End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "train_loss": 3.0590674640896083,
-    "train_runtime": 25.7247,
+    "train_loss": 1.781525854233865,
+    "train_runtime": 119.0355,
     "train_samples": 3527,
-    "train_samples_per_second": 411.317,
-    "train_steps_per_second": 12.945
+    "train_samples_per_second": 88.889,
+    "train_steps_per_second": 2.797
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
-    "train_loss": 3.0590674640896083,
-    "train_runtime": 25.7247,
+    "train_loss": 1.781525854233865,
+    "train_runtime": 119.0355,
     "train_samples": 3527,
-    "train_samples_per_second": 411.317,
-    "train_steps_per_second": 12.945
+    "train_samples_per_second": 88.889,
+    "train_steps_per_second": 2.797
 }

trainer_state.json

@@ -10,2014 +10,2014 @@
     {
       "epoch": 0.01,
       "learning_rate": 1.993993993993994e-05,
-      "loss": 7.239,
+      "loss": 9.3363,
       "step": 1
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.987987987987988e-05,
-      "loss": 6.4074,
+      "loss": 6.3834,
       "step": 2
     },
     {
       "epoch": 0.03,
       "learning_rate": 1.981981981981982e-05,
-      "loss": 6.4224,
+      "loss": 5.1305,
       "step": 3
     },
     {
       "epoch": 0.04,
       "learning_rate": 1.9759759759759763e-05,
-      "loss": 5.6383,
+      "loss": 4.4537,
       "step": 4
     },
     {
       "epoch": 0.05,
       "learning_rate": 1.9699699699699702e-05,
-      "loss": 5.4807,
+      "loss": 4.0593,
       "step": 5
     },
     {
       "epoch": 0.05,
       "learning_rate": 1.963963963963964e-05,
-      "loss": 5.4554,
+      "loss": 3.8647,
       "step": 6
     },
     {
       "epoch": 0.06,
       "learning_rate": 1.957957957957958e-05,
-      "loss": 5.2014,
+      "loss": 3.5701,
       "step": 7
     },
     {
       "epoch": 0.07,
       "learning_rate": 1.951951951951952e-05,
-      "loss": 5.3549,
+      "loss": 3.594,
       "step": 8
     },
     {
       "epoch": 0.08,
       "learning_rate": 1.9459459459459463e-05,
-      "loss": 5.0557,
+      "loss": 3.5468,
       "step": 9
     },
     {
       "epoch": 0.09,
       "learning_rate": 1.9399399399399402e-05,
-      "loss": 4.7239,
+      "loss": 3.2149,
       "step": 10
     },
     {
       "epoch": 0.1,
       "learning_rate": 1.9339339339339342e-05,
-      "loss": 4.5006,
+      "loss": 3.2163,
       "step": 11
     },
     {
       "epoch": 0.11,
       "learning_rate": 1.927927927927928e-05,
-      "loss": 4.6071,
+      "loss": 3.0711,
       "step": 12
     },
     {
       "epoch": 0.12,
       "learning_rate": 1.921921921921922e-05,
-      "loss": 4.6655,
+      "loss": 2.9672,
       "step": 13
     },
     {
       "epoch": 0.13,
       "learning_rate": 1.915915915915916e-05,
-      "loss": 4.3521,
+      "loss": 2.9493,
       "step": 14
     },
     {
       "epoch": 0.14,
       "learning_rate": 1.90990990990991e-05,
-      "loss": 4.8466,
+      "loss": 3.1754,
       "step": 15
     },
     {
       "epoch": 0.14,
       "learning_rate": 1.903903903903904e-05,
-      "loss": 4.5095,
+      "loss": 2.8729,
       "step": 16
     },
     {
       "epoch": 0.15,
       "learning_rate": 1.8978978978978982e-05,
-      "loss": 4.4407,
+      "loss": 2.8177,
       "step": 17
     },
     {
       "epoch": 0.16,
       "learning_rate": 1.891891891891892e-05,
-      "loss": 4.4312,
+      "loss": 2.7751,
       "step": 18
     },
     {
       "epoch": 0.17,
       "learning_rate": 1.885885885885886e-05,
-      "loss": 4.314,
+      "loss": 2.7206,
       "step": 19
     },
     {
       "epoch": 0.18,
       "learning_rate": 1.87987987987988e-05,
-      "loss": 4.274,
+      "loss": 2.6308,
       "step": 20
     },
     {
       "epoch": 0.19,
       "learning_rate": 1.873873873873874e-05,
-      "loss": 4.3241,
+      "loss": 2.6611,
       "step": 21
     },
     {
       "epoch": 0.2,
       "learning_rate": 1.8678678678678682e-05,
-      "loss": 4.3871,
+      "loss": 2.6,
       "step": 22
     },
     {
       "epoch": 0.21,
       "learning_rate": 1.861861861861862e-05,
-      "loss": 4.0151,
+      "loss": 2.3343,
       "step": 23
     },
     {
       "epoch": 0.22,
       "learning_rate": 1.855855855855856e-05,
-      "loss": 4.3852,
+      "loss": 2.5656,
       "step": 24
     },
     {
       "epoch": 0.23,
       "learning_rate": 1.84984984984985e-05,
-      "loss": 4.2063,
+      "loss": 2.6288,
       "step": 25
     },
     {
       "epoch": 0.23,
       "learning_rate": 1.843843843843844e-05,
-      "loss": 4.327,
+      "loss": 2.6641,
       "step": 26
     },
     {
       "epoch": 0.24,
       "learning_rate": 1.8378378378378383e-05,
-      "loss": 3.8483,
+      "loss": 2.2362,
       "step": 27
     },
     {
       "epoch": 0.25,
       "learning_rate": 1.831831831831832e-05,
-      "loss": 4.0045,
+      "loss": 2.2673,
       "step": 28
     },
     {
       "epoch": 0.26,
       "learning_rate": 1.8258258258258258e-05,
-      "loss": 4.1035,
+      "loss": 2.5034,
       "step": 29
     },
     {
       "epoch": 0.27,
       "learning_rate": 1.81981981981982e-05,
-      "loss": 4.0994,
+      "loss": 2.4348,
       "step": 30
     },
     {
       "epoch": 0.28,
       "learning_rate": 1.813813813813814e-05,
-      "loss": 3.8518,
+      "loss": 2.3081,
       "step": 31
     },
     {
       "epoch": 0.29,
       "learning_rate": 1.807807807807808e-05,
-      "loss": 3.8938,
+      "loss": 2.2955,
       "step": 32
     },
     {
       "epoch": 0.3,
       "learning_rate": 1.801801801801802e-05,
-      "loss": 3.7891,
+      "loss": 2.2767,
       "step": 33
     },
     {
       "epoch": 0.31,
       "learning_rate": 1.795795795795796e-05,
-      "loss": 3.978,
+      "loss": 2.3092,
       "step": 34
     },
     {
       "epoch": 0.32,
       "learning_rate": 1.78978978978979e-05,
-      "loss": 3.9747,
+      "loss": 2.3261,
       "step": 35
     },
     {
       "epoch": 0.32,
       "learning_rate": 1.783783783783784e-05,
-      "loss": 3.8688,
+      "loss": 2.4289,
       "step": 36
     },
     {
       "epoch": 0.33,
       "learning_rate": 1.7777777777777777e-05,
-      "loss": 3.8895,
+      "loss": 2.4217,
       "step": 37
     },
     {
       "epoch": 0.34,
       "learning_rate": 1.771771771771772e-05,
-      "loss": 3.7801,
+      "loss": 2.3543,
       "step": 38
     },
     {
       "epoch": 0.35,
       "learning_rate": 1.765765765765766e-05,
-      "loss": 3.7968,
+      "loss": 2.2967,
       "step": 39
     },
     {
       "epoch": 0.36,
       "learning_rate": 1.7597597597597598e-05,
-      "loss": 3.8245,
+      "loss": 2.3744,
       "step": 40
     },
     {
       "epoch": 0.37,
       "learning_rate": 1.7537537537537538e-05,
-      "loss": 3.7415,
+      "loss": 2.2404,
       "step": 41
     },
     {
       "epoch": 0.38,
       "learning_rate": 1.7477477477477477e-05,
-      "loss": 3.5251,
+      "loss": 2.0521,
       "step": 42
     },
     {
       "epoch": 0.39,
       "learning_rate": 1.741741741741742e-05,
-      "loss": 3.667,
+      "loss": 2.1582,
       "step": 43
     },
     {
       "epoch": 0.4,
       "learning_rate": 1.735735735735736e-05,
-      "loss": 3.6625,
+      "loss": 2.2699,
       "step": 44
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.72972972972973e-05,
-      "loss": 3.675,
+      "loss": 2.1781,
       "step": 45
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.7237237237237238e-05,
-      "loss": 3.8588,
+      "loss": 2.2529,
       "step": 46
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.7177177177177177e-05,
-      "loss": 3.5836,
+      "loss": 2.0092,
       "step": 47
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.711711711711712e-05,
-      "loss": 3.7319,
+      "loss": 2.2386,
       "step": 48
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.705705705705706e-05,
-      "loss": 3.5722,
+      "loss": 2.1246,
       "step": 49
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.6996996996997e-05,
-      "loss": 3.573,
+      "loss": 2.0728,
       "step": 50
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.693693693693694e-05,
-      "loss": 3.5678,
+      "loss": 2.1956,
       "step": 51
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.6876876876876878e-05,
-      "loss": 3.6492,
+      "loss": 2.1301,
       "step": 52
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6816816816816817e-05,
-      "loss": 3.6264,
+      "loss": 2.2538,
       "step": 53
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6756756756756757e-05,
-      "loss": 3.5554,
+      "loss": 2.0491,
       "step": 54
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.6696696696696696e-05,
-      "loss": 3.4947,
+      "loss": 2.143,
       "step": 55
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.663663663663664e-05,
-      "loss": 3.2952,
+      "loss": 2.0769,
       "step": 56
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.6576576576576578e-05,
-      "loss": 3.4397,
+      "loss": 2.0398,
       "step": 57
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.6516516516516518e-05,
-      "loss": 3.6218,
+      "loss": 2.1762,
       "step": 58
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.6456456456456457e-05,
-      "loss": 3.6569,
+      "loss": 2.1344,
       "step": 59
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.6396396396396396e-05,
-      "loss": 3.2034,
+      "loss": 1.8858,
       "step": 60
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.633633633633634e-05,
-      "loss": 3.3751,
+      "loss": 2.0582,
       "step": 61
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.627627627627628e-05,
-      "loss": 3.4071,
+      "loss": 2.0495,
       "step": 62
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.6216216216216218e-05,
-      "loss": 3.3431,
+      "loss": 2.0432,
       "step": 63
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.6156156156156157e-05,
-      "loss": 3.2513,
+      "loss": 2.0205,
       "step": 64
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.6096096096096097e-05,
-      "loss": 3.4148,
+      "loss": 2.0229,
       "step": 65
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.6036036036036036e-05,
-      "loss": 3.3335,
+      "loss": 1.9601,
       "step": 66
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.5975975975975976e-05,
-      "loss": 3.3857,
+      "loss": 2.0695,
       "step": 67
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.591591591591592e-05,
-      "loss": 3.1585,
+      "loss": 1.9346,
       "step": 68
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.5855855855855858e-05,
-      "loss": 3.4215,
+      "loss": 2.11,
       "step": 69
     },
     {
       "epoch": 0.63,
       "learning_rate": 1.5795795795795797e-05,
-      "loss": 3.3944,
+      "loss": 2.0759,
       "step": 70
     },
     {
       "epoch": 0.64,
       "learning_rate": 1.5735735735735737e-05,
-      "loss": 3.1508,
+      "loss": 1.9352,
       "step": 71
     },
     {
       "epoch": 0.65,
       "learning_rate": 1.5675675675675676e-05,
-      "loss": 3.47,
+      "loss": 2.1723,
       "step": 72
     },
     {
       "epoch": 0.66,
       "learning_rate": 1.5615615615615616e-05,
-      "loss": 3.1349,
+      "loss": 2.0719,
       "step": 73
     },
     {
       "epoch": 0.67,
       "learning_rate": 1.555555555555556e-05,
-      "loss": 3.0709,
+      "loss": 1.9658,
       "step": 74
     },
     {
       "epoch": 0.68,
       "learning_rate": 1.5495495495495498e-05,
-      "loss": 3.0129,
+      "loss": 1.7428,
       "step": 75
     },
     {
       "epoch": 0.68,
       "learning_rate": 1.5435435435435437e-05,
-      "loss": 3.0602,
+      "loss": 1.8585,
       "step": 76
     },
     {
       "epoch": 0.69,
       "learning_rate": 1.5375375375375377e-05,
-      "loss": 3.1955,
+      "loss": 1.9221,
       "step": 77
     },
     {
       "epoch": 0.7,
       "learning_rate": 1.5315315315315316e-05,
-      "loss": 3.0284,
+      "loss": 1.9126,
       "step": 78
     },
     {
       "epoch": 0.71,
       "learning_rate": 1.5255255255255257e-05,
-      "loss": 3.0934,
+      "loss": 1.8657,
       "step": 79
     },
     {
       "epoch": 0.72,
       "learning_rate": 1.5195195195195196e-05,
-      "loss": 3.1247,
+      "loss": 1.9183,
       "step": 80
     },
     {
       "epoch": 0.73,
       "learning_rate": 1.5135135135135138e-05,
-      "loss": 3.2246,
+      "loss": 1.983,
       "step": 81
     },
     {
       "epoch": 0.74,
       "learning_rate": 1.5075075075075077e-05,
-      "loss": 3.1494,
+      "loss": 1.9661,
       "step": 82
     },
     {
       "epoch": 0.75,
       "learning_rate": 1.5015015015015015e-05,
-      "loss": 3.2155,
+      "loss": 1.9937,
       "step": 83
     },
     {
       "epoch": 0.76,
       "learning_rate": 1.4954954954954957e-05,
-      "loss": 3.0095,
+      "loss": 1.8473,
       "step": 84
     },
     {
       "epoch": 0.77,
       "learning_rate": 1.4894894894894895e-05,
-      "loss": 3.0313,
+      "loss": 1.9417,
       "step": 85
     },
     {
       "epoch": 0.77,
       "learning_rate": 1.4834834834834836e-05,
-      "loss": 3.2005,
+      "loss": 1.8969,
       "step": 86
     },
     {
       "epoch": 0.78,
       "learning_rate": 1.4774774774774776e-05,
-      "loss": 3.1127,
+      "loss": 1.8054,
       "step": 87
     },
     {
       "epoch": 0.79,
       "learning_rate": 1.4714714714714715e-05,
-      "loss": 3.1082,
+      "loss": 1.7336,
       "step": 88
     },
     {
       "epoch": 0.8,
       "learning_rate": 1.4654654654654656e-05,
-      "loss": 3.0101,
+      "loss": 1.8264,
       "step": 89
     },
     {
       "epoch": 0.81,
       "learning_rate": 1.4594594594594596e-05,
-      "loss": 3.0918,
+      "loss": 1.8013,
       "step": 90
     },
     {
       "epoch": 0.82,
       "learning_rate": 1.4534534534534537e-05,
-      "loss": 2.9955,
+      "loss": 1.9437,
       "step": 91
     },
     {
       "epoch": 0.83,
       "learning_rate": 1.4474474474474476e-05,
-      "loss": 3.0981,
+      "loss": 1.893,
       "step": 92
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.4414414414414416e-05,
-      "loss": 3.1136,
+      "loss": 1.83,
       "step": 93
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.4354354354354357e-05,
-      "loss": 3.2654,
+      "loss": 1.966,
       "step": 94
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4294294294294296e-05,
-      "loss": 3.0342,
+      "loss": 1.9087,
       "step": 95
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4234234234234234e-05,
-      "loss": 2.9786,
+      "loss": 1.8374,
       "step": 96
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.4174174174174175e-05,
-      "loss": 3.1309,
+      "loss": 2.1122,
       "step": 97
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.4114114114114114e-05,
-      "loss": 3.178,
+      "loss": 2.0771,
       "step": 98
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.4054054054054055e-05,
-      "loss": 2.9293,
+      "loss": 1.8904,
       "step": 99
     },
     {
       "epoch": 0.9,
       "learning_rate": 1.3993993993993995e-05,
-      "loss": 3.1669,
+      "loss": 1.9853,
       "step": 100
     },
     {
       "epoch": 0.91,
       "learning_rate": 1.3933933933933934e-05,
-      "loss": 2.8725,
+      "loss": 1.8186,
       "step": 101
     },
     {
       "epoch": 0.92,
       "learning_rate": 1.3873873873873875e-05,
-      "loss": 3.0613,
+      "loss": 1.8105,
       "step": 102
     },
     {
       "epoch": 0.93,
       "learning_rate": 1.3813813813813815e-05,
-      "loss": 2.8206,
+      "loss": 1.6838,
       "step": 103
     },
     {
       "epoch": 0.94,
       "learning_rate": 1.3753753753753756e-05,
-      "loss": 2.9075,
+      "loss": 1.7476,
       "step": 104
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.3693693693693695e-05,
-      "loss": 2.9944,
+      "loss": 1.8371,
       "step": 105
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.3633633633633635e-05,
-      "loss": 2.955,
+      "loss": 1.8751,
       "step": 106
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.3573573573573576e-05,
-      "loss": 3.0356,
+      "loss": 1.7548,
       "step": 107
     },
     {
       "epoch": 0.97,
       "learning_rate": 1.3513513513513515e-05,
-      "loss": 2.7716,
+      "loss": 1.6747,
       "step": 108
     },
     {
       "epoch": 0.98,
       "learning_rate": 1.3453453453453456e-05,
-      "loss": 2.8735,
+      "loss": 1.7632,
       "step": 109
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.3393393393393394e-05,
-      "loss": 2.9966,
+      "loss": 1.8086,
       "step": 110
     },
     {
       "epoch": 1.0,
       "learning_rate": 1.3333333333333333e-05,
-      "loss": 3.0701,
+      "loss": 2.2582,
       "step": 111
     },
     {
       "epoch": 1.01,
       "learning_rate": 1.3273273273273274e-05,
-      "loss": 2.8179,
+      "loss": 1.6199,
       "step": 112
     },
     {
       "epoch": 1.02,
       "learning_rate": 1.3213213213213214e-05,
-      "loss": 2.8501,
+      "loss": 1.6633,
       "step": 113
     },
     {
       "epoch": 1.03,
       "learning_rate": 1.3153153153153155e-05,
-      "loss": 2.7673,
+      "loss": 1.6855,
       "step": 114
     },
     {
       "epoch": 1.04,
       "learning_rate": 1.3093093093093094e-05,
-      "loss": 2.9732,
+      "loss": 1.7482,
       "step": 115
     },
     {
       "epoch": 1.05,
       "learning_rate": 1.3033033033033034e-05,
-      "loss": 2.8297,
+      "loss": 1.6648,
       "step": 116
     },
     {
       "epoch": 1.05,
       "learning_rate": 1.2972972972972975e-05,
-      "loss": 2.7962,
+      "loss": 1.7411,
       "step": 117
     },
     {
       "epoch": 1.06,
       "learning_rate": 1.2912912912912914e-05,
-      "loss": 3.0255,
+      "loss": 1.7804,
       "step": 118
     },
     {
       "epoch": 1.07,
       "learning_rate": 1.2852852852852854e-05,
-      "loss": 2.9122,
+      "loss": 1.6331,
       "step": 119
     },
     {
       "epoch": 1.08,
       "learning_rate": 1.2792792792792795e-05,
-      "loss": 2.9703,
+      "loss": 1.7026,
       "step": 120
     },
     {
       "epoch": 1.09,
       "learning_rate": 1.2732732732732732e-05,
-      "loss": 2.7686,
+      "loss": 1.5627,
       "step": 121
     },
     {
       "epoch": 1.1,
       "learning_rate": 1.2672672672672675e-05,
-      "loss": 2.9087,
+      "loss": 1.7268,
       "step": 122
     },
     {
       "epoch": 1.11,
       "learning_rate": 1.2612612612612613e-05,
-      "loss": 3.0105,
+      "loss": 1.7191,
       "step": 123
     },
     {
       "epoch": 1.12,
       "learning_rate": 1.2552552552552552e-05,
-      "loss": 2.9715,
+      "loss": 1.7727,
       "step": 124
     },
     {
       "epoch": 1.13,
       "learning_rate": 1.2492492492492493e-05,
-      "loss": 2.9058,
+      "loss": 1.6784,
       "step": 125
     },
     {
       "epoch": 1.14,
       "learning_rate": 1.2432432432432433e-05,
-      "loss": 2.7668,
+      "loss": 1.7173,
       "step": 126
     },
     {
       "epoch": 1.14,
       "learning_rate": 1.2372372372372374e-05,
-      "loss": 3.0009,
+      "loss": 1.7755,
       "step": 127
     },
     {
       "epoch": 1.15,
       "learning_rate": 1.2312312312312313e-05,
-      "loss": 2.7362,
+      "loss": 1.5832,
       "step": 128
     },
     {
       "epoch": 1.16,
       "learning_rate": 1.2252252252252253e-05,
-      "loss": 2.8804,
+      "loss": 1.6627,
       "step": 129
     },
     {
       "epoch": 1.17,
       "learning_rate": 1.2192192192192194e-05,
-      "loss": 2.8841,
+      "loss": 1.7832,
       "step": 130
     },
     {
       "epoch": 1.18,
       "learning_rate": 1.2132132132132133e-05,
-      "loss": 2.8075,
+      "loss": 1.6711,
       "step": 131
     },
     {
       "epoch": 1.19,
       "learning_rate": 1.2072072072072074e-05,
-      "loss": 2.7388,
+      "loss": 1.6217,
       "step": 132
     },
     {
       "epoch": 1.2,
       "learning_rate": 1.2012012012012014e-05,
-      "loss": 2.8681,
+      "loss": 1.6746,
       "step": 133
     },
     {
       "epoch": 1.21,
       "learning_rate": 1.1951951951951951e-05,
-      "loss": 2.7851,
+      "loss": 1.602,
       "step": 134
     },
     {
       "epoch": 1.22,
       "learning_rate": 1.1891891891891894e-05,
-      "loss": 2.9474,
+      "loss": 1.7648,
       "step": 135
     },
     {
       "epoch": 1.23,
       "learning_rate": 1.1831831831831832e-05,
-      "loss": 2.7493,
+      "loss": 1.5384,
       "step": 136
     },
     {
       "epoch": 1.23,
       "learning_rate": 1.1771771771771771e-05,
-      "loss": 2.8871,
+      "loss": 1.6899,
       "step": 137
     },
     {
       "epoch": 1.24,
       "learning_rate": 1.1711711711711713e-05,
-      "loss": 2.9732,
+      "loss": 1.7466,
       "step": 138
     },
     {
       "epoch": 1.25,
       "learning_rate": 1.1651651651651652e-05,
-      "loss": 2.9281,
+      "loss": 1.5775,
       "step": 139
     },
     {
       "epoch": 1.26,
       "learning_rate": 1.1591591591591593e-05,
-      "loss": 2.8014,
+      "loss": 1.5667,
       "step": 140
     },
     {
       "epoch": 1.27,
       "learning_rate": 1.1531531531531532e-05,
-      "loss": 2.8444,
+      "loss": 1.6807,
       "step": 141
     },
     {
       "epoch": 1.28,
       "learning_rate": 1.1471471471471472e-05,
-      "loss": 2.994,
+      "loss": 1.7766,
       "step": 142
     },
     {
       "epoch": 1.29,
       "learning_rate": 1.1411411411411413e-05,
-      "loss": 2.9153,
+      "loss": 1.7,
       "step": 143
     },
     {
       "epoch": 1.3,
       "learning_rate": 1.1351351351351352e-05,
-      "loss": 2.7083,
+      "loss": 1.5558,
       "step": 144
     },
     {
       "epoch": 1.31,
       "learning_rate": 1.1291291291291293e-05,
-      "loss": 2.8247,
+      "loss": 1.7148,
       "step": 145
     },
     {
       "epoch": 1.32,
       "learning_rate": 1.1231231231231233e-05,
-      "loss": 2.6741,
+      "loss": 1.5432,
       "step": 146
     },
     {
       "epoch": 1.32,
       "learning_rate": 1.117117117117117e-05,
-      "loss": 2.8198,
+      "loss": 1.5042,
       "step": 147
     },
     {
       "epoch": 1.33,
       "learning_rate": 1.1111111111111113e-05,
-      "loss": 2.9513,
+      "loss": 1.8052,
       "step": 148
     },
     {
       "epoch": 1.34,
       "learning_rate": 1.1051051051051051e-05,
-      "loss": 2.9997,
+      "loss": 1.8016,
       "step": 149
     },
     {
       "epoch": 1.35,
       "learning_rate": 1.0990990990990992e-05,
-      "loss": 2.6673,
+      "loss": 1.4395,
       "step": 150
     },
     {
       "epoch": 1.36,
       "learning_rate": 1.0930930930930932e-05,
-      "loss": 2.8524,
+      "loss": 1.6007,
       "step": 151
     },
     {
       "epoch": 1.37,
       "learning_rate": 1.0870870870870871e-05,
-      "loss": 2.8598,
+      "loss": 1.6131,
       "step": 152
     },
     {
       "epoch": 1.38,
       "learning_rate": 1.0810810810810812e-05,
-      "loss": 2.8718,
+      "loss": 1.612,
       "step": 153
     },
     {
       "epoch": 1.39,
       "learning_rate": 1.0750750750750751e-05,
-      "loss": 2.782,
+      "loss": 1.5766,
       "step": 154
     },
     {
       "epoch": 1.4,
       "learning_rate": 1.0690690690690693e-05,
-      "loss": 2.8597,
+      "loss": 1.577,
       "step": 155
     },
     {
       "epoch": 1.41,
       "learning_rate": 1.0630630630630632e-05,
-      "loss": 2.842,
+      "loss": 1.6121,
       "step": 156
     },
     {
       "epoch": 1.41,
       "learning_rate": 1.0570570570570571e-05,
-      "loss": 2.8293,
+      "loss": 1.7291,
       "step": 157
     },
     {
       "epoch": 1.42,
       "learning_rate": 1.0510510510510512e-05,
-      "loss": 2.7173,
+      "loss": 1.5699,
       "step": 158
     },
     {
       "epoch": 1.43,
       "learning_rate": 1.0450450450450452e-05,
-      "loss": 2.723,
+      "loss": 1.564,
       "step": 159
     },
     {
       "epoch": 1.44,
       "learning_rate": 1.039039039039039e-05,
-      "loss": 2.8006,
+      "loss": 1.5114,
       "step": 160
     },
     {
       "epoch": 1.45,
       "learning_rate": 1.033033033033033e-05,
-      "loss": 2.75,
+      "loss": 1.5204,
       "step": 161
     },
     {
       "epoch": 1.46,
       "learning_rate": 1.027027027027027e-05,
-      "loss": 2.5134,
+      "loss": 1.5483,
       "step": 162
     },
     {
       "epoch": 1.47,
       "learning_rate": 1.0210210210210211e-05,
-      "loss": 2.8993,
+      "loss": 1.7533,
       "step": 163
     },
     {
       "epoch": 1.48,
       "learning_rate": 1.015015015015015e-05,
-      "loss": 2.8422,
+      "loss": 1.7088,
       "step": 164
     },
     {
       "epoch": 1.49,
       "learning_rate": 1.009009009009009e-05,
-      "loss": 2.513,
+      "loss": 1.3951,
       "step": 165
     },
     {
       "epoch": 1.5,
       "learning_rate": 1.0030030030030031e-05,
-      "loss": 2.7341,
+      "loss": 1.533,
       "step": 166
     },
     {
       "epoch": 1.5,
       "learning_rate": 9.96996996996997e-06,
-      "loss": 2.9636,
+      "loss": 1.6597,
       "step": 167
     },
     {
       "epoch": 1.51,
       "learning_rate": 9.90990990990991e-06,
-      "loss": 2.7541,
+      "loss": 1.6057,
       "step": 168
     },
     {
       "epoch": 1.52,
       "learning_rate": 9.849849849849851e-06,
-      "loss": 3.0172,
+      "loss": 1.6473,
       "step": 169
     },
     {
       "epoch": 1.53,
       "learning_rate": 9.78978978978979e-06,
-      "loss": 2.7282,
+      "loss": 1.3704,
       "step": 170
     },
     {
       "epoch": 1.54,
       "learning_rate": 9.729729729729732e-06,
-      "loss": 2.7824,
+      "loss": 1.5277,
       "step": 171
     },
     {
       "epoch": 1.55,
       "learning_rate": 9.669669669669671e-06,
-      "loss": 2.7859,
+      "loss": 1.4736,
       "step": 172
     },
     {
       "epoch": 1.56,
       "learning_rate": 9.60960960960961e-06,
-      "loss": 2.6787,
+      "loss": 1.5006,
       "step": 173
     },
     {
       "epoch": 1.57,
       "learning_rate": 9.54954954954955e-06,
-      "loss": 2.6765,
+      "loss": 1.4437,
       "step": 174
     },
     {
       "epoch": 1.58,
       "learning_rate": 9.489489489489491e-06,
-      "loss": 2.6947,
+      "loss": 1.553,
       "step": 175
     },
     {
       "epoch": 1.59,
       "learning_rate": 9.42942942942943e-06,
-      "loss": 2.8773,
+      "loss": 1.5653,
       "step": 176
     },
     {
       "epoch": 1.59,
       "learning_rate": 9.36936936936937e-06,
-      "loss": 2.7026,
+      "loss": 1.4045,
       "step": 177
     },
     {
       "epoch": 1.6,
       "learning_rate": 9.30930930930931e-06,
-      "loss": 2.9494,
+      "loss": 1.6598,
       "step": 178
     },
     {
       "epoch": 1.61,
       "learning_rate": 9.24924924924925e-06,
-      "loss": 2.6875,
+      "loss": 1.4961,
       "step": 179
     },
     {
       "epoch": 1.62,
       "learning_rate": 9.189189189189191e-06,
-      "loss": 2.663,
+      "loss": 1.4487,
       "step": 180
     },
     {
       "epoch": 1.63,
       "learning_rate": 9.129129129129129e-06,
-      "loss": 2.6553,
+      "loss": 1.4329,
       "step": 181
     },
     {
       "epoch": 1.64,
       "learning_rate": 9.06906906906907e-06,
-      "loss": 2.8712,
+      "loss": 1.5354,
       "step": 182
     },
     {
       "epoch": 1.65,
       "learning_rate": 9.00900900900901e-06,
-      "loss": 3.0305,
+      "loss": 1.7155,
       "step": 183
     },
     {
       "epoch": 1.66,
       "learning_rate": 8.94894894894895e-06,
-      "loss": 3.1108,
+      "loss": 1.6296,
       "step": 184
     },
     {
       "epoch": 1.67,
       "learning_rate": 8.888888888888888e-06,
-      "loss": 2.8379,
+      "loss": 1.5776,
       "step": 185
     },
     {
       "epoch": 1.68,
       "learning_rate": 8.82882882882883e-06,
-      "loss": 2.7338,
+      "loss": 1.5153,
       "step": 186
     },
     {
       "epoch": 1.68,
       "learning_rate": 8.768768768768769e-06,
-      "loss": 2.5357,
+      "loss": 1.353,
       "step": 187
     },
     {
       "epoch": 1.69,
       "learning_rate": 8.70870870870871e-06,
-      "loss": 2.6869,
+      "loss": 1.5484,
       "step": 188
     },
     {
       "epoch": 1.7,
       "learning_rate": 8.64864864864865e-06,
-      "loss": 2.6221,
+      "loss": 1.3934,
       "step": 189
     },
     {
       "epoch": 1.71,
       "learning_rate": 8.588588588588589e-06,
-      "loss": 2.7291,
+      "loss": 1.5505,
       "step": 190
     },
     {
       "epoch": 1.72,
       "learning_rate": 8.52852852852853e-06,
-      "loss": 2.7433,
+      "loss": 1.3282,
       "step": 191
     },
     {
       "epoch": 1.73,
       "learning_rate": 8.46846846846847e-06,
-      "loss": 2.742,
+      "loss": 1.4492,
       "step": 192
     },
     {
       "epoch": 1.74,
       "learning_rate": 8.408408408408409e-06,
-      "loss": 2.7337,
+      "loss": 1.4575,
       "step": 193
     },
     {
       "epoch": 1.75,
       "learning_rate": 8.348348348348348e-06,
-      "loss": 2.7747,
+      "loss": 1.4301,
       "step": 194
     },
     {
       "epoch": 1.76,
       "learning_rate": 8.288288288288289e-06,
-      "loss": 2.708,
+      "loss": 1.4052,
       "step": 195
     },
     {
       "epoch": 1.77,
       "learning_rate": 8.228228228228229e-06,
-      "loss": 3.1017,
+      "loss": 1.7595,
       "step": 196
     },
     {
       "epoch": 1.77,
       "learning_rate": 8.16816816816817e-06,
-      "loss": 2.8754,
+      "loss": 1.5578,
       "step": 197
     },
     {
       "epoch": 1.78,
       "learning_rate": 8.108108108108109e-06,
-      "loss": 2.8458,
+      "loss": 1.4915,
       "step": 198
     },
     {
       "epoch": 1.79,
       "learning_rate": 8.048048048048048e-06,
-      "loss": 2.5941,
+      "loss": 1.3835,
       "step": 199
     },
     {
       "epoch": 1.8,
       "learning_rate": 7.987987987987988e-06,
-      "loss": 2.5742,
+      "loss": 1.3505,
       "step": 200
     },
     {
       "epoch": 1.81,
       "learning_rate": 7.927927927927929e-06,
-      "loss": 2.8937,
+      "loss": 1.5297,
       "step": 201
     },
     {
       "epoch": 1.82,
       "learning_rate": 7.867867867867868e-06,
-      "loss": 2.9912,
+      "loss": 1.6206,
       "step": 202
     },
     {
       "epoch": 1.83,
       "learning_rate": 7.807807807807808e-06,
-      "loss": 2.762,
+      "loss": 1.4387,
       "step": 203
     },
     {
       "epoch": 1.84,
       "learning_rate": 7.747747747747749e-06,
-      "loss": 2.6746,
+      "loss": 1.5027,
       "step": 204
     },
     {
       "epoch": 1.85,
       "learning_rate": 7.687687687687688e-06,
-      "loss": 2.6511,
+      "loss": 1.468,
       "step": 205
     },
     {
       "epoch": 1.86,
       "learning_rate": 7.6276276276276285e-06,
-      "loss": 2.5135,
+      "loss": 1.4163,
       "step": 206
     },
     {
       "epoch": 1.86,
       "learning_rate": 7.567567567567569e-06,
-      "loss": 2.4414,
+      "loss": 1.225,
       "step": 207
     },
     {
       "epoch": 1.87,
       "learning_rate": 7.507507507507507e-06,
-      "loss": 2.7932,
+      "loss": 1.5404,
       "step": 208
     },
     {
       "epoch": 1.88,
       "learning_rate": 7.447447447447448e-06,
-      "loss": 3.0052,
+      "loss": 1.7203,
       "step": 209
     },
     {
       "epoch": 1.89,
       "learning_rate": 7.387387387387388e-06,
-      "loss": 2.8108,
+      "loss": 1.5402,
       "step": 210
     },
     {
       "epoch": 1.9,
       "learning_rate": 7.327327327327328e-06,
-      "loss": 2.7098,
+      "loss": 1.4962,
       "step": 211
     },
     {
       "epoch": 1.91,
       "learning_rate": 7.267267267267268e-06,
-      "loss": 2.7694,
+      "loss": 1.4352,
       "step": 212
     },
     {
       "epoch": 1.92,
       "learning_rate": 7.207207207207208e-06,
-      "loss": 2.6369,
+      "loss": 1.4584,
       "step": 213
     },
     {
       "epoch": 1.93,
       "learning_rate": 7.147147147147148e-06,
-      "loss": 2.6996,
+      "loss": 1.3992,
       "step": 214
     },
     {
       "epoch": 1.94,
       "learning_rate": 7.087087087087087e-06,
-      "loss": 2.8253,
+      "loss": 1.4541,
       "step": 215
     },
     {
       "epoch": 1.95,
       "learning_rate": 7.027027027027028e-06,
-      "loss": 2.5087,
+      "loss": 1.3759,
       "step": 216
     },
     {
       "epoch": 1.95,
       "learning_rate": 6.966966966966967e-06,
-      "loss": 2.7107,
+      "loss": 1.4622,
       "step": 217
     },
     {
       "epoch": 1.96,
       "learning_rate": 6.906906906906907e-06,
-      "loss": 2.6624,
+      "loss": 1.4799,
       "step": 218
     },
     {
       "epoch": 1.97,
       "learning_rate": 6.846846846846848e-06,
-      "loss": 2.5835,
+      "loss": 1.4265,
       "step": 219
     },
     {
       "epoch": 1.98,
       "learning_rate": 6.786786786786788e-06,
-      "loss": 2.6851,
+      "loss": 1.3938,
       "step": 220
     },
     {
       "epoch": 1.99,
       "learning_rate": 6.726726726726728e-06,
-      "loss": 2.7938,
+      "loss": 1.5023,
       "step": 221
     },
     {
       "epoch": 2.0,
       "learning_rate": 6.666666666666667e-06,
-      "loss": 2.4996,
+      "loss": 1.2256,
       "step": 222
     },
     {
       "epoch": 2.01,
       "learning_rate": 6.606606606606607e-06,
-      "loss": 2.62,
+      "loss": 1.3189,
       "step": 223
     },
     {
       "epoch": 2.02,
       "learning_rate": 6.546546546546547e-06,
-      "loss": 2.7648,
+      "loss": 1.4738,
       "step": 224
     },
     {
       "epoch": 2.03,
       "learning_rate": 6.486486486486487e-06,
-      "loss": 2.8287,
+      "loss": 1.4976,
       "step": 225
     },
     {
       "epoch": 2.04,
       "learning_rate": 6.426426426426427e-06,
-      "loss": 2.6175,
+      "loss": 1.4145,
       "step": 226
     },
     {
       "epoch": 2.05,
       "learning_rate": 6.366366366366366e-06,
-      "loss": 2.4404,
+      "loss": 1.3549,
       "step": 227
     },
     {
       "epoch": 2.05,
       "learning_rate": 6.3063063063063065e-06,
-      "loss": 2.5263,
+      "loss": 1.4151,
       "step": 228
     },
     {
       "epoch": 2.06,
       "learning_rate": 6.246246246246247e-06,
-      "loss": 2.9257,
+      "loss": 1.6266,
       "step": 229
     },
     {
       "epoch": 2.07,
       "learning_rate": 6.186186186186187e-06,
-      "loss": 2.9413,
+      "loss": 1.4899,
       "step": 230
     },
     {
       "epoch": 2.08,
       "learning_rate": 6.126126126126126e-06,
-      "loss": 2.5352,
+      "loss": 1.2746,
       "step": 231
     },
     {
       "epoch": 2.09,
       "learning_rate": 6.066066066066067e-06,
-      "loss": 2.6219,
+      "loss": 1.3838,
       "step": 232
     },
     {
       "epoch": 2.1,
       "learning_rate": 6.006006006006007e-06,
-      "loss": 2.6584,
+      "loss": 1.3595,
       "step": 233
     },
     {
       "epoch": 2.11,
       "learning_rate": 5.945945945945947e-06,
-      "loss": 2.7308,
+      "loss": 1.3993,
       "step": 234
     },
     {
       "epoch": 2.12,
       "learning_rate": 5.885885885885886e-06,
-      "loss": 2.5438,
+      "loss": 1.2979,
       "step": 235
     },
     {
       "epoch": 2.13,
       "learning_rate": 5.825825825825826e-06,
-      "loss": 2.5946,
+      "loss": 1.424,
       "step": 236
     },
     {
       "epoch": 2.14,
       "learning_rate": 5.765765765765766e-06,
-      "loss": 2.597,
+      "loss": 1.3078,
       "step": 237
     },
     {
       "epoch": 2.14,
       "learning_rate": 5.7057057057057065e-06,
-      "loss": 2.6751,
+      "loss": 1.4321,
       "step": 238
     },
     {
       "epoch": 2.15,
       "learning_rate": 5.645645645645647e-06,
-      "loss": 2.7518,
+      "loss": 1.3732,
       "step": 239
     },
     {
       "epoch": 2.16,
       "learning_rate": 5.585585585585585e-06,
-      "loss": 2.6249,
+      "loss": 1.5107,
       "step": 240
     },
     {
       "epoch": 2.17,
       "learning_rate": 5.5255255255255255e-06,
-      "loss": 2.7765,
+      "loss": 1.5524,
       "step": 241
     },
     {
       "epoch": 2.18,
       "learning_rate": 5.465465465465466e-06,
-      "loss": 2.7938,
+      "loss": 1.4516,
       "step": 242
     },
     {
       "epoch": 2.19,
       "learning_rate": 5.405405405405406e-06,
-      "loss": 2.3933,
+      "loss": 1.2053,
       "step": 243
     },
     {
       "epoch": 2.2,
       "learning_rate": 5.345345345345346e-06,
-      "loss": 2.8995,
+      "loss": 1.459,
       "step": 244
     },
     {
       "epoch": 2.21,
       "learning_rate": 5.285285285285286e-06,
-      "loss": 2.7298,
+      "loss": 1.3932,
       "step": 245
     },
     {
       "epoch": 2.22,
       "learning_rate": 5.225225225225226e-06,
-      "loss": 2.5716,
+      "loss": 1.3219,
       "step": 246
     },
     {
       "epoch": 2.23,
       "learning_rate": 5.165165165165165e-06,
-      "loss": 2.8012,
+      "loss": 1.3934,
       "step": 247
     },
     {
       "epoch": 2.23,
       "learning_rate": 5.105105105105106e-06,
-      "loss": 2.5597,
+      "loss": 1.3035,
       "step": 248
     },
     {
       "epoch": 2.24,
       "learning_rate": 5.045045045045045e-06,
-      "loss": 2.5855,
+      "loss": 1.3401,
       "step": 249
     },
     {
       "epoch": 2.25,
       "learning_rate": 4.984984984984985e-06,
-      "loss": 2.6125,
+      "loss": 1.3076,
       "step": 250
     },
     {
       "epoch": 2.26,
       "learning_rate": 4.9249249249249255e-06,
-      "loss": 2.5589,
+      "loss": 1.3364,
       "step": 251
     },
     {
       "epoch": 2.27,
       "learning_rate": 4.864864864864866e-06,
-      "loss": 2.4177,
+      "loss": 1.2897,
       "step": 252
     },
     {
       "epoch": 2.28,
       "learning_rate": 4.804804804804805e-06,
-      "loss": 2.7882,
+      "loss": 1.2572,
       "step": 253
     },
     {
       "epoch": 2.29,
       "learning_rate": 4.7447447447447454e-06,
-      "loss": 2.6047,
+      "loss": 1.3396,
       "step": 254
     },
     {
       "epoch": 2.3,
       "learning_rate": 4.684684684684685e-06,
-      "loss": 2.6191,
+      "loss": 1.4288,
       "step": 255
     },
     {
       "epoch": 2.31,
       "learning_rate": 4.624624624624625e-06,
-      "loss": 2.5648,
+      "loss": 1.3392,
       "step": 256
     },
     {
       "epoch": 2.32,
       "learning_rate": 4.5645645645645645e-06,
-      "loss": 2.6584,
+      "loss": 1.4231,
       "step": 257
     },
     {
       "epoch": 2.32,
       "learning_rate": 4.504504504504505e-06,
-      "loss": 2.5868,
+      "loss": 1.3351,
       "step": 258
     },
     {
       "epoch": 2.33,
       "learning_rate": 4.444444444444444e-06,
-      "loss": 2.5987,
+      "loss": 1.4397,
       "step": 259
     },
     {
       "epoch": 2.34,
       "learning_rate": 4.384384384384384e-06,
-      "loss": 2.6715,
+      "loss": 1.4746,
       "step": 260
     },
     {
       "epoch": 2.35,
       "learning_rate": 4.324324324324325e-06,
-      "loss": 2.546,
+      "loss": 1.4103,
       "step": 261
     },
     {
       "epoch": 2.36,
       "learning_rate": 4.264264264264265e-06,
-      "loss": 2.7366,
+      "loss": 1.325,
       "step": 262
     },
     {
       "epoch": 2.37,
       "learning_rate": 4.204204204204204e-06,
-      "loss": 2.7874,
+      "loss": 1.4474,
       "step": 263
     },
     {
       "epoch": 2.38,
       "learning_rate": 4.1441441441441446e-06,
-      "loss": 2.5763,
+      "loss": 1.243,
       "step": 264
     },
     {
       "epoch": 2.39,
       "learning_rate": 4.084084084084085e-06,
-      "loss": 2.7863,
+      "loss": 1.3342,
       "step": 265
     },
     {
       "epoch": 2.4,
       "learning_rate": 4.024024024024024e-06,
-      "loss": 2.5058,
+      "loss": 1.2141,
       "step": 266
     },
     {
       "epoch": 2.41,
       "learning_rate": 3.9639639639639645e-06,
-      "loss": 2.623,
+      "loss": 1.389,
       "step": 267
     },
     {
       "epoch": 2.41,
       "learning_rate": 3.903903903903904e-06,
-      "loss": 2.6655,
+      "loss": 1.4364,
       "step": 268
     },
     {
       "epoch": 2.42,
       "learning_rate": 3.843843843843844e-06,
-      "loss": 2.8604,
+      "loss": 1.4544,
       "step": 269
     },
     {
       "epoch": 2.43,
       "learning_rate": 3.7837837837837844e-06,
-      "loss": 2.8491,
+      "loss": 1.4788,
       "step": 270
     },
     {
       "epoch": 2.44,
       "learning_rate": 3.723723723723724e-06,
-      "loss": 2.6519,
+      "loss": 1.3652,
       "step": 271
     },
     {
       "epoch": 2.45,
       "learning_rate": 3.663663663663664e-06,
-      "loss": 2.3903,
+      "loss": 1.2965,
       "step": 272
     },
     {
       "epoch": 2.46,
       "learning_rate": 3.603603603603604e-06,
-      "loss": 2.5994,
+      "loss": 1.4887,
       "step": 273
     },
     {
       "epoch": 2.47,
       "learning_rate": 3.5435435435435437e-06,
-      "loss": 2.5526,
+      "loss": 1.2402,
       "step": 274
     },
     {
       "epoch": 2.48,
       "learning_rate": 3.4834834834834835e-06,
-      "loss": 2.5653,
+      "loss": 1.265,
       "step": 275
     },
     {
       "epoch": 2.49,
       "learning_rate": 3.423423423423424e-06,
-      "loss": 2.5769,
+      "loss": 1.3136,
       "step": 276
     },
     {
       "epoch": 2.5,
       "learning_rate": 3.363363363363364e-06,
-      "loss": 2.7054,
+      "loss": 1.5174,
       "step": 277
     },
     {
       "epoch": 2.5,
       "learning_rate": 3.3033033033033035e-06,
-      "loss": 2.7291,
+      "loss": 1.3669,
       "step": 278
     },
     {
       "epoch": 2.51,
       "learning_rate": 3.2432432432432437e-06,
-      "loss": 2.5145,
+      "loss": 1.2474,
       "step": 279
     },
     {
       "epoch": 2.52,
       "learning_rate": 3.183183183183183e-06,
-      "loss": 2.7142,
+      "loss": 1.3761,
       "step": 280
     },
     {
       "epoch": 2.53,
       "learning_rate": 3.1231231231231234e-06,
-      "loss": 2.6612,
+      "loss": 1.3245,
       "step": 281
     },
     {
       "epoch": 2.54,
       "learning_rate": 3.063063063063063e-06,
-      "loss": 2.7143,
+      "loss": 1.3889,
       "step": 282
     },
     {
       "epoch": 2.55,
       "learning_rate": 3.0030030030030034e-06,
-      "loss": 2.6341,
+      "loss": 1.2964,
       "step": 283
     },
     {
       "epoch": 2.56,
       "learning_rate": 2.942942942942943e-06,
-      "loss": 2.5404,
+      "loss": 1.4232,
       "step": 284
     },
     {
       "epoch": 2.57,
       "learning_rate": 2.882882882882883e-06,
-      "loss": 2.7922,
+      "loss": 1.5877,
       "step": 285
     },
     {
       "epoch": 2.58,
       "learning_rate": 2.8228228228228234e-06,
-      "loss": 3.1301,
+      "loss": 1.7226,
       "step": 286
     },
     {
       "epoch": 2.59,
       "learning_rate": 2.7627627627627628e-06,
-      "loss": 2.8461,
+      "loss": 1.3672,
       "step": 287
     },
     {
       "epoch": 2.59,
       "learning_rate": 2.702702702702703e-06,
-      "loss": 2.6317,
+      "loss": 1.3058,
       "step": 288
     },
     {
       "epoch": 2.6,
       "learning_rate": 2.642642642642643e-06,
-      "loss": 2.7555,
+      "loss": 1.5852,
       "step": 289
     },
     {
       "epoch": 2.61,
       "learning_rate": 2.5825825825825827e-06,
-      "loss": 2.3511,
+      "loss": 1.1589,
       "step": 290
     },
     {
       "epoch": 2.62,
       "learning_rate": 2.5225225225225225e-06,
-      "loss": 2.6216,
+      "loss": 1.4041,
       "step": 291
     },
     {
       "epoch": 2.63,
       "learning_rate": 2.4624624624624628e-06,
-      "loss": 2.6718,
+      "loss": 1.3072,
       "step": 292
     },
     {
       "epoch": 2.64,
       "learning_rate": 2.4024024024024026e-06,
-      "loss": 2.6301,
+      "loss": 1.3421,
       "step": 293
     },
     {
       "epoch": 2.65,
       "learning_rate": 2.3423423423423424e-06,
-      "loss": 2.7771,
+      "loss": 1.5669,
       "step": 294
     },
     {
       "epoch": 2.66,
       "learning_rate": 2.2822822822822822e-06,
-      "loss": 2.5465,
+      "loss": 1.4368,
       "step": 295
     },
     {
       "epoch": 2.67,
       "learning_rate": 2.222222222222222e-06,
-      "loss": 2.5168,
+      "loss": 1.1976,
       "step": 296
     },
     {
       "epoch": 2.68,
       "learning_rate": 2.1621621621621623e-06,
-      "loss": 2.6376,
+      "loss": 1.4212,
       "step": 297
     },
     {
       "epoch": 2.68,
       "learning_rate": 2.102102102102102e-06,
-      "loss": 2.4557,
+      "loss": 1.3067,
       "step": 298
     },
     {
       "epoch": 2.69,
       "learning_rate": 2.0420420420420424e-06,
-      "loss": 2.5339,
+      "loss": 1.3672,
       "step": 299
     },
     {
       "epoch": 2.7,
       "learning_rate": 1.9819819819819822e-06,
-      "loss": 2.6562,
+      "loss": 1.3335,
       "step": 300
     },
     {
       "epoch": 2.71,
       "learning_rate": 1.921921921921922e-06,
-      "loss": 2.7499,
+      "loss": 1.467,
       "step": 301
     },
     {
       "epoch": 2.72,
       "learning_rate": 1.861861861861862e-06,
-      "loss": 2.8368,
+      "loss": 1.535,
       "step": 302
     },
     {
       "epoch": 2.73,
       "learning_rate": 1.801801801801802e-06,
-      "loss": 2.6257,
+      "loss": 1.3377,
       "step": 303
     },
     {
       "epoch": 2.74,
       "learning_rate": 1.7417417417417418e-06,
-      "loss": 2.6908,
+      "loss": 1.5501,
       "step": 304
     },
     {
       "epoch": 2.75,
       "learning_rate": 1.681681681681682e-06,
-      "loss": 2.8002,
+      "loss": 1.5146,
       "step": 305
     },
     {
       "epoch": 2.76,
       "learning_rate": 1.6216216216216219e-06,
-      "loss": 2.6201,
+      "loss": 1.3974,
       "step": 306
     },
     {
       "epoch": 2.77,
       "learning_rate": 1.5615615615615617e-06,
-      "loss": 2.6824,
+      "loss": 1.4609,
       "step": 307
     },
     {
       "epoch": 2.77,
       "learning_rate": 1.5015015015015017e-06,
-      "loss": 2.7193,
+      "loss": 1.4186,
       "step": 308
     },
     {
       "epoch": 2.78,
       "learning_rate": 1.4414414414414416e-06,
-      "loss": 2.7957,
+      "loss": 1.4664,
       "step": 309
     },
     {
       "epoch": 2.79,
       "learning_rate": 1.3813813813813814e-06,
-      "loss": 2.6111,
+      "loss": 1.3609,
       "step": 310
     },
     {
       "epoch": 2.8,
       "learning_rate": 1.3213213213213214e-06,
-      "loss": 2.7703,
+      "loss": 1.4355,
       "step": 311
     },
     {
       "epoch": 2.81,
       "learning_rate": 1.2612612612612613e-06,
-      "loss": 2.6248,
+      "loss": 1.3592,
       "step": 312
     },
     {
       "epoch": 2.82,
       "learning_rate": 1.2012012012012013e-06,
-      "loss": 2.6614,
+      "loss": 1.3873,
       "step": 313
     },
     {
       "epoch": 2.83,
       "learning_rate": 1.1411411411411411e-06,
-      "loss": 2.7111,
+      "loss": 1.3438,
       "step": 314
     },
     {
       "epoch": 2.84,
       "learning_rate": 1.0810810810810812e-06,
-      "loss": 2.3901,
+      "loss": 1.2623,
       "step": 315
     },
     {
       "epoch": 2.85,
       "learning_rate": 1.0210210210210212e-06,
-      "loss": 2.7881,
+      "loss": 1.4811,
       "step": 316
     },
     {
       "epoch": 2.86,
       "learning_rate": 9.60960960960961e-07,
-      "loss": 2.651,
+      "loss": 1.2461,
       "step": 317
     },
     {
       "epoch": 2.86,
       "learning_rate": 9.00900900900901e-07,
-      "loss": 2.5288,
+      "loss": 1.2419,
       "step": 318
     },
     {
       "epoch": 2.87,
       "learning_rate": 8.40840840840841e-07,
-      "loss": 2.8173,
+      "loss": 1.5783,
       "step": 319
     },
     {
       "epoch": 2.88,
       "learning_rate": 7.807807807807808e-07,
-      "loss": 2.5508,
+      "loss": 1.2684,
       "step": 320
     },
     {
       "epoch": 2.89,
       "learning_rate": 7.207207207207208e-07,
-      "loss": 2.6362,
+      "loss": 1.3713,
       "step": 321
     },
     {
       "epoch": 2.9,
       "learning_rate": 6.606606606606607e-07,
-      "loss": 2.8817,
+      "loss": 1.5109,
       "step": 322
     },
     {
       "epoch": 2.91,
       "learning_rate": 6.006006006006006e-07,
-      "loss": 2.7706,
+      "loss": 1.3602,
       "step": 323
     },
     {
       "epoch": 2.92,
       "learning_rate": 5.405405405405406e-07,
-      "loss": 2.7205,
+      "loss": 1.5054,
       "step": 324
     },
     {
       "epoch": 2.93,
       "learning_rate": 4.804804804804805e-07,
-      "loss": 2.8032,
+      "loss": 1.5096,
       "step": 325
     },
     {
       "epoch": 2.94,
       "learning_rate": 4.204204204204205e-07,
-      "loss": 2.6215,
+      "loss": 1.3165,
       "step": 326
     },
     {
       "epoch": 2.95,
       "learning_rate": 3.603603603603604e-07,
-      "loss": 2.831,
+      "loss": 1.4575,
       "step": 327
     },
     {
       "epoch": 2.95,
       "learning_rate": 3.003003003003003e-07,
-      "loss": 2.9592,
+      "loss": 1.6184,
       "step": 328
     },
     {
       "epoch": 2.96,
       "learning_rate": 2.4024024024024026e-07,
-      "loss": 2.4744,
+      "loss": 1.2022,
       "step": 329
     },
     {
       "epoch": 2.97,
       "learning_rate": 1.801801801801802e-07,
-      "loss": 2.7128,
+      "loss": 1.4417,
       "step": 330
     },
     {
       "epoch": 2.98,
       "learning_rate": 1.2012012012012013e-07,
-      "loss": 2.4625,
+      "loss": 1.2554,
       "step": 331
     },
     {
       "epoch": 2.99,
       "learning_rate": 6.006006006006006e-08,
-      "loss": 2.4875,
+      "loss": 1.28,
       "step": 332
     },
     {
       "epoch": 3.0,
       "learning_rate": 0.0,
-      "loss": 2.5134,
+      "loss": 1.2557,
       "step": 333
     },
     {
       "epoch": 3.0,
       "step": 333,
-      "total_flos": 119723569053696.0,
-      "train_loss": 3.0590674640896083,
-      "train_runtime": 25.7247,
-      "train_samples_per_second": 411.317,
-      "train_steps_per_second": 12.945
+      "total_flos": 1915204220928000.0,
+      "train_loss": 1.781525854233865,
+      "train_runtime": 119.0355,
+      "train_samples_per_second": 88.889,
+      "train_steps_per_second": 2.797
     }
   ],
   "max_steps": 333,
   "num_train_epochs": 3,
-  "total_flos": 119723569053696.0,
+  "total_flos": 1915204220928000.0,
   "trial_name": null,
   "trial_params": null
 }