
+      "grad_norm": 0.6529800295829773,
+      "learning_rate": 3.973279356282159e-05,
+      "loss": 1.5458,
+      "step": 239
+    },
+    {
+      "epoch": 0.14791987673343607,
+      "grad_norm": 0.6552094221115112,
+      "learning_rate": 3.9725731774217316e-05,
+      "loss": 1.5856,
+      "step": 240
+    },
+    {
+      "epoch": 0.14853620955315872,
+      "grad_norm": 0.6603219509124756,
+      "learning_rate": 3.971857852752518e-05,
+      "loss": 1.6069,
+      "step": 241
+    },
+    {
+      "epoch": 0.14915254237288136,
+      "grad_norm": 0.6710034608840942,
+      "learning_rate": 3.971133385591113e-05,
+      "loss": 1.6145,
+      "step": 242
+    },
+    {
+      "epoch": 0.149768875192604,
+      "grad_norm": 0.648938775062561,
+      "learning_rate": 3.970399779296498e-05,
+      "loss": 1.5991,
+      "step": 243
+    },
+    {
+      "epoch": 0.15038520801232666,
+      "grad_norm": 0.6794759035110474,
+      "learning_rate": 3.9696570372700284e-05,
+      "loss": 1.5898,
+      "step": 244
+    },
+    {
+      "epoch": 0.1510015408320493,
+      "grad_norm": 0.6564098596572876,
+      "learning_rate": 3.968905162955417e-05,
+      "loss": 1.6538,
+      "step": 245
+    },
+    {
+      "epoch": 0.15161787365177196,
+      "grad_norm": 0.6441242694854736,
+      "learning_rate": 3.968144159838719e-05,
+      "loss": 1.5993,
+      "step": 246
+    },
+    {
+      "epoch": 0.1522342064714946,
+      "grad_norm": 0.6316927671432495,
+      "learning_rate": 3.967374031448317e-05,
+      "loss": 1.5541,
+      "step": 247
+    },
+    {
+      "epoch": 0.15285053929121725,
+      "grad_norm": 0.6498556137084961,
+      "learning_rate": 3.966594781354899e-05,
+      "loss": 1.5799,
+      "step": 248
+    },
+    {
+      "epoch": 0.1534668721109399,
+      "grad_norm": 0.6671091914176941,
+      "learning_rate": 3.965806413171447e-05,
+      "loss": 1.5254,
+      "step": 249
+    },
+    {
+      "epoch": 0.15408320493066255,
+      "grad_norm": 0.6488400101661682,
+      "learning_rate": 3.965008930553221e-05,
+      "loss": 1.5769,
+      "step": 250
+    },
+    {
+      "epoch": 0.1546995377503852,
+      "grad_norm": 0.6346517205238342,
+      "learning_rate": 3.964202337197737e-05,
+      "loss": 1.5823,
+      "step": 251
+    },
+    {
+      "epoch": 0.15531587057010784,
+      "grad_norm": 0.6554057598114014,
+      "learning_rate": 3.963386636844755e-05,
+      "loss": 1.7108,
+      "step": 252
+    },
+    {
+      "epoch": 0.15593220338983052,
+      "grad_norm": 0.6496933102607727,
+      "learning_rate": 3.9625618332762583e-05,
+      "loss": 1.6616,
+      "step": 253
+    },
+    {
+      "epoch": 0.15654853620955317,
+      "grad_norm": 0.6630948185920715,
+      "learning_rate": 3.9617279303164374e-05,
+      "loss": 1.6548,
+      "step": 254
+    },
+    {
+      "epoch": 0.15716486902927582,
+      "grad_norm": 0.6255600452423096,
+      "learning_rate": 3.960884931831673e-05,
+      "loss": 1.5547,
+      "step": 255
+    },
+    {
+      "epoch": 0.15778120184899846,
+      "grad_norm": 0.6128103733062744,
+      "learning_rate": 3.960032841730514e-05,
+      "loss": 1.5376,
+      "step": 256
+    },
+    {
+      "epoch": 0.1583975346687211,
+      "grad_norm": 0.6571413278579712,
+      "learning_rate": 3.959171663963667e-05,
+      "loss": 1.6524,
+      "step": 257
+    },
+    {
+      "epoch": 0.15901386748844376,
+      "grad_norm": 0.6260244846343994,
+      "learning_rate": 3.958301402523969e-05,
+      "loss": 1.5668,
+      "step": 258
+    },
+    {
+      "epoch": 0.1596302003081664,
+      "grad_norm": 0.6612125039100647,
+      "learning_rate": 3.957422061446377e-05,
+      "loss": 1.6241,
+      "step": 259
+    },
+    {
+      "epoch": 0.16024653312788906,
+      "grad_norm": 0.6285637021064758,
+      "learning_rate": 3.956533644807943e-05,
+      "loss": 1.5283,
+      "step": 260
+    },
+    {
+      "epoch": 0.1608628659476117,
+      "grad_norm": 0.6467995047569275,
+      "learning_rate": 3.9556361567277995e-05,
+      "loss": 1.5803,
+      "step": 261
+    },
+    {
+      "epoch": 0.16147919876733435,
+      "grad_norm": 0.6415068507194519,
+      "learning_rate": 3.9547296013671365e-05,
+      "loss": 1.6484,
+      "step": 262
+    },
+    {
+      "epoch": 0.162095531587057,
+      "grad_norm": 0.6287012100219727,
+      "learning_rate": 3.953813982929187e-05,
+      "loss": 1.5626,
+      "step": 263
+    },
+    {
+      "epoch": 0.16271186440677965,
+      "grad_norm": 0.6504274606704712,
+      "learning_rate": 3.952889305659203e-05,
+      "loss": 1.5822,
+      "step": 264
+    },
+    {
+      "epoch": 0.1633281972265023,
+      "grad_norm": 0.6381985545158386,
+      "learning_rate": 3.951955573844437e-05,
+      "loss": 1.5835,
+      "step": 265
+    },
+    {
+      "epoch": 0.16394453004622497,
+      "grad_norm": 0.6651602387428284,
+      "learning_rate": 3.951012791814126e-05,
+      "loss": 1.6438,
+      "step": 266
+    },
+    {
+      "epoch": 0.16456086286594762,
+      "grad_norm": 0.6480642557144165,
+      "learning_rate": 3.950060963939465e-05,
+      "loss": 1.5801,
+      "step": 267
+    },
+    {
+      "epoch": 0.16517719568567027,
+      "grad_norm": 0.6354721188545227,
+      "learning_rate": 3.9491000946335904e-05,
+      "loss": 1.5894,
+      "step": 268
+    },
+    {
+      "epoch": 0.16579352850539292,
+      "grad_norm": 0.6333348751068115,
+      "learning_rate": 3.9481301883515605e-05,
+      "loss": 1.589,
+      "step": 269
+    },
+    {
+      "epoch": 0.16640986132511557,
+      "grad_norm": 0.6304234266281128,
+      "learning_rate": 3.947151249590333e-05,
+      "loss": 1.5163,
+      "step": 270
+    },
+    {
+      "epoch": 0.1670261941448382,
+      "grad_norm": 0.6365682482719421,
+      "learning_rate": 3.946163282888743e-05,
+      "loss": 1.5537,
+      "step": 271
+    },
+    {
+      "epoch": 0.16764252696456086,
+      "grad_norm": 0.6680180430412292,
+      "learning_rate": 3.945166292827486e-05,
+      "loss": 1.5785,
+      "step": 272
+    },
+    {
+      "epoch": 0.1682588597842835,
+      "grad_norm": 0.6543309092521667,
+      "learning_rate": 3.944160284029092e-05,
+      "loss": 1.6562,
+      "step": 273
+    },
+    {
+      "epoch": 0.16887519260400616,
+      "grad_norm": 0.6407747268676758,
+      "learning_rate": 3.943145261157908e-05,
+      "loss": 1.592,
+      "step": 274
+    },
+    {
+      "epoch": 0.1694915254237288,
+      "grad_norm": 0.6593438982963562,
+      "learning_rate": 3.942121228920072e-05,
+      "loss": 1.567,
+      "step": 275
+    },
+    {
+      "epoch": 0.17010785824345145,
+      "grad_norm": 0.6593472361564636,
+      "learning_rate": 3.9410881920634974e-05,
+      "loss": 1.6531,
+      "step": 276
+    },
+    {
+      "epoch": 0.1707241910631741,
+      "grad_norm": 0.6622447371482849,
+      "learning_rate": 3.9400461553778455e-05,
+      "loss": 1.5938,
+      "step": 277
+    },
+    {
+      "epoch": 0.17134052388289678,
+      "grad_norm": 0.6343011856079102,
+      "learning_rate": 3.938995123694504e-05,
+      "loss": 1.5481,
+      "step": 278
+    },
+    {
+      "epoch": 0.17195685670261943,
+      "grad_norm": 0.6643282771110535,
+      "learning_rate": 3.937935101886568e-05,
+      "loss": 1.5363,
+      "step": 279
+    },
+    {
+      "epoch": 0.17257318952234207,
+      "grad_norm": 0.6595264077186584,
+      "learning_rate": 3.9368660948688135e-05,
+      "loss": 1.5831,
+      "step": 280
+    },
+    {
+      "epoch": 0.17318952234206472,
+      "grad_norm": 0.6727797389030457,
+      "learning_rate": 3.935788107597678e-05,
+      "loss": 1.5718,
+      "step": 281
+    },
+    {
+      "epoch": 0.17380585516178737,
+      "grad_norm": 0.6271126866340637,
+      "learning_rate": 3.934701145071235e-05,
+      "loss": 1.6389,
+      "step": 282
+    },
+    {
+      "epoch": 0.17442218798151002,
+      "grad_norm": 0.6499379277229309,
+      "learning_rate": 3.93360521232917e-05,
+      "loss": 1.5553,
+      "step": 283
+    },
+    {
+      "epoch": 0.17503852080123267,
+      "grad_norm": 0.6241937875747681,
+      "learning_rate": 3.932500314452762e-05,
+      "loss": 1.5052,
+      "step": 284
+    },
+    {
+      "epoch": 0.17565485362095531,
+      "grad_norm": 0.6471588015556335,
+      "learning_rate": 3.931386456564854e-05,
+      "loss": 1.6057,
+      "step": 285
+    },
+    {
+      "epoch": 0.17627118644067796,
+      "grad_norm": 0.654523491859436,
+      "learning_rate": 3.9302636438298334e-05,
+      "loss": 1.6026,
+      "step": 286
+    },
+    {
+      "epoch": 0.1768875192604006,
+      "grad_norm": 0.6462265253067017,
+      "learning_rate": 3.929131881453606e-05,
+      "loss": 1.6083,
+      "step": 287
+    },
+    {
+      "epoch": 0.17750385208012326,
+      "grad_norm": 0.6588643789291382,
+      "learning_rate": 3.9279911746835726e-05,
+      "loss": 1.6134,
+      "step": 288
+    },
+    {
+      "epoch": 0.1781201848998459,
+      "grad_norm": 0.6225916147232056,
+      "learning_rate": 3.926841528808604e-05,
+      "loss": 1.5746,
+      "step": 289
+    },
+    {
+      "epoch": 0.17873651771956856,
+      "grad_norm": 0.647373616695404,
+      "learning_rate": 3.925682949159019e-05,
+      "loss": 1.6009,
+      "step": 290
+    },
+    {
+      "epoch": 0.17935285053929123,
+      "grad_norm": 0.635344922542572,
+      "learning_rate": 3.9245154411065566e-05,
+      "loss": 1.583,
+      "step": 291
+    },
+    {
+      "epoch": 0.17996918335901388,
+      "grad_norm": 0.6447535753250122,
+      "learning_rate": 3.9233390100643515e-05,
+      "loss": 1.6787,
+      "step": 292
+    },
+    {
+      "epoch": 0.18058551617873653,
+      "grad_norm": 0.6271030306816101,
+      "learning_rate": 3.92215366148691e-05,
+      "loss": 1.5895,
+      "step": 293
+    },
+    {
+      "epoch": 0.18120184899845918,
+      "grad_norm": 0.652759850025177,
+      "learning_rate": 3.9209594008700846e-05,
+      "loss": 1.6372,
+      "step": 294
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 0.6784529089927673,
+      "learning_rate": 3.9197562337510485e-05,
+      "loss": 1.6218,
+      "step": 295
+    },
+    {
+      "epoch": 0.18243451463790447,
+      "grad_norm": 0.637200117111206,
+      "learning_rate": 3.91854416570827e-05,
+      "loss": 1.5875,
+      "step": 296
+    },
+    {
+      "epoch": 0.18305084745762712,
+      "grad_norm": 0.63194340467453,
+      "learning_rate": 3.9173232023614866e-05,
+      "loss": 1.5507,
+      "step": 297
+    },
+    {
+      "epoch": 0.18366718027734977,
+      "grad_norm": 0.6260672211647034,
+      "learning_rate": 3.916093349371678e-05,
+      "loss": 1.5402,
+      "step": 298
+    },
+    {
+      "epoch": 0.18428351309707242,
+      "grad_norm": 0.6279014348983765,
+      "learning_rate": 3.91485461244104e-05,
+      "loss": 1.5537,
+      "step": 299
+    },
+    {
+      "epoch": 0.18489984591679506,
+      "grad_norm": 0.6495095491409302,
+      "learning_rate": 3.913606997312961e-05,
+      "loss": 1.6009,
+      "step": 300
+    },
+    {
+      "epoch": 0.1855161787365177,
+      "grad_norm": 0.6379627585411072,
+      "learning_rate": 3.9123505097719914e-05,
+      "loss": 1.6117,
+      "step": 301
+    },
+    {
+      "epoch": 0.18613251155624036,
+      "grad_norm": 0.6391900181770325,
+      "learning_rate": 3.9110851556438174e-05,
+      "loss": 1.6022,
+      "step": 302
+    },
+    {
+      "epoch": 0.186748844375963,
+      "grad_norm": 0.6260542273521423,
+      "learning_rate": 3.909810940795237e-05,
+      "loss": 1.5987,
+      "step": 303
+    },
+    {
+      "epoch": 0.18736517719568568,
+      "grad_norm": 0.6432803273200989,
+      "learning_rate": 3.90852787113413e-05,
+      "loss": 1.6286,
+      "step": 304
+    },
+    {
+      "epoch": 0.18798151001540833,
+      "grad_norm": 0.6564033031463623,
+      "learning_rate": 3.9072359526094315e-05,
+      "loss": 1.5856,
+      "step": 305
+    },
+    {
+      "epoch": 0.18859784283513098,
+      "grad_norm": 0.6385253071784973,
+      "learning_rate": 3.9059351912111044e-05,
+      "loss": 1.5412,
+      "step": 306
+    },
+    {
+      "epoch": 0.18921417565485363,
+      "grad_norm": 0.6348027586936951,
+      "learning_rate": 3.9046255929701116e-05,
+      "loss": 1.6031,
+      "step": 307
+    },
+    {
+      "epoch": 0.18983050847457628,
+      "grad_norm": 0.6338834762573242,
+      "learning_rate": 3.903307163958386e-05,
+      "loss": 1.5937,
+      "step": 308
+    },
+    {
+      "epoch": 0.19044684129429892,
+      "grad_norm": 0.6354376077651978,
+      "learning_rate": 3.901979910288809e-05,
+      "loss": 1.5642,
+      "step": 309
+    },
+    {
+      "epoch": 0.19106317411402157,
+      "grad_norm": 0.6317285299301147,
+      "learning_rate": 3.900643838115171e-05,
+      "loss": 1.5423,
+      "step": 310
+    },
+    {
+      "epoch": 0.19167950693374422,
+      "grad_norm": 0.6343933343887329,
+      "learning_rate": 3.899298953632154e-05,
+      "loss": 1.5861,
+      "step": 311
+    },
+    {
+      "epoch": 0.19229583975346687,
+      "grad_norm": 0.6410644054412842,
+      "learning_rate": 3.897945263075296e-05,
+      "loss": 1.5917,
+      "step": 312
+    },
+    {
+      "epoch": 0.19291217257318952,
+      "grad_norm": 0.6140625476837158,
+      "learning_rate": 3.896582772720966e-05,
+      "loss": 1.5459,
+      "step": 313
+    },
+    {
+      "epoch": 0.19352850539291216,
+      "grad_norm": 0.6385317444801331,
+      "learning_rate": 3.8952114888863304e-05,
+      "loss": 1.5529,
+      "step": 314
+    },
+    {
+      "epoch": 0.1941448382126348,
+      "grad_norm": 0.6224085688591003,
+      "learning_rate": 3.8938314179293305e-05,
+      "loss": 1.5449,
+      "step": 315
+    },
+    {
+      "epoch": 0.19476117103235746,
+      "grad_norm": 0.6194934248924255,
+      "learning_rate": 3.892442566248644e-05,
+      "loss": 1.5461,
+      "step": 316
+    },
+    {
+      "epoch": 0.19537750385208014,
+      "grad_norm": 0.624581515789032,
+      "learning_rate": 3.8910449402836645e-05,
+      "loss": 1.5707,
+      "step": 317
+    },
+    {
+      "epoch": 0.19599383667180278,
+      "grad_norm": 0.6206932663917542,
+      "learning_rate": 3.889638546514465e-05,
+      "loss": 1.5833,
+      "step": 318
+    },
+    {
+      "epoch": 0.19661016949152543,
+      "grad_norm": 0.636967122554779,
+      "learning_rate": 3.8882233914617715e-05,
+      "loss": 1.6226,
+      "step": 319
+    },
+    {
+      "epoch": 0.19722650231124808,
+      "grad_norm": 0.6057730913162231,
+      "learning_rate": 3.8867994816869316e-05,
+      "loss": 1.5537,
+      "step": 320
+    },
+    {
+      "epoch": 0.19784283513097073,
+      "grad_norm": 0.6264992356300354,
+      "learning_rate": 3.885366823791883e-05,
+      "loss": 1.5785,
+      "step": 321
+    },
+    {
+      "epoch": 0.19845916795069338,
+      "grad_norm": 0.6459900736808777,
+      "learning_rate": 3.883925424419126e-05,
+      "loss": 1.6063,
+      "step": 322
+    },
+    {
+      "epoch": 0.19907550077041603,
+      "grad_norm": 0.6335688829421997,
+      "learning_rate": 3.882475290251686e-05,
+      "loss": 1.5388,
+      "step": 323
+    },
+    {
+      "epoch": 0.19969183359013867,
+      "grad_norm": 0.6406273245811462,
+      "learning_rate": 3.8810164280130944e-05,
+      "loss": 1.5848,
+      "step": 324
+    },
+    {
+      "epoch": 0.20030816640986132,
+      "grad_norm": 0.6211891770362854,
+      "learning_rate": 3.879548844467343e-05,
+      "loss": 1.4908,
+      "step": 325
+    },
+    {
+      "epoch": 0.20092449922958397,
+      "grad_norm": 0.666120707988739,
+      "learning_rate": 3.878072546418865e-05,
+      "loss": 1.5351,
+      "step": 326
+    },
+    {
+      "epoch": 0.20154083204930662,
+      "grad_norm": 0.6317727565765381,
+      "learning_rate": 3.8765875407124956e-05,
+      "loss": 1.5363,
+      "step": 327
+    },
+    {
+      "epoch": 0.20215716486902927,
+      "grad_norm": 0.6246560215950012,
+      "learning_rate": 3.8750938342334436e-05,
+      "loss": 1.5637,
+      "step": 328
+    },
+    {
+      "epoch": 0.2027734976887519,
+      "grad_norm": 0.6172101497650146,
+      "learning_rate": 3.873591433907259e-05,
+      "loss": 1.5253,
+      "step": 329
+    },
+    {
+      "epoch": 0.2033898305084746,
+      "grad_norm": 0.6300083994865417,
+      "learning_rate": 3.8720803466998e-05,
+      "loss": 1.5946,
+      "step": 330
+    },
+    {
+      "epoch": 0.20400616332819724,
+      "grad_norm": 0.6064188480377197,
+      "learning_rate": 3.8705605796172025e-05,
+      "loss": 1.5502,
+      "step": 331
+    },
+    {
+      "epoch": 0.20462249614791989,
+      "grad_norm": 0.6302444338798523,
+      "learning_rate": 3.869032139705846e-05,
+      "loss": 1.5739,
+      "step": 332
+    },
+    {
+      "epoch": 0.20523882896764253,
+      "grad_norm": 0.6242470741271973,
+      "learning_rate": 3.8674950340523204e-05,
+      "loss": 1.556,
+      "step": 333
+    },
+    {
+      "epoch": 0.20585516178736518,
+      "grad_norm": 0.6166197061538696,
+      "learning_rate": 3.8659492697833965e-05,
+      "loss": 1.5887,
+      "step": 334
+    },
+    {
+      "epoch": 0.20647149460708783,
+      "grad_norm": 0.6221072673797607,
+      "learning_rate": 3.864394854065989e-05,
+      "loss": 1.5845,
+      "step": 335
+    },
+    {
+      "epoch": 0.20708782742681048,
+      "grad_norm": 0.6159368753433228,
+      "learning_rate": 3.8628317941071254e-05,
+      "loss": 1.5656,
+      "step": 336
+    },
+    {
+      "epoch": 0.20770416024653313,
+      "grad_norm": 0.6356878280639648,
+      "learning_rate": 3.861260097153912e-05,
+      "loss": 1.5358,
+      "step": 337
+    },
+    {
+      "epoch": 0.20832049306625577,
+      "grad_norm": 0.6479353904724121,
+      "learning_rate": 3.859679770493499e-05,
+      "loss": 1.5515,
+      "step": 338
+    },
+    {
+      "epoch": 0.20893682588597842,
+      "grad_norm": 0.6261966824531555,
+      "learning_rate": 3.858090821453052e-05,
+      "loss": 1.5361,
+      "step": 339
+    },
+    {
+      "epoch": 0.20955315870570107,
+      "grad_norm": 0.6239880323410034,
+      "learning_rate": 3.8564932573997096e-05,
+      "loss": 1.6195,
+      "step": 340
+    },
+    {
+      "epoch": 0.21016949152542372,
+      "grad_norm": 0.6246241927146912,
+      "learning_rate": 3.854887085740557e-05,
+      "loss": 1.5846,
+      "step": 341
+    },
+    {
+      "epoch": 0.21078582434514637,
+      "grad_norm": 0.6406985521316528,
+      "learning_rate": 3.853272313922587e-05,
+      "loss": 1.5648,
+      "step": 342
+    },
+    {
+      "epoch": 0.21140215716486904,
+      "grad_norm": 0.6283363699913025,
+      "learning_rate": 3.851648949432667e-05,
+      "loss": 1.5856,
+      "step": 343
+    },
+    {
+      "epoch": 0.2120184899845917,
+      "grad_norm": 0.6151086091995239,
+      "learning_rate": 3.8500169997975055e-05,
+      "loss": 1.5237,
+      "step": 344
+    },
+    {
+      "epoch": 0.21263482280431434,
+      "grad_norm": 0.6174421310424805,
+      "learning_rate": 3.848376472583613e-05,
+      "loss": 1.5409,
+      "step": 345
+    },
+    {
+      "epoch": 0.213251155624037,
+      "grad_norm": 0.6334297060966492,
+      "learning_rate": 3.8467273753972734e-05,
+      "loss": 1.5677,
+      "step": 346
+    },
+    {
+      "epoch": 0.21386748844375963,
+      "grad_norm": 0.6260072588920593,
+      "learning_rate": 3.845069715884502e-05,
+      "loss": 1.5605,
+      "step": 347
+    },
+    {
+      "epoch": 0.21448382126348228,
+      "grad_norm": 0.6221638321876526,
+      "learning_rate": 3.843403501731016e-05,
+      "loss": 1.5543,
+      "step": 348
+    },
+    {
+      "epoch": 0.21510015408320493,
+      "grad_norm": 0.6135300993919373,
+      "learning_rate": 3.841728740662194e-05,
+      "loss": 1.499,
+      "step": 349
+    },
+    {
+      "epoch": 0.21571648690292758,
+      "grad_norm": 0.6266738772392273,
+      "learning_rate": 3.840045440443043e-05,
+      "loss": 1.5937,
+      "step": 350
+    },
+    {
+      "epoch": 0.21633281972265023,
+      "grad_norm": 0.6476675271987915,
+      "learning_rate": 3.8383536088781624e-05,
+      "loss": 1.5651,
+      "step": 351
+    },
+    {
+      "epoch": 0.21694915254237288,
+      "grad_norm": 0.6242912411689758,
+      "learning_rate": 3.836653253811704e-05,
+      "loss": 1.6536,
+      "step": 352
+    },
+    {
+      "epoch": 0.21756548536209552,
+      "grad_norm": 0.6051536202430725,
+      "learning_rate": 3.834944383127344e-05,
+      "loss": 1.4944,
+      "step": 353
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 0.6046092510223389,
+      "learning_rate": 3.833227004748237e-05,
+      "loss": 1.5516,
+      "step": 354
+    },
+    {
+      "epoch": 0.21879815100154082,
+      "grad_norm": 0.6210584044456482,
+      "learning_rate": 3.831501126636983e-05,
+      "loss": 1.6006,
+      "step": 355
+    },
+    {
+      "epoch": 0.2194144838212635,
+      "grad_norm": 0.6627335548400879,
+      "learning_rate": 3.829766756795595e-05,
+      "loss": 1.6129,
+      "step": 356
+    },
+    {
+      "epoch": 0.22003081664098614,
+      "grad_norm": 0.6231202483177185,
+      "learning_rate": 3.8280239032654536e-05,
+      "loss": 1.5539,
+      "step": 357
+    },
+    {
+      "epoch": 0.2206471494607088,
+      "grad_norm": 0.6369742751121521,
+      "learning_rate": 3.826272574127276e-05,
+      "loss": 1.5917,
+      "step": 358
+    },
+    {
+      "epoch": 0.22126348228043144,
+      "grad_norm": 0.6062166094779968,
+      "learning_rate": 3.8245127775010764e-05,
+      "loss": 1.5284,
+      "step": 359
+    },
+    {
+      "epoch": 0.2218798151001541,
+      "grad_norm": 0.6495872139930725,
+      "learning_rate": 3.822744521546127e-05,
+      "loss": 1.5648,
+      "step": 360
+    },
+    {
+      "epoch": 0.22249614791987674,
+      "grad_norm": 0.6449238061904907,
+      "learning_rate": 3.820967814460923e-05,
+      "loss": 1.5595,
+      "step": 361
+    },
+    {
+      "epoch": 0.22311248073959938,
+      "grad_norm": 0.6381934881210327,
+      "learning_rate": 3.819182664483144e-05,
+      "loss": 1.5733,
+      "step": 362
+    },
+    {
+      "epoch": 0.22372881355932203,
+      "grad_norm": 0.6340575218200684,
+      "learning_rate": 3.817389079889612e-05,
+      "loss": 1.5488,
+      "step": 363
+    },
+    {
+      "epoch": 0.22434514637904468,
+      "grad_norm": 0.627213716506958,
+      "learning_rate": 3.815587068996258e-05,
+      "loss": 1.5744,
+      "step": 364
+    },
+    {
+      "epoch": 0.22496147919876733,
+      "grad_norm": 0.6066584587097168,
+      "learning_rate": 3.8137766401580814e-05,
+      "loss": 1.5482,
+      "step": 365
+    },
+    {
+      "epoch": 0.22557781201848998,
+      "grad_norm": 0.6272307634353638,
+      "learning_rate": 3.811957801769111e-05,
+      "loss": 1.5628,
+      "step": 366
+    },
+    {
+      "epoch": 0.22619414483821262,
+      "grad_norm": 0.6388189196586609,
+      "learning_rate": 3.810130562262366e-05,
+      "loss": 1.5816,
+      "step": 367
+    },
+    {
+      "epoch": 0.22681047765793527,
+      "grad_norm": 0.613997220993042,
+      "learning_rate": 3.808294930109818e-05,
+      "loss": 1.5346,
+      "step": 368
+    },
+    {
+      "epoch": 0.22742681047765795,
+      "grad_norm": 0.623281717300415,
+      "learning_rate": 3.80645091382235e-05,
+      "loss": 1.5748,
+      "step": 369
+    },
+    {
+      "epoch": 0.2280431432973806,
+      "grad_norm": 0.6061915755271912,
+      "learning_rate": 3.804598521949718e-05,
+      "loss": 1.539,
+      "step": 370
+    },
+    {
+      "epoch": 0.22865947611710324,
+      "grad_norm": 0.6243897080421448,
+      "learning_rate": 3.802737763080513e-05,
+      "loss": 1.5695,
+      "step": 371
+    },
+    {
+      "epoch": 0.2292758089368259,
+      "grad_norm": 0.6500725150108337,
+      "learning_rate": 3.800868645842118e-05,
+      "loss": 1.6556,
+      "step": 372
+    },
+    {
+      "epoch": 0.22989214175654854,
+      "grad_norm": 0.6233870983123779,
+      "learning_rate": 3.798991178900671e-05,
+      "loss": 1.5392,
+      "step": 373
+    },
+    {
+      "epoch": 0.2305084745762712,
+      "grad_norm": 0.6300720572471619,
+      "learning_rate": 3.797105370961019e-05,
+      "loss": 1.5576,
+      "step": 374
+    },
+    {
+      "epoch": 0.23112480739599384,
+      "grad_norm": 0.6239476203918457,
+      "learning_rate": 3.795211230766687e-05,
+      "loss": 1.5261,
+      "step": 375
+    },
+    {
+      "epoch": 0.23174114021571648,
+      "grad_norm": 0.6265757083892822,
+      "learning_rate": 3.793308767099829e-05,
+      "loss": 1.5849,
+      "step": 376
+    },
+    {
+      "epoch": 0.23235747303543913,
+      "grad_norm": 0.6283276677131653,
+      "learning_rate": 3.791397988781194e-05,
+      "loss": 1.5687,
+      "step": 377
+    },
+    {
+      "epoch": 0.23297380585516178,
+      "grad_norm": 0.624076247215271,
+      "learning_rate": 3.789478904670078e-05,
+      "loss": 1.5565,
+      "step": 378
+    },
+    {
+      "epoch": 0.23359013867488443,
+      "grad_norm": 0.6197231411933899,
+      "learning_rate": 3.7875515236642894e-05,
+      "loss": 1.5715,
+      "step": 379
+    },
+    {
+      "epoch": 0.23420647149460708,
+      "grad_norm": 0.6188759207725525,
+      "learning_rate": 3.785615854700104e-05,
+      "loss": 1.5728,
+      "step": 380
+    },
+    {
+      "epoch": 0.23482280431432973,
+      "grad_norm": 0.6138371229171753,
+      "learning_rate": 3.7836719067522236e-05,
+      "loss": 1.5571,
+      "step": 381
+    },
+    {
+      "epoch": 0.2354391371340524,
+      "grad_norm": 0.6327232718467712,
+      "learning_rate": 3.781719688833739e-05,
+      "loss": 1.5941,
+      "step": 382
+    },
+    {
+      "epoch": 0.23605546995377505,
+      "grad_norm": 0.6079659461975098,
+      "learning_rate": 3.779759209996079e-05,
+      "loss": 1.5373,
+      "step": 383
+    },
+    {
+      "epoch": 0.2366718027734977,
+      "grad_norm": 0.6112452745437622,
+      "learning_rate": 3.777790479328979e-05,
+      "loss": 1.5547,
+      "step": 384
+    },
+    {
+      "epoch": 0.23728813559322035,
+      "grad_norm": 0.6212001442909241,
+      "learning_rate": 3.7758135059604315e-05,
+      "loss": 1.537,
+      "step": 385
+    },
+    {
+      "epoch": 0.237904468412943,
+      "grad_norm": 0.6198443174362183,
+      "learning_rate": 3.773828299056647e-05,
+      "loss": 1.532,
+      "step": 386
+    },
+    {
+      "epoch": 0.23852080123266564,
+      "grad_norm": 0.6353772282600403,
+      "learning_rate": 3.77183486782201e-05,
+      "loss": 1.5347,
+      "step": 387
+    },
+    {
+      "epoch": 0.2391371340523883,
+      "grad_norm": 0.6267209053039551,
+      "learning_rate": 3.7698332214990385e-05,
+      "loss": 1.527,
+      "step": 388
+    },
+    {
+      "epoch": 0.23975346687211094,
+      "grad_norm": 0.634272575378418,
+      "learning_rate": 3.767823369368336e-05,
+      "loss": 1.5845,
+      "step": 389
+    },
+    {
+      "epoch": 0.24036979969183359,
+      "grad_norm": 0.6304289102554321,
+      "learning_rate": 3.765805320748558e-05,
+      "loss": 1.5901,
+      "step": 390
+    },
+    {
+      "epoch": 0.24098613251155623,
+      "grad_norm": 0.6337018013000488,
+      "learning_rate": 3.763779084996357e-05,
+      "loss": 1.5677,
+      "step": 391
+    },
+    {
+      "epoch": 0.24160246533127888,
+      "grad_norm": 0.6266562342643738,
+      "learning_rate": 3.761744671506349e-05,
+      "loss": 1.6029,
+      "step": 392
+    },
+    {
+      "epoch": 0.24221879815100153,
+      "grad_norm": 0.6246030926704407,
+      "learning_rate": 3.7597020897110634e-05,
+      "loss": 1.586,
+      "step": 393
+    },
+    {
+      "epoch": 0.24283513097072418,
+      "grad_norm": 0.6081348657608032,
+      "learning_rate": 3.7576513490809044e-05,
+      "loss": 1.5734,
+      "step": 394
+    },
+    {
+      "epoch": 0.24345146379044685,
+      "grad_norm": 0.6142352819442749,
+      "learning_rate": 3.7555924591241015e-05,
+      "loss": 1.5099,
+      "step": 395
+    },
+    {
+      "epoch": 0.2440677966101695,
+      "grad_norm": 0.6177673935890198,
+      "learning_rate": 3.753525429386671e-05,
+      "loss": 1.5074,
+      "step": 396
+    },
+    {
+      "epoch": 0.24468412942989215,
+      "grad_norm": 0.6194700002670288,
+      "learning_rate": 3.751450269452368e-05,
+      "loss": 1.5212,
+      "step": 397
+    },
+    {
+      "epoch": 0.2453004622496148,
+      "grad_norm": 0.6167085766792297,
+      "learning_rate": 3.749366988942645e-05,
+      "loss": 1.6025,
+      "step": 398
+    },
+    {
+      "epoch": 0.24591679506933745,
+      "grad_norm": 0.6241602897644043,
+      "learning_rate": 3.747275597516601e-05,
+      "loss": 1.5204,
+      "step": 399
+    },
+    {
+      "epoch": 0.2465331278890601,
+      "grad_norm": 0.6154288053512573,
+      "learning_rate": 3.745176104870945e-05,
+      "loss": 1.5505,
+      "step": 400
+    },
+    {
+      "epoch": 0.24714946070878274,
+      "grad_norm": 0.6391974091529846,
+      "learning_rate": 3.743068520739947e-05,
+      "loss": 1.5339,
+      "step": 401
+    },
+    {
+      "epoch": 0.2477657935285054,
+      "grad_norm": 0.6121848225593567,
+      "learning_rate": 3.740952854895392e-05,
+      "loss": 1.4906,
+      "step": 402
+    },
+    {
+      "epoch": 0.24838212634822804,
+      "grad_norm": 0.6097393035888672,
+      "learning_rate": 3.738829117146535e-05,
+      "loss": 1.4699,
+      "step": 403
+    },
+    {
+      "epoch": 0.2489984591679507,
+      "grad_norm": 0.6633890271186829,
+      "learning_rate": 3.736697317340059e-05,
+      "loss": 1.5798,
+      "step": 404
+    },
+    {
+      "epoch": 0.24961479198767333,
+      "grad_norm": 0.6471389532089233,
+      "learning_rate": 3.7345574653600234e-05,
+      "loss": 1.4527,
+      "step": 405
+    },
+    {
+      "epoch": 0.250231124807396,
+      "grad_norm": 0.7222802639007568,
+      "learning_rate": 3.7324095711278235e-05,
+      "loss": 1.5641,
+      "step": 406
+    },
+    {
+      "epoch": 0.25084745762711863,
+      "grad_norm": 0.6318141222000122,
+      "learning_rate": 3.730253644602142e-05,
+      "loss": 1.5671,
+      "step": 407
+    },
+    {
+      "epoch": 0.2514637904468413,
+      "grad_norm": 0.6795703172683716,
+      "learning_rate": 3.7280896957789024e-05,
+      "loss": 1.5776,
+      "step": 408
+    },
+    {
+      "epoch": 0.2520801232665639,
+      "grad_norm": 0.6108106374740601,
+      "learning_rate": 3.725917734691224e-05,
+      "loss": 1.5307,
+      "step": 409
+    },
+    {
+      "epoch": 0.2526964560862866,
+      "grad_norm": 0.6504859328269958,
+      "learning_rate": 3.7237377714093756e-05,
+      "loss": 1.5231,
+      "step": 410
+    },
+    {
+      "epoch": 0.2533127889060092,
+      "grad_norm": 0.6572467684745789,
+      "learning_rate": 3.7215498160407275e-05,
+      "loss": 1.5766,
+      "step": 411
+    },
+    {
+      "epoch": 0.2539291217257319,
+      "grad_norm": 0.6699705123901367,
+      "learning_rate": 3.719353878729704e-05,
+      "loss": 1.6119,
+      "step": 412
+    },
+    {
+      "epoch": 0.2545454545454545,
+      "grad_norm": 0.6572472453117371,
+      "learning_rate": 3.717149969657739e-05,
+      "loss": 1.5487,
+      "step": 413
+    },
+    {
+      "epoch": 0.2551617873651772,
+      "grad_norm": 0.6471747756004333,
+      "learning_rate": 3.7149380990432284e-05,
+      "loss": 1.5511,
+      "step": 414
+    },
+    {
+      "epoch": 0.25577812018489987,
+      "grad_norm": 0.6392568945884705,
+      "learning_rate": 3.712718277141479e-05,
+      "loss": 1.5491,
+      "step": 415
+    },
+    {
+      "epoch": 0.2563944530046225,
+      "grad_norm": 0.6324256658554077,
+      "learning_rate": 3.710490514244664e-05,
+      "loss": 1.5234,
+      "step": 416
+    },
+    {
+      "epoch": 0.25701078582434517,
+      "grad_norm": 0.6545376181602478,
+      "learning_rate": 3.708254820681778e-05,
+      "loss": 1.5703,
+      "step": 417
+    },
+    {
+      "epoch": 0.2576271186440678,
+      "grad_norm": 0.6385619044303894,
+      "learning_rate": 3.7060112068185834e-05,
+      "loss": 1.5831,
+      "step": 418
+    },
+    {
+      "epoch": 0.25824345146379046,
+      "grad_norm": 0.6294272541999817,
+      "learning_rate": 3.703759683057564e-05,
+      "loss": 1.5387,
+      "step": 419
+    },
+    {
+      "epoch": 0.2588597842835131,
+      "grad_norm": 0.6257953643798828,
+      "learning_rate": 3.7015002598378805e-05,
+      "loss": 1.4815,
+      "step": 420
+    },
+    {
+      "epoch": 0.25947611710323576,
+      "grad_norm": 0.6191906332969666,
+      "learning_rate": 3.6992329476353184e-05,
+      "loss": 1.5048,
+      "step": 421
+    },
+    {
+      "epoch": 0.2600924499229584,
+      "grad_norm": 0.6304193139076233,
+      "learning_rate": 3.696957756962238e-05,
+      "loss": 1.4664,
+      "step": 422
+    },
+    {
+      "epoch": 0.26070878274268106,
+      "grad_norm": 0.6742473244667053,
+      "learning_rate": 3.694674698367532e-05,
+      "loss": 1.581,
+      "step": 423
+    },
+    {
+      "epoch": 0.2613251155624037,
+      "grad_norm": 0.6408929824829102,
+      "learning_rate": 3.6923837824365704e-05,
+      "loss": 1.579,
+      "step": 424
+    },
+    {
+      "epoch": 0.26194144838212635,
+      "grad_norm": 0.6516648530960083,
+      "learning_rate": 3.6900850197911526e-05,
+      "loss": 1.5845,
+      "step": 425
+    },
+    {
+      "epoch": 0.26255778120184897,
+      "grad_norm": 0.6402381658554077,
+      "learning_rate": 3.687778421089461e-05,
+      "loss": 1.6036,
+      "step": 426
+    },
+    {
+      "epoch": 0.26317411402157165,
+      "grad_norm": 0.6402097344398499,
+      "learning_rate": 3.68546399702601e-05,
+      "loss": 1.5358,
+      "step": 427
+    },
+    {
+      "epoch": 0.2637904468412943,
+      "grad_norm": 0.6428090929985046,
+      "learning_rate": 3.683141758331594e-05,
+      "loss": 1.5649,
+      "step": 428
+    },
+    {
+      "epoch": 0.26440677966101694,
+      "grad_norm": 0.6168109774589539,
+      "learning_rate": 3.680811715773243e-05,
+      "loss": 1.4968,
+      "step": 429
+    },
+    {
+      "epoch": 0.2650231124807396,
+      "grad_norm": 0.6354753971099854,
+      "learning_rate": 3.678473880154166e-05,
+      "loss": 1.5276,
+      "step": 430
+    },
+    {
+      "epoch": 0.26563944530046224,
+      "grad_norm": 0.6141413450241089,
+      "learning_rate": 3.676128262313707e-05,
+      "loss": 1.5411,
+      "step": 431
+    },
+    {
+      "epoch": 0.2662557781201849,
+      "grad_norm": 0.6401364207267761,
+      "learning_rate": 3.673774873127291e-05,
+      "loss": 1.6012,
+      "step": 432
+    },
+    {
+      "epoch": 0.26687211093990754,
+      "grad_norm": 0.6499303579330444,
+      "learning_rate": 3.671413723506375e-05,
+      "loss": 1.5878,
+      "step": 433
+    },
+    {
+      "epoch": 0.2674884437596302,
+      "grad_norm": 0.626645565032959,
+      "learning_rate": 3.669044824398397e-05,
+      "loss": 1.53,
+      "step": 434
+    },
+    {
+      "epoch": 0.26810477657935283,
+      "grad_norm": 0.6259933114051819,
+      "learning_rate": 3.666668186786726e-05,
+      "loss": 1.5172,
+      "step": 435
+    },
+    {
+      "epoch": 0.2687211093990755,
+      "grad_norm": 0.6535432934761047,
+      "learning_rate": 3.664283821690611e-05,
+      "loss": 1.5639,
+      "step": 436
+    },
+    {
+      "epoch": 0.26933744221879813,
+      "grad_norm": 0.627822756767273,
+      "learning_rate": 3.661891740165127e-05,
+      "loss": 1.6058,
+      "step": 437
+    },
+    {
+      "epoch": 0.2699537750385208,
+      "grad_norm": 0.6237924695014954,
+      "learning_rate": 3.6594919533011274e-05,
+      "loss": 1.5589,
+      "step": 438
+    },
+    {
+      "epoch": 0.2705701078582434,
+      "grad_norm": 0.6268482804298401,
+      "learning_rate": 3.657084472225193e-05,
+      "loss": 1.5538,
+      "step": 439
+    },
+    {
+      "epoch": 0.2711864406779661,
+      "grad_norm": 0.6410224437713623,
+      "learning_rate": 3.654669308099575e-05,
+      "loss": 1.5947,
+      "step": 440
+    },
+    {
+      "epoch": 0.2718027734976888,
+      "grad_norm": 0.6285886168479919,
+      "learning_rate": 3.6522464721221513e-05,
+      "loss": 1.5114,
+      "step": 441
+    },
+    {
+      "epoch": 0.2724191063174114,
+      "grad_norm": 0.6456573605537415,
+      "learning_rate": 3.6498159755263667e-05,
+      "loss": 1.5966,
+      "step": 442
+    },
+    {
+      "epoch": 0.2730354391371341,
+      "grad_norm": 0.6269593834877014,
+      "learning_rate": 3.647377829581186e-05,
+      "loss": 1.6099,
+      "step": 443
+    },
+    {
+      "epoch": 0.2736517719568567,
+      "grad_norm": 0.63033127784729,
+      "learning_rate": 3.64493204559104e-05,
+      "loss": 1.5018,
+      "step": 444
+    },
+    {
+      "epoch": 0.27426810477657937,
+      "grad_norm": 0.6265948414802551,
+      "learning_rate": 3.6424786348957726e-05,
+      "loss": 1.5309,
+      "step": 445
+    },
+    {
+      "epoch": 0.274884437596302,
+      "grad_norm": 0.6247050166130066,
+      "learning_rate": 3.6400176088705903e-05,
+      "loss": 1.4771,
+      "step": 446
+    },
+    {
+      "epoch": 0.27550077041602467,
+      "grad_norm": 0.6381626725196838,
+      "learning_rate": 3.637548978926005e-05,
+      "loss": 1.5488,
+      "step": 447
+    },
+    {
+      "epoch": 0.2761171032357473,
+      "grad_norm": 0.6295813918113708,
+      "learning_rate": 3.6350727565077876e-05,
+      "loss": 1.556,
+      "step": 448
+    },
+    {
+      "epoch": 0.27673343605546996,
+      "grad_norm": 0.6318914294242859,
+      "learning_rate": 3.6325889530969095e-05,
+      "loss": 1.52,
+      "step": 449
+    },
+    {
+      "epoch": 0.2773497688751926,
+      "grad_norm": 0.622588574886322,
+      "learning_rate": 3.6300975802094904e-05,
+      "loss": 1.5562,
+      "step": 450
+    },
+    {
+      "epoch": 0.27796610169491526,
+      "grad_norm": 0.6113806962966919,
+      "learning_rate": 3.627598649396748e-05,
+      "loss": 1.5129,
+      "step": 451
+    },
+    {
+      "epoch": 0.2785824345146379,
+      "grad_norm": 0.628746747970581,
+      "learning_rate": 3.6250921722449417e-05,
+      "loss": 1.5225,
+      "step": 452
+    },
+    {
+      "epoch": 0.27919876733436055,
+      "grad_norm": 0.642052948474884,
+      "learning_rate": 3.6225781603753175e-05,
+      "loss": 1.5206,
+      "step": 453
+    },
+    {
+      "epoch": 0.27981510015408323,
+      "grad_norm": 0.6341302394866943,
+      "learning_rate": 3.620056625444057e-05,
+      "loss": 1.5517,
+      "step": 454
+    },
+    {
+      "epoch": 0.28043143297380585,
+      "grad_norm": 0.631202757358551,
+      "learning_rate": 3.617527579142224e-05,
+      "loss": 1.5652,
+      "step": 455
+    },
+    {
+      "epoch": 0.2810477657935285,
+      "grad_norm": 0.6409715414047241,
+      "learning_rate": 3.6149910331957076e-05,
+      "loss": 1.5595,
+      "step": 456
+    },
+    {
+      "epoch": 0.28166409861325115,
+      "grad_norm": 0.6304307579994202,
+      "learning_rate": 3.612446999365167e-05,
+      "loss": 1.5613,
+      "step": 457
+    },
+    {
+      "epoch": 0.2822804314329738,
+      "grad_norm": 0.6630215644836426,
+      "learning_rate": 3.609895489445983e-05,
+      "loss": 1.5297,
+      "step": 458
+    },
+    {
+      "epoch": 0.28289676425269644,
+      "grad_norm": 0.6226741075515747,
+      "learning_rate": 3.607336515268194e-05,
+      "loss": 1.4715,
+      "step": 459
+    },
+    {
+      "epoch": 0.2835130970724191,
+      "grad_norm": 0.64727783203125,
+      "learning_rate": 3.60477008869645e-05,
+      "loss": 1.6325,
+      "step": 460
+    },
+    {
+      "epoch": 0.28412942989214174,
+      "grad_norm": 0.6112896203994751,
+      "learning_rate": 3.6021962216299544e-05,
+      "loss": 1.5171,
+      "step": 461
+    },
+    {
+      "epoch": 0.2847457627118644,
+      "grad_norm": 0.6166220903396606,
+      "learning_rate": 3.599614926002406e-05,
+      "loss": 1.5326,
+      "step": 462
+    },
+    {
+      "epoch": 0.28536209553158703,
+      "grad_norm": 0.6457735300064087,
+      "learning_rate": 3.5970262137819456e-05,
+      "loss": 1.5413,
+      "step": 463
+    },
+    {
+      "epoch": 0.2859784283513097,
+      "grad_norm": 0.6203924417495728,
+      "learning_rate": 3.594430096971104e-05,
+      "loss": 1.534,
+      "step": 464
+    },
+    {
+      "epoch": 0.28659476117103233,
+      "grad_norm": 0.6635544300079346,
+      "learning_rate": 3.591826587606742e-05,
+      "loss": 1.5581,
+      "step": 465
+    },
+    {
+      "epoch": 0.287211093990755,
+      "grad_norm": 0.6290273666381836,
+      "learning_rate": 3.589215697759995e-05,
+      "loss": 1.507,
+      "step": 466
+    },
+    {
+      "epoch": 0.2878274268104777,
+      "grad_norm": 0.6142866015434265,
+      "learning_rate": 3.586597439536216e-05,
+      "loss": 1.4979,
+      "step": 467
+    },
+    {
+      "epoch": 0.2884437596302003,
+      "grad_norm": 0.6286876201629639,
+      "learning_rate": 3.583971825074927e-05,
+      "loss": 1.4555,
+      "step": 468
+    },
+    {
+      "epoch": 0.289060092449923,
+      "grad_norm": 0.6208572387695312,
+      "learning_rate": 3.581338866549752e-05,
+      "loss": 1.5307,
+      "step": 469
+    },
+    {
+      "epoch": 0.2896764252696456,
+      "grad_norm": 0.6354894638061523,
+      "learning_rate": 3.5786985761683685e-05,
+      "loss": 1.4588,
+      "step": 470
+    },
+    {
+      "epoch": 0.2902927580893683,
+      "grad_norm": 0.6662725210189819,
+      "learning_rate": 3.576050966172447e-05,
+      "loss": 1.4847,
+      "step": 471
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 0.6339364647865295,
+      "learning_rate": 3.5733960488375946e-05,
+      "loss": 1.5006,
+      "step": 472
+    },
+    {
+      "epoch": 0.29152542372881357,
+      "grad_norm": 0.6413895487785339,
+      "learning_rate": 3.5707338364733e-05,
+      "loss": 1.5562,
+      "step": 473
+    },
+    {
+      "epoch": 0.2921417565485362,
+      "grad_norm": 0.6387337446212769,
+      "learning_rate": 3.568064341422874e-05,
+      "loss": 1.549,
+      "step": 474
+    },
+    {
+      "epoch": 0.29275808936825887,
+      "grad_norm": 0.6315402388572693,
+      "learning_rate": 3.5653875760633946e-05,
+      "loss": 1.5277,
+      "step": 475
+    },
+    {
+      "epoch": 0.2933744221879815,
+      "grad_norm": 0.649841845035553,
+      "learning_rate": 3.562703552805649e-05,
+      "loss": 1.5936,
+      "step": 476
+    },
+    {
+      "epoch": 0.29399075500770416,
+      "grad_norm": 0.681604266166687,
+      "learning_rate": 3.560012284094074e-05,
+      "loss": 1.5763,
+      "step": 477
+    },
+    {
+      "epoch": 0.2946070878274268,
+      "grad_norm": 0.6153651475906372,
+      "learning_rate": 3.5573137824067e-05,
+      "loss": 1.5296,
+      "step": 478
+    },
+    {
+      "epoch": 0.29522342064714946,
+      "grad_norm": 0.6196587681770325,
+      "learning_rate": 3.5546080602550935e-05,
+      "loss": 1.5017,
+      "step": 479
+    },
+    {
+      "epoch": 0.29583975346687214,
+      "grad_norm": 0.6492789387702942,
+      "learning_rate": 3.5518951301843e-05,
+      "loss": 1.4908,
+      "step": 480
+    },
+    {
+      "epoch": 0.29645608628659476,
+      "grad_norm": 0.6555952429771423,
+      "learning_rate": 3.5491750047727814e-05,
+      "loss": 1.5554,
+      "step": 481
+    },
+    {
+      "epoch": 0.29707241910631743,
+      "grad_norm": 0.6530175805091858,
+      "learning_rate": 3.5464476966323626e-05,
+      "loss": 1.5217,
+      "step": 482
+    },
+    {
+      "epoch": 0.29768875192604005,
+      "grad_norm": 0.6493626236915588,
+      "learning_rate": 3.543713218408171e-05,
+      "loss": 1.5261,
+      "step": 483
+    },
+    {
+      "epoch": 0.2983050847457627,
+      "grad_norm": 0.6442253589630127,
+      "learning_rate": 3.540971582778578e-05,
+      "loss": 1.4622,
+      "step": 484
+    },
+    {
+      "epoch": 0.29892141756548535,
+      "grad_norm": 0.6336924433708191,
+      "learning_rate": 3.538222802455139e-05,
+      "loss": 1.5069,
+      "step": 485
+    },
+    {
+      "epoch": 0.299537750385208,
+      "grad_norm": 0.6494876146316528,
+      "learning_rate": 3.535466890182539e-05,
+      "loss": 1.5268,
+      "step": 486
+    },
+    {
+      "epoch": 0.30015408320493064,
+      "grad_norm": 0.6452393531799316,
+      "learning_rate": 3.532703858738525e-05,
+      "loss": 1.586,
+      "step": 487
+    },
+    {
+      "epoch": 0.3007704160246533,
+      "grad_norm": 0.639706015586853,
+      "learning_rate": 3.5299337209338566e-05,
+      "loss": 1.5198,
+      "step": 488
+    },
+    {
+      "epoch": 0.30138674884437594,
+      "grad_norm": 0.6566674113273621,
+      "learning_rate": 3.52715648961224e-05,
+      "loss": 1.6219,
+      "step": 489
+    },
+    {
+      "epoch": 0.3020030816640986,
+      "grad_norm": 0.6216531991958618,
+      "learning_rate": 3.52437217765027e-05,
+      "loss": 1.4874,
+      "step": 490
+    },
+    {
+      "epoch": 0.30261941448382124,
+      "grad_norm": 0.6326376795768738,
+      "learning_rate": 3.5215807979573724e-05,
+      "loss": 1.5384,
+      "step": 491
+    },
+    {
+      "epoch": 0.3032357473035439,
+      "grad_norm": 0.6411398649215698,
+      "learning_rate": 3.5187823634757395e-05,
+      "loss": 1.5534,
+      "step": 492
+    },
+    {
+      "epoch": 0.3038520801232666,
+      "grad_norm": 0.6091376543045044,
+      "learning_rate": 3.5159768871802766e-05,
+      "loss": 1.5432,
+      "step": 493
+    },
+    {
+      "epoch": 0.3044684129429892,
+      "grad_norm": 0.6145322322845459,
+      "learning_rate": 3.513164382078535e-05,
+      "loss": 1.5224,
+      "step": 494
+    },
+    {
+      "epoch": 0.3050847457627119,
+      "grad_norm": 0.6333751678466797,
+      "learning_rate": 3.5103448612106575e-05,
+      "loss": 1.5535,
+      "step": 495
+    },
+    {
+      "epoch": 0.3057010785824345,
+      "grad_norm": 0.6383638978004456,
+      "learning_rate": 3.507518337649314e-05,
+      "loss": 1.5962,
+      "step": 496
+    },
+    {
+      "epoch": 0.3063174114021572,
+      "grad_norm": 0.6318841576576233,
+      "learning_rate": 3.504684824499642e-05,
+      "loss": 1.5213,
+      "step": 497
+    },
+    {
+      "epoch": 0.3069337442218798,
+      "grad_norm": 0.6321629881858826,
+      "learning_rate": 3.501844334899188e-05,
+      "loss": 1.5732,
+      "step": 498
+    },
+    {
+      "epoch": 0.3075500770416025,
+      "grad_norm": 0.6273744702339172,
+      "learning_rate": 3.498996882017843e-05,
+      "loss": 1.4679,
+      "step": 499
+    },
+    {
+      "epoch": 0.3081664098613251,
+      "grad_norm": 0.6182377934455872,
+      "learning_rate": 3.4961424790577835e-05,
+      "loss": 1.4724,
+      "step": 500
+    },
+    {
+      "epoch": 0.3087827426810478,
+      "grad_norm": 0.6165096759796143,
+      "learning_rate": 3.4932811392534104e-05,
+      "loss": 1.5005,
+      "step": 501
+    },
+    {
+      "epoch": 0.3093990755007704,
+      "grad_norm": 0.5994207859039307,
+      "learning_rate": 3.490412875871287e-05,
+      "loss": 1.4517,
+      "step": 502
+    },
+    {
+      "epoch": 0.31001540832049307,
+      "grad_norm": 0.6823679804801941,
+      "learning_rate": 3.487537702210077e-05,
+      "loss": 1.5238,
+      "step": 503
+    },
+    {
+      "epoch": 0.3106317411402157,
+      "grad_norm": 0.6435548067092896,
+      "learning_rate": 3.4846556316004844e-05,
+      "loss": 1.5634,
+      "step": 504
+    },
+    {
+      "epoch": 0.31124807395993837,
+      "grad_norm": 0.6374288201332092,
+      "learning_rate": 3.481766677405191e-05,
+      "loss": 1.5566,
+      "step": 505
+    },
+    {
+      "epoch": 0.31186440677966104,
+      "grad_norm": 0.6297028064727783,
+      "learning_rate": 3.478870853018792e-05,
+      "loss": 1.5282,
+      "step": 506
+    },
+    {
+      "epoch": 0.31248073959938366,
+      "grad_norm": 0.6188157200813293,
+      "learning_rate": 3.475968171867741e-05,
+      "loss": 1.5,
+      "step": 507
+    },
+    {
+      "epoch": 0.31309707241910634,
+      "grad_norm": 0.6542200446128845,
+      "learning_rate": 3.473058647410277e-05,
+      "loss": 1.5111,
+      "step": 508
+    },
+    {
+      "epoch": 0.31371340523882896,
+      "grad_norm": 0.6281660199165344,
+      "learning_rate": 3.47014229313637e-05,
+      "loss": 1.5577,
+      "step": 509
+    },
+    {
+      "epoch": 0.31432973805855163,
+      "grad_norm": 0.6386356353759766,
+      "learning_rate": 3.467219122567661e-05,
+      "loss": 1.5276,
+      "step": 510
+    },
+    {
+      "epoch": 0.31494607087827425,
+      "grad_norm": 0.6302286982536316,
+      "learning_rate": 3.464289149257387e-05,
+      "loss": 1.5313,
+      "step": 511
+    },
+    {
+      "epoch": 0.31556240369799693,
+      "grad_norm": 0.6208285093307495,
+      "learning_rate": 3.4613523867903324e-05,
+      "loss": 1.5475,
+      "step": 512
+    },
+    {
+      "epoch": 0.31617873651771955,
+      "grad_norm": 0.6123420000076294,
+      "learning_rate": 3.458408848782754e-05,
+      "loss": 1.4761,
+      "step": 513
+    },
+    {
+      "epoch": 0.3167950693374422,
+      "grad_norm": 0.648521900177002,
+      "learning_rate": 3.4554585488823265e-05,
+      "loss": 1.4817,
+      "step": 514
+    },
+    {
+      "epoch": 0.31741140215716485,
+      "grad_norm": 0.67547607421875,
+      "learning_rate": 3.452501500768076e-05,
+      "loss": 1.5742,
+      "step": 515
+    },
+    {
+      "epoch": 0.3180277349768875,
+      "grad_norm": 0.6527137160301208,
+      "learning_rate": 3.449537718150316e-05,
+      "loss": 1.4839,
+      "step": 516
+    },
+    {
+      "epoch": 0.31864406779661014,
+      "grad_norm": 0.6243513822555542,
+      "learning_rate": 3.4465672147705844e-05,
+      "loss": 1.4877,
+      "step": 517
+    },
+    {
+      "epoch": 0.3192604006163328,
+      "grad_norm": 0.6275103092193604,
+      "learning_rate": 3.4435900044015796e-05,
+      "loss": 1.5742,
+      "step": 518
+    },
+    {
+      "epoch": 0.3198767334360555,
+      "grad_norm": 0.6227298378944397,
+      "learning_rate": 3.440606100847096e-05,
+      "loss": 1.5862,
+      "step": 519
+    },
+    {
+      "epoch": 0.3204930662557781,
+      "grad_norm": 0.6253839731216431,
+      "learning_rate": 3.437615517941964e-05,
+      "loss": 1.5021,
+      "step": 520
+    },
+    {
+      "epoch": 0.3211093990755008,
+      "grad_norm": 0.6459096074104309,
+      "learning_rate": 3.43461826955198e-05,
+      "loss": 1.5576,
+      "step": 521
+    },
+    {
+      "epoch": 0.3217257318952234,
+      "grad_norm": 0.642366349697113,
+      "learning_rate": 3.431614369573844e-05,
+      "loss": 1.5324,
+      "step": 522
+    },
+    {
+      "epoch": 0.3223420647149461,
+      "grad_norm": 0.6393497586250305,
+      "learning_rate": 3.4286038319351e-05,
+      "loss": 1.4959,
+      "step": 523
+    },
+    {
+      "epoch": 0.3229583975346687,
+      "grad_norm": 0.6141563057899475,
+      "learning_rate": 3.4255866705940635e-05,
+      "loss": 1.5329,
+      "step": 524
+    },
+    {
+      "epoch": 0.3235747303543914,
+      "grad_norm": 0.6277351379394531,
+      "learning_rate": 3.4225628995397615e-05,
+      "loss": 1.4421,
+      "step": 525
+    },
+    {
+      "epoch": 0.324191063174114,
+      "grad_norm": 0.6193028092384338,
+      "learning_rate": 3.419532532791869e-05,
+      "loss": 1.4832,
+      "step": 526
+    },
+    {
+      "epoch": 0.3248073959938367,
+      "grad_norm": 0.6569095849990845,
+      "learning_rate": 3.4164955844006396e-05,
+      "loss": 1.5458,
+      "step": 527
+    },
+    {
+      "epoch": 0.3254237288135593,
+      "grad_norm": 0.6416183114051819,
+      "learning_rate": 3.413452068446843e-05,
+      "loss": 1.5214,
+      "step": 528
+    },
+    {
+      "epoch": 0.326040061633282,
+      "grad_norm": 0.6409138441085815,
+      "learning_rate": 3.410401999041701e-05,
+      "loss": 1.5498,
+      "step": 529
+    },
+    {
+      "epoch": 0.3266563944530046,
+      "grad_norm": 0.6253319382667542,
+      "learning_rate": 3.4073453903268166e-05,
+      "loss": 1.5245,
+      "step": 530
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 0.6222389340400696,
+      "learning_rate": 3.404282256474117e-05,
+      "loss": 1.4386,
+      "step": 531
+    },
+    {
+      "epoch": 0.32788906009244995,
+      "grad_norm": 0.6330371499061584,
+      "learning_rate": 3.40121261168578e-05,
+      "loss": 1.4607,
+      "step": 532
+    },
+    {
+      "epoch": 0.32850539291217257,
+      "grad_norm": 0.6958795189857483,
+      "learning_rate": 3.3981364701941714e-05,
+      "loss": 1.5438,
+      "step": 533
+    },
+    {
+      "epoch": 0.32912172573189524,
+      "grad_norm": 0.6625787019729614,
+      "learning_rate": 3.39505384626178e-05,
+      "loss": 1.5514,
+      "step": 534
+    },
+    {
+      "epoch": 0.32973805855161786,
+      "grad_norm": 0.6192840337753296,
+      "learning_rate": 3.39196475418115e-05,
+      "loss": 1.4313,
+      "step": 535
+    },
+    {
+      "epoch": 0.33035439137134054,
+      "grad_norm": 0.6188201308250427,
+      "learning_rate": 3.388869208274815e-05,
+      "loss": 1.4991,
+      "step": 536
+    },
+    {
+      "epoch": 0.33097072419106316,
+      "grad_norm": 0.6207196712493896,
+      "learning_rate": 3.385767222895232e-05,
+      "loss": 1.5068,
+      "step": 537
+    },
+    {
+      "epoch": 0.33158705701078584,
+      "grad_norm": 0.6164208650588989,
+      "learning_rate": 3.382658812424712e-05,
+      "loss": 1.4934,
+      "step": 538
+    },
+    {
+      "epoch": 0.33220338983050846,
+      "grad_norm": 0.6260819435119629,
+      "learning_rate": 3.379543991275361e-05,
+      "loss": 1.4864,
+      "step": 539
+    },
+    {
+      "epoch": 0.33281972265023113,
+      "grad_norm": 0.663669228553772,
+      "learning_rate": 3.3764227738890026e-05,
+      "loss": 1.4923,
+      "step": 540
+    },
+    {
+      "epoch": 0.33343605546995375,
+      "grad_norm": 0.6321974992752075,
+      "learning_rate": 3.37329517473712e-05,
+      "loss": 1.4754,
+      "step": 541
+    },
+    {
+      "epoch": 0.3340523882896764,
+      "grad_norm": 0.6340722441673279,
+      "learning_rate": 3.370161208320784e-05,
+      "loss": 1.5135,
+      "step": 542
+    },
+    {
+      "epoch": 0.3346687211093991,
+      "grad_norm": 0.6164632439613342,
+      "learning_rate": 3.367020889170588e-05,
+      "loss": 1.5456,
+      "step": 543
+    },
+    {
+      "epoch": 0.3352850539291217,
+      "grad_norm": 0.6347816586494446,
+      "learning_rate": 3.363874231846579e-05,
+      "loss": 1.5307,
+      "step": 544
+    },
+    {
+      "epoch": 0.3359013867488444,
+      "grad_norm": 0.6264616847038269,
+      "learning_rate": 3.36072125093819e-05,
+      "loss": 1.4914,
+      "step": 545
+    },
+    {
+      "epoch": 0.336517719568567,
+      "grad_norm": 0.6221004724502563,
+      "learning_rate": 3.357561961064175e-05,
+      "loss": 1.5095,
+      "step": 546
+    },
+    {
+      "epoch": 0.3371340523882897,
+      "grad_norm": 0.6659384965896606,
+      "learning_rate": 3.354396376872538e-05,
+      "loss": 1.5756,
+      "step": 547
+    },
+    {
+      "epoch": 0.3377503852080123,
+      "grad_norm": 0.6458001732826233,
+      "learning_rate": 3.3512245130404683e-05,
+      "loss": 1.5201,
+      "step": 548
+    },
+    {
+      "epoch": 0.338366718027735,
+      "grad_norm": 0.6084640622138977,
+      "learning_rate": 3.348046384274269e-05,
+      "loss": 1.5696,
+      "step": 549
+    },
+    {
+      "epoch": 0.3389830508474576,
+      "grad_norm": 0.6264820098876953,
+      "learning_rate": 3.344862005309291e-05,
+      "loss": 1.4964,
+      "step": 550
+    },
+    {
+      "epoch": 0.3395993836671803,
+      "grad_norm": 0.6191871166229248,
+      "learning_rate": 3.341671390909864e-05,
+      "loss": 1.5339,
+      "step": 551
+    },
+    {
+      "epoch": 0.3402157164869029,
+      "grad_norm": 0.6216069459915161,
+      "learning_rate": 3.338474555869228e-05,
+      "loss": 1.4627,
+      "step": 552
+    },
+    {
+      "epoch": 0.3408320493066256,
+      "grad_norm": 0.6235500574111938,
+      "learning_rate": 3.335271515009467e-05,
+      "loss": 1.473,
+      "step": 553
+    },
+    {
+      "epoch": 0.3414483821263482,
+      "grad_norm": 0.6313657164573669,
+      "learning_rate": 3.332062283181434e-05,
+      "loss": 1.4555,
+      "step": 554
+    },
+    {
+      "epoch": 0.3420647149460709,
+      "grad_norm": 0.6590183973312378,
+      "learning_rate": 3.328846875264691e-05,
+      "loss": 1.5455,
+      "step": 555
+    },
+    {
+      "epoch": 0.34268104776579356,
+      "grad_norm": 0.6327791810035706,
+      "learning_rate": 3.3256253061674326e-05,
+      "loss": 1.5265,
+      "step": 556
+    },
+    {
+      "epoch": 0.3432973805855162,
+      "grad_norm": 0.6291818618774414,
+      "learning_rate": 3.32239759082642e-05,
+      "loss": 1.5322,
+      "step": 557
+    },
+    {
+      "epoch": 0.34391371340523885,
+      "grad_norm": 0.6328380703926086,
+      "learning_rate": 3.3191637442069115e-05,
+      "loss": 1.4937,
+      "step": 558
+    },
+    {
+      "epoch": 0.3445300462249615,
+      "grad_norm": 0.6432162523269653,
+      "learning_rate": 3.315923781302592e-05,
+      "loss": 1.4924,
+      "step": 559
+    },
+    {
+      "epoch": 0.34514637904468415,
+      "grad_norm": 0.6319786310195923,
+      "learning_rate": 3.3126777171355083e-05,
+      "loss": 1.4655,
+      "step": 560
+    },
+    {
+      "epoch": 0.34576271186440677,
+      "grad_norm": 0.6463144421577454,
+      "learning_rate": 3.309425566755991e-05,
+      "loss": 1.5037,
+      "step": 561
+    },
+    {
+      "epoch": 0.34637904468412944,
+      "grad_norm": 0.6314020752906799,
+      "learning_rate": 3.306167345242591e-05,
+      "loss": 1.5287,
+      "step": 562
+    },
+    {
+      "epoch": 0.34699537750385206,
+      "grad_norm": 0.6312958002090454,
+      "learning_rate": 3.30290306770201e-05,
+      "loss": 1.5239,
+      "step": 563
+    },
+    {
+      "epoch": 0.34761171032357474,
+      "grad_norm": 0.631670355796814,
+      "learning_rate": 3.299632749269025e-05,
+      "loss": 1.5446,
+      "step": 564
+    },
+    {
+      "epoch": 0.34822804314329736,
+      "grad_norm": 0.6398617029190063,
+      "learning_rate": 3.296356405106423e-05,
+      "loss": 1.5112,
+      "step": 565
+    },
+    {
+      "epoch": 0.34884437596302004,
+      "grad_norm": 0.62071293592453,
+      "learning_rate": 3.293074050404929e-05,
+      "loss": 1.4851,
+      "step": 566
+    },
+    {
+      "epoch": 0.34946070878274266,
+      "grad_norm": 0.6425772309303284,
+      "learning_rate": 3.289785700383138e-05,
+      "loss": 1.486,
+      "step": 567
+    },
+    {
+      "epoch": 0.35007704160246533,
+      "grad_norm": 0.638628363609314,
+      "learning_rate": 3.2864913702874384e-05,
+      "loss": 1.5042,
+      "step": 568
+    },
+    {
+      "epoch": 0.350693374422188,
+      "grad_norm": 0.6274445056915283,
+      "learning_rate": 3.2831910753919477e-05,
+      "loss": 1.5057,
+      "step": 569
+    },
+    {
+      "epoch": 0.35130970724191063,
+      "grad_norm": 0.6332955956459045,
+      "learning_rate": 3.2798848309984385e-05,
+      "loss": 1.5038,
+      "step": 570
+    },
+    {
+      "epoch": 0.3519260400616333,
+      "grad_norm": 0.624282717704773,
+      "learning_rate": 3.276572652436268e-05,
+      "loss": 1.5155,
+      "step": 571
+    },
+    {
+      "epoch": 0.3525423728813559,
+      "grad_norm": 0.6485286951065063,
+      "learning_rate": 3.2732545550623066e-05,
+      "loss": 1.5301,
+      "step": 572
+    },
+    {
+      "epoch": 0.3531587057010786,
+      "grad_norm": 0.6295774579048157,
+      "learning_rate": 3.269930554260869e-05,
+      "loss": 1.5521,
+      "step": 573
+    },
+    {
+      "epoch": 0.3537750385208012,
+      "grad_norm": 0.6386708617210388,
+      "learning_rate": 3.266600665443638e-05,
+      "loss": 1.499,
+      "step": 574
+    },
+    {
+      "epoch": 0.3543913713405239,
+      "grad_norm": 0.6355401873588562,
+      "learning_rate": 3.2632649040496005e-05,
+      "loss": 1.5603,
+      "step": 575
+    },
+    {
+      "epoch": 0.3550077041602465,
+      "grad_norm": 0.6240369081497192,
+      "learning_rate": 3.259923285544966e-05,
+      "loss": 1.496,
+      "step": 576
+    },
+    {
+      "epoch": 0.3556240369799692,
+      "grad_norm": 0.6588268876075745,
+      "learning_rate": 3.256575825423106e-05,
+      "loss": 1.5202,
+      "step": 577
+    },
+    {
+      "epoch": 0.3562403697996918,
+      "grad_norm": 0.6418221592903137,
+      "learning_rate": 3.253222539204472e-05,
+      "loss": 1.4892,
+      "step": 578
+    },
+    {
+      "epoch": 0.3568567026194145,
+      "grad_norm": 0.6440759301185608,
+      "learning_rate": 3.249863442436532e-05,
+      "loss": 1.5012,
+      "step": 579
+    },
+    {
+      "epoch": 0.3574730354391371,
+      "grad_norm": 0.6364291906356812,
+      "learning_rate": 3.2464985506936914e-05,
+      "loss": 1.4539,
+      "step": 580
+    },
+    {
+      "epoch": 0.3580893682588598,
+      "grad_norm": 0.6213045120239258,
+      "learning_rate": 3.243127879577225e-05,
+      "loss": 1.4471,
+      "step": 581
+    },
+    {
+      "epoch": 0.35870570107858246,
+      "grad_norm": 0.6480481028556824,
+      "learning_rate": 3.2397514447152035e-05,
+      "loss": 1.4998,
+      "step": 582
+    },
+    {
+      "epoch": 0.3593220338983051,
+      "grad_norm": 0.6508931517601013,
+      "learning_rate": 3.2363692617624226e-05,
+      "loss": 1.5233,
+      "step": 583
+    },
+    {
+      "epoch": 0.35993836671802776,
+      "grad_norm": 0.6378174424171448,
+      "learning_rate": 3.232981346400327e-05,
+      "loss": 1.5147,
+      "step": 584
+    },
+    {
+      "epoch": 0.3605546995377504,
+      "grad_norm": 0.6446996927261353,
+      "learning_rate": 3.2295877143369394e-05,
+      "loss": 1.5091,
+      "step": 585
+    },
+    {
+      "epoch": 0.36117103235747305,
+      "grad_norm": 0.6296038031578064,
+      "learning_rate": 3.2261883813067906e-05,
+      "loss": 1.573,
+      "step": 586
+    },
+    {
+      "epoch": 0.3617873651771957,
+      "grad_norm": 0.6350198984146118,
+      "learning_rate": 3.222783363070841e-05,
+      "loss": 1.4667,
+      "step": 587
+    },
+    {
+      "epoch": 0.36240369799691835,
+      "grad_norm": 0.6288630962371826,
+      "learning_rate": 3.2193726754164114e-05,
+      "loss": 1.4666,
+      "step": 588
+    },
+    {
+      "epoch": 0.36302003081664097,
+      "grad_norm": 0.6324213743209839,
+      "learning_rate": 3.21595633415711e-05,
+      "loss": 1.5423,
+      "step": 589
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.6186695694923401,
+      "learning_rate": 3.2125343551327545e-05,
+      "loss": 1.4596,
+      "step": 590
+    },
+    {
+      "epoch": 0.36425269645608627,
+      "grad_norm": 0.6489487290382385,
+      "learning_rate": 3.209106754209306e-05,
+      "loss": 1.5015,
+      "step": 591
+    },
+    {
+      "epoch": 0.36486902927580894,
+      "grad_norm": 0.6401476263999939,
+      "learning_rate": 3.2056735472787896e-05,
+      "loss": 1.5509,
+      "step": 592
+    },
+    {
+      "epoch": 0.36548536209553156,
+      "grad_norm": 0.6308756470680237,
+      "learning_rate": 3.202234750259222e-05,
+      "loss": 1.4969,
+      "step": 593
+    },
+    {
+      "epoch": 0.36610169491525424,
+      "grad_norm": 0.62753826379776,
+      "learning_rate": 3.19879037909454e-05,
+      "loss": 1.4621,
+      "step": 594
+    },
+    {
+      "epoch": 0.3667180277349769,
+      "grad_norm": 0.6245593428611755,
+      "learning_rate": 3.1953404497545224e-05,
+      "loss": 1.4296,
+      "step": 595
+    },
+    {
+      "epoch": 0.36733436055469953,
+      "grad_norm": 0.6454522609710693,
+      "learning_rate": 3.191884978234721e-05,
+      "loss": 1.4412,
+      "step": 596
+    },
+    {
+      "epoch": 0.3679506933744222,
+      "grad_norm": 0.6471550464630127,
+      "learning_rate": 3.188423980556382e-05,
+      "loss": 1.5135,
+      "step": 597
+    },
+    {
+      "epoch": 0.36856702619414483,
+      "grad_norm": 0.6443375945091248,
+      "learning_rate": 3.1849574727663734e-05,
+      "loss": 1.5463,
+      "step": 598
+    },
+    {
+      "epoch": 0.3691833590138675,
+      "grad_norm": 0.6420419216156006,
+      "learning_rate": 3.181485470937114e-05,
+      "loss": 1.5464,
+      "step": 599
+    },
+    {
+      "epoch": 0.3697996918335901,
+      "grad_norm": 0.6404433250427246,
+      "learning_rate": 3.17800799116649e-05,
+      "loss": 1.5228,
+      "step": 600
+    },
+    {
+      "epoch": 0.3704160246533128,
+      "grad_norm": 0.6368747353553772,
+      "learning_rate": 3.174525049577793e-05,
+      "loss": 1.5595,
+      "step": 601
+    },
+    {
+      "epoch": 0.3710323574730354,
+      "grad_norm": 0.6601626873016357,
+      "learning_rate": 3.171036662319631e-05,
+      "loss": 1.4279,
+      "step": 602
+    },
+    {
+      "epoch": 0.3716486902927581,
+      "grad_norm": 0.6518347859382629,
+      "learning_rate": 3.167542845565866e-05,
+      "loss": 1.5115,
+      "step": 603
+    },
+    {
+      "epoch": 0.3722650231124807,
+      "grad_norm": 0.6589585542678833,
+      "learning_rate": 3.164043615515531e-05,
+      "loss": 1.4223,
+      "step": 604
+    },
+    {
+      "epoch": 0.3728813559322034,
+      "grad_norm": 0.6309995055198669,
+      "learning_rate": 3.160538988392759e-05,
+      "loss": 1.4763,
+      "step": 605
+    },
+    {
+      "epoch": 0.373497688751926,
+      "grad_norm": 0.6362992525100708,
+      "learning_rate": 3.157028980446707e-05,
+      "loss": 1.4877,
+      "step": 606
+    },
+    {
+      "epoch": 0.3741140215716487,
+      "grad_norm": 0.6353477239608765,
+      "learning_rate": 3.1535136079514785e-05,
+      "loss": 1.4983,
+      "step": 607
+    },
+    {
+      "epoch": 0.37473035439137137,
+      "grad_norm": 0.6715290546417236,
+      "learning_rate": 3.1499928872060504e-05,
+      "loss": 1.5051,
+      "step": 608
+    },
+    {
+      "epoch": 0.375346687211094,
+      "grad_norm": 0.6573690176010132,
+      "learning_rate": 3.146466834534197e-05,
+      "loss": 1.5581,
+      "step": 609
+    },
+    {
+      "epoch": 0.37596302003081666,
+      "grad_norm": 0.6558327674865723,
+      "learning_rate": 3.142935466284413e-05,
+      "loss": 1.4989,
+      "step": 610
+    },
+    {
+      "epoch": 0.3765793528505393,
+      "grad_norm": 0.6386260390281677,
+      "learning_rate": 3.13939879882984e-05,
+      "loss": 1.491,
+      "step": 611
+    },
+    {
+      "epoch": 0.37719568567026196,
+      "grad_norm": 0.6379230618476868,
+      "learning_rate": 3.135856848568188e-05,
+      "loss": 1.5419,
+      "step": 612
+    },
+    {
+      "epoch": 0.3778120184899846,
+      "grad_norm": 0.6291123032569885,
+      "learning_rate": 3.132309631921662e-05,
+      "loss": 1.4688,
+      "step": 613
+    },
+    {
+      "epoch": 0.37842835130970726,
+      "grad_norm": 0.6273422241210938,
+      "learning_rate": 3.128757165336885e-05,
+      "loss": 1.4147,
+      "step": 614
+    },
+    {
+      "epoch": 0.3790446841294299,
+      "grad_norm": 0.6356596946716309,
+      "learning_rate": 3.125199465284817e-05,
+      "loss": 1.5015,
+      "step": 615
+    },
+    {
+      "epoch": 0.37966101694915255,
+      "grad_norm": 0.6427242159843445,
+      "learning_rate": 3.1216365482606894e-05,
+      "loss": 1.5036,
+      "step": 616
+    },
+    {
+      "epoch": 0.3802773497688752,
+      "grad_norm": 0.6553143858909607,
+      "learning_rate": 3.118068430783917e-05,
+      "loss": 1.5595,
+      "step": 617
+    },
+    {
+      "epoch": 0.38089368258859785,
+      "grad_norm": 0.6268466114997864,
+      "learning_rate": 3.114495129398028e-05,
+      "loss": 1.4878,
+      "step": 618
+    },
+    {
+      "epoch": 0.38151001540832047,
+      "grad_norm": 0.6335549354553223,
+      "learning_rate": 3.1109166606705865e-05,
+      "loss": 1.537,
+      "step": 619
+    },
+    {
+      "epoch": 0.38212634822804314,
+      "grad_norm": 0.6393272876739502,
+      "learning_rate": 3.107333041193114e-05,
+      "loss": 1.5054,
+      "step": 620
+    },
+    {
+      "epoch": 0.3827426810477658,
+      "grad_norm": 0.6280071139335632,
+      "learning_rate": 3.103744287581012e-05,
+      "loss": 1.4897,
+      "step": 621
+    },
+    {
+      "epoch": 0.38335901386748844,
+      "grad_norm": 0.6517126560211182,
+      "learning_rate": 3.100150416473491e-05,
+      "loss": 1.4843,
+      "step": 622
+    },
+    {
+      "epoch": 0.3839753466872111,
+      "grad_norm": 0.6666373610496521,
+      "learning_rate": 3.096551444533483e-05,
+      "loss": 1.5471,
+      "step": 623
+    },
+    {
+      "epoch": 0.38459167950693374,
+      "grad_norm": 0.6469865441322327,
+      "learning_rate": 3.092947388447574e-05,
+      "loss": 1.549,
+      "step": 624
+    },
+    {
+      "epoch": 0.3852080123266564,
+      "grad_norm": 0.6489653587341309,
+      "learning_rate": 3.089338264925921e-05,
+      "loss": 1.5347,
+      "step": 625
+    },
+    {
+      "epoch": 0.38582434514637903,
+      "grad_norm": 0.6462615728378296,
+      "learning_rate": 3.085724090702177e-05,
+      "loss": 1.5098,
+      "step": 626
+    },
+    {
+      "epoch": 0.3864406779661017,
+      "grad_norm": 0.648750901222229,
+      "learning_rate": 3.08210488253341e-05,
+      "loss": 1.545,
+      "step": 627
+    },
+    {
+      "epoch": 0.38705701078582433,
+      "grad_norm": 0.6393259167671204,
+      "learning_rate": 3.07848065720003e-05,
+      "loss": 1.4958,
+      "step": 628
+    },
+    {
+      "epoch": 0.387673343605547,
+      "grad_norm": 0.6603264808654785,
+      "learning_rate": 3.0748514315057087e-05,
+      "loss": 1.4882,
+      "step": 629
+    },
+    {
+      "epoch": 0.3882896764252696,
+      "grad_norm": 0.6588591933250427,
+      "learning_rate": 3.071217222277303e-05,
+      "loss": 1.4983,
+      "step": 630
+    },
+    {
+      "epoch": 0.3889060092449923,
+      "grad_norm": 0.6373116970062256,
+      "learning_rate": 3.0675780463647737e-05,
+      "loss": 1.513,
+      "step": 631
+    },
+    {
+      "epoch": 0.3895223420647149,
+      "grad_norm": 0.6421071290969849,
+      "learning_rate": 3.0639339206411104e-05,
+      "loss": 1.539,
+      "step": 632
+    },
+    {
+      "epoch": 0.3901386748844376,
+      "grad_norm": 0.6400432586669922,
+      "learning_rate": 3.060284862002253e-05,
+      "loss": 1.4893,
+      "step": 633
+    },
+    {
+      "epoch": 0.3907550077041603,
+      "grad_norm": 0.6378886103630066,
+      "learning_rate": 3.056630887367013e-05,
+      "loss": 1.4926,
+      "step": 634
+    },
+    {
+      "epoch": 0.3913713405238829,
+      "grad_norm": 0.6466785669326782,
+      "learning_rate": 3.0529720136769924e-05,
+      "loss": 1.5619,
+      "step": 635
+    },
+    {
+      "epoch": 0.39198767334360557,
+      "grad_norm": 0.6507505774497986,
+      "learning_rate": 3.0493082578965114e-05,
+      "loss": 1.4699,
+      "step": 636
+    },
+    {
+      "epoch": 0.3926040061633282,
+      "grad_norm": 0.6481563448905945,
+      "learning_rate": 3.0456396370125233e-05,
+      "loss": 1.5298,
+      "step": 637
+    },
+    {
+      "epoch": 0.39322033898305087,
+      "grad_norm": 0.6414892077445984,
+      "learning_rate": 3.041966168034539e-05,
+      "loss": 1.4908,
+      "step": 638
+    },
+    {
+      "epoch": 0.3938366718027735,
+      "grad_norm": 0.6348177790641785,
+      "learning_rate": 3.038287867994547e-05,
+      "loss": 1.5126,
+      "step": 639
+    },
+    {
+      "epoch": 0.39445300462249616,
+      "grad_norm": 0.6342907547950745,
+      "learning_rate": 3.034604753946936e-05,
+      "loss": 1.4447,
+      "step": 640
+    },
+    {
+      "epoch": 0.3950693374422188,
+      "grad_norm": 0.6346410512924194,
+      "learning_rate": 3.0309168429684148e-05,
+      "loss": 1.4592,
+      "step": 641
+    },
+    {
+      "epoch": 0.39568567026194146,
+      "grad_norm": 0.642825186252594,
+      "learning_rate": 3.027224152157932e-05,
+      "loss": 1.5082,
+      "step": 642
+    },
+    {
+      "epoch": 0.3963020030816641,
+      "grad_norm": 0.6557008028030396,
+      "learning_rate": 3.0235266986366e-05,
+      "loss": 1.4764,
+      "step": 643
+    },
+    {
+      "epoch": 0.39691833590138675,
+      "grad_norm": 0.6577332019805908,
+      "learning_rate": 3.0198244995476087e-05,
+      "loss": 1.5069,
+      "step": 644
+    },
+    {
+      "epoch": 0.3975346687211094,
+      "grad_norm": 0.6441546678543091,
+      "learning_rate": 3.0161175720561574e-05,
+      "loss": 1.5226,
+      "step": 645
+    },
+    {
+      "epoch": 0.39815100154083205,
+      "grad_norm": 0.6265890598297119,
+      "learning_rate": 3.012405933349363e-05,
+      "loss": 1.454,
+      "step": 646
+    },
+    {
+      "epoch": 0.3987673343605547,
+      "grad_norm": 0.617996871471405,
+      "learning_rate": 3.0086896006361886e-05,
+      "loss": 1.3763,
+      "step": 647
+    },
+    {
+      "epoch": 0.39938366718027735,
+      "grad_norm": 0.7179849743843079,
+      "learning_rate": 3.0049685911473608e-05,
+      "loss": 1.5737,
+      "step": 648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6592563390731812,
+      "learning_rate": 3.0012429221352898e-05,
+      "loss": 1.489,
+      "step": 649
+    },
+    {
+      "epoch": 0.40061633281972264,
+      "grad_norm": 0.6569672226905823,
+      "learning_rate": 2.9975126108739875e-05,
+      "loss": 1.5116,
+      "step": 650
+    },
+    {
+      "epoch": 0.4012326656394453,
+      "grad_norm": 0.6473748683929443,
+      "learning_rate": 2.9937776746589945e-05,
+      "loss": 1.5001,
+      "step": 651
+    },
+    {
+      "epoch": 0.40184899845916794,
+      "grad_norm": 0.6282142400741577,
+      "learning_rate": 2.9900381308072894e-05,
+      "loss": 1.4954,
+      "step": 652
+    },
+    {
+      "epoch": 0.4024653312788906,
+      "grad_norm": 0.6429985761642456,
+      "learning_rate": 2.986293996657218e-05,
+      "loss": 1.4735,
+      "step": 653
+    },
+    {
+      "epoch": 0.40308166409861323,
+      "grad_norm": 0.6365038156509399,
+      "learning_rate": 2.9825452895684073e-05,
+      "loss": 1.4961,
+      "step": 654
+    },
+    {
+      "epoch": 0.4036979969183359,
+      "grad_norm": 0.6219319105148315,
+      "learning_rate": 2.978792026921686e-05,
+      "loss": 1.4487,
+      "step": 655
+    },
+    {
+      "epoch": 0.40431432973805853,
+      "grad_norm": 0.6654730439186096,
+      "learning_rate": 2.9750342261190063e-05,
+      "loss": 1.5337,
+      "step": 656
+    },
+    {
+      "epoch": 0.4049306625577812,
+      "grad_norm": 0.6424438953399658,
+      "learning_rate": 2.9712719045833603e-05,
+      "loss": 1.4577,
+      "step": 657
+    },
+    {
+      "epoch": 0.4055469953775038,
+      "grad_norm": 0.645980715751648,
+      "learning_rate": 2.967505079758701e-05,
+      "loss": 1.5231,
+      "step": 658
+    },
+    {
+      "epoch": 0.4061633281972265,
+      "grad_norm": 0.6435828804969788,
+      "learning_rate": 2.963733769109861e-05,
+      "loss": 1.4566,
+      "step": 659
+    },
+    {
+      "epoch": 0.4067796610169492,
+      "grad_norm": 0.6797378659248352,
+      "learning_rate": 2.95995799012247e-05,
+      "loss": 1.4502,
+      "step": 660
+    },
+    {
+      "epoch": 0.4073959938366718,
+      "grad_norm": 0.6313846707344055,
+      "learning_rate": 2.956177760302876e-05,
+      "loss": 1.4217,
+      "step": 661
+    },
+    {
+      "epoch": 0.4080123266563945,
+      "grad_norm": 0.6844395995140076,
+      "learning_rate": 2.952393097178065e-05,
+      "loss": 1.4947,
+      "step": 662
+    },
+    {
+      "epoch": 0.4086286594761171,
+      "grad_norm": 0.6478317379951477,
+      "learning_rate": 2.9486040182955737e-05,
+      "loss": 1.5423,
+      "step": 663
+    },
+    {
+      "epoch": 0.40924499229583977,
+      "grad_norm": 0.6268302202224731,
+      "learning_rate": 2.9448105412234178e-05,
+      "loss": 1.4477,
+      "step": 664
+    },
+    {
+      "epoch": 0.4098613251155624,
+      "grad_norm": 0.6396114826202393,
+      "learning_rate": 2.94101268355e-05,
+      "loss": 1.435,
+      "step": 665
+    },
+    {
+      "epoch": 0.41047765793528507,
+      "grad_norm": 0.6264893412590027,
+      "learning_rate": 2.937210462884037e-05,
+      "loss": 1.4967,
+      "step": 666
+    },
+    {
+      "epoch": 0.4110939907550077,
+      "grad_norm": 0.651495635509491,
+      "learning_rate": 2.9334038968544738e-05,
+      "loss": 1.5682,
+      "step": 667
+    },
+    {
+      "epoch": 0.41171032357473036,
+      "grad_norm": 0.6641359925270081,
+      "learning_rate": 2.9295930031104006e-05,
+      "loss": 1.551,
+      "step": 668
+    },
+    {
+      "epoch": 0.412326656394453,
+      "grad_norm": 0.6501958966255188,
+      "learning_rate": 2.9257777993209756e-05,
+      "loss": 1.5307,
+      "step": 669
+    },
+    {
+      "epoch": 0.41294298921417566,
+      "grad_norm": 0.651185154914856,
+      "learning_rate": 2.9219583031753392e-05,
+      "loss": 1.4893,
+      "step": 670
+    },
+    {
+      "epoch": 0.4135593220338983,
+      "grad_norm": 0.6622418761253357,
+      "learning_rate": 2.9181345323825342e-05,
+      "loss": 1.55,
+      "step": 671
+    },
+    {
+      "epoch": 0.41417565485362096,
+      "grad_norm": 0.665107786655426,
+      "learning_rate": 2.9143065046714202e-05,
+      "loss": 1.4776,
+      "step": 672
+    },
+    {
+      "epoch": 0.41479198767334363,
+      "grad_norm": 0.6447892785072327,
+      "learning_rate": 2.910474237790598e-05,
+      "loss": 1.5724,
+      "step": 673
+    },
+    {
+      "epoch": 0.41540832049306625,
+      "grad_norm": 0.6534367203712463,
+      "learning_rate": 2.9066377495083192e-05,
+      "loss": 1.4929,
+      "step": 674
+    },
+    {
+      "epoch": 0.41602465331278893,
+      "grad_norm": 0.6608549356460571,
+      "learning_rate": 2.9027970576124108e-05,
+      "loss": 1.5393,
+      "step": 675
+    },
+    {
+      "epoch": 0.41664098613251155,
+      "grad_norm": 0.6643514037132263,
+      "learning_rate": 2.898952179910189e-05,
+      "loss": 1.464,
+      "step": 676
+    },
+    {
+      "epoch": 0.4172573189522342,
+      "grad_norm": 0.6583855748176575,
+      "learning_rate": 2.8951031342283757e-05,
+      "loss": 1.4958,
+      "step": 677
+    },
+    {
+      "epoch": 0.41787365177195684,
+      "grad_norm": 0.6277612447738647,
+      "learning_rate": 2.8912499384130196e-05,
+      "loss": 1.4656,
+      "step": 678
+    },
+    {
+      "epoch": 0.4184899845916795,
+      "grad_norm": 0.6526576280593872,
+      "learning_rate": 2.8873926103294115e-05,
+      "loss": 1.4862,
+      "step": 679
+    },
+    {
+      "epoch": 0.41910631741140214,
+      "grad_norm": 0.6871560215950012,
+      "learning_rate": 2.8835311678619997e-05,
+      "loss": 1.5168,
+      "step": 680
+    },
+    {
+      "epoch": 0.4197226502311248,
+      "grad_norm": 0.6310210824012756,
+      "learning_rate": 2.8796656289143114e-05,
+      "loss": 1.4582,
+      "step": 681
+    },
+    {
+      "epoch": 0.42033898305084744,
+      "grad_norm": 0.6626084446907043,
+      "learning_rate": 2.875796011408863e-05,
+      "loss": 1.5576,
+      "step": 682
+    },
+    {
+      "epoch": 0.4209553158705701,
+      "grad_norm": 0.6169494390487671,
+      "learning_rate": 2.8719223332870864e-05,
+      "loss": 1.4072,
+      "step": 683
+    },
+    {
+      "epoch": 0.42157164869029273,
+      "grad_norm": 0.6335439085960388,
+      "learning_rate": 2.868044612509236e-05,
+      "loss": 1.4381,
+      "step": 684
+    },
+    {
+      "epoch": 0.4221879815100154,
+      "grad_norm": 0.6794674396514893,
+      "learning_rate": 2.8641628670543116e-05,
+      "loss": 1.5033,
+      "step": 685
+    },
+    {
+      "epoch": 0.4228043143297381,
+      "grad_norm": 0.6509531736373901,
+      "learning_rate": 2.860277114919974e-05,
+      "loss": 1.5414,
+      "step": 686
+    },
+    {
+      "epoch": 0.4234206471494607,
+      "grad_norm": 0.6503686904907227,
+      "learning_rate": 2.8563873741224602e-05,
+      "loss": 1.5238,
+      "step": 687
+    },
+    {
+      "epoch": 0.4240369799691834,
+      "grad_norm": 0.6381911039352417,
+      "learning_rate": 2.8524936626964996e-05,
+      "loss": 1.5677,
+      "step": 688
+    },
+    {
+      "epoch": 0.424653312788906,
+      "grad_norm": 0.652459979057312,
+      "learning_rate": 2.8485959986952344e-05,
+      "loss": 1.5336,
+      "step": 689
+    },
+    {
+      "epoch": 0.4252696456086287,
+      "grad_norm": 0.6382361054420471,
+      "learning_rate": 2.8446944001901298e-05,
+      "loss": 1.4167,
+      "step": 690
+    },
+    {
+      "epoch": 0.4258859784283513,
+      "grad_norm": 0.6898105144500732,
+      "learning_rate": 2.8407888852708932e-05,
+      "loss": 1.4322,
+      "step": 691
+    },
+    {
+      "epoch": 0.426502311248074,
+      "grad_norm": 0.687101423740387,
+      "learning_rate": 2.8368794720453935e-05,
+      "loss": 1.4322,
+      "step": 692
+    },
+    {
+      "epoch": 0.4271186440677966,
+      "grad_norm": 0.6702199578285217,
+      "learning_rate": 2.832966178639571e-05,
+      "loss": 1.4865,
+      "step": 693
+    },
+    {
+      "epoch": 0.42773497688751927,
+      "grad_norm": 0.6446210741996765,
+      "learning_rate": 2.8290490231973575e-05,
+      "loss": 1.4972,
+      "step": 694
+    },
+    {
+      "epoch": 0.4283513097072419,
+      "grad_norm": 0.6693007946014404,
+      "learning_rate": 2.8251280238805913e-05,
+      "loss": 1.4929,
+      "step": 695
+    },
+    {
+      "epoch": 0.42896764252696457,
+      "grad_norm": 0.6394990682601929,
+      "learning_rate": 2.8212031988689333e-05,
+      "loss": 1.4326,
+      "step": 696
+    },
+    {
+      "epoch": 0.4295839753466872,
+      "grad_norm": 0.6386460065841675,
+      "learning_rate": 2.81727456635978e-05,
+      "loss": 1.4112,
+      "step": 697
+    },
+    {
+      "epoch": 0.43020030816640986,
+      "grad_norm": 0.6588038802146912,
+      "learning_rate": 2.8133421445681843e-05,
+      "loss": 1.4654,
+      "step": 698
+    },
+    {
+      "epoch": 0.43081664098613254,
+      "grad_norm": 0.6661253571510315,
+      "learning_rate": 2.8094059517267646e-05,
+      "loss": 1.5444,
+      "step": 699
+    },
+    {
+      "epoch": 0.43143297380585516,
+      "grad_norm": 0.6428791284561157,
+      "learning_rate": 2.805466006085628e-05,
+      "loss": 1.4637,
+      "step": 700
+    },
+    {
+      "epoch": 0.43204930662557783,
+      "grad_norm": 0.6493973135948181,
+      "learning_rate": 2.801522325912277e-05,
+      "loss": 1.4362,
+      "step": 701
+    },
+    {
+      "epoch": 0.43266563944530045,
+      "grad_norm": 0.6501812934875488,
+      "learning_rate": 2.7975749294915315e-05,
+      "loss": 1.4655,
+      "step": 702
+    },
+    {
+      "epoch": 0.43328197226502313,
+      "grad_norm": 0.6604790687561035,
+      "learning_rate": 2.793623835125443e-05,
+      "loss": 1.4996,
+      "step": 703
+    },
+    {
+      "epoch": 0.43389830508474575,
+      "grad_norm": 0.6836243867874146,
+      "learning_rate": 2.7896690611332053e-05,
+      "loss": 1.4963,
+      "step": 704
+    },
+    {
+      "epoch": 0.4345146379044684,
+      "grad_norm": 0.6414328813552856,
+      "learning_rate": 2.785710625851075e-05,
+      "loss": 1.4454,
+      "step": 705
+    },
+    {
+      "epoch": 0.43513097072419105,
+      "grad_norm": 0.651253342628479,
+      "learning_rate": 2.781748547632284e-05,
+      "loss": 1.5247,
+      "step": 706
+    },
+    {
+      "epoch": 0.4357473035439137,
+      "grad_norm": 0.6446231007575989,
+      "learning_rate": 2.7777828448469536e-05,
+      "loss": 1.4538,
+      "step": 707
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 0.6915920972824097,
+      "learning_rate": 2.7738135358820114e-05,
+      "loss": 1.5617,
+      "step": 708
+    },
+    {
+      "epoch": 0.436979969183359,
+      "grad_norm": 0.6558811068534851,
+      "learning_rate": 2.769840639141104e-05,
+      "loss": 1.4662,
+      "step": 709
+    },
+    {
+      "epoch": 0.43759630200308164,
+      "grad_norm": 0.6515608429908752,
+      "learning_rate": 2.7658641730445142e-05,
+      "loss": 1.4382,
+      "step": 710
+    },
+    {
+      "epoch": 0.4382126348228043,
+      "grad_norm": 0.6421152353286743,
+      "learning_rate": 2.7618841560290722e-05,
+      "loss": 1.4465,
+      "step": 711
+    },
+    {
+      "epoch": 0.438828967642527,
+      "grad_norm": 0.6631745100021362,
+      "learning_rate": 2.7579006065480737e-05,
+      "loss": 1.4599,
+      "step": 712
+    },
+    {
+      "epoch": 0.4394453004622496,
+      "grad_norm": 0.6604713797569275,
+      "learning_rate": 2.7539135430711908e-05,
+      "loss": 1.4806,
+      "step": 713
+    },
+    {
+      "epoch": 0.4400616332819723,
+      "grad_norm": 0.6700551509857178,
+      "learning_rate": 2.74992298408439e-05,
+      "loss": 1.4934,
+      "step": 714
+    },
+    {
+      "epoch": 0.4406779661016949,
+      "grad_norm": 0.6602810621261597,
+      "learning_rate": 2.7459289480898443e-05,
+      "loss": 1.5138,
+      "step": 715
+    },
+    {
+      "epoch": 0.4412942989214176,
+      "grad_norm": 0.6400670409202576,
+      "learning_rate": 2.7419314536058467e-05,
+      "loss": 1.4378,
+      "step": 716
+    },
+    {
+      "epoch": 0.4419106317411402,
+      "grad_norm": 0.6445081830024719,
+      "learning_rate": 2.7379305191667263e-05,
+      "loss": 1.4935,
+      "step": 717
+    },
+    {
+      "epoch": 0.4425269645608629,
+      "grad_norm": 0.6491249799728394,
+      "learning_rate": 2.7339261633227612e-05,
+      "loss": 1.4695,
+      "step": 718
+    },
+    {
+      "epoch": 0.4431432973805855,
+      "grad_norm": 0.6502711176872253,
+      "learning_rate": 2.7299184046400936e-05,
+      "loss": 1.4849,
+      "step": 719
+    },
+    {
+      "epoch": 0.4437596302003082,
+      "grad_norm": 0.6470304131507874,
+      "learning_rate": 2.725907261700642e-05,
+      "loss": 1.4573,
+      "step": 720
+    },
+    {
+      "epoch": 0.4443759630200308,
+      "grad_norm": 0.6609827280044556,
+      "learning_rate": 2.721892753102016e-05,
+      "loss": 1.481,
+      "step": 721
+    },
+    {
+      "epoch": 0.44499229583975347,
+      "grad_norm": 0.6376433372497559,
+      "learning_rate": 2.71787489745743e-05,
+      "loss": 1.4438,
+      "step": 722
+    },
+    {
+      "epoch": 0.4456086286594761,
+      "grad_norm": 0.6593858003616333,
+      "learning_rate": 2.7138537133956186e-05,
+      "loss": 1.5037,
+      "step": 723
+    },
+    {
+      "epoch": 0.44622496147919877,
+      "grad_norm": 0.6915332078933716,
+      "learning_rate": 2.7098292195607456e-05,
+      "loss": 1.4216,
+      "step": 724
+    },
+    {
+      "epoch": 0.44684129429892144,
+      "grad_norm": 0.6552739143371582,
+      "learning_rate": 2.7058014346123225e-05,
+      "loss": 1.4874,
+      "step": 725
+    },
+    {
+      "epoch": 0.44745762711864406,
+      "grad_norm": 0.6488803625106812,
+      "learning_rate": 2.70177037722512e-05,
+      "loss": 1.5355,
+      "step": 726
+    },
+    {
+      "epoch": 0.44807395993836674,
+      "grad_norm": 0.6355862021446228,
+      "learning_rate": 2.69773606608908e-05,
+      "loss": 1.4827,
+      "step": 727
+    },
+    {
+      "epoch": 0.44869029275808936,
+      "grad_norm": 0.652966320514679,
+      "learning_rate": 2.6936985199092335e-05,
+      "loss": 1.4875,
+      "step": 728
+    },
+    {
+      "epoch": 0.44930662557781204,
+      "grad_norm": 0.6532738208770752,
+      "learning_rate": 2.6896577574056065e-05,
+      "loss": 1.5267,
+      "step": 729
+    },
+    {
+      "epoch": 0.44992295839753466,
+      "grad_norm": 0.6622675657272339,
+      "learning_rate": 2.6856137973131403e-05,
+      "loss": 1.449,
+      "step": 730
+    },
+    {
+      "epoch": 0.45053929121725733,
+      "grad_norm": 0.641776442527771,
+      "learning_rate": 2.681566658381601e-05,
+      "loss": 1.472,
+      "step": 731
+    },
+    {
+      "epoch": 0.45115562403697995,
+      "grad_norm": 0.6786929965019226,
+      "learning_rate": 2.6775163593754937e-05,
+      "loss": 1.5539,
+      "step": 732
+    },
+    {
+      "epoch": 0.45177195685670263,
+      "grad_norm": 0.6311996579170227,
+      "learning_rate": 2.6734629190739744e-05,
+      "loss": 1.4317,
+      "step": 733
+    },
+    {
+      "epoch": 0.45238828967642525,
+      "grad_norm": 0.6562221050262451,
+      "learning_rate": 2.669406356270764e-05,
+      "loss": 1.5049,
+      "step": 734
+    },
+    {
+      "epoch": 0.4530046224961479,
+      "grad_norm": 0.6616199612617493,
+      "learning_rate": 2.6653466897740607e-05,
+      "loss": 1.4478,
+      "step": 735
+    },
+    {
+      "epoch": 0.45362095531587054,
+      "grad_norm": 0.6472796201705933,
+      "learning_rate": 2.6612839384064546e-05,
+      "loss": 1.4283,
+      "step": 736
+    },
+    {
+      "epoch": 0.4542372881355932,
+      "grad_norm": 0.6530714631080627,
+      "learning_rate": 2.657218121004836e-05,
+      "loss": 1.4406,
+      "step": 737
+    },
+    {
+      "epoch": 0.4548536209553159,
+      "grad_norm": 0.709610104560852,
+      "learning_rate": 2.653149256420312e-05,
+      "loss": 1.4922,
+      "step": 738
+    },
+    {
+      "epoch": 0.4554699537750385,
+      "grad_norm": 0.6500923037528992,
+      "learning_rate": 2.6490773635181188e-05,
+      "loss": 1.4193,
+      "step": 739
+    },
+    {
+      "epoch": 0.4560862865947612,
+      "grad_norm": 0.6618855595588684,
+      "learning_rate": 2.6450024611775318e-05,
+      "loss": 1.4259,
+      "step": 740
+    },
+    {
+      "epoch": 0.4567026194144838,
+      "grad_norm": 0.6506283283233643,
+      "learning_rate": 2.6409245682917814e-05,
+      "loss": 1.5005,
+      "step": 741
+    },
+    {
+      "epoch": 0.4573189522342065,
+      "grad_norm": 0.6780362725257874,
+      "learning_rate": 2.6368437037679625e-05,
+      "loss": 1.4403,
+      "step": 742
+    },
+    {
+      "epoch": 0.4579352850539291,
+      "grad_norm": 0.6576229333877563,
+      "learning_rate": 2.632759886526947e-05,
+      "loss": 1.4727,
+      "step": 743
+    },
+    {
+      "epoch": 0.4585516178736518,
+      "grad_norm": 0.671320915222168,
+      "learning_rate": 2.6286731355033e-05,
+      "loss": 1.4791,
+      "step": 744
+    },
+    {
+      "epoch": 0.4591679506933744,
+      "grad_norm": 0.6556963920593262,
+      "learning_rate": 2.6245834696451857e-05,
+      "loss": 1.4887,
+      "step": 745
+    },
+    {
+      "epoch": 0.4597842835130971,
+      "grad_norm": 0.651106059551239,
+      "learning_rate": 2.6204909079142847e-05,
+      "loss": 1.4462,
+      "step": 746
+    },
+    {
+      "epoch": 0.4604006163328197,
+      "grad_norm": 0.6544069051742554,
+      "learning_rate": 2.6163954692857047e-05,
+      "loss": 1.5272,
+      "step": 747
+    },
+    {
+      "epoch": 0.4610169491525424,
+      "grad_norm": 0.6927053332328796,
+      "learning_rate": 2.61229717274789e-05,
+      "loss": 1.4843,
+      "step": 748
+    },
+    {
+      "epoch": 0.461633281972265,
+      "grad_norm": 0.6328755617141724,
+      "learning_rate": 2.6081960373025378e-05,
+      "loss": 1.3956,
+      "step": 749
+    },
+    {
+      "epoch": 0.4622496147919877,
+      "grad_norm": 0.6715585589408875,
+      "learning_rate": 2.604092081964507e-05,
+      "loss": 1.4925,
+      "step": 750
+    },
+    {
+      "epoch": 0.46286594761171035,
+      "grad_norm": 0.6592444181442261,
+      "learning_rate": 2.5999853257617315e-05,
+      "loss": 1.5232,
+      "step": 751
+    },
+    {
+      "epoch": 0.46348228043143297,
+      "grad_norm": 0.6313751339912415,
+      "learning_rate": 2.5958757877351294e-05,
+      "loss": 1.4398,
+      "step": 752
+    },
+    {
+      "epoch": 0.46409861325115565,
+      "grad_norm": 0.6490797996520996,
+      "learning_rate": 2.5917634869385185e-05,
+      "loss": 1.4558,
+      "step": 753
+    },
+    {
+      "epoch": 0.46471494607087827,
+      "grad_norm": 0.6704680919647217,
+      "learning_rate": 2.5876484424385263e-05,
+      "loss": 1.5176,
+      "step": 754
+    },
+    {
+      "epoch": 0.46533127889060094,
+      "grad_norm": 0.6712368726730347,
+      "learning_rate": 2.5835306733144996e-05,
+      "loss": 1.4243,
+      "step": 755
+    },
+    {
+      "epoch": 0.46594761171032356,
+      "grad_norm": 0.6699070334434509,
+      "learning_rate": 2.57941019865842e-05,
+      "loss": 1.5105,
+      "step": 756
+    },
+    {
+      "epoch": 0.46656394453004624,
+      "grad_norm": 0.6497899293899536,
+      "learning_rate": 2.575287037574812e-05,
+      "loss": 1.4815,
+      "step": 757
+    },
+    {
+      "epoch": 0.46718027734976886,
+      "grad_norm": 0.6557024717330933,
+      "learning_rate": 2.5711612091806566e-05,
+      "loss": 1.4426,
+      "step": 758
+    },
+    {
+      "epoch": 0.46779661016949153,
+      "grad_norm": 0.6698507070541382,
+      "learning_rate": 2.567032732605301e-05,
+      "loss": 1.4753,
+      "step": 759
+    },
+    {
+      "epoch": 0.46841294298921415,
+      "grad_norm": 0.6543661952018738,
+      "learning_rate": 2.5629016269903706e-05,
+      "loss": 1.498,
+      "step": 760
+    },
+    {
+      "epoch": 0.46902927580893683,
+      "grad_norm": 0.6658725738525391,
+      "learning_rate": 2.5587679114896806e-05,
+      "loss": 1.4045,
+      "step": 761
+    },
+    {
+      "epoch": 0.46964560862865945,
+      "grad_norm": 0.6653224229812622,
+      "learning_rate": 2.5546316052691467e-05,
+      "loss": 1.469,
+      "step": 762
+    },
+    {
+      "epoch": 0.4702619414483821,
+      "grad_norm": 0.6615314483642578,
+      "learning_rate": 2.550492727506696e-05,
+      "loss": 1.5018,
+      "step": 763
+    },
+    {
+      "epoch": 0.4708782742681048,
+      "grad_norm": 0.6513172388076782,
+      "learning_rate": 2.5463512973921814e-05,
+      "loss": 1.4181,
+      "step": 764
+    },
+    {
+      "epoch": 0.4714946070878274,
+      "grad_norm": 0.6554753184318542,
+      "learning_rate": 2.5422073341272852e-05,
+      "loss": 1.3937,
+      "step": 765
+    },
+    {
+      "epoch": 0.4721109399075501,
+      "grad_norm": 0.6689674258232117,
+      "learning_rate": 2.538060856925437e-05,
+      "loss": 1.4254,
+      "step": 766
+    },
+    {
+      "epoch": 0.4727272727272727,
+      "grad_norm": 0.6607903242111206,
+      "learning_rate": 2.5339118850117244e-05,
+      "loss": 1.4113,
+      "step": 767
+    },
+    {
+      "epoch": 0.4733436055469954,
+      "grad_norm": 0.6706930994987488,
+      "learning_rate": 2.5297604376227963e-05,
+      "loss": 1.4264,
+      "step": 768
+    },
+    {
+      "epoch": 0.473959938366718,
+      "grad_norm": 0.6535710692405701,
+      "learning_rate": 2.5256065340067843e-05,
+      "loss": 1.4139,
+      "step": 769
+    },
+    {
+      "epoch": 0.4745762711864407,
+      "grad_norm": 0.6666339039802551,
+      "learning_rate": 2.521450193423206e-05,
+      "loss": 1.4982,
+      "step": 770
+    },
+    {
+      "epoch": 0.4751926040061633,
+      "grad_norm": 0.6591233611106873,
+      "learning_rate": 2.5172914351428765e-05,
+      "loss": 1.472,
+      "step": 771
+    },
+    {
+      "epoch": 0.475808936825886,
+      "grad_norm": 0.6556514501571655,
+      "learning_rate": 2.513130278447824e-05,
+      "loss": 1.4035,
+      "step": 772
+    },
+    {
+      "epoch": 0.4764252696456086,
+      "grad_norm": 0.681161642074585,
+      "learning_rate": 2.508966742631195e-05,
+      "loss": 1.5286,
+      "step": 773
+    },
+    {
+      "epoch": 0.4770416024653313,
+      "grad_norm": 0.651854395866394,
+      "learning_rate": 2.504800846997165e-05,
+      "loss": 1.4774,
+      "step": 774
+    },
+    {
+      "epoch": 0.4776579352850539,
+      "grad_norm": 0.6415034532546997,
+      "learning_rate": 2.500632610860855e-05,
+      "loss": 1.4045,
+      "step": 775
+    },
+    {
+      "epoch": 0.4782742681047766,
+      "grad_norm": 0.6623095273971558,
+      "learning_rate": 2.4964620535482336e-05,
+      "loss": 1.4454,
+      "step": 776
+    },
+    {
+      "epoch": 0.47889060092449925,
+      "grad_norm": 0.6609742045402527,
+      "learning_rate": 2.492289194396034e-05,
+      "loss": 1.4954,
+      "step": 777
+    },
+    {
+      "epoch": 0.4795069337442219,
+      "grad_norm": 0.6443544030189514,
+      "learning_rate": 2.4881140527516613e-05,
+      "loss": 1.4089,
+      "step": 778
+    },
+    {
+      "epoch": 0.48012326656394455,
+      "grad_norm": 0.6567139029502869,
+      "learning_rate": 2.483936647973103e-05,
+      "loss": 1.4591,
+      "step": 779
+    },
+    {
+      "epoch": 0.48073959938366717,
+      "grad_norm": 0.6636015772819519,
+      "learning_rate": 2.4797569994288385e-05,
+      "loss": 1.4427,
+      "step": 780
+    },
+    {
+      "epoch": 0.48135593220338985,
+      "grad_norm": 0.6707240343093872,
+      "learning_rate": 2.475575126497753e-05,
+      "loss": 1.5628,
+      "step": 781
+    },
+    {
+      "epoch": 0.48197226502311247,
+      "grad_norm": 0.667204737663269,
+      "learning_rate": 2.4713910485690428e-05,
+      "loss": 1.4458,
+      "step": 782
+    },
+    {
+      "epoch": 0.48258859784283514,
+      "grad_norm": 0.6726914644241333,
+      "learning_rate": 2.4672047850421295e-05,
+      "loss": 1.4424,
+      "step": 783
+    },
+    {
+      "epoch": 0.48320493066255776,
+      "grad_norm": 0.6603702902793884,
+      "learning_rate": 2.4630163553265657e-05,
+      "loss": 1.4882,
+      "step": 784
+    },
+    {
+      "epoch": 0.48382126348228044,
+      "grad_norm": 0.6715510487556458,
+      "learning_rate": 2.4588257788419503e-05,
+      "loss": 1.4697,
+      "step": 785
+    },
+    {
+      "epoch": 0.48443759630200306,
+      "grad_norm": 0.6559399366378784,
+      "learning_rate": 2.454633075017834e-05,
+      "loss": 1.4835,
+      "step": 786
+    },
+    {
+      "epoch": 0.48505392912172574,
+      "grad_norm": 0.6464762687683105,
+      "learning_rate": 2.4504382632936303e-05,
+      "loss": 1.4401,
+      "step": 787
+    },
+    {
+      "epoch": 0.48567026194144836,
+      "grad_norm": 0.6455738544464111,
+      "learning_rate": 2.446241363118528e-05,
+      "loss": 1.39,
+      "step": 788
+    },
+    {
+      "epoch": 0.48628659476117103,
+      "grad_norm": 0.6899600625038147,
+      "learning_rate": 2.4420423939513973e-05,
+      "loss": 1.4898,
+      "step": 789
+    },
+    {
+      "epoch": 0.4869029275808937,
+      "grad_norm": 0.6539801955223083,
+      "learning_rate": 2.4378413752607025e-05,
+      "loss": 1.4481,
+      "step": 790
+    },
+    {
+      "epoch": 0.4875192604006163,
+      "grad_norm": 0.6555554866790771,
+      "learning_rate": 2.4336383265244088e-05,
+      "loss": 1.4525,
+      "step": 791
+    },
+    {
+      "epoch": 0.488135593220339,
+      "grad_norm": 0.6425526142120361,
+      "learning_rate": 2.4294332672298947e-05,
+      "loss": 1.3769,
+      "step": 792
+    },
+    {
+      "epoch": 0.4887519260400616,
+      "grad_norm": 0.6495502591133118,
+      "learning_rate": 2.4252262168738617e-05,
+      "loss": 1.4287,
+      "step": 793
+    },
+    {
+      "epoch": 0.4893682588597843,
+      "grad_norm": 0.6833025217056274,
+      "learning_rate": 2.4210171949622408e-05,
+      "loss": 1.4681,
+      "step": 794
+    },
+    {
+      "epoch": 0.4899845916795069,
+      "grad_norm": 0.6515230536460876,
+      "learning_rate": 2.416806221010105e-05,
+      "loss": 1.4676,
+      "step": 795
+    },
+    {
+      "epoch": 0.4906009244992296,
+      "grad_norm": 0.6599568128585815,
+      "learning_rate": 2.412593314541578e-05,
+      "loss": 1.4813,
+      "step": 796
+    },
+    {
+      "epoch": 0.4912172573189522,
+      "grad_norm": 0.6508007645606995,
+      "learning_rate": 2.408378495089744e-05,
+      "loss": 1.4416,
+      "step": 797
+    },
+    {
+      "epoch": 0.4918335901386749,
+      "grad_norm": 0.6619030237197876,
+      "learning_rate": 2.404161782196556e-05,
+      "loss": 1.4391,
+      "step": 798
+    },
+    {
+      "epoch": 0.4924499229583975,
+      "grad_norm": 0.6672455668449402,
+      "learning_rate": 2.3999431954127457e-05,
+      "loss": 1.5233,
+      "step": 799
+    },
+    {
+      "epoch": 0.4930662557781202,
+      "grad_norm": 0.6848391890525818,
+      "learning_rate": 2.3957227542977334e-05,
+      "loss": 1.5218,
+      "step": 800
+    },
+    {
+      "epoch": 0.4936825885978428,
+      "grad_norm": 0.6891319751739502,
+      "learning_rate": 2.3915004784195386e-05,
+      "loss": 1.5312,
+      "step": 801
+    },
+    {
+      "epoch": 0.4942989214175655,
+      "grad_norm": 0.664033055305481,
+      "learning_rate": 2.3872763873546835e-05,
+      "loss": 1.4585,
+      "step": 802
+    },
+    {
+      "epoch": 0.49491525423728816,
+      "grad_norm": 0.6545723676681519,
+      "learning_rate": 2.3830505006881125e-05,
+      "loss": 1.4515,
+      "step": 803
+    },
+    {
+      "epoch": 0.4955315870570108,
+      "grad_norm": 0.6647583246231079,
+      "learning_rate": 2.3788228380130884e-05,
+      "loss": 1.4261,
+      "step": 804
+    },
+    {
+      "epoch": 0.49614791987673346,
+      "grad_norm": 0.6671835780143738,
+      "learning_rate": 2.374593418931114e-05,
+      "loss": 1.5139,
+      "step": 805
+    },
+    {
+      "epoch": 0.4967642526964561,
+      "grad_norm": 0.6669169664382935,
+      "learning_rate": 2.370362263051833e-05,
+      "loss": 1.4518,
+      "step": 806
+    },
+    {
+      "epoch": 0.49738058551617875,
+      "grad_norm": 0.6806299686431885,
+      "learning_rate": 2.3661293899929417e-05,
+      "loss": 1.4638,
+      "step": 807
+    },
+    {
+      "epoch": 0.4979969183359014,
+      "grad_norm": 0.6680865287780762,
+      "learning_rate": 2.3618948193800996e-05,
+      "loss": 1.4511,
+      "step": 808
+    },
+    {
+      "epoch": 0.49861325115562405,
+      "grad_norm": 0.6399280428886414,
+      "learning_rate": 2.3576585708468355e-05,
+      "loss": 1.3566,
+      "step": 809
+    },
+    {
+      "epoch": 0.49922958397534667,
+      "grad_norm": 0.6574198603630066,
+      "learning_rate": 2.353420664034457e-05,
+      "loss": 1.5564,
+      "step": 810
+    },
+    {
+      "epoch": 0.49984591679506934,
+      "grad_norm": 0.6828180551528931,
+      "learning_rate": 2.3491811185919644e-05,
+      "loss": 1.4455,
+      "step": 811
+    },
+    {
+      "epoch": 0.500462249614792,
+      "grad_norm": 0.6625897288322449,
+      "learning_rate": 2.3449399541759498e-05,
+      "loss": 1.3979,
+      "step": 812
+    },
+    {
+      "epoch": 0.5010785824345146,
+      "grad_norm": 0.6758010983467102,
+      "learning_rate": 2.3406971904505155e-05,
+      "loss": 1.4646,
+      "step": 813
+    },
+    {
+      "epoch": 0.5016949152542373,
+      "grad_norm": 0.6834417581558228,
+      "learning_rate": 2.336452847087179e-05,
+      "loss": 1.4211,
+      "step": 814
+    },
+    {
+      "epoch": 0.50231124807396,
+      "grad_norm": 0.6549422144889832,
+      "learning_rate": 2.3322069437647795e-05,
+      "loss": 1.3809,
+      "step": 815
+    },
+    {
+      "epoch": 0.5029275808936826,
+      "grad_norm": 0.6464531421661377,
+      "learning_rate": 2.3279595001693903e-05,
+      "loss": 1.4303,
+      "step": 816
+    },
+    {
+      "epoch": 0.5035439137134052,
+      "grad_norm": 0.6724410653114319,
+      "learning_rate": 2.3237105359942272e-05,
+      "loss": 1.5311,
+      "step": 817
+    },
+    {
+      "epoch": 0.5041602465331279,
+      "grad_norm": 0.6607643365859985,
+      "learning_rate": 2.319460070939553e-05,
+      "loss": 1.4646,
+      "step": 818
+    },
+    {
+      "epoch": 0.5047765793528506,
+      "grad_norm": 0.6936811208724976,
+      "learning_rate": 2.315208124712593e-05,
+      "loss": 1.491,
+      "step": 819
+    },
+    {
+      "epoch": 0.5053929121725732,
+      "grad_norm": 0.6466771364212036,
+      "learning_rate": 2.3109547170274373e-05,
+      "loss": 1.415,
+      "step": 820
+    },
+    {
+      "epoch": 0.5060092449922958,
+      "grad_norm": 0.6754940152168274,
+      "learning_rate": 2.306699867604953e-05,
+      "loss": 1.4786,
+      "step": 821
+    },
+    {
+      "epoch": 0.5066255778120184,
+      "grad_norm": 0.6594055891036987,
+      "learning_rate": 2.3024435961726926e-05,
+      "loss": 1.4387,
+      "step": 822
+    },
+    {
+      "epoch": 0.5072419106317412,
+      "grad_norm": 0.6514335870742798,
+      "learning_rate": 2.2981859224647997e-05,
+      "loss": 1.4451,
+      "step": 823
+    },
+    {
+      "epoch": 0.5078582434514638,
+      "grad_norm": 0.659854531288147,
+      "learning_rate": 2.2939268662219213e-05,
+      "loss": 1.4717,
+      "step": 824
+    },
+    {
+      "epoch": 0.5084745762711864,
+      "grad_norm": 0.6732680201530457,
+      "learning_rate": 2.289666447191114e-05,
+      "loss": 1.5037,
+      "step": 825
+    },
+    {
+      "epoch": 0.509090909090909,
+      "grad_norm": 0.657238781452179,
+      "learning_rate": 2.2854046851257525e-05,
+      "loss": 1.4832,
+      "step": 826
+    },
+    {
+      "epoch": 0.5097072419106318,
+      "grad_norm": 0.6703278422355652,
+      "learning_rate": 2.281141599785439e-05,
+      "loss": 1.4503,
+      "step": 827
+    },
+    {
+      "epoch": 0.5103235747303544,
+      "grad_norm": 0.6670629382133484,
+      "learning_rate": 2.2768772109359114e-05,
+      "loss": 1.4201,
+      "step": 828
+    },
+    {
+      "epoch": 0.510939907550077,
+      "grad_norm": 0.6621464490890503,
+      "learning_rate": 2.2726115383489495e-05,
+      "loss": 1.4666,
+      "step": 829
+    },
+    {
+      "epoch": 0.5115562403697997,
+      "grad_norm": 0.6738835573196411,
+      "learning_rate": 2.2683446018022882e-05,
+      "loss": 1.5386,
+      "step": 830
+    },
+    {
+      "epoch": 0.5121725731895224,
+      "grad_norm": 0.6561502814292908,
+      "learning_rate": 2.2640764210795187e-05,
+      "loss": 1.4534,
+      "step": 831
+    },
+    {
+      "epoch": 0.512788906009245,
+      "grad_norm": 0.6672171354293823,
+      "learning_rate": 2.2598070159700043e-05,
+      "loss": 1.4285,
+      "step": 832
+    },
+    {
+      "epoch": 0.5134052388289676,
+      "grad_norm": 0.6675829291343689,
+      "learning_rate": 2.2555364062687844e-05,
+      "loss": 1.4009,
+      "step": 833
+    },
+    {
+      "epoch": 0.5140215716486903,
+      "grad_norm": 0.6768284440040588,
+      "learning_rate": 2.2512646117764828e-05,
+      "loss": 1.3978,
+      "step": 834
+    },
+    {
+      "epoch": 0.514637904468413,
+      "grad_norm": 0.6675807237625122,
+      "learning_rate": 2.2469916522992155e-05,
+      "loss": 1.4627,
+      "step": 835
+    },
+    {
+      "epoch": 0.5152542372881356,
+      "grad_norm": 0.6885505318641663,
+      "learning_rate": 2.242717547648502e-05,
+      "loss": 1.4702,
+      "step": 836
+    },
+    {
+      "epoch": 0.5158705701078582,
+      "grad_norm": 0.6633325219154358,
+      "learning_rate": 2.238442317641171e-05,
+      "loss": 1.4615,
+      "step": 837
+    },
+    {
+      "epoch": 0.5164869029275809,
+      "grad_norm": 0.6807160973548889,
+      "learning_rate": 2.234165982099268e-05,
+      "loss": 1.5114,
+      "step": 838
+    },
+    {
+      "epoch": 0.5171032357473035,
+      "grad_norm": 0.6771484017372131,
+      "learning_rate": 2.229888560849964e-05,
+      "loss": 1.4715,
+      "step": 839
+    },
+    {
+      "epoch": 0.5177195685670262,
+      "grad_norm": 0.6473548412322998,
+      "learning_rate": 2.2256100737254657e-05,
+      "loss": 1.3995,
+      "step": 840
+    },
+    {
+      "epoch": 0.5183359013867489,
+      "grad_norm": 0.7034615874290466,
+      "learning_rate": 2.22133054056292e-05,
+      "loss": 1.5146,
+      "step": 841
+    },
+    {
+      "epoch": 0.5189522342064715,
+      "grad_norm": 0.7064915895462036,
+      "learning_rate": 2.217049981204325e-05,
+      "loss": 1.4643,
+      "step": 842
+    },
+    {
+      "epoch": 0.5195685670261941,
+      "grad_norm": 0.859814465045929,
+      "learning_rate": 2.2127684154964357e-05,
+      "loss": 1.4419,
+      "step": 843
+    },
+    {
+      "epoch": 0.5201848998459168,
+      "grad_norm": 0.6511772871017456,
+      "learning_rate": 2.2084858632906733e-05,
+      "loss": 1.4144,
+      "step": 844
+    },
+    {
+      "epoch": 0.5208012326656395,
+      "grad_norm": 0.6685275435447693,
+      "learning_rate": 2.2042023444430333e-05,
+      "loss": 1.4596,
+      "step": 845
+    },
+    {
+      "epoch": 0.5214175654853621,
+      "grad_norm": 0.6598917841911316,
+      "learning_rate": 2.199917878813992e-05,
+      "loss": 1.4823,
+      "step": 846
+    },
+    {
+      "epoch": 0.5220338983050847,
+      "grad_norm": 0.6725688576698303,
+      "learning_rate": 2.195632486268418e-05,
+      "loss": 1.5023,
+      "step": 847
+    },
+    {
+      "epoch": 0.5226502311248074,
+      "grad_norm": 0.7149176597595215,
+      "learning_rate": 2.1913461866754747e-05,
+      "loss": 1.4431,
+      "step": 848
+    },
+    {
+      "epoch": 0.5232665639445301,
+      "grad_norm": 0.6676431894302368,
+      "learning_rate": 2.1870589999085302e-05,
+      "loss": 1.3993,
+      "step": 849
+    },
+    {
+      "epoch": 0.5238828967642527,
+      "grad_norm": 0.6679561734199524,
+      "learning_rate": 2.182770945845071e-05,
+      "loss": 1.4462,
+      "step": 850
+    },
+    {
+      "epoch": 0.5244992295839753,
+      "grad_norm": 0.6490969061851501,
+      "learning_rate": 2.178482044366599e-05,
+      "loss": 1.435,
+      "step": 851
+    },
+    {
+      "epoch": 0.5251155624036979,
+      "grad_norm": 0.6597004532814026,
+      "learning_rate": 2.1741923153585497e-05,
+      "loss": 1.4516,
+      "step": 852
+    },
+    {
+      "epoch": 0.5257318952234207,
+      "grad_norm": 0.6608878374099731,
+      "learning_rate": 2.1699017787101924e-05,
+      "loss": 1.4543,
+      "step": 853
+    },
+    {
+      "epoch": 0.5263482280431433,
+      "grad_norm": 0.6896352767944336,
+      "learning_rate": 2.165610454314542e-05,
+      "loss": 1.4563,
+      "step": 854
+    },
+    {
+      "epoch": 0.5269645608628659,
+      "grad_norm": 0.6917398571968079,
+      "learning_rate": 2.1613183620682662e-05,
+      "loss": 1.3637,
+      "step": 855
+    },
+    {
+      "epoch": 0.5275808936825886,
+      "grad_norm": 0.6737967133522034,
+      "learning_rate": 2.157025521871593e-05,
+      "loss": 1.3772,
+      "step": 856
+    },
+    {
+      "epoch": 0.5281972265023113,
+      "grad_norm": 0.6648988723754883,
+      "learning_rate": 2.152731953628216e-05,
+      "loss": 1.4338,
+      "step": 857
+    },
+    {
+      "epoch": 0.5288135593220339,
+      "grad_norm": 0.6846292018890381,
+      "learning_rate": 2.1484376772452087e-05,
+      "loss": 1.499,
+      "step": 858
+    },
+    {
+      "epoch": 0.5294298921417565,
+      "grad_norm": 0.661718487739563,
+      "learning_rate": 2.144142712632923e-05,
+      "loss": 1.4039,
+      "step": 859
+    },
+    {
+      "epoch": 0.5300462249614792,
+      "grad_norm": 0.679577648639679,
+      "learning_rate": 2.1398470797049055e-05,
+      "loss": 1.4868,
+      "step": 860
+    },
+    {
+      "epoch": 0.5306625577812019,
+      "grad_norm": 0.6697185635566711,
+      "learning_rate": 2.1355507983778e-05,
+      "loss": 1.4118,
+      "step": 861
+    },
+    {
+      "epoch": 0.5312788906009245,
+      "grad_norm": 0.6589436531066895,
+      "learning_rate": 2.131253888571255e-05,
+      "loss": 1.4483,
+      "step": 862
+    },
+    {
+      "epoch": 0.5318952234206471,
+      "grad_norm": 0.6604962348937988,
+      "learning_rate": 2.126956370207836e-05,
+      "loss": 1.4177,
+      "step": 863
+    },
+    {
+      "epoch": 0.5325115562403698,
+      "grad_norm": 0.6640836596488953,
+      "learning_rate": 2.1226582632129286e-05,
+      "loss": 1.4351,
+      "step": 864
+    },
+    {
+      "epoch": 0.5331278890600925,
+      "grad_norm": 0.659049928188324,
+      "learning_rate": 2.1183595875146467e-05,
+      "loss": 1.4109,
+      "step": 865
+    },
+    {
+      "epoch": 0.5337442218798151,
+      "grad_norm": 0.6786544322967529,
+      "learning_rate": 2.1140603630437434e-05,
+      "loss": 1.4525,
+      "step": 866
+    },
+    {
+      "epoch": 0.5343605546995378,
+      "grad_norm": 0.6710315346717834,
+      "learning_rate": 2.1097606097335133e-05,
+      "loss": 1.4424,
+      "step": 867
+    },
+    {
+      "epoch": 0.5349768875192604,
+      "grad_norm": 0.658591091632843,
+      "learning_rate": 2.1054603475197045e-05,
+      "loss": 1.4215,
+      "step": 868
+    },
+    {
+      "epoch": 0.535593220338983,
+      "grad_norm": 0.6709784865379333,
+      "learning_rate": 2.1011595963404253e-05,
+      "loss": 1.4753,
+      "step": 869
+    },
+    {
+      "epoch": 0.5362095531587057,
+      "grad_norm": 0.6819916367530823,
+      "learning_rate": 2.096858376136049e-05,
+      "loss": 1.4439,
+      "step": 870
+    },
+    {
+      "epoch": 0.5368258859784284,
+      "grad_norm": 0.6897315382957458,
+      "learning_rate": 2.092556706849125e-05,
+      "loss": 1.4687,
+      "step": 871
+    },
+    {
+      "epoch": 0.537442218798151,
+      "grad_norm": 0.6903764009475708,
+      "learning_rate": 2.0882546084242856e-05,
+      "loss": 1.4768,
+      "step": 872
+    },
+    {
+      "epoch": 0.5380585516178736,
+      "grad_norm": 0.677721381187439,
+      "learning_rate": 2.0839521008081508e-05,
+      "loss": 1.4324,
+      "step": 873
+    },
+    {
+      "epoch": 0.5386748844375963,
+      "grad_norm": 0.6687971353530884,
+      "learning_rate": 2.0796492039492394e-05,
+      "loss": 1.4655,
+      "step": 874
+    },
+    {
+      "epoch": 0.539291217257319,
+      "grad_norm": 0.6661518216133118,
+      "learning_rate": 2.0753459377978737e-05,
+      "loss": 1.4712,
+      "step": 875
+    },
+    {
+      "epoch": 0.5399075500770416,
+      "grad_norm": 0.6647510528564453,
+      "learning_rate": 2.071042322306089e-05,
+      "loss": 1.423,
+      "step": 876
+    },
+    {
+      "epoch": 0.5405238828967642,
+      "grad_norm": 0.7283526062965393,
+      "learning_rate": 2.066738377427541e-05,
+      "loss": 1.4547,
+      "step": 877
+    },
+    {
+      "epoch": 0.5411402157164868,
+      "grad_norm": 0.679385781288147,
+      "learning_rate": 2.0624341231174113e-05,
+      "loss": 1.4217,
+      "step": 878
+    },
+    {
+      "epoch": 0.5417565485362096,
+      "grad_norm": 0.6685990691184998,
+      "learning_rate": 2.0581295793323156e-05,
+      "loss": 1.4135,
+      "step": 879
+    },
+    {
+      "epoch": 0.5423728813559322,
+      "grad_norm": 0.6681914329528809,
+      "learning_rate": 2.0538247660302143e-05,
+      "loss": 1.4704,
+      "step": 880
+    },
+    {
+      "epoch": 0.5429892141756548,
+      "grad_norm": 0.6571676135063171,
+      "learning_rate": 2.0495197031703154e-05,
+      "loss": 1.3697,
+      "step": 881
+    },
+    {
+      "epoch": 0.5436055469953776,
+      "grad_norm": 0.6493473052978516,
+      "learning_rate": 2.0452144107129844e-05,
+      "loss": 1.4209,
+      "step": 882
+    },
+    {
+      "epoch": 0.5442218798151002,
+      "grad_norm": 0.6795789003372192,
+      "learning_rate": 2.0409089086196506e-05,
+      "loss": 1.451,
+      "step": 883
+    },
+    {
+      "epoch": 0.5448382126348228,
+      "grad_norm": 0.6869490742683411,
+      "learning_rate": 2.0366032168527178e-05,
+      "loss": 1.5008,
+      "step": 884
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.6746764183044434,
+      "learning_rate": 2.032297355375465e-05,
+      "loss": 1.3978,
+      "step": 885
+    },
+    {
+      "epoch": 0.5460708782742681,
+      "grad_norm": 0.6667096018791199,
+      "learning_rate": 2.027991344151963e-05,
+      "loss": 1.4117,
+      "step": 886
+    },
+    {
+      "epoch": 0.5466872110939908,
+      "grad_norm": 0.6780974268913269,
+      "learning_rate": 2.0236852031469735e-05,
+      "loss": 1.4576,
+      "step": 887
+    },
+    {
+      "epoch": 0.5473035439137134,
+      "grad_norm": 0.6567050814628601,
+      "learning_rate": 2.0193789523258608e-05,
+      "loss": 1.4527,
+      "step": 888
+    },
+    {
+      "epoch": 0.547919876733436,
+      "grad_norm": 0.6834569573402405,
+      "learning_rate": 2.0150726116544982e-05,
+      "loss": 1.4345,
+      "step": 889
+    },
+    {
+      "epoch": 0.5485362095531587,
+      "grad_norm": 0.6610395908355713,
+      "learning_rate": 2.0107662010991764e-05,
+      "loss": 1.4347,
+      "step": 890
+    },
+    {
+      "epoch": 0.5491525423728814,
+      "grad_norm": 0.6741618514060974,
+      "learning_rate": 2.006459740626509e-05,
+      "loss": 1.452,
+      "step": 891
+    },
+    {
+      "epoch": 0.549768875192604,
+      "grad_norm": 0.6818397045135498,
+      "learning_rate": 2.002153250203342e-05,
+      "loss": 1.408,
+      "step": 892
+    },
+    {
+      "epoch": 0.5503852080123267,
+      "grad_norm": 0.6544389128684998,
+      "learning_rate": 1.997846749796659e-05,
+      "loss": 1.4037,
+      "step": 893
+    },
+    {
+      "epoch": 0.5510015408320493,
+      "grad_norm": 0.6719116568565369,
+      "learning_rate": 1.9935402593734915e-05,
+      "loss": 1.4347,
+      "step": 894
+    },
+    {
+      "epoch": 0.551617873651772,
+      "grad_norm": 0.6798861026763916,
+      "learning_rate": 1.9892337989008243e-05,
+      "loss": 1.4861,
+      "step": 895
+    },
+    {
+      "epoch": 0.5522342064714946,
+      "grad_norm": 0.6727917790412903,
+      "learning_rate": 1.984927388345502e-05,
+      "loss": 1.4481,
+      "step": 896
+    },
+    {
+      "epoch": 0.5528505392912173,
+      "grad_norm": 0.6841053366661072,
+      "learning_rate": 1.9806210476741396e-05,
+      "loss": 1.4536,
+      "step": 897
+    },
+    {
+      "epoch": 0.5534668721109399,
+      "grad_norm": 0.6897479891777039,
+      "learning_rate": 1.976314796853027e-05,
+      "loss": 1.4889,
+      "step": 898
+    },
+    {
+      "epoch": 0.5540832049306625,
+      "grad_norm": 0.657565176486969,
+      "learning_rate": 1.9720086558480372e-05,
+      "loss": 1.4495,
+      "step": 899
+    },
+    {
+      "epoch": 0.5546995377503852,
+      "grad_norm": 0.6826606392860413,
+      "learning_rate": 1.9677026446245352e-05,
+      "loss": 1.4127,
+      "step": 900
+    },
+    {
+      "epoch": 0.5553158705701079,
+      "grad_norm": 0.6761550307273865,
+      "learning_rate": 1.9633967831472835e-05,
+      "loss": 1.4465,
+      "step": 901
+    },
+    {
+      "epoch": 0.5559322033898305,
+      "grad_norm": 0.6841764450073242,
+      "learning_rate": 1.95909109138035e-05,
+      "loss": 1.4282,
+      "step": 902
+    },
+    {
+      "epoch": 0.5565485362095531,
+      "grad_norm": 0.6865301728248596,
+      "learning_rate": 1.9547855892870166e-05,
+      "loss": 1.4903,
+      "step": 903
+    },
+    {
+      "epoch": 0.5571648690292758,
+      "grad_norm": 0.6651366353034973,
+      "learning_rate": 1.9504802968296853e-05,
+      "loss": 1.4387,
+      "step": 904
+    },
+    {
+      "epoch": 0.5577812018489985,
+      "grad_norm": 0.6853525042533875,
+      "learning_rate": 1.9461752339697863e-05,
+      "loss": 1.4799,
+      "step": 905
+    },
+    {
+      "epoch": 0.5583975346687211,
+      "grad_norm": 0.6708539724349976,
+      "learning_rate": 1.9418704206676847e-05,
+      "loss": 1.4576,
+      "step": 906
+    },
+    {
+      "epoch": 0.5590138674884437,
+      "grad_norm": 0.6524003744125366,
+      "learning_rate": 1.937565876882589e-05,
+      "loss": 1.4474,
+      "step": 907
+    },
+    {
+      "epoch": 0.5596302003081665,
+      "grad_norm": 0.660274863243103,
+      "learning_rate": 1.93326162257246e-05,
+      "loss": 1.4079,
+      "step": 908
+    },
+    {
+      "epoch": 0.5602465331278891,
+      "grad_norm": 0.665828287601471,
+      "learning_rate": 1.9289576776939114e-05,
+      "loss": 1.4248,
+      "step": 909
+    },
+    {
+      "epoch": 0.5608628659476117,
+      "grad_norm": 0.6802533268928528,
+      "learning_rate": 1.9246540622021266e-05,
+      "loss": 1.4001,
+      "step": 910
+    },
+    {
+      "epoch": 0.5614791987673343,
+      "grad_norm": 0.674877941608429,
+      "learning_rate": 1.920350796050762e-05,
+      "loss": 1.4604,
+      "step": 911
+    },
+    {
+      "epoch": 0.562095531587057,
+      "grad_norm": 0.6802154183387756,
+      "learning_rate": 1.9160478991918496e-05,
+      "loss": 1.441,
+      "step": 912
+    },
+    {
+      "epoch": 0.5627118644067797,
+      "grad_norm": 0.6677930951118469,
+      "learning_rate": 1.9117453915757148e-05,
+      "loss": 1.3754,
+      "step": 913
+    },
+    {
+      "epoch": 0.5633281972265023,
+      "grad_norm": 0.683760941028595,
+      "learning_rate": 1.907443293150876e-05,
+      "loss": 1.4508,
+      "step": 914
+    },
+    {
+      "epoch": 0.5639445300462249,
+      "grad_norm": 0.6907650232315063,
+      "learning_rate": 1.9031416238639518e-05,
+      "loss": 1.4885,
+      "step": 915
+    },
+    {
+      "epoch": 0.5645608628659476,
+      "grad_norm": 0.672909140586853,
+      "learning_rate": 1.8988404036595757e-05,
+      "loss": 1.3758,
+      "step": 916
+    },
+    {
+      "epoch": 0.5651771956856703,
+      "grad_norm": 0.6761062741279602,
+      "learning_rate": 1.894539652480296e-05,
+      "loss": 1.4423,
+      "step": 917
+    },
+    {
+      "epoch": 0.5657935285053929,
+      "grad_norm": 0.7031134366989136,
+      "learning_rate": 1.8902393902664873e-05,
+      "loss": 1.3982,
+      "step": 918
+    },
+    {
+      "epoch": 0.5664098613251156,
+      "grad_norm": 0.6689425706863403,
+      "learning_rate": 1.885939636956257e-05,
+      "loss": 1.4437,
+      "step": 919
+    },
+    {
+      "epoch": 0.5670261941448382,
+      "grad_norm": 0.6898466944694519,
+      "learning_rate": 1.881640412485353e-05,
+      "loss": 1.4017,
+      "step": 920
+    },
+    {
+      "epoch": 0.5676425269645609,
+      "grad_norm": 0.6723157167434692,
+      "learning_rate": 1.877341736787072e-05,
+      "loss": 1.4168,
+      "step": 921
+    },
+    {
+      "epoch": 0.5682588597842835,
+      "grad_norm": 0.6810134649276733,
+      "learning_rate": 1.8730436297921645e-05,
+      "loss": 1.4349,
+      "step": 922
+    },
+    {
+      "epoch": 0.5688751926040062,
+      "grad_norm": 0.6878021955490112,
+      "learning_rate": 1.8687461114287455e-05,
+      "loss": 1.333,
+      "step": 923
+    },
+    {
+      "epoch": 0.5694915254237288,
+      "grad_norm": 0.7006982564926147,
+      "learning_rate": 1.8644492016222015e-05,
+      "loss": 1.4425,
+      "step": 924
+    },
+    {
+      "epoch": 0.5701078582434514,
+      "grad_norm": 0.6941596269607544,
+      "learning_rate": 1.8601529202950948e-05,
+      "loss": 1.4304,
+      "step": 925
+    },
+    {
+      "epoch": 0.5707241910631741,
+      "grad_norm": 0.6846546530723572,
+      "learning_rate": 1.8558572873670774e-05,
+      "loss": 1.4505,
+      "step": 926
+    },
+    {
+      "epoch": 0.5713405238828968,
+      "grad_norm": 0.6761274337768555,
+      "learning_rate": 1.8515623227547923e-05,
+      "loss": 1.4106,
+      "step": 927
+    },
+    {
+      "epoch": 0.5719568567026194,
+      "grad_norm": 0.6644502878189087,
+      "learning_rate": 1.8472680463717842e-05,
+      "loss": 1.4591,
+      "step": 928
+    },
+    {
+      "epoch": 0.572573189522342,
+      "grad_norm": 0.6804108023643494,
+      "learning_rate": 1.8429744781284072e-05,
+      "loss": 1.4747,
+      "step": 929
+    },
+    {
+      "epoch": 0.5731895223420647,
+      "grad_norm": 0.6691780686378479,
+      "learning_rate": 1.8386816379317344e-05,
+      "loss": 1.4035,
+      "step": 930
+    },
+    {
+      "epoch": 0.5738058551617874,
+      "grad_norm": 0.7028989195823669,
+      "learning_rate": 1.8343895456854588e-05,
+      "loss": 1.448,
+      "step": 931
+    },
+    {
+      "epoch": 0.57442218798151,
+      "grad_norm": 0.6800969243049622,
+      "learning_rate": 1.830098221289808e-05,
+      "loss": 1.3695,
+      "step": 932
+    },
+    {
+      "epoch": 0.5750385208012326,
+      "grad_norm": 0.6787469983100891,
+      "learning_rate": 1.8258076846414506e-05,
+      "loss": 1.4078,
+      "step": 933
+    },
+    {
+      "epoch": 0.5756548536209554,
+      "grad_norm": 0.6691258549690247,
+      "learning_rate": 1.8215179556334012e-05,
+      "loss": 1.3818,
+      "step": 934
+    },
+    {
+      "epoch": 0.576271186440678,
+      "grad_norm": 0.677056074142456,
+      "learning_rate": 1.8172290541549297e-05,
+      "loss": 1.42,
+      "step": 935
+    },
+    {
+      "epoch": 0.5768875192604006,
+      "grad_norm": 0.6807578206062317,
+      "learning_rate": 1.8129410000914698e-05,
+      "loss": 1.4405,
+      "step": 936
+    },
+    {
+      "epoch": 0.5775038520801232,
+      "grad_norm": 0.6643014550209045,
+      "learning_rate": 1.8086538133245267e-05,
+      "loss": 1.3877,
+      "step": 937
+    },
+    {
+      "epoch": 0.578120184899846,
+      "grad_norm": 0.6755670309066772,
+      "learning_rate": 1.8043675137315824e-05,
+      "loss": 1.4315,
+      "step": 938
+    },
+    {
+      "epoch": 0.5787365177195686,
+      "grad_norm": 0.6702075004577637,
+      "learning_rate": 1.8000821211860077e-05,
+      "loss": 1.4442,
+      "step": 939
+    },
+    {
+      "epoch": 0.5793528505392912,
+      "grad_norm": 0.6891489028930664,
+      "learning_rate": 1.7957976555569677e-05,
+      "loss": 1.4554,
+      "step": 940
+    },
+    {
+      "epoch": 0.5799691833590138,
+      "grad_norm": 0.6698282361030579,
+      "learning_rate": 1.7915141367093274e-05,
+      "loss": 1.4076,
+      "step": 941
+    },
+    {
+      "epoch": 0.5805855161787365,
+      "grad_norm": 0.6736328601837158,
+      "learning_rate": 1.787231584503565e-05,
+      "loss": 1.4102,
+      "step": 942
+    },
+    {
+      "epoch": 0.5812018489984592,
+      "grad_norm": 0.6851081848144531,
+      "learning_rate": 1.782950018795676e-05,
+      "loss": 1.4017,
+      "step": 943
+    },
+    {
+      "epoch": 0.5818181818181818,
+      "grad_norm": 0.6863130331039429,
+      "learning_rate": 1.7786694594370806e-05,
+      "loss": 1.4394,
+      "step": 944
+    },
+    {
+      "epoch": 0.5824345146379045,
+      "grad_norm": 0.6876046061515808,
+      "learning_rate": 1.7743899262745353e-05,
+      "loss": 1.3686,
+      "step": 945
+    },
+    {
+      "epoch": 0.5830508474576271,
+      "grad_norm": 0.6842202544212341,
+      "learning_rate": 1.7701114391500364e-05,
+      "loss": 1.3856,
+      "step": 946
+    },
+    {
+      "epoch": 0.5836671802773498,
+      "grad_norm": 0.6818761229515076,
+      "learning_rate": 1.7658340179007335e-05,
+      "loss": 1.3733,
+      "step": 947
+    },
+    {
+      "epoch": 0.5842835130970724,
+      "grad_norm": 0.6609686613082886,
+      "learning_rate": 1.7615576823588294e-05,
+      "loss": 1.3586,
+      "step": 948
+    },
+    {
+      "epoch": 0.5848998459167951,
+      "grad_norm": 0.6683045625686646,
+      "learning_rate": 1.7572824523514982e-05,
+      "loss": 1.454,
+      "step": 949
+    },
+    {
+      "epoch": 0.5855161787365177,
+      "grad_norm": 0.6885985732078552,
+      "learning_rate": 1.7530083477007855e-05,
+      "loss": 1.443,
+      "step": 950
+    },
+    {
+      "epoch": 0.5861325115562404,
+      "grad_norm": 0.67160564661026,
+      "learning_rate": 1.748735388223518e-05,
+      "loss": 1.4051,
+      "step": 951
+    },
+    {
+      "epoch": 0.586748844375963,
+      "grad_norm": 0.6979023218154907,
+      "learning_rate": 1.7444635937312156e-05,
+      "loss": 1.4371,
+      "step": 952
+    },
+    {
+      "epoch": 0.5873651771956857,
+      "grad_norm": 0.6684184074401855,
+      "learning_rate": 1.740192984029996e-05,
+      "loss": 1.4285,
+      "step": 953
+    },
+    {
+      "epoch": 0.5879815100154083,
+      "grad_norm": 0.6790677309036255,
+      "learning_rate": 1.735923578920482e-05,
+      "loss": 1.4607,
+      "step": 954
+    },
+    {
+      "epoch": 0.588597842835131,
+      "grad_norm": 0.692985475063324,
+      "learning_rate": 1.7316553981977128e-05,
+      "loss": 1.3717,
+      "step": 955
+    },
+    {
+      "epoch": 0.5892141756548536,
+      "grad_norm": 0.6927673816680908,
+      "learning_rate": 1.7273884616510505e-05,
+      "loss": 1.4044,
+      "step": 956
+    },
+    {
+      "epoch": 0.5898305084745763,
+      "grad_norm": 0.686356246471405,
+      "learning_rate": 1.7231227890640893e-05,
+      "loss": 1.3903,
+      "step": 957
+    },
+    {
+      "epoch": 0.5904468412942989,
+      "grad_norm": 0.6897464394569397,
+      "learning_rate": 1.7188584002145612e-05,
+      "loss": 1.4586,
+      "step": 958
+    },
+    {
+      "epoch": 0.5910631741140215,
+      "grad_norm": 0.6666277050971985,
+      "learning_rate": 1.7145953148742478e-05,
+      "loss": 1.4593,
+      "step": 959
+    },
+    {
+      "epoch": 0.5916795069337443,
+      "grad_norm": 0.6714256405830383,
+      "learning_rate": 1.7103335528088866e-05,
+      "loss": 1.3904,
+      "step": 960
+    },
+    {
+      "epoch": 0.5922958397534669,
+      "grad_norm": 0.6808894872665405,
+      "learning_rate": 1.7060731337780794e-05,
+      "loss": 1.4326,
+      "step": 961
+    },
+    {
+      "epoch": 0.5929121725731895,
+      "grad_norm": 0.6975859999656677,
+      "learning_rate": 1.7018140775352007e-05,
+      "loss": 1.4409,
+      "step": 962
+    },
+    {
+      "epoch": 0.5935285053929121,
+      "grad_norm": 0.6830939054489136,
+      "learning_rate": 1.6975564038273084e-05,
+      "loss": 1.3683,
+      "step": 963
+    },
+    {
+      "epoch": 0.5941448382126349,
+      "grad_norm": 0.6693740487098694,
+      "learning_rate": 1.6933001323950476e-05,
+      "loss": 1.3316,
+      "step": 964
+    },
+    {
+      "epoch": 0.5947611710323575,
+      "grad_norm": 0.6849698424339294,
+      "learning_rate": 1.689045282972563e-05,
+      "loss": 1.3855,
+      "step": 965
+    },
+    {
+      "epoch": 0.5953775038520801,
+      "grad_norm": 0.7089459300041199,
+      "learning_rate": 1.6847918752874082e-05,
+      "loss": 1.4248,
+      "step": 966
+    },
+    {
+      "epoch": 0.5959938366718027,
+      "grad_norm": 0.6675121784210205,
+      "learning_rate": 1.680539929060448e-05,
+      "loss": 1.4441,
+      "step": 967
+    },
+    {
+      "epoch": 0.5966101694915255,
+      "grad_norm": 0.7075740098953247,
+      "learning_rate": 1.6762894640057735e-05,
+      "loss": 1.4399,
+      "step": 968
+    },
+    {
+      "epoch": 0.5972265023112481,
+      "grad_norm": 0.6729096174240112,
+      "learning_rate": 1.6720404998306094e-05,
+      "loss": 1.4364,
+      "step": 969
+    },
+    {
+      "epoch": 0.5978428351309707,
+      "grad_norm": 0.6988831162452698,
+      "learning_rate": 1.6677930562352208e-05,
+      "loss": 1.4147,
+      "step": 970
+    },
+    {
+      "epoch": 0.5984591679506934,
+      "grad_norm": 0.6925584673881531,
+      "learning_rate": 1.6635471529128215e-05,
+      "loss": 1.4554,
+      "step": 971
+    },
+    {
+      "epoch": 0.599075500770416,
+      "grad_norm": 0.6823211312294006,
+      "learning_rate": 1.6593028095494842e-05,
+      "loss": 1.4105,
+      "step": 972
+    },
+    {
+      "epoch": 0.5996918335901387,
+      "grad_norm": 0.6767581105232239,
+      "learning_rate": 1.655060045824051e-05,
+      "loss": 1.4272,
+      "step": 973
+    },
+    {
+      "epoch": 0.6003081664098613,
+      "grad_norm": 0.6897989511489868,
+      "learning_rate": 1.6508188814080367e-05,
+      "loss": 1.4132,
+      "step": 974
+    },
+    {
+      "epoch": 0.600924499229584,
+      "grad_norm": 0.6938064694404602,
+      "learning_rate": 1.6465793359655433e-05,
+      "loss": 1.4464,
+      "step": 975
+    },
+    {
+      "epoch": 0.6015408320493066,
+      "grad_norm": 0.6732364892959595,
+      "learning_rate": 1.6423414291531655e-05,
+      "loss": 1.4384,
+      "step": 976
+    },
+    {
+      "epoch": 0.6021571648690293,
+      "grad_norm": 0.6800574660301208,
+      "learning_rate": 1.638105180619901e-05,
+      "loss": 1.4428,
+      "step": 977
+    },
+    {
+      "epoch": 0.6027734976887519,
+      "grad_norm": 0.6989326477050781,
+      "learning_rate": 1.6338706100070586e-05,
+      "loss": 1.4086,
+      "step": 978
+    },
+    {
+      "epoch": 0.6033898305084746,
+      "grad_norm": 0.6868564486503601,
+      "learning_rate": 1.629637736948168e-05,
+      "loss": 1.4247,
+      "step": 979
+    },
+    {
+      "epoch": 0.6040061633281972,
+      "grad_norm": 0.6902734041213989,
+      "learning_rate": 1.6254065810688866e-05,
+      "loss": 1.3884,
+      "step": 980
+    },
+    {
+      "epoch": 0.6046224961479199,
+      "grad_norm": 0.6859354972839355,
+      "learning_rate": 1.6211771619869123e-05,
+      "loss": 1.3885,
+      "step": 981
+    },
+    {
+      "epoch": 0.6052388289676425,
+      "grad_norm": 0.6829075813293457,
+      "learning_rate": 1.616949499311888e-05,
+      "loss": 1.4694,
+      "step": 982
+    },
+    {
+      "epoch": 0.6058551617873652,
+      "grad_norm": 0.6718934774398804,
+      "learning_rate": 1.6127236126453168e-05,
+      "loss": 1.4215,
+      "step": 983
+    },
+    {
+      "epoch": 0.6064714946070878,
+      "grad_norm": 0.6713712215423584,
+      "learning_rate": 1.608499521580462e-05,
+      "loss": 1.4091,
+      "step": 984
+    },
+    {
+      "epoch": 0.6070878274268104,
+      "grad_norm": 0.6932412385940552,
+      "learning_rate": 1.6042772457022666e-05,
+      "loss": 1.4868,
+      "step": 985
+    },
+    {
+      "epoch": 0.6077041602465332,
+      "grad_norm": 0.7100671529769897,
+      "learning_rate": 1.6000568045872556e-05,
+      "loss": 1.4582,
+      "step": 986
+    },
+    {
+      "epoch": 0.6083204930662558,
+      "grad_norm": 0.6786303520202637,
+      "learning_rate": 1.5958382178034447e-05,
+      "loss": 1.4771,
+      "step": 987
+    },
+    {
+      "epoch": 0.6089368258859784,
+      "grad_norm": 0.6749539375305176,
+      "learning_rate": 1.5916215049102564e-05,
+      "loss": 1.3796,
+      "step": 988
+    },
+    {
+      "epoch": 0.609553158705701,
+      "grad_norm": 0.6852856874465942,
+      "learning_rate": 1.5874066854584224e-05,
+      "loss": 1.3981,
+      "step": 989
+    },
+    {
+      "epoch": 0.6101694915254238,
+      "grad_norm": 0.6997564435005188,
+      "learning_rate": 1.5831937789898957e-05,
+      "loss": 1.4493,
+      "step": 990
+    },
+    {
+      "epoch": 0.6107858243451464,
+      "grad_norm": 0.6845008134841919,
+      "learning_rate": 1.57898280503776e-05,
+      "loss": 1.3884,
+      "step": 991
+    },
+    {
+      "epoch": 0.611402157164869,
+      "grad_norm": 0.7117551565170288,
+      "learning_rate": 1.5747737831261393e-05,
+      "loss": 1.453,
+      "step": 992
+    },
+    {
+      "epoch": 0.6120184899845916,
+      "grad_norm": 0.6957306861877441,
+      "learning_rate": 1.5705667327701056e-05,
+      "loss": 1.4064,
+      "step": 993
+    },
+    {
+      "epoch": 0.6126348228043144,
+      "grad_norm": 0.6860647797584534,
+      "learning_rate": 1.566361673475592e-05,
+      "loss": 1.3506,
+      "step": 994
+    },
+    {
+      "epoch": 0.613251155624037,
+      "grad_norm": 0.6879005432128906,
+      "learning_rate": 1.562158624739298e-05,
+      "loss": 1.4843,
+      "step": 995
+    },
+    {
+      "epoch": 0.6138674884437596,
+      "grad_norm": 0.7000822424888611,
+      "learning_rate": 1.557957606048603e-05,
+      "loss": 1.407,
+      "step": 996
+    },
+    {
+      "epoch": 0.6144838212634823,
+      "grad_norm": 0.698491632938385,
+      "learning_rate": 1.5537586368814727e-05,
+      "loss": 1.3917,
+      "step": 997
+    },
+    {
+      "epoch": 0.615100154083205,
+      "grad_norm": 0.7287001609802246,
+      "learning_rate": 1.54956173670637e-05,
+      "loss": 1.4017,
+      "step": 998
+    },
+    {
+      "epoch": 0.6157164869029276,
+      "grad_norm": 0.7118046879768372,
+      "learning_rate": 1.545366924982167e-05,
+      "loss": 1.402,
+      "step": 999
+    },
+    {
+      "epoch": 0.6163328197226502,
+      "grad_norm": 0.6940359473228455,
+      "learning_rate": 1.5411742211580504e-05,
+      "loss": 1.3956,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6169491525423729,
+      "grad_norm": 0.6867558360099792,
+      "learning_rate": 1.5369836446734343e-05,
+      "loss": 1.4436,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6175654853620955,
+      "grad_norm": 0.6873385310173035,
+      "learning_rate": 1.532795214957872e-05,
+      "loss": 1.4537,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6181818181818182,
+      "grad_norm": 0.6997383832931519,
+      "learning_rate": 1.528608951430958e-05,
+      "loss": 1.463,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6187981510015408,
+      "grad_norm": 0.687658965587616,
+      "learning_rate": 1.5244248735022473e-05,
+      "loss": 1.4255,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6194144838212635,
+      "grad_norm": 0.6884608268737793,
+      "learning_rate": 1.5202430005711625e-05,
+      "loss": 1.4706,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6200308166409861,
+      "grad_norm": 0.7008056640625,
+      "learning_rate": 1.5160633520268978e-05,
+      "loss": 1.5052,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6206471494607088,
+      "grad_norm": 0.7026891708374023,
+      "learning_rate": 1.5118859472483389e-05,
+      "loss": 1.441,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6212634822804314,
+      "grad_norm": 0.6971073150634766,
+      "learning_rate": 1.507710805603966e-05,
+      "loss": 1.3532,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6218798151001541,
+      "grad_norm": 0.7128373384475708,
+      "learning_rate": 1.5035379464517667e-05,
+      "loss": 1.4875,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6224961479198767,
+      "grad_norm": 0.6831132173538208,
+      "learning_rate": 1.4993673891391456e-05,
+      "loss": 1.4559,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6231124807395994,
+      "grad_norm": 0.6850417852401733,
+      "learning_rate": 1.4951991530028352e-05,
+      "loss": 1.3441,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6237288135593221,
+      "grad_norm": 0.6863908767700195,
+      "learning_rate": 1.4910332573688062e-05,
+      "loss": 1.4899,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6243451463790447,
+      "grad_norm": 0.6850789785385132,
+      "learning_rate": 1.4868697215521764e-05,
+      "loss": 1.4014,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6249614791987673,
+      "grad_norm": 0.7038957476615906,
+      "learning_rate": 1.482708564857124e-05,
+      "loss": 1.486,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6255778120184899,
+      "grad_norm": 0.6841074824333191,
+      "learning_rate": 1.4785498065767953e-05,
+      "loss": 1.3738,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6261941448382127,
+      "grad_norm": 0.7064815759658813,
+      "learning_rate": 1.4743934659932162e-05,
+      "loss": 1.5064,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6268104776579353,
+      "grad_norm": 0.6969325542449951,
+      "learning_rate": 1.4702395623772042e-05,
+      "loss": 1.4584,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6274268104776579,
+      "grad_norm": 0.681100606918335,
+      "learning_rate": 1.4660881149882768e-05,
+      "loss": 1.4532,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6280431432973805,
+      "grad_norm": 0.6736499071121216,
+      "learning_rate": 1.461939143074563e-05,
+      "loss": 1.4733,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6286594761171033,
+      "grad_norm": 0.7187907695770264,
+      "learning_rate": 1.4577926658727155e-05,
+      "loss": 1.4732,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6292758089368259,
+      "grad_norm": 0.6810031533241272,
+      "learning_rate": 1.4536487026078188e-05,
+      "loss": 1.4505,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6298921417565485,
+      "grad_norm": 0.7185724973678589,
+      "learning_rate": 1.449507272493304e-05,
+      "loss": 1.41,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6305084745762712,
+      "grad_norm": 0.6889476180076599,
+      "learning_rate": 1.4453683947308538e-05,
+      "loss": 1.3529,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6311248073959939,
+      "grad_norm": 0.7004572749137878,
+      "learning_rate": 1.44123208851032e-05,
+      "loss": 1.4009,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6317411402157165,
+      "grad_norm": 0.7015326619148254,
+      "learning_rate": 1.4370983730096307e-05,
+      "loss": 1.4323,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6323574730354391,
+      "grad_norm": 0.689512312412262,
+      "learning_rate": 1.4329672673946995e-05,
+      "loss": 1.3846,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6329738058551618,
+      "grad_norm": 0.7007417678833008,
+      "learning_rate": 1.4288387908193434e-05,
+      "loss": 1.4277,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6335901386748845,
+      "grad_norm": 0.701409101486206,
+      "learning_rate": 1.4247129624251883e-05,
+      "loss": 1.4366,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6342064714946071,
+      "grad_norm": 0.6901214122772217,
+      "learning_rate": 1.4205898013415804e-05,
+      "loss": 1.4107,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6348228043143297,
+      "grad_norm": 0.6846891641616821,
+      "learning_rate": 1.416469326685501e-05,
+      "loss": 1.4469,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6354391371340524,
+      "grad_norm": 0.6826861500740051,
+      "learning_rate": 1.412351557561475e-05,
+      "loss": 1.3618,
+      "step": 1031
+    },
+    {
+      "epoch": 0.636055469953775,
+      "grad_norm": 0.712218701839447,
+      "learning_rate": 1.4082365130614824e-05,
+      "loss": 1.4264,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6366718027734977,
+      "grad_norm": 0.6926703453063965,
+      "learning_rate": 1.4041242122648714e-05,
+      "loss": 1.3978,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6372881355932203,
+      "grad_norm": 0.6992529034614563,
+      "learning_rate": 1.4000146742382692e-05,
+      "loss": 1.4868,
+      "step": 1034
+    },
+    {
+      "epoch": 0.637904468412943,
+      "grad_norm": 0.6841732263565063,
+      "learning_rate": 1.3959079180354933e-05,
+      "loss": 1.3619,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6385208012326656,
+      "grad_norm": 0.6914638876914978,
+      "learning_rate": 1.3918039626974626e-05,
+      "loss": 1.4077,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6391371340523883,
+      "grad_norm": 0.6879279613494873,
+      "learning_rate": 1.3877028272521102e-05,
+      "loss": 1.3889,
+      "step": 1037
+    },
+    {
+      "epoch": 0.639753466872111,
+      "grad_norm": 0.7122974395751953,
+      "learning_rate": 1.3836045307142967e-05,
+      "loss": 1.4497,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6403697996918336,
+      "grad_norm": 0.6926254630088806,
+      "learning_rate": 1.3795090920857161e-05,
+      "loss": 1.3819,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6409861325115562,
+      "grad_norm": 0.6959458589553833,
+      "learning_rate": 1.3754165303548148e-05,
+      "loss": 1.4077,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6416024653312788,
+      "grad_norm": 0.6710784435272217,
+      "learning_rate": 1.3713268644967012e-05,
+      "loss": 1.3921,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6422187981510016,
+      "grad_norm": 0.6844033002853394,
+      "learning_rate": 1.367240113473053e-05,
+      "loss": 1.4421,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6428351309707242,
+      "grad_norm": 0.6891530752182007,
+      "learning_rate": 1.3631562962320382e-05,
+      "loss": 1.3631,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6434514637904468,
+      "grad_norm": 0.6909316182136536,
+      "learning_rate": 1.3590754317082196e-05,
+      "loss": 1.3655,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6440677966101694,
+      "grad_norm": 0.7180591225624084,
+      "learning_rate": 1.3549975388224686e-05,
+      "loss": 1.4067,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6446841294298922,
+      "grad_norm": 0.7052772641181946,
+      "learning_rate": 1.3509226364818817e-05,
+      "loss": 1.3907,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6453004622496148,
+      "grad_norm": 0.7179615497589111,
+      "learning_rate": 1.3468507435796885e-05,
+      "loss": 1.3709,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6459167950693374,
+      "grad_norm": 0.6939482092857361,
+      "learning_rate": 1.3427818789951647e-05,
+      "loss": 1.3833,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6465331278890601,
+      "grad_norm": 0.6862821578979492,
+      "learning_rate": 1.338716061593546e-05,
+      "loss": 1.3881,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6471494607087828,
+      "grad_norm": 0.6936004161834717,
+      "learning_rate": 1.3346533102259391e-05,
+      "loss": 1.3849,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6477657935285054,
+      "grad_norm": 0.698454737663269,
+      "learning_rate": 1.3305936437292366e-05,
+      "loss": 1.3328,
+      "step": 1051
+    },
+    {
+      "epoch": 0.648382126348228,
+      "grad_norm": 0.6913847327232361,
+      "learning_rate": 1.3265370809260265e-05,
+      "loss": 1.4123,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6489984591679507,
+      "grad_norm": 0.7037560343742371,
+      "learning_rate": 1.322483640624507e-05,
+      "loss": 1.41,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6496147919876734,
+      "grad_norm": 0.7199169397354126,
+      "learning_rate": 1.3184333416183996e-05,
+      "loss": 1.3898,
+      "step": 1054
+    },
+    {
+      "epoch": 0.650231124807396,
+      "grad_norm": 0.7095668315887451,
+      "learning_rate": 1.3143862026868604e-05,
+      "loss": 1.431,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6508474576271186,
+      "grad_norm": 0.691674530506134,
+      "learning_rate": 1.3103422425943942e-05,
+      "loss": 1.4326,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6514637904468413,
+      "grad_norm": 0.6917114853858948,
+      "learning_rate": 1.3063014800907677e-05,
+      "loss": 1.4941,
+      "step": 1057
+    },
+    {
+      "epoch": 0.652080123266564,
+      "grad_norm": 0.7097762227058411,
+      "learning_rate": 1.3022639339109201e-05,
+      "loss": 1.4593,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6526964560862866,
+      "grad_norm": 0.6939966082572937,
+      "learning_rate": 1.2982296227748804e-05,
+      "loss": 1.4865,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6533127889060092,
+      "grad_norm": 0.6821755170822144,
+      "learning_rate": 1.2941985653876778e-05,
+      "loss": 1.4389,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6539291217257319,
+      "grad_norm": 0.7058581709861755,
+      "learning_rate": 1.2901707804392556e-05,
+      "loss": 1.4054,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6545454545454545,
+      "grad_norm": 0.6868889927864075,
+      "learning_rate": 1.286146286604382e-05,
+      "loss": 1.4012,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6551617873651772,
+      "grad_norm": 0.6890780925750732,
+      "learning_rate": 1.2821251025425698e-05,
+      "loss": 1.4019,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6557781201848999,
+      "grad_norm": 0.7047430872917175,
+      "learning_rate": 1.2781072468979848e-05,
+      "loss": 1.4217,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6563944530046225,
+      "grad_norm": 0.6973077654838562,
+      "learning_rate": 1.2740927382993586e-05,
+      "loss": 1.4174,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6570107858243451,
+      "grad_norm": 0.696974515914917,
+      "learning_rate": 1.2700815953599067e-05,
+      "loss": 1.4529,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6576271186440678,
+      "grad_norm": 0.6864145398139954,
+      "learning_rate": 1.2660738366772391e-05,
+      "loss": 1.4023,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6582434514637905,
+      "grad_norm": 0.6852766871452332,
+      "learning_rate": 1.2620694808332743e-05,
+      "loss": 1.3387,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6588597842835131,
+      "grad_norm": 0.7034934163093567,
+      "learning_rate": 1.2580685463941538e-05,
+      "loss": 1.3917,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6594761171032357,
+      "grad_norm": 0.7008872628211975,
+      "learning_rate": 1.2540710519101564e-05,
+      "loss": 1.4584,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6600924499229583,
+      "grad_norm": 0.6999143362045288,
+      "learning_rate": 1.2500770159156104e-05,
+      "loss": 1.4401,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6607087827426811,
+      "grad_norm": 0.6936800479888916,
+      "learning_rate": 1.2460864569288095e-05,
+      "loss": 1.3586,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6613251155624037,
+      "grad_norm": 0.6965266466140747,
+      "learning_rate": 1.2420993934519268e-05,
+      "loss": 1.3917,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6619414483821263,
+      "grad_norm": 0.7005475163459778,
+      "learning_rate": 1.2381158439709285e-05,
+      "loss": 1.4211,
+      "step": 1074
+    },
+    {
+      "epoch": 0.662557781201849,
+      "grad_norm": 0.6943978071212769,
+      "learning_rate": 1.2341358269554866e-05,
+      "loss": 1.3793,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6631741140215717,
+      "grad_norm": 0.6871793866157532,
+      "learning_rate": 1.2301593608588958e-05,
+      "loss": 1.3604,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6637904468412943,
+      "grad_norm": 0.7159802913665771,
+      "learning_rate": 1.2261864641179894e-05,
+      "loss": 1.4797,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6644067796610169,
+      "grad_norm": 0.7132952809333801,
+      "learning_rate": 1.2222171551530466e-05,
+      "loss": 1.423,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6650231124807396,
+      "grad_norm": 0.7126781344413757,
+      "learning_rate": 1.2182514523677163e-05,
+      "loss": 1.3755,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6656394453004623,
+      "grad_norm": 0.7203714847564697,
+      "learning_rate": 1.2142893741489257e-05,
+      "loss": 1.4024,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6662557781201849,
+      "grad_norm": 0.7033785581588745,
+      "learning_rate": 1.210330938866795e-05,
+      "loss": 1.408,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6668721109399075,
+      "grad_norm": 0.6952004432678223,
+      "learning_rate": 1.2063761648745573e-05,
+      "loss": 1.4091,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6674884437596302,
+      "grad_norm": 0.6984594464302063,
+      "learning_rate": 1.2024250705084683e-05,
+      "loss": 1.4586,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6681047765793529,
+      "grad_norm": 0.6936200857162476,
+      "learning_rate": 1.1984776740877239e-05,
+      "loss": 1.3687,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6687211093990755,
+      "grad_norm": 0.6909807324409485,
+      "learning_rate": 1.194533993914373e-05,
+      "loss": 1.3636,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6693374422187982,
+      "grad_norm": 0.6899669766426086,
+      "learning_rate": 1.1905940482732357e-05,
+      "loss": 1.4395,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6699537750385208,
+      "grad_norm": 0.6890370845794678,
+      "learning_rate": 1.1866578554318168e-05,
+      "loss": 1.396,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6705701078582434,
+      "grad_norm": 0.7054473161697388,
+      "learning_rate": 1.1827254336402205e-05,
+      "loss": 1.4086,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6711864406779661,
+      "grad_norm": 0.679929256439209,
+      "learning_rate": 1.1787968011310675e-05,
+      "loss": 1.4122,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6718027734976888,
+      "grad_norm": 0.7349550127983093,
+      "learning_rate": 1.1748719761194093e-05,
+      "loss": 1.5266,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6724191063174114,
+      "grad_norm": 0.7072716951370239,
+      "learning_rate": 1.1709509768026432e-05,
+      "loss": 1.3337,
+      "step": 1091
+    },
+    {
+      "epoch": 0.673035439137134,
+      "grad_norm": 0.6961799263954163,
+      "learning_rate": 1.1670338213604299e-05,
+      "loss": 1.4086,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6736517719568567,
+      "grad_norm": 0.705967903137207,
+      "learning_rate": 1.1631205279546079e-05,
+      "loss": 1.4231,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6742681047765794,
+      "grad_norm": 0.6936699748039246,
+      "learning_rate": 1.1592111147291071e-05,
+      "loss": 1.3823,
+      "step": 1094
+    },
+    {
+      "epoch": 0.674884437596302,
+      "grad_norm": 0.6913039684295654,
+      "learning_rate": 1.1553055998098715e-05,
+      "loss": 1.381,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6755007704160246,
+      "grad_norm": 0.6969724893569946,
+      "learning_rate": 1.1514040013047656e-05,
+      "loss": 1.3794,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6761171032357473,
+      "grad_norm": 0.7696253657341003,
+      "learning_rate": 1.1475063373035002e-05,
+      "loss": 1.44,
+      "step": 1097
+    },
+    {
+      "epoch": 0.67673343605547,
+      "grad_norm": 0.7233097553253174,
+      "learning_rate": 1.1436126258775406e-05,
+      "loss": 1.3911,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6773497688751926,
+      "grad_norm": 0.7103934288024902,
+      "learning_rate": 1.1397228850800262e-05,
+      "loss": 1.4205,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 0.6893105506896973,
+      "learning_rate": 1.1358371329456887e-05,
+      "loss": 1.37,
+      "step": 1100
+    },
+    {
+      "epoch": 0.678582434514638,
+      "grad_norm": 0.695289671421051,
+      "learning_rate": 1.131955387490765e-05,
+      "loss": 1.4481,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6791987673343606,
+      "grad_norm": 0.7296865582466125,
+      "learning_rate": 1.1280776667129141e-05,
+      "loss": 1.3886,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6798151001540832,
+      "grad_norm": 0.6744548678398132,
+      "learning_rate": 1.1242039885911372e-05,
+      "loss": 1.3429,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6804314329738058,
+      "grad_norm": 0.7026019096374512,
+      "learning_rate": 1.12033437108569e-05,
+      "loss": 1.418,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6810477657935285,
+      "grad_norm": 0.6900532245635986,
+      "learning_rate": 1.1164688321380006e-05,
+      "loss": 1.4008,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6816640986132512,
+      "grad_norm": 0.7204012870788574,
+      "learning_rate": 1.1126073896705893e-05,
+      "loss": 1.4799,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6822804314329738,
+      "grad_norm": 0.6975680589675903,
+      "learning_rate": 1.1087500615869806e-05,
+      "loss": 1.4047,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6828967642526964,
+      "grad_norm": 0.7122611999511719,
+      "learning_rate": 1.1048968657716253e-05,
+      "loss": 1.395,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6835130970724191,
+      "grad_norm": 0.7107416391372681,
+      "learning_rate": 1.1010478200898116e-05,
+      "loss": 1.4052,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6841294298921418,
+      "grad_norm": 0.7212474346160889,
+      "learning_rate": 1.0972029423875894e-05,
+      "loss": 1.4165,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6847457627118644,
+      "grad_norm": 0.685422956943512,
+      "learning_rate": 1.0933622504916812e-05,
+      "loss": 1.3463,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6853620955315871,
+      "grad_norm": 0.7066401839256287,
+      "learning_rate": 1.0895257622094024e-05,
+      "loss": 1.4016,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6859784283513097,
+      "grad_norm": 0.6854957938194275,
+      "learning_rate": 1.0856934953285801e-05,
+      "loss": 1.4201,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6865947611710324,
+      "grad_norm": 0.7080780863761902,
+      "learning_rate": 1.0818654676174672e-05,
+      "loss": 1.4075,
+      "step": 1114
+    },
+    {
+      "epoch": 0.687211093990755,
+      "grad_norm": 0.7332090139389038,
+      "learning_rate": 1.078041696824661e-05,
+      "loss": 1.444,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6878274268104777,
+      "grad_norm": 0.7206618189811707,
+      "learning_rate": 1.074222200679025e-05,
+      "loss": 1.3857,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6884437596302003,
+      "grad_norm": 0.723366379737854,
+      "learning_rate": 1.0704069968896004e-05,
+      "loss": 1.414,
+      "step": 1117
+    },
+    {
+      "epoch": 0.689060092449923,
+      "grad_norm": 0.7411608099937439,
+      "learning_rate": 1.0665961031455269e-05,
+      "loss": 1.4459,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6896764252696456,
+      "grad_norm": 0.7127360701560974,
+      "learning_rate": 1.0627895371159634e-05,
+      "loss": 1.3778,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6902927580893683,
+      "grad_norm": 0.695058286190033,
+      "learning_rate": 1.0589873164500007e-05,
+      "loss": 1.3622,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6909090909090909,
+      "grad_norm": 0.711351752281189,
+      "learning_rate": 1.0551894587765827e-05,
+      "loss": 1.3547,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6915254237288135,
+      "grad_norm": 0.7054039835929871,
+      "learning_rate": 1.0513959817044256e-05,
+      "loss": 1.4178,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6921417565485362,
+      "grad_norm": 0.7050431370735168,
+      "learning_rate": 1.0476069028219365e-05,
+      "loss": 1.3451,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6927580893682589,
+      "grad_norm": 0.7250397801399231,
+      "learning_rate": 1.0438222396971245e-05,
+      "loss": 1.3704,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6933744221879815,
+      "grad_norm": 0.7442151308059692,
+      "learning_rate": 1.0400420098775306e-05,
+      "loss": 1.3714,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6939907550077041,
+      "grad_norm": 0.7112275958061218,
+      "learning_rate": 1.0362662308901406e-05,
+      "loss": 1.4388,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6946070878274269,
+      "grad_norm": 0.6967577338218689,
+      "learning_rate": 1.0324949202412994e-05,
+      "loss": 1.3365,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6952234206471495,
+      "grad_norm": 0.7117252945899963,
+      "learning_rate": 1.0287280954166399e-05,
+      "loss": 1.4273,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6958397534668721,
+      "grad_norm": 0.6936394572257996,
+      "learning_rate": 1.0249657738809943e-05,
+      "loss": 1.3806,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6964560862865947,
+      "grad_norm": 0.7187258005142212,
+      "learning_rate": 1.0212079730783148e-05,
+      "loss": 1.3803,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6970724191063175,
+      "grad_norm": 0.7474151849746704,
+      "learning_rate": 1.0174547104315933e-05,
+      "loss": 1.4681,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6976887519260401,
+      "grad_norm": 0.7287842035293579,
+      "learning_rate": 1.0137060033427825e-05,
+      "loss": 1.3954,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6983050847457627,
+      "grad_norm": 0.7195358276367188,
+      "learning_rate": 1.0099618691927113e-05,
+      "loss": 1.3253,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6989214175654853,
+      "grad_norm": 0.6951313614845276,
+      "learning_rate": 1.006222325341006e-05,
+      "loss": 1.3478,
+      "step": 1134
+    },
+    {
+      "epoch": 0.699537750385208,
+      "grad_norm": 0.7087445259094238,
+      "learning_rate": 1.0024873891260128e-05,
+      "loss": 1.3921,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7001540832049307,
+      "grad_norm": 0.7034221291542053,
+      "learning_rate": 9.987570778647116e-06,
+      "loss": 1.3998,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7007704160246533,
+      "grad_norm": 0.7146585583686829,
+      "learning_rate": 9.950314088526397e-06,
+      "loss": 1.4499,
+      "step": 1137
+    },
+    {
+      "epoch": 0.701386748844376,
+      "grad_norm": 0.7069058418273926,
+      "learning_rate": 9.913103993638111e-06,
+      "loss": 1.4594,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7020030816640986,
+      "grad_norm": 0.7051481008529663,
+      "learning_rate": 9.875940666506381e-06,
+      "loss": 1.4391,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7026194144838213,
+      "grad_norm": 0.7115175127983093,
+      "learning_rate": 9.838824279438433e-06,
+      "loss": 1.446,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7032357473035439,
+      "grad_norm": 0.6992084980010986,
+      "learning_rate": 9.80175500452391e-06,
+      "loss": 1.3959,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7038520801232666,
+      "grad_norm": 0.7067825198173523,
+      "learning_rate": 9.764733013634015e-06,
+      "loss": 1.3272,
+      "step": 1142
+    },
+    {
+      "epoch": 0.7044684129429892,
+      "grad_norm": 0.7206598520278931,
+      "learning_rate": 9.727758478420682e-06,
+      "loss": 1.3917,
+      "step": 1143
+    },
+    {
+      "epoch": 0.7050847457627119,
+      "grad_norm": 0.7401341795921326,
+      "learning_rate": 9.69083157031585e-06,
+      "loss": 1.3576,
+      "step": 1144
+    },
+    {
+      "epoch": 0.7057010785824345,
+      "grad_norm": 0.7194449305534363,
+      "learning_rate": 9.65395246053065e-06,
+      "loss": 1.4549,
+      "step": 1145
+    },
+    {
+      "epoch": 0.7063174114021572,
+      "grad_norm": 0.7008496522903442,
+      "learning_rate": 9.617121320054537e-06,
+      "loss": 1.3303,
+      "step": 1146
+    },
+    {
+      "epoch": 0.7069337442218798,
+      "grad_norm": 0.7107238173484802,
+      "learning_rate": 9.580338319654614e-06,
+      "loss": 1.4444,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7075500770416024,
+      "grad_norm": 0.7031033635139465,
+      "learning_rate": 9.543603629874771e-06,
+      "loss": 1.4088,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7081664098613251,
+      "grad_norm": 0.6872963309288025,
+      "learning_rate": 9.50691742103489e-06,
+      "loss": 1.3579,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7087827426810478,
+      "grad_norm": 0.709552526473999,
+      "learning_rate": 9.470279863230075e-06,
+      "loss": 1.4079,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7093990755007704,
+      "grad_norm": 0.7044284343719482,
+      "learning_rate": 9.433691126329878e-06,
+      "loss": 1.3765,
+      "step": 1151
+    },
+    {
+      "epoch": 0.710015408320493,
+      "grad_norm": 0.7167053818702698,
+      "learning_rate": 9.397151379977474e-06,
+      "loss": 1.4557,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7106317411402158,
+      "grad_norm": 0.7098035216331482,
+      "learning_rate": 9.360660793588898e-06,
+      "loss": 1.4247,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7112480739599384,
+      "grad_norm": 0.7159917950630188,
+      "learning_rate": 9.32421953635227e-06,
+      "loss": 1.3613,
+      "step": 1154
+    },
+    {
+      "epoch": 0.711864406779661,
+      "grad_norm": 0.7129877209663391,
+      "learning_rate": 9.287827777226978e-06,
+      "loss": 1.4031,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7124807395993836,
+      "grad_norm": 0.7346654534339905,
+      "learning_rate": 9.251485684942915e-06,
+      "loss": 1.3784,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7130970724191064,
+      "grad_norm": 0.7175418138504028,
+      "learning_rate": 9.215193427999711e-06,
+      "loss": 1.4222,
+      "step": 1157
+    },
+    {
+      "epoch": 0.713713405238829,
+      "grad_norm": 0.7117372751235962,
+      "learning_rate": 9.178951174665918e-06,
+      "loss": 1.4215,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7143297380585516,
+      "grad_norm": 0.7046418190002441,
+      "learning_rate": 9.142759092978242e-06,
+      "loss": 1.4494,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7149460708782742,
+      "grad_norm": 0.7309653759002686,
+      "learning_rate": 9.106617350740785e-06,
+      "loss": 1.4566,
+      "step": 1160
+    },
+    {
+      "epoch": 0.715562403697997,
+      "grad_norm": 0.7295824289321899,
+      "learning_rate": 9.070526115524259e-06,
+      "loss": 1.3975,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7161787365177196,
+      "grad_norm": 0.7102481126785278,
+      "learning_rate": 9.034485554665174e-06,
+      "loss": 1.4925,
+      "step": 1162
+    },
+    {
+      "epoch": 0.7167950693374422,
+      "grad_norm": 0.715046763420105,
+      "learning_rate": 8.998495835265093e-06,
+      "loss": 1.3405,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7174114021571649,
+      "grad_norm": 0.7069038152694702,
+      "learning_rate": 8.962557124189879e-06,
+      "loss": 1.3604,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7180277349768875,
+      "grad_norm": 0.6958725452423096,
+      "learning_rate": 8.926669588068872e-06,
+      "loss": 1.3526,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7186440677966102,
+      "grad_norm": 0.7394919991493225,
+      "learning_rate": 8.890833393294138e-06,
+      "loss": 1.3391,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7192604006163328,
+      "grad_norm": 0.7283654808998108,
+      "learning_rate": 8.855048706019724e-06,
+      "loss": 1.3693,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7198767334360555,
+      "grad_norm": 0.7168674468994141,
+      "learning_rate": 8.819315692160836e-06,
+      "loss": 1.4116,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7204930662557781,
+      "grad_norm": 0.7166938185691833,
+      "learning_rate": 8.783634517393108e-06,
+      "loss": 1.428,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7211093990755008,
+      "grad_norm": 0.7017228603363037,
+      "learning_rate": 8.748005347151829e-06,
+      "loss": 1.4062,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7217257318952234,
+      "grad_norm": 0.7194585800170898,
+      "learning_rate": 8.712428346631164e-06,
+      "loss": 1.453,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7223420647149461,
+      "grad_norm": 0.710802435874939,
+      "learning_rate": 8.67690368078338e-06,
+      "loss": 1.3687,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7229583975346687,
+      "grad_norm": 0.7056416869163513,
+      "learning_rate": 8.641431514318126e-06,
+      "loss": 1.3782,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7235747303543913,
+      "grad_norm": 0.710121214389801,
+      "learning_rate": 8.606012011701607e-06,
+      "loss": 1.3897,
+      "step": 1174
+    },
+    {
+      "epoch": 0.724191063174114,
+      "grad_norm": 0.696091890335083,
+      "learning_rate": 8.570645337155878e-06,
+      "loss": 1.4023,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7248073959938367,
+      "grad_norm": 0.6936500072479248,
+      "learning_rate": 8.535331654658044e-06,
+      "loss": 1.3682,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7254237288135593,
+      "grad_norm": 0.7060485482215881,
+      "learning_rate": 8.500071127939503e-06,
+      "loss": 1.3287,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7260400616332819,
+      "grad_norm": 0.7032719254493713,
+      "learning_rate": 8.464863920485222e-06,
+      "loss": 1.3692,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7266563944530047,
+      "grad_norm": 0.7065621614456177,
+      "learning_rate": 8.429710195532938e-06,
+      "loss": 1.4197,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.6944799423217773,
+      "learning_rate": 8.394610116072413e-06,
+      "loss": 1.3146,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7278890600924499,
+      "grad_norm": 0.7001040577888489,
+      "learning_rate": 8.359563844844698e-06,
+      "loss": 1.4202,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7285053929121725,
+      "grad_norm": 0.7114636898040771,
+      "learning_rate": 8.324571544341347e-06,
+      "loss": 1.3622,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7291217257318953,
+      "grad_norm": 0.7142324447631836,
+      "learning_rate": 8.289633376803696e-06,
+      "loss": 1.3781,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7297380585516179,
+      "grad_norm": 0.6933879256248474,
+      "learning_rate": 8.254749504222081e-06,
+      "loss": 1.3588,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7303543913713405,
+      "grad_norm": 0.7201752066612244,
+      "learning_rate": 8.219920088335096e-06,
+      "loss": 1.4058,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7309707241910631,
+      "grad_norm": 0.7171357870101929,
+      "learning_rate": 8.185145290628868e-06,
+      "loss": 1.3599,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7315870570107859,
+      "grad_norm": 0.7056859135627747,
+      "learning_rate": 8.150425272336262e-06,
+      "loss": 1.3929,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7322033898305085,
+      "grad_norm": 0.7151781916618347,
+      "learning_rate": 8.115760194436186e-06,
+      "loss": 1.4267,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7328197226502311,
+      "grad_norm": 0.7176246047019958,
+      "learning_rate": 8.081150217652798e-06,
+      "loss": 1.3788,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7334360554699538,
+      "grad_norm": 0.7289698719978333,
+      "learning_rate": 8.046595502454777e-06,
+      "loss": 1.4488,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7340523882896764,
+      "grad_norm": 0.711230456829071,
+      "learning_rate": 8.012096209054607e-06,
+      "loss": 1.3588,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7346687211093991,
+      "grad_norm": 0.7277358770370483,
+      "learning_rate": 7.977652497407784e-06,
+      "loss": 1.4182,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7352850539291217,
+      "grad_norm": 0.7100339531898499,
+      "learning_rate": 7.943264527212106e-06,
+      "loss": 1.3666,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7359013867488444,
+      "grad_norm": 0.7025458216667175,
+      "learning_rate": 7.908932457906945e-06,
+      "loss": 1.3813,
+      "step": 1194
+    },
+    {
+      "epoch": 0.736517719568567,
+      "grad_norm": 0.729390025138855,
+      "learning_rate": 7.874656448672463e-06,
+      "loss": 1.3758,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7371340523882897,
+      "grad_norm": 0.7198243737220764,
+      "learning_rate": 7.840436658428912e-06,
+      "loss": 1.4209,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7377503852080123,
+      "grad_norm": 0.7019217014312744,
+      "learning_rate": 7.806273245835886e-06,
+      "loss": 1.3795,
+      "step": 1197
+    },
+    {
+      "epoch": 0.738366718027735,
+      "grad_norm": 0.7048054933547974,
+      "learning_rate": 7.7721663692916e-06,
+      "loss": 1.3679,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7389830508474576,
+      "grad_norm": 0.7118720412254333,
+      "learning_rate": 7.7381161869321e-06,
+      "loss": 1.457,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7395993836671803,
+      "grad_norm": 0.7205626964569092,
+      "learning_rate": 7.704122856630603e-06,
+      "loss": 1.4363,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7402157164869029,
+      "grad_norm": 0.7166395783424377,
+      "learning_rate": 7.670186535996735e-06,
+      "loss": 1.3607,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7408320493066256,
+      "grad_norm": 0.7080506086349487,
+      "learning_rate": 7.636307382375778e-06,
+      "loss": 1.3268,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7414483821263482,
+      "grad_norm": 0.7311720848083496,
+      "learning_rate": 7.602485552847963e-06,
+      "loss": 1.3585,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7420647149460708,
+      "grad_norm": 0.71082603931427,
+      "learning_rate": 7.568721204227756e-06,
+      "loss": 1.3972,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7426810477657936,
+      "grad_norm": 0.7301634550094604,
+      "learning_rate": 7.535014493063093e-06,
+      "loss": 1.397,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7432973805855162,
+      "grad_norm": 0.7354387640953064,
+      "learning_rate": 7.50136557563468e-06,
+      "loss": 1.4075,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7439137134052388,
+      "grad_norm": 0.7137809991836548,
+      "learning_rate": 7.4677746079552785e-06,
+      "loss": 1.4476,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7445300462249614,
+      "grad_norm": 0.7131066918373108,
+      "learning_rate": 7.434241745768947e-06,
+      "loss": 1.3502,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7451463790446842,
+      "grad_norm": 0.7096465229988098,
+      "learning_rate": 7.400767144550338e-06,
+      "loss": 1.3659,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7457627118644068,
+      "grad_norm": 0.7059075832366943,
+      "learning_rate": 7.3673509595040046e-06,
+      "loss": 1.3727,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7463790446841294,
+      "grad_norm": 0.7131897211074829,
+      "learning_rate": 7.333993345563626e-06,
+      "loss": 1.3898,
+      "step": 1211
+    },
+    {
+      "epoch": 0.746995377503852,
+      "grad_norm": 0.7129470109939575,
+      "learning_rate": 7.300694457391317e-06,
+      "loss": 1.4215,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7476117103235748,
+      "grad_norm": 0.7073593735694885,
+      "learning_rate": 7.267454449376934e-06,
+      "loss": 1.4434,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7482280431432974,
+      "grad_norm": 0.7063706517219543,
+      "learning_rate": 7.234273475637332e-06,
+      "loss": 1.3978,
+      "step": 1214
+    },
+    {
+      "epoch": 0.74884437596302,
+      "grad_norm": 0.7119913101196289,
+      "learning_rate": 7.20115169001562e-06,
+      "loss": 1.3667,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7494607087827427,
+      "grad_norm": 0.7262555360794067,
+      "learning_rate": 7.168089246080521e-06,
+      "loss": 1.3771,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7500770416024654,
+      "grad_norm": 0.7164336442947388,
+      "learning_rate": 7.135086297125624e-06,
+      "loss": 1.3527,
+      "step": 1217
+    },
+    {
+      "epoch": 0.750693374422188,
+      "grad_norm": 0.7244095802307129,
+      "learning_rate": 7.102142996168624e-06,
+      "loss": 1.4139,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7513097072419106,
+      "grad_norm": 0.7116941809654236,
+      "learning_rate": 7.069259495950704e-06,
+      "loss": 1.4362,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7519260400616333,
+      "grad_norm": 0.7252678275108337,
+      "learning_rate": 7.036435948935782e-06,
+      "loss": 1.3703,
+      "step": 1220
+    },
+    {
+      "epoch": 0.752542372881356,
+      "grad_norm": 0.7081559896469116,
+      "learning_rate": 7.003672507309758e-06,
+      "loss": 1.3299,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7531587057010786,
+      "grad_norm": 0.7235808968544006,
+      "learning_rate": 6.970969322979901e-06,
+      "loss": 1.3618,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7537750385208012,
+      "grad_norm": 0.7001596689224243,
+      "learning_rate": 6.9383265475740876e-06,
+      "loss": 1.323,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7543913713405239,
+      "grad_norm": 0.7268368601799011,
+      "learning_rate": 6.905744332440096e-06,
+      "loss": 1.4239,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7550077041602465,
+      "grad_norm": 0.713495671749115,
+      "learning_rate": 6.873222828644919e-06,
+      "loss": 1.3722,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7556240369799692,
+      "grad_norm": 0.7131713032722473,
+      "learning_rate": 6.840762186974082e-06,
+      "loss": 1.4169,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7562403697996918,
+      "grad_norm": 0.7363885045051575,
+      "learning_rate": 6.808362557930899e-06,
+      "loss": 1.4106,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7568567026194145,
+      "grad_norm": 0.7581199407577515,
+      "learning_rate": 6.7760240917358085e-06,
+      "loss": 1.4574,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7574730354391371,
+      "grad_norm": 0.7095124125480652,
+      "learning_rate": 6.743746938325682e-06,
+      "loss": 1.3803,
+      "step": 1229
+    },
+    {
+      "epoch": 0.7580893682588598,
+      "grad_norm": 0.7028020024299622,
+      "learning_rate": 6.711531247353098e-06,
+      "loss": 1.421,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7587057010785825,
+      "grad_norm": 0.7266805768013,
+      "learning_rate": 6.679377168185661e-06,
+      "loss": 1.4083,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7593220338983051,
+      "grad_norm": 0.7083789110183716,
+      "learning_rate": 6.647284849905335e-06,
+      "loss": 1.3774,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7599383667180277,
+      "grad_norm": 0.7222945094108582,
+      "learning_rate": 6.615254441307728e-06,
+      "loss": 1.3642,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7605546995377503,
+      "grad_norm": 0.7144194841384888,
+      "learning_rate": 6.583286090901369e-06,
+      "loss": 1.3527,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7611710323574731,
+      "grad_norm": 0.6895611882209778,
+      "learning_rate": 6.551379946907093e-06,
+      "loss": 1.3261,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7617873651771957,
+      "grad_norm": 0.7057611346244812,
+      "learning_rate": 6.519536157257313e-06,
+      "loss": 1.3808,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7624036979969183,
+      "grad_norm": 0.7023035883903503,
+      "learning_rate": 6.48775486959532e-06,
+      "loss": 1.3392,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7630200308166409,
+      "grad_norm": 0.7267735600471497,
+      "learning_rate": 6.456036231274619e-06,
+      "loss": 1.3959,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7636363636363637,
+      "grad_norm": 0.7065931558609009,
+      "learning_rate": 6.424380389358256e-06,
+      "loss": 1.3883,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7642526964560863,
+      "grad_norm": 0.7238292098045349,
+      "learning_rate": 6.392787490618109e-06,
+      "loss": 1.367,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7648690292758089,
+      "grad_norm": 0.7149064540863037,
+      "learning_rate": 6.361257681534214e-06,
+      "loss": 1.3419,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7654853620955316,
+      "grad_norm": 0.7724552154541016,
+      "learning_rate": 6.329791108294121e-06,
+      "loss": 1.4499,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7661016949152543,
+      "grad_norm": 0.7136510610580444,
+      "learning_rate": 6.298387916792162e-06,
+      "loss": 1.4062,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7667180277349769,
+      "grad_norm": 0.7175265550613403,
+      "learning_rate": 6.2670482526288e-06,
+      "loss": 1.3944,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7673343605546995,
+      "grad_norm": 0.7174100279808044,
+      "learning_rate": 6.23577226110998e-06,
+      "loss": 1.4153,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7679506933744222,
+      "grad_norm": 0.6994783878326416,
+      "learning_rate": 6.204560087246401e-06,
+      "loss": 1.3657,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7685670261941449,
+      "grad_norm": 0.7514929175376892,
+      "learning_rate": 6.173411875752882e-06,
+      "loss": 1.4144,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7691833590138675,
+      "grad_norm": 0.7208922505378723,
+      "learning_rate": 6.142327771047691e-06,
+      "loss": 1.3717,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7697996918335901,
+      "grad_norm": 0.7243021130561829,
+      "learning_rate": 6.111307917251852e-06,
+      "loss": 1.329,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7704160246533128,
+      "grad_norm": 0.7086784243583679,
+      "learning_rate": 6.080352458188503e-06,
+      "loss": 1.3561,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7710323574730354,
+      "grad_norm": 0.7406588196754456,
+      "learning_rate": 6.049461537382207e-06,
+      "loss": 1.3816,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7716486902927581,
+      "grad_norm": 0.7102490663528442,
+      "learning_rate": 6.018635298058289e-06,
+      "loss": 1.4063,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7722650231124807,
+      "grad_norm": 0.7220368385314941,
+      "learning_rate": 5.987873883142208e-06,
+      "loss": 1.43,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7728813559322034,
+      "grad_norm": 0.7251654863357544,
+      "learning_rate": 5.957177435258832e-06,
+      "loss": 1.3865,
+      "step": 1254
+    },
+    {
+      "epoch": 0.773497688751926,
+      "grad_norm": 0.7314308285713196,
+      "learning_rate": 5.926546096731836e-06,
+      "loss": 1.4022,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7741140215716487,
+      "grad_norm": 0.7118254899978638,
+      "learning_rate": 5.895980009583002e-06,
+      "loss": 1.3715,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7747303543913714,
+      "grad_norm": 0.7330620288848877,
+      "learning_rate": 5.86547931553157e-06,
+      "loss": 1.4047,
+      "step": 1257
+    },
+    {
+      "epoch": 0.775346687211094,
+      "grad_norm": 0.707216203212738,
+      "learning_rate": 5.83504415599361e-06,
+      "loss": 1.3342,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7759630200308166,
+      "grad_norm": 0.7439261674880981,
+      "learning_rate": 5.804674672081316e-06,
+      "loss": 1.4486,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7765793528505393,
+      "grad_norm": 0.7214643359184265,
+      "learning_rate": 5.774371004602386e-06,
+      "loss": 1.3898,
+      "step": 1260
+    },
+    {
+      "epoch": 0.777195685670262,
+      "grad_norm": 0.7158306241035461,
+      "learning_rate": 5.744133294059375e-06,
+      "loss": 1.3912,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7778120184899846,
+      "grad_norm": 0.7160073518753052,
+      "learning_rate": 5.713961680649003e-06,
+      "loss": 1.3828,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7784283513097072,
+      "grad_norm": 0.7245885729789734,
+      "learning_rate": 5.683856304261561e-06,
+      "loss": 1.4208,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7790446841294298,
+      "grad_norm": 0.731977105140686,
+      "learning_rate": 5.653817304480211e-06,
+      "loss": 1.3954,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7796610169491526,
+      "grad_norm": 0.7258724570274353,
+      "learning_rate": 5.623844820580362e-06,
+      "loss": 1.3735,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7802773497688752,
+      "grad_norm": 0.731006920337677,
+      "learning_rate": 5.593938991529042e-06,
+      "loss": 1.3741,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7808936825885978,
+      "grad_norm": 0.7350090146064758,
+      "learning_rate": 5.5640999559842165e-06,
+      "loss": 1.3616,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7815100154083205,
+      "grad_norm": 0.6930884718894958,
+      "learning_rate": 5.534327852294161e-06,
+      "loss": 1.3928,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7821263482280432,
+      "grad_norm": 0.7102153301239014,
+      "learning_rate": 5.5046228184968434e-06,
+      "loss": 1.3629,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7827426810477658,
+      "grad_norm": 0.7317716479301453,
+      "learning_rate": 5.474984992319246e-06,
+      "loss": 1.3752,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7833590138674884,
+      "grad_norm": 0.7138091325759888,
+      "learning_rate": 5.445414511176739e-06,
+      "loss": 1.3422,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7839753466872111,
+      "grad_norm": 0.7453138828277588,
+      "learning_rate": 5.415911512172469e-06,
+      "loss": 1.4376,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7845916795069338,
+      "grad_norm": 0.7103058099746704,
+      "learning_rate": 5.386476132096683e-06,
+      "loss": 1.4404,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7852080123266564,
+      "grad_norm": 0.7179279327392578,
+      "learning_rate": 5.35710850742613e-06,
+      "loss": 1.3607,
+      "step": 1274
+    },
+    {
+      "epoch": 0.785824345146379,
+      "grad_norm": 0.7095497846603394,
+      "learning_rate": 5.327808774323391e-06,
+      "loss": 1.4056,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7864406779661017,
+      "grad_norm": 0.7179814577102661,
+      "learning_rate": 5.298577068636294e-06,
+      "loss": 1.3701,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7870570107858244,
+      "grad_norm": 0.696064829826355,
+      "learning_rate": 5.269413525897242e-06,
+      "loss": 1.3474,
+      "step": 1277
+    },
+    {
+      "epoch": 0.787673343605547,
+      "grad_norm": 0.818098783493042,
+      "learning_rate": 5.240318281322596e-06,
+      "loss": 1.3964,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7882896764252696,
+      "grad_norm": 0.7271025776863098,
+      "learning_rate": 5.211291469812078e-06,
+      "loss": 1.3838,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7889060092449923,
+      "grad_norm": 0.7381759881973267,
+      "learning_rate": 5.182333225948099e-06,
+      "loss": 1.4363,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7895223420647149,
+      "grad_norm": 0.7190137505531311,
+      "learning_rate": 5.153443683995156e-06,
+      "loss": 1.3693,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7901386748844376,
+      "grad_norm": 0.7235825657844543,
+      "learning_rate": 5.124622977899234e-06,
+      "loss": 1.4,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7907550077041603,
+      "grad_norm": 0.747995138168335,
+      "learning_rate": 5.095871241287136e-06,
+      "loss": 1.4039,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7913713405238829,
+      "grad_norm": 0.7190261483192444,
+      "learning_rate": 5.067188607465896e-06,
+      "loss": 1.3254,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7919876733436055,
+      "grad_norm": 0.7184574007987976,
+      "learning_rate": 5.038575209422165e-06,
+      "loss": 1.3675,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7926040061633282,
+      "grad_norm": 0.7203155159950256,
+      "learning_rate": 5.010031179821575e-06,
+      "loss": 1.3917,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7932203389830509,
+      "grad_norm": 0.7268801331520081,
+      "learning_rate": 4.981556651008122e-06,
+      "loss": 1.4372,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7938366718027735,
+      "grad_norm": 0.7292035818099976,
+      "learning_rate": 4.953151755003578e-06,
+      "loss": 1.4315,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7944530046224961,
+      "grad_norm": 0.715397298336029,
+      "learning_rate": 4.924816623506872e-06,
+      "loss": 1.378,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7950693374422187,
+      "grad_norm": 0.7268989086151123,
+      "learning_rate": 4.8965513878934315e-06,
+      "loss": 1.4453,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7956856702619415,
+      "grad_norm": 0.7248528599739075,
+      "learning_rate": 4.86835617921465e-06,
+      "loss": 1.392,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7963020030816641,
+      "grad_norm": 0.7248679995536804,
+      "learning_rate": 4.840231128197246e-06,
+      "loss": 1.4362,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7969183359013867,
+      "grad_norm": 0.7252503037452698,
+      "learning_rate": 4.812176365242609e-06,
+      "loss": 1.3433,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7975346687211095,
+      "grad_norm": 0.7331882119178772,
+      "learning_rate": 4.78419202042628e-06,
+      "loss": 1.3657,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7981510015408321,
+      "grad_norm": 0.7273738980293274,
+      "learning_rate": 4.756278223497299e-06,
+      "loss": 1.3729,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7987673343605547,
+      "grad_norm": 0.7178515791893005,
+      "learning_rate": 4.728435103877604e-06,
+      "loss": 1.3206,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7993836671802773,
+      "grad_norm": 0.7424306273460388,
+      "learning_rate": 4.700662790661434e-06,
+      "loss": 1.4229,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.742459237575531,
+      "learning_rate": 4.672961412614754e-06,
+      "loss": 1.4205,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8006163328197227,
+      "grad_norm": 0.7272717952728271,
+      "learning_rate": 4.645331098174621e-06,
+      "loss": 1.4361,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8012326656394453,
+      "grad_norm": 0.7326829433441162,
+      "learning_rate": 4.61777197544861e-06,
+      "loss": 1.4317,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8018489984591679,
+      "grad_norm": 0.7059509754180908,
+      "learning_rate": 4.590284172214228e-06,
+      "loss": 1.4159,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8024653312788906,
+      "grad_norm": 0.7047575116157532,
+      "learning_rate": 4.5628678159182995e-06,
+      "loss": 1.4321,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8030816640986133,
+      "grad_norm": 0.7133081555366516,
+      "learning_rate": 4.535523033676381e-06,
+      "loss": 1.3675,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8036979969183359,
+      "grad_norm": 0.7335922122001648,
+      "learning_rate": 4.508249952272194e-06,
+      "loss": 1.4238,
+      "step": 1304
+    },
+    {
+      "epoch": 0.8043143297380585,
+      "grad_norm": 0.7125181555747986,
+      "learning_rate": 4.481048698157011e-06,
+      "loss": 1.3518,
+      "step": 1305
+    },
+    {
+      "epoch": 0.8049306625577812,
+      "grad_norm": 0.7158137559890747,
+      "learning_rate": 4.453919397449067e-06,
+      "loss": 1.3763,
+      "step": 1306
+    },
+    {
+      "epoch": 0.8055469953775038,
+      "grad_norm": 0.7084963321685791,
+      "learning_rate": 4.426862175933002e-06,
+      "loss": 1.3512,
+      "step": 1307
+    },
+    {
+      "epoch": 0.8061633281972265,
+      "grad_norm": 0.7073696255683899,
+      "learning_rate": 4.399877159059269e-06,
+      "loss": 1.3259,
+      "step": 1308
+    },
+    {
+      "epoch": 0.8067796610169492,
+      "grad_norm": 0.7452852129936218,
+      "learning_rate": 4.372964471943512e-06,
+      "loss": 1.4189,
+      "step": 1309
+    },
+    {
+      "epoch": 0.8073959938366718,
+      "grad_norm": 0.7733001708984375,
+      "learning_rate": 4.346124239366052e-06,
+      "loss": 1.3823,
+      "step": 1310
+    },
+    {
+      "epoch": 0.8080123266563944,
+      "grad_norm": 0.7213237285614014,
+      "learning_rate": 4.31935658577127e-06,
+      "loss": 1.3737,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8086286594761171,
+      "grad_norm": 0.7072203755378723,
+      "learning_rate": 4.2926616352670105e-06,
+      "loss": 1.3493,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8092449922958398,
+      "grad_norm": 0.7354212403297424,
+      "learning_rate": 4.2660395116240585e-06,
+      "loss": 1.392,
+      "step": 1313
+    },
+    {
+      "epoch": 0.8098613251155624,
+      "grad_norm": 0.7316009402275085,
+      "learning_rate": 4.239490338275536e-06,
+      "loss": 1.3756,
+      "step": 1314
+    },
+    {
+      "epoch": 0.810477657935285,
+      "grad_norm": 0.7126972675323486,
+      "learning_rate": 4.213014238316317e-06,
+      "loss": 1.3844,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8110939907550077,
+      "grad_norm": 0.7217592000961304,
+      "learning_rate": 4.186611334502479e-06,
+      "loss": 1.3882,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8117103235747304,
+      "grad_norm": 0.7217358350753784,
+      "learning_rate": 4.160281749250734e-06,
+      "loss": 1.3673,
+      "step": 1317
+    },
+    {
+      "epoch": 0.812326656394453,
+      "grad_norm": 0.7270700931549072,
+      "learning_rate": 4.134025604637846e-06,
+      "loss": 1.3842,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8129429892141756,
+      "grad_norm": 0.7280433773994446,
+      "learning_rate": 4.107843022400062e-06,
+      "loss": 1.3252,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8135593220338984,
+      "grad_norm": 0.7191141843795776,
+      "learning_rate": 4.0817341239325854e-06,
+      "loss": 1.3664,
+      "step": 1320
+    },
+    {
+      "epoch": 0.814175654853621,
+      "grad_norm": 0.7274618744850159,
+      "learning_rate": 4.055699030288962e-06,
+      "loss": 1.418,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8147919876733436,
+      "grad_norm": 0.7222168445587158,
+      "learning_rate": 4.029737862180545e-06,
+      "loss": 1.4273,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8154083204930662,
+      "grad_norm": 0.7185543775558472,
+      "learning_rate": 4.003850739975952e-06,
+      "loss": 1.3778,
+      "step": 1323
+    },
+    {
+      "epoch": 0.816024653312789,
+      "grad_norm": 0.7258328199386597,
+      "learning_rate": 3.978037783700459e-06,
+      "loss": 1.3482,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8166409861325116,
+      "grad_norm": 0.7273615598678589,
+      "learning_rate": 3.9522991130355024e-06,
+      "loss": 1.3393,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8172573189522342,
+      "grad_norm": 0.7251265048980713,
+      "learning_rate": 3.926634847318065e-06,
+      "loss": 1.3487,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8178736517719568,
+      "grad_norm": 0.7428770661354065,
+      "learning_rate": 3.90104510554018e-06,
+      "loss": 1.4398,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8184899845916795,
+      "grad_norm": 0.7304401397705078,
+      "learning_rate": 3.8755300063483315e-06,
+      "loss": 1.4503,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8191063174114022,
+      "grad_norm": 0.7172685265541077,
+      "learning_rate": 3.850089668042927e-06,
+      "loss": 1.3497,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8197226502311248,
+      "grad_norm": 0.7083370089530945,
+      "learning_rate": 3.824724208577762e-06,
+      "loss": 1.3713,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8203389830508474,
+      "grad_norm": 0.7312684059143066,
+      "learning_rate": 3.7994337455594334e-06,
+      "loss": 1.3897,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8209553158705701,
+      "grad_norm": 0.723172664642334,
+      "learning_rate": 3.7742183962468314e-06,
+      "loss": 1.3794,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8215716486902928,
+      "grad_norm": 0.7249532341957092,
+      "learning_rate": 3.7490782775505906e-06,
+      "loss": 1.3421,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8221879815100154,
+      "grad_norm": 0.7260609865188599,
+      "learning_rate": 3.7240135060325243e-06,
+      "loss": 1.3069,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8228043143297381,
+      "grad_norm": 0.7325504422187805,
+      "learning_rate": 3.699024197905099e-06,
+      "loss": 1.4338,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8234206471494607,
+      "grad_norm": 0.7493659853935242,
+      "learning_rate": 3.6741104690309136e-06,
+      "loss": 1.3945,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8240369799691833,
+      "grad_norm": 0.7268967032432556,
+      "learning_rate": 3.6492724349221263e-06,
+      "loss": 1.3236,
+      "step": 1337
+    },
+    {
+      "epoch": 0.824653312788906,
+      "grad_norm": 0.7218689322471619,
+      "learning_rate": 3.6245102107399533e-06,
+      "loss": 1.3569,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8252696456086287,
+      "grad_norm": 0.727701723575592,
+      "learning_rate": 3.599823911294107e-06,
+      "loss": 1.4089,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8258859784283513,
+      "grad_norm": 0.7251731753349304,
+      "learning_rate": 3.575213651042275e-06,
+      "loss": 1.4189,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8265023112480739,
+      "grad_norm": 0.7352141737937927,
+      "learning_rate": 3.550679544089606e-06,
+      "loss": 1.3759,
+      "step": 1341
+    },
+    {
+      "epoch": 0.8271186440677966,
+      "grad_norm": 0.7323986291885376,
+      "learning_rate": 3.5262217041881465e-06,
+      "loss": 1.3801,
+      "step": 1342
+    },
+    {
+      "epoch": 0.8277349768875193,
+      "grad_norm": 0.7270162105560303,
+      "learning_rate": 3.5018402447363363e-06,
+      "loss": 1.3617,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8283513097072419,
+      "grad_norm": 0.7232922911643982,
+      "learning_rate": 3.4775352787784944e-06,
+      "loss": 1.4649,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8289676425269645,
+      "grad_norm": 0.7182289958000183,
+      "learning_rate": 3.4533069190042557e-06,
+      "loss": 1.344,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8295839753466873,
+      "grad_norm": 0.7160165309906006,
+      "learning_rate": 3.4291552777480775e-06,
+      "loss": 1.3756,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8302003081664099,
+      "grad_norm": 0.7182633280754089,
+      "learning_rate": 3.4050804669887304e-06,
+      "loss": 1.408,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8308166409861325,
+      "grad_norm": 0.7277814745903015,
+      "learning_rate": 3.3810825983487348e-06,
+      "loss": 1.3208,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8314329738058551,
+      "grad_norm": 0.7269110679626465,
+      "learning_rate": 3.3571617830938963e-06,
+      "loss": 1.4085,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8320493066255779,
+      "grad_norm": 0.7359941601753235,
+      "learning_rate": 3.333318132132737e-06,
+      "loss": 1.4223,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8326656394453005,
+      "grad_norm": 0.7269524335861206,
+      "learning_rate": 3.3095517560160317e-06,
+      "loss": 1.4026,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8332819722650231,
+      "grad_norm": 0.7317273616790771,
+      "learning_rate": 3.2858627649362583e-06,
+      "loss": 1.415,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8338983050847457,
+      "grad_norm": 0.7236542105674744,
+      "learning_rate": 3.262251268727097e-06,
+      "loss": 1.3936,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8345146379044684,
+      "grad_norm": 0.7345942854881287,
+      "learning_rate": 3.238717376862939e-06,
+      "loss": 1.3643,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8351309707241911,
+      "grad_norm": 0.7178893685340881,
+      "learning_rate": 3.215261198458348e-06,
+      "loss": 1.309,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8357473035439137,
+      "grad_norm": 0.7405028343200684,
+      "learning_rate": 3.1918828422675753e-06,
+      "loss": 1.3716,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8363636363636363,
+      "grad_norm": 0.7354982495307922,
+      "learning_rate": 3.1685824166840606e-06,
+      "loss": 1.4047,
+      "step": 1357
+    },
+    {
+      "epoch": 0.836979969183359,
+      "grad_norm": 0.7350180745124817,
+      "learning_rate": 3.1453600297399055e-06,
+      "loss": 1.3633,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8375963020030817,
+      "grad_norm": 0.7463111877441406,
+      "learning_rate": 3.1222157891053895e-06,
+      "loss": 1.4208,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8382126348228043,
+      "grad_norm": 0.7327929735183716,
+      "learning_rate": 3.099149802088479e-06,
+      "loss": 1.4247,
+      "step": 1360
+    },
+    {
+      "epoch": 0.838828967642527,
+      "grad_norm": 0.7054534554481506,
+      "learning_rate": 3.076162175634303e-06,
+      "loss": 1.3707,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8394453004622496,
+      "grad_norm": 0.7115305662155151,
+      "learning_rate": 3.05325301632468e-06,
+      "loss": 1.342,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8400616332819723,
+      "grad_norm": 0.7385708093643188,
+      "learning_rate": 3.0304224303776154e-06,
+      "loss": 1.4044,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8406779661016949,
+      "grad_norm": 0.7390673160552979,
+      "learning_rate": 3.007670523646826e-06,
+      "loss": 1.3919,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8412942989214176,
+      "grad_norm": 0.7156866788864136,
+      "learning_rate": 2.9849974016211945e-06,
+      "loss": 1.3119,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8419106317411402,
+      "grad_norm": 0.6909548044204712,
+      "learning_rate": 2.9624031694243594e-06,
+      "loss": 1.3135,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8425269645608628,
+      "grad_norm": 0.7377358675003052,
+      "learning_rate": 2.9398879318141714e-06,
+      "loss": 1.4042,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8431432973805855,
+      "grad_norm": 0.7310072183609009,
+      "learning_rate": 2.917451793182222e-06,
+      "loss": 1.4501,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8437596302003082,
+      "grad_norm": 0.7252581119537354,
+      "learning_rate": 2.8950948575533598e-06,
+      "loss": 1.3258,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8443759630200308,
+      "grad_norm": 0.7234041690826416,
+      "learning_rate": 2.8728172285852185e-06,
+      "loss": 1.3633,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8449922958397534,
+      "grad_norm": 0.7043598890304565,
+      "learning_rate": 2.8506190095677235e-06,
+      "loss": 1.4535,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8456086286594762,
+      "grad_norm": 0.7169867753982544,
+      "learning_rate": 2.8285003034226098e-06,
+      "loss": 1.3718,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8462249614791988,
+      "grad_norm": 0.724925696849823,
+      "learning_rate": 2.8064612127029667e-06,
+      "loss": 1.4027,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8468412942989214,
+      "grad_norm": 0.7323075532913208,
+      "learning_rate": 2.784501839592735e-06,
+      "loss": 1.3886,
+      "step": 1374
+    },
+    {
+      "epoch": 0.847457627118644,
+      "grad_norm": 0.7221584320068359,
+      "learning_rate": 2.7626222859062468e-06,
+      "loss": 1.3797,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8480739599383668,
+      "grad_norm": 0.7247506976127625,
+      "learning_rate": 2.7408226530877645e-06,
+      "loss": 1.3669,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8486902927580894,
+      "grad_norm": 0.7102174758911133,
+      "learning_rate": 2.7191030422109842e-06,
+      "loss": 1.3623,
+      "step": 1377
+    },
+    {
+      "epoch": 0.849306625577812,
+      "grad_norm": 0.7192674875259399,
+      "learning_rate": 2.6974635539785855e-06,
+      "loss": 1.3987,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8499229583975346,
+      "grad_norm": 0.7591804265975952,
+      "learning_rate": 2.6759042887217644e-06,
+      "loss": 1.3459,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8505392912172574,
+      "grad_norm": 0.788594126701355,
+      "learning_rate": 2.6544253463997714e-06,
+      "loss": 1.3891,
+      "step": 1380
+    },
+    {
+      "epoch": 0.85115562403698,
+      "grad_norm": 0.7144513130187988,
+      "learning_rate": 2.633026826599412e-06,
+      "loss": 1.3943,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8517719568567026,
+      "grad_norm": 0.7241362929344177,
+      "learning_rate": 2.611708828534647e-06,
+      "loss": 1.3679,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8523882896764252,
+      "grad_norm": 0.7267704010009766,
+      "learning_rate": 2.5904714510460882e-06,
+      "loss": 1.3619,
+      "step": 1383
+    },
+    {
+      "epoch": 0.853004622496148,
+      "grad_norm": 0.7231332063674927,
+      "learning_rate": 2.5693147926005324e-06,
+      "loss": 1.4001,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8536209553158706,
+      "grad_norm": 0.7272068858146667,
+      "learning_rate": 2.5482389512905513e-06,
+      "loss": 1.4356,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8542372881355932,
+      "grad_norm": 0.7333012223243713,
+      "learning_rate": 2.5272440248340015e-06,
+      "loss": 1.4343,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8548536209553159,
+      "grad_norm": 0.7473307847976685,
+      "learning_rate": 2.5063301105735605e-06,
+      "loss": 1.4214,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8554699537750385,
+      "grad_norm": 0.7091383934020996,
+      "learning_rate": 2.4854973054763163e-06,
+      "loss": 1.4001,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8560862865947612,
+      "grad_norm": 0.7252072691917419,
+      "learning_rate": 2.4647457061332915e-06,
+      "loss": 1.3426,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8567026194144838,
+      "grad_norm": 0.724292516708374,
+      "learning_rate": 2.44407540875899e-06,
+      "loss": 1.3994,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8573189522342065,
+      "grad_norm": 0.7236889600753784,
+      "learning_rate": 2.423486509190962e-06,
+      "loss": 1.3364,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8579352850539291,
+      "grad_norm": 0.7115809321403503,
+      "learning_rate": 2.402979102889369e-06,
+      "loss": 1.3956,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8585516178736518,
+      "grad_norm": 0.7352230548858643,
+      "learning_rate": 2.3825532849365175e-06,
+      "loss": 1.4102,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8591679506933744,
+      "grad_norm": 0.7262681722640991,
+      "learning_rate": 2.362209150036432e-06,
+      "loss": 1.3586,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8597842835130971,
+      "grad_norm": 0.7057443857192993,
+      "learning_rate": 2.3419467925144246e-06,
+      "loss": 1.3251,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8604006163328197,
+      "grad_norm": 0.7192476391792297,
+      "learning_rate": 2.3217663063166395e-06,
+      "loss": 1.3178,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8610169491525423,
+      "grad_norm": 0.7523620128631592,
+      "learning_rate": 2.3016677850096226e-06,
+      "loss": 1.419,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8616332819722651,
+      "grad_norm": 0.7219938039779663,
+      "learning_rate": 2.281651321779903e-06,
+      "loss": 1.3614,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8622496147919877,
+      "grad_norm": 0.7282921671867371,
+      "learning_rate": 2.2617170094335374e-06,
+      "loss": 1.3872,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8628659476117103,
+      "grad_norm": 0.7162911891937256,
+      "learning_rate": 2.2418649403956906e-06,
+      "loss": 1.3392,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8634822804314329,
+      "grad_norm": 0.7239651679992676,
+      "learning_rate": 2.222095206710215e-06,
+      "loss": 1.3218,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8640986132511557,
+      "grad_norm": 0.7120274901390076,
+      "learning_rate": 2.202407900039216e-06,
+      "loss": 1.3747,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8647149460708783,
+      "grad_norm": 0.7461383938789368,
+      "learning_rate": 2.1828031116626212e-06,
+      "loss": 1.4185,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8653312788906009,
+      "grad_norm": 0.7477744817733765,
+      "learning_rate": 2.1632809324777626e-06,
+      "loss": 1.3865,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8659476117103235,
+      "grad_norm": 0.7402583360671997,
+      "learning_rate": 2.143841452998965e-06,
+      "loss": 1.3869,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8665639445300463,
+      "grad_norm": 0.7182440757751465,
+      "learning_rate": 2.1244847633571087e-06,
+      "loss": 1.3914,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8671802773497689,
+      "grad_norm": 0.7290626764297485,
+      "learning_rate": 2.1052109532992195e-06,
+      "loss": 1.4571,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8677966101694915,
+      "grad_norm": 0.7254422307014465,
+      "learning_rate": 2.086020112188065e-06,
+      "loss": 1.3781,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8684129429892141,
+      "grad_norm": 0.7252994179725647,
+      "learning_rate": 2.0669123290017112e-06,
+      "loss": 1.4259,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8690292758089369,
+      "grad_norm": 0.7205610871315002,
+      "learning_rate": 2.047887692333137e-06,
+      "loss": 1.3647,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8696456086286595,
+      "grad_norm": 0.7031440138816833,
+      "learning_rate": 2.028946290389817e-06,
+      "loss": 1.3082,
+      "step": 1411
+    },
+    {
+      "epoch": 0.8702619414483821,
+      "grad_norm": 0.7062574625015259,
+      "learning_rate": 2.010088210993304e-06,
+      "loss": 1.4072,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8708782742681048,
+      "grad_norm": 0.738295316696167,
+      "learning_rate": 1.991313541578819e-06,
+      "loss": 1.3904,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8714946070878274,
+      "grad_norm": 0.7189211845397949,
+      "learning_rate": 1.972622369194872e-06,
+      "loss": 1.3463,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8721109399075501,
+      "grad_norm": 0.7320471405982971,
+      "learning_rate": 1.954014780502822e-06,
+      "loss": 1.3921,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8727272727272727,
+      "grad_norm": 0.7337237000465393,
+      "learning_rate": 1.935490861776512e-06,
+      "loss": 1.3637,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8733436055469954,
+      "grad_norm": 0.7296890616416931,
+      "learning_rate": 1.917050698901832e-06,
+      "loss": 1.3858,
+      "step": 1417
+    },
+    {
+      "epoch": 0.873959938366718,
+      "grad_norm": 0.7268068194389343,
+      "learning_rate": 1.8986943773763445e-06,
+      "loss": 1.3827,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8745762711864407,
+      "grad_norm": 0.7310439944267273,
+      "learning_rate": 1.8804219823088955e-06,
+      "loss": 1.4079,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8751926040061633,
+      "grad_norm": 0.7287306189537048,
+      "learning_rate": 1.8622335984191875e-06,
+      "loss": 1.3907,
+      "step": 1420
+    },
+    {
+      "epoch": 0.875808936825886,
+      "grad_norm": 0.7096817493438721,
+      "learning_rate": 1.8441293100374214e-06,
+      "loss": 1.3793,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8764252696456086,
+      "grad_norm": 0.7398281693458557,
+      "learning_rate": 1.826109201103885e-06,
+      "loss": 1.4189,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8770416024653312,
+      "grad_norm": 0.7173163890838623,
+      "learning_rate": 1.808173355168561e-06,
+      "loss": 1.3524,
+      "step": 1423
+    },
+    {
+      "epoch": 0.877657935285054,
+      "grad_norm": 0.7151859402656555,
+      "learning_rate": 1.7903218553907664e-06,
+      "loss": 1.3286,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8782742681047766,
+      "grad_norm": 0.7404639720916748,
+      "learning_rate": 1.7725547845387336e-06,
+      "loss": 1.384,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8788906009244992,
+      "grad_norm": 0.7217142581939697,
+      "learning_rate": 1.7548722249892413e-06,
+      "loss": 1.417,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8795069337442218,
+      "grad_norm": 0.7019501328468323,
+      "learning_rate": 1.737274258727244e-06,
+      "loss": 1.3569,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8801232665639446,
+      "grad_norm": 0.7110302448272705,
+      "learning_rate": 1.7197609673454651e-06,
+      "loss": 1.3772,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8807395993836672,
+      "grad_norm": 0.7107219696044922,
+      "learning_rate": 1.7023324320440537e-06,
+      "loss": 1.3858,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8813559322033898,
+      "grad_norm": 0.7200949788093567,
+      "learning_rate": 1.6849887336301729e-06,
+      "loss": 1.3171,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8819722650231124,
+      "grad_norm": 0.7224409580230713,
+      "learning_rate": 1.6677299525176383e-06,
+      "loss": 1.3848,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8825885978428352,
+      "grad_norm": 0.7271050214767456,
+      "learning_rate": 1.6505561687265626e-06,
+      "loss": 1.3439,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8832049306625578,
+      "grad_norm": 0.7470753788948059,
+      "learning_rate": 1.6334674618829604e-06,
+      "loss": 1.4063,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8838212634822804,
+      "grad_norm": 0.7314844727516174,
+      "learning_rate": 1.6164639112183843e-06,
+      "loss": 1.3953,
+      "step": 1434
+    },
+    {
+      "epoch": 0.884437596302003,
+      "grad_norm": 0.7257526516914368,
+      "learning_rate": 1.5995455955695738e-06,
+      "loss": 1.3384,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8850539291217258,
+      "grad_norm": 0.7165493369102478,
+      "learning_rate": 1.5827125933780663e-06,
+      "loss": 1.3162,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8856702619414484,
+      "grad_norm": 0.7319618463516235,
+      "learning_rate": 1.5659649826898426e-06,
+      "loss": 1.3572,
+      "step": 1437
+    },
+    {
+      "epoch": 0.886286594761171,
+      "grad_norm": 0.7209432721138,
+      "learning_rate": 1.5493028411549826e-06,
+      "loss": 1.4002,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8869029275808937,
+      "grad_norm": 0.7280626893043518,
+      "learning_rate": 1.5327262460272741e-06,
+      "loss": 1.4002,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8875192604006163,
+      "grad_norm": 0.7359567880630493,
+      "learning_rate": 1.516235274163873e-06,
+      "loss": 1.3863,
+      "step": 1440
+    },
+    {
+      "epoch": 0.888135593220339,
+      "grad_norm": 0.733280599117279,
+      "learning_rate": 1.499830002024949e-06,
+      "loss": 1.3872,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8887519260400616,
+      "grad_norm": 0.7490953803062439,
+      "learning_rate": 1.4835105056733289e-06,
+      "loss": 1.3993,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8893682588597843,
+      "grad_norm": 0.7335138916969299,
+      "learning_rate": 1.467276860774134e-06,
+      "loss": 1.3683,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8899845916795069,
+      "grad_norm": 0.7324069738388062,
+      "learning_rate": 1.4511291425944295e-06,
+      "loss": 1.382,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8906009244992296,
+      "grad_norm": 0.72662752866745,
+      "learning_rate": 1.4350674260029051e-06,
+      "loss": 1.3592,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8912172573189522,
+      "grad_norm": 0.7359530925750732,
+      "learning_rate": 1.4190917854694863e-06,
+      "loss": 1.3572,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8918335901386749,
+      "grad_norm": 0.7337392568588257,
+      "learning_rate": 1.4032022950650093e-06,
+      "loss": 1.3581,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8924499229583975,
+      "grad_norm": 0.7240508198738098,
+      "learning_rate": 1.3873990284608873e-06,
+      "loss": 1.3726,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8930662557781202,
+      "grad_norm": 0.733126699924469,
+      "learning_rate": 1.3716820589287516e-06,
+      "loss": 1.391,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8936825885978429,
+      "grad_norm": 0.7423774600028992,
+      "learning_rate": 1.3560514593401109e-06,
+      "loss": 1.4264,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8942989214175655,
+      "grad_norm": 0.7311149835586548,
+      "learning_rate": 1.3405073021660343e-06,
+      "loss": 1.4013,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8949152542372881,
+      "grad_norm": 0.7221912741661072,
+      "learning_rate": 1.3250496594768004e-06,
+      "loss": 1.4485,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8955315870570107,
+      "grad_norm": 0.7111760377883911,
+      "learning_rate": 1.3096786029415464e-06,
+      "loss": 1.3742,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8961479198767335,
+      "grad_norm": 0.7323029637336731,
+      "learning_rate": 1.2943942038279778e-06,
+      "loss": 1.3676,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8967642526964561,
+      "grad_norm": 0.7225942015647888,
+      "learning_rate": 1.2791965330020072e-06,
+      "loss": 1.3482,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8973805855161787,
+      "grad_norm": 0.722963809967041,
+      "learning_rate": 1.2640856609274166e-06,
+      "loss": 1.373,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8979969183359013,
+      "grad_norm": 0.7386912107467651,
+      "learning_rate": 1.2490616576655646e-06,
+      "loss": 1.4276,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8986132511556241,
+      "grad_norm": 0.7295930981636047,
+      "learning_rate": 1.2341245928750477e-06,
+      "loss": 1.3937,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8992295839753467,
+      "grad_norm": 0.7337785959243774,
+      "learning_rate": 1.2192745358113523e-06,
+      "loss": 1.4112,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8998459167950693,
+      "grad_norm": 0.7204175591468811,
+      "learning_rate": 1.2045115553265684e-06,
+      "loss": 1.3429,
+      "step": 1460
+    },
+    {
+      "epoch": 0.900462249614792,
+      "grad_norm": 0.7215947508811951,
+      "learning_rate": 1.1898357198690614e-06,
+      "loss": 1.3339,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9010785824345147,
+      "grad_norm": 0.7369790077209473,
+      "learning_rate": 1.1752470974831387e-06,
+      "loss": 1.4019,
+      "step": 1462
+    },
+    {
+      "epoch": 0.9016949152542373,
+      "grad_norm": 0.7801219820976257,
+      "learning_rate": 1.1607457558087475e-06,
+      "loss": 1.4512,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9023112480739599,
+      "grad_norm": 0.7241122722625732,
+      "learning_rate": 1.1463317620811699e-06,
+      "loss": 1.3793,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9029275808936826,
+      "grad_norm": 0.7354309558868408,
+      "learning_rate": 1.1320051831306866e-06,
+      "loss": 1.365,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9035439137134053,
+      "grad_norm": 0.7350800037384033,
+      "learning_rate": 1.1177660853822836e-06,
+      "loss": 1.4047,
+      "step": 1466
+    },
+    {
+      "epoch": 0.9041602465331279,
+      "grad_norm": 0.7345896363258362,
+      "learning_rate": 1.1036145348553528e-06,
+      "loss": 1.352,
+      "step": 1467
+    },
+    {
+      "epoch": 0.9047765793528505,
+      "grad_norm": 0.7281753420829773,
+      "learning_rate": 1.0895505971633623e-06,
+      "loss": 1.3789,
+      "step": 1468
+    },
+    {
+      "epoch": 0.9053929121725732,
+      "grad_norm": 0.742215633392334,
+      "learning_rate": 1.075574337513563e-06,
+      "loss": 1.3924,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9060092449922958,
+      "grad_norm": 0.7282941937446594,
+      "learning_rate": 1.0616858207067037e-06,
+      "loss": 1.322,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9066255778120185,
+      "grad_norm": 0.7248411178588867,
+      "learning_rate": 1.0478851111366972e-06,
+      "loss": 1.3778,
+      "step": 1471
+    },
+    {
+      "epoch": 0.9072419106317411,
+      "grad_norm": 0.7386523485183716,
+      "learning_rate": 1.0341722727903459e-06,
+      "loss": 1.4193,
+      "step": 1472
+    },
+    {
+      "epoch": 0.9078582434514638,
+      "grad_norm": 0.7255223989486694,
+      "learning_rate": 1.0205473692470403e-06,
+      "loss": 1.4317,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9084745762711864,
+      "grad_norm": 0.7541740536689758,
+      "learning_rate": 1.0070104636784683e-06,
+      "loss": 1.4586,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.7295917272567749,
+      "learning_rate": 9.935616188482955e-07,
+      "loss": 1.3724,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9097072419106318,
+      "grad_norm": 0.7236862778663635,
+      "learning_rate": 9.802008971119137e-07,
+      "loss": 1.389,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9103235747303544,
+      "grad_norm": 0.7203603386878967,
+      "learning_rate": 9.66928360416135e-07,
+      "loss": 1.3698,
+      "step": 1477
+    },
+    {
+      "epoch": 0.910939907550077,
+      "grad_norm": 0.7180744409561157,
+      "learning_rate": 9.537440702988877e-07,
+      "loss": 1.3836,
+      "step": 1478
+    },
+    {
+      "epoch": 0.9115562403697997,
+      "grad_norm": 0.7191655039787292,
+      "learning_rate": 9.406480878889557e-07,
+      "loss": 1.3987,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9121725731895224,
+      "grad_norm": 0.7223641872406006,
+      "learning_rate": 9.276404739056888e-07,
+      "loss": 1.3684,
+      "step": 1480
+    },
+    {
+      "epoch": 0.912788906009245,
+      "grad_norm": 0.7153388857841492,
+      "learning_rate": 9.147212886587064e-07,
+      "loss": 1.3463,
+      "step": 1481
+    },
+    {
+      "epoch": 0.9134052388289676,
+      "grad_norm": 0.7190892696380615,
+      "learning_rate": 9.018905920476361e-07,
+      "loss": 1.3721,
+      "step": 1482
+    },
+    {
+      "epoch": 0.9140215716486902,
+      "grad_norm": 0.7367353439331055,
+      "learning_rate": 8.891484435618336e-07,
+      "loss": 1.3341,
+      "step": 1483
+    },
+    {
+      "epoch": 0.914637904468413,
+      "grad_norm": 0.7566488981246948,
+      "learning_rate": 8.764949022800961e-07,
+      "loss": 1.4075,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9152542372881356,
+      "grad_norm": 0.7184884548187256,
+      "learning_rate": 8.639300268703921e-07,
+      "loss": 1.3457,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9158705701078582,
+      "grad_norm": 0.7262673377990723,
+      "learning_rate": 8.514538755895985e-07,
+      "loss": 1.3941,
+      "step": 1486
+    },
+    {
+      "epoch": 0.916486902927581,
+      "grad_norm": 0.7452720999717712,
+      "learning_rate": 8.390665062832238e-07,
+      "loss": 1.3973,
+      "step": 1487
+    },
+    {
+      "epoch": 0.9171032357473036,
+      "grad_norm": 0.7270230650901794,
+      "learning_rate": 8.267679763851344e-07,
+      "loss": 1.3902,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9177195685670262,
+      "grad_norm": 0.7322908639907837,
+      "learning_rate": 8.145583429172998e-07,
+      "loss": 1.3399,
+      "step": 1489
+    },
+    {
+      "epoch": 0.9183359013867488,
+      "grad_norm": 0.7247247695922852,
+      "learning_rate": 8.024376624895169e-07,
+      "loss": 1.3568,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9189522342064715,
+      "grad_norm": 0.7169971466064453,
+      "learning_rate": 7.90405991299159e-07,
+      "loss": 1.3358,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9195685670261942,
+      "grad_norm": 0.7295762300491333,
+      "learning_rate": 7.784633851309053e-07,
+      "loss": 1.4511,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9201848998459168,
+      "grad_norm": 0.7147249579429626,
+      "learning_rate": 7.666098993564897e-07,
+      "loss": 1.2799,
+      "step": 1493
+    },
+    {
+      "epoch": 0.9208012326656394,
+      "grad_norm": 0.7116222381591797,
+      "learning_rate": 7.548455889344364e-07,
+      "loss": 1.3776,
+      "step": 1494
+    },
+    {
+      "epoch": 0.9214175654853621,
+      "grad_norm": 0.7110794186592102,
+      "learning_rate": 7.431705084098073e-07,
+      "loss": 1.4168,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9220338983050848,
+      "grad_norm": 0.7403594255447388,
+      "learning_rate": 7.315847119139618e-07,
+      "loss": 1.3906,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9226502311248074,
+      "grad_norm": 0.7179705500602722,
+      "learning_rate": 7.200882531642839e-07,
+      "loss": 1.3312,
+      "step": 1497
+    },
+    {
+      "epoch": 0.92326656394453,
+      "grad_norm": 0.7244460582733154,
+      "learning_rate": 7.086811854639486e-07,
+      "loss": 1.4023,
+      "step": 1498
+    },
+    {
+      "epoch": 0.9238828967642527,
+      "grad_norm": 0.7258676290512085,
+      "learning_rate": 6.973635617016717e-07,
+      "loss": 1.4067,
+      "step": 1499
+    },
+    {
+      "epoch": 0.9244992295839753,
+      "grad_norm": 0.7246628403663635,
+      "learning_rate": 6.861354343514648e-07,
+      "loss": 1.4011,
+      "step": 1500
+    },
+    {
+      "epoch": 0.925115562403698,
+      "grad_norm": 0.7228919267654419,
+      "learning_rate": 6.749968554723829e-07,
+      "loss": 1.3562,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9257318952234207,
+      "grad_norm": 0.7195237874984741,
+      "learning_rate": 6.639478767082997e-07,
+      "loss": 1.3637,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9263482280431433,
+      "grad_norm": 0.7245622873306274,
+      "learning_rate": 6.52988549287652e-07,
+      "loss": 1.3663,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9269645608628659,
+      "grad_norm": 0.7418311834335327,
+      "learning_rate": 6.421189240232184e-07,
+      "loss": 1.3975,
+      "step": 1504
+    },
+    {
+      "epoch": 0.9275808936825886,
+      "grad_norm": 0.727361798286438,
+      "learning_rate": 6.313390513118656e-07,
+      "loss": 1.4014,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9281972265023113,
+      "grad_norm": 0.7426782846450806,
+      "learning_rate": 6.206489811343241e-07,
+      "loss": 1.3709,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9288135593220339,
+      "grad_norm": 0.743525505065918,
+      "learning_rate": 6.100487630549645e-07,
+      "loss": 1.407,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9294298921417565,
+      "grad_norm": 0.7232701778411865,
+      "learning_rate": 5.995384462215525e-07,
+      "loss": 1.391,
+      "step": 1508
+    },
+    {
+      "epoch": 0.9300462249614792,
+      "grad_norm": 0.7274662852287292,
+      "learning_rate": 5.891180793650275e-07,
+      "loss": 1.3877,
+      "step": 1509
+    },
+    {
+      "epoch": 0.9306625577812019,
+      "grad_norm": 0.7287861704826355,
+      "learning_rate": 5.787877107992823e-07,
+      "loss": 1.4199,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9312788906009245,
+      "grad_norm": 0.7321040034294128,
+      "learning_rate": 5.685473884209303e-07,
+      "loss": 1.354,
+      "step": 1511
+    },
+    {
+      "epoch": 0.9318952234206471,
+      "grad_norm": 0.737404465675354,
+      "learning_rate": 5.583971597090831e-07,
+      "loss": 1.4415,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9325115562403699,
+      "grad_norm": 0.7138884663581848,
+      "learning_rate": 5.483370717251446e-07,
+      "loss": 1.3111,
+      "step": 1513
+    },
+    {
+      "epoch": 0.9331278890600925,
+      "grad_norm": 0.7342374324798584,
+      "learning_rate": 5.383671711125704e-07,
+      "loss": 1.3704,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9337442218798151,
+      "grad_norm": 0.7225253582000732,
+      "learning_rate": 5.284875040966752e-07,
+      "loss": 1.3201,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9343605546995377,
+      "grad_norm": 0.7454358339309692,
+      "learning_rate": 5.186981164843951e-07,
+      "loss": 1.3992,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9349768875192604,
+      "grad_norm": 0.7335894703865051,
+      "learning_rate": 5.089990536640988e-07,
+      "loss": 1.3962,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9355932203389831,
+      "grad_norm": 0.7320942282676697,
+      "learning_rate": 4.993903606053562e-07,
+      "loss": 1.3594,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9362095531587057,
+      "grad_norm": 0.712775707244873,
+      "learning_rate": 4.898720818587421e-07,
+      "loss": 1.3608,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9368258859784283,
+      "grad_norm": 0.7340325117111206,
+      "learning_rate": 4.804442615556304e-07,
+      "loss": 1.4351,
+      "step": 1520
+    },
+    {
+      "epoch": 0.937442218798151,
+      "grad_norm": 0.7273707389831543,
+      "learning_rate": 4.7110694340797735e-07,
+      "loss": 1.3953,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9380585516178737,
+      "grad_norm": 0.7152740955352783,
+      "learning_rate": 4.618601707081349e-07,
+      "loss": 1.3717,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9386748844375963,
+      "grad_norm": 0.7229000926017761,
+      "learning_rate": 4.5270398632863754e-07,
+      "loss": 1.338,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9392912172573189,
+      "grad_norm": 0.7413216829299927,
+      "learning_rate": 4.4363843272201337e-07,
+      "loss": 1.4145,
+      "step": 1524
+    },
+    {
+      "epoch": 0.9399075500770416,
+      "grad_norm": 0.7118443846702576,
+      "learning_rate": 4.346635519205711e-07,
+      "loss": 1.316,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9405238828967643,
+      "grad_norm": 0.7365329265594482,
+      "learning_rate": 4.2577938553623357e-07,
+      "loss": 1.4335,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9411402157164869,
+      "grad_norm": 0.7154534459114075,
+      "learning_rate": 4.169859747603111e-07,
+      "loss": 1.3357,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9417565485362096,
+      "grad_norm": 0.6909504532814026,
+      "learning_rate": 4.0828336036333517e-07,
+      "loss": 1.2937,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9423728813559322,
+      "grad_norm": 0.7170702219009399,
+      "learning_rate": 3.996715826948605e-07,
+      "loss": 1.4113,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9429892141756548,
+      "grad_norm": 0.7334873676300049,
+      "learning_rate": 3.9115068168327886e-07,
+      "loss": 1.4169,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9436055469953775,
+      "grad_norm": 0.7421221137046814,
+      "learning_rate": 3.8272069683562787e-07,
+      "loss": 1.3436,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9442218798151002,
+      "grad_norm": 0.7343822121620178,
+      "learning_rate": 3.743816672374179e-07,
+      "loss": 1.3902,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9448382126348228,
+      "grad_norm": 0.7070733308792114,
+      "learning_rate": 3.661336315524544e-07,
+      "loss": 1.2661,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9454545454545454,
+      "grad_norm": 0.7308067083358765,
+      "learning_rate": 3.5797662802263376e-07,
+      "loss": 1.3827,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9460708782742681,
+      "grad_norm": 0.734627902507782,
+      "learning_rate": 3.4991069446779434e-07,
+      "loss": 1.3972,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9466872110939908,
+      "grad_norm": 0.7203368544578552,
+      "learning_rate": 3.4193586828553226e-07,
+      "loss": 1.3535,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9473035439137134,
+      "grad_norm": 0.7200612425804138,
+      "learning_rate": 3.3405218645101714e-07,
+      "loss": 1.3178,
+      "step": 1537
+    },
+    {
+      "epoch": 0.947919876733436,
+      "grad_norm": 0.734172523021698,
+      "learning_rate": 3.262596855168343e-07,
+      "loss": 1.3855,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9485362095531588,
+      "grad_norm": 0.7278779745101929,
+      "learning_rate": 3.1855840161280736e-07,
+      "loss": 1.3976,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9491525423728814,
+      "grad_norm": 0.7094461917877197,
+      "learning_rate": 3.1094837044583603e-07,
+      "loss": 1.3658,
+      "step": 1540
+    },
+    {
+      "epoch": 0.949768875192604,
+      "grad_norm": 0.722000002861023,
+      "learning_rate": 3.034296272997228e-07,
+      "loss": 1.3635,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9503852080123266,
+      "grad_norm": 0.7232699990272522,
+      "learning_rate": 2.9600220703502435e-07,
+      "loss": 1.288,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9510015408320494,
+      "grad_norm": 0.7094044089317322,
+      "learning_rate": 2.886661440888716e-07,
+      "loss": 1.2866,
+      "step": 1543
+    },
+    {
+      "epoch": 0.951617873651772,
+      "grad_norm": 0.7381419539451599,
+      "learning_rate": 2.8142147247481875e-07,
+      "loss": 1.4083,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9522342064714946,
+      "grad_norm": 0.7336975932121277,
+      "learning_rate": 2.742682257826879e-07,
+      "loss": 1.4225,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9528505392912172,
+      "grad_norm": 0.7240797877311707,
+      "learning_rate": 2.6720643717841554e-07,
+      "loss": 1.4069,
+      "step": 1546
+    },
+    {
+      "epoch": 0.95346687211094,
+      "grad_norm": 0.7330515384674072,
+      "learning_rate": 2.6023613940388217e-07,
+      "loss": 1.3468,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9540832049306626,
+      "grad_norm": 0.7307139039039612,
+      "learning_rate": 2.533573647767806e-07,
+      "loss": 1.3646,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9546995377503852,
+      "grad_norm": 0.7152190208435059,
+      "learning_rate": 2.465701451904612e-07,
+      "loss": 1.3491,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9553158705701078,
+      "grad_norm": 0.7295586466789246,
+      "learning_rate": 2.398745121137669e-07,
+      "loss": 1.362,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9559322033898305,
+      "grad_norm": 0.7249663472175598,
+      "learning_rate": 2.3327049659091604e-07,
+      "loss": 1.3736,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9565485362095532,
+      "grad_norm": 0.7256446480751038,
+      "learning_rate": 2.2675812924133566e-07,
+      "loss": 1.3341,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9571648690292758,
+      "grad_norm": 0.7275270223617554,
+      "learning_rate": 2.203374402595282e-07,
+      "loss": 1.2793,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9577812018489985,
+      "grad_norm": 0.7114280462265015,
+      "learning_rate": 2.140084594149272e-07,
+      "loss": 1.3562,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9583975346687211,
+      "grad_norm": 0.7545391917228699,
+      "learning_rate": 2.0777121605177532e-07,
+      "loss": 1.3658,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9590138674884437,
+      "grad_norm": 0.7280948162078857,
+      "learning_rate": 2.0162573908895978e-07,
+      "loss": 1.4011,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9596302003081664,
+      "grad_norm": 0.7233501076698303,
+      "learning_rate": 1.9557205701990156e-07,
+      "loss": 1.354,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9602465331278891,
+      "grad_norm": 0.7393321990966797,
+      "learning_rate": 1.8961019791241763e-07,
+      "loss": 1.3834,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9608628659476117,
+      "grad_norm": 0.7296978831291199,
+      "learning_rate": 1.8374018940858773e-07,
+      "loss": 1.3679,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9614791987673343,
+      "grad_norm": 0.7477917075157166,
+      "learning_rate": 1.7796205872462558e-07,
+      "loss": 1.3805,
+      "step": 1560
+    },
+    {
+      "epoch": 0.962095531587057,
+      "grad_norm": 0.7210617065429688,
+      "learning_rate": 1.7227583265075453e-07,
+      "loss": 1.4369,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9627118644067797,
+      "grad_norm": 0.7462349534034729,
+      "learning_rate": 1.666815375510855e-07,
+      "loss": 1.3717,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9633281972265023,
+      "grad_norm": 0.7392804026603699,
+      "learning_rate": 1.6117919936349037e-07,
+      "loss": 1.3412,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9639445300462249,
+      "grad_norm": 0.7448540329933167,
+      "learning_rate": 1.5576884359948864e-07,
+      "loss": 1.3827,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9645608628659477,
+      "grad_norm": 0.7214841842651367,
+      "learning_rate": 1.5045049534412103e-07,
+      "loss": 1.3621,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9651771956856703,
+      "grad_norm": 0.7396149635314941,
+      "learning_rate": 1.4522417925583398e-07,
+      "loss": 1.3917,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9657935285053929,
+      "grad_norm": 0.7189009785652161,
+      "learning_rate": 1.4008991956637298e-07,
+      "loss": 1.4358,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9664098613251155,
+      "grad_norm": 0.7343153953552246,
+      "learning_rate": 1.350477400806649e-07,
+      "loss": 1.38,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9670261941448383,
+      "grad_norm": 0.7388443350791931,
+      "learning_rate": 1.300976641767049e-07,
+      "loss": 1.3686,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9676425269645609,
+      "grad_norm": 0.7136210203170776,
+      "learning_rate": 1.2523971480545183e-07,
+      "loss": 1.3045,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9682588597842835,
+      "grad_norm": 0.7327091693878174,
+      "learning_rate": 1.2047391449072632e-07,
+      "loss": 1.4277,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9688751926040061,
+      "grad_norm": 0.7424311637878418,
+      "learning_rate": 1.1580028532909072e-07,
+      "loss": 1.4042,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9694915254237289,
+      "grad_norm": 0.7276663780212402,
+      "learning_rate": 1.1121884898976697e-07,
+      "loss": 1.4313,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9701078582434515,
+      "grad_norm": 0.7247490882873535,
+      "learning_rate": 1.0672962671452347e-07,
+      "loss": 1.3336,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9707241910631741,
+      "grad_norm": 0.7185050249099731,
+      "learning_rate": 1.023326393175772e-07,
+      "loss": 1.3625,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9713405238828967,
+      "grad_norm": 0.7435563206672668,
+      "learning_rate": 9.802790718549838e-08,
+      "loss": 1.4238,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9719568567026194,
+      "grad_norm": 0.7177332043647766,
+      "learning_rate": 9.381545027712157e-08,
+      "loss": 1.3614,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9725731895223421,
+      "grad_norm": 0.7312375903129578,
+      "learning_rate": 8.969528812344585e-08,
+      "loss": 1.4499,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9731895223420647,
+      "grad_norm": 0.7222129106521606,
+      "learning_rate": 8.56674398275481e-08,
+      "loss": 1.4005,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9738058551617874,
+      "grad_norm": 0.7181993722915649,
+      "learning_rate": 8.173192406448982e-08,
+      "loss": 1.2664,
+      "step": 1580
+    },
+    {
+      "epoch": 0.97442218798151,
+      "grad_norm": 0.7184217572212219,
+      "learning_rate": 7.788875908123938e-08,
+      "loss": 1.3528,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9750385208012327,
+      "grad_norm": 0.7295069098472595,
+      "learning_rate": 7.413796269657658e-08,
+      "loss": 1.3755,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9756548536209553,
+      "grad_norm": 0.7309151291847229,
+      "learning_rate": 7.047955230101933e-08,
+      "loss": 1.3595,
+      "step": 1583
+    },
+    {
+      "epoch": 0.976271186440678,
+      "grad_norm": 0.7414847016334534,
+      "learning_rate": 6.691354485673707e-08,
+      "loss": 1.4012,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9768875192604006,
+      "grad_norm": 0.7242807149887085,
+      "learning_rate": 6.343995689747528e-08,
+      "loss": 1.3901,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9775038520801232,
+      "grad_norm": 0.7461392879486084,
+      "learning_rate": 6.005880452847557e-08,
+      "loss": 1.3785,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9781201848998459,
+      "grad_norm": 0.7206755876541138,
+      "learning_rate": 5.6770103426408986e-08,
+      "loss": 1.3596,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9787365177195686,
+      "grad_norm": 0.7150706052780151,
+      "learning_rate": 5.357386883929172e-08,
+      "loss": 1.3272,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9793528505392912,
+      "grad_norm": 0.7407107353210449,
+      "learning_rate": 5.0470115586422894e-08,
+      "loss": 1.345,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9799691833590138,
+      "grad_norm": 0.7264348268508911,
+      "learning_rate": 4.7458858058309076e-08,
+      "loss": 1.3619,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9805855161787366,
+      "grad_norm": 0.7331915497779846,
+      "learning_rate": 4.4540110216608754e-08,
+      "loss": 1.355,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9812018489984592,
+      "grad_norm": 0.7333071231842041,
+      "learning_rate": 4.171388559405465e-08,
+      "loss": 1.366,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9818181818181818,
+      "grad_norm": 0.7220433354377747,
+      "learning_rate": 3.898019729440039e-08,
+      "loss": 1.3133,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9824345146379044,
+      "grad_norm": 0.7353851199150085,
+      "learning_rate": 3.6339057992351713e-08,
+      "loss": 1.3854,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9830508474576272,
+      "grad_norm": 0.7174954414367676,
+      "learning_rate": 3.3790479933519806e-08,
+      "loss": 1.3554,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9836671802773498,
+      "grad_norm": 0.7186242341995239,
+      "learning_rate": 3.133447493434805e-08,
+      "loss": 1.3541,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9842835130970724,
+      "grad_norm": 0.732376754283905,
+      "learning_rate": 2.897105438207204e-08,
+      "loss": 1.3404,
+      "step": 1597
+    },
+    {
+      "epoch": 0.984899845916795,
+      "grad_norm": 0.7284975647926331,
+      "learning_rate": 2.6700229234659646e-08,
+      "loss": 1.3989,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9855161787365178,
+      "grad_norm": 0.7639544606208801,
+      "learning_rate": 2.4522010020762153e-08,
+      "loss": 1.4025,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9861325115562404,
+      "grad_norm": 0.7221969366073608,
+      "learning_rate": 2.243640683965875e-08,
+      "loss": 1.3988,
+      "step": 1600
+    },
+    {
+      "epoch": 0.986748844375963,
+      "grad_norm": 0.7278079986572266,
+      "learning_rate": 2.0443429361223233e-08,
+      "loss": 1.355,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9873651771956856,
+      "grad_norm": 0.7300804257392883,
+      "learning_rate": 1.8543086825870693e-08,
+      "loss": 1.3995,
+      "step": 1602
+    },
+    {
+      "epoch": 0.9879815100154083,
+      "grad_norm": 0.7292181849479675,
+      "learning_rate": 1.6735388044508695e-08,
+      "loss": 1.3497,
+      "step": 1603
+    },
+    {
+      "epoch": 0.988597842835131,
+      "grad_norm": 0.7365246415138245,
+      "learning_rate": 1.5020341398512826e-08,
+      "loss": 1.4073,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9892141756548536,
+      "grad_norm": 0.7097652554512024,
+      "learning_rate": 1.3397954839671212e-08,
+      "loss": 1.2937,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9898305084745763,
+      "grad_norm": 0.7248865962028503,
+      "learning_rate": 1.1868235890155621e-08,
+      "loss": 1.359,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9904468412942989,
+      "grad_norm": 0.7467084527015686,
+      "learning_rate": 1.0431191642490401e-08,
+      "loss": 1.4447,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9910631741140216,
+      "grad_norm": 0.7339733839035034,
+      "learning_rate": 9.08682875951028e-09,
+      "loss": 1.3643,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9916795069337442,
+      "grad_norm": 0.7175866365432739,
+      "learning_rate": 7.835153474335944e-09,
+      "loss": 1.3987,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9922958397534669,
+      "grad_norm": 0.730383038520813,
+      "learning_rate": 6.67617159034295e-09,
+      "loss": 1.3112,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9929121725731895,
+      "grad_norm": 0.744522750377655,
+      "learning_rate": 5.609888481132863e-09,
+      "loss": 1.2988,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9935285053929122,
+      "grad_norm": 0.7123790383338928,
+      "learning_rate": 4.6363090905177096e-09,
+      "loss": 1.3524,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9941448382126348,
+      "grad_norm": 0.7403290867805481,
+      "learning_rate": 3.75543793248001e-09,
+      "loss": 1.3811,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9947611710323575,
+      "grad_norm": 0.7189465165138245,
+      "learning_rate": 2.96727909116834e-09,
+      "loss": 1.3414,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9953775038520801,
+      "grad_norm": 0.7205127477645874,
+      "learning_rate": 2.2718362208729026e-09,
+      "loss": 1.3561,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9959938366718027,
+      "grad_norm": 0.7263943552970886,
+      "learning_rate": 1.6691125460033263e-09,
+      "loss": 1.4006,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9966101694915255,
+      "grad_norm": 0.7321056127548218,
+      "learning_rate": 1.159110861079782e-09,
+      "loss": 1.3213,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9972265023112481,
+      "grad_norm": 0.7304533123970032,
+      "learning_rate": 7.418335307174396e-10,
+      "loss": 1.3884,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9978428351309707,
+      "grad_norm": 0.7297084927558899,
+      "learning_rate": 4.1728248961758754e-10,
+      "loss": 1.3961,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9984591679506933,
+      "grad_norm": 0.7325052618980408,
+      "learning_rate": 1.8545924255652935e-10,
+      "loss": 1.316,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9990755007704161,
+      "grad_norm": 0.7296575903892517,
+      "learning_rate": 4.636486438114318e-11,
+      "loss": 1.3544,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9996918335901387,
+      "grad_norm": 0.7358527183532715,
+      "learning_rate": 0.0,
+      "loss": 1.4344,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9996918335901387,
+      "step": 1622,
+      "total_flos": 2.085852585718815e+19,
+      "train_loss": 1.4798993557213855,
+      "train_runtime": 22030.8952,
+      "train_samples_per_second": 2.357,
+      "train_steps_per_second": 0.074
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1622,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.085852585718815e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}