diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21760 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.094240837696335,
+  "global_step": 3600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-06,
+      "loss": 0.7153,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-06,
+      "loss": 0.6958,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7011,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 0.6957,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.5e-06,
+      "loss": 0.6746,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9e-06,
+      "loss": 0.7119,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.05e-05,
+      "loss": 0.6807,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7032,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.6606,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5e-05,
+      "loss": 0.6767,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.65e-05,
+      "loss": 0.73,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8e-05,
+      "loss": 0.7163,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95e-05,
+      "loss": 0.6514,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6609,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.25e-05,
+      "loss": 0.723,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6772,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.55e-05,
+      "loss": 0.649,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.6923,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 0.6637,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-05,
+      "loss": 0.6858,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9999997193849462e-05,
+      "loss": 0.6661,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9999988775398905e-05,
+      "loss": 0.6535,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9999974744651475e-05,
+      "loss": 0.6824,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9999955101612418e-05,
+      "loss": 0.6422,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9999929846289086e-05,
+      "loss": 0.6556,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999898978690932e-05,
+      "loss": 0.6515,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999862498829503e-05,
+      "loss": 0.6177,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999820406718447e-05,
+      "loss": 0.6386,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999772702373513e-05,
+      "loss": 0.6396,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999719385812552e-05,
+      "loss": 0.6225,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999660457055506e-05,
+      "loss": 0.6362,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999595916124435e-05,
+      "loss": 0.6346,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999525763043476e-05,
+      "loss": 0.6425,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999449997838886e-05,
+      "loss": 0.6719,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999368620539002e-05,
+      "loss": 0.597,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999281631174282e-05,
+      "loss": 0.6304,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999189029777273e-05,
+      "loss": 0.6627,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.999909081638261e-05,
+      "loss": 0.592,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9998986991027054e-05,
+      "loss": 0.6063,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9998877553749445e-05,
+      "loss": 0.6304,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9998762504590728e-05,
+      "loss": 0.6231,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9998641843593956e-05,
+      "loss": 0.6547,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9998515570804267e-05,
+      "loss": 0.6284,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999838368626891e-05,
+      "loss": 0.5911,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9998246190037228e-05,
+      "loss": 0.6095,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999810308216067e-05,
+      "loss": 0.6341,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9997954362692773e-05,
+      "loss": 0.6128,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999780003168919e-05,
+      "loss": 0.6455,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9997640089207655e-05,
+      "loss": 0.6394,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999747453530802e-05,
+      "loss": 0.6309,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999730337005222e-05,
+      "loss": 0.5972,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9997126593504303e-05,
+      "loss": 0.6112,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999694420573041e-05,
+      "loss": 0.6476,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9996756206798774e-05,
+      "loss": 0.6162,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9996562596779742e-05,
+      "loss": 0.63,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9996363375745762e-05,
+      "loss": 0.6257,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9996158543771358e-05,
+      "loss": 0.5764,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9995948100933174e-05,
+      "loss": 0.6339,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.999573204730995e-05,
+      "loss": 0.6279,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.9995510382982523e-05,
+      "loss": 0.5701,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9995283108033832e-05,
+      "loss": 0.6096,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9995050222548905e-05,
+      "loss": 0.588,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.999481172661488e-05,
+      "loss": 0.6113,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9994567620320993e-05,
+      "loss": 0.594,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.999431790375858e-05,
+      "loss": 0.6021,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9994062577021067e-05,
+      "loss": 0.5902,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9993801640203988e-05,
+      "loss": 0.6281,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9993535093404975e-05,
+      "loss": 0.5983,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9993262936723754e-05,
+      "loss": 0.6028,
+      "step": 69
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9992985170262162e-05,
+      "loss": 0.598,
+      "step": 70
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9992701794124113e-05,
+      "loss": 0.5827,
+      "step": 71
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9992412808415638e-05,
+      "loss": 0.6696,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9992118213244867e-05,
+      "loss": 0.607,
+      "step": 73
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9991818008722023e-05,
+      "loss": 0.5694,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9991512194959424e-05,
+      "loss": 0.6125,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9991200772071492e-05,
+      "loss": 0.5923,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.9990883740174748e-05,
+      "loss": 0.5588,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9990561099387814e-05,
+      "loss": 0.5921,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9990232849831406e-05,
+      "loss": 0.5705,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.998989899162833e-05,
+      "loss": 0.5575,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.998955952490351e-05,
+      "loss": 0.5739,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.998921444978396e-05,
+      "loss": 0.5689,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9988863766398784e-05,
+      "loss": 0.5833,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9988507474879198e-05,
+      "loss": 0.588,
+      "step": 84
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.99881455753585e-05,
+      "loss": 0.6027,
+      "step": 85
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9987778067972115e-05,
+      "loss": 0.5751,
+      "step": 86
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.998740495285752e-05,
+      "loss": 0.5958,
+      "step": 87
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9987026230154343e-05,
+      "loss": 0.5866,
+      "step": 88
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9986641900004267e-05,
+      "loss": 0.6029,
+      "step": 89
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.99862519625511e-05,
+      "loss": 0.58,
+      "step": 90
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9985856417940737e-05,
+      "loss": 0.561,
+      "step": 91
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9985455266321168e-05,
+      "loss": 0.5607,
+      "step": 92
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.9985048507842486e-05,
+      "loss": 0.647,
+      "step": 93
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.998463614265689e-05,
+      "loss": 0.606,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.998421817091865e-05,
+      "loss": 0.5647,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9983794592784166e-05,
+      "loss": 0.5931,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.998336540841192e-05,
+      "loss": 0.5699,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.998293061796249e-05,
+      "loss": 0.5598,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9982490221598553e-05,
+      "loss": 0.5817,
+      "step": 99
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.998204421948489e-05,
+      "loss": 0.5742,
+      "step": 100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9981592611788366e-05,
+      "loss": 0.5659,
+      "step": 101
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.998113539867796e-05,
+      "loss": 0.579,
+      "step": 102
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.998067258032473e-05,
+      "loss": 0.5934,
+      "step": 103
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9980204156901855e-05,
+      "loss": 0.5678,
+      "step": 104
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9979730128584584e-05,
+      "loss": 0.5705,
+      "step": 105
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9979250495550283e-05,
+      "loss": 0.5535,
+      "step": 106
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9978765257978408e-05,
+      "loss": 0.5874,
+      "step": 107
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9978274416050508e-05,
+      "loss": 0.6145,
+      "step": 108
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.997777796995024e-05,
+      "loss": 0.5545,
+      "step": 109
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.9977275919863344e-05,
+      "loss": 0.5431,
+      "step": 110
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.997676826597767e-05,
+      "loss": 0.5219,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9976255008483153e-05,
+      "loss": 0.5579,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.997573614757184e-05,
+      "loss": 0.5905,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9975211683437845e-05,
+      "loss": 0.5382,
+      "step": 114
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9974681616277416e-05,
+      "loss": 0.5732,
+      "step": 115
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.997414594628888e-05,
+      "loss": 0.5442,
+      "step": 116
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9973604673672643e-05,
+      "loss": 0.5479,
+      "step": 117
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9973057798631237e-05,
+      "loss": 0.5926,
+      "step": 118
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.997250532136928e-05,
+      "loss": 0.571,
+      "step": 119
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9971947242093474e-05,
+      "loss": 0.5335,
+      "step": 120
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9971383561012627e-05,
+      "loss": 0.5788,
+      "step": 121
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9970814278337648e-05,
+      "loss": 0.6215,
+      "step": 122
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9970239394281534e-05,
+      "loss": 0.5664,
+      "step": 123
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9969658909059378e-05,
+      "loss": 0.571,
+      "step": 124
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9969072822888376e-05,
+      "loss": 0.611,
+      "step": 125
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9968481135987807e-05,
+      "loss": 0.6318,
+      "step": 126
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9967883848579052e-05,
+      "loss": 0.547,
+      "step": 127
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.9967280960885596e-05,
+      "loss": 0.6071,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9966672473133003e-05,
+      "loss": 0.561,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.996605838554895e-05,
+      "loss": 0.5855,
+      "step": 130
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.996543869836319e-05,
+      "loss": 0.5734,
+      "step": 131
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.996481341180759e-05,
+      "loss": 0.5751,
+      "step": 132
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9964182526116095e-05,
+      "loss": 0.5681,
+      "step": 133
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.996354604152476e-05,
+      "loss": 0.5651,
+      "step": 134
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9962903958271723e-05,
+      "loss": 0.5713,
+      "step": 135
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9962256276597226e-05,
+      "loss": 0.5482,
+      "step": 136
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9961602996743595e-05,
+      "loss": 0.5739,
+      "step": 137
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9960944118955264e-05,
+      "loss": 0.5621,
+      "step": 138
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.996027964347875e-05,
+      "loss": 0.5392,
+      "step": 139
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.995960957056267e-05,
+      "loss": 0.5664,
+      "step": 140
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9958933900457733e-05,
+      "loss": 0.5568,
+      "step": 141
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9958252633416747e-05,
+      "loss": 0.5748,
+      "step": 142
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.99575657696946e-05,
+      "loss": 0.5356,
+      "step": 143
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.99568733095483e-05,
+      "loss": 0.5731,
+      "step": 144
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9956175253236923e-05,
+      "loss": 0.5612,
+      "step": 145
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.9955471601021645e-05,
+      "loss": 0.5187,
+      "step": 146
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9954762353165755e-05,
+      "loss": 0.5654,
+      "step": 147
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.995404750993461e-05,
+      "loss": 0.5358,
+      "step": 148
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.995332707159567e-05,
+      "loss": 0.6109,
+      "step": 149
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9952601038418495e-05,
+      "loss": 0.5554,
+      "step": 150
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9951869410674725e-05,
+      "loss": 0.5489,
+      "step": 151
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9951132188638113e-05,
+      "loss": 0.5647,
+      "step": 152
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.995038937258448e-05,
+      "loss": 0.5214,
+      "step": 153
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994964096279176e-05,
+      "loss": 0.5398,
+      "step": 154
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9948886959539982e-05,
+      "loss": 0.549,
+      "step": 155
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9948127363111236e-05,
+      "loss": 0.5433,
+      "step": 156
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994736217378975e-05,
+      "loss": 0.562,
+      "step": 157
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994659139186181e-05,
+      "loss": 0.548,
+      "step": 158
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994581501761581e-05,
+      "loss": 0.5539,
+      "step": 159
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994503305134223e-05,
+      "loss": 0.5789,
+      "step": 160
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9944245493333656e-05,
+      "loss": 0.5707,
+      "step": 161
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994345234388474e-05,
+      "loss": 0.6,
+      "step": 162
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.994265360329225e-05,
+      "loss": 0.573,
+      "step": 163
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9941849271855037e-05,
+      "loss": 0.58,
+      "step": 164
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9941039349874047e-05,
+      "loss": 0.5133,
+      "step": 165
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.994022383765231e-05,
+      "loss": 0.5487,
+      "step": 166
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9939402735494953e-05,
+      "loss": 0.585,
+      "step": 167
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.99385760437092e-05,
+      "loss": 0.5624,
+      "step": 168
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9937743762604356e-05,
+      "loss": 0.5281,
+      "step": 169
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.993690589249182e-05,
+      "loss": 0.5221,
+      "step": 170
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.993606243368509e-05,
+      "loss": 0.5578,
+      "step": 171
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9935213386499745e-05,
+      "loss": 0.5366,
+      "step": 172
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9934358751253454e-05,
+      "loss": 0.5637,
+      "step": 173
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.993349852826599e-05,
+      "loss": 0.5536,
+      "step": 174
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9932632717859206e-05,
+      "loss": 0.5197,
+      "step": 175
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9931761320357046e-05,
+      "loss": 0.5176,
+      "step": 176
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9930884336085548e-05,
+      "loss": 0.573,
+      "step": 177
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.993000176537284e-05,
+      "loss": 0.5581,
+      "step": 178
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9929113608549134e-05,
+      "loss": 0.558,
+      "step": 179
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.992821986594674e-05,
+      "loss": 0.5415,
+      "step": 180
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.992732053790006e-05,
+      "loss": 0.5378,
+      "step": 181
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.992641562474557e-05,
+      "loss": 0.5651,
+      "step": 182
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9925505126821853e-05,
+      "loss": 0.6225,
+      "step": 183
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.992458904446957e-05,
+      "loss": 0.5733,
+      "step": 184
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9923667378031488e-05,
+      "loss": 0.5399,
+      "step": 185
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9922740127852434e-05,
+      "loss": 0.5528,
+      "step": 186
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.992180729427936e-05,
+      "loss": 0.5378,
+      "step": 187
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9920868877661275e-05,
+      "loss": 0.5455,
+      "step": 188
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9919924878349297e-05,
+      "loss": 0.5761,
+      "step": 189
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9918975296696628e-05,
+      "loss": 0.5807,
+      "step": 190
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.991802013305855e-05,
+      "loss": 0.5471,
+      "step": 191
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.991705938779245e-05,
+      "loss": 0.5698,
+      "step": 192
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.991609306125779e-05,
+      "loss": 0.5581,
+      "step": 193
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9915121153816123e-05,
+      "loss": 0.5207,
+      "step": 194
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.991414366583109e-05,
+      "loss": 0.5443,
+      "step": 195
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9913160597668427e-05,
+      "loss": 0.544,
+      "step": 196
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9912171949695945e-05,
+      "loss": 0.503,
+      "step": 197
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.991117772228356e-05,
+      "loss": 0.553,
+      "step": 198
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9910177915803255e-05,
+      "loss": 0.5494,
+      "step": 199
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9909172530629118e-05,
+      "loss": 0.5578,
+      "step": 200
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.5553032755851746,
+      "eval_runtime": 1629.7738,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 200
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9908161567137313e-05,
+      "loss": 0.5242,
+      "step": 201
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.99071450257061e-05,
+      "loss": 0.5323,
+      "step": 202
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.990612290671582e-05,
+      "loss": 0.571,
+      "step": 203
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9905095210548896e-05,
+      "loss": 0.5539,
+      "step": 204
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9904061937589853e-05,
+      "loss": 0.5443,
+      "step": 205
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.990302308822529e-05,
+      "loss": 0.5483,
+      "step": 206
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9901978662843897e-05,
+      "loss": 0.5501,
+      "step": 207
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9900928661836448e-05,
+      "loss": 0.5454,
+      "step": 208
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9899873085595804e-05,
+      "loss": 0.5254,
+      "step": 209
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.989881193451691e-05,
+      "loss": 0.5579,
+      "step": 210
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9897745208996808e-05,
+      "loss": 0.5194,
+      "step": 211
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9896672909434604e-05,
+      "loss": 0.5196,
+      "step": 212
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9895595036231518e-05,
+      "loss": 0.5588,
+      "step": 213
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9894511589790827e-05,
+      "loss": 0.5517,
+      "step": 214
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.989342257051791e-05,
+      "loss": 0.553,
+      "step": 215
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.989232797882023e-05,
+      "loss": 0.5449,
+      "step": 216
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.989122781510733e-05,
+      "loss": 0.5151,
+      "step": 217
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9890122079790843e-05,
+      "loss": 0.5466,
+      "step": 218
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.988901077328448e-05,
+      "loss": 0.5422,
+      "step": 219
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9887893896004045e-05,
+      "loss": 0.547,
+      "step": 220
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.988677144836741e-05,
+      "loss": 0.562,
+      "step": 221
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9885643430794555e-05,
+      "loss": 0.5839,
+      "step": 222
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9884509843707526e-05,
+      "loss": 0.5191,
+      "step": 223
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9883370687530457e-05,
+      "loss": 0.536,
+      "step": 224
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9882225962689572e-05,
+      "loss": 0.5508,
+      "step": 225
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9881075669613168e-05,
+      "loss": 0.5377,
+      "step": 226
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9879919808731642e-05,
+      "loss": 0.6046,
+      "step": 227
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.987875838047745e-05,
+      "loss": 0.552,
+      "step": 228
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.987759138528515e-05,
+      "loss": 0.5302,
+      "step": 229
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9876418823591375e-05,
+      "loss": 0.5597,
+      "step": 230
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.987524069583485e-05,
+      "loss": 0.5477,
+      "step": 231
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.987405700245637e-05,
+      "loss": 0.5216,
+      "step": 232
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9872867743898818e-05,
+      "loss": 0.6011,
+      "step": 233
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9871672920607158e-05,
+      "loss": 0.5426,
+      "step": 234
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9870472533028437e-05,
+      "loss": 0.529,
+      "step": 235
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.986926658161179e-05,
+      "loss": 0.5509,
+      "step": 236
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9868055066808422e-05,
+      "loss": 0.547,
+      "step": 237
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.986683798907163e-05,
+      "loss": 0.5642,
+      "step": 238
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.986561534885678e-05,
+      "loss": 0.5161,
+      "step": 239
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9864387146621338e-05,
+      "loss": 0.5567,
+      "step": 240
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9863153382824828e-05,
+      "loss": 0.5433,
+      "step": 241
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9861914057928876e-05,
+      "loss": 0.5279,
+      "step": 242
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9860669172397174e-05,
+      "loss": 0.533,
+      "step": 243
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9859418726695505e-05,
+      "loss": 0.5691,
+      "step": 244
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9858162721291724e-05,
+      "loss": 0.5244,
+      "step": 245
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9856901156655774e-05,
+      "loss": 0.5634,
+      "step": 246
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.985563403325967e-05,
+      "loss": 0.5427,
+      "step": 247
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9854361351577508e-05,
+      "loss": 0.5268,
+      "step": 248
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.9853083112085472e-05,
+      "loss": 0.5552,
+      "step": 249
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9851799315261816e-05,
+      "loss": 0.5549,
+      "step": 250
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9850509961586875e-05,
+      "loss": 0.5448,
+      "step": 251
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.984921505154307e-05,
+      "loss": 0.5988,
+      "step": 252
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9847914585614898e-05,
+      "loss": 0.5334,
+      "step": 253
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9846608564288926e-05,
+      "loss": 0.543,
+      "step": 254
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.984529698805381e-05,
+      "loss": 0.5511,
+      "step": 255
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9843979857400275e-05,
+      "loss": 0.5266,
+      "step": 256
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.984265717282114e-05,
+      "loss": 0.5667,
+      "step": 257
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.984132893481129e-05,
+      "loss": 0.5459,
+      "step": 258
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9839995143867674e-05,
+      "loss": 0.5309,
+      "step": 259
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9838655800489357e-05,
+      "loss": 0.5355,
+      "step": 260
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.983731090517745e-05,
+      "loss": 0.546,
+      "step": 261
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9835960458435143e-05,
+      "loss": 0.5449,
+      "step": 262
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9834604460767717e-05,
+      "loss": 0.571,
+      "step": 263
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.9833242912682523e-05,
+      "loss": 0.5427,
+      "step": 264
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.983187581468899e-05,
+      "loss": 0.5316,
+      "step": 265
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.983050316729862e-05,
+      "loss": 0.5359,
+      "step": 266
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9829124971024996e-05,
+      "loss": 0.5508,
+      "step": 267
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9827741226383766e-05,
+      "loss": 0.5425,
+      "step": 268
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9826351933892675e-05,
+      "loss": 0.5584,
+      "step": 269
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.982495709407153e-05,
+      "loss": 0.5437,
+      "step": 270
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9823556707442205e-05,
+      "loss": 0.5134,
+      "step": 271
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.982215077452867e-05,
+      "loss": 0.5138,
+      "step": 272
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9820739295856954e-05,
+      "loss": 0.5367,
+      "step": 273
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9819322271955165e-05,
+      "loss": 0.5209,
+      "step": 274
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.981789970335349e-05,
+      "loss": 0.5088,
+      "step": 275
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9816471590584194e-05,
+      "loss": 0.5189,
+      "step": 276
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9815037934181605e-05,
+      "loss": 0.5432,
+      "step": 277
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.981359873468212e-05,
+      "loss": 0.5173,
+      "step": 278
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.981215399262424e-05,
+      "loss": 0.5256,
+      "step": 279
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.981070370854851e-05,
+      "loss": 0.5448,
+      "step": 280
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9809247882997554e-05,
+      "loss": 0.564,
+      "step": 281
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.980778651651608e-05,
+      "loss": 0.5377,
+      "step": 282
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9806319609650868e-05,
+      "loss": 0.5481,
+      "step": 283
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9804847162950753e-05,
+      "loss": 0.5195,
+      "step": 284
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9803369176966665e-05,
+      "loss": 0.5378,
+      "step": 285
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.98018856522516e-05,
+      "loss": 0.5286,
+      "step": 286
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.980039658936062e-05,
+      "loss": 0.5757,
+      "step": 287
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.979890198885086e-05,
+      "loss": 0.5602,
+      "step": 288
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9797401851281532e-05,
+      "loss": 0.53,
+      "step": 289
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9795896177213922e-05,
+      "loss": 0.5484,
+      "step": 290
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9794384967211378e-05,
+      "loss": 0.5415,
+      "step": 291
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9792868221839328e-05,
+      "loss": 0.5367,
+      "step": 292
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.979134594166526e-05,
+      "loss": 0.5366,
+      "step": 293
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.978981812725875e-05,
+      "loss": 0.5177,
+      "step": 294
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.978828477919143e-05,
+      "loss": 0.5421,
+      "step": 295
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9786745898037006e-05,
+      "loss": 0.5189,
+      "step": 296
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9785201484371258e-05,
+      "loss": 0.5597,
+      "step": 297
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9783651538772033e-05,
+      "loss": 0.5072,
+      "step": 298
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.978209606181925e-05,
+      "loss": 0.5602,
+      "step": 299
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.978053505409489e-05,
+      "loss": 0.4924,
+      "step": 300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.977896851618302e-05,
+      "loss": 0.5515,
+      "step": 301
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9777396448669757e-05,
+      "loss": 0.5502,
+      "step": 302
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9775818852143297e-05,
+      "loss": 0.5238,
+      "step": 303
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9774235727193905e-05,
+      "loss": 0.5559,
+      "step": 304
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.977264707441391e-05,
+      "loss": 0.5197,
+      "step": 305
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9771052894397714e-05,
+      "loss": 0.5262,
+      "step": 306
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9769453187741785e-05,
+      "loss": 0.5035,
+      "step": 307
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.976784795504466e-05,
+      "loss": 0.528,
+      "step": 308
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9766237196906935e-05,
+      "loss": 0.6093,
+      "step": 309
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9764620913931292e-05,
+      "loss": 0.5225,
+      "step": 310
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.976299910672246e-05,
+      "loss": 0.5564,
+      "step": 311
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.976137177588724e-05,
+      "loss": 0.5432,
+      "step": 312
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.975973892203452e-05,
+      "loss": 0.5538,
+      "step": 313
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.975810054577522e-05,
+      "loss": 0.5423,
+      "step": 314
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9756456647722356e-05,
+      "loss": 0.5278,
+      "step": 315
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.975480722849099e-05,
+      "loss": 0.5234,
+      "step": 316
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9753152288698265e-05,
+      "loss": 0.55,
+      "step": 317
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9751491828963376e-05,
+      "loss": 0.5303,
+      "step": 318
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9749825849907597e-05,
+      "loss": 0.5517,
+      "step": 319
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9748154352154255e-05,
+      "loss": 0.5476,
+      "step": 320
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.974647733632874e-05,
+      "loss": 0.5132,
+      "step": 321
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9744794803058528e-05,
+      "loss": 0.541,
+      "step": 322
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.974310675297313e-05,
+      "loss": 0.5356,
+      "step": 323
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9741413186704146e-05,
+      "loss": 0.5263,
+      "step": 324
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9739714104885228e-05,
+      "loss": 0.5445,
+      "step": 325
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9738009508152084e-05,
+      "loss": 0.5355,
+      "step": 326
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.973629939714251e-05,
+      "loss": 0.5317,
+      "step": 327
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9734583772496333e-05,
+      "loss": 0.532,
+      "step": 328
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.973286263485547e-05,
+      "loss": 0.5283,
+      "step": 329
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9731135984863893e-05,
+      "loss": 0.5179,
+      "step": 330
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.972940382316763e-05,
+      "loss": 0.5234,
+      "step": 331
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9727666150414773e-05,
+      "loss": 0.5502,
+      "step": 332
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.972592296725548e-05,
+      "loss": 0.6085,
+      "step": 333
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9724174274341972e-05,
+      "loss": 0.5254,
+      "step": 334
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9722420072328523e-05,
+      "loss": 0.5377,
+      "step": 335
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9720660361871478e-05,
+      "loss": 0.5125,
+      "step": 336
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9718895143629233e-05,
+      "loss": 0.5418,
+      "step": 337
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9717124418262257e-05,
+      "loss": 0.5222,
+      "step": 338
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.971534818643307e-05,
+      "loss": 0.522,
+      "step": 339
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9713566448806255e-05,
+      "loss": 0.518,
+      "step": 340
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9711779206048457e-05,
+      "loss": 0.6221,
+      "step": 341
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9709986458828376e-05,
+      "loss": 0.5239,
+      "step": 342
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9708188207816777e-05,
+      "loss": 0.518,
+      "step": 343
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.970638445368648e-05,
+      "loss": 0.5749,
+      "step": 344
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9704575197112368e-05,
+      "loss": 0.5346,
+      "step": 345
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9702760438771378e-05,
+      "loss": 0.5317,
+      "step": 346
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.970094017934251e-05,
+      "loss": 0.5671,
+      "step": 347
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9699114419506816e-05,
+      "loss": 0.5148,
+      "step": 348
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.9697283159947418e-05,
+      "loss": 0.5225,
+      "step": 349
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.969544640134948e-05,
+      "loss": 0.5393,
+      "step": 350
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.969360414440024e-05,
+      "loss": 0.5097,
+      "step": 351
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.969175638978898e-05,
+      "loss": 0.5125,
+      "step": 352
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9689903138207035e-05,
+      "loss": 0.5507,
+      "step": 353
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.968804439034782e-05,
+      "loss": 0.5253,
+      "step": 354
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9686180146906785e-05,
+      "loss": 0.5582,
+      "step": 355
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.968431040858144e-05,
+      "loss": 0.5137,
+      "step": 356
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.968243517607136e-05,
+      "loss": 0.5406,
+      "step": 357
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9680554450078172e-05,
+      "loss": 0.5376,
+      "step": 358
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9678668231305545e-05,
+      "loss": 0.5297,
+      "step": 359
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9676776520459223e-05,
+      "loss": 0.5328,
+      "step": 360
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9674879318246997e-05,
+      "loss": 0.5912,
+      "step": 361
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9672976625378706e-05,
+      "loss": 0.5234,
+      "step": 362
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.967106844256625e-05,
+      "loss": 0.5302,
+      "step": 363
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.966915477052358e-05,
+      "loss": 0.5232,
+      "step": 364
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.966723560996671e-05,
+      "loss": 0.5562,
+      "step": 365
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9665310961613696e-05,
+      "loss": 0.5443,
+      "step": 366
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.966338082618465e-05,
+      "loss": 0.4979,
+      "step": 367
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.966144520440175e-05,
+      "loss": 0.5292,
+      "step": 368
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.96595040969892e-05,
+      "loss": 0.5348,
+      "step": 369
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.965755750467328e-05,
+      "loss": 0.4987,
+      "step": 370
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9655605428182312e-05,
+      "loss": 0.5364,
+      "step": 371
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.965364786824667e-05,
+      "loss": 0.5363,
+      "step": 372
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9651684825598788e-05,
+      "loss": 0.6316,
+      "step": 373
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9649716300973143e-05,
+      "loss": 0.5826,
+      "step": 374
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.964774229510626e-05,
+      "loss": 0.5029,
+      "step": 375
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.964576280873673e-05,
+      "loss": 0.5316,
+      "step": 376
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9643777842605173e-05,
+      "loss": 0.5527,
+      "step": 377
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.964178739745428e-05,
+      "loss": 0.5195,
+      "step": 378
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9639791474028774e-05,
+      "loss": 0.5524,
+      "step": 379
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9637790073075442e-05,
+      "loss": 0.5197,
+      "step": 380
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9635783195343116e-05,
+      "loss": 0.5208,
+      "step": 381
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.963377084158268e-05,
+      "loss": 0.5519,
+      "step": 382
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9631753012547053e-05,
+      "loss": 0.5178,
+      "step": 383
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9629729708991213e-05,
+      "loss": 0.5469,
+      "step": 384
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.96277009316722e-05,
+      "loss": 0.5482,
+      "step": 385
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9625666681349068e-05,
+      "loss": 0.523,
+      "step": 386
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9623626958782956e-05,
+      "loss": 0.5295,
+      "step": 387
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9621581764737028e-05,
+      "loss": 0.5225,
+      "step": 388
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.961953109997649e-05,
+      "loss": 0.5253,
+      "step": 389
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9617474965268617e-05,
+      "loss": 0.5334,
+      "step": 390
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.961541336138272e-05,
+      "loss": 0.575,
+      "step": 391
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9613346289090145e-05,
+      "loss": 0.5269,
+      "step": 392
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.96112737491643e-05,
+      "loss": 0.5224,
+      "step": 393
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9609195742380635e-05,
+      "loss": 0.5284,
+      "step": 394
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.960711226951664e-05,
+      "loss": 0.5372,
+      "step": 395
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9605023331351854e-05,
+      "loss": 0.5677,
+      "step": 396
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.960292892866786e-05,
+      "loss": 0.5102,
+      "step": 397
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9600829062248288e-05,
+      "loss": 0.5743,
+      "step": 398
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9598723732878806e-05,
+      "loss": 0.5257,
+      "step": 399
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9596612941347135e-05,
+      "loss": 0.519,
+      "step": 400
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5348415970802307,
+      "eval_runtime": 1629.4084,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9594496688443035e-05,
+      "loss": 0.5256,
+      "step": 401
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.9592374974958304e-05,
+      "loss": 0.5181,
+      "step": 402
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.959024780168679e-05,
+      "loss": 0.529,
+      "step": 403
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9588115169424384e-05,
+      "loss": 0.5219,
+      "step": 404
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9585977078969015e-05,
+      "loss": 0.5013,
+      "step": 405
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9583833531120666e-05,
+      "loss": 0.4982,
+      "step": 406
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9581684526681337e-05,
+      "loss": 0.527,
+      "step": 407
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9579530066455097e-05,
+      "loss": 0.5104,
+      "step": 408
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9577370151248043e-05,
+      "loss": 0.5498,
+      "step": 409
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.957520478186831e-05,
+      "loss": 0.5508,
+      "step": 410
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9573033959126086e-05,
+      "loss": 0.5333,
+      "step": 411
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9570857683833582e-05,
+      "loss": 0.4636,
+      "step": 412
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.956867595680507e-05,
+      "loss": 0.5251,
+      "step": 413
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9566488778856837e-05,
+      "loss": 0.5452,
+      "step": 414
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9564296150807237e-05,
+      "loss": 0.5275,
+      "step": 415
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9562098073476643e-05,
+      "loss": 0.503,
+      "step": 416
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9559894547687467e-05,
+      "loss": 0.528,
+      "step": 417
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9557685574264176e-05,
+      "loss": 0.5122,
+      "step": 418
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.955547115403326e-05,
+      "loss": 0.5117,
+      "step": 419
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9553251287823252e-05,
+      "loss": 0.5163,
+      "step": 420
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9551025976464722e-05,
+      "loss": 0.5107,
+      "step": 421
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.954879522079028e-05,
+      "loss": 0.5438,
+      "step": 422
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.954655902163457e-05,
+      "loss": 0.5317,
+      "step": 423
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.954431737983427e-05,
+      "loss": 0.5186,
+      "step": 424
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9542070296228107e-05,
+      "loss": 0.4889,
+      "step": 425
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.953981777165682e-05,
+      "loss": 0.5441,
+      "step": 426
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9537559806963218e-05,
+      "loss": 0.5163,
+      "step": 427
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.953529640299211e-05,
+      "loss": 0.5241,
+      "step": 428
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9533027560590366e-05,
+      "loss": 0.5062,
+      "step": 429
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9530753280606874e-05,
+      "loss": 0.5219,
+      "step": 430
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9528473563892567e-05,
+      "loss": 0.5253,
+      "step": 431
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.952618841130041e-05,
+      "loss": 0.5015,
+      "step": 432
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9523897823685398e-05,
+      "loss": 0.5185,
+      "step": 433
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9521601801904565e-05,
+      "loss": 0.5382,
+      "step": 434
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9519300346816973e-05,
+      "loss": 0.5243,
+      "step": 435
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9516993459283724e-05,
+      "loss": 0.5288,
+      "step": 436
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9514681140167936e-05,
+      "loss": 0.5063,
+      "step": 437
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.9512363390334786e-05,
+      "loss": 0.5433,
+      "step": 438
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9510040210651456e-05,
+      "loss": 0.5453,
+      "step": 439
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9507711601987182e-05,
+      "loss": 0.5252,
+      "step": 440
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.950537756521321e-05,
+      "loss": 0.5176,
+      "step": 441
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9503038101202835e-05,
+      "loss": 0.5284,
+      "step": 442
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9500693210831377e-05,
+      "loss": 0.5311,
+      "step": 443
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9498342894976174e-05,
+      "loss": 0.5601,
+      "step": 444
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.949598715451662e-05,
+      "loss": 0.5318,
+      "step": 445
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9493625990334113e-05,
+      "loss": 0.5262,
+      "step": 446
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9491259403312093e-05,
+      "loss": 0.5487,
+      "step": 447
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9488887394336025e-05,
+      "loss": 0.5668,
+      "step": 448
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9486509964293404e-05,
+      "loss": 0.551,
+      "step": 449
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.948412711407376e-05,
+      "loss": 0.525,
+      "step": 450
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9481738844568637e-05,
+      "loss": 0.5279,
+      "step": 451
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9479345156671622e-05,
+      "loss": 0.5294,
+      "step": 452
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.947694605127831e-05,
+      "loss": 0.5248,
+      "step": 453
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9474541529286344e-05,
+      "loss": 0.5096,
+      "step": 454
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.947213159159538e-05,
+      "loss": 0.5576,
+      "step": 455
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.946971623910711e-05,
+      "loss": 0.5699,
+      "step": 456
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9467295472725237e-05,
+      "loss": 0.5303,
+      "step": 457
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9464869293355504e-05,
+      "loss": 0.5051,
+      "step": 458
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9462437701905672e-05,
+      "loss": 0.5,
+      "step": 459
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9460000699285535e-05,
+      "loss": 0.5412,
+      "step": 460
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9457558286406904e-05,
+      "loss": 0.5274,
+      "step": 461
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9455110464183612e-05,
+      "loss": 0.4832,
+      "step": 462
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.945265723353152e-05,
+      "loss": 0.5304,
+      "step": 463
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9450198595368518e-05,
+      "loss": 0.5129,
+      "step": 464
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.944773455061451e-05,
+      "loss": 0.5178,
+      "step": 465
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9445265100191426e-05,
+      "loss": 0.5294,
+      "step": 466
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.944279024502322e-05,
+      "loss": 0.5169,
+      "step": 467
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9440309986035873e-05,
+      "loss": 0.5127,
+      "step": 468
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9437824324157375e-05,
+      "loss": 0.5163,
+      "step": 469
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9435333260317744e-05,
+      "loss": 0.5073,
+      "step": 470
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.943283679544903e-05,
+      "loss": 0.5269,
+      "step": 471
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9430334930485284e-05,
+      "loss": 0.5265,
+      "step": 472
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.942782766636259e-05,
+      "loss": 0.4911,
+      "step": 473
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.942531500401905e-05,
+      "loss": 0.5274,
+      "step": 474
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9422796944394785e-05,
+      "loss": 0.529,
+      "step": 475
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9420273488431937e-05,
+      "loss": 0.5125,
+      "step": 476
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9417744637074664e-05,
+      "loss": 0.5004,
+      "step": 477
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9415210391269143e-05,
+      "loss": 0.5534,
+      "step": 478
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.941267075196357e-05,
+      "loss": 0.5778,
+      "step": 479
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.941012572010817e-05,
+      "loss": 0.5124,
+      "step": 480
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.940757529665516e-05,
+      "loss": 0.541,
+      "step": 481
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9405019482558803e-05,
+      "loss": 0.4948,
+      "step": 482
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9402458278775355e-05,
+      "loss": 0.5138,
+      "step": 483
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9399891686263107e-05,
+      "loss": 0.535,
+      "step": 484
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.939731970598235e-05,
+      "loss": 0.5169,
+      "step": 485
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.939474233889541e-05,
+      "loss": 0.5578,
+      "step": 486
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.939215958596661e-05,
+      "loss": 0.5484,
+      "step": 487
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9389571448162298e-05,
+      "loss": 0.5493,
+      "step": 488
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.9386977926450833e-05,
+      "loss": 0.4896,
+      "step": 489
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9384379021802593e-05,
+      "loss": 0.5146,
+      "step": 490
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9381774735189964e-05,
+      "loss": 0.5524,
+      "step": 491
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9379165067587346e-05,
+      "loss": 0.48,
+      "step": 492
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9376550019971166e-05,
+      "loss": 0.5063,
+      "step": 493
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9373929593319843e-05,
+      "loss": 0.5296,
+      "step": 494
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9371303788613823e-05,
+      "loss": 0.523,
+      "step": 495
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9368672606835554e-05,
+      "loss": 0.5097,
+      "step": 496
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9366036048969507e-05,
+      "loss": 0.5188,
+      "step": 497
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9363394116002163e-05,
+      "loss": 0.5205,
+      "step": 498
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9360746808922e-05,
+      "loss": 0.5386,
+      "step": 499
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9358094128719524e-05,
+      "loss": 0.5087,
+      "step": 500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9355436076387242e-05,
+      "loss": 0.5583,
+      "step": 501
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.935277265291968e-05,
+      "loss": 0.5414,
+      "step": 502
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.935010385931335e-05,
+      "loss": 0.4974,
+      "step": 503
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9347429696566806e-05,
+      "loss": 0.5196,
+      "step": 504
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.934475016568059e-05,
+      "loss": 0.5188,
+      "step": 505
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.9342065267657257e-05,
+      "loss": 0.5025,
+      "step": 506
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.933937500350137e-05,
+      "loss": 0.5469,
+      "step": 507
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9336679374219508e-05,
+      "loss": 0.5176,
+      "step": 508
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.933397838082024e-05,
+      "loss": 0.526,
+      "step": 509
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9331272024314154e-05,
+      "loss": 0.5169,
+      "step": 510
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9328560305713848e-05,
+      "loss": 0.5216,
+      "step": 511
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9325843226033917e-05,
+      "loss": 0.5269,
+      "step": 512
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9323120786290963e-05,
+      "loss": 0.5348,
+      "step": 513
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.93203929875036e-05,
+      "loss": 0.5347,
+      "step": 514
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.931765983069244e-05,
+      "loss": 0.504,
+      "step": 515
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9314921316880108e-05,
+      "loss": 0.5144,
+      "step": 516
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.931217744709122e-05,
+      "loss": 0.5093,
+      "step": 517
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9309428222352414e-05,
+      "loss": 0.5379,
+      "step": 518
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9306673643692312e-05,
+      "loss": 0.5435,
+      "step": 519
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.930391371214156e-05,
+      "loss": 0.5079,
+      "step": 520
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9301148428732788e-05,
+      "loss": 0.5309,
+      "step": 521
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9298377794500636e-05,
+      "loss": 0.5126,
+      "step": 522
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.929560181048175e-05,
+      "loss": 0.5361,
+      "step": 523
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.929282047771477e-05,
+      "loss": 0.5097,
+      "step": 524
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9290033797240345e-05,
+      "loss": 0.489,
+      "step": 525
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9287241770101117e-05,
+      "loss": 0.509,
+      "step": 526
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.928444439734174e-05,
+      "loss": 0.5081,
+      "step": 527
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9281641680008855e-05,
+      "loss": 0.5296,
+      "step": 528
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.92788336191511e-05,
+      "loss": 0.4862,
+      "step": 529
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.927602021581914e-05,
+      "loss": 0.5074,
+      "step": 530
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9273201471065602e-05,
+      "loss": 0.5053,
+      "step": 531
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9270377385945135e-05,
+      "loss": 0.5295,
+      "step": 532
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9267547961514386e-05,
+      "loss": 0.5248,
+      "step": 533
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.926471319883198e-05,
+      "loss": 0.5275,
+      "step": 534
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9261873098958566e-05,
+      "loss": 0.5551,
+      "step": 535
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9259027662956773e-05,
+      "loss": 0.547,
+      "step": 536
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.925617689189123e-05,
+      "loss": 0.5087,
+      "step": 537
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9253320786828564e-05,
+      "loss": 0.541,
+      "step": 538
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9250459348837392e-05,
+      "loss": 0.5192,
+      "step": 539
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9247592578988337e-05,
+      "loss": 0.5181,
+      "step": 540
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9244720478354003e-05,
+      "loss": 0.5715,
+      "step": 541
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9241843048009e-05,
+      "loss": 0.5647,
+      "step": 542
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9238960289029932e-05,
+      "loss": 0.5856,
+      "step": 543
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9236072202495396e-05,
+      "loss": 0.5069,
+      "step": 544
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.923317878948597e-05,
+      "loss": 0.5262,
+      "step": 545
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.923028005108423e-05,
+      "loss": 0.5094,
+      "step": 546
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9227375988374766e-05,
+      "loss": 0.5849,
+      "step": 547
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9224466602444128e-05,
+      "loss": 0.5476,
+      "step": 548
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9221551894380884e-05,
+      "loss": 0.5769,
+      "step": 549
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9218631865275577e-05,
+      "loss": 0.4851,
+      "step": 550
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.921570651622074e-05,
+      "loss": 0.493,
+      "step": 551
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9212775848310907e-05,
+      "loss": 0.4699,
+      "step": 552
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.92098398626426e-05,
+      "loss": 0.5248,
+      "step": 553
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.920689856031433e-05,
+      "loss": 0.5258,
+      "step": 554
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9203951942426587e-05,
+      "loss": 0.5431,
+      "step": 555
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9201000010081864e-05,
+      "loss": 0.5293,
+      "step": 556
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9198042764384628e-05,
+      "loss": 0.5036,
+      "step": 557
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9195080206441356e-05,
+      "loss": 0.4952,
+      "step": 558
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9192112337360498e-05,
+      "loss": 0.4974,
+      "step": 559
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9189139158252478e-05,
+      "loss": 0.5169,
+      "step": 560
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.918616067022973e-05,
+      "loss": 0.4749,
+      "step": 561
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9183176874406673e-05,
+      "loss": 0.5039,
+      "step": 562
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.918018777189969e-05,
+      "loss": 0.5663,
+      "step": 563
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9177193363827173e-05,
+      "loss": 0.4877,
+      "step": 564
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9174193651309488e-05,
+      "loss": 0.5285,
+      "step": 565
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9171188635468983e-05,
+      "loss": 0.4874,
+      "step": 566
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9168178317430004e-05,
+      "loss": 0.5472,
+      "step": 567
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9165162698318864e-05,
+      "loss": 0.4999,
+      "step": 568
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9162141779263876e-05,
+      "loss": 0.4976,
+      "step": 569
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9159115561395312e-05,
+      "loss": 0.5347,
+      "step": 570
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9156084045845458e-05,
+      "loss": 0.5122,
+      "step": 571
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9153047233748556e-05,
+      "loss": 0.5175,
+      "step": 572
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9150005126240846e-05,
+      "loss": 0.5036,
+      "step": 573
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9146957724460537e-05,
+      "loss": 0.5217,
+      "step": 574
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9143905029547827e-05,
+      "loss": 0.5016,
+      "step": 575
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9140847042644886e-05,
+      "loss": 0.502,
+      "step": 576
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9137783764895885e-05,
+      "loss": 0.525,
+      "step": 577
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.913471519744695e-05,
+      "loss": 0.51,
+      "step": 578
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9131641341446194e-05,
+      "loss": 0.5272,
+      "step": 579
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9128562198043717e-05,
+      "loss": 0.5134,
+      "step": 580
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9125477768391587e-05,
+      "loss": 0.5292,
+      "step": 581
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.912238805364385e-05,
+      "loss": 0.4876,
+      "step": 582
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9119293054956546e-05,
+      "loss": 0.4858,
+      "step": 583
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9116192773487667e-05,
+      "loss": 0.5244,
+      "step": 584
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.91130872103972e-05,
+      "loss": 0.5226,
+      "step": 585
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9109976366847098e-05,
+      "loss": 0.477,
+      "step": 586
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.91068602440013e-05,
+      "loss": 0.5384,
+      "step": 587
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9103738843025703e-05,
+      "loss": 0.4927,
+      "step": 588
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9100612165088202e-05,
+      "loss": 0.4778,
+      "step": 589
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9097480211358643e-05,
+      "loss": 0.4951,
+      "step": 590
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9094342983008873e-05,
+      "loss": 0.523,
+      "step": 591
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.909120048121268e-05,
+      "loss": 0.505,
+      "step": 592
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9088052707145848e-05,
+      "loss": 0.5311,
+      "step": 593
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9084899661986132e-05,
+      "loss": 0.5571,
+      "step": 594
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9081741346913245e-05,
+      "loss": 0.4918,
+      "step": 595
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9078577763108892e-05,
+      "loss": 0.5002,
+      "step": 596
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9075408911756732e-05,
+      "loss": 0.5797,
+      "step": 597
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.90722347940424e-05,
+      "loss": 0.5315,
+      "step": 598
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.906905541115351e-05,
+      "loss": 0.497,
+      "step": 599
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.906587076427963e-05,
+      "loss": 0.5132,
+      "step": 600
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5215901732444763,
+      "eval_runtime": 1631.7624,
+      "eval_samples_per_second": 0.602,
+      "eval_steps_per_second": 0.022,
+      "step": 600
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9062680854612314e-05,
+      "loss": 0.483,
+      "step": 601
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9059485683345074e-05,
+      "loss": 0.4987,
+      "step": 602
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9056285251673398e-05,
+      "loss": 0.5016,
+      "step": 603
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.905307956079473e-05,
+      "loss": 0.5128,
+      "step": 604
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.90498686119085e-05,
+      "loss": 0.5089,
+      "step": 605
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.904665240621609e-05,
+      "loss": 0.5249,
+      "step": 606
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9043430944920858e-05,
+      "loss": 0.529,
+      "step": 607
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.904020422922812e-05,
+      "loss": 0.5109,
+      "step": 608
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9036972260345165e-05,
+      "loss": 0.5474,
+      "step": 609
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.903373503948124e-05,
+      "loss": 0.5187,
+      "step": 610
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9030492567847575e-05,
+      "loss": 0.5026,
+      "step": 611
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.902724484665734e-05,
+      "loss": 0.5024,
+      "step": 612
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9023991877125684e-05,
+      "loss": 0.5098,
+      "step": 613
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9020733660469717e-05,
+      "loss": 0.4863,
+      "step": 614
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.901747019790852e-05,
+      "loss": 0.4984,
+      "step": 615
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.901420149066311e-05,
+      "loss": 0.5251,
+      "step": 616
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9010927539956505e-05,
+      "loss": 0.4805,
+      "step": 617
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.900764834701365e-05,
+      "loss": 0.568,
+      "step": 618
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9004363913061475e-05,
+      "loss": 0.5233,
+      "step": 619
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9001074239328856e-05,
+      "loss": 0.5043,
+      "step": 620
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.8997779327046642e-05,
+      "loss": 0.5575,
+      "step": 621
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.8994479177447633e-05,
+      "loss": 0.5403,
+      "step": 622
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.8991173791766586e-05,
+      "loss": 0.5322,
+      "step": 623
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.898786317124023e-05,
+      "loss": 0.5417,
+      "step": 624
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.8984547317107243e-05,
+      "loss": 0.4864,
+      "step": 625
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.898122623060826e-05,
+      "loss": 0.4948,
+      "step": 626
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.8977899912985877e-05,
+      "loss": 0.5284,
+      "step": 627
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8974568365484652e-05,
+      "loss": 0.5207,
+      "step": 628
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.897123158935109e-05,
+      "loss": 0.5251,
+      "step": 629
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.896788958583366e-05,
+      "loss": 0.523,
+      "step": 630
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.896454235618278e-05,
+      "loss": 0.4924,
+      "step": 631
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8961189901650834e-05,
+      "loss": 0.5162,
+      "step": 632
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.895783222349215e-05,
+      "loss": 0.5134,
+      "step": 633
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.895446932296301e-05,
+      "loss": 0.508,
+      "step": 634
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8951101201321667e-05,
+      "loss": 0.5431,
+      "step": 635
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8947727859828306e-05,
+      "loss": 0.5241,
+      "step": 636
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8944349299745072e-05,
+      "loss": 0.482,
+      "step": 637
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8940965522336074e-05,
+      "loss": 0.5096,
+      "step": 638
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.893757652886736e-05,
+      "loss": 0.5119,
+      "step": 639
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8934182320606923e-05,
+      "loss": 0.5207,
+      "step": 640
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8930782898824734e-05,
+      "loss": 0.5283,
+      "step": 641
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.892737826479269e-05,
+      "loss": 0.5286,
+      "step": 642
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.892396841978465e-05,
+      "loss": 0.5333,
+      "step": 643
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.8920553365076415e-05,
+      "loss": 0.5341,
+      "step": 644
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.891713310194574e-05,
+      "loss": 0.4886,
+      "step": 645
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.891370763167233e-05,
+      "loss": 0.5213,
+      "step": 646
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8910276955537835e-05,
+      "loss": 0.5021,
+      "step": 647
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8906841074825855e-05,
+      "loss": 0.5571,
+      "step": 648
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8903399990821932e-05,
+      "loss": 0.5303,
+      "step": 649
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.889995370481357e-05,
+      "loss": 0.5144,
+      "step": 650
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8896502218090197e-05,
+      "loss": 0.5137,
+      "step": 651
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8893045531943205e-05,
+      "loss": 0.5088,
+      "step": 652
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8889583647665927e-05,
+      "loss": 0.5299,
+      "step": 653
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8886116566553635e-05,
+      "loss": 0.4963,
+      "step": 654
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8882644289903542e-05,
+      "loss": 0.5051,
+      "step": 655
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8879166819014825e-05,
+      "loss": 0.5363,
+      "step": 656
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8875684155188587e-05,
+      "loss": 0.5091,
+      "step": 657
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8872196299727874e-05,
+      "loss": 0.5121,
+      "step": 658
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.8868703253937684e-05,
+      "loss": 0.5055,
+      "step": 659
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.886520501912495e-05,
+      "loss": 0.5193,
+      "step": 660
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.886170159659855e-05,
+      "loss": 0.4886,
+      "step": 661
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8858192987669303e-05,
+      "loss": 0.5151,
+      "step": 662
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8854679193649963e-05,
+      "loss": 0.5498,
+      "step": 663
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8851160215855227e-05,
+      "loss": 0.4924,
+      "step": 664
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8847636055601736e-05,
+      "loss": 0.5054,
+      "step": 665
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8844106714208064e-05,
+      "loss": 0.4716,
+      "step": 666
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8840572192994736e-05,
+      "loss": 0.5087,
+      "step": 667
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8837032493284188e-05,
+      "loss": 0.505,
+      "step": 668
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8833487616400823e-05,
+      "loss": 0.5402,
+      "step": 669
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8829937563670964e-05,
+      "loss": 0.509,
+      "step": 670
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.882638233642288e-05,
+      "loss": 0.5205,
+      "step": 671
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8822821935986764e-05,
+      "loss": 0.4988,
+      "step": 672
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8819256363694758e-05,
+      "loss": 0.5718,
+      "step": 673
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.881568562088093e-05,
+      "loss": 0.4871,
+      "step": 674
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8812109708881283e-05,
+      "loss": 0.5681,
+      "step": 675
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8808528629033765e-05,
+      "loss": 0.4768,
+      "step": 676
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8804942382678245e-05,
+      "loss": 0.5375,
+      "step": 677
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.8801350971156527e-05,
+      "loss": 0.5003,
+      "step": 678
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.879775439581235e-05,
+      "loss": 0.5206,
+      "step": 679
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.879415265799139e-05,
+      "loss": 0.5009,
+      "step": 680
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.879054575904125e-05,
+      "loss": 0.5174,
+      "step": 681
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.878693370031145e-05,
+      "loss": 0.5027,
+      "step": 682
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8783316483153465e-05,
+      "loss": 0.5342,
+      "step": 683
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.877969410892069e-05,
+      "loss": 0.5656,
+      "step": 684
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8776066578968447e-05,
+      "loss": 0.4984,
+      "step": 685
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8772433894653986e-05,
+      "loss": 0.5308,
+      "step": 686
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8768796057336494e-05,
+      "loss": 0.4966,
+      "step": 687
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8765153068377072e-05,
+      "loss": 0.5408,
+      "step": 688
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.876150492913876e-05,
+      "loss": 0.5347,
+      "step": 689
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8757851640986525e-05,
+      "loss": 0.5102,
+      "step": 690
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8754193205287258e-05,
+      "loss": 0.5088,
+      "step": 691
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8750529623409766e-05,
+      "loss": 0.4898,
+      "step": 692
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.87468608967248e-05,
+      "loss": 0.4889,
+      "step": 693
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8743187026605028e-05,
+      "loss": 0.5001,
+      "step": 694
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.8739508014425032e-05,
+      "loss": 0.5044,
+      "step": 695
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.873582386156133e-05,
+      "loss": 0.5105,
+      "step": 696
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8732134569392364e-05,
+      "loss": 0.5035,
+      "step": 697
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8728440139298492e-05,
+      "loss": 0.5131,
+      "step": 698
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8724740572662e-05,
+      "loss": 0.5045,
+      "step": 699
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8721035870867093e-05,
+      "loss": 0.5726,
+      "step": 700
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8717326035299893e-05,
+      "loss": 0.522,
+      "step": 701
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.871361106734845e-05,
+      "loss": 0.4899,
+      "step": 702
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.870989096840274e-05,
+      "loss": 0.4879,
+      "step": 703
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8706165739854637e-05,
+      "loss": 0.5533,
+      "step": 704
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.870243538309796e-05,
+      "loss": 0.5284,
+      "step": 705
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8698699899528428e-05,
+      "loss": 0.4662,
+      "step": 706
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8694959290543685e-05,
+      "loss": 0.5003,
+      "step": 707
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8691213557543293e-05,
+      "loss": 0.4761,
+      "step": 708
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8687462701928738e-05,
+      "loss": 0.5092,
+      "step": 709
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8683706725103406e-05,
+      "loss": 0.5361,
+      "step": 710
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.867994562847261e-05,
+      "loss": 0.5231,
+      "step": 711
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.8676179413443582e-05,
+      "loss": 0.4997,
+      "step": 712
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.867240808142546e-05,
+      "loss": 0.4798,
+      "step": 713
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8668631633829296e-05,
+      "loss": 0.5145,
+      "step": 714
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.866485007206807e-05,
+      "loss": 0.5365,
+      "step": 715
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8661063397556667e-05,
+      "loss": 0.4624,
+      "step": 716
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.865727161171188e-05,
+      "loss": 0.5068,
+      "step": 717
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8653474715952415e-05,
+      "loss": 0.5229,
+      "step": 718
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8649672711698896e-05,
+      "loss": 0.5048,
+      "step": 719
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8645865600373857e-05,
+      "loss": 0.4996,
+      "step": 720
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.864205338340174e-05,
+      "loss": 0.5277,
+      "step": 721
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8638236062208903e-05,
+      "loss": 0.5211,
+      "step": 722
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8634413638223612e-05,
+      "loss": 0.529,
+      "step": 723
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.863058611287603e-05,
+      "loss": 0.5075,
+      "step": 724
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.862675348759824e-05,
+      "loss": 0.5293,
+      "step": 725
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.862291576382424e-05,
+      "loss": 0.503,
+      "step": 726
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.861907294298992e-05,
+      "loss": 0.5223,
+      "step": 727
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8615225026533095e-05,
+      "loss": 0.4908,
+      "step": 728
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.861137201589346e-05,
+      "loss": 0.5373,
+      "step": 729
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.8607513912512645e-05,
+      "loss": 0.5279,
+      "step": 730
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8603650717834163e-05,
+      "loss": 0.5185,
+      "step": 731
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8599782433303452e-05,
+      "loss": 0.5178,
+      "step": 732
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8595909060367838e-05,
+      "loss": 0.495,
+      "step": 733
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8592030600476556e-05,
+      "loss": 0.5135,
+      "step": 734
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8588147055080743e-05,
+      "loss": 0.5071,
+      "step": 735
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8584258425633445e-05,
+      "loss": 0.5201,
+      "step": 736
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8580364713589605e-05,
+      "loss": 0.5175,
+      "step": 737
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.857646592040607e-05,
+      "loss": 0.5619,
+      "step": 738
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8572562047541585e-05,
+      "loss": 0.522,
+      "step": 739
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8568653096456793e-05,
+      "loss": 0.4892,
+      "step": 740
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.856473906861425e-05,
+      "loss": 0.4797,
+      "step": 741
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8560819965478397e-05,
+      "loss": 0.544,
+      "step": 742
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8556895788515585e-05,
+      "loss": 0.5029,
+      "step": 743
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.855296653919405e-05,
+      "loss": 0.5326,
+      "step": 744
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8549032218983934e-05,
+      "loss": 0.4968,
+      "step": 745
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.8545092829357288e-05,
+      "loss": 0.4814,
+      "step": 746
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.854114837178804e-05,
+      "loss": 0.5239,
+      "step": 747
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8537198847752024e-05,
+      "loss": 0.5122,
+      "step": 748
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8533244258726963e-05,
+      "loss": 0.518,
+      "step": 749
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8529284606192486e-05,
+      "loss": 0.4914,
+      "step": 750
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8525319891630108e-05,
+      "loss": 0.5039,
+      "step": 751
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8521350116523244e-05,
+      "loss": 0.487,
+      "step": 752
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8517375282357187e-05,
+      "loss": 0.5394,
+      "step": 753
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.851339539061915e-05,
+      "loss": 0.5296,
+      "step": 754
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8509410442798216e-05,
+      "loss": 0.4966,
+      "step": 755
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8505420440385366e-05,
+      "loss": 0.515,
+      "step": 756
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.850142538487348e-05,
+      "loss": 0.5056,
+      "step": 757
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.849742527775731e-05,
+      "loss": 0.5252,
+      "step": 758
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8493420120533516e-05,
+      "loss": 0.4955,
+      "step": 759
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8489409914700637e-05,
+      "loss": 0.5247,
+      "step": 760
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8485394661759114e-05,
+      "loss": 0.4882,
+      "step": 761
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8481374363211262e-05,
+      "loss": 0.4766,
+      "step": 762
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8477349020561288e-05,
+      "loss": 0.5222,
+      "step": 763
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.8473318635315293e-05,
+      "loss": 0.509,
+      "step": 764
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.846928320898125e-05,
+      "loss": 0.4936,
+      "step": 765
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8465242743069043e-05,
+      "loss": 0.5225,
+      "step": 766
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8461197239090407e-05,
+      "loss": 0.4973,
+      "step": 767
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8457146698558996e-05,
+      "loss": 0.5122,
+      "step": 768
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8453091122990325e-05,
+      "loss": 0.4924,
+      "step": 769
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8449030513901805e-05,
+      "loss": 0.4984,
+      "step": 770
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8444964872812728e-05,
+      "loss": 0.4971,
+      "step": 771
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8440894201244262e-05,
+      "loss": 0.5301,
+      "step": 772
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8436818500719472e-05,
+      "loss": 0.5045,
+      "step": 773
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.843273777276329e-05,
+      "loss": 0.5107,
+      "step": 774
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8428652018902526e-05,
+      "loss": 0.5004,
+      "step": 775
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8424561240665888e-05,
+      "loss": 0.4782,
+      "step": 776
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8420465439583958e-05,
+      "loss": 0.4626,
+      "step": 777
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.841636461718919e-05,
+      "loss": 0.5025,
+      "step": 778
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8412258775015917e-05,
+      "loss": 0.4845,
+      "step": 779
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8408147914600354e-05,
+      "loss": 0.4984,
+      "step": 780
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.84040320374806e-05,
+      "loss": 0.5067,
+      "step": 781
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.8399911145196615e-05,
+      "loss": 0.5073,
+      "step": 782
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8395785239290256e-05,
+      "loss": 0.4973,
+      "step": 783
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8391654321305237e-05,
+      "loss": 0.5176,
+      "step": 784
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8387518392787157e-05,
+      "loss": 0.4818,
+      "step": 785
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.838337745528349e-05,
+      "loss": 0.5061,
+      "step": 786
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8379231510343574e-05,
+      "loss": 0.5173,
+      "step": 787
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8375080559518636e-05,
+      "loss": 0.4983,
+      "step": 788
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8370924604361767e-05,
+      "loss": 0.5029,
+      "step": 789
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8366763646427932e-05,
+      "loss": 0.4968,
+      "step": 790
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.836259768727397e-05,
+      "loss": 0.4888,
+      "step": 791
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8358426728458584e-05,
+      "loss": 0.4987,
+      "step": 792
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.835425077154235e-05,
+      "loss": 0.5167,
+      "step": 793
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8350069818087724e-05,
+      "loss": 0.5204,
+      "step": 794
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.834588386965902e-05,
+      "loss": 0.5094,
+      "step": 795
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8341692927822426e-05,
+      "loss": 0.4998,
+      "step": 796
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8337496994145996e-05,
+      "loss": 0.5111,
+      "step": 797
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.833329607019965e-05,
+      "loss": 0.4872,
+      "step": 798
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.8329090157555188e-05,
+      "loss": 0.4954,
+      "step": 799
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.832487925778625e-05,
+      "loss": 0.4949,
+      "step": 800
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5112899541854858,
+      "eval_runtime": 1630.2396,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 800
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8320663372468376e-05,
+      "loss": 0.5574,
+      "step": 801
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8316442503178947e-05,
+      "loss": 0.4964,
+      "step": 802
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8312216651497208e-05,
+      "loss": 0.4954,
+      "step": 803
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8307985819004288e-05,
+      "loss": 0.4832,
+      "step": 804
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8303750007283164e-05,
+      "loss": 0.5011,
+      "step": 805
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8299509217918674e-05,
+      "loss": 0.5239,
+      "step": 806
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.829526345249753e-05,
+      "loss": 0.4579,
+      "step": 807
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8291012712608294e-05,
+      "loss": 0.5357,
+      "step": 808
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.82867569998414e-05,
+      "loss": 0.4833,
+      "step": 809
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.828249631578913e-05,
+      "loss": 0.575,
+      "step": 810
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8278230662045643e-05,
+      "loss": 0.4873,
+      "step": 811
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8273960040206944e-05,
+      "loss": 0.4915,
+      "step": 812
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.82696844518709e-05,
+      "loss": 0.4766,
+      "step": 813
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8265403898637234e-05,
+      "loss": 0.5139,
+      "step": 814
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8261118382107532e-05,
+      "loss": 0.5609,
+      "step": 815
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.8256827903885237e-05,
+      "loss": 0.5198,
+      "step": 816
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8252532465575646e-05,
+      "loss": 0.4819,
+      "step": 817
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8248232068785905e-05,
+      "loss": 0.4931,
+      "step": 818
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8243926715125032e-05,
+      "loss": 0.5121,
+      "step": 819
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.823961640620388e-05,
+      "loss": 0.4745,
+      "step": 820
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8235301143635178e-05,
+      "loss": 0.4768,
+      "step": 821
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.823098092903348e-05,
+      "loss": 0.5352,
+      "step": 822
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8226655764015227e-05,
+      "loss": 0.4871,
+      "step": 823
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8222325650198676e-05,
+      "loss": 0.4734,
+      "step": 824
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.821799058920397e-05,
+      "loss": 0.4659,
+      "step": 825
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8213650582653077e-05,
+      "loss": 0.4733,
+      "step": 826
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.820930563216983e-05,
+      "loss": 0.4811,
+      "step": 827
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8204955739379903e-05,
+      "loss": 0.4776,
+      "step": 828
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8200600905910823e-05,
+      "loss": 0.4601,
+      "step": 829
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8196241133391973e-05,
+      "loss": 0.4638,
+      "step": 830
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8191876423454567e-05,
+      "loss": 0.468,
+      "step": 831
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8187506777731683e-05,
+      "loss": 0.4947,
+      "step": 832
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.8183132197858235e-05,
+      "loss": 0.4937,
+      "step": 833
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.817875268547099e-05,
+      "loss": 0.483,
+      "step": 834
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8174368242208555e-05,
+      "loss": 0.5328,
+      "step": 835
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8169978869711387e-05,
+      "loss": 0.5059,
+      "step": 836
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8165584569621783e-05,
+      "loss": 0.4958,
+      "step": 837
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8161185343583885e-05,
+      "loss": 0.531,
+      "step": 838
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8156781193243678e-05,
+      "loss": 0.5285,
+      "step": 839
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.815237212024899e-05,
+      "loss": 0.5,
+      "step": 840
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8147958126249492e-05,
+      "loss": 0.4926,
+      "step": 841
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8143539212896692e-05,
+      "loss": 0.535,
+      "step": 842
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8139115381843944e-05,
+      "loss": 0.5149,
+      "step": 843
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8134686634746445e-05,
+      "loss": 0.4912,
+      "step": 844
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.813025297326121e-05,
+      "loss": 0.5049,
+      "step": 845
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8125814399047128e-05,
+      "loss": 0.5139,
+      "step": 846
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8121370913764892e-05,
+      "loss": 0.4923,
+      "step": 847
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.811692251907706e-05,
+      "loss": 0.5524,
+      "step": 848
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8112469216647996e-05,
+      "loss": 0.5068,
+      "step": 849
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.8108011008143933e-05,
+      "loss": 0.5308,
+      "step": 850
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8103547895232918e-05,
+      "loss": 0.4902,
+      "step": 851
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8099079879584843e-05,
+      "loss": 0.5003,
+      "step": 852
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8094606962871435e-05,
+      "loss": 0.4908,
+      "step": 853
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.809012914676624e-05,
+      "loss": 0.4936,
+      "step": 854
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.808564643294466e-05,
+      "loss": 0.4993,
+      "step": 855
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.808115882308391e-05,
+      "loss": 0.4972,
+      "step": 856
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8076666318863044e-05,
+      "loss": 0.4849,
+      "step": 857
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8072168921962957e-05,
+      "loss": 0.4506,
+      "step": 858
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8067666634066356e-05,
+      "loss": 0.5216,
+      "step": 859
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.806315945685779e-05,
+      "loss": 0.4861,
+      "step": 860
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8058647392023637e-05,
+      "loss": 0.4945,
+      "step": 861
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8054130441252095e-05,
+      "loss": 0.5441,
+      "step": 862
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8049608606233205e-05,
+      "loss": 0.4766,
+      "step": 863
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8045081888658825e-05,
+      "loss": 0.5003,
+      "step": 864
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8040550290222633e-05,
+      "loss": 0.4828,
+      "step": 865
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8036013812620153e-05,
+      "loss": 0.534,
+      "step": 866
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8031472457548717e-05,
+      "loss": 0.4976,
+      "step": 867
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.8026926226707493e-05,
+      "loss": 0.5125,
+      "step": 868
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.8022375121797462e-05,
+      "loss": 0.5327,
+      "step": 869
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.8017819144521446e-05,
+      "loss": 0.4871,
+      "step": 870
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.8013258296584074e-05,
+      "loss": 0.5067,
+      "step": 871
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.80086925796918e-05,
+      "loss": 0.5195,
+      "step": 872
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.80041219955529e-05,
+      "loss": 0.4923,
+      "step": 873
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7999546545877482e-05,
+      "loss": 0.5229,
+      "step": 874
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7994966232377464e-05,
+      "loss": 0.5058,
+      "step": 875
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7990381056766583e-05,
+      "loss": 0.4982,
+      "step": 876
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7985791020760396e-05,
+      "loss": 0.5356,
+      "step": 877
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7981196126076285e-05,
+      "loss": 0.5139,
+      "step": 878
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7976596374433447e-05,
+      "loss": 0.5208,
+      "step": 879
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.797199176755289e-05,
+      "loss": 0.5185,
+      "step": 880
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7967382307157447e-05,
+      "loss": 0.5042,
+      "step": 881
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.796276799497176e-05,
+      "loss": 0.5093,
+      "step": 882
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.7958148832722292e-05,
+      "loss": 0.5336,
+      "step": 883
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.795352482213732e-05,
+      "loss": 0.4919,
+      "step": 884
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.794889596494693e-05,
+      "loss": 0.5067,
+      "step": 885
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.794426226288302e-05,
+      "loss": 0.5325,
+      "step": 886
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.793962371767932e-05,
+      "loss": 0.4855,
+      "step": 887
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7934980331071343e-05,
+      "loss": 0.5194,
+      "step": 888
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.793033210479643e-05,
+      "loss": 0.5146,
+      "step": 889
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.792567904059374e-05,
+      "loss": 0.479,
+      "step": 890
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7921021140204224e-05,
+      "loss": 0.5041,
+      "step": 891
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7916358405370657e-05,
+      "loss": 0.5042,
+      "step": 892
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7911690837837614e-05,
+      "loss": 0.5675,
+      "step": 893
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.790701843935148e-05,
+      "loss": 0.5313,
+      "step": 894
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.790234121166045e-05,
+      "loss": 0.5271,
+      "step": 895
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7897659156514524e-05,
+      "loss": 0.471,
+      "step": 896
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7892972275665513e-05,
+      "loss": 0.4869,
+      "step": 897
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7888280570867027e-05,
+      "loss": 0.5246,
+      "step": 898
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.788358404387448e-05,
+      "loss": 0.4897,
+      "step": 899
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7878882696445095e-05,
+      "loss": 0.5294,
+      "step": 900
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.78741765303379e-05,
+      "loss": 0.4842,
+      "step": 901
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.7869465547313723e-05,
+      "loss": 0.5021,
+      "step": 902
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7864749749135192e-05,
+      "loss": 0.4658,
+      "step": 903
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7860029137566736e-05,
+      "loss": 0.5332,
+      "step": 904
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7855303714374593e-05,
+      "loss": 0.477,
+      "step": 905
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7850573481326798e-05,
+      "loss": 0.5066,
+      "step": 906
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.784583844019318e-05,
+      "loss": 0.5256,
+      "step": 907
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7841098592745374e-05,
+      "loss": 0.4969,
+      "step": 908
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7836353940756802e-05,
+      "loss": 0.5157,
+      "step": 909
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.78316044860027e-05,
+      "loss": 0.5084,
+      "step": 910
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.782685023026009e-05,
+      "loss": 0.4951,
+      "step": 911
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.782209117530779e-05,
+      "loss": 0.4799,
+      "step": 912
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.781732732292642e-05,
+      "loss": 0.506,
+      "step": 913
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7812558674898394e-05,
+      "loss": 0.5588,
+      "step": 914
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.7807785233007912e-05,
+      "loss": 0.4531,
+      "step": 915
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.780300699904098e-05,
+      "loss": 0.5101,
+      "step": 916
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.779822397478538e-05,
+      "loss": 0.5326,
+      "step": 917
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.779343616203071e-05,
+      "loss": 0.4904,
+      "step": 918
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.778864356256834e-05,
+      "loss": 0.4965,
+      "step": 919
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7783846178191435e-05,
+      "loss": 0.472,
+      "step": 920
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7779044010694956e-05,
+      "loss": 0.5043,
+      "step": 921
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.777423706187565e-05,
+      "loss": 0.4982,
+      "step": 922
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7769425333532054e-05,
+      "loss": 0.4861,
+      "step": 923
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.776460882746449e-05,
+      "loss": 0.5035,
+      "step": 924
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7759787545475072e-05,
+      "loss": 0.4874,
+      "step": 925
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.77549614893677e-05,
+      "loss": 0.5231,
+      "step": 926
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7750130660948058e-05,
+      "loss": 0.5217,
+      "step": 927
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.774529506202361e-05,
+      "loss": 0.4819,
+      "step": 928
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7740454694403625e-05,
+      "loss": 0.5021,
+      "step": 929
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7735609559899136e-05,
+      "loss": 0.5153,
+      "step": 930
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.773075966032297e-05,
+      "loss": 0.5474,
+      "step": 931
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7725904997489725e-05,
+      "loss": 0.5044,
+      "step": 932
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7721045573215795e-05,
+      "loss": 0.4858,
+      "step": 933
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7716181389319357e-05,
+      "loss": 0.4867,
+      "step": 934
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.771131244762035e-05,
+      "loss": 0.4969,
+      "step": 935
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.7706438749940512e-05,
+      "loss": 0.5205,
+      "step": 936
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7701560298103356e-05,
+      "loss": 0.5533,
+      "step": 937
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.769667709393416e-05,
+      "loss": 0.5089,
+      "step": 938
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7691789139260004e-05,
+      "loss": 0.5003,
+      "step": 939
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7686896435909724e-05,
+      "loss": 0.4951,
+      "step": 940
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7681998985713947e-05,
+      "loss": 0.4928,
+      "step": 941
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7677096790505074e-05,
+      "loss": 0.5476,
+      "step": 942
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7672189852117268e-05,
+      "loss": 0.4931,
+      "step": 943
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7667278172386484e-05,
+      "loss": 0.4897,
+      "step": 944
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.766236175315044e-05,
+      "loss": 0.5092,
+      "step": 945
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7657440596248633e-05,
+      "loss": 0.5143,
+      "step": 946
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.765251470352233e-05,
+      "loss": 0.5069,
+      "step": 947
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7647584076814574e-05,
+      "loss": 0.4904,
+      "step": 948
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7642648717970168e-05,
+      "loss": 0.4671,
+      "step": 949
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7637708628835698e-05,
+      "loss": 0.5185,
+      "step": 950
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.763276381125952e-05,
+      "loss": 0.4688,
+      "step": 951
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7627814267091743e-05,
+      "loss": 0.4829,
+      "step": 952
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.7622859998184266e-05,
+      "loss": 0.4949,
+      "step": 953
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.761790100639074e-05,
+      "loss": 0.4868,
+      "step": 954
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7612937293566592e-05,
+      "loss": 0.4979,
+      "step": 955
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.760796886156901e-05,
+      "loss": 0.5411,
+      "step": 956
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7602995712256947e-05,
+      "loss": 0.4982,
+      "step": 957
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.759801784749113e-05,
+      "loss": 0.5205,
+      "step": 958
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.759303526913404e-05,
+      "loss": 0.5018,
+      "step": 959
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7588047979049926e-05,
+      "loss": 0.5087,
+      "step": 960
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7583055979104798e-05,
+      "loss": 0.4786,
+      "step": 961
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7578059271166438e-05,
+      "loss": 0.531,
+      "step": 962
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7573057857104373e-05,
+      "loss": 0.5015,
+      "step": 963
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7568051738789903e-05,
+      "loss": 0.5041,
+      "step": 964
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7563040918096074e-05,
+      "loss": 0.5154,
+      "step": 965
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.755802539689772e-05,
+      "loss": 0.4986,
+      "step": 966
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7553005177071403e-05,
+      "loss": 0.4991,
+      "step": 967
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7547980260495452e-05,
+      "loss": 0.5344,
+      "step": 968
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7542950649049967e-05,
+      "loss": 0.5083,
+      "step": 969
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.753791634461679e-05,
+      "loss": 0.5125,
+      "step": 970
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.7532877349079522e-05,
+      "loss": 0.5064,
+      "step": 971
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.752783366432352e-05,
+      "loss": 0.4551,
+      "step": 972
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.75227852922359e-05,
+      "loss": 0.5013,
+      "step": 973
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7517732234705525e-05,
+      "loss": 0.4964,
+      "step": 974
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7512674493623013e-05,
+      "loss": 0.5013,
+      "step": 975
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7507612070880734e-05,
+      "loss": 0.4954,
+      "step": 976
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7502544968372812e-05,
+      "loss": 0.4882,
+      "step": 977
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7497473187995115e-05,
+      "loss": 0.5363,
+      "step": 978
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.749239673164528e-05,
+      "loss": 0.5142,
+      "step": 979
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7487315601222673e-05,
+      "loss": 0.506,
+      "step": 980
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7482229798628418e-05,
+      "loss": 0.4956,
+      "step": 981
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7477139325765377e-05,
+      "loss": 0.5096,
+      "step": 982
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.747204418453818e-05,
+      "loss": 0.499,
+      "step": 983
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7466944376853185e-05,
+      "loss": 0.485,
+      "step": 984
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.74618399046185e-05,
+      "loss": 0.5828,
+      "step": 985
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7456730769743986e-05,
+      "loss": 0.4825,
+      "step": 986
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.7451616974141243e-05,
+      "loss": 0.5137,
+      "step": 987
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.744649851972361e-05,
+      "loss": 0.5041,
+      "step": 988
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7441375408406178e-05,
+      "loss": 0.5158,
+      "step": 989
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.743624764210578e-05,
+      "loss": 0.5209,
+      "step": 990
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7431115222740974e-05,
+      "loss": 0.487,
+      "step": 991
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.742597815223209e-05,
+      "loss": 0.4833,
+      "step": 992
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7420836432501168e-05,
+      "loss": 0.4905,
+      "step": 993
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7415690065472003e-05,
+      "loss": 0.5091,
+      "step": 994
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7410539053070126e-05,
+      "loss": 0.4966,
+      "step": 995
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.740538339722281e-05,
+      "loss": 0.4731,
+      "step": 996
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7400223099859054e-05,
+      "loss": 0.5372,
+      "step": 997
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7395058162909603e-05,
+      "loss": 0.5074,
+      "step": 998
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7389888588306944e-05,
+      "loss": 0.5313,
+      "step": 999
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7384714377985275e-05,
+      "loss": 0.5144,
+      "step": 1000
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5035184025764465,
+      "eval_runtime": 1631.0565,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 1000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7379535533880557e-05,
+      "loss": 0.4913,
+      "step": 1001
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7374352057930473e-05,
+      "loss": 0.4703,
+      "step": 1002
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.736916395207443e-05,
+      "loss": 0.5134,
+      "step": 1003
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7363971218253575e-05,
+      "loss": 0.4859,
+      "step": 1004
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.7358773858410798e-05,
+      "loss": 0.5031,
+      "step": 1005
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7353571874490695e-05,
+      "loss": 0.5078,
+      "step": 1006
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7348365268439616e-05,
+      "loss": 0.4782,
+      "step": 1007
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7343154042205633e-05,
+      "loss": 0.5191,
+      "step": 1008
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7337938197738532e-05,
+      "loss": 0.5135,
+      "step": 1009
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7332717736989845e-05,
+      "loss": 0.5635,
+      "step": 1010
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7327492661912826e-05,
+      "loss": 0.506,
+      "step": 1011
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.732226297446246e-05,
+      "loss": 0.4912,
+      "step": 1012
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.731702867659544e-05,
+      "loss": 0.5476,
+      "step": 1013
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7311789770270205e-05,
+      "loss": 0.4945,
+      "step": 1014
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7306546257446912e-05,
+      "loss": 0.5544,
+      "step": 1015
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.730129814008743e-05,
+      "loss": 0.4811,
+      "step": 1016
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7296045420155366e-05,
+      "loss": 0.5254,
+      "step": 1017
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.729078809961604e-05,
+      "loss": 0.4748,
+      "step": 1018
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7285526180436503e-05,
+      "loss": 0.4723,
+      "step": 1019
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7280259664585508e-05,
+      "loss": 0.4957,
+      "step": 1020
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.7274988554033554e-05,
+      "loss": 0.4698,
+      "step": 1021
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.726971285075283e-05,
+      "loss": 0.492,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7264432556717274e-05,
+      "loss": 0.5154,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7259147673902516e-05,
+      "loss": 0.4808,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7253858204285914e-05,
+      "loss": 0.4829,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7248564149846542e-05,
+      "loss": 0.4957,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.724326551256519e-05,
+      "loss": 0.5079,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7237962294424358e-05,
+      "loss": 0.4745,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.723265449740827e-05,
+      "loss": 0.4869,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7227342123502846e-05,
+      "loss": 0.5056,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7222025174695734e-05,
+      "loss": 0.4967,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7216703652976294e-05,
+      "loss": 0.4786,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.721137756033558e-05,
+      "loss": 0.5112,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7206046898766378e-05,
+      "loss": 0.5793,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7200711670263167e-05,
+      "loss": 0.4977,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7195371876822145e-05,
+      "loss": 0.489,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.719002752044121e-05,
+      "loss": 0.5364,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.718467860311997e-05,
+      "loss": 0.4929,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.7179325126859746e-05,
+      "loss": 0.4857,
+      "step": 1039
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7173967093663562e-05,
+      "loss": 0.4991,
+      "step": 1040
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7168604505536135e-05,
+      "loss": 0.4859,
+      "step": 1041
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7163237364483897e-05,
+      "loss": 0.4866,
+      "step": 1042
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7157865672514992e-05,
+      "loss": 0.5176,
+      "step": 1043
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7152489431639243e-05,
+      "loss": 0.5109,
+      "step": 1044
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.71471086438682e-05,
+      "loss": 0.5356,
+      "step": 1045
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7141723311215095e-05,
+      "loss": 0.5206,
+      "step": 1046
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7136333435694876e-05,
+      "loss": 0.5056,
+      "step": 1047
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7130939019324177e-05,
+      "loss": 0.4634,
+      "step": 1048
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7125540064121342e-05,
+      "loss": 0.5066,
+      "step": 1049
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7120136572106405e-05,
+      "loss": 0.4937,
+      "step": 1050
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7114728545301102e-05,
+      "loss": 0.4766,
+      "step": 1051
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.710931598572887e-05,
+      "loss": 0.4927,
+      "step": 1052
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7103898895414823e-05,
+      "loss": 0.5093,
+      "step": 1053
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7098477276385796e-05,
+      "loss": 0.4686,
+      "step": 1054
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7093051130670305e-05,
+      "loss": 0.4694,
+      "step": 1055
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7087620460298553e-05,
+      "loss": 0.5193,
+      "step": 1056
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.7082185267302455e-05,
+      "loss": 0.5151,
+      "step": 1057
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7076745553715596e-05,
+      "loss": 0.4973,
+      "step": 1058
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.707130132157327e-05,
+      "loss": 0.5221,
+      "step": 1059
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7065852572912455e-05,
+      "loss": 0.4882,
+      "step": 1060
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.706039930977182e-05,
+      "loss": 0.4973,
+      "step": 1061
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7054941534191712e-05,
+      "loss": 0.4738,
+      "step": 1062
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.704947924821419e-05,
+      "loss": 0.4684,
+      "step": 1063
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7044012453882978e-05,
+      "loss": 0.4647,
+      "step": 1064
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7038541153243497e-05,
+      "loss": 0.4938,
+      "step": 1065
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7033065348342857e-05,
+      "loss": 0.5245,
+      "step": 1066
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7027585041229845e-05,
+      "loss": 0.4952,
+      "step": 1067
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.702210023395494e-05,
+      "loss": 0.5025,
+      "step": 1068
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7016610928570296e-05,
+      "loss": 0.4788,
+      "step": 1069
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7011117127129755e-05,
+      "loss": 0.5095,
+      "step": 1070
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.700561883168885e-05,
+      "loss": 0.4873,
+      "step": 1071
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.7000116044304776e-05,
+      "loss": 0.5131,
+      "step": 1072
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.6994608767036434e-05,
+      "loss": 0.5118,
+      "step": 1073
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.6989097001944375e-05,
+      "loss": 0.5126,
+      "step": 1074
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6983580751090852e-05,
+      "loss": 0.4846,
+      "step": 1075
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.697806001653979e-05,
+      "loss": 0.4926,
+      "step": 1076
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6972534800356793e-05,
+      "loss": 0.4725,
+      "step": 1077
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.696700510460913e-05,
+      "loss": 0.5247,
+      "step": 1078
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6961470931365767e-05,
+      "loss": 0.4697,
+      "step": 1079
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6955932282697326e-05,
+      "loss": 0.4868,
+      "step": 1080
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6950389160676108e-05,
+      "loss": 0.5054,
+      "step": 1081
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.69448415673761e-05,
+      "loss": 0.5127,
+      "step": 1082
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.693928950487295e-05,
+      "loss": 0.4921,
+      "step": 1083
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6933732975243978e-05,
+      "loss": 0.5009,
+      "step": 1084
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6928171980568184e-05,
+      "loss": 0.4941,
+      "step": 1085
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6922606522926227e-05,
+      "loss": 0.4972,
+      "step": 1086
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6917036604400446e-05,
+      "loss": 0.5107,
+      "step": 1087
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6911462227074843e-05,
+      "loss": 0.4664,
+      "step": 1088
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.690588339303509e-05,
+      "loss": 0.4857,
+      "step": 1089
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6900300104368527e-05,
+      "loss": 0.5355,
+      "step": 1090
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.6894712363164168e-05,
+      "loss": 0.473,
+      "step": 1091
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.688912017151267e-05,
+      "loss": 0.53,
+      "step": 1092
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6883523531506385e-05,
+      "loss": 0.4739,
+      "step": 1093
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6877922445239304e-05,
+      "loss": 0.4809,
+      "step": 1094
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.68723169148071e-05,
+      "loss": 0.5019,
+      "step": 1095
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.68667069423071e-05,
+      "loss": 0.4787,
+      "step": 1096
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6861092529838293e-05,
+      "loss": 0.5131,
+      "step": 1097
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6855473679501326e-05,
+      "loss": 0.4975,
+      "step": 1098
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.684985039339852e-05,
+      "loss": 0.5132,
+      "step": 1099
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.684422267363384e-05,
+      "loss": 0.4694,
+      "step": 1100
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6838590522312926e-05,
+      "loss": 0.4816,
+      "step": 1101
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6832953941543047e-05,
+      "loss": 0.5073,
+      "step": 1102
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.682731293343317e-05,
+      "loss": 0.5082,
+      "step": 1103
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.682166750009389e-05,
+      "loss": 0.5585,
+      "step": 1104
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6816017643637455e-05,
+      "loss": 0.4792,
+      "step": 1105
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6810363366177793e-05,
+      "loss": 0.4796,
+      "step": 1106
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.6804704669830465e-05,
+      "loss": 0.486,
+      "step": 1107
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.679904155671269e-05,
+      "loss": 0.4714,
+      "step": 1108
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6793374028943345e-05,
+      "loss": 0.4926,
+      "step": 1109
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.678770208864295e-05,
+      "loss": 0.4736,
+      "step": 1110
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6782025737933684e-05,
+      "loss": 0.4895,
+      "step": 1111
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6776344978939372e-05,
+      "loss": 0.4958,
+      "step": 1112
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.677065981378549e-05,
+      "loss": 0.4905,
+      "step": 1113
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.676497024459916e-05,
+      "loss": 0.4753,
+      "step": 1114
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6759276273509155e-05,
+      "loss": 0.5016,
+      "step": 1115
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6753577902645893e-05,
+      "loss": 0.5038,
+      "step": 1116
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.674787513414144e-05,
+      "loss": 0.4836,
+      "step": 1117
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6742167970129504e-05,
+      "loss": 0.458,
+      "step": 1118
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6736456412745444e-05,
+      "loss": 0.4877,
+      "step": 1119
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.673074046412626e-05,
+      "loss": 0.5022,
+      "step": 1120
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6725020126410588e-05,
+      "loss": 0.5097,
+      "step": 1121
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6719295401738707e-05,
+      "loss": 0.4768,
+      "step": 1122
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.6713566292252555e-05,
+      "loss": 0.5192,
+      "step": 1123
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.670783280009569e-05,
+      "loss": 0.5332,
+      "step": 1124
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.670209492741332e-05,
+      "loss": 0.4763,
+      "step": 1125
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6696352676352288e-05,
+      "loss": 0.4712,
+      "step": 1126
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6690606049061078e-05,
+      "loss": 0.46,
+      "step": 1127
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.668485504768981e-05,
+      "loss": 0.4748,
+      "step": 1128
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6679099674390242e-05,
+      "loss": 0.5125,
+      "step": 1129
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6673339931315762e-05,
+      "loss": 0.4807,
+      "step": 1130
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6667575820621406e-05,
+      "loss": 0.5366,
+      "step": 1131
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6661807344463826e-05,
+      "loss": 0.4882,
+      "step": 1132
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.665603450500132e-05,
+      "loss": 0.4923,
+      "step": 1133
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6650257304393823e-05,
+      "loss": 0.4947,
+      "step": 1134
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6644475744802884e-05,
+      "loss": 0.4646,
+      "step": 1135
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6638689828391696e-05,
+      "loss": 0.4886,
+      "step": 1136
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.663289955732508e-05,
+      "loss": 0.4793,
+      "step": 1137
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6627104933769493e-05,
+      "loss": 0.5417,
+      "step": 1138
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6621305959893002e-05,
+      "loss": 0.5176,
+      "step": 1139
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6615502637865315e-05,
+      "loss": 0.4679,
+      "step": 1140
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.660969496985777e-05,
+      "loss": 0.4947,
+      "step": 1141
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.6603882958043322e-05,
+      "loss": 0.5032,
+      "step": 1142
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.659806660459656e-05,
+      "loss": 0.4881,
+      "step": 1143
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6592245911693676e-05,
+      "loss": 0.4949,
+      "step": 1144
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6586420881512526e-05,
+      "loss": 0.4721,
+      "step": 1145
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6580591516232547e-05,
+      "loss": 0.5047,
+      "step": 1146
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6574757818034826e-05,
+      "loss": 0.4896,
+      "step": 1147
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6568919789102054e-05,
+      "loss": 0.4922,
+      "step": 1148
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6563077431618557e-05,
+      "loss": 0.4913,
+      "step": 1149
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6557230747770263e-05,
+      "loss": 0.5079,
+      "step": 1150
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6551379739744736e-05,
+      "loss": 0.4832,
+      "step": 1151
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.654552440973115e-05,
+      "loss": 0.4734,
+      "step": 1152
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6539664759920292e-05,
+      "loss": 0.4847,
+      "step": 1153
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6533800792504578e-05,
+      "loss": 0.508,
+      "step": 1154
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.652793250967802e-05,
+      "loss": 0.4814,
+      "step": 1155
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6522059913636265e-05,
+      "loss": 0.4917,
+      "step": 1156
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6516183006576566e-05,
+      "loss": 0.4957,
+      "step": 1157
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.651030179069778e-05,
+      "loss": 0.499,
+      "step": 1158
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.650441626820039e-05,
+      "loss": 0.4818,
+      "step": 1159
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.6498526441286484e-05,
+      "loss": 0.5876,
+      "step": 1160
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6492632312159754e-05,
+      "loss": 0.5103,
+      "step": 1161
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6486733883025517e-05,
+      "loss": 0.457,
+      "step": 1162
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6480831156090686e-05,
+      "loss": 0.5095,
+      "step": 1163
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.647492413356379e-05,
+      "loss": 0.477,
+      "step": 1164
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.646901281765496e-05,
+      "loss": 0.4921,
+      "step": 1165
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6463097210575927e-05,
+      "loss": 0.484,
+      "step": 1166
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.645717731454005e-05,
+      "loss": 0.4916,
+      "step": 1167
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6451253131762267e-05,
+      "loss": 0.5341,
+      "step": 1168
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6445324664459137e-05,
+      "loss": 0.4444,
+      "step": 1169
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.643939191484881e-05,
+      "loss": 0.5556,
+      "step": 1170
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.643345488515105e-05,
+      "loss": 0.4702,
+      "step": 1171
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.642751357758722e-05,
+      "loss": 0.4825,
+      "step": 1172
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6421567994380272e-05,
+      "loss": 0.5436,
+      "step": 1173
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6415618137754768e-05,
+      "loss": 0.5018,
+      "step": 1174
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6409664009936873e-05,
+      "loss": 0.53,
+      "step": 1175
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6403705613154336e-05,
+      "loss": 0.5095,
+      "step": 1176
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.6397742949636523e-05,
+      "loss": 0.4866,
+      "step": 1177
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6391776021614375e-05,
+      "loss": 0.5112,
+      "step": 1178
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6385804831320434e-05,
+      "loss": 0.4779,
+      "step": 1179
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6379829380988857e-05,
+      "loss": 0.4846,
+      "step": 1180
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.637384967285537e-05,
+      "loss": 0.4866,
+      "step": 1181
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6367865709157294e-05,
+      "loss": 0.4538,
+      "step": 1182
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6361877492133565e-05,
+      "loss": 0.4674,
+      "step": 1183
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6355885024024682e-05,
+      "loss": 0.4919,
+      "step": 1184
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6349888307072758e-05,
+      "loss": 0.4912,
+      "step": 1185
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.634388734352147e-05,
+      "loss": 0.4933,
+      "step": 1186
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.633788213561612e-05,
+      "loss": 0.5238,
+      "step": 1187
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6331872685603556e-05,
+      "loss": 0.5094,
+      "step": 1188
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6325858995732244e-05,
+      "loss": 0.4644,
+      "step": 1189
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.631984106825223e-05,
+      "loss": 0.4594,
+      "step": 1190
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6313818905415135e-05,
+      "loss": 0.502,
+      "step": 1191
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6307792509474174e-05,
+      "loss": 0.5018,
+      "step": 1192
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6301761882684147e-05,
+      "loss": 0.5318,
+      "step": 1193
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.6295727027301427e-05,
+      "loss": 0.4925,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.628968794558398e-05,
+      "loss": 0.471,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.628364463979135e-05,
+      "loss": 0.5119,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.627759711218466e-05,
+      "loss": 0.4793,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.627154536502661e-05,
+      "loss": 0.4589,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6265489400581486e-05,
+      "loss": 0.5019,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6259429221115143e-05,
+      "loss": 0.4954,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.496025025844574,
+      "eval_runtime": 1630.9694,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.625336482889502e-05,
+      "loss": 0.4832,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.624729622619013e-05,
+      "loss": 0.5603,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6241223415271064e-05,
+      "loss": 0.5035,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6235146398409985e-05,
+      "loss": 0.5029,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6229065177880622e-05,
+      "loss": 0.513,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.622297975595829e-05,
+      "loss": 0.4954,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6216890134919872e-05,
+      "loss": 0.481,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6210796317043818e-05,
+      "loss": 0.4825,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.620469830461015e-05,
+      "loss": 0.5194,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6198596099900454e-05,
+      "loss": 0.4937,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.6192489705197907e-05,
+      "loss": 0.5401,
+      "step": 1211
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6186379122787224e-05,
+      "loss": 0.4953,
+      "step": 1212
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6180264354954705e-05,
+      "loss": 0.5451,
+      "step": 1213
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6174145403988207e-05,
+      "loss": 0.4905,
+      "step": 1214
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6168022272177166e-05,
+      "loss": 0.5325,
+      "step": 1215
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6161894961812564e-05,
+      "loss": 0.4749,
+      "step": 1216
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6155763475186957e-05,
+      "loss": 0.4917,
+      "step": 1217
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.614962781459446e-05,
+      "loss": 0.4683,
+      "step": 1218
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6143487982330756e-05,
+      "loss": 0.4703,
+      "step": 1219
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.613734398069308e-05,
+      "loss": 0.4963,
+      "step": 1220
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6131195811980238e-05,
+      "loss": 0.5282,
+      "step": 1221
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6125043478492573e-05,
+      "loss": 0.4625,
+      "step": 1222
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.611888698253202e-05,
+      "loss": 0.4841,
+      "step": 1223
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6112726326402038e-05,
+      "loss": 0.4842,
+      "step": 1224
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6106561512407667e-05,
+      "loss": 0.4782,
+      "step": 1225
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.610039254285549e-05,
+      "loss": 0.489,
+      "step": 1226
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6094219420053644e-05,
+      "loss": 0.4526,
+      "step": 1227
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.6088042146311833e-05,
+      "loss": 0.4985,
+      "step": 1228
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.60818607239413e-05,
+      "loss": 0.4997,
+      "step": 1229
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6075675155254844e-05,
+      "loss": 0.4696,
+      "step": 1230
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.606948544256681e-05,
+      "loss": 0.5039,
+      "step": 1231
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.606329158819312e-05,
+      "loss": 0.4879,
+      "step": 1232
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6057093594451204e-05,
+      "loss": 0.4852,
+      "step": 1233
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6050891463660072e-05,
+      "loss": 0.4859,
+      "step": 1234
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.604468519814028e-05,
+      "loss": 0.4751,
+      "step": 1235
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.603847480021391e-05,
+      "loss": 0.478,
+      "step": 1236
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6032260272204614e-05,
+      "loss": 0.4807,
+      "step": 1237
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.602604161643757e-05,
+      "loss": 0.4797,
+      "step": 1238
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6019818835239508e-05,
+      "loss": 0.4812,
+      "step": 1239
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6013591930938716e-05,
+      "loss": 0.4758,
+      "step": 1240
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6007360905865e-05,
+      "loss": 0.475,
+      "step": 1241
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.6001125762349723e-05,
+      "loss": 0.5012,
+      "step": 1242
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.5994886502725783e-05,
+      "loss": 0.504,
+      "step": 1243
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.5988643129327624e-05,
+      "loss": 0.4745,
+      "step": 1244
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.5982395644491225e-05,
+      "loss": 0.4887,
+      "step": 1245
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.5976144050554098e-05,
+      "loss": 0.4762,
+      "step": 1246
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5969888349855304e-05,
+      "loss": 0.4861,
+      "step": 1247
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.596362854473544e-05,
+      "loss": 0.5072,
+      "step": 1248
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.595736463753662e-05,
+      "loss": 0.5,
+      "step": 1249
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5951096630602506e-05,
+      "loss": 0.5376,
+      "step": 1250
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.594482452627831e-05,
+      "loss": 0.4842,
+      "step": 1251
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.593854832691075e-05,
+      "loss": 0.4922,
+      "step": 1252
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5932268034848086e-05,
+      "loss": 0.4518,
+      "step": 1253
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5925983652440117e-05,
+      "loss": 0.4868,
+      "step": 1254
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5919695182038166e-05,
+      "loss": 0.5333,
+      "step": 1255
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.591340262599508e-05,
+      "loss": 0.5102,
+      "step": 1256
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.590710598666524e-05,
+      "loss": 0.5093,
+      "step": 1257
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.590080526640456e-05,
+      "loss": 0.4836,
+      "step": 1258
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5894500467570475e-05,
+      "loss": 0.4793,
+      "step": 1259
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5888191592521945e-05,
+      "loss": 0.4939,
+      "step": 1260
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5881878643619456e-05,
+      "loss": 0.529,
+      "step": 1261
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5875561623225023e-05,
+      "loss": 0.5008,
+      "step": 1262
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.5869240533702178e-05,
+      "loss": 0.4601,
+      "step": 1263
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5862915377415984e-05,
+      "loss": 0.4661,
+      "step": 1264
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5856586156733012e-05,
+      "loss": 0.4914,
+      "step": 1265
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.585025287402136e-05,
+      "loss": 0.4741,
+      "step": 1266
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5843915531650657e-05,
+      "loss": 0.4875,
+      "step": 1267
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5837574131992033e-05,
+      "loss": 0.5044,
+      "step": 1268
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5831228677418148e-05,
+      "loss": 0.4744,
+      "step": 1269
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5824879170303177e-05,
+      "loss": 0.4847,
+      "step": 1270
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5818525613022805e-05,
+      "loss": 0.513,
+      "step": 1271
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5812168007954238e-05,
+      "loss": 0.5,
+      "step": 1272
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.58058063574762e-05,
+      "loss": 0.4887,
+      "step": 1273
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.579944066396892e-05,
+      "loss": 0.4788,
+      "step": 1274
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5793070929814145e-05,
+      "loss": 0.5257,
+      "step": 1275
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.578669715739513e-05,
+      "loss": 0.4896,
+      "step": 1276
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.578031934909665e-05,
+      "loss": 0.5051,
+      "step": 1277
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.577393750730498e-05,
+      "loss": 0.4949,
+      "step": 1278
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5767551634407907e-05,
+      "loss": 0.5006,
+      "step": 1279
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.5761161732794725e-05,
+      "loss": 0.4866,
+      "step": 1280
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5754767804856242e-05,
+      "loss": 0.4772,
+      "step": 1281
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.574836985298477e-05,
+      "loss": 0.4711,
+      "step": 1282
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.574196787957412e-05,
+      "loss": 0.4465,
+      "step": 1283
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.573556188701961e-05,
+      "loss": 0.5006,
+      "step": 1284
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.572915187771807e-05,
+      "loss": 0.4794,
+      "step": 1285
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5722737854067825e-05,
+      "loss": 0.4931,
+      "step": 1286
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5716319818468697e-05,
+      "loss": 0.4635,
+      "step": 1287
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5709897773322024e-05,
+      "loss": 0.4693,
+      "step": 1288
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5703471721030636e-05,
+      "loss": 0.4935,
+      "step": 1289
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5697041663998855e-05,
+      "loss": 0.4992,
+      "step": 1290
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5690607604632518e-05,
+      "loss": 0.4863,
+      "step": 1291
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.568416954533894e-05,
+      "loss": 0.4513,
+      "step": 1292
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5677727488526957e-05,
+      "loss": 0.487,
+      "step": 1293
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.567128143660687e-05,
+      "loss": 0.4773,
+      "step": 1294
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5664831391990504e-05,
+      "loss": 0.4847,
+      "step": 1295
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.565837735709116e-05,
+      "loss": 0.4694,
+      "step": 1296
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.5651919334323636e-05,
+      "loss": 0.5088,
+      "step": 1297
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5645457326104228e-05,
+      "loss": 0.4486,
+      "step": 1298
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.563899133485071e-05,
+      "loss": 0.4763,
+      "step": 1299
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5632521362982363e-05,
+      "loss": 0.5388,
+      "step": 1300
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5626047412919946e-05,
+      "loss": 0.4778,
+      "step": 1301
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.561956948708571e-05,
+      "loss": 0.5251,
+      "step": 1302
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5613087587903388e-05,
+      "loss": 0.5112,
+      "step": 1303
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5606601717798212e-05,
+      "loss": 0.512,
+      "step": 1304
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.560011187919689e-05,
+      "loss": 0.4677,
+      "step": 1305
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5593618074527618e-05,
+      "loss": 0.4747,
+      "step": 1306
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5587120306220072e-05,
+      "loss": 0.483,
+      "step": 1307
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5580618576705414e-05,
+      "loss": 0.4563,
+      "step": 1308
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.557411288841629e-05,
+      "loss": 0.5022,
+      "step": 1309
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5567603243786826e-05,
+      "loss": 0.4815,
+      "step": 1310
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.556108964525263e-05,
+      "loss": 0.4731,
+      "step": 1311
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5554572095250784e-05,
+      "loss": 0.4838,
+      "step": 1312
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.554805059621985e-05,
+      "loss": 0.4833,
+      "step": 1313
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.554152515059987e-05,
+      "loss": 0.5018,
+      "step": 1314
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.5534995760832364e-05,
+      "loss": 0.5034,
+      "step": 1315
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5528462429360324e-05,
+      "loss": 0.4926,
+      "step": 1316
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.552192515862822e-05,
+      "loss": 0.4666,
+      "step": 1317
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.551538395108199e-05,
+      "loss": 0.4964,
+      "step": 1318
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.550883880916905e-05,
+      "loss": 0.4734,
+      "step": 1319
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5502289735338292e-05,
+      "loss": 0.4652,
+      "step": 1320
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5495736732040065e-05,
+      "loss": 0.5503,
+      "step": 1321
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5489179801726204e-05,
+      "loss": 0.4757,
+      "step": 1322
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.548261894685001e-05,
+      "loss": 0.4728,
+      "step": 1323
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5476054169866246e-05,
+      "loss": 0.5671,
+      "step": 1324
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.546948547323114e-05,
+      "loss": 0.4679,
+      "step": 1325
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5462912859402402e-05,
+      "loss": 0.5197,
+      "step": 1326
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5456336330839194e-05,
+      "loss": 0.4813,
+      "step": 1327
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.544975589000214e-05,
+      "loss": 0.4797,
+      "step": 1328
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5443171539353343e-05,
+      "loss": 0.4522,
+      "step": 1329
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.543658328135636e-05,
+      "loss": 0.4904,
+      "step": 1330
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.54299911184762e-05,
+      "loss": 0.5025,
+      "step": 1331
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5423395053179355e-05,
+      "loss": 0.4538,
+      "step": 1332
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5416795087933756e-05,
+      "loss": 0.4849,
+      "step": 1333
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5410191225208813e-05,
+      "loss": 0.5095,
+      "step": 1334
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5403583467475375e-05,
+      "loss": 0.5178,
+      "step": 1335
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5396971817205764e-05,
+      "loss": 0.4827,
+      "step": 1336
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5390356276873745e-05,
+      "loss": 0.4854,
+      "step": 1337
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5383736848954544e-05,
+      "loss": 0.4945,
+      "step": 1338
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.537711353592485e-05,
+      "loss": 0.4987,
+      "step": 1339
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5370486340262793e-05,
+      "loss": 0.5006,
+      "step": 1340
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.536385526444796e-05,
+      "loss": 0.499,
+      "step": 1341
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5357220310961394e-05,
+      "loss": 0.4705,
+      "step": 1342
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5350581482285583e-05,
+      "loss": 0.4772,
+      "step": 1343
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.534393878090447e-05,
+      "loss": 0.4824,
+      "step": 1344
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5337292209303437e-05,
+      "loss": 0.4781,
+      "step": 1345
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5330641769969334e-05,
+      "loss": 0.5041,
+      "step": 1346
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5323987465390433e-05,
+      "loss": 0.476,
+      "step": 1347
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5317329298056473e-05,
+      "loss": 0.4534,
+      "step": 1348
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5310667270458624e-05,
+      "loss": 0.4503,
+      "step": 1349
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5304001385089516e-05,
+      "loss": 0.5065,
+      "step": 1350
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.52973316444432e-05,
+      "loss": 0.4835,
+      "step": 1351
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5290658051015192e-05,
+      "loss": 0.4946,
+      "step": 1352
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5283980607302432e-05,
+      "loss": 0.4744,
+      "step": 1353
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.527729931580332e-05,
+      "loss": 0.4806,
+      "step": 1354
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5270614179017666e-05,
+      "loss": 0.4774,
+      "step": 1355
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5263925199446754e-05,
+      "loss": 0.4557,
+      "step": 1356
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.525723237959328e-05,
+      "loss": 0.4737,
+      "step": 1357
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5250535721961384e-05,
+      "loss": 0.4859,
+      "step": 1358
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5243835229056648e-05,
+      "loss": 0.489,
+      "step": 1359
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5237130903386085e-05,
+      "loss": 0.4817,
+      "step": 1360
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5230422747458138e-05,
+      "loss": 0.5028,
+      "step": 1361
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5223710763782684e-05,
+      "loss": 0.468,
+      "step": 1362
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.521699495487104e-05,
+      "loss": 0.5186,
+      "step": 1363
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5210275323235945e-05,
+      "loss": 0.4754,
+      "step": 1364
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5203551871391568e-05,
+      "loss": 0.4559,
+      "step": 1365
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.5196824601853523e-05,
+      "loss": 0.4951,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5190093517138834e-05,
+      "loss": 0.4783,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5183358619765963e-05,
+      "loss": 0.4557,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5176619912254783e-05,
+      "loss": 0.472,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5169877397126618e-05,
+      "loss": 0.4583,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.51631310769042e-05,
+      "loss": 0.4989,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5156380954111685e-05,
+      "loss": 0.5026,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.514962703127465e-05,
+      "loss": 0.4825,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5142869310920104e-05,
+      "loss": 0.5233,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5136107795576476e-05,
+      "loss": 0.4622,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5129342487773597e-05,
+      "loss": 0.4974,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5122573390042743e-05,
+      "loss": 0.493,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5115800504916585e-05,
+      "loss": 0.4843,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.510902383492923e-05,
+      "loss": 0.4855,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.510224338261619e-05,
+      "loss": 0.5305,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5095459150514382e-05,
+      "loss": 0.4557,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.508867114116216e-05,
+      "loss": 0.4977,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.5081879357099286e-05,
+      "loss": 0.4786,
+      "step": 1383
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5075083800866924e-05,
+      "loss": 0.4871,
+      "step": 1384
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5068284475007652e-05,
+      "loss": 0.4829,
+      "step": 1385
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.506148138206546e-05,
+      "loss": 0.4813,
+      "step": 1386
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.505467452458575e-05,
+      "loss": 0.48,
+      "step": 1387
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5047863905115338e-05,
+      "loss": 0.4722,
+      "step": 1388
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5041049526202427e-05,
+      "loss": 0.4805,
+      "step": 1389
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5034231390396655e-05,
+      "loss": 0.4948,
+      "step": 1390
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5027409500249037e-05,
+      "loss": 0.4569,
+      "step": 1391
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5020583858312017e-05,
+      "loss": 0.4881,
+      "step": 1392
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5013754467139428e-05,
+      "loss": 0.4583,
+      "step": 1393
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5006921329286508e-05,
+      "loss": 0.4616,
+      "step": 1394
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.5000084447309898e-05,
+      "loss": 0.483,
+      "step": 1395
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.499324382376765e-05,
+      "loss": 0.4691,
+      "step": 1396
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.498639946121919e-05,
+      "loss": 0.4658,
+      "step": 1397
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.4979551362225373e-05,
+      "loss": 0.4751,
+      "step": 1398
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.497269952934844e-05,
+      "loss": 0.4726,
+      "step": 1399
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.4965843965152015e-05,
+      "loss": 0.5112,
+      "step": 1400
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.4895597994327545,
+      "eval_runtime": 1629.6202,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 1400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4958984672201145e-05,
+      "loss": 0.4913,
+      "step": 1401
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.495212165306225e-05,
+      "loss": 0.5168,
+      "step": 1402
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.494525491030315e-05,
+      "loss": 0.4978,
+      "step": 1403
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4938384446493067e-05,
+      "loss": 0.4821,
+      "step": 1404
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4931510264202598e-05,
+      "loss": 0.4667,
+      "step": 1405
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.492463236600375e-05,
+      "loss": 0.5034,
+      "step": 1406
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4917750754469912e-05,
+      "loss": 0.503,
+      "step": 1407
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.491086543217586e-05,
+      "loss": 0.4823,
+      "step": 1408
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4903976401697764e-05,
+      "loss": 0.4564,
+      "step": 1409
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4897083665613173e-05,
+      "loss": 0.493,
+      "step": 1410
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4890187226501034e-05,
+      "loss": 0.462,
+      "step": 1411
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4883287086941668e-05,
+      "loss": 0.484,
+      "step": 1412
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4876383249516785e-05,
+      "loss": 0.4696,
+      "step": 1413
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4869475716809483e-05,
+      "loss": 0.4755,
+      "step": 1414
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4862564491404236e-05,
+      "loss": 0.5009,
+      "step": 1415
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4855649575886907e-05,
+      "loss": 0.5153,
+      "step": 1416
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.484873097284473e-05,
+      "loss": 0.4697,
+      "step": 1417
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.4841808684866326e-05,
+      "loss": 0.4481,
+      "step": 1418
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4834882714541696e-05,
+      "loss": 0.446,
+      "step": 1419
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4827953064462207e-05,
+      "loss": 0.5499,
+      "step": 1420
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4821019737220618e-05,
+      "loss": 0.4743,
+      "step": 1421
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4814082735411055e-05,
+      "loss": 0.4743,
+      "step": 1422
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4807142061629023e-05,
+      "loss": 0.484,
+      "step": 1423
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4800197718471396e-05,
+      "loss": 0.4728,
+      "step": 1424
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4793249708536422e-05,
+      "loss": 0.4989,
+      "step": 1425
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.478629803442373e-05,
+      "loss": 0.5548,
+      "step": 1426
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4779342698734304e-05,
+      "loss": 0.5066,
+      "step": 1427
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4772383704070507e-05,
+      "loss": 0.468,
+      "step": 1428
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4765421053036076e-05,
+      "loss": 0.4629,
+      "step": 1429
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.475845474823611e-05,
+      "loss": 0.5146,
+      "step": 1430
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4751484792277073e-05,
+      "loss": 0.4534,
+      "step": 1431
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.4744511187766798e-05,
+      "loss": 0.467,
+      "step": 1432
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.473753393731448e-05,
+      "loss": 0.4708,
+      "step": 1433
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.473055304353069e-05,
+      "loss": 0.5501,
+      "step": 1434
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.472356850902735e-05,
+      "loss": 0.4774,
+      "step": 1435
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4716580336417738e-05,
+      "loss": 0.4746,
+      "step": 1436
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.470958852831651e-05,
+      "loss": 0.5016,
+      "step": 1437
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.470259308733968e-05,
+      "loss": 0.4796,
+      "step": 1438
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4695594016104605e-05,
+      "loss": 0.499,
+      "step": 1439
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4688591317230022e-05,
+      "loss": 0.4772,
+      "step": 1440
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4681584993336006e-05,
+      "loss": 0.4989,
+      "step": 1441
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4674575047044e-05,
+      "loss": 0.4797,
+      "step": 1442
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.46675614809768e-05,
+      "loss": 0.5104,
+      "step": 1443
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.466054429775856e-05,
+      "loss": 0.4721,
+      "step": 1444
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4653523500014767e-05,
+      "loss": 0.5009,
+      "step": 1445
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4646499090372296e-05,
+      "loss": 0.4882,
+      "step": 1446
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4639471071459346e-05,
+      "loss": 0.4717,
+      "step": 1447
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.463243944590547e-05,
+      "loss": 0.4905,
+      "step": 1448
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4625404216341577e-05,
+      "loss": 0.4552,
+      "step": 1449
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.461836538539993e-05,
+      "loss": 0.4931,
+      "step": 1450
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4611322955714118e-05,
+      "loss": 0.5269,
+      "step": 1451
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.46042769299191e-05,
+      "loss": 0.5111,
+      "step": 1452
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.459722731065117e-05,
+      "loss": 0.4703,
+      "step": 1453
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.459017410054796e-05,
+      "loss": 0.5178,
+      "step": 1454
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.458311730224846e-05,
+      "loss": 0.5054,
+      "step": 1455
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.457605691839299e-05,
+      "loss": 0.477,
+      "step": 1456
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4568992951623223e-05,
+      "loss": 0.4661,
+      "step": 1457
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4561925404582158e-05,
+      "loss": 0.4978,
+      "step": 1458
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.455485427991415e-05,
+      "loss": 0.4601,
+      "step": 1459
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4547779580264876e-05,
+      "loss": 0.5148,
+      "step": 1460
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4540701308281364e-05,
+      "loss": 0.5011,
+      "step": 1461
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4533619466611972e-05,
+      "loss": 0.4883,
+      "step": 1462
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4526534057906397e-05,
+      "loss": 0.4987,
+      "step": 1463
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.451944508481567e-05,
+      "loss": 0.4701,
+      "step": 1464
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.451235254999215e-05,
+      "loss": 0.468,
+      "step": 1465
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.450525645608954e-05,
+      "loss": 0.4841,
+      "step": 1466
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4498156805762854e-05,
+      "loss": 0.4967,
+      "step": 1467
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4491053601668466e-05,
+      "loss": 0.5091,
+      "step": 1468
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4483946846464053e-05,
+      "loss": 0.4731,
+      "step": 1469
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4476836542808638e-05,
+      "loss": 0.5005,
+      "step": 1470
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4469722693362564e-05,
+      "loss": 0.4833,
+      "step": 1471
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.44626053007875e-05,
+      "loss": 0.5057,
+      "step": 1472
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4455484367746445e-05,
+      "loss": 0.5795,
+      "step": 1473
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.444835989690372e-05,
+      "loss": 0.481,
+      "step": 1474
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.444123189092497e-05,
+      "loss": 0.4586,
+      "step": 1475
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4434100352477156e-05,
+      "loss": 0.4685,
+      "step": 1476
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4426965284228576e-05,
+      "loss": 0.4605,
+      "step": 1477
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.441982668884884e-05,
+      "loss": 0.5113,
+      "step": 1478
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4412684569008877e-05,
+      "loss": 0.4913,
+      "step": 1479
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4405538927380927e-05,
+      "loss": 0.4912,
+      "step": 1480
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.439838976663857e-05,
+      "loss": 0.4754,
+      "step": 1481
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4391237089456677e-05,
+      "loss": 0.4683,
+      "step": 1482
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4384080898511454e-05,
+      "loss": 0.5139,
+      "step": 1483
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4376921196480406e-05,
+      "loss": 0.4634,
+      "step": 1484
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4369757986042368e-05,
+      "loss": 0.5328,
+      "step": 1485
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.4362591269877482e-05,
+      "loss": 0.4401,
+      "step": 1486
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4355421050667188e-05,
+      "loss": 0.4924,
+      "step": 1487
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4348247331094256e-05,
+      "loss": 0.4758,
+      "step": 1488
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4341070113842757e-05,
+      "loss": 0.4836,
+      "step": 1489
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4333889401598067e-05,
+      "loss": 0.5002,
+      "step": 1490
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.432670519704688e-05,
+      "loss": 0.4769,
+      "step": 1491
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4319517502877186e-05,
+      "loss": 0.4861,
+      "step": 1492
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4312326321778288e-05,
+      "loss": 0.4918,
+      "step": 1493
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4305131656440783e-05,
+      "loss": 0.4761,
+      "step": 1494
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4297933509556593e-05,
+      "loss": 0.5001,
+      "step": 1495
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.429073188381892e-05,
+      "loss": 0.4724,
+      "step": 1496
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.428352678192228e-05,
+      "loss": 0.4936,
+      "step": 1497
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4276318206562485e-05,
+      "loss": 0.5435,
+      "step": 1498
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4269106160436654e-05,
+      "loss": 0.4767,
+      "step": 1499
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4261890646243182e-05,
+      "loss": 0.5082,
+      "step": 1500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.42546716666818e-05,
+      "loss": 0.472,
+      "step": 1501
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4247449224453494e-05,
+      "loss": 0.4899,
+      "step": 1502
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4240223322260578e-05,
+      "loss": 0.5039,
+      "step": 1503
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.4232993962806645e-05,
+      "loss": 0.4567,
+      "step": 1504
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4225761148796588e-05,
+      "loss": 0.4438,
+      "step": 1505
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4218524882936578e-05,
+      "loss": 0.5171,
+      "step": 1506
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.42112851679341e-05,
+      "loss": 0.4848,
+      "step": 1507
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.420404200649791e-05,
+      "loss": 0.4681,
+      "step": 1508
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4196795401338066e-05,
+      "loss": 0.4792,
+      "step": 1509
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4189545355165908e-05,
+      "loss": 0.479,
+      "step": 1510
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4182291870694066e-05,
+      "loss": 0.4843,
+      "step": 1511
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.417503495063646e-05,
+      "loss": 0.4838,
+      "step": 1512
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4167774597708286e-05,
+      "loss": 0.5005,
+      "step": 1513
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.416051081462603e-05,
+      "loss": 0.4797,
+      "step": 1514
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.415324360410746e-05,
+      "loss": 0.4736,
+      "step": 1515
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4145972968871632e-05,
+      "loss": 0.4731,
+      "step": 1516
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4138698911638875e-05,
+      "loss": 0.4795,
+      "step": 1517
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.413142143513081e-05,
+      "loss": 0.4606,
+      "step": 1518
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4124140542070323e-05,
+      "loss": 0.4582,
+      "step": 1519
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.4116856235181584e-05,
+      "loss": 0.5023,
+      "step": 1520
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.410956851719004e-05,
+      "loss": 0.4398,
+      "step": 1521
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4102277390822423e-05,
+      "loss": 0.502,
+      "step": 1522
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4094982858806725e-05,
+      "loss": 0.4624,
+      "step": 1523
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4087684923872225e-05,
+      "loss": 0.4558,
+      "step": 1524
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.408038358874947e-05,
+      "loss": 0.4989,
+      "step": 1525
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4073078856170273e-05,
+      "loss": 0.4962,
+      "step": 1526
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4065770728867735e-05,
+      "loss": 0.4796,
+      "step": 1527
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4058459209576207e-05,
+      "loss": 0.4922,
+      "step": 1528
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4051144301031326e-05,
+      "loss": 0.4624,
+      "step": 1529
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4043826005969986e-05,
+      "loss": 0.5094,
+      "step": 1530
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4036504327130355e-05,
+      "loss": 0.491,
+      "step": 1531
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4029179267251854e-05,
+      "loss": 0.4894,
+      "step": 1532
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.4021850829075193e-05,
+      "loss": 0.463,
+      "step": 1533
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.401451901534233e-05,
+      "loss": 0.486,
+      "step": 1534
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.400718382879648e-05,
+      "loss": 0.4871,
+      "step": 1535
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.399984527218213e-05,
+      "loss": 0.5024,
+      "step": 1536
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3992503348245032e-05,
+      "loss": 0.4615,
+      "step": 1537
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.3985158059732188e-05,
+      "loss": 0.4855,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.397780940939186e-05,
+      "loss": 0.4872,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3970457399973575e-05,
+      "loss": 0.4736,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3963102034228113e-05,
+      "loss": 0.4882,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3955743314907503e-05,
+      "loss": 0.4752,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3948381244765038e-05,
+      "loss": 0.4862,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.394101582655527e-05,
+      "loss": 0.473,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.393364706303398e-05,
+      "loss": 0.4733,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3926274956958224e-05,
+      "loss": 0.461,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3918899511086304e-05,
+      "loss": 0.4989,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3911520728177763e-05,
+      "loss": 0.4792,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3904138610993397e-05,
+      "loss": 0.4751,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.389675316229525e-05,
+      "loss": 0.4801,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3889364384846625e-05,
+      "loss": 0.4565,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.388197228141204e-05,
+      "loss": 0.4679,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3874576854757286e-05,
+      "loss": 0.4675,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3867178107649384e-05,
+      "loss": 0.476,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.3859776042856607e-05,
+      "loss": 0.4517,
+      "step": 1555
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3852370663148454e-05,
+      "loss": 0.4537,
+      "step": 1556
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3844961971295672e-05,
+      "loss": 0.4474,
+      "step": 1557
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3837549970070257e-05,
+      "loss": 0.4656,
+      "step": 1558
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.383013466224543e-05,
+      "loss": 0.4925,
+      "step": 1559
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3822716050595655e-05,
+      "loss": 0.483,
+      "step": 1560
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3815294137896627e-05,
+      "loss": 0.5114,
+      "step": 1561
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3807868926925283e-05,
+      "loss": 0.5426,
+      "step": 1562
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3800440420459792e-05,
+      "loss": 0.4832,
+      "step": 1563
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3793008621279554e-05,
+      "loss": 0.4576,
+      "step": 1564
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3785573532165198e-05,
+      "loss": 0.5538,
+      "step": 1565
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3778135155898594e-05,
+      "loss": 0.4748,
+      "step": 1566
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.377069349526283e-05,
+      "loss": 0.4661,
+      "step": 1567
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3763248553042236e-05,
+      "loss": 0.4704,
+      "step": 1568
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3755800332022357e-05,
+      "loss": 0.4937,
+      "step": 1569
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.374834883498997e-05,
+      "loss": 0.4894,
+      "step": 1570
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3740894064733082e-05,
+      "loss": 0.4632,
+      "step": 1571
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.3733436024040917e-05,
+      "loss": 0.548,
+      "step": 1572
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3725974715703926e-05,
+      "loss": 0.4858,
+      "step": 1573
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3718510142513785e-05,
+      "loss": 0.4894,
+      "step": 1574
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.371104230726339e-05,
+      "loss": 0.4441,
+      "step": 1575
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3703571212746855e-05,
+      "loss": 0.4867,
+      "step": 1576
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3696096861759514e-05,
+      "loss": 0.4534,
+      "step": 1577
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.368861925709793e-05,
+      "loss": 0.4944,
+      "step": 1578
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.368113840155986e-05,
+      "loss": 0.4571,
+      "step": 1579
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3673654297944308e-05,
+      "loss": 0.4714,
+      "step": 1580
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.366616694905146e-05,
+      "loss": 0.5014,
+      "step": 1581
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3658676357682744e-05,
+      "loss": 0.5014,
+      "step": 1582
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3651182526640786e-05,
+      "loss": 0.4715,
+      "step": 1583
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.364368545872943e-05,
+      "loss": 0.4727,
+      "step": 1584
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.363618515675373e-05,
+      "loss": 0.5346,
+      "step": 1585
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.362868162351995e-05,
+      "loss": 0.4732,
+      "step": 1586
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3621174861835555e-05,
+      "loss": 0.4662,
+      "step": 1587
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.361366487450924e-05,
+      "loss": 0.4778,
+      "step": 1588
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.3606151664350875e-05,
+      "loss": 0.5011,
+      "step": 1589
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.359863523417157e-05,
+      "loss": 0.4623,
+      "step": 1590
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3591115586783612e-05,
+      "loss": 0.4358,
+      "step": 1591
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.35835927250005e-05,
+      "loss": 0.4907,
+      "step": 1592
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3576066651636954e-05,
+      "loss": 0.4868,
+      "step": 1593
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3568537369508868e-05,
+      "loss": 0.4756,
+      "step": 1594
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.356100488143335e-05,
+      "loss": 0.4981,
+      "step": 1595
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.35534691902287e-05,
+      "loss": 0.5876,
+      "step": 1596
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3545930298714436e-05,
+      "loss": 0.4984,
+      "step": 1597
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.353838820971125e-05,
+      "loss": 0.4976,
+      "step": 1598
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3530842926041043e-05,
+      "loss": 0.4644,
+      "step": 1599
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.352329445052691e-05,
+      "loss": 0.4632,
+      "step": 1600
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.4851336181163788,
+      "eval_runtime": 1632.0561,
+      "eval_samples_per_second": 0.602,
+      "eval_steps_per_second": 0.022,
+      "step": 1600
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3515742785993138e-05,
+      "loss": 0.4626,
+      "step": 1601
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3508187935265206e-05,
+      "loss": 0.4894,
+      "step": 1602
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3500629901169794e-05,
+      "loss": 0.4739,
+      "step": 1603
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.349306868653476e-05,
+      "loss": 0.4572,
+      "step": 1604
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3485504294189154e-05,
+      "loss": 0.4774,
+      "step": 1605
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3477936726963228e-05,
+      "loss": 0.472,
+      "step": 1606
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.347036598768841e-05,
+      "loss": 0.4955,
+      "step": 1607
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3462792079197313e-05,
+      "loss": 0.5247,
+      "step": 1608
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.345521500432375e-05,
+      "loss": 0.4865,
+      "step": 1609
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3447634765902704e-05,
+      "loss": 0.4606,
+      "step": 1610
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3440051366770344e-05,
+      "loss": 0.4805,
+      "step": 1611
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.343246480976403e-05,
+      "loss": 0.4272,
+      "step": 1612
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3424875097722298e-05,
+      "loss": 0.4895,
+      "step": 1613
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3417282233484857e-05,
+      "loss": 0.4685,
+      "step": 1614
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3409686219892615e-05,
+      "loss": 0.461,
+      "step": 1615
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3402087059787636e-05,
+      "loss": 0.4951,
+      "step": 1616
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.339448475601318e-05,
+      "loss": 0.4505,
+      "step": 1617
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3386879311413666e-05,
+      "loss": 0.4593,
+      "step": 1618
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3379270728834703e-05,
+      "loss": 0.4958,
+      "step": 1619
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.337165901112307e-05,
+      "loss": 0.4674,
+      "step": 1620
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3364044161126706e-05,
+      "loss": 0.4679,
+      "step": 1621
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3356426181694747e-05,
+      "loss": 0.498,
+      "step": 1622
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.3348805075677477e-05,
+      "loss": 0.4698,
+      "step": 1623
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.334118084592636e-05,
+      "loss": 0.4714,
+      "step": 1624
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3333553495294033e-05,
+      "loss": 0.4613,
+      "step": 1625
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.332592302663429e-05,
+      "loss": 0.4721,
+      "step": 1626
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.33182894428021e-05,
+      "loss": 0.4727,
+      "step": 1627
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3310652746653586e-05,
+      "loss": 0.4871,
+      "step": 1628
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.330301294104606e-05,
+      "loss": 0.4782,
+      "step": 1629
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3295370028837964e-05,
+      "loss": 0.492,
+      "step": 1630
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.328772401288893e-05,
+      "loss": 0.4393,
+      "step": 1631
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3280074896059736e-05,
+      "loss": 0.4629,
+      "step": 1632
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3272422681212327e-05,
+      "loss": 0.4883,
+      "step": 1633
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3264767371209805e-05,
+      "loss": 0.4817,
+      "step": 1634
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3257108968916435e-05,
+      "loss": 0.4647,
+      "step": 1635
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3249447477197627e-05,
+      "loss": 0.4443,
+      "step": 1636
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.324178289891996e-05,
+      "loss": 0.4335,
+      "step": 1637
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3234115236951154e-05,
+      "loss": 0.4876,
+      "step": 1638
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.32264444941601e-05,
+      "loss": 0.4844,
+      "step": 1639
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3218770673416824e-05,
+      "loss": 0.45,
+      "step": 1640
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.321109377759253e-05,
+      "loss": 0.4928,
+      "step": 1641
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.320341380955953e-05,
+      "loss": 0.4815,
+      "step": 1642
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.319573077219132e-05,
+      "loss": 0.4642,
+      "step": 1643
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.318804466836254e-05,
+      "loss": 0.5012,
+      "step": 1644
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3180355500948972e-05,
+      "loss": 0.4792,
+      "step": 1645
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3172663272827535e-05,
+      "loss": 0.512,
+      "step": 1646
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3164967986876307e-05,
+      "loss": 0.4482,
+      "step": 1647
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.315726964597451e-05,
+      "loss": 0.4591,
+      "step": 1648
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.31495682530025e-05,
+      "loss": 0.5224,
+      "step": 1649
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.314186381084178e-05,
+      "loss": 0.4622,
+      "step": 1650
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3134156322374993e-05,
+      "loss": 0.536,
+      "step": 1651
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.312644579048592e-05,
+      "loss": 0.468,
+      "step": 1652
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3118732218059488e-05,
+      "loss": 0.499,
+      "step": 1653
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.311101560798175e-05,
+      "loss": 0.4827,
+      "step": 1654
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.310329596313991e-05,
+      "loss": 0.469,
+      "step": 1655
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.309557328642229e-05,
+      "loss": 0.4844,
+      "step": 1656
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.3087847580718367e-05,
+      "loss": 0.5253,
+      "step": 1657
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.308011884891873e-05,
+      "loss": 0.4704,
+      "step": 1658
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3072387093915114e-05,
+      "loss": 0.5155,
+      "step": 1659
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3064652318600383e-05,
+      "loss": 0.4661,
+      "step": 1660
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3056914525868528e-05,
+      "loss": 0.4529,
+      "step": 1661
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3049173718614667e-05,
+      "loss": 0.4776,
+      "step": 1662
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.304142989973505e-05,
+      "loss": 0.5016,
+      "step": 1663
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3033683072127067e-05,
+      "loss": 0.4497,
+      "step": 1664
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.30259332386892e-05,
+      "loss": 0.5008,
+      "step": 1665
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.301818040232109e-05,
+      "loss": 0.5012,
+      "step": 1666
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.301042456592348e-05,
+      "loss": 0.5089,
+      "step": 1667
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.3002665732398246e-05,
+      "loss": 0.4737,
+      "step": 1668
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.2994903904648376e-05,
+      "loss": 0.4582,
+      "step": 1669
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.2987139085577987e-05,
+      "loss": 0.4933,
+      "step": 1670
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.2979371278092324e-05,
+      "loss": 0.4873,
+      "step": 1671
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.2971600485097723e-05,
+      "loss": 0.4824,
+      "step": 1672
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.2963826709501657e-05,
+      "loss": 0.5121,
+      "step": 1673
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.295604995421271e-05,
+      "loss": 0.4759,
+      "step": 1674
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.294827022214059e-05,
+      "loss": 0.4535,
+      "step": 1675
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.29404875161961e-05,
+      "loss": 0.5377,
+      "step": 1676
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.293270183929117e-05,
+      "loss": 0.4434,
+      "step": 1677
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2924913194338835e-05,
+      "loss": 0.462,
+      "step": 1678
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2917121584253246e-05,
+      "loss": 0.4775,
+      "step": 1679
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.290932701194966e-05,
+      "loss": 0.485,
+      "step": 1680
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2901529480344446e-05,
+      "loss": 0.4783,
+      "step": 1681
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2893728992355067e-05,
+      "loss": 0.4752,
+      "step": 1682
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2885925550900118e-05,
+      "loss": 0.478,
+      "step": 1683
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2878119158899267e-05,
+      "loss": 0.474,
+      "step": 1684
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.287030981927331e-05,
+      "loss": 0.5183,
+      "step": 1685
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.286249753494413e-05,
+      "loss": 0.4767,
+      "step": 1686
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2854682308834732e-05,
+      "loss": 0.4894,
+      "step": 1687
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.284686414386921e-05,
+      "loss": 0.4574,
+      "step": 1688
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2839043042972743e-05,
+      "loss": 0.4621,
+      "step": 1689
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2831219009071636e-05,
+      "loss": 0.4559,
+      "step": 1690
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.282339204509327e-05,
+      "loss": 0.4806,
+      "step": 1691
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2815562153966136e-05,
+      "loss": 0.4747,
+      "step": 1692
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.2807729338619806e-05,
+      "loss": 0.4919,
+      "step": 1693
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2799893601984963e-05,
+      "loss": 0.4381,
+      "step": 1694
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.279205494699337e-05,
+      "loss": 0.4816,
+      "step": 1695
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2784213376577885e-05,
+      "loss": 0.4869,
+      "step": 1696
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.277636889367247e-05,
+      "loss": 0.4917,
+      "step": 1697
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2768521501212148e-05,
+      "loss": 0.5204,
+      "step": 1698
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.276067120213306e-05,
+      "loss": 0.4778,
+      "step": 1699
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2752817999372412e-05,
+      "loss": 0.4844,
+      "step": 1700
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2744961895868517e-05,
+      "loss": 0.4695,
+      "step": 1701
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2737102894560757e-05,
+      "loss": 0.5202,
+      "step": 1702
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2729240998389605e-05,
+      "loss": 0.4794,
+      "step": 1703
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2721376210296617e-05,
+      "loss": 0.4682,
+      "step": 1704
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2713508533224432e-05,
+      "loss": 0.4815,
+      "step": 1705
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.270563797011676e-05,
+      "loss": 0.5304,
+      "step": 1706
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2697764523918412e-05,
+      "loss": 0.4887,
+      "step": 1707
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2689888197575254e-05,
+      "loss": 0.4757,
+      "step": 1708
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.2682008994034244e-05,
+      "loss": 0.4933,
+      "step": 1709
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.267412691624341e-05,
+      "loss": 0.4783,
+      "step": 1710
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2666241967151864e-05,
+      "loss": 0.4891,
+      "step": 1711
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2658354149709788e-05,
+      "loss": 0.5174,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2650463466868424e-05,
+      "loss": 0.4794,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2642569921580108e-05,
+      "loss": 0.5172,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.263467351679823e-05,
+      "loss": 0.4977,
+      "step": 1715
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.262677425547727e-05,
+      "loss": 0.4729,
+      "step": 1716
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.261887214057275e-05,
+      "loss": 0.4895,
+      "step": 1717
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.261096717504127e-05,
+      "loss": 0.4759,
+      "step": 1718
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.260305936184052e-05,
+      "loss": 0.4691,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2595148703929212e-05,
+      "loss": 0.4572,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.258723520426716e-05,
+      "loss": 0.4594,
+      "step": 1721
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2579318865815224e-05,
+      "loss": 0.4461,
+      "step": 1722
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.257139969153533e-05,
+      "loss": 0.4659,
+      "step": 1723
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2563477684390457e-05,
+      "loss": 0.4097,
+      "step": 1724
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.255555284734466e-05,
+      "loss": 0.4854,
+      "step": 1725
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.254762518336303e-05,
+      "loss": 0.419,
+      "step": 1726
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.253969469541175e-05,
+      "loss": 0.4379,
+      "step": 1727
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2531761386458023e-05,
+      "loss": 0.4417,
+      "step": 1728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.252382525947013e-05,
+      "loss": 0.4635,
+      "step": 1729
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.25158863174174e-05,
+      "loss": 0.438,
+      "step": 1730
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2507944563270214e-05,
+      "loss": 0.4348,
+      "step": 1731
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.25e-05,
+      "loss": 0.4245,
+      "step": 1732
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2492052630579258e-05,
+      "loss": 0.461,
+      "step": 1733
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2484102457981508e-05,
+      "loss": 0.4388,
+      "step": 1734
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2476149485181343e-05,
+      "loss": 0.4497,
+      "step": 1735
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2468193715154393e-05,
+      "loss": 0.4504,
+      "step": 1736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.246023515087734e-05,
+      "loss": 0.4544,
+      "step": 1737
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2452273795327903e-05,
+      "loss": 0.4175,
+      "step": 1738
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.244430965148485e-05,
+      "loss": 0.4389,
+      "step": 1739
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2436342722327995e-05,
+      "loss": 0.4635,
+      "step": 1740
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.242837301083819e-05,
+      "loss": 0.4361,
+      "step": 1741
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.242040051999733e-05,
+      "loss": 0.4659,
+      "step": 1742
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.241242525278835e-05,
+      "loss": 0.4334,
+      "step": 1743
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.2404447212195225e-05,
+      "loss": 0.443,
+      "step": 1744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2396466401202965e-05,
+      "loss": 0.4544,
+      "step": 1745
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2388482822797615e-05,
+      "loss": 0.4507,
+      "step": 1746
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.238049647996626e-05,
+      "loss": 0.4582,
+      "step": 1747
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2372507375697017e-05,
+      "loss": 0.4251,
+      "step": 1748
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.236451551297904e-05,
+      "loss": 0.4593,
+      "step": 1749
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2356520894802505e-05,
+      "loss": 0.4674,
+      "step": 1750
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2348523524158632e-05,
+      "loss": 0.4649,
+      "step": 1751
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2340523404039663e-05,
+      "loss": 0.4564,
+      "step": 1752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2332520537438866e-05,
+      "loss": 0.4793,
+      "step": 1753
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2324514927350547e-05,
+      "loss": 0.4185,
+      "step": 1754
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2316506576770025e-05,
+      "loss": 0.4422,
+      "step": 1755
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2308495488693658e-05,
+      "loss": 0.4413,
+      "step": 1756
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2300481666118818e-05,
+      "loss": 0.4489,
+      "step": 1757
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2292465112043907e-05,
+      "loss": 0.4566,
+      "step": 1758
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.228444582946834e-05,
+      "loss": 0.4381,
+      "step": 1759
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2276423821392574e-05,
+      "loss": 0.4405,
+      "step": 1760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2268399090818052e-05,
+      "loss": 0.4763,
+      "step": 1761
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2260371640747263e-05,
+      "loss": 0.4611,
+      "step": 1762
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2252341474183704e-05,
+      "loss": 0.4391,
+      "step": 1763
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2244308594131893e-05,
+      "loss": 0.4589,
+      "step": 1764
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.223627300359735e-05,
+      "loss": 0.4724,
+      "step": 1765
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2228234705586625e-05,
+      "loss": 0.4599,
+      "step": 1766
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2220193703107283e-05,
+      "loss": 0.4207,
+      "step": 1767
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2212149999167872e-05,
+      "loss": 0.4576,
+      "step": 1768
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.220410359677799e-05,
+      "loss": 0.4041,
+      "step": 1769
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2196054498948213e-05,
+      "loss": 0.4441,
+      "step": 1770
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2188002708690146e-05,
+      "loss": 0.4309,
+      "step": 1771
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2179948229016393e-05,
+      "loss": 0.4216,
+      "step": 1772
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2171891062940564e-05,
+      "loss": 0.4468,
+      "step": 1773
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2163831213477267e-05,
+      "loss": 0.4274,
+      "step": 1774
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2155768683642136e-05,
+      "loss": 0.4574,
+      "step": 1775
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2147703476451787e-05,
+      "loss": 0.4594,
+      "step": 1776
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2139635594923846e-05,
+      "loss": 0.4582,
+      "step": 1777
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2131565042076935e-05,
+      "loss": 0.4327,
+      "step": 1778
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.2123491820930686e-05,
+      "loss": 0.457,
+      "step": 1779
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2115415934505714e-05,
+      "loss": 0.4441,
+      "step": 1780
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2107337385823642e-05,
+      "loss": 0.4111,
+      "step": 1781
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2099256177907094e-05,
+      "loss": 0.461,
+      "step": 1782
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.209117231377967e-05,
+      "loss": 0.4496,
+      "step": 1783
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.208308579646598e-05,
+      "loss": 0.4342,
+      "step": 1784
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.207499662899162e-05,
+      "loss": 0.4577,
+      "step": 1785
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.206690481438319e-05,
+      "loss": 0.4359,
+      "step": 1786
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2058810355668257e-05,
+      "loss": 0.4497,
+      "step": 1787
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2050713255875397e-05,
+      "loss": 0.4201,
+      "step": 1788
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2042613518034163e-05,
+      "loss": 0.4669,
+      "step": 1789
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2034511145175105e-05,
+      "loss": 0.4416,
+      "step": 1790
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2026406140329747e-05,
+      "loss": 0.4864,
+      "step": 1791
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.201829850653061e-05,
+      "loss": 0.4757,
+      "step": 1792
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2010188246811193e-05,
+      "loss": 0.5288,
+      "step": 1793
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.2002075364205972e-05,
+      "loss": 0.4263,
+      "step": 1794
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.1993959861750413e-05,
+      "loss": 0.4354,
+      "step": 1795
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.1985841742480954e-05,
+      "loss": 0.4543,
+      "step": 1796
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1977721009435028e-05,
+      "loss": 0.4349,
+      "step": 1797
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1969597665651022e-05,
+      "loss": 0.4321,
+      "step": 1798
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1961471714168322e-05,
+      "loss": 0.4313,
+      "step": 1799
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1953343158027277e-05,
+      "loss": 0.4316,
+      "step": 1800
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 0.4807283580303192,
+      "eval_runtime": 1631.6739,
+      "eval_samples_per_second": 0.602,
+      "eval_steps_per_second": 0.022,
+      "step": 1800
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1945212000269218e-05,
+      "loss": 0.4844,
+      "step": 1801
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.193707824393644e-05,
+      "loss": 0.4597,
+      "step": 1802
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1928941892072215e-05,
+      "loss": 0.4294,
+      "step": 1803
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1920802947720788e-05,
+      "loss": 0.444,
+      "step": 1804
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1912661413927377e-05,
+      "loss": 0.473,
+      "step": 1805
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1904517293738158e-05,
+      "loss": 0.3968,
+      "step": 1806
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1896370590200283e-05,
+      "loss": 0.4251,
+      "step": 1807
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.188822130636188e-05,
+      "loss": 0.4454,
+      "step": 1808
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1880069445272016e-05,
+      "loss": 0.443,
+      "step": 1809
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.187191500998074e-05,
+      "loss": 0.4505,
+      "step": 1810
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.1863758003539067e-05,
+      "loss": 0.4713,
+      "step": 1811
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.185559842899897e-05,
+      "loss": 0.4512,
+      "step": 1812
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.184743628941338e-05,
+      "loss": 0.4358,
+      "step": 1813
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.183927158783618e-05,
+      "loss": 0.4438,
+      "step": 1814
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1831104327322234e-05,
+      "loss": 0.4273,
+      "step": 1815
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1822934510927347e-05,
+      "loss": 0.4701,
+      "step": 1816
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1814762141708276e-05,
+      "loss": 0.4404,
+      "step": 1817
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1806587222722753e-05,
+      "loss": 0.4246,
+      "step": 1818
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.179840975702944e-05,
+      "loss": 0.434,
+      "step": 1819
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1790229747687973e-05,
+      "loss": 0.4471,
+      "step": 1820
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.178204719775892e-05,
+      "loss": 0.4214,
+      "step": 1821
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.177386211030382e-05,
+      "loss": 0.4509,
+      "step": 1822
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1765674488385146e-05,
+      "loss": 0.4488,
+      "step": 1823
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1757484335066328e-05,
+      "loss": 0.5065,
+      "step": 1824
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1749291653411736e-05,
+      "loss": 0.4621,
+      "step": 1825
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1741096446486693e-05,
+      "loss": 0.4698,
+      "step": 1826
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1732898717357463e-05,
+      "loss": 0.4619,
+      "step": 1827
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.172469846909125e-05,
+      "loss": 0.4649,
+      "step": 1828
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.171649570475621e-05,
+      "loss": 0.4141,
+      "step": 1829
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.1708290427421434e-05,
+      "loss": 0.4513,
+      "step": 1830
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1700082640156954e-05,
+      "loss": 0.4283,
+      "step": 1831
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.169187234603374e-05,
+      "loss": 0.4334,
+      "step": 1832
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.16836595481237e-05,
+      "loss": 0.4295,
+      "step": 1833
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.167544424949969e-05,
+      "loss": 0.43,
+      "step": 1834
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.166722645323548e-05,
+      "loss": 0.4354,
+      "step": 1835
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.165900616240579e-05,
+      "loss": 0.4503,
+      "step": 1836
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.165078338008628e-05,
+      "loss": 0.4507,
+      "step": 1837
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1642558109353516e-05,
+      "loss": 0.4496,
+      "step": 1838
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.163433035328502e-05,
+      "loss": 0.4255,
+      "step": 1839
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1626100114959233e-05,
+      "loss": 0.4598,
+      "step": 1840
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.161786739745553e-05,
+      "loss": 0.4281,
+      "step": 1841
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1609632203854205e-05,
+      "loss": 0.4562,
+      "step": 1842
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1601394537236488e-05,
+      "loss": 0.4654,
+      "step": 1843
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1593154400684524e-05,
+      "loss": 0.4022,
+      "step": 1844
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1584911797281397e-05,
+      "loss": 0.4294,
+      "step": 1845
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1576666730111098e-05,
+      "loss": 0.4297,
+      "step": 1846
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.1568419202258546e-05,
+      "loss": 0.4516,
+      "step": 1847
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1560169216809588e-05,
+      "loss": 0.4381,
+      "step": 1848
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1551916776850982e-05,
+      "loss": 0.4568,
+      "step": 1849
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1543661885470406e-05,
+      "loss": 0.4481,
+      "step": 1850
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1535404545756455e-05,
+      "loss": 0.4665,
+      "step": 1851
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.152714476079864e-05,
+      "loss": 0.4706,
+      "step": 1852
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1518882533687392e-05,
+      "loss": 0.4178,
+      "step": 1853
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.151061786751404e-05,
+      "loss": 0.434,
+      "step": 1854
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1502350765370852e-05,
+      "loss": 0.4905,
+      "step": 1855
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1494081230350986e-05,
+      "loss": 0.4419,
+      "step": 1856
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1485809265548513e-05,
+      "loss": 0.4339,
+      "step": 1857
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.147753487405842e-05,
+      "loss": 0.4281,
+      "step": 1858
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1469258058976603e-05,
+      "loss": 0.4801,
+      "step": 1859
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1460978823399854e-05,
+      "loss": 0.4192,
+      "step": 1860
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.145269717042588e-05,
+      "loss": 0.4462,
+      "step": 1861
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1444413103153285e-05,
+      "loss": 0.4483,
+      "step": 1862
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1436126624681592e-05,
+      "loss": 0.4953,
+      "step": 1863
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.1427837738111205e-05,
+      "loss": 0.4267,
+      "step": 1864
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.141954644654344e-05,
+      "loss": 0.4397,
+      "step": 1865
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1411252753080515e-05,
+      "loss": 0.4239,
+      "step": 1866
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.140295666082555e-05,
+      "loss": 0.4723,
+      "step": 1867
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.139465817288254e-05,
+      "loss": 0.4282,
+      "step": 1868
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1386357292356408e-05,
+      "loss": 0.4149,
+      "step": 1869
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1378054022352944e-05,
+      "loss": 0.4319,
+      "step": 1870
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1369748365978858e-05,
+      "loss": 0.414,
+      "step": 1871
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1361440326341728e-05,
+      "loss": 0.4506,
+      "step": 1872
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1353129906550042e-05,
+      "loss": 0.434,
+      "step": 1873
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1344817109713168e-05,
+      "loss": 0.4457,
+      "step": 1874
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1336501938941376e-05,
+      "loss": 0.4334,
+      "step": 1875
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.13281843973458e-05,
+      "loss": 0.461,
+      "step": 1876
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.131986448803849e-05,
+      "loss": 0.4458,
+      "step": 1877
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1311542214132364e-05,
+      "loss": 0.4635,
+      "step": 1878
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1303217578741227e-05,
+      "loss": 0.4409,
+      "step": 1879
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1294890584979772e-05,
+      "loss": 0.4562,
+      "step": 1880
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1286561235963573e-05,
+      "loss": 0.4284,
+      "step": 1881
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.1278229534809083e-05,
+      "loss": 0.4541,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1269895484633632e-05,
+      "loss": 0.4207,
+      "step": 1883
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.126155908855544e-05,
+      "loss": 0.4357,
+      "step": 1884
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.125322034969359e-05,
+      "loss": 0.4438,
+      "step": 1885
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.124487927116806e-05,
+      "loss": 0.4299,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1236535856099688e-05,
+      "loss": 0.4395,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.122819010761019e-05,
+      "loss": 0.4798,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1219842028822164e-05,
+      "loss": 0.4855,
+      "step": 1889
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1211491622859054e-05,
+      "loss": 0.4194,
+      "step": 1890
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.120313889284521e-05,
+      "loss": 0.4659,
+      "step": 1891
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1194783841905826e-05,
+      "loss": 0.4498,
+      "step": 1892
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.118642647316698e-05,
+      "loss": 0.4188,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.11780667897556e-05,
+      "loss": 0.4196,
+      "step": 1894
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1169704794799498e-05,
+      "loss": 0.452,
+      "step": 1895
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1161340491427333e-05,
+      "loss": 0.4539,
+      "step": 1896
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1152973882768656e-05,
+      "loss": 0.4323,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1144604971953842e-05,
+      "loss": 0.4447,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.113623376211416e-05,
+      "loss": 0.4453,
+      "step": 1899
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.112786025638172e-05,
+      "loss": 0.4561,
+      "step": 1900
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1119484457889507e-05,
+      "loss": 0.483,
+      "step": 1901
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.111110636977134e-05,
+      "loss": 0.4346,
+      "step": 1902
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1102725995161922e-05,
+      "loss": 0.4263,
+      "step": 1903
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.10943433371968e-05,
+      "loss": 0.453,
+      "step": 1904
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.108595839901236e-05,
+      "loss": 0.4371,
+      "step": 1905
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1077571183745865e-05,
+      "loss": 0.4957,
+      "step": 1906
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1069181694535415e-05,
+      "loss": 0.4458,
+      "step": 1907
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1060789934519975e-05,
+      "loss": 0.4412,
+      "step": 1908
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1052395906839343e-05,
+      "loss": 0.4274,
+      "step": 1909
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1043999614634172e-05,
+      "loss": 0.4403,
+      "step": 1910
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.103560106104597e-05,
+      "loss": 0.4339,
+      "step": 1911
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1027200249217077e-05,
+      "loss": 0.4469,
+      "step": 1912
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1018797182290694e-05,
+      "loss": 0.4468,
+      "step": 1913
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1010391863410844e-05,
+      "loss": 0.4603,
+      "step": 1914
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.1001984295722424e-05,
+      "loss": 0.4459,
+      "step": 1915
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.099357448237114e-05,
+      "loss": 0.4471,
+      "step": 1916
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0985162426503555e-05,
+      "loss": 0.4353,
+      "step": 1917
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0976748131267067e-05,
+      "loss": 0.4618,
+      "step": 1918
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.096833159980992e-05,
+      "loss": 0.4372,
+      "step": 1919
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.095991283528118e-05,
+      "loss": 0.4435,
+      "step": 1920
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.095149184083076e-05,
+      "loss": 0.4241,
+      "step": 1921
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0943068619609407e-05,
+      "loss": 0.4283,
+      "step": 1922
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0934643174768696e-05,
+      "loss": 0.4438,
+      "step": 1923
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.092621550946103e-05,
+      "loss": 0.4461,
+      "step": 1924
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.091778562683966e-05,
+      "loss": 0.4254,
+      "step": 1925
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.090935353005864e-05,
+      "loss": 0.4305,
+      "step": 1926
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.090091922227288e-05,
+      "loss": 0.4453,
+      "step": 1927
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0892482706638098e-05,
+      "loss": 0.441,
+      "step": 1928
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0884043986310852e-05,
+      "loss": 0.4408,
+      "step": 1929
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.087560306444852e-05,
+      "loss": 0.4231,
+      "step": 1930
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0867159944209287e-05,
+      "loss": 0.4454,
+      "step": 1931
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0858714628752186e-05,
+      "loss": 0.4827,
+      "step": 1932
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.0850267121237058e-05,
+      "loss": 0.4224,
+      "step": 1933
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0841817424824568e-05,
+      "loss": 0.4762,
+      "step": 1934
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0833365542676198e-05,
+      "loss": 0.5049,
+      "step": 1935
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.082491147795425e-05,
+      "loss": 0.4378,
+      "step": 1936
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.081645523382184e-05,
+      "loss": 0.4613,
+      "step": 1937
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0807996813442897e-05,
+      "loss": 0.4637,
+      "step": 1938
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0799536219982173e-05,
+      "loss": 0.4486,
+      "step": 1939
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0791073456605224e-05,
+      "loss": 0.4589,
+      "step": 1940
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0782608526478426e-05,
+      "loss": 0.4653,
+      "step": 1941
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0774141432768954e-05,
+      "loss": 0.4298,
+      "step": 1942
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.076567217864481e-05,
+      "loss": 0.5152,
+      "step": 1943
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0757200767274792e-05,
+      "loss": 0.446,
+      "step": 1944
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0748727201828494e-05,
+      "loss": 0.5328,
+      "step": 1945
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.074025148547635e-05,
+      "loss": 0.4807,
+      "step": 1946
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.073177362138956e-05,
+      "loss": 0.4217,
+      "step": 1947
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0723293612740157e-05,
+      "loss": 0.4548,
+      "step": 1948
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0714811462700964e-05,
+      "loss": 0.4105,
+      "step": 1949
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0706327174445605e-05,
+      "loss": 0.4452,
+      "step": 1950
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.0697840751148495e-05,
+      "loss": 0.505,
+      "step": 1951
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0689352195984875e-05,
+      "loss": 0.4335,
+      "step": 1952
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.068086151213076e-05,
+      "loss": 0.4588,
+      "step": 1953
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.067236870276297e-05,
+      "loss": 0.4713,
+      "step": 1954
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.066387377105911e-05,
+      "loss": 0.4682,
+      "step": 1955
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.06553767201976e-05,
+      "loss": 0.444,
+      "step": 1956
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.064687755335763e-05,
+      "loss": 0.4632,
+      "step": 1957
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0638376273719195e-05,
+      "loss": 0.4293,
+      "step": 1958
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0629872884463083e-05,
+      "loss": 0.4902,
+      "step": 1959
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.062136738877086e-05,
+      "loss": 0.4435,
+      "step": 1960
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0612859789824893e-05,
+      "loss": 0.4254,
+      "step": 1961
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.060435009080832e-05,
+      "loss": 0.4635,
+      "step": 1962
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0595838294905082e-05,
+      "loss": 0.4404,
+      "step": 1963
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0587324405299893e-05,
+      "loss": 0.4424,
+      "step": 1964
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.057880842517825e-05,
+      "loss": 0.4405,
+      "step": 1965
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.057029035772644e-05,
+      "loss": 0.4245,
+      "step": 1966
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0561770206131527e-05,
+      "loss": 0.4919,
+      "step": 1967
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.0553247973581357e-05,
+      "loss": 0.4444,
+      "step": 1968
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0544723663264544e-05,
+      "loss": 0.4372,
+      "step": 1969
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0536197278370504e-05,
+      "loss": 0.4519,
+      "step": 1970
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.052766882208939e-05,
+      "loss": 0.4503,
+      "step": 1971
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0519138297612174e-05,
+      "loss": 0.439,
+      "step": 1972
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.051060570813057e-05,
+      "loss": 0.4427,
+      "step": 1973
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0502071056837075e-05,
+      "loss": 0.4289,
+      "step": 1974
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0493534346924962e-05,
+      "loss": 0.4463,
+      "step": 1975
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0484995581588266e-05,
+      "loss": 0.4322,
+      "step": 1976
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.04764547640218e-05,
+      "loss": 0.4399,
+      "step": 1977
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0467911897421143e-05,
+      "loss": 0.4387,
+      "step": 1978
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0459366984982623e-05,
+      "loss": 0.4613,
+      "step": 1979
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0450820029903362e-05,
+      "loss": 0.4676,
+      "step": 1980
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0442271035381227e-05,
+      "loss": 0.4341,
+      "step": 1981
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0433720004614854e-05,
+      "loss": 0.4168,
+      "step": 1982
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.0425166940803637e-05,
+      "loss": 0.4393,
+      "step": 1983
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.041661184714774e-05,
+      "loss": 0.4309,
+      "step": 1984
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.040805472684808e-05,
+      "loss": 0.4477,
+      "step": 1985
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0399495583106327e-05,
+      "loss": 0.4888,
+      "step": 1986
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0390934419124917e-05,
+      "loss": 0.4378,
+      "step": 1987
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.038237123810704e-05,
+      "loss": 0.4416,
+      "step": 1988
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.037380604325664e-05,
+      "loss": 0.444,
+      "step": 1989
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.036523883777841e-05,
+      "loss": 0.456,
+      "step": 1990
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0356669624877797e-05,
+      "loss": 0.4399,
+      "step": 1991
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.034809840776101e-05,
+      "loss": 0.4096,
+      "step": 1992
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0339525189634992e-05,
+      "loss": 0.4543,
+      "step": 1993
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0330949973707446e-05,
+      "loss": 0.4578,
+      "step": 1994
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0322372763186813e-05,
+      "loss": 0.4465,
+      "step": 1995
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0313793561282296e-05,
+      "loss": 0.4462,
+      "step": 1996
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.030521237120382e-05,
+      "loss": 0.4284,
+      "step": 1997
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0296629196162067e-05,
+      "loss": 0.4505,
+      "step": 1998
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.028804403936847e-05,
+      "loss": 0.4101,
+      "step": 1999
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0279456904035192e-05,
+      "loss": 0.4393,
+      "step": 2000
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 0.4775671362876892,
+      "eval_runtime": 1629.0485,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 2000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0270867793375137e-05,
+      "loss": 0.4443,
+      "step": 2001
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.0262276710601952e-05,
+      "loss": 0.4202,
+      "step": 2002
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0253683658930015e-05,
+      "loss": 0.436,
+      "step": 2003
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0245088641574458e-05,
+      "loss": 0.4623,
+      "step": 2004
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0236491661751118e-05,
+      "loss": 0.4619,
+      "step": 2005
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0227892722676593e-05,
+      "loss": 0.4308,
+      "step": 2006
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0219291827568208e-05,
+      "loss": 0.4558,
+      "step": 2007
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0210688979644015e-05,
+      "loss": 0.4252,
+      "step": 2008
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.02020841821228e-05,
+      "loss": 0.4944,
+      "step": 2009
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0193477438224072e-05,
+      "loss": 0.4347,
+      "step": 2010
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0184868751168082e-05,
+      "loss": 0.4555,
+      "step": 2011
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0176258124175793e-05,
+      "loss": 0.4559,
+      "step": 2012
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0167645560468904e-05,
+      "loss": 0.4312,
+      "step": 2013
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0159031063269827e-05,
+      "loss": 0.4409,
+      "step": 2014
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.015041463580172e-05,
+      "loss": 0.4448,
+      "step": 2015
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0141796281288434e-05,
+      "loss": 0.4349,
+      "step": 2016
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.0133176002954566e-05,
+      "loss": 0.4708,
+      "step": 2017
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.012455380402542e-05,
+      "loss": 0.4895,
+      "step": 2018
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.011592968772702e-05,
+      "loss": 0.4475,
+      "step": 2019
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0107303657286106e-05,
+      "loss": 0.4414,
+      "step": 2020
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.009867571593014e-05,
+      "loss": 0.425,
+      "step": 2021
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.009004586688729e-05,
+      "loss": 0.4451,
+      "step": 2022
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.008141411338645e-05,
+      "loss": 0.4358,
+      "step": 2023
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0072780458657222e-05,
+      "loss": 0.4463,
+      "step": 2024
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.006414490592991e-05,
+      "loss": 0.4367,
+      "step": 2025
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0055507458435546e-05,
+      "loss": 0.44,
+      "step": 2026
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0046868119405847e-05,
+      "loss": 0.4767,
+      "step": 2027
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0038226892073255e-05,
+      "loss": 0.4359,
+      "step": 2028
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.002958377967092e-05,
+      "loss": 0.4641,
+      "step": 2029
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.002093878543269e-05,
+      "loss": 0.4793,
+      "step": 2030
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.001229191259311e-05,
+      "loss": 0.4415,
+      "step": 2031
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.0003643164387455e-05,
+      "loss": 0.4319,
+      "step": 2032
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.999499254405167e-05,
+      "loss": 0.425,
+      "step": 2033
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.998634005482242e-05,
+      "loss": 0.4358,
+      "step": 2034
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.9977685699937056e-05,
+      "loss": 0.4505,
+      "step": 2035
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.996902948263364e-05,
+      "loss": 0.4582,
+      "step": 2036
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.996037140615093e-05,
+      "loss": 0.4653,
+      "step": 2037
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.995171147372836e-05,
+      "loss": 0.414,
+      "step": 2038
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9943049688606086e-05,
+      "loss": 0.4627,
+      "step": 2039
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9934386054024942e-05,
+      "loss": 0.4621,
+      "step": 2040
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.992572057322645e-05,
+      "loss": 0.4797,
+      "step": 2041
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.991705324945284e-05,
+      "loss": 0.4731,
+      "step": 2042
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9908384085947014e-05,
+      "loss": 0.4425,
+      "step": 2043
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.989971308595257e-05,
+      "loss": 0.4363,
+      "step": 2044
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.989104025271379e-05,
+      "loss": 0.4638,
+      "step": 2045
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9882365589475648e-05,
+      "loss": 0.4293,
+      "step": 2046
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.987368909948379e-05,
+      "loss": 0.5125,
+      "step": 2047
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.986501078598457e-05,
+      "loss": 0.4747,
+      "step": 2048
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9856330652225e-05,
+      "loss": 0.4484,
+      "step": 2049
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9847648701452775e-05,
+      "loss": 0.4571,
+      "step": 2050
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9838964936916285e-05,
+      "loss": 0.437,
+      "step": 2051
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9830279361864592e-05,
+      "loss": 0.457,
+      "step": 2052
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.9821591979547425e-05,
+      "loss": 0.4527,
+      "step": 2053
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.98129027932152e-05,
+      "loss": 0.4227,
+      "step": 2054
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9804211806119016e-05,
+      "loss": 0.4375,
+      "step": 2055
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.979551902151062e-05,
+      "loss": 0.4321,
+      "step": 2056
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9786824442642455e-05,
+      "loss": 0.4258,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9778128072767628e-05,
+      "loss": 0.4249,
+      "step": 2058
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9769429915139916e-05,
+      "loss": 0.4229,
+      "step": 2059
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9760729973013758e-05,
+      "loss": 0.4144,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9752028249644276e-05,
+      "loss": 0.4619,
+      "step": 2061
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9743324748287236e-05,
+      "loss": 0.4387,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9734619472199097e-05,
+      "loss": 0.4599,
+      "step": 2063
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.972591242463696e-05,
+      "loss": 0.4228,
+      "step": 2064
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.97172036088586e-05,
+      "loss": 0.4378,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.970849302812245e-05,
+      "loss": 0.4455,
+      "step": 2066
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9699780685687593e-05,
+      "loss": 0.43,
+      "step": 2067
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.96910665848138e-05,
+      "loss": 0.4749,
+      "step": 2068
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.968235072876147e-05,
+      "loss": 0.4303,
+      "step": 2069
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9673633120791673e-05,
+      "loss": 0.4279,
+      "step": 2070
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.9664913764166123e-05,
+      "loss": 0.4425,
+      "step": 2071
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.965619266214721e-05,
+      "loss": 0.416,
+      "step": 2072
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.964746981799796e-05,
+      "loss": 0.4365,
+      "step": 2073
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.963874523498205e-05,
+      "loss": 0.4349,
+      "step": 2074
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.963001891636382e-05,
+      "loss": 0.4304,
+      "step": 2075
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9621290865408243e-05,
+      "loss": 0.4577,
+      "step": 2076
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9612561085380958e-05,
+      "loss": 0.4481,
+      "step": 2077
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9603829579548243e-05,
+      "loss": 0.4529,
+      "step": 2078
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9595096351177004e-05,
+      "loss": 0.4337,
+      "step": 2079
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9586361403534823e-05,
+      "loss": 0.4153,
+      "step": 2080
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9577624739889913e-05,
+      "loss": 0.4395,
+      "step": 2081
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9568886363511118e-05,
+      "loss": 0.4408,
+      "step": 2082
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.956014627766793e-05,
+      "loss": 0.4318,
+      "step": 2083
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.955140448563049e-05,
+      "loss": 0.4444,
+      "step": 2084
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.954266099066956e-05,
+      "loss": 0.4506,
+      "step": 2085
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9533915796056557e-05,
+      "loss": 0.4246,
+      "step": 2086
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.9525168905063516e-05,
+      "loss": 0.4403,
+      "step": 2087
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.951642032096313e-05,
+      "loss": 0.4804,
+      "step": 2088
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.95076700470287e-05,
+      "loss": 0.4404,
+      "step": 2089
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.949891808653417e-05,
+      "loss": 0.4523,
+      "step": 2090
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9490164442754127e-05,
+      "loss": 0.4433,
+      "step": 2091
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9481409118963766e-05,
+      "loss": 0.4563,
+      "step": 2092
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9472652118438922e-05,
+      "loss": 0.4515,
+      "step": 2093
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.946389344445606e-05,
+      "loss": 0.429,
+      "step": 2094
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9455133100292265e-05,
+      "loss": 0.4174,
+      "step": 2095
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.944637108922526e-05,
+      "loss": 0.4543,
+      "step": 2096
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.943760741453336e-05,
+      "loss": 0.4448,
+      "step": 2097
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9428842079495544e-05,
+      "loss": 0.4801,
+      "step": 2098
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9420075087391383e-05,
+      "loss": 0.445,
+      "step": 2099
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9411306441501085e-05,
+      "loss": 0.4497,
+      "step": 2100
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9402536145105453e-05,
+      "loss": 0.4461,
+      "step": 2101
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.939376420148594e-05,
+      "loss": 0.4694,
+      "step": 2102
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9384990613924585e-05,
+      "loss": 0.4533,
+      "step": 2103
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.937621538570406e-05,
+      "loss": 0.4092,
+      "step": 2104
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.9367438520107653e-05,
+      "loss": 0.4802,
+      "step": 2105
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.935866002041925e-05,
+      "loss": 0.4333,
+      "step": 2106
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9349879889923364e-05,
+      "loss": 0.4807,
+      "step": 2107
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9341098131905105e-05,
+      "loss": 0.4387,
+      "step": 2108
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9332314749650192e-05,
+      "loss": 0.4516,
+      "step": 2109
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.932352974644497e-05,
+      "loss": 0.4409,
+      "step": 2110
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9314743125576372e-05,
+      "loss": 0.4357,
+      "step": 2111
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9305954890331945e-05,
+      "loss": 0.4816,
+      "step": 2112
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.929716504399983e-05,
+      "loss": 0.412,
+      "step": 2113
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9288373589868787e-05,
+      "loss": 0.4348,
+      "step": 2114
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9279580531228164e-05,
+      "loss": 0.4532,
+      "step": 2115
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.927078587136791e-05,
+      "loss": 0.4602,
+      "step": 2116
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9261989613578584e-05,
+      "loss": 0.4272,
+      "step": 2117
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.9253191761151337e-05,
+      "loss": 0.4072,
+      "step": 2118
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.92443923173779e-05,
+      "loss": 0.4262,
+      "step": 2119
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.923559128555063e-05,
+      "loss": 0.4609,
+      "step": 2120
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.922678866896246e-05,
+      "loss": 0.4547,
+      "step": 2121
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.921798447090692e-05,
+      "loss": 0.448,
+      "step": 2122
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.920917869467813e-05,
+      "loss": 0.4366,
+      "step": 2123
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9200371343570794e-05,
+      "loss": 0.4524,
+      "step": 2124
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.919156242088022e-05,
+      "loss": 0.4313,
+      "step": 2125
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.91827519299023e-05,
+      "loss": 0.4488,
+      "step": 2126
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9173939873933493e-05,
+      "loss": 0.46,
+      "step": 2127
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9165126256270877e-05,
+      "loss": 0.4386,
+      "step": 2128
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.915631108021209e-05,
+      "loss": 0.4462,
+      "step": 2129
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9147494349055365e-05,
+      "loss": 0.5029,
+      "step": 2130
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.91386760660995e-05,
+      "loss": 0.4448,
+      "step": 2131
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.91298562346439e-05,
+      "loss": 0.4212,
+      "step": 2132
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9121034857988524e-05,
+      "loss": 0.4754,
+      "step": 2133
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9112211939433928e-05,
+      "loss": 0.4103,
+      "step": 2134
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9103387482281226e-05,
+      "loss": 0.4271,
+      "step": 2135
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.909456148983213e-05,
+      "loss": 0.4656,
+      "step": 2136
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9085733965388915e-05,
+      "loss": 0.4331,
+      "step": 2137
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9076904912254424e-05,
+      "loss": 0.4532,
+      "step": 2138
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9068074333732076e-05,
+      "loss": 0.464,
+      "step": 2139
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.9059242233125866e-05,
+      "loss": 0.4598,
+      "step": 2140
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9050408613740356e-05,
+      "loss": 0.4556,
+      "step": 2141
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9041573478880668e-05,
+      "loss": 0.4463,
+      "step": 2142
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.90327368318525e-05,
+      "loss": 0.4675,
+      "step": 2143
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.9023898675962125e-05,
+      "loss": 0.4918,
+      "step": 2144
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.901505901451635e-05,
+      "loss": 0.4784,
+      "step": 2145
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.900621785082258e-05,
+      "loss": 0.4796,
+      "step": 2146
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8997375188188756e-05,
+      "loss": 0.4595,
+      "step": 2147
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.89885310299234e-05,
+      "loss": 0.4849,
+      "step": 2148
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8979685379335573e-05,
+      "loss": 0.4497,
+      "step": 2149
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8970838239734917e-05,
+      "loss": 0.4417,
+      "step": 2150
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8961989614431607e-05,
+      "loss": 0.4416,
+      "step": 2151
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8953139506736397e-05,
+      "loss": 0.4194,
+      "step": 2152
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.894428791996058e-05,
+      "loss": 0.4531,
+      "step": 2153
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8935434857416007e-05,
+      "loss": 0.479,
+      "step": 2154
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8926580322415087e-05,
+      "loss": 0.4442,
+      "step": 2155
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8917724318270762e-05,
+      "loss": 0.4395,
+      "step": 2156
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8908866848296553e-05,
+      "loss": 0.4695,
+      "step": 2157
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8900007915806498e-05,
+      "loss": 0.4438,
+      "step": 2158
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8891147524115204e-05,
+      "loss": 0.4388,
+      "step": 2159
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.888228567653781e-05,
+      "loss": 0.4667,
+      "step": 2160
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8873422376390018e-05,
+      "loss": 0.4212,
+      "step": 2161
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8864557626988056e-05,
+      "loss": 0.4392,
+      "step": 2162
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8855691431648696e-05,
+      "loss": 0.428,
+      "step": 2163
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8846823793689262e-05,
+      "loss": 0.4798,
+      "step": 2164
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.88379547164276e-05,
+      "loss": 0.4648,
+      "step": 2165
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.882908420318212e-05,
+      "loss": 0.4803,
+      "step": 2166
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8820212257271746e-05,
+      "loss": 0.4754,
+      "step": 2167
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8811338882015948e-05,
+      "loss": 0.438,
+      "step": 2168
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.880246408073473e-05,
+      "loss": 0.4146,
+      "step": 2169
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8793587856748636e-05,
+      "loss": 0.4959,
+      "step": 2170
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.878471021337872e-05,
+      "loss": 0.4542,
+      "step": 2171
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8775831153946597e-05,
+      "loss": 0.4457,
+      "step": 2172
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.8766950681774388e-05,
+      "loss": 0.4724,
+      "step": 2173
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.875806880018476e-05,
+      "loss": 0.4414,
+      "step": 2174
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.874918551250089e-05,
+      "loss": 0.4388,
+      "step": 2175
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8740300822046496e-05,
+      "loss": 0.4341,
+      "step": 2176
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8731414732145816e-05,
+      "loss": 0.4333,
+      "step": 2177
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8722527246123607e-05,
+      "loss": 0.3993,
+      "step": 2178
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8713638367305155e-05,
+      "loss": 0.4333,
+      "step": 2179
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8704748099016263e-05,
+      "loss": 0.4897,
+      "step": 2180
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.869585644458326e-05,
+      "loss": 0.4282,
+      "step": 2181
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.868696340733298e-05,
+      "loss": 0.432,
+      "step": 2182
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8678068990592783e-05,
+      "loss": 0.4141,
+      "step": 2183
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8669173197690558e-05,
+      "loss": 0.4778,
+      "step": 2184
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8660276031954683e-05,
+      "loss": 0.4382,
+      "step": 2185
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8651377496714072e-05,
+      "loss": 0.4302,
+      "step": 2186
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8642477595298134e-05,
+      "loss": 0.439,
+      "step": 2187
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8633576331036808e-05,
+      "loss": 0.4376,
+      "step": 2188
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.862467370726052e-05,
+      "loss": 0.4566,
+      "step": 2189
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8615769727300226e-05,
+      "loss": 0.4715,
+      "step": 2190
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8606864394487375e-05,
+      "loss": 0.4653,
+      "step": 2191
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.859795771215393e-05,
+      "loss": 0.4439,
+      "step": 2192
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8589049683632355e-05,
+      "loss": 0.4617,
+      "step": 2193
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8580140312255628e-05,
+      "loss": 0.425,
+      "step": 2194
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8571229601357206e-05,
+      "loss": 0.4337,
+      "step": 2195
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8562317554271073e-05,
+      "loss": 0.4224,
+      "step": 2196
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.85534041743317e-05,
+      "loss": 0.4266,
+      "step": 2197
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8544489464874048e-05,
+      "loss": 0.4185,
+      "step": 2198
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8535573429233602e-05,
+      "loss": 0.4511,
+      "step": 2199
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8526656070746316e-05,
+      "loss": 0.4144,
+      "step": 2200
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 0.4745250642299652,
+      "eval_runtime": 1630.0077,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 2200
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.851773739274865e-05,
+      "loss": 0.454,
+      "step": 2201
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8508817398577565e-05,
+      "loss": 0.4466,
+      "step": 2202
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8499896091570507e-05,
+      "loss": 0.4521,
+      "step": 2203
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8490973475065407e-05,
+      "loss": 0.4238,
+      "step": 2204
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8482049552400693e-05,
+      "loss": 0.4332,
+      "step": 2205
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8473124326915283e-05,
+      "loss": 0.443,
+      "step": 2206
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8464197801948582e-05,
+      "loss": 0.4416,
+      "step": 2207
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8455269980840478e-05,
+      "loss": 0.4394,
+      "step": 2208
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8446340866931346e-05,
+      "loss": 0.417,
+      "step": 2209
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8437410463562045e-05,
+      "loss": 0.415,
+      "step": 2210
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8428478774073915e-05,
+      "loss": 0.4698,
+      "step": 2211
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8419545801808776e-05,
+      "loss": 0.4584,
+      "step": 2212
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8410611550108935e-05,
+      "loss": 0.4359,
+      "step": 2213
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.840167602231717e-05,
+      "loss": 0.4554,
+      "step": 2214
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8392739221776733e-05,
+      "loss": 0.4335,
+      "step": 2215
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.838380115183137e-05,
+      "loss": 0.4322,
+      "step": 2216
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8374861815825285e-05,
+      "loss": 0.4487,
+      "step": 2217
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8365921217103164e-05,
+      "loss": 0.4238,
+      "step": 2218
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8356979359010154e-05,
+      "loss": 0.4338,
+      "step": 2219
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8348036244891885e-05,
+      "loss": 0.4515,
+      "step": 2220
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.833909187809446e-05,
+      "loss": 0.461,
+      "step": 2221
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8330146261964436e-05,
+      "loss": 0.4842,
+      "step": 2222
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.832119939984885e-05,
+      "loss": 0.4177,
+      "step": 2223
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8312251295095203e-05,
+      "loss": 0.4291,
+      "step": 2224
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8303301951051457e-05,
+      "loss": 0.479,
+      "step": 2225
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.8294351371066033e-05,
+      "loss": 0.4586,
+      "step": 2226
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.828539955848783e-05,
+      "loss": 0.4427,
+      "step": 2227
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8276446516666194e-05,
+      "loss": 0.4558,
+      "step": 2228
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8267492248950944e-05,
+      "loss": 0.4375,
+      "step": 2229
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8258536758692337e-05,
+      "loss": 0.4417,
+      "step": 2230
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.824958004924111e-05,
+      "loss": 0.44,
+      "step": 2231
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8240622123948445e-05,
+      "loss": 0.4364,
+      "step": 2232
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8231662986165976e-05,
+      "loss": 0.442,
+      "step": 2233
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8222702639245794e-05,
+      "loss": 0.4353,
+      "step": 2234
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8213741086540452e-05,
+      "loss": 0.4659,
+      "step": 2235
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8204778331402944e-05,
+      "loss": 0.4358,
+      "step": 2236
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.81958143771867e-05,
+      "loss": 0.4727,
+      "step": 2237
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8186849227245635e-05,
+      "loss": 0.4438,
+      "step": 2238
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8177882884934075e-05,
+      "loss": 0.4307,
+      "step": 2239
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8168915353606815e-05,
+      "loss": 0.4205,
+      "step": 2240
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8159946636619085e-05,
+      "loss": 0.4484,
+      "step": 2241
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8150976737326565e-05,
+      "loss": 0.43,
+      "step": 2242
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.814200565908537e-05,
+      "loss": 0.4424,
+      "step": 2243
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8133033405252056e-05,
+      "loss": 0.4187,
+      "step": 2244
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.812405997918363e-05,
+      "loss": 0.4026,
+      "step": 2245
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.811508538423752e-05,
+      "loss": 0.4783,
+      "step": 2246
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.810610962377161e-05,
+      "loss": 0.4512,
+      "step": 2247
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.809713270114421e-05,
+      "loss": 0.4316,
+      "step": 2248
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8088154619714063e-05,
+      "loss": 0.4154,
+      "step": 2249
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8079175382840352e-05,
+      "loss": 0.4285,
+      "step": 2250
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.807019499388269e-05,
+      "loss": 0.4316,
+      "step": 2251
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.806121345620111e-05,
+      "loss": 0.4238,
+      "step": 2252
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8052230773156097e-05,
+      "loss": 0.4358,
+      "step": 2253
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8043246948108548e-05,
+      "loss": 0.5003,
+      "step": 2254
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.803426198441979e-05,
+      "loss": 0.4419,
+      "step": 2255
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8025275885451577e-05,
+      "loss": 0.417,
+      "step": 2256
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8016288654566095e-05,
+      "loss": 0.4501,
+      "step": 2257
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8007300295125943e-05,
+      "loss": 0.4302,
+      "step": 2258
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.7998310810494146e-05,
+      "loss": 0.4777,
+      "step": 2259
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.798932020403415e-05,
+      "loss": 0.4419,
+      "step": 2260
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.798032847910982e-05,
+      "loss": 0.443,
+      "step": 2261
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7971335639085445e-05,
+      "loss": 0.452,
+      "step": 2262
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7962341687325713e-05,
+      "loss": 0.4626,
+      "step": 2263
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.795334662719576e-05,
+      "loss": 0.4628,
+      "step": 2264
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7944350462061118e-05,
+      "loss": 0.4182,
+      "step": 2265
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7935353195287718e-05,
+      "loss": 0.4604,
+      "step": 2266
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7926354830241928e-05,
+      "loss": 0.5283,
+      "step": 2267
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.791735537029051e-05,
+      "loss": 0.4319,
+      "step": 2268
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7908354818800648e-05,
+      "loss": 0.4012,
+      "step": 2269
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7899353179139935e-05,
+      "loss": 0.4391,
+      "step": 2270
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.789035045467635e-05,
+      "loss": 0.479,
+      "step": 2271
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7881346648778303e-05,
+      "loss": 0.4508,
+      "step": 2272
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.78723417648146e-05,
+      "loss": 0.437,
+      "step": 2273
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.786333580615444e-05,
+      "loss": 0.4263,
+      "step": 2274
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7854328776167443e-05,
+      "loss": 0.4475,
+      "step": 2275
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.7845320678223616e-05,
+      "loss": 0.4116,
+      "step": 2276
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.783631151569337e-05,
+      "loss": 0.4531,
+      "step": 2277
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7827301291947514e-05,
+      "loss": 0.4576,
+      "step": 2278
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.781829001035725e-05,
+      "loss": 0.4273,
+      "step": 2279
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.780927767429418e-05,
+      "loss": 0.4709,
+      "step": 2280
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.780026428713031e-05,
+      "loss": 0.4248,
+      "step": 2281
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.779124985223802e-05,
+      "loss": 0.4314,
+      "step": 2282
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.778223437299009e-05,
+      "loss": 0.4294,
+      "step": 2283
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7773217852759704e-05,
+      "loss": 0.4158,
+      "step": 2284
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.776420029492041e-05,
+      "loss": 0.457,
+      "step": 2285
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7755181702846164e-05,
+      "loss": 0.4365,
+      "step": 2286
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7746162079911303e-05,
+      "loss": 0.4334,
+      "step": 2287
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7737141429490552e-05,
+      "loss": 0.4383,
+      "step": 2288
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7728119754959015e-05,
+      "loss": 0.4277,
+      "step": 2289
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.771909705969218e-05,
+      "loss": 0.5118,
+      "step": 2290
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.771007334706593e-05,
+      "loss": 0.434,
+      "step": 2291
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.77010486204565e-05,
+      "loss": 0.4458,
+      "step": 2292
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7692022883240536e-05,
+      "loss": 0.4313,
+      "step": 2293
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.7682996138795045e-05,
+      "loss": 0.4799,
+      "step": 2294
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.767396839049741e-05,
+      "loss": 0.4527,
+      "step": 2295
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7664939641725403e-05,
+      "loss": 0.4741,
+      "step": 2296
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7655909895857152e-05,
+      "loss": 0.465,
+      "step": 2297
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.764687915627117e-05,
+      "loss": 0.4264,
+      "step": 2298
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.763784742634634e-05,
+      "loss": 0.4748,
+      "step": 2299
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7628814709461913e-05,
+      "loss": 0.4466,
+      "step": 2300
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7619781008997514e-05,
+      "loss": 0.429,
+      "step": 2301
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.761074632833313e-05,
+      "loss": 0.4402,
+      "step": 2302
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.760171067084912e-05,
+      "loss": 0.4177,
+      "step": 2303
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7592674039926196e-05,
+      "loss": 0.4307,
+      "step": 2304
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7583636438945457e-05,
+      "loss": 0.4436,
+      "step": 2305
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7574597871288354e-05,
+      "loss": 0.4173,
+      "step": 2306
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.756555834033668e-05,
+      "loss": 0.4264,
+      "step": 2307
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.7556517849472626e-05,
+      "loss": 0.5214,
+      "step": 2308
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.754747640207871e-05,
+      "loss": 0.4604,
+      "step": 2309
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.753843400153783e-05,
+      "loss": 0.4752,
+      "step": 2310
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.752939065123322e-05,
+      "loss": 0.4365,
+      "step": 2311
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.752034635454848e-05,
+      "loss": 0.4387,
+      "step": 2312
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.751130111486758e-05,
+      "loss": 0.4414,
+      "step": 2313
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7502254935574817e-05,
+      "loss": 0.469,
+      "step": 2314
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7493207820054842e-05,
+      "loss": 0.4459,
+      "step": 2315
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.748415977169268e-05,
+      "loss": 0.416,
+      "step": 2316
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.747511079387368e-05,
+      "loss": 0.4568,
+      "step": 2317
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7466060889983548e-05,
+      "loss": 0.4309,
+      "step": 2318
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7457010063408334e-05,
+      "loss": 0.4433,
+      "step": 2319
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.744795831753444e-05,
+      "loss": 0.4341,
+      "step": 2320
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7438905655748612e-05,
+      "loss": 0.4563,
+      "step": 2321
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.742985208143792e-05,
+      "loss": 0.4463,
+      "step": 2322
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7420797597989804e-05,
+      "loss": 0.453,
+      "step": 2323
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7411742208792025e-05,
+      "loss": 0.5068,
+      "step": 2324
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7402685917232687e-05,
+      "loss": 0.4747,
+      "step": 2325
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.739362872670023e-05,
+      "loss": 0.4269,
+      "step": 2326
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7384570640583436e-05,
+      "loss": 0.4247,
+      "step": 2327
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7375511662271416e-05,
+      "loss": 0.4384,
+      "step": 2328
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.7366451795153622e-05,
+      "loss": 0.4665,
+      "step": 2329
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7357391042619828e-05,
+      "loss": 0.4298,
+      "step": 2330
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.734832940806015e-05,
+      "loss": 0.429,
+      "step": 2331
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7339266894865035e-05,
+      "loss": 0.4239,
+      "step": 2332
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7330203506425238e-05,
+      "loss": 0.4433,
+      "step": 2333
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.732113924613187e-05,
+      "loss": 0.4351,
+      "step": 2334
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7312074117376346e-05,
+      "loss": 0.4333,
+      "step": 2335
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.730300812355042e-05,
+      "loss": 0.4364,
+      "step": 2336
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7293941268046164e-05,
+      "loss": 0.4148,
+      "step": 2337
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.728487355425597e-05,
+      "loss": 0.4284,
+      "step": 2338
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.727580498557256e-05,
+      "loss": 0.4371,
+      "step": 2339
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7266735565388964e-05,
+      "loss": 0.4476,
+      "step": 2340
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7257665297098538e-05,
+      "loss": 0.4439,
+      "step": 2341
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7248594184094954e-05,
+      "loss": 0.4277,
+      "step": 2342
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7239522229772195e-05,
+      "loss": 0.4565,
+      "step": 2343
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7230449437524573e-05,
+      "loss": 0.4624,
+      "step": 2344
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7221375810746702e-05,
+      "loss": 0.497,
+      "step": 2345
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7212301352833503e-05,
+      "loss": 0.4811,
+      "step": 2346
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7203226067180225e-05,
+      "loss": 0.5048,
+      "step": 2347
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7194149957182412e-05,
+      "loss": 0.4452,
+      "step": 2348
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7185073026235923e-05,
+      "loss": 0.4156,
+      "step": 2349
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7175995277736912e-05,
+      "loss": 0.4573,
+      "step": 2350
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7166916715081874e-05,
+      "loss": 0.4255,
+      "step": 2351
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7157837341667553e-05,
+      "loss": 0.4332,
+      "step": 2352
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7148757160891052e-05,
+      "loss": 0.4282,
+      "step": 2353
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7139676176149743e-05,
+      "loss": 0.4272,
+      "step": 2354
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7130594390841305e-05,
+      "loss": 0.4517,
+      "step": 2355
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7121511808363724e-05,
+      "loss": 0.4168,
+      "step": 2356
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7112428432115277e-05,
+      "loss": 0.4611,
+      "step": 2357
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.710334426549454e-05,
+      "loss": 0.4696,
+      "step": 2358
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.709425931190038e-05,
+      "loss": 0.4678,
+      "step": 2359
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7085173574731966e-05,
+      "loss": 0.4058,
+      "step": 2360
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7076087057388768e-05,
+      "loss": 0.43,
+      "step": 2361
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.706699976327052e-05,
+      "loss": 0.4257,
+      "step": 2362
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.7057911695777276e-05,
+      "loss": 0.4483,
+      "step": 2363
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7048822858309367e-05,
+      "loss": 0.4411,
+      "step": 2364
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.703973325426741e-05,
+      "loss": 0.4185,
+      "step": 2365
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7030642887052307e-05,
+      "loss": 0.4591,
+      "step": 2366
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7021551760065255e-05,
+      "loss": 0.4229,
+      "step": 2367
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7012459876707728e-05,
+      "loss": 0.4196,
+      "step": 2368
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.7003367240381486e-05,
+      "loss": 0.4167,
+      "step": 2369
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.699427385448857e-05,
+      "loss": 0.4216,
+      "step": 2370
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.69851797224313e-05,
+      "loss": 0.4443,
+      "step": 2371
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6976084847612282e-05,
+      "loss": 0.4525,
+      "step": 2372
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.696698923343439e-05,
+      "loss": 0.3858,
+      "step": 2373
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6957892883300778e-05,
+      "loss": 0.4547,
+      "step": 2374
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6948795800614873e-05,
+      "loss": 0.4711,
+      "step": 2375
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6939697988780386e-05,
+      "loss": 0.4724,
+      "step": 2376
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6930599451201292e-05,
+      "loss": 0.4293,
+      "step": 2377
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6921500191281843e-05,
+      "loss": 0.4308,
+      "step": 2378
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6912400212426553e-05,
+      "loss": 0.4149,
+      "step": 2379
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.6903299518040213e-05,
+      "loss": 0.4697,
+      "step": 2380
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6894198111527876e-05,
+      "loss": 0.4625,
+      "step": 2381
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6885095996294866e-05,
+      "loss": 0.4373,
+      "step": 2382
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.687599317574677e-05,
+      "loss": 0.4399,
+      "step": 2383
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6866889653289444e-05,
+      "loss": 0.4402,
+      "step": 2384
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6857785432328993e-05,
+      "loss": 0.4269,
+      "step": 2385
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6848680516271798e-05,
+      "loss": 0.43,
+      "step": 2386
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6839574908524496e-05,
+      "loss": 0.471,
+      "step": 2387
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6830468612493976e-05,
+      "loss": 0.4431,
+      "step": 2388
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.682136163158739e-05,
+      "loss": 0.436,
+      "step": 2389
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6812253969212154e-05,
+      "loss": 0.4615,
+      "step": 2390
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6803145628775923e-05,
+      "loss": 0.4322,
+      "step": 2391
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6794036613686613e-05,
+      "loss": 0.4161,
+      "step": 2392
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6784926927352396e-05,
+      "loss": 0.4121,
+      "step": 2393
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.677581657318169e-05,
+      "loss": 0.4272,
+      "step": 2394
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6766705554583174e-05,
+      "loss": 0.4431,
+      "step": 2395
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6757593874965754e-05,
+      "loss": 0.4402,
+      "step": 2396
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.67484815377386e-05,
+      "loss": 0.4526,
+      "step": 2397
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.6739368546311128e-05,
+      "loss": 0.4467,
+      "step": 2398
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6730254904092995e-05,
+      "loss": 0.4337,
+      "step": 2399
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6721140614494093e-05,
+      "loss": 0.4342,
+      "step": 2400
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 0.47000622749328613,
+      "eval_runtime": 1631.6533,
+      "eval_samples_per_second": 0.602,
+      "eval_steps_per_second": 0.022,
+      "step": 2400
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6712025680924572e-05,
+      "loss": 0.3977,
+      "step": 2401
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.670291010679481e-05,
+      "loss": 0.4126,
+      "step": 2402
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.669379389551544e-05,
+      "loss": 0.4358,
+      "step": 2403
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6684677050497316e-05,
+      "loss": 0.4836,
+      "step": 2404
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.667555957515153e-05,
+      "loss": 0.4342,
+      "step": 2405
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6666441472889434e-05,
+      "loss": 0.4022,
+      "step": 2406
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.665732274712258e-05,
+      "loss": 0.4389,
+      "step": 2407
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6648203401262776e-05,
+      "loss": 0.4826,
+      "step": 2408
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.663908343872206e-05,
+      "loss": 0.4487,
+      "step": 2409
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6629962862912695e-05,
+      "loss": 0.4401,
+      "step": 2410
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6620841677247176e-05,
+      "loss": 0.429,
+      "step": 2411
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6611719885138223e-05,
+      "loss": 0.4252,
+      "step": 2412
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.660259748999879e-05,
+      "loss": 0.4233,
+      "step": 2413
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.659347449524205e-05,
+      "loss": 0.4404,
+      "step": 2414
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.6584350904281397e-05,
+      "loss": 0.422,
+      "step": 2415
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6575226720530458e-05,
+      "loss": 0.4537,
+      "step": 2416
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6566101947403085e-05,
+      "loss": 0.4102,
+      "step": 2417
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6556976588313332e-05,
+      "loss": 0.4684,
+      "step": 2418
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6547850646675487e-05,
+      "loss": 0.4889,
+      "step": 2419
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6538724125904052e-05,
+      "loss": 0.4437,
+      "step": 2420
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6529597029413748e-05,
+      "loss": 0.4489,
+      "step": 2421
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.652046936061951e-05,
+      "loss": 0.4156,
+      "step": 2422
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.651134112293648e-05,
+      "loss": 0.3894,
+      "step": 2423
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6502212319780022e-05,
+      "loss": 0.452,
+      "step": 2424
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6493082954565708e-05,
+      "loss": 0.4308,
+      "step": 2425
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6483953030709327e-05,
+      "loss": 0.4423,
+      "step": 2426
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6474822551626867e-05,
+      "loss": 0.471,
+      "step": 2427
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6465691520734527e-05,
+      "loss": 0.4336,
+      "step": 2428
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6456559941448714e-05,
+      "loss": 0.4317,
+      "step": 2429
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.644742781718604e-05,
+      "loss": 0.4778,
+      "step": 2430
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6438295151363315e-05,
+      "loss": 0.4246,
+      "step": 2431
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.6429161947397565e-05,
+      "loss": 0.4406,
+      "step": 2432
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.642002820870601e-05,
+      "loss": 0.4429,
+      "step": 2433
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.641089393870606e-05,
+      "loss": 0.4246,
+      "step": 2434
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6401759140815345e-05,
+      "loss": 0.515,
+      "step": 2435
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6392623818451673e-05,
+      "loss": 0.4313,
+      "step": 2436
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6383487975033054e-05,
+      "loss": 0.4492,
+      "step": 2437
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.63743516139777e-05,
+      "loss": 0.4558,
+      "step": 2438
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6365214738704015e-05,
+      "loss": 0.4203,
+      "step": 2439
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6356077352630578e-05,
+      "loss": 0.4638,
+      "step": 2440
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.634693945917619e-05,
+      "loss": 0.438,
+      "step": 2441
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6337801061759812e-05,
+      "loss": 0.4489,
+      "step": 2442
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6328662163800615e-05,
+      "loss": 0.4199,
+      "step": 2443
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6319522768717947e-05,
+      "loss": 0.43,
+      "step": 2444
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.631038287993134e-05,
+      "loss": 0.4396,
+      "step": 2445
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6301242500860513e-05,
+      "loss": 0.4602,
+      "step": 2446
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.629210163492538e-05,
+      "loss": 0.4173,
+      "step": 2447
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6282960285546018e-05,
+      "loss": 0.4166,
+      "step": 2448
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.6273818456142695e-05,
+      "loss": 0.5097,
+      "step": 2449
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6264676150135866e-05,
+      "loss": 0.4344,
+      "step": 2450
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6255533370946144e-05,
+      "loss": 0.4814,
+      "step": 2451
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6246390121994337e-05,
+      "loss": 0.4634,
+      "step": 2452
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.623724640670142e-05,
+      "loss": 0.4225,
+      "step": 2453
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6228102228488554e-05,
+      "loss": 0.4268,
+      "step": 2454
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.621895759077705e-05,
+      "loss": 0.4595,
+      "step": 2455
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6209812496988418e-05,
+      "loss": 0.5869,
+      "step": 2456
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.620066695054432e-05,
+      "loss": 0.4175,
+      "step": 2457
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6191520954866595e-05,
+      "loss": 0.4417,
+      "step": 2458
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.618237451337725e-05,
+      "loss": 0.4348,
+      "step": 2459
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6173227629498447e-05,
+      "loss": 0.4548,
+      "step": 2460
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.616408030665254e-05,
+      "loss": 0.4359,
+      "step": 2461
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.615493254826202e-05,
+      "loss": 0.4083,
+      "step": 2462
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6145784357749553e-05,
+      "loss": 0.4823,
+      "step": 2463
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6136635738537964e-05,
+      "loss": 0.485,
+      "step": 2464
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.612748669405025e-05,
+      "loss": 0.4446,
+      "step": 2465
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.6118337227709547e-05,
+      "loss": 0.4221,
+      "step": 2466
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6109187342939164e-05,
+      "loss": 0.4431,
+      "step": 2467
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.610003704316256e-05,
+      "loss": 0.3977,
+      "step": 2468
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.609088633180335e-05,
+      "loss": 0.4142,
+      "step": 2469
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6081735212285306e-05,
+      "loss": 0.4346,
+      "step": 2470
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.607258368803235e-05,
+      "loss": 0.4394,
+      "step": 2471
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.606343176246855e-05,
+      "loss": 0.4624,
+      "step": 2472
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6054279439018138e-05,
+      "loss": 0.4216,
+      "step": 2473
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.604512672110549e-05,
+      "loss": 0.5044,
+      "step": 2474
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6035973612155112e-05,
+      "loss": 0.4351,
+      "step": 2475
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6026820115591685e-05,
+      "loss": 0.4252,
+      "step": 2476
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6017666234840015e-05,
+      "loss": 0.4578,
+      "step": 2477
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.6008511973325056e-05,
+      "loss": 0.4173,
+      "step": 2478
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.599935733447191e-05,
+      "loss": 0.4365,
+      "step": 2479
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.5990202321705814e-05,
+      "loss": 0.4597,
+      "step": 2480
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.5981046938452147e-05,
+      "loss": 0.4201,
+      "step": 2481
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.5971891188136427e-05,
+      "loss": 0.4583,
+      "step": 2482
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.5962735074184316e-05,
+      "loss": 0.4672,
+      "step": 2483
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5953578600021595e-05,
+      "loss": 0.4134,
+      "step": 2484
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.594442176907419e-05,
+      "loss": 0.4624,
+      "step": 2485
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.593526458476817e-05,
+      "loss": 0.4436,
+      "step": 2486
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.592610705052972e-05,
+      "loss": 0.4351,
+      "step": 2487
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.591694916978516e-05,
+      "loss": 0.4237,
+      "step": 2488
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5907790945960947e-05,
+      "loss": 0.4739,
+      "step": 2489
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5898632382483665e-05,
+      "loss": 0.4338,
+      "step": 2490
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5889473482780015e-05,
+      "loss": 0.4308,
+      "step": 2491
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5880314250276832e-05,
+      "loss": 0.4444,
+      "step": 2492
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5871154688401077e-05,
+      "loss": 0.4298,
+      "step": 2493
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.586199480057983e-05,
+      "loss": 0.4258,
+      "step": 2494
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.585283459024029e-05,
+      "loss": 0.4345,
+      "step": 2495
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5843674060809784e-05,
+      "loss": 0.4611,
+      "step": 2496
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.5834513215715755e-05,
+      "loss": 0.4486,
+      "step": 2497
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.582535205838577e-05,
+      "loss": 0.4317,
+      "step": 2498
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.58161905922475e-05,
+      "loss": 0.458,
+      "step": 2499
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.580702882072874e-05,
+      "loss": 0.4581,
+      "step": 2500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.57978667472574e-05,
+      "loss": 0.4873,
+      "step": 2501
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5788704375261502e-05,
+      "loss": 0.425,
+      "step": 2502
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.577954170816917e-05,
+      "loss": 0.4485,
+      "step": 2503
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5770378749408653e-05,
+      "loss": 0.4389,
+      "step": 2504
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.576121550240831e-05,
+      "loss": 0.4315,
+      "step": 2505
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5752051970596593e-05,
+      "loss": 0.4209,
+      "step": 2506
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.574288815740207e-05,
+      "loss": 0.4394,
+      "step": 2507
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5733724066253413e-05,
+      "loss": 0.4069,
+      "step": 2508
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5724559700579405e-05,
+      "loss": 0.4427,
+      "step": 2509
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.571539506380891e-05,
+      "loss": 0.42,
+      "step": 2510
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.570623015937092e-05,
+      "loss": 0.4396,
+      "step": 2511
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.569706499069451e-05,
+      "loss": 0.475,
+      "step": 2512
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5687899561208862e-05,
+      "loss": 0.4144,
+      "step": 2513
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5678733874343248e-05,
+      "loss": 0.4327,
+      "step": 2514
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5669567933527042e-05,
+      "loss": 0.4216,
+      "step": 2515
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5660401742189716e-05,
+      "loss": 0.4159,
+      "step": 2516
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5651235303760834e-05,
+      "loss": 0.4259,
+      "step": 2517
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.5642068621670043e-05,
+      "loss": 0.4435,
+      "step": 2518
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5632901699347088e-05,
+      "loss": 0.4647,
+      "step": 2519
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.562373454022181e-05,
+      "loss": 0.4594,
+      "step": 2520
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.561456714772413e-05,
+      "loss": 0.5265,
+      "step": 2521
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5605399525284055e-05,
+      "loss": 0.436,
+      "step": 2522
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5596231676331695e-05,
+      "loss": 0.4329,
+      "step": 2523
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.558706360429722e-05,
+      "loss": 0.4496,
+      "step": 2524
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5577895312610897e-05,
+      "loss": 0.4263,
+      "step": 2525
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5568726804703072e-05,
+      "loss": 0.4264,
+      "step": 2526
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5559558084004182e-05,
+      "loss": 0.4367,
+      "step": 2527
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5550389153944724e-05,
+      "loss": 0.4423,
+      "step": 2528
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5541220017955297e-05,
+      "loss": 0.4337,
+      "step": 2529
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.553205067946655e-05,
+      "loss": 0.4859,
+      "step": 2530
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5522881141909235e-05,
+      "loss": 0.4402,
+      "step": 2531
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5513711408714163e-05,
+      "loss": 0.416,
+      "step": 2532
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.550454148331221e-05,
+      "loss": 0.4419,
+      "step": 2533
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5495371369134348e-05,
+      "loss": 0.4476,
+      "step": 2534
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.5486201069611604e-05,
+      "loss": 0.471,
+      "step": 2535
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5477030588175067e-05,
+      "loss": 0.4571,
+      "step": 2536
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5467859928255915e-05,
+      "loss": 0.4628,
+      "step": 2537
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.545868909328538e-05,
+      "loss": 0.426,
+      "step": 2538
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5449518086694758e-05,
+      "loss": 0.4446,
+      "step": 2539
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.544034691191541e-05,
+      "loss": 0.4435,
+      "step": 2540
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.543117557237877e-05,
+      "loss": 0.4359,
+      "step": 2541
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.542200407151632e-05,
+      "loss": 0.4237,
+      "step": 2542
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5412832412759613e-05,
+      "loss": 0.414,
+      "step": 2543
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5403660599540247e-05,
+      "loss": 0.4334,
+      "step": 2544
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5394488635289893e-05,
+      "loss": 0.4123,
+      "step": 2545
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.538531652344028e-05,
+      "loss": 0.431,
+      "step": 2546
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5376144267423172e-05,
+      "loss": 0.4371,
+      "step": 2547
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.536697187067041e-05,
+      "loss": 0.4443,
+      "step": 2548
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5357799336613866e-05,
+      "loss": 0.431,
+      "step": 2549
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5348626668685493e-05,
+      "loss": 0.4493,
+      "step": 2550
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5339453870317255e-05,
+      "loss": 0.4246,
+      "step": 2551
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.5330280944941194e-05,
+      "loss": 0.4251,
+      "step": 2552
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5321107895989406e-05,
+      "loss": 0.4079,
+      "step": 2553
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5311934726893997e-05,
+      "loss": 0.4979,
+      "step": 2554
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5302761441087147e-05,
+      "loss": 0.4201,
+      "step": 2555
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.529358804200107e-05,
+      "loss": 0.4069,
+      "step": 2556
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5284414533068036e-05,
+      "loss": 0.4445,
+      "step": 2557
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.527524091772033e-05,
+      "loss": 0.4769,
+      "step": 2558
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.52660671993903e-05,
+      "loss": 0.4562,
+      "step": 2559
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5256893381510318e-05,
+      "loss": 0.4467,
+      "step": 2560
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5247719467512807e-05,
+      "loss": 0.4302,
+      "step": 2561
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5238545460830211e-05,
+      "loss": 0.448,
+      "step": 2562
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5229371364895017e-05,
+      "loss": 0.4426,
+      "step": 2563
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5220197183139749e-05,
+      "loss": 0.445,
+      "step": 2564
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5211022918996957e-05,
+      "loss": 0.4584,
+      "step": 2565
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5201848575899219e-05,
+      "loss": 0.4734,
+      "step": 2566
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5192674157279148e-05,
+      "loss": 0.4601,
+      "step": 2567
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.5183499666569386e-05,
+      "loss": 0.3944,
+      "step": 2568
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.51743251072026e-05,
+      "loss": 0.4172,
+      "step": 2569
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5165150482611477e-05,
+      "loss": 0.4361,
+      "step": 2570
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5155975796228743e-05,
+      "loss": 0.4446,
+      "step": 2571
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5146801051487136e-05,
+      "loss": 0.4107,
+      "step": 2572
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.513762625181941e-05,
+      "loss": 0.4059,
+      "step": 2573
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5128451400658358e-05,
+      "loss": 0.4161,
+      "step": 2574
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5119276501436775e-05,
+      "loss": 0.4087,
+      "step": 2575
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5110101557587482e-05,
+      "loss": 0.4186,
+      "step": 2576
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5100926572543317e-05,
+      "loss": 0.3996,
+      "step": 2577
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.509175154973713e-05,
+      "loss": 0.4617,
+      "step": 2578
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5082576492601793e-05,
+      "loss": 0.4153,
+      "step": 2579
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5073401404570176e-05,
+      "loss": 0.4534,
+      "step": 2580
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5064226289075176e-05,
+      "loss": 0.435,
+      "step": 2581
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5055051149549686e-05,
+      "loss": 0.43,
+      "step": 2582
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5045875989426629e-05,
+      "loss": 0.4412,
+      "step": 2583
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5036700812138908e-05,
+      "loss": 0.4284,
+      "step": 2584
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5027525621119453e-05,
+      "loss": 0.4186,
+      "step": 2585
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5018350419801197e-05,
+      "loss": 0.4559,
+      "step": 2586
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.5009175211617067e-05,
+      "loss": 0.4318,
+      "step": 2587
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.5e-05,
+      "loss": 0.42,
+      "step": 2588
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4990824788382935e-05,
+      "loss": 0.4327,
+      "step": 2589
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4981649580198805e-05,
+      "loss": 0.4265,
+      "step": 2590
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.497247437888055e-05,
+      "loss": 0.436,
+      "step": 2591
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4963299187861096e-05,
+      "loss": 0.422,
+      "step": 2592
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4954124010573377e-05,
+      "loss": 0.4424,
+      "step": 2593
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4944948850450316e-05,
+      "loss": 0.4111,
+      "step": 2594
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.493577371092483e-05,
+      "loss": 0.4313,
+      "step": 2595
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4926598595429828e-05,
+      "loss": 0.4222,
+      "step": 2596
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4917423507398211e-05,
+      "loss": 0.4204,
+      "step": 2597
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4908248450262869e-05,
+      "loss": 0.4091,
+      "step": 2598
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4899073427456683e-05,
+      "loss": 0.4413,
+      "step": 2599
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4889898442412517e-05,
+      "loss": 0.4126,
+      "step": 2600
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.4672556519508362,
+      "eval_runtime": 1630.5926,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 2600
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4880723498563228e-05,
+      "loss": 0.4466,
+      "step": 2601
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4871548599341646e-05,
+      "loss": 0.4377,
+      "step": 2602
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.486237374818059e-05,
+      "loss": 0.4749,
+      "step": 2603
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4853198948512868e-05,
+      "loss": 0.426,
+      "step": 2604
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.484402420377126e-05,
+      "loss": 0.4719,
+      "step": 2605
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4834849517388524e-05,
+      "loss": 0.4108,
+      "step": 2606
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4825674892797404e-05,
+      "loss": 0.4223,
+      "step": 2607
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4816500333430618e-05,
+      "loss": 0.4322,
+      "step": 2608
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4807325842720856e-05,
+      "loss": 0.4427,
+      "step": 2609
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4798151424100787e-05,
+      "loss": 0.4215,
+      "step": 2610
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4788977081003049e-05,
+      "loss": 0.4481,
+      "step": 2611
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4779802816860254e-05,
+      "loss": 0.4412,
+      "step": 2612
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.477062863510498e-05,
+      "loss": 0.4009,
+      "step": 2613
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4761454539169788e-05,
+      "loss": 0.4485,
+      "step": 2614
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4752280532487192e-05,
+      "loss": 0.4469,
+      "step": 2615
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4743106618489683e-05,
+      "loss": 0.4349,
+      "step": 2616
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4733932800609703e-05,
+      "loss": 0.4226,
+      "step": 2617
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4724759082279672e-05,
+      "loss": 0.4363,
+      "step": 2618
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4715585466931966e-05,
+      "loss": 0.4566,
+      "step": 2619
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4706411957998931e-05,
+      "loss": 0.4483,
+      "step": 2620
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.4697238558912858e-05,
+      "loss": 0.4426,
+      "step": 2621
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4688065273106008e-05,
+      "loss": 0.4373,
+      "step": 2622
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.46788921040106e-05,
+      "loss": 0.3995,
+      "step": 2623
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4669719055058805e-05,
+      "loss": 0.5261,
+      "step": 2624
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.466054612968275e-05,
+      "loss": 0.4056,
+      "step": 2625
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4651373331314514e-05,
+      "loss": 0.4176,
+      "step": 2626
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4642200663386133e-05,
+      "loss": 0.4089,
+      "step": 2627
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4633028129329599e-05,
+      "loss": 0.4439,
+      "step": 2628
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4623855732576825e-05,
+      "loss": 0.3849,
+      "step": 2629
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4614683476559723e-05,
+      "loss": 0.4197,
+      "step": 2630
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4605511364710104e-05,
+      "loss": 0.4758,
+      "step": 2631
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4596339400459753e-05,
+      "loss": 0.4333,
+      "step": 2632
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.458716758724039e-05,
+      "loss": 0.4251,
+      "step": 2633
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4577995928483683e-05,
+      "loss": 0.4421,
+      "step": 2634
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4568824427621232e-05,
+      "loss": 0.423,
+      "step": 2635
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.455965308808459e-05,
+      "loss": 0.4421,
+      "step": 2636
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4550481913305243e-05,
+      "loss": 0.4087,
+      "step": 2637
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.4541310906714624e-05,
+      "loss": 0.4231,
+      "step": 2638
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4532140071744087e-05,
+      "loss": 0.4096,
+      "step": 2639
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4522969411824935e-05,
+      "loss": 0.4806,
+      "step": 2640
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4513798930388402e-05,
+      "loss": 0.4085,
+      "step": 2641
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4504628630865653e-05,
+      "loss": 0.4221,
+      "step": 2642
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4495458516687794e-05,
+      "loss": 0.4368,
+      "step": 2643
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4486288591285846e-05,
+      "loss": 0.4317,
+      "step": 2644
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4477118858090766e-05,
+      "loss": 0.4233,
+      "step": 2645
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4467949320533449e-05,
+      "loss": 0.5205,
+      "step": 2646
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4458779982044706e-05,
+      "loss": 0.447,
+      "step": 2647
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4449610846055273e-05,
+      "loss": 0.4462,
+      "step": 2648
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4440441915995822e-05,
+      "loss": 0.4509,
+      "step": 2649
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4431273195296929e-05,
+      "loss": 0.4121,
+      "step": 2650
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4422104687389107e-05,
+      "loss": 0.4488,
+      "step": 2651
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4412936395702784e-05,
+      "loss": 0.4234,
+      "step": 2652
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4403768323668311e-05,
+      "loss": 0.4039,
+      "step": 2653
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4394600474715944e-05,
+      "loss": 0.4498,
+      "step": 2654
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.4385432852275872e-05,
+      "loss": 0.4536,
+      "step": 2655
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4376265459778192e-05,
+      "loss": 0.4005,
+      "step": 2656
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4367098300652916e-05,
+      "loss": 0.4278,
+      "step": 2657
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4357931378329963e-05,
+      "loss": 0.4389,
+      "step": 2658
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.434876469623917e-05,
+      "loss": 0.4126,
+      "step": 2659
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4339598257810285e-05,
+      "loss": 0.446,
+      "step": 2660
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4330432066472957e-05,
+      "loss": 0.4224,
+      "step": 2661
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4321266125656753e-05,
+      "loss": 0.4527,
+      "step": 2662
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.431210043879114e-05,
+      "loss": 0.4406,
+      "step": 2663
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4302935009305493e-05,
+      "loss": 0.4887,
+      "step": 2664
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4293769840629082e-05,
+      "loss": 0.4356,
+      "step": 2665
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4284604936191091e-05,
+      "loss": 0.4496,
+      "step": 2666
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4275440299420597e-05,
+      "loss": 0.419,
+      "step": 2667
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4266275933746586e-05,
+      "loss": 0.4258,
+      "step": 2668
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.425711184259793e-05,
+      "loss": 0.4243,
+      "step": 2669
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4247948029403406e-05,
+      "loss": 0.4329,
+      "step": 2670
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.423878449759169e-05,
+      "loss": 0.4298,
+      "step": 2671
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4229621250591347e-05,
+      "loss": 0.4363,
+      "step": 2672
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.4220458291830835e-05,
+      "loss": 0.4472,
+      "step": 2673
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4211295624738507e-05,
+      "loss": 0.464,
+      "step": 2674
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4202133252742605e-05,
+      "loss": 0.4095,
+      "step": 2675
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4192971179271266e-05,
+      "loss": 0.4224,
+      "step": 2676
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4183809407752499e-05,
+      "loss": 0.4239,
+      "step": 2677
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4174647941614232e-05,
+      "loss": 0.4323,
+      "step": 2678
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4165486784284244e-05,
+      "loss": 0.4469,
+      "step": 2679
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4156325939190215e-05,
+      "loss": 0.4282,
+      "step": 2680
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4147165409759712e-05,
+      "loss": 0.4327,
+      "step": 2681
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4138005199420177e-05,
+      "loss": 0.4237,
+      "step": 2682
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4128845311598926e-05,
+      "loss": 0.4196,
+      "step": 2683
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.411968574972317e-05,
+      "loss": 0.4715,
+      "step": 2684
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4110526517219987e-05,
+      "loss": 0.4065,
+      "step": 2685
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.410136761751634e-05,
+      "loss": 0.4686,
+      "step": 2686
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4092209054039055e-05,
+      "loss": 0.4221,
+      "step": 2687
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4083050830214844e-05,
+      "loss": 0.4199,
+      "step": 2688
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4073892949470285e-05,
+      "loss": 0.41,
+      "step": 2689
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.4064735415231836e-05,
+      "loss": 0.4307,
+      "step": 2690
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4055578230925814e-05,
+      "loss": 0.4384,
+      "step": 2691
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4046421399978415e-05,
+      "loss": 0.409,
+      "step": 2692
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4037264925815688e-05,
+      "loss": 0.434,
+      "step": 2693
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4028108811863572e-05,
+      "loss": 0.4223,
+      "step": 2694
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4018953061547854e-05,
+      "loss": 0.4611,
+      "step": 2695
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4009797678294187e-05,
+      "loss": 0.4475,
+      "step": 2696
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.4000642665528094e-05,
+      "loss": 0.4848,
+      "step": 2697
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3991488026674947e-05,
+      "loss": 0.4202,
+      "step": 2698
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.398233376515999e-05,
+      "loss": 0.4521,
+      "step": 2699
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3973179884408316e-05,
+      "loss": 0.4336,
+      "step": 2700
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.396402638784489e-05,
+      "loss": 0.4255,
+      "step": 2701
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3954873278894516e-05,
+      "loss": 0.4115,
+      "step": 2702
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3945720560981863e-05,
+      "loss": 0.4419,
+      "step": 2703
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3936568237531451e-05,
+      "loss": 0.4269,
+      "step": 2704
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3927416311967657e-05,
+      "loss": 0.4653,
+      "step": 2705
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.39182647877147e-05,
+      "loss": 0.4271,
+      "step": 2706
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3909113668196654e-05,
+      "loss": 0.4338,
+      "step": 2707
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3899962956837443e-05,
+      "loss": 0.4219,
+      "step": 2708
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3890812657060835e-05,
+      "loss": 0.405,
+      "step": 2709
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3881662772290452e-05,
+      "loss": 0.5019,
+      "step": 2710
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.387251330594975e-05,
+      "loss": 0.4751,
+      "step": 2711
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3863364261462035e-05,
+      "loss": 0.4673,
+      "step": 2712
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3854215642250452e-05,
+      "loss": 0.4253,
+      "step": 2713
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3845067451737984e-05,
+      "loss": 0.4969,
+      "step": 2714
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3835919693347463e-05,
+      "loss": 0.4099,
+      "step": 2715
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3826772370501555e-05,
+      "loss": 0.4471,
+      "step": 2716
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3817625486622757e-05,
+      "loss": 0.4351,
+      "step": 2717
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3808479045133408e-05,
+      "loss": 0.4269,
+      "step": 2718
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3799333049455683e-05,
+      "loss": 0.4151,
+      "step": 2719
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3790187503011586e-05,
+      "loss": 0.4583,
+      "step": 2720
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.378104240922295e-05,
+      "loss": 0.4169,
+      "step": 2721
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3771897771511452e-05,
+      "loss": 0.4408,
+      "step": 2722
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3762753593298581e-05,
+      "loss": 0.4477,
+      "step": 2723
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3753609878005669e-05,
+      "loss": 0.4214,
+      "step": 2724
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3744466629053855e-05,
+      "loss": 0.4203,
+      "step": 2725
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3735323849864138e-05,
+      "loss": 0.4238,
+      "step": 2726
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3726181543857305e-05,
+      "loss": 0.4131,
+      "step": 2727
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3717039714453984e-05,
+      "loss": 0.4311,
+      "step": 2728
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.370789836507462e-05,
+      "loss": 0.5067,
+      "step": 2729
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3698757499139488e-05,
+      "loss": 0.4683,
+      "step": 2730
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3689617120068665e-05,
+      "loss": 0.4371,
+      "step": 2731
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3680477231282058e-05,
+      "loss": 0.4314,
+      "step": 2732
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3671337836199388e-05,
+      "loss": 0.4335,
+      "step": 2733
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3662198938240192e-05,
+      "loss": 0.4534,
+      "step": 2734
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3653060540823817e-05,
+      "loss": 0.4226,
+      "step": 2735
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3643922647369425e-05,
+      "loss": 0.4318,
+      "step": 2736
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3634785261295992e-05,
+      "loss": 0.4286,
+      "step": 2737
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3625648386022306e-05,
+      "loss": 0.4328,
+      "step": 2738
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3616512024966952e-05,
+      "loss": 0.4694,
+      "step": 2739
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3607376181548336e-05,
+      "loss": 0.489,
+      "step": 2740
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3598240859184657e-05,
+      "loss": 0.4744,
+      "step": 2741
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3589106061293938e-05,
+      "loss": 0.4007,
+      "step": 2742
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3579971791293991e-05,
+      "loss": 0.4381,
+      "step": 2743
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3570838052602433e-05,
+      "loss": 0.396,
+      "step": 2744
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3561704848636686e-05,
+      "loss": 0.4014,
+      "step": 2745
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3552572182813964e-05,
+      "loss": 0.4509,
+      "step": 2746
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3543440058551289e-05,
+      "loss": 0.4147,
+      "step": 2747
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3534308479265475e-05,
+      "loss": 0.4415,
+      "step": 2748
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3525177448373138e-05,
+      "loss": 0.4245,
+      "step": 2749
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3516046969290676e-05,
+      "loss": 0.4336,
+      "step": 2750
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3506917045434293e-05,
+      "loss": 0.4273,
+      "step": 2751
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3497787680219982e-05,
+      "loss": 0.4471,
+      "step": 2752
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3488658877063527e-05,
+      "loss": 0.4864,
+      "step": 2753
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3479530639380497e-05,
+      "loss": 0.5016,
+      "step": 2754
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3470402970586255e-05,
+      "loss": 0.44,
+      "step": 2755
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.346127587409595e-05,
+      "loss": 0.4438,
+      "step": 2756
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3452149353324512e-05,
+      "loss": 0.422,
+      "step": 2757
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3443023411686669e-05,
+      "loss": 0.4594,
+      "step": 2758
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3433898052596916e-05,
+      "loss": 0.4443,
+      "step": 2759
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.342477327946954e-05,
+      "loss": 0.4432,
+      "step": 2760
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3415649095718604e-05,
+      "loss": 0.4462,
+      "step": 2761
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3406525504757953e-05,
+      "loss": 0.4184,
+      "step": 2762
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.339740251000121e-05,
+      "loss": 0.4334,
+      "step": 2763
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.338828011486178e-05,
+      "loss": 0.4394,
+      "step": 2764
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3379158322752825e-05,
+      "loss": 0.4189,
+      "step": 2765
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3370037137087306e-05,
+      "loss": 0.4784,
+      "step": 2766
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3360916561277939e-05,
+      "loss": 0.41,
+      "step": 2767
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3351796598737228e-05,
+      "loss": 0.4095,
+      "step": 2768
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3342677252877425e-05,
+      "loss": 0.4257,
+      "step": 2769
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3333558527110572e-05,
+      "loss": 0.4326,
+      "step": 2770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3324440424848472e-05,
+      "loss": 0.4441,
+      "step": 2771
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3315322949502693e-05,
+      "loss": 0.4192,
+      "step": 2772
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.330620610448456e-05,
+      "loss": 0.4952,
+      "step": 2773
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3297089893205188e-05,
+      "loss": 0.4435,
+      "step": 2774
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3287974319075428e-05,
+      "loss": 0.4372,
+      "step": 2775
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3278859385505906e-05,
+      "loss": 0.4344,
+      "step": 2776
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3269745095907005e-05,
+      "loss": 0.4434,
+      "step": 2777
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3260631453688873e-05,
+      "loss": 0.4543,
+      "step": 2778
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.32515184622614e-05,
+      "loss": 0.4213,
+      "step": 2779
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3242406125034249e-05,
+      "loss": 0.3946,
+      "step": 2780
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3233294445416828e-05,
+      "loss": 0.4457,
+      "step": 2781
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3224183426818311e-05,
+      "loss": 0.4217,
+      "step": 2782
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3215073072647606e-05,
+      "loss": 0.4245,
+      "step": 2783
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.320596338631339e-05,
+      "loss": 0.419,
+      "step": 2784
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3196854371224081e-05,
+      "loss": 0.4237,
+      "step": 2785
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3187746030787854e-05,
+      "loss": 0.4326,
+      "step": 2786
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3178638368412612e-05,
+      "loss": 0.4326,
+      "step": 2787
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.316953138750603e-05,
+      "loss": 0.4306,
+      "step": 2788
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3160425091475506e-05,
+      "loss": 0.4312,
+      "step": 2789
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3151319483728201e-05,
+      "loss": 0.4316,
+      "step": 2790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3142214567671008e-05,
+      "loss": 0.4769,
+      "step": 2791
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3133110346710557e-05,
+      "loss": 0.4367,
+      "step": 2792
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3124006824253231e-05,
+      "loss": 0.4184,
+      "step": 2793
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3114904003705137e-05,
+      "loss": 0.4198,
+      "step": 2794
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3105801888472127e-05,
+      "loss": 0.398,
+      "step": 2795
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.309670048195979e-05,
+      "loss": 0.4368,
+      "step": 2796
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3087599787573452e-05,
+      "loss": 0.4468,
+      "step": 2797
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3078499808718161e-05,
+      "loss": 0.4387,
+      "step": 2798
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3069400548798708e-05,
+      "loss": 0.4091,
+      "step": 2799
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3060302011219615e-05,
+      "loss": 0.4322,
+      "step": 2800
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 0.4646455943584442,
+      "eval_runtime": 1630.2219,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 2800
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3051204199385133e-05,
+      "loss": 0.4334,
+      "step": 2801
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.304210711669923e-05,
+      "loss": 0.4332,
+      "step": 2802
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3033010766565616e-05,
+      "loss": 0.4478,
+      "step": 2803
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.302391515238772e-05,
+      "loss": 0.4602,
+      "step": 2804
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3014820277568696e-05,
+      "loss": 0.4827,
+      "step": 2805
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3005726145511427e-05,
+      "loss": 0.4171,
+      "step": 2806
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2996632759618516e-05,
+      "loss": 0.4479,
+      "step": 2807
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2987540123292274e-05,
+      "loss": 0.4245,
+      "step": 2808
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2978448239934748e-05,
+      "loss": 0.4718,
+      "step": 2809
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.2969357112947695e-05,
+      "loss": 0.4231,
+      "step": 2810
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2960266745732595e-05,
+      "loss": 0.4211,
+      "step": 2811
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2951177141690637e-05,
+      "loss": 0.4498,
+      "step": 2812
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2942088304222725e-05,
+      "loss": 0.44,
+      "step": 2813
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2933000236729481e-05,
+      "loss": 0.4174,
+      "step": 2814
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2923912942611237e-05,
+      "loss": 0.4353,
+      "step": 2815
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2914826425268036e-05,
+      "loss": 0.4363,
+      "step": 2816
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2905740688099626e-05,
+      "loss": 0.4097,
+      "step": 2817
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2896655734505469e-05,
+      "loss": 0.4217,
+      "step": 2818
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2887571567884728e-05,
+      "loss": 0.4226,
+      "step": 2819
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2878488191636282e-05,
+      "loss": 0.4898,
+      "step": 2820
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2869405609158692e-05,
+      "loss": 0.4018,
+      "step": 2821
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.286032382385026e-05,
+      "loss": 0.4456,
+      "step": 2822
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2851242839108949e-05,
+      "loss": 0.46,
+      "step": 2823
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2842162658332445e-05,
+      "loss": 0.4265,
+      "step": 2824
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2833083284918132e-05,
+      "loss": 0.4214,
+      "step": 2825
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2824004722263087e-05,
+      "loss": 0.4213,
+      "step": 2826
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.2814926973764083e-05,
+      "loss": 0.4441,
+      "step": 2827
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.280585004281759e-05,
+      "loss": 0.4499,
+      "step": 2828
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2796773932819774e-05,
+      "loss": 0.4407,
+      "step": 2829
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2787698647166498e-05,
+      "loss": 0.4586,
+      "step": 2830
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.27786241892533e-05,
+      "loss": 0.4271,
+      "step": 2831
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2769550562475427e-05,
+      "loss": 0.422,
+      "step": 2832
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2760477770227806e-05,
+      "loss": 0.441,
+      "step": 2833
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2751405815905054e-05,
+      "loss": 0.4077,
+      "step": 2834
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.274233470290147e-05,
+      "loss": 0.4538,
+      "step": 2835
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2733264434611042e-05,
+      "loss": 0.4983,
+      "step": 2836
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2724195014427443e-05,
+      "loss": 0.4642,
+      "step": 2837
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2715126445744028e-05,
+      "loss": 0.4301,
+      "step": 2838
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2706058731953837e-05,
+      "loss": 0.4208,
+      "step": 2839
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2696991876449581e-05,
+      "loss": 0.4001,
+      "step": 2840
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2687925882623657e-05,
+      "loss": 0.3954,
+      "step": 2841
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2678860753868134e-05,
+      "loss": 0.4255,
+      "step": 2842
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.2669796493574764e-05,
+      "loss": 0.4222,
+      "step": 2843
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.266073310513497e-05,
+      "loss": 0.4448,
+      "step": 2844
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.265167059193985e-05,
+      "loss": 0.4196,
+      "step": 2845
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2642608957380173e-05,
+      "loss": 0.4438,
+      "step": 2846
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.263354820484638e-05,
+      "loss": 0.418,
+      "step": 2847
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2624488337728585e-05,
+      "loss": 0.4088,
+      "step": 2848
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.261542935941657e-05,
+      "loss": 0.4088,
+      "step": 2849
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2606371273299776e-05,
+      "loss": 0.4497,
+      "step": 2850
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.259731408276732e-05,
+      "loss": 0.4182,
+      "step": 2851
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2588257791207979e-05,
+      "loss": 0.4206,
+      "step": 2852
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2579202402010194e-05,
+      "loss": 0.4066,
+      "step": 2853
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2570147918562076e-05,
+      "loss": 0.4849,
+      "step": 2854
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2561094344251394e-05,
+      "loss": 0.4243,
+      "step": 2855
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.255204168246556e-05,
+      "loss": 0.43,
+      "step": 2856
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2542989936591669e-05,
+      "loss": 0.4046,
+      "step": 2857
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2533939110016456e-05,
+      "loss": 0.4341,
+      "step": 2858
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2524889206126324e-05,
+      "loss": 0.4161,
+      "step": 2859
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2515840228307325e-05,
+      "loss": 0.4148,
+      "step": 2860
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2506792179945158e-05,
+      "loss": 0.4152,
+      "step": 2861
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.2497745064425189e-05,
+      "loss": 0.4266,
+      "step": 2862
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2488698885132422e-05,
+      "loss": 0.4296,
+      "step": 2863
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.247965364545152e-05,
+      "loss": 0.3862,
+      "step": 2864
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2470609348766787e-05,
+      "loss": 0.4325,
+      "step": 2865
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2461565998462178e-05,
+      "loss": 0.4504,
+      "step": 2866
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2452523597921292e-05,
+      "loss": 0.3977,
+      "step": 2867
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.244348215052738e-05,
+      "loss": 0.4206,
+      "step": 2868
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2434441659663316e-05,
+      "loss": 0.3946,
+      "step": 2869
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.242540212871165e-05,
+      "loss": 0.414,
+      "step": 2870
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.241636356105454e-05,
+      "loss": 0.4536,
+      "step": 2871
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2407325960073803e-05,
+      "loss": 0.4541,
+      "step": 2872
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2398289329150884e-05,
+      "loss": 0.4241,
+      "step": 2873
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2389253671666874e-05,
+      "loss": 0.4228,
+      "step": 2874
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2380218991002488e-05,
+      "loss": 0.4714,
+      "step": 2875
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2371185290538088e-05,
+      "loss": 0.4554,
+      "step": 2876
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.236215257365366e-05,
+      "loss": 0.435,
+      "step": 2877
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2353120843728834e-05,
+      "loss": 0.4251,
+      "step": 2878
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.2344090104142852e-05,
+      "loss": 0.4312,
+      "step": 2879
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2335060358274602e-05,
+      "loss": 0.4328,
+      "step": 2880
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2326031609502591e-05,
+      "loss": 0.4499,
+      "step": 2881
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2317003861204961e-05,
+      "loss": 0.4091,
+      "step": 2882
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2307977116759469e-05,
+      "loss": 0.4175,
+      "step": 2883
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2298951379543503e-05,
+      "loss": 0.4382,
+      "step": 2884
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2289926652934075e-05,
+      "loss": 0.4503,
+      "step": 2885
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2280902940307819e-05,
+      "loss": 0.4233,
+      "step": 2886
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2271880245040986e-05,
+      "loss": 0.4136,
+      "step": 2887
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2262858570509446e-05,
+      "loss": 0.4222,
+      "step": 2888
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2253837920088698e-05,
+      "loss": 0.4368,
+      "step": 2889
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2244818297153838e-05,
+      "loss": 0.4323,
+      "step": 2890
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2235799705079593e-05,
+      "loss": 0.4525,
+      "step": 2891
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2226782147240298e-05,
+      "loss": 0.3862,
+      "step": 2892
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2217765627009911e-05,
+      "loss": 0.4332,
+      "step": 2893
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2208750147761984e-05,
+      "loss": 0.4103,
+      "step": 2894
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2199735712869691e-05,
+      "loss": 0.4436,
+      "step": 2895
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.2190722325705819e-05,
+      "loss": 0.458,
+      "step": 2896
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2181709989642755e-05,
+      "loss": 0.4173,
+      "step": 2897
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2172698708052492e-05,
+      "loss": 0.4667,
+      "step": 2898
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2163688484306633e-05,
+      "loss": 0.4492,
+      "step": 2899
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2154679321776385e-05,
+      "loss": 0.393,
+      "step": 2900
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2145671223832555e-05,
+      "loss": 0.44,
+      "step": 2901
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2136664193845555e-05,
+      "loss": 0.4417,
+      "step": 2902
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2127658235185404e-05,
+      "loss": 0.4532,
+      "step": 2903
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2118653351221698e-05,
+      "loss": 0.4317,
+      "step": 2904
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2109649545323651e-05,
+      "loss": 0.4376,
+      "step": 2905
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.210064682086007e-05,
+      "loss": 0.4016,
+      "step": 2906
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2091645181199354e-05,
+      "loss": 0.4035,
+      "step": 2907
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2082644629709494e-05,
+      "loss": 0.4188,
+      "step": 2908
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2073645169758078e-05,
+      "loss": 0.4435,
+      "step": 2909
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2064646804712285e-05,
+      "loss": 0.443,
+      "step": 2910
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2055649537938886e-05,
+      "loss": 0.4458,
+      "step": 2911
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.204665337280424e-05,
+      "loss": 0.4135,
+      "step": 2912
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.2037658312674286e-05,
+      "loss": 0.4008,
+      "step": 2913
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.2028664360914563e-05,
+      "loss": 0.4202,
+      "step": 2914
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.2019671520890185e-05,
+      "loss": 0.442,
+      "step": 2915
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.2010679795965859e-05,
+      "loss": 0.4313,
+      "step": 2916
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.200168918950586e-05,
+      "loss": 0.3969,
+      "step": 2917
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.199269970487406e-05,
+      "loss": 0.4122,
+      "step": 2918
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1983711345433905e-05,
+      "loss": 0.4349,
+      "step": 2919
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.197472411454842e-05,
+      "loss": 0.4273,
+      "step": 2920
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1965738015580209e-05,
+      "loss": 0.5154,
+      "step": 2921
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1956753051891456e-05,
+      "loss": 0.4172,
+      "step": 2922
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1947769226843905e-05,
+      "loss": 0.4347,
+      "step": 2923
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1938786543798891e-05,
+      "loss": 0.4531,
+      "step": 2924
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1929805006117315e-05,
+      "loss": 0.4256,
+      "step": 2925
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.192082461715965e-05,
+      "loss": 0.4306,
+      "step": 2926
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1911845380285938e-05,
+      "loss": 0.4186,
+      "step": 2927
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1902867298855791e-05,
+      "loss": 0.4485,
+      "step": 2928
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.189389037622839e-05,
+      "loss": 0.4111,
+      "step": 2929
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.1884914615762483e-05,
+      "loss": 0.4011,
+      "step": 2930
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1875940020816377e-05,
+      "loss": 0.4242,
+      "step": 2931
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1866966594747949e-05,
+      "loss": 0.4647,
+      "step": 2932
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1857994340914636e-05,
+      "loss": 0.4358,
+      "step": 2933
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1849023262673436e-05,
+      "loss": 0.4529,
+      "step": 2934
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1840053363380912e-05,
+      "loss": 0.4391,
+      "step": 2935
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1831084646393184e-05,
+      "loss": 0.4282,
+      "step": 2936
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1822117115065927e-05,
+      "loss": 0.4224,
+      "step": 2937
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.181315077275437e-05,
+      "loss": 0.4292,
+      "step": 2938
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.18041856228133e-05,
+      "loss": 0.4391,
+      "step": 2939
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1795221668597062e-05,
+      "loss": 0.441,
+      "step": 2940
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1786258913459552e-05,
+      "loss": 0.4463,
+      "step": 2941
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1777297360754207e-05,
+      "loss": 0.4567,
+      "step": 2942
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1768337013834028e-05,
+      "loss": 0.4056,
+      "step": 2943
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1759377876051559e-05,
+      "loss": 0.4265,
+      "step": 2944
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1750419950758895e-05,
+      "loss": 0.4193,
+      "step": 2945
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1741463241307665e-05,
+      "loss": 0.4022,
+      "step": 2946
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1732507751049062e-05,
+      "loss": 0.446,
+      "step": 2947
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.1723553483333807e-05,
+      "loss": 0.427,
+      "step": 2948
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1714600441512176e-05,
+      "loss": 0.448,
+      "step": 2949
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1705648628933966e-05,
+      "loss": 0.4321,
+      "step": 2950
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1696698048948549e-05,
+      "loss": 0.4821,
+      "step": 2951
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.16877487049048e-05,
+      "loss": 0.4368,
+      "step": 2952
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.167880060015115e-05,
+      "loss": 0.4311,
+      "step": 2953
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1669853738035565e-05,
+      "loss": 0.4395,
+      "step": 2954
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1660908121905547e-05,
+      "loss": 0.4444,
+      "step": 2955
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1651963755108119e-05,
+      "loss": 0.4307,
+      "step": 2956
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1643020640989852e-05,
+      "loss": 0.4227,
+      "step": 2957
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1634078782896841e-05,
+      "loss": 0.4237,
+      "step": 2958
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1625138184174718e-05,
+      "loss": 0.4336,
+      "step": 2959
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.161619884816863e-05,
+      "loss": 0.4321,
+      "step": 2960
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1607260778223267e-05,
+      "loss": 0.4408,
+      "step": 2961
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1598323977682834e-05,
+      "loss": 0.4178,
+      "step": 2962
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1589388449891067e-05,
+      "loss": 0.4468,
+      "step": 2963
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1580454198191228e-05,
+      "loss": 0.4729,
+      "step": 2964
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.1571521225926091e-05,
+      "loss": 0.427,
+      "step": 2965
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1562589536437956e-05,
+      "loss": 0.5399,
+      "step": 2966
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1553659133068656e-05,
+      "loss": 0.4539,
+      "step": 2967
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1544730019159523e-05,
+      "loss": 0.4137,
+      "step": 2968
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1535802198051419e-05,
+      "loss": 0.428,
+      "step": 2969
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1526875673084721e-05,
+      "loss": 0.4714,
+      "step": 2970
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1517950447599311e-05,
+      "loss": 0.4482,
+      "step": 2971
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1509026524934597e-05,
+      "loss": 0.4285,
+      "step": 2972
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1500103908429495e-05,
+      "loss": 0.4369,
+      "step": 2973
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1491182601422436e-05,
+      "loss": 0.4446,
+      "step": 2974
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1482262607251352e-05,
+      "loss": 0.4155,
+      "step": 2975
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1473343929253688e-05,
+      "loss": 0.4493,
+      "step": 2976
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1464426570766403e-05,
+      "loss": 0.4554,
+      "step": 2977
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1455510535125956e-05,
+      "loss": 0.4355,
+      "step": 2978
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.144659582566831e-05,
+      "loss": 0.4112,
+      "step": 2979
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1437682445728933e-05,
+      "loss": 0.391,
+      "step": 2980
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1428770398642797e-05,
+      "loss": 0.4122,
+      "step": 2981
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.1419859687744375e-05,
+      "loss": 0.4228,
+      "step": 2982
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.141095031636764e-05,
+      "loss": 0.4316,
+      "step": 2983
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1402042287846068e-05,
+      "loss": 0.4333,
+      "step": 2984
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1393135605512627e-05,
+      "loss": 0.4482,
+      "step": 2985
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1384230272699778e-05,
+      "loss": 0.4118,
+      "step": 2986
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1375326292739481e-05,
+      "loss": 0.4181,
+      "step": 2987
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1366423668963194e-05,
+      "loss": 0.3909,
+      "step": 2988
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1357522404701867e-05,
+      "loss": 0.4312,
+      "step": 2989
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1348622503285932e-05,
+      "loss": 0.4222,
+      "step": 2990
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1339723968045318e-05,
+      "loss": 0.4242,
+      "step": 2991
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1330826802309444e-05,
+      "loss": 0.4433,
+      "step": 2992
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.132193100940722e-05,
+      "loss": 0.5265,
+      "step": 2993
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1313036592667027e-05,
+      "loss": 0.4041,
+      "step": 2994
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1304143555416747e-05,
+      "loss": 0.4428,
+      "step": 2995
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1295251900983741e-05,
+      "loss": 0.4248,
+      "step": 2996
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.128636163269485e-05,
+      "loss": 0.4069,
+      "step": 2997
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1277472753876393e-05,
+      "loss": 0.4607,
+      "step": 2998
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.1268585267854188e-05,
+      "loss": 0.4808,
+      "step": 2999
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1259699177953507e-05,
+      "loss": 0.4227,
+      "step": 3000
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.4606461226940155,
+      "eval_runtime": 1630.7829,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 3000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1250814487499112e-05,
+      "loss": 0.4572,
+      "step": 3001
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1241931199815243e-05,
+      "loss": 0.5127,
+      "step": 3002
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1233049318225616e-05,
+      "loss": 0.4315,
+      "step": 3003
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1224168846053407e-05,
+      "loss": 0.4668,
+      "step": 3004
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.121528978662128e-05,
+      "loss": 0.4429,
+      "step": 3005
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1206412143251369e-05,
+      "loss": 0.437,
+      "step": 3006
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.119753591926527e-05,
+      "loss": 0.4005,
+      "step": 3007
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1188661117984055e-05,
+      "loss": 0.4347,
+      "step": 3008
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1179787742728255e-05,
+      "loss": 0.4234,
+      "step": 3009
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1170915796817882e-05,
+      "loss": 0.4307,
+      "step": 3010
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1162045283572401e-05,
+      "loss": 0.4454,
+      "step": 3011
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1153176206310747e-05,
+      "loss": 0.4213,
+      "step": 3012
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1144308568351311e-05,
+      "loss": 0.4449,
+      "step": 3013
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1135442373011948e-05,
+      "loss": 0.4531,
+      "step": 3014
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1126577623609982e-05,
+      "loss": 0.4389,
+      "step": 3015
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.1117714323462188e-05,
+      "loss": 0.451,
+      "step": 3016
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1108852475884799e-05,
+      "loss": 0.4159,
+      "step": 3017
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1099992084193506e-05,
+      "loss": 0.4244,
+      "step": 3018
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1091133151703452e-05,
+      "loss": 0.4434,
+      "step": 3019
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1082275681729237e-05,
+      "loss": 0.4138,
+      "step": 3020
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1073419677584916e-05,
+      "loss": 0.4533,
+      "step": 3021
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1064565142583995e-05,
+      "loss": 0.4271,
+      "step": 3022
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1055712080039422e-05,
+      "loss": 0.4322,
+      "step": 3023
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1046860493263604e-05,
+      "loss": 0.4176,
+      "step": 3024
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1038010385568394e-05,
+      "loss": 0.4086,
+      "step": 3025
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.102916176026509e-05,
+      "loss": 0.444,
+      "step": 3026
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.102031462066443e-05,
+      "loss": 0.426,
+      "step": 3027
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1011468970076606e-05,
+      "loss": 0.4476,
+      "step": 3028
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.1002624811811246e-05,
+      "loss": 0.4247,
+      "step": 3029
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.099378214917742e-05,
+      "loss": 0.4246,
+      "step": 3030
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.0984940985483649e-05,
+      "loss": 0.4322,
+      "step": 3031
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.0976101324037878e-05,
+      "loss": 0.4346,
+      "step": 3032
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.0967263168147498e-05,
+      "loss": 0.4033,
+      "step": 3033
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.0958426521119332e-05,
+      "loss": 0.4331,
+      "step": 3034
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0949591386259646e-05,
+      "loss": 0.4301,
+      "step": 3035
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0940757766874133e-05,
+      "loss": 0.4148,
+      "step": 3036
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0931925666267925e-05,
+      "loss": 0.4455,
+      "step": 3037
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.092309508774558e-05,
+      "loss": 0.4575,
+      "step": 3038
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0914266034611088e-05,
+      "loss": 0.4299,
+      "step": 3039
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.090543851016787e-05,
+      "loss": 0.4124,
+      "step": 3040
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0896612517718776e-05,
+      "loss": 0.4041,
+      "step": 3041
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.088778806056608e-05,
+      "loss": 0.412,
+      "step": 3042
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0878965142011478e-05,
+      "loss": 0.4322,
+      "step": 3043
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0870143765356105e-05,
+      "loss": 0.4127,
+      "step": 3044
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0861323933900503e-05,
+      "loss": 0.4313,
+      "step": 3045
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0852505650944636e-05,
+      "loss": 0.4486,
+      "step": 3046
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0843688919787909e-05,
+      "loss": 0.4316,
+      "step": 3047
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.083487374372912e-05,
+      "loss": 0.4247,
+      "step": 3048
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0826060126066506e-05,
+      "loss": 0.4401,
+      "step": 3049
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0817248070097706e-05,
+      "loss": 0.4363,
+      "step": 3050
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.0808437579119781e-05,
+      "loss": 0.4242,
+      "step": 3051
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0799628656429208e-05,
+      "loss": 0.4336,
+      "step": 3052
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0790821305321875e-05,
+      "loss": 0.4114,
+      "step": 3053
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.078201552909308e-05,
+      "loss": 0.5086,
+      "step": 3054
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0773211331037542e-05,
+      "loss": 0.4013,
+      "step": 3055
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0764408714449372e-05,
+      "loss": 0.4079,
+      "step": 3056
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0755607682622104e-05,
+      "loss": 0.4888,
+      "step": 3057
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0746808238848672e-05,
+      "loss": 0.4251,
+      "step": 3058
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0738010386421421e-05,
+      "loss": 0.4298,
+      "step": 3059
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0729214128632093e-05,
+      "loss": 0.4111,
+      "step": 3060
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0720419468771843e-05,
+      "loss": 0.4601,
+      "step": 3061
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0711626410131212e-05,
+      "loss": 0.4686,
+      "step": 3062
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0702834956000166e-05,
+      "loss": 0.4174,
+      "step": 3063
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0694045109668055e-05,
+      "loss": 0.4056,
+      "step": 3064
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0685256874423625e-05,
+      "loss": 0.4486,
+      "step": 3065
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.067647025355503e-05,
+      "loss": 0.4372,
+      "step": 3066
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0667685250349809e-05,
+      "loss": 0.4812,
+      "step": 3067
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.0658901868094901e-05,
+      "loss": 0.4129,
+      "step": 3068
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0650120110076639e-05,
+      "loss": 0.4259,
+      "step": 3069
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0641339979580752e-05,
+      "loss": 0.42,
+      "step": 3070
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0632561479892353e-05,
+      "loss": 0.4302,
+      "step": 3071
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.062378461429594e-05,
+      "loss": 0.4031,
+      "step": 3072
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0615009386075419e-05,
+      "loss": 0.4297,
+      "step": 3073
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.060623579851407e-05,
+      "loss": 0.4284,
+      "step": 3074
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.059746385489455e-05,
+      "loss": 0.4272,
+      "step": 3075
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0588693558498922e-05,
+      "loss": 0.3922,
+      "step": 3076
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0579924912608616e-05,
+      "loss": 0.4323,
+      "step": 3077
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0571157920504452e-05,
+      "loss": 0.441,
+      "step": 3078
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0562392585466634e-05,
+      "loss": 0.4247,
+      "step": 3079
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0553628910774741e-05,
+      "loss": 0.4317,
+      "step": 3080
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0544866899707734e-05,
+      "loss": 0.4147,
+      "step": 3081
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.053610655554394e-05,
+      "loss": 0.4289,
+      "step": 3082
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0527347881561082e-05,
+      "loss": 0.4165,
+      "step": 3083
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0518590881036237e-05,
+      "loss": 0.4215,
+      "step": 3084
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.0509835557245877e-05,
+      "loss": 0.431,
+      "step": 3085
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.050108191346583e-05,
+      "loss": 0.4398,
+      "step": 3086
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0492329952971303e-05,
+      "loss": 0.4227,
+      "step": 3087
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0483579679036873e-05,
+      "loss": 0.4657,
+      "step": 3088
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0474831094936484e-05,
+      "loss": 0.4374,
+      "step": 3089
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0466084203943447e-05,
+      "loss": 0.4407,
+      "step": 3090
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0457339009330443e-05,
+      "loss": 0.433,
+      "step": 3091
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0448595514369515e-05,
+      "loss": 0.4469,
+      "step": 3092
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0439853722332075e-05,
+      "loss": 0.4398,
+      "step": 3093
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0431113636488883e-05,
+      "loss": 0.4237,
+      "step": 3094
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.042237526011009e-05,
+      "loss": 0.4518,
+      "step": 3095
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0413638596465174e-05,
+      "loss": 0.4303,
+      "step": 3096
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0404903648822996e-05,
+      "loss": 0.4207,
+      "step": 3097
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0396170420451761e-05,
+      "loss": 0.3916,
+      "step": 3098
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0387438914619043e-05,
+      "loss": 0.4432,
+      "step": 3099
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0378709134591757e-05,
+      "loss": 0.424,
+      "step": 3100
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.0369981083636183e-05,
+      "loss": 0.4437,
+      "step": 3101
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.036125476501795e-05,
+      "loss": 0.4205,
+      "step": 3102
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0352530182002047e-05,
+      "loss": 0.4275,
+      "step": 3103
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0343807337852794e-05,
+      "loss": 0.4269,
+      "step": 3104
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0335086235833882e-05,
+      "loss": 0.4248,
+      "step": 3105
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0326366879208336e-05,
+      "loss": 0.435,
+      "step": 3106
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0317649271238539e-05,
+      "loss": 0.4201,
+      "step": 3107
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0308933415186207e-05,
+      "loss": 0.4085,
+      "step": 3108
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0300219314312409e-05,
+      "loss": 0.443,
+      "step": 3109
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0291506971877555e-05,
+      "loss": 0.4392,
+      "step": 3110
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0282796391141402e-05,
+      "loss": 0.4511,
+      "step": 3111
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.027408757536304e-05,
+      "loss": 0.4098,
+      "step": 3112
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0265380527800904e-05,
+      "loss": 0.4209,
+      "step": 3113
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0256675251712765e-05,
+      "loss": 0.432,
+      "step": 3114
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.024797175035573e-05,
+      "loss": 0.4289,
+      "step": 3115
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0239270026986243e-05,
+      "loss": 0.4429,
+      "step": 3116
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0230570084860087e-05,
+      "loss": 0.4247,
+      "step": 3117
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0221871927232377e-05,
+      "loss": 0.4362,
+      "step": 3118
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.0213175557357546e-05,
+      "loss": 0.4135,
+      "step": 3119
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0204480978489381e-05,
+      "loss": 0.431,
+      "step": 3120
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0195788193880989e-05,
+      "loss": 0.4539,
+      "step": 3121
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0187097206784803e-05,
+      "loss": 0.4239,
+      "step": 3122
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.017840802045258e-05,
+      "loss": 0.4098,
+      "step": 3123
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0169720638135415e-05,
+      "loss": 0.3993,
+      "step": 3124
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0161035063083716e-05,
+      "loss": 0.4461,
+      "step": 3125
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0152351298547224e-05,
+      "loss": 0.4579,
+      "step": 3126
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0143669347775002e-05,
+      "loss": 0.4079,
+      "step": 3127
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.013498921401543e-05,
+      "loss": 0.4459,
+      "step": 3128
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0126310900516209e-05,
+      "loss": 0.446,
+      "step": 3129
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0117634410524356e-05,
+      "loss": 0.4189,
+      "step": 3130
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0108959747286214e-05,
+      "loss": 0.4277,
+      "step": 3131
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0100286914047433e-05,
+      "loss": 0.4395,
+      "step": 3132
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.009161591405299e-05,
+      "loss": 0.4566,
+      "step": 3133
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0082946750547162e-05,
+      "loss": 0.4393,
+      "step": 3134
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.007427942677355e-05,
+      "loss": 0.4224,
+      "step": 3135
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0065613945975064e-05,
+      "loss": 0.4751,
+      "step": 3136
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.0056950311393918e-05,
+      "loss": 0.4322,
+      "step": 3137
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.0048288526271644e-05,
+      "loss": 0.4621,
+      "step": 3138
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.003962859384908e-05,
+      "loss": 0.4123,
+      "step": 3139
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.0030970517366363e-05,
+      "loss": 0.478,
+      "step": 3140
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.002231430006295e-05,
+      "loss": 0.4005,
+      "step": 3141
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.001365994517758e-05,
+      "loss": 0.4329,
+      "step": 3142
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.000500745594833e-05,
+      "loss": 0.4112,
+      "step": 3143
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.996356835612544e-06,
+      "loss": 0.4066,
+      "step": 3144
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.987708087406884e-06,
+      "loss": 0.4465,
+      "step": 3145
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.97906121456731e-06,
+      "loss": 0.4167,
+      "step": 3146
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.970416220329081e-06,
+      "loss": 0.4266,
+      "step": 3147
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.961773107926746e-06,
+      "loss": 0.4793,
+      "step": 3148
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.953131880594156e-06,
+      "loss": 0.4459,
+      "step": 3149
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.944492541564456e-06,
+      "loss": 0.4755,
+      "step": 3150
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.93585509407009e-06,
+      "loss": 0.4236,
+      "step": 3151
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.927219541342777e-06,
+      "loss": 0.4199,
+      "step": 3152
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.918585886613548e-06,
+      "loss": 0.4356,
+      "step": 3153
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.909954133112709e-06,
+      "loss": 0.4439,
+      "step": 3154
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.901324284069868e-06,
+      "loss": 0.4552,
+      "step": 3155
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.892696342713901e-06,
+      "loss": 0.4534,
+      "step": 3156
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.884070312272986e-06,
+      "loss": 0.4457,
+      "step": 3157
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.87544619597458e-06,
+      "loss": 0.4277,
+      "step": 3158
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.866823997045433e-06,
+      "loss": 0.4425,
+      "step": 3159
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.858203718711564e-06,
+      "loss": 0.4345,
+      "step": 3160
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.84958536419828e-06,
+      "loss": 0.3949,
+      "step": 3161
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.840968936730172e-06,
+      "loss": 0.4319,
+      "step": 3162
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.832354439531102e-06,
+      "loss": 0.425,
+      "step": 3163
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.82374187582421e-06,
+      "loss": 0.4191,
+      "step": 3164
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.81513124883192e-06,
+      "loss": 0.4147,
+      "step": 3165
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.80652256177593e-06,
+      "loss": 0.4242,
+      "step": 3166
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.797915817877205e-06,
+      "loss": 0.4153,
+      "step": 3167
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.789311020355987e-06,
+      "loss": 0.4696,
+      "step": 3168
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.780708172431793e-06,
+      "loss": 0.4079,
+      "step": 3169
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.772107277323412e-06,
+      "loss": 0.4221,
+      "step": 3170
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.763508338248888e-06,
+      "loss": 0.4138,
+      "step": 3171
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.754911358425551e-06,
+      "loss": 0.4304,
+      "step": 3172
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.746316341069985e-06,
+      "loss": 0.4182,
+      "step": 3173
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.737723289398049e-06,
+      "loss": 0.416,
+      "step": 3174
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.72913220662486e-06,
+      "loss": 0.4233,
+      "step": 3175
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.720543095964808e-06,
+      "loss": 0.425,
+      "step": 3176
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.711955960631529e-06,
+      "loss": 0.4071,
+      "step": 3177
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.703370803837933e-06,
+      "loss": 0.4525,
+      "step": 3178
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.694787628796186e-06,
+      "loss": 0.4487,
+      "step": 3179
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.68620643871771e-06,
+      "loss": 0.4075,
+      "step": 3180
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.67762723681319e-06,
+      "loss": 0.4191,
+      "step": 3181
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.669050026292558e-06,
+      "loss": 0.44,
+      "step": 3182
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.66047481036501e-06,
+      "loss": 0.4205,
+      "step": 3183
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.651901592238993e-06,
+      "loss": 0.4284,
+      "step": 3184
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.643330375122207e-06,
+      "loss": 0.4908,
+      "step": 3185
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.634761162221597e-06,
+      "loss": 0.4276,
+      "step": 3186
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.626193956743365e-06,
+      "loss": 0.4016,
+      "step": 3187
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.617628761892964e-06,
+      "loss": 0.4093,
+      "step": 3188
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.609065580875088e-06,
+      "loss": 0.4133,
+      "step": 3189
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.600504416893674e-06,
+      "loss": 0.4244,
+      "step": 3190
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.591945273151924e-06,
+      "loss": 0.4716,
+      "step": 3191
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.58338815285226e-06,
+      "loss": 0.383,
+      "step": 3192
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.574833059196362e-06,
+      "loss": 0.465,
+      "step": 3193
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.566279995385146e-06,
+      "loss": 0.4089,
+      "step": 3194
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.557728964618776e-06,
+      "loss": 0.4289,
+      "step": 3195
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.54917997009664e-06,
+      "loss": 0.3979,
+      "step": 3196
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.540633015017378e-06,
+      "loss": 0.4159,
+      "step": 3197
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.532088102578861e-06,
+      "loss": 0.424,
+      "step": 3198
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.5235452359782e-06,
+      "loss": 0.4485,
+      "step": 3199
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.515004418411735e-06,
+      "loss": 0.4364,
+      "step": 3200
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 0.4586215317249298,
+      "eval_runtime": 1632.4982,
+      "eval_samples_per_second": 0.602,
+      "eval_steps_per_second": 0.022,
+      "step": 3200
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.506465653075042e-06,
+      "loss": 0.4219,
+      "step": 3201
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.497928943162927e-06,
+      "loss": 0.4408,
+      "step": 3202
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.489394291869435e-06,
+      "loss": 0.4767,
+      "step": 3203
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.480861702387832e-06,
+      "loss": 0.4448,
+      "step": 3204
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.472331177910611e-06,
+      "loss": 0.401,
+      "step": 3205
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.4638027216295e-06,
+      "loss": 0.4482,
+      "step": 3206
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.455276336735455e-06,
+      "loss": 0.4474,
+      "step": 3207
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.446752026418644e-06,
+      "loss": 0.4127,
+      "step": 3208
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.438229793868472e-06,
+      "loss": 0.4445,
+      "step": 3209
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.429709642273563e-06,
+      "loss": 0.4621,
+      "step": 3210
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.421191574821755e-06,
+      "loss": 0.4057,
+      "step": 3211
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.412675594700113e-06,
+      "loss": 0.4308,
+      "step": 3212
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.404161705094922e-06,
+      "loss": 0.4307,
+      "step": 3213
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.395649909191684e-06,
+      "loss": 0.413,
+      "step": 3214
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.387140210175113e-06,
+      "loss": 0.4295,
+      "step": 3215
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.37863261122914e-06,
+      "loss": 0.4543,
+      "step": 3216
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.37012711553692e-06,
+      "loss": 0.4448,
+      "step": 3217
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.36162372628081e-06,
+      "loss": 0.4033,
+      "step": 3218
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.353122446642374e-06,
+      "loss": 0.4788,
+      "step": 3219
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.344623279802406e-06,
+      "loss": 0.4408,
+      "step": 3220
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.336126228940893e-06,
+      "loss": 0.4326,
+      "step": 3221
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.327631297237031e-06,
+      "loss": 0.3969,
+      "step": 3222
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.319138487869239e-06,
+      "loss": 0.4215,
+      "step": 3223
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.310647804015124e-06,
+      "loss": 0.4393,
+      "step": 3224
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.302159248851504e-06,
+      "loss": 0.4153,
+      "step": 3225
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.2936728255544e-06,
+      "loss": 0.3998,
+      "step": 3226
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.285188537299037e-06,
+      "loss": 0.4366,
+      "step": 3227
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.276706387259845e-06,
+      "loss": 0.4446,
+      "step": 3228
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.268226378610442e-06,
+      "loss": 0.4202,
+      "step": 3229
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.259748514523654e-06,
+      "loss": 0.4175,
+      "step": 3230
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.251272798171507e-06,
+      "loss": 0.4308,
+      "step": 3231
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.242799232725212e-06,
+      "loss": 0.4436,
+      "step": 3232
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.234327821355194e-06,
+      "loss": 0.4303,
+      "step": 3233
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.225858567231047e-06,
+      "loss": 0.4021,
+      "step": 3234
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.217391473521577e-06,
+      "loss": 0.4461,
+      "step": 3235
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.208926543394777e-06,
+      "loss": 0.4331,
+      "step": 3236
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.200463780017833e-06,
+      "loss": 0.4208,
+      "step": 3237
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.192003186557107e-06,
+      "loss": 0.4558,
+      "step": 3238
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.183544766178162e-06,
+      "loss": 0.4157,
+      "step": 3239
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.175088522045748e-06,
+      "loss": 0.425,
+      "step": 3240
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.1666344573238e-06,
+      "loss": 0.4213,
+      "step": 3241
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.158182575175431e-06,
+      "loss": 0.4281,
+      "step": 3242
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.149732878762943e-06,
+      "loss": 0.5155,
+      "step": 3243
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.141285371247816e-06,
+      "loss": 0.4439,
+      "step": 3244
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.132840055790717e-06,
+      "loss": 0.41,
+      "step": 3245
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.124396935551486e-06,
+      "loss": 0.405,
+      "step": 3246
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.115956013689149e-06,
+      "loss": 0.4441,
+      "step": 3247
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.107517293361903e-06,
+      "loss": 0.4146,
+      "step": 3248
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.099080777727123e-06,
+      "loss": 0.3994,
+      "step": 3249
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.090646469941362e-06,
+      "loss": 0.408,
+      "step": 3250
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.082214373160349e-06,
+      "loss": 0.4466,
+      "step": 3251
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.073784490538973e-06,
+      "loss": 0.4337,
+      "step": 3252
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.06535682523131e-06,
+      "loss": 0.4365,
+      "step": 3253
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.056931380390594e-06,
+      "loss": 0.4296,
+      "step": 3254
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.048508159169236e-06,
+      "loss": 0.4476,
+      "step": 3255
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.04008716471882e-06,
+      "loss": 0.4796,
+      "step": 3256
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.03166840019008e-06,
+      "loss": 0.3946,
+      "step": 3257
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.023251868732935e-06,
+      "loss": 0.4522,
+      "step": 3258
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.014837573496449e-06,
+      "loss": 0.433,
+      "step": 3259
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.006425517628864e-06,
+      "loss": 0.4221,
+      "step": 3260
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.998015704277577e-06,
+      "loss": 0.4328,
+      "step": 3261
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.989608136589156e-06,
+      "loss": 0.4318,
+      "step": 3262
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.98120281770931e-06,
+      "loss": 0.4099,
+      "step": 3263
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.972799750782924e-06,
+      "loss": 0.4222,
+      "step": 3264
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.964398938954033e-06,
+      "loss": 0.4297,
+      "step": 3265
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.956000385365832e-06,
+      "loss": 0.4281,
+      "step": 3266
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.947604093160661e-06,
+      "loss": 0.4212,
+      "step": 3267
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.939210065480029e-06,
+      "loss": 0.3999,
+      "step": 3268
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.930818305464586e-06,
+      "loss": 0.412,
+      "step": 3269
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.922428816254143e-06,
+      "loss": 0.4541,
+      "step": 3270
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.91404160098764e-06,
+      "loss": 0.4417,
+      "step": 3271
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.905656662803205e-06,
+      "loss": 0.4881,
+      "step": 3272
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.897274004838075e-06,
+      "loss": 0.4254,
+      "step": 3273
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.888893630228655e-06,
+      "loss": 0.4415,
+      "step": 3274
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.880515542110494e-06,
+      "loss": 0.4147,
+      "step": 3275
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.872139743618279e-06,
+      "loss": 0.4388,
+      "step": 3276
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.86376623788584e-06,
+      "loss": 0.4403,
+      "step": 3277
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.855395028046158e-06,
+      "loss": 0.3969,
+      "step": 3278
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.847026117231348e-06,
+      "loss": 0.4459,
+      "step": 3279
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.838659508572668e-06,
+      "loss": 0.421,
+      "step": 3280
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.830295205200508e-06,
+      "loss": 0.4524,
+      "step": 3281
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.821933210244406e-06,
+      "loss": 0.4033,
+      "step": 3282
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.813573526833025e-06,
+      "loss": 0.4155,
+      "step": 3283
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.805216158094177e-06,
+      "loss": 0.4216,
+      "step": 3284
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.796861107154796e-06,
+      "loss": 0.4127,
+      "step": 3285
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.78850837714095e-06,
+      "loss": 0.4943,
+      "step": 3286
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.780157971177842e-06,
+      "loss": 0.4298,
+      "step": 3287
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.771809892389809e-06,
+      "loss": 0.401,
+      "step": 3288
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.76346414390031e-06,
+      "loss": 0.4153,
+      "step": 3289
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.755120728831937e-06,
+      "loss": 0.441,
+      "step": 3290
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.746779650306406e-06,
+      "loss": 0.4397,
+      "step": 3291
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.73844091144456e-06,
+      "loss": 0.4188,
+      "step": 3292
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.730104515366372e-06,
+      "loss": 0.4411,
+      "step": 3293
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.721770465190925e-06,
+      "loss": 0.4315,
+      "step": 3294
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.713438764036432e-06,
+      "loss": 0.3859,
+      "step": 3295
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.705109415020232e-06,
+      "loss": 0.418,
+      "step": 3296
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.696782421258774e-06,
+      "loss": 0.4103,
+      "step": 3297
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.688457785867638e-06,
+      "loss": 0.4255,
+      "step": 3298
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.68013551196151e-06,
+      "loss": 0.4223,
+      "step": 3299
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.671815602654198e-06,
+      "loss": 0.4247,
+      "step": 3300
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.663498061058633e-06,
+      "loss": 0.4423,
+      "step": 3301
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.655182890286834e-06,
+      "loss": 0.4055,
+      "step": 3302
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.646870093449957e-06,
+      "loss": 0.4399,
+      "step": 3303
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.638559673658271e-06,
+      "loss": 0.4202,
+      "step": 3304
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.630251634021143e-06,
+      "loss": 0.4277,
+      "step": 3305
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.621945977647052e-06,
+      "loss": 0.4358,
+      "step": 3306
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.613642707643593e-06,
+      "loss": 0.4191,
+      "step": 3307
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.605341827117464e-06,
+      "loss": 0.4313,
+      "step": 3308
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.597043339174459e-06,
+      "loss": 0.4149,
+      "step": 3309
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.588747246919487e-06,
+      "loss": 0.4262,
+      "step": 3310
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.580453553456562e-06,
+      "loss": 0.4257,
+      "step": 3311
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.572162261888801e-06,
+      "loss": 0.3933,
+      "step": 3312
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.563873375318412e-06,
+      "loss": 0.4219,
+      "step": 3313
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.555586896846716e-06,
+      "loss": 0.4478,
+      "step": 3314
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.547302829574122e-06,
+      "loss": 0.399,
+      "step": 3315
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.539021176600152e-06,
+      "loss": 0.427,
+      "step": 3316
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.530741941023404e-06,
+      "loss": 0.4465,
+      "step": 3317
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.522465125941583e-06,
+      "loss": 0.3887,
+      "step": 3318
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.514190734451484e-06,
+      "loss": 0.4375,
+      "step": 3319
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.505918769649014e-06,
+      "loss": 0.4184,
+      "step": 3320
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.497649234629145e-06,
+      "loss": 0.4255,
+      "step": 3321
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.489382132485963e-06,
+      "loss": 0.4273,
+      "step": 3322
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.481117466312616e-06,
+      "loss": 0.459,
+      "step": 3323
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.472855239201362e-06,
+      "loss": 0.4224,
+      "step": 3324
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.464595454243549e-06,
+      "loss": 0.4286,
+      "step": 3325
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.456338114529596e-06,
+      "loss": 0.4198,
+      "step": 3326
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.448083223149017e-06,
+      "loss": 0.3882,
+      "step": 3327
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.439830783190413e-06,
+      "loss": 0.4311,
+      "step": 3328
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.431580797741455e-06,
+      "loss": 0.3998,
+      "step": 3329
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.423333269888912e-06,
+      "loss": 0.4118,
+      "step": 3330
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.415088202718612e-06,
+      "loss": 0.4171,
+      "step": 3331
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.406845599315482e-06,
+      "loss": 0.4226,
+      "step": 3332
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.39860546276352e-06,
+      "loss": 0.4475,
+      "step": 3333
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.390367796145801e-06,
+      "loss": 0.4227,
+      "step": 3334
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.38213260254447e-06,
+      "loss": 0.4096,
+      "step": 3335
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.373899885040764e-06,
+      "loss": 0.4572,
+      "step": 3336
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.365669646714983e-06,
+      "loss": 0.4375,
+      "step": 3337
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.357441890646487e-06,
+      "loss": 0.4574,
+      "step": 3338
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.349216619913726e-06,
+      "loss": 0.4246,
+      "step": 3339
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.340993837594208e-06,
+      "loss": 0.4334,
+      "step": 3340
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.332773546764521e-06,
+      "loss": 0.4073,
+      "step": 3341
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.324555750500312e-06,
+      "loss": 0.4531,
+      "step": 3342
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.316340451876297e-06,
+      "loss": 0.4043,
+      "step": 3343
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.308127653966263e-06,
+      "loss": 0.4085,
+      "step": 3344
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.299917359843055e-06,
+      "loss": 0.4676,
+      "step": 3345
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.291709572578572e-06,
+      "loss": 0.4085,
+      "step": 3346
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.283504295243795e-06,
+      "loss": 0.4402,
+      "step": 3347
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.275301530908755e-06,
+      "loss": 0.4348,
+      "step": 3348
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.267101282642542e-06,
+      "loss": 0.4351,
+      "step": 3349
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.258903553513312e-06,
+      "loss": 0.4222,
+      "step": 3350
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.250708346588261e-06,
+      "loss": 0.432,
+      "step": 3351
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.242515664933675e-06,
+      "loss": 0.4314,
+      "step": 3352
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.234325511614854e-06,
+      "loss": 0.4462,
+      "step": 3353
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.226137889696181e-06,
+      "loss": 0.4365,
+      "step": 3354
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.21795280224108e-06,
+      "loss": 0.3954,
+      "step": 3355
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.209770252312032e-06,
+      "loss": 0.4263,
+      "step": 3356
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.201590242970561e-06,
+      "loss": 0.4257,
+      "step": 3357
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.19341277727725e-06,
+      "loss": 0.4314,
+      "step": 3358
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.185237858291721e-06,
+      "loss": 0.4397,
+      "step": 3359
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.17706548907266e-06,
+      "loss": 0.4446,
+      "step": 3360
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.168895672677768e-06,
+      "loss": 0.3986,
+      "step": 3361
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.160728412163823e-06,
+      "loss": 0.4228,
+      "step": 3362
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.152563710586627e-06,
+      "loss": 0.417,
+      "step": 3363
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.144401571001033e-06,
+      "loss": 0.4379,
+      "step": 3364
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.136241996460934e-06,
+      "loss": 0.4365,
+      "step": 3365
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.128084990019262e-06,
+      "loss": 0.4288,
+      "step": 3366
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.11993055472799e-06,
+      "loss": 0.4329,
+      "step": 3367
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.111778693638124e-06,
+      "loss": 0.4307,
+      "step": 3368
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.103629409799714e-06,
+      "loss": 0.4479,
+      "step": 3369
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.095482706261843e-06,
+      "loss": 0.4437,
+      "step": 3370
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.087338586072625e-06,
+      "loss": 0.4035,
+      "step": 3371
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.079197052279213e-06,
+      "loss": 0.4252,
+      "step": 3372
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.071058107927786e-06,
+      "loss": 0.4018,
+      "step": 3373
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.062921756063568e-06,
+      "loss": 0.4312,
+      "step": 3374
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.054787999730788e-06,
+      "loss": 0.4425,
+      "step": 3375
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.046656841972727e-06,
+      "loss": 0.4423,
+      "step": 3376
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.03852828583168e-06,
+      "loss": 0.4236,
+      "step": 3377
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.03040233434898e-06,
+      "loss": 0.4516,
+      "step": 3378
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.022278990564977e-06,
+      "loss": 0.4779,
+      "step": 3379
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.014158257519046e-06,
+      "loss": 0.4429,
+      "step": 3380
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.006040138249592e-06,
+      "loss": 0.4347,
+      "step": 3381
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.997924635794032e-06,
+      "loss": 0.4077,
+      "step": 3382
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.989811753188811e-06,
+      "loss": 0.427,
+      "step": 3383
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.98170149346939e-06,
+      "loss": 0.468,
+      "step": 3384
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.973593859670252e-06,
+      "loss": 0.4215,
+      "step": 3385
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.965488854824896e-06,
+      "loss": 0.4413,
+      "step": 3386
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.957386481965838e-06,
+      "loss": 0.4352,
+      "step": 3387
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.949286744124604e-06,
+      "loss": 0.4374,
+      "step": 3388
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.941189644331747e-06,
+      "loss": 0.4266,
+      "step": 3389
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.933095185616815e-06,
+      "loss": 0.4346,
+      "step": 3390
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.925003371008379e-06,
+      "loss": 0.4253,
+      "step": 3391
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.916914203534023e-06,
+      "loss": 0.4379,
+      "step": 3392
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.908827686220334e-06,
+      "loss": 0.4271,
+      "step": 3393
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.90074382209291e-06,
+      "loss": 0.4017,
+      "step": 3394
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.892662614176357e-06,
+      "loss": 0.4539,
+      "step": 3395
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.884584065494287e-06,
+      "loss": 0.4345,
+      "step": 3396
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.876508179069323e-06,
+      "loss": 0.4326,
+      "step": 3397
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.86843495792307e-06,
+      "loss": 0.4197,
+      "step": 3398
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.860364405076155e-06,
+      "loss": 0.437,
+      "step": 3399
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.852296523548212e-06,
+      "loss": 0.3954,
+      "step": 3400
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 0.45593467354774475,
+      "eval_runtime": 1631.2328,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 3400
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.844231316357863e-06,
+      "loss": 0.4367,
+      "step": 3401
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.83616878652273e-06,
+      "loss": 0.3946,
+      "step": 3402
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.828108937059436e-06,
+      "loss": 0.4229,
+      "step": 3403
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.820051770983611e-06,
+      "loss": 0.4177,
+      "step": 3404
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.811997291309858e-06,
+      "loss": 0.4599,
+      "step": 3405
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.803945501051789e-06,
+      "loss": 0.4691,
+      "step": 3406
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.795896403222015e-06,
+      "loss": 0.507,
+      "step": 3407
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.787850000832129e-06,
+      "loss": 0.4347,
+      "step": 3408
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.779806296892721e-06,
+      "loss": 0.4427,
+      "step": 3409
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.77176529441337e-06,
+      "loss": 0.3959,
+      "step": 3410
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.76372699640265e-06,
+      "loss": 0.4363,
+      "step": 3411
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.755691405868113e-06,
+      "loss": 0.4161,
+      "step": 3412
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.7476585258163e-06,
+      "loss": 0.4183,
+      "step": 3413
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.73962835925274e-06,
+      "loss": 0.4095,
+      "step": 3414
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.731600909181949e-06,
+      "loss": 0.4328,
+      "step": 3415
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.723576178607428e-06,
+      "loss": 0.4165,
+      "step": 3416
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.715554170531653e-06,
+      "loss": 0.433,
+      "step": 3417
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.707534887956096e-06,
+      "loss": 0.4385,
+      "step": 3418
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.699518333881185e-06,
+      "loss": 0.3981,
+      "step": 3419
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.691504511306345e-06,
+      "loss": 0.429,
+      "step": 3420
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.683493423229978e-06,
+      "loss": 0.4127,
+      "step": 3421
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.675485072649457e-06,
+      "loss": 0.4198,
+      "step": 3422
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.667479462561135e-06,
+      "loss": 0.42,
+      "step": 3423
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.65947659596034e-06,
+      "loss": 0.435,
+      "step": 3424
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.651476475841369e-06,
+      "loss": 0.425,
+      "step": 3425
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.643479105197499e-06,
+      "loss": 0.4131,
+      "step": 3426
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.635484487020968e-06,
+      "loss": 0.4351,
+      "step": 3427
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.627492624302986e-06,
+      "loss": 0.4224,
+      "step": 3428
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.619503520033745e-06,
+      "loss": 0.4326,
+      "step": 3429
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.611517177202392e-06,
+      "loss": 0.4018,
+      "step": 3430
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.603533598797035e-06,
+      "loss": 0.4323,
+      "step": 3431
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.595552787804773e-06,
+      "loss": 0.4313,
+      "step": 3432
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.587574747211654e-06,
+      "loss": 0.4132,
+      "step": 3433
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.579599480002673e-06,
+      "loss": 0.4207,
+      "step": 3434
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.571626989161814e-06,
+      "loss": 0.4219,
+      "step": 3435
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.563657277672009e-06,
+      "loss": 0.4363,
+      "step": 3436
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.555690348515152e-06,
+      "loss": 0.4269,
+      "step": 3437
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.547726204672098e-06,
+      "loss": 0.4094,
+      "step": 3438
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.539764849122661e-06,
+      "loss": 0.3924,
+      "step": 3439
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.531806284845604e-06,
+      "loss": 0.4628,
+      "step": 3440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.523850514818661e-06,
+      "loss": 0.4437,
+      "step": 3441
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.515897542018496e-06,
+      "loss": 0.4012,
+      "step": 3442
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.507947369420749e-06,
+      "loss": 0.4031,
+      "step": 3443
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.500000000000004e-06,
+      "loss": 0.4293,
+      "step": 3444
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.492055436729793e-06,
+      "loss": 0.3983,
+      "step": 3445
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.4841136825826065e-06,
+      "loss": 0.3959,
+      "step": 3446
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.476174740529873e-06,
+      "loss": 0.3807,
+      "step": 3447
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.468238613541981e-06,
+      "loss": 0.3847,
+      "step": 3448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.4603053045882535e-06,
+      "loss": 0.3755,
+      "step": 3449
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.45237481663697e-06,
+      "loss": 0.3756,
+      "step": 3450
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.444447152655346e-06,
+      "loss": 0.3691,
+      "step": 3451
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.436522315609546e-06,
+      "loss": 0.3973,
+      "step": 3452
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.428600308464674e-06,
+      "loss": 0.3715,
+      "step": 3453
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.420681134184777e-06,
+      "loss": 0.4007,
+      "step": 3454
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.412764795732838e-06,
+      "loss": 0.3938,
+      "step": 3455
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.404851296070792e-06,
+      "loss": 0.4111,
+      "step": 3456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.396940638159488e-06,
+      "loss": 0.3645,
+      "step": 3457
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.3890328249587316e-06,
+      "loss": 0.375,
+      "step": 3458
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.381127859427258e-06,
+      "loss": 0.3798,
+      "step": 3459
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.3732257445227356e-06,
+      "loss": 0.3819,
+      "step": 3460
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.36532648320177e-06,
+      "loss": 0.3904,
+      "step": 3461
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.357430078419895e-06,
+      "loss": 0.4115,
+      "step": 3462
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 7.349536533131578e-06,
+      "loss": 0.3791,
+      "step": 3463
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.341645850290217e-06,
+      "loss": 0.3778,
+      "step": 3464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.333758032848135e-06,
+      "loss": 0.4028,
+      "step": 3465
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.32587308375659e-06,
+      "loss": 0.3702,
+      "step": 3466
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.317991005965758e-06,
+      "loss": 0.3732,
+      "step": 3467
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.310111802424748e-06,
+      "loss": 0.4023,
+      "step": 3468
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.302235476081589e-06,
+      "loss": 0.4135,
+      "step": 3469
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.294362029883243e-06,
+      "loss": 0.385,
+      "step": 3470
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.286491466775574e-06,
+      "loss": 0.3973,
+      "step": 3471
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.278623789703386e-06,
+      "loss": 0.3955,
+      "step": 3472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.270759001610398e-06,
+      "loss": 0.4817,
+      "step": 3473
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.262897105439246e-06,
+      "loss": 0.3879,
+      "step": 3474
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.255038104131486e-06,
+      "loss": 0.3924,
+      "step": 3475
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.2471820006275885e-06,
+      "loss": 0.4011,
+      "step": 3476
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.2393287978669455e-06,
+      "loss": 0.3978,
+      "step": 3477
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.231478498787859e-06,
+      "loss": 0.3827,
+      "step": 3478
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.223631106327534e-06,
+      "loss": 0.404,
+      "step": 3479
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 7.215786623422113e-06,
+      "loss": 0.3751,
+      "step": 3480
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.2079450530066305e-06,
+      "loss": 0.3799,
+      "step": 3481
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.200106398015039e-06,
+      "loss": 0.4018,
+      "step": 3482
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.192270661380194e-06,
+      "loss": 0.3847,
+      "step": 3483
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.184437846033866e-06,
+      "loss": 0.4234,
+      "step": 3484
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.176607954906735e-06,
+      "loss": 0.3813,
+      "step": 3485
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.168780990928368e-06,
+      "loss": 0.4112,
+      "step": 3486
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.1609569570272585e-06,
+      "loss": 0.3907,
+      "step": 3487
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.1531358561307935e-06,
+      "loss": 0.4601,
+      "step": 3488
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.145317691165267e-06,
+      "loss": 0.3767,
+      "step": 3489
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.137502465055869e-06,
+      "loss": 0.3878,
+      "step": 3490
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.129690180726694e-06,
+      "loss": 0.3757,
+      "step": 3491
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.1218808411007374e-06,
+      "loss": 0.3562,
+      "step": 3492
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.1140744490998925e-06,
+      "loss": 0.3921,
+      "step": 3493
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.1062710076449356e-06,
+      "loss": 0.3802,
+      "step": 3494
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.098470519655554e-06,
+      "loss": 0.3993,
+      "step": 3495
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.090672988050338e-06,
+      "loss": 0.3815,
+      "step": 3496
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.082878415746751e-06,
+      "loss": 0.3887,
+      "step": 3497
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 7.075086805661163e-06,
+      "loss": 0.3886,
+      "step": 3498
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.0672981607088355e-06,
+      "loss": 0.3905,
+      "step": 3499
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.059512483803904e-06,
+      "loss": 0.3644,
+      "step": 3500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.0517297778594145e-06,
+      "loss": 0.3809,
+      "step": 3501
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.043950045787291e-06,
+      "loss": 0.3895,
+      "step": 3502
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.036173290498346e-06,
+      "loss": 0.3757,
+      "step": 3503
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.0283995149022816e-06,
+      "loss": 0.3916,
+      "step": 3504
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.020628721907679e-06,
+      "loss": 0.3879,
+      "step": 3505
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.012860914422009e-06,
+      "loss": 0.3833,
+      "step": 3506
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 7.005096095351624e-06,
+      "loss": 0.3626,
+      "step": 3507
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.9973342676017625e-06,
+      "loss": 0.3923,
+      "step": 3508
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.989575434076525e-06,
+      "loss": 0.3968,
+      "step": 3509
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.981819597678914e-06,
+      "loss": 0.3587,
+      "step": 3510
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.974066761310796e-06,
+      "loss": 0.386,
+      "step": 3511
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.9663169278729325e-06,
+      "loss": 0.3626,
+      "step": 3512
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.958570100264944e-06,
+      "loss": 0.3921,
+      "step": 3513
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.950826281385338e-06,
+      "loss": 0.3788,
+      "step": 3514
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 6.943085474131477e-06,
+      "loss": 0.3983,
+      "step": 3515
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.935347681399622e-06,
+      "loss": 0.3896,
+      "step": 3516
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.927612906084888e-06,
+      "loss": 0.4057,
+      "step": 3517
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.919881151081272e-06,
+      "loss": 0.4107,
+      "step": 3518
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.912152419281636e-06,
+      "loss": 0.4083,
+      "step": 3519
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.904426713577709e-06,
+      "loss": 0.3753,
+      "step": 3520
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.896704036860091e-06,
+      "loss": 0.3714,
+      "step": 3521
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.888984392018254e-06,
+      "loss": 0.4431,
+      "step": 3522
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.881267781940519e-06,
+      "loss": 0.3622,
+      "step": 3523
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.873554209514085e-06,
+      "loss": 0.3741,
+      "step": 3524
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.865843677625014e-06,
+      "loss": 0.3565,
+      "step": 3525
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.858136189158227e-06,
+      "loss": 0.3587,
+      "step": 3526
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.8504317469974985e-06,
+      "loss": 0.4032,
+      "step": 3527
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.842730354025488e-06,
+      "loss": 0.4623,
+      "step": 3528
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.835032013123695e-06,
+      "loss": 0.3816,
+      "step": 3529
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.827336727172469e-06,
+      "loss": 0.3886,
+      "step": 3530
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.819644499051033e-06,
+      "loss": 0.3874,
+      "step": 3531
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 6.811955331637463e-06,
+      "loss": 0.3805,
+      "step": 3532
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.804269227808682e-06,
+      "loss": 0.3735,
+      "step": 3533
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.7965861904404744e-06,
+      "loss": 0.3847,
+      "step": 3534
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.788906222407477e-06,
+      "loss": 0.394,
+      "step": 3535
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.781229326583172e-06,
+      "loss": 0.4161,
+      "step": 3536
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.7735555058399065e-06,
+      "loss": 0.366,
+      "step": 3537
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.76588476304885e-06,
+      "loss": 0.3843,
+      "step": 3538
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.758217101080047e-06,
+      "loss": 0.4121,
+      "step": 3539
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.750552522802377e-06,
+      "loss": 0.4126,
+      "step": 3540
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.742891031083569e-06,
+      "loss": 0.3698,
+      "step": 3541
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.735232628790196e-06,
+      "loss": 0.3741,
+      "step": 3542
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.727577318787676e-06,
+      "loss": 0.3772,
+      "step": 3543
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.719925103940267e-06,
+      "loss": 0.4146,
+      "step": 3544
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.7122759871110735e-06,
+      "loss": 0.3963,
+      "step": 3545
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.704629971162039e-06,
+      "loss": 0.3859,
+      "step": 3546
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.696987058953944e-06,
+      "loss": 0.3656,
+      "step": 3547
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.689347253346413e-06,
+      "loss": 0.3549,
+      "step": 3548
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 6.681710557197901e-06,
+      "loss": 0.4073,
+      "step": 3549
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.674076973365709e-06,
+      "loss": 0.3893,
+      "step": 3550
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.666446504705971e-06,
+      "loss": 0.3694,
+      "step": 3551
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.658819154073642e-06,
+      "loss": 0.4029,
+      "step": 3552
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.651194924322527e-06,
+      "loss": 0.4131,
+      "step": 3553
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.643573818305258e-06,
+      "loss": 0.3772,
+      "step": 3554
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.635955838873297e-06,
+      "loss": 0.3831,
+      "step": 3555
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.6283409888769355e-06,
+      "loss": 0.3763,
+      "step": 3556
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.6207292711653e-06,
+      "loss": 0.4013,
+      "step": 3557
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.613120688586336e-06,
+      "loss": 0.3996,
+      "step": 3558
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.605515243986823e-06,
+      "loss": 0.3743,
+      "step": 3559
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.597912940212364e-06,
+      "loss": 0.3572,
+      "step": 3560
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.590313780107386e-06,
+      "loss": 0.3964,
+      "step": 3561
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.582717766515142e-06,
+      "loss": 0.3597,
+      "step": 3562
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.575124902277705e-06,
+      "loss": 0.36,
+      "step": 3563
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.56753519023597e-06,
+      "loss": 0.39,
+      "step": 3564
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.5599486332296546e-06,
+      "loss": 0.3641,
+      "step": 3565
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 6.552365234097302e-06,
+      "loss": 0.3921,
+      "step": 3566
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.544784995676254e-06,
+      "loss": 0.3819,
+      "step": 3567
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.537207920802688e-06,
+      "loss": 0.3767,
+      "step": 3568
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.529634012311595e-06,
+      "loss": 0.3378,
+      "step": 3569
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.522063273036775e-06,
+      "loss": 0.3965,
+      "step": 3570
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.514495705810848e-06,
+      "loss": 0.4016,
+      "step": 3571
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.506931313465245e-06,
+      "loss": 0.4001,
+      "step": 3572
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.499370098830208e-06,
+      "loss": 0.3866,
+      "step": 3573
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.491812064734798e-06,
+      "loss": 0.3429,
+      "step": 3574
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.484257214006868e-06,
+      "loss": 0.3721,
+      "step": 3575
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.476705549473089e-06,
+      "loss": 0.3835,
+      "step": 3576
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.469157073958957e-06,
+      "loss": 0.4137,
+      "step": 3577
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.461611790288751e-06,
+      "loss": 0.4142,
+      "step": 3578
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.454069701285565e-06,
+      "loss": 0.3897,
+      "step": 3579
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.446530809771299e-06,
+      "loss": 0.4277,
+      "step": 3580
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.438995118566658e-06,
+      "loss": 0.3797,
+      "step": 3581
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.431462630491137e-06,
+      "loss": 0.3683,
+      "step": 3582
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.423933348363049e-06,
+      "loss": 0.4109,
+      "step": 3583
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 6.416407274999498e-06,
+      "loss": 0.41,
+      "step": 3584
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.408884413216391e-06,
+      "loss": 0.3913,
+      "step": 3585
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.401364765828433e-06,
+      "loss": 0.3719,
+      "step": 3586
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.393848335649124e-06,
+      "loss": 0.3724,
+      "step": 3587
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.3863351254907645e-06,
+      "loss": 0.3671,
+      "step": 3588
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.378825138164449e-06,
+      "loss": 0.4044,
+      "step": 3589
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.371318376480058e-06,
+      "loss": 0.3915,
+      "step": 3590
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.363814843246276e-06,
+      "loss": 0.383,
+      "step": 3591
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.356314541270569e-06,
+      "loss": 0.3632,
+      "step": 3592
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.348817473359213e-06,
+      "loss": 0.3684,
+      "step": 3593
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.341323642317255e-06,
+      "loss": 0.4015,
+      "step": 3594
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.3338330509485436e-06,
+      "loss": 0.3915,
+      "step": 3595
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.326345702055698e-06,
+      "loss": 0.4091,
+      "step": 3596
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.318861598440139e-06,
+      "loss": 0.3875,
+      "step": 3597
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.311380742902072e-06,
+      "loss": 0.4245,
+      "step": 3598
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.303903138240485e-06,
+      "loss": 0.3857,
+      "step": 3599
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 6.296428787253147e-06,
+      "loss": 0.3894,
+      "step": 3600
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 0.4590950608253479,
+      "eval_runtime": 1630.5433,
+      "eval_samples_per_second": 0.603,
+      "eval_steps_per_second": 0.022,
+      "step": 3600
+    }
+  ],
+  "max_steps": 5157,
+  "num_train_epochs": 3,
+  "total_flos": 1.5371932973825786e+19,
+  "trial_name": null,
+  "trial_params": null
+}