diff --git "a/checkpoint-3131/trainer_state.json" "b/checkpoint-3131/trainer_state.json"
--- "a/checkpoint-3131/trainer_state.json"
+++ "b/checkpoint-3131/trainer_state.json"
@@ -11,1605 +11,1605 @@
     {
       "epoch": 0.0,
       "learning_rate": 3.0000000000000004e-07,
-      "loss": 3.6419,
+      "loss": 6.0535,
       "step": 1
     },
     {
       "epoch": 0.0,
-      "eval_loss": 3.714031457901001,
-      "eval_runtime": 28.1012,
-      "eval_samples_per_second": 187.643,
-      "eval_steps_per_second": 5.872,
+      "eval_loss": 6.236783504486084,
+      "eval_runtime": 28.6757,
+      "eval_samples_per_second": 183.884,
+      "eval_steps_per_second": 11.508,
       "step": 1
     },
     {
       "epoch": 0.0,
       "learning_rate": 6.000000000000001e-07,
-      "loss": 3.8158,
+      "loss": 6.1527,
       "step": 2
     },
     {
       "epoch": 0.0,
       "learning_rate": 9e-07,
-      "loss": 3.6399,
+      "loss": 6.1004,
       "step": 3
     },
     {
       "epoch": 0.0,
       "learning_rate": 1.2000000000000002e-06,
-      "loss": 3.4312,
+      "loss": 5.6802,
       "step": 4
     },
     {
       "epoch": 0.0,
       "learning_rate": 1.5e-06,
-      "loss": 3.6597,
+      "loss": 6.105,
       "step": 5
     },
     {
       "epoch": 0.0,
       "learning_rate": 1.8e-06,
-      "loss": 3.5666,
+      "loss": 5.8833,
       "step": 6
     },
     {
       "epoch": 0.0,
       "learning_rate": 2.1000000000000002e-06,
-      "loss": 3.7699,
+      "loss": 6.0386,
       "step": 7
     },
     {
       "epoch": 0.0,
       "learning_rate": 2.4000000000000003e-06,
-      "loss": 3.5036,
+      "loss": 5.6186,
       "step": 8
     },
     {
       "epoch": 0.0,
       "learning_rate": 2.7e-06,
-      "loss": 3.4987,
+      "loss": 5.6635,
       "step": 9
     },
     {
       "epoch": 0.0,
       "learning_rate": 3e-06,
-      "loss": 3.5354,
+      "loss": 5.5451,
       "step": 10
     },
     {
       "epoch": 0.0,
       "learning_rate": 3.3e-06,
-      "loss": 3.6143,
+      "loss": 5.8442,
       "step": 11
     },
     {
       "epoch": 0.0,
       "learning_rate": 3.6e-06,
-      "loss": 3.3061,
+      "loss": 5.2233,
       "step": 12
     },
     {
       "epoch": 0.0,
       "learning_rate": 3.9e-06,
-      "loss": 3.2166,
+      "loss": 5.2118,
       "step": 13
     },
     {
       "epoch": 0.0,
       "learning_rate": 4.2000000000000004e-06,
-      "loss": 3.0829,
+      "loss": 5.0138,
       "step": 14
     },
     {
       "epoch": 0.0,
       "learning_rate": 4.5e-06,
-      "loss": 3.2437,
+      "loss": 4.8483,
       "step": 15
     },
     {
       "epoch": 0.01,
       "learning_rate": 4.800000000000001e-06,
-      "loss": 3.0641,
+      "loss": 4.5374,
       "step": 16
     },
     {
       "epoch": 0.01,
       "learning_rate": 5.1e-06,
-      "loss": 2.8823,
+      "loss": 4.4414,
       "step": 17
     },
     {
       "epoch": 0.01,
       "learning_rate": 5.4e-06,
-      "loss": 2.6965,
+      "loss": 4.0476,
       "step": 18
     },
     {
       "epoch": 0.01,
       "learning_rate": 5.7000000000000005e-06,
-      "loss": 2.8709,
+      "loss": 4.1573,
       "step": 19
     },
     {
       "epoch": 0.01,
       "learning_rate": 6e-06,
-      "loss": 2.7647,
+      "loss": 3.9995,
       "step": 20
     },
     {
       "epoch": 0.01,
       "learning_rate": 6.3e-06,
-      "loss": 2.6207,
+      "loss": 3.6553,
       "step": 21
     },
     {
       "epoch": 0.01,
       "learning_rate": 6.6e-06,
-      "loss": 2.3536,
+      "loss": 3.4045,
       "step": 22
     },
     {
       "epoch": 0.01,
       "learning_rate": 6.900000000000001e-06,
-      "loss": 2.3904,
+      "loss": 3.3223,
       "step": 23
     },
     {
       "epoch": 0.01,
       "learning_rate": 7.2e-06,
-      "loss": 2.194,
+      "loss": 3.0992,
       "step": 24
     },
     {
       "epoch": 0.01,
       "learning_rate": 7.5e-06,
-      "loss": 2.0802,
+      "loss": 2.9959,
       "step": 25
     },
     {
       "epoch": 0.01,
       "learning_rate": 7.8e-06,
-      "loss": 1.8686,
+      "loss": 2.8946,
       "step": 26
     },
     {
       "epoch": 0.01,
       "learning_rate": 8.1e-06,
-      "loss": 1.2353,
+      "loss": 2.9426,
       "step": 27
     },
     {
       "epoch": 0.01,
       "learning_rate": 8.400000000000001e-06,
-      "loss": 0.8353,
+      "loss": 2.5529,
       "step": 28
     },
     {
       "epoch": 0.01,
       "learning_rate": 8.7e-06,
-      "loss": 0.6401,
+      "loss": 2.2701,
       "step": 29
     },
     {
       "epoch": 0.01,
       "learning_rate": 9e-06,
-      "loss": 0.7808,
+      "loss": 2.0024,
       "step": 30
     },
     {
       "epoch": 0.01,
       "learning_rate": 9.3e-06,
-      "loss": 0.7208,
+      "loss": 1.7888,
       "step": 31
     },
     {
       "epoch": 0.01,
       "learning_rate": 9.600000000000001e-06,
-      "loss": 0.556,
+      "loss": 1.1965,
       "step": 32
     },
     {
       "epoch": 0.01,
       "learning_rate": 9.9e-06,
-      "loss": 0.5941,
+      "loss": 1.1853,
       "step": 33
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.02e-05,
-      "loss": 0.5093,
+      "loss": 0.9086,
       "step": 34
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.05e-05,
-      "loss": 0.5861,
+      "loss": 0.7273,
       "step": 35
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.08e-05,
-      "loss": 0.4627,
+      "loss": 0.6449,
       "step": 36
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.11e-05,
-      "loss": 0.4513,
+      "loss": 0.5931,
       "step": 37
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.1400000000000001e-05,
-      "loss": 0.3879,
+      "loss": 0.4779,
       "step": 38
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.1700000000000001e-05,
-      "loss": 0.3766,
+      "loss": 0.4716,
       "step": 39
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.2e-05,
-      "loss": 0.3213,
+      "loss": 0.4238,
       "step": 40
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.2299999999999999e-05,
-      "loss": 0.3489,
+      "loss": 0.4042,
       "step": 41
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.26e-05,
-      "loss": 0.2836,
+      "loss": 0.3709,
       "step": 42
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.29e-05,
-      "loss": 0.3482,
+      "loss": 0.4412,
       "step": 43
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.32e-05,
-      "loss": 0.2433,
+      "loss": 0.2741,
       "step": 44
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.3500000000000001e-05,
-      "loss": 0.2226,
+      "loss": 0.2435,
       "step": 45
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.3800000000000002e-05,
-      "loss": 0.2462,
+      "loss": 0.2673,
       "step": 46
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.4099999999999999e-05,
-      "loss": 0.2534,
+      "loss": 0.2521,
       "step": 47
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.44e-05,
-      "loss": 0.2298,
+      "loss": 0.2228,
       "step": 48
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.47e-05,
-      "loss": 0.215,
+      "loss": 0.2301,
       "step": 49
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.5e-05,
-      "loss": 0.2234,
+      "loss": 0.2527,
       "step": 50
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.53e-05,
-      "loss": 0.1899,
+      "loss": 0.2003,
       "step": 51
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.56e-05,
-      "loss": 0.1684,
+      "loss": 0.2002,
       "step": 52
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.59e-05,
-      "loss": 0.138,
+      "loss": 0.1483,
       "step": 53
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.62e-05,
-      "loss": 0.1327,
+      "loss": 0.1477,
       "step": 54
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.65e-05,
-      "loss": 0.1816,
+      "loss": 0.203,
       "step": 55
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.6800000000000002e-05,
-      "loss": 0.1412,
+      "loss": 0.1541,
       "step": 56
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.71e-05,
-      "loss": 0.18,
+      "loss": 0.2026,
       "step": 57
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.74e-05,
-      "loss": 0.1254,
+      "loss": 0.1226,
       "step": 58
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.77e-05,
-      "loss": 0.1552,
+      "loss": 0.2942,
       "step": 59
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.8e-05,
-      "loss": 0.1943,
+      "loss": 0.1515,
       "step": 60
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.83e-05,
-      "loss": 0.1746,
+      "loss": 0.1453,
       "step": 61
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.86e-05,
-      "loss": 0.1195,
+      "loss": 0.1109,
       "step": 62
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.8900000000000002e-05,
-      "loss": 0.1555,
+      "loss": 0.1693,
       "step": 63
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.9200000000000003e-05,
-      "loss": 0.1173,
+      "loss": 0.1435,
       "step": 64
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.95e-05,
-      "loss": 0.0829,
+      "loss": 0.1054,
       "step": 65
     },
     {
       "epoch": 0.02,
       "learning_rate": 1.98e-05,
-      "loss": 0.1057,
+      "loss": 0.1141,
       "step": 66
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.01e-05,
-      "loss": 0.1058,
+      "loss": 0.1126,
       "step": 67
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.04e-05,
-      "loss": 0.1056,
+      "loss": 0.0971,
       "step": 68
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.07e-05,
-      "loss": 0.0856,
+      "loss": 0.0835,
       "step": 69
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.1e-05,
-      "loss": 0.1165,
+      "loss": 0.0912,
       "step": 70
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.13e-05,
-      "loss": 0.0946,
+      "loss": 0.0894,
       "step": 71
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.16e-05,
-      "loss": 0.0911,
+      "loss": 0.0879,
       "step": 72
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.19e-05,
-      "loss": 0.0712,
+      "loss": 0.0721,
       "step": 73
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.22e-05,
-      "loss": 0.0583,
+      "loss": 0.078,
       "step": 74
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.25e-05,
-      "loss": 0.0931,
+      "loss": 0.0975,
       "step": 75
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.2800000000000002e-05,
-      "loss": 0.0844,
+      "loss": 0.0823,
       "step": 76
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.3100000000000002e-05,
-      "loss": 0.13,
+      "loss": 0.1045,
       "step": 77
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.3400000000000003e-05,
-      "loss": 0.09,
+      "loss": 0.0814,
       "step": 78
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.37e-05,
-      "loss": 0.1056,
+      "loss": 0.0841,
       "step": 79
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.4e-05,
-      "loss": 0.0711,
+      "loss": 0.0564,
       "step": 80
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.43e-05,
-      "loss": 0.0853,
+      "loss": 0.0791,
       "step": 81
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.4599999999999998e-05,
-      "loss": 0.1072,
+      "loss": 0.1116,
       "step": 82
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.49e-05,
-      "loss": 0.0877,
+      "loss": 0.0904,
       "step": 83
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.52e-05,
-      "loss": 0.0563,
+      "loss": 0.0594,
       "step": 84
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.55e-05,
-      "loss": 0.0686,
+      "loss": 0.0668,
       "step": 85
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.58e-05,
-      "loss": 0.0587,
+      "loss": 0.0762,
       "step": 86
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.61e-05,
-      "loss": 0.0599,
+      "loss": 0.0609,
       "step": 87
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.64e-05,
-      "loss": 0.0516,
+      "loss": 0.0735,
       "step": 88
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.6700000000000002e-05,
-      "loss": 0.0403,
+      "loss": 0.0629,
       "step": 89
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.7000000000000002e-05,
-      "loss": 0.054,
+      "loss": 0.0607,
       "step": 90
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.7300000000000003e-05,
-      "loss": 0.0513,
+      "loss": 0.0571,
       "step": 91
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.7600000000000003e-05,
-      "loss": 0.0581,
+      "loss": 0.0704,
       "step": 92
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.79e-05,
-      "loss": 0.0563,
+      "loss": 0.0817,
       "step": 93
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.8199999999999998e-05,
-      "loss": 0.0392,
+      "loss": 0.0452,
       "step": 94
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.8499999999999998e-05,
-      "loss": 0.06,
+      "loss": 0.0467,
       "step": 95
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.88e-05,
-      "loss": 0.065,
+      "loss": 0.0532,
       "step": 96
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.91e-05,
-      "loss": 0.0489,
+      "loss": 0.0515,
       "step": 97
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.94e-05,
-      "loss": 0.0729,
+      "loss": 0.0583,
       "step": 98
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.97e-05,
-      "loss": 0.0617,
+      "loss": 0.0603,
       "step": 99
     },
     {
       "epoch": 0.03,
       "learning_rate": 3e-05,
-      "loss": 0.0513,
+      "loss": 0.0545,
       "step": 100
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.9999991942708107e-05,
-      "loss": 0.0567,
+      "loss": 0.0683,
       "step": 101
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.9999967770841087e-05,
-      "loss": 0.0494,
+      "loss": 0.0468,
       "step": 102
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.9999927484424903e-05,
-      "loss": 0.0631,
+      "loss": 0.0847,
       "step": 103
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.9999871083502844e-05,
-      "loss": 0.0511,
+      "loss": 0.0513,
       "step": 104
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.999979856813549e-05,
-      "loss": 0.0804,
+      "loss": 0.0589,
       "step": 105
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.9999709938400753e-05,
-      "loss": 0.0539,
+      "loss": 0.0577,
       "step": 106
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.9999605194393848e-05,
-      "loss": 0.0434,
+      "loss": 0.0405,
       "step": 107
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.999948433622729e-05,
-      "loss": 0.0561,
+      "loss": 0.0559,
       "step": 108
     },
     {
       "epoch": 0.03,
       "learning_rate": 2.999934736403094e-05,
-      "loss": 0.0588,
+      "loss": 0.0502,
       "step": 109
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9999194277951926e-05,
-      "loss": 0.0858,
+      "loss": 0.0711,
       "step": 110
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9999025078154727e-05,
-      "loss": 0.0351,
+      "loss": 0.0359,
       "step": 111
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9998839764821103e-05,
-      "loss": 0.044,
+      "loss": 0.0572,
       "step": 112
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999863833815014e-05,
-      "loss": 0.0394,
+      "loss": 0.0377,
       "step": 113
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9998420798358235e-05,
-      "loss": 0.0569,
+      "loss": 0.0546,
       "step": 114
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999818714567909e-05,
-      "loss": 0.0497,
+      "loss": 0.0439,
       "step": 115
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999793738036372e-05,
-      "loss": 0.0935,
+      "loss": 0.0686,
       "step": 116
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9997671502680447e-05,
-      "loss": 0.0571,
+      "loss": 0.0653,
       "step": 117
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999738951291491e-05,
-      "loss": 0.0513,
+      "loss": 0.0451,
       "step": 118
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9997091411370048e-05,
-      "loss": 0.0547,
+      "loss": 0.0441,
       "step": 119
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9996777198366115e-05,
-      "loss": 0.0661,
+      "loss": 0.073,
       "step": 120
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999644687424067e-05,
-      "loss": 0.0441,
+      "loss": 0.0439,
       "step": 121
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9996100439348582e-05,
-      "loss": 0.0459,
+      "loss": 0.0425,
       "step": 122
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999573789406203e-05,
-      "loss": 0.0433,
+      "loss": 0.0375,
       "step": 123
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.99953592387705e-05,
-      "loss": 0.06,
+      "loss": 0.0405,
       "step": 124
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9994964473880776e-05,
-      "loss": 0.0458,
+      "loss": 0.0481,
       "step": 125
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9994553599816963e-05,
-      "loss": 0.051,
+      "loss": 0.0527,
       "step": 126
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9994126617020462e-05,
-      "loss": 0.0558,
+      "loss": 0.0509,
       "step": 127
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9993683525949985e-05,
-      "loss": 0.0334,
+      "loss": 0.0476,
       "step": 128
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9993224327081544e-05,
-      "loss": 0.0557,
+      "loss": 0.0646,
       "step": 129
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9992749020908465e-05,
-      "loss": 0.0272,
+      "loss": 0.035,
       "step": 130
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.999225760794137e-05,
-      "loss": 0.0475,
+      "loss": 0.0769,
       "step": 131
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9991750088708185e-05,
-      "loss": 0.0385,
+      "loss": 0.0521,
       "step": 132
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9991226463754136e-05,
-      "loss": 0.0545,
+      "loss": 0.0789,
       "step": 133
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9990686733641767e-05,
-      "loss": 0.0463,
+      "loss": 0.0498,
       "step": 134
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9990130898950905e-05,
-      "loss": 0.0491,
+      "loss": 0.0568,
       "step": 135
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9989558960278684e-05,
-      "loss": 0.0381,
+      "loss": 0.0403,
       "step": 136
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9988970918239553e-05,
-      "loss": 0.0427,
+      "loss": 0.0638,
       "step": 137
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9988366773465235e-05,
-      "loss": 0.0351,
+      "loss": 0.0348,
       "step": 138
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9987746526604774e-05,
-      "loss": 0.0318,
+      "loss": 0.0336,
       "step": 139
     },
     {
       "epoch": 0.04,
       "learning_rate": 2.9987110178324505e-05,
-      "loss": 0.0783,
+      "loss": 0.0693,
       "step": 140
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.998645772930806e-05,
-      "loss": 0.0405,
+      "loss": 0.0417,
       "step": 141
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9985789180256363e-05,
-      "loss": 0.047,
+      "loss": 0.0548,
       "step": 142
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9985104531887647e-05,
-      "loss": 0.0467,
+      "loss": 0.0431,
       "step": 143
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9984403784937428e-05,
-      "loss": 0.0421,
+      "loss": 0.0409,
       "step": 144
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.998368694015853e-05,
-      "loss": 0.0414,
+      "loss": 0.0375,
       "step": 145
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9982953998321054e-05,
-      "loss": 0.0325,
+      "loss": 0.0366,
       "step": 146
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9982204960212412e-05,
-      "loss": 0.0308,
+      "loss": 0.0332,
       "step": 147
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.998143982663729e-05,
-      "loss": 0.0598,
+      "loss": 0.0604,
       "step": 148
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9980658598417685e-05,
-      "loss": 0.0414,
+      "loss": 0.044,
       "step": 149
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997986127639287e-05,
-      "loss": 0.0437,
+      "loss": 0.0364,
       "step": 150
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997904786141941e-05,
-      "loss": 0.122,
+      "loss": 0.0614,
       "step": 151
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997821835437117e-05,
-      "loss": 0.0243,
+      "loss": 0.0333,
       "step": 152
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9977372756139287e-05,
-      "loss": 0.0841,
+      "loss": 0.0386,
       "step": 153
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9976511067632193e-05,
-      "loss": 0.0478,
+      "loss": 0.0415,
       "step": 154
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997563328977561e-05,
-      "loss": 0.0586,
+      "loss": 0.0488,
       "step": 155
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997473942351253e-05,
-      "loss": 0.0345,
+      "loss": 0.0311,
       "step": 156
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9973829469803243e-05,
-      "loss": 0.0341,
+      "loss": 0.0382,
       "step": 157
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997290342962532e-05,
-      "loss": 0.0518,
+      "loss": 0.0657,
       "step": 158
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9971961303973606e-05,
-      "loss": 0.0557,
+      "loss": 0.0697,
       "step": 159
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.997100309386024e-05,
-      "loss": 0.0346,
+      "loss": 0.0342,
       "step": 160
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9970028800314624e-05,
-      "loss": 0.035,
+      "loss": 0.0317,
       "step": 161
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9969038424383454e-05,
-      "loss": 0.0318,
+      "loss": 0.032,
       "step": 162
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9968031967130686e-05,
-      "loss": 0.0257,
+      "loss": 0.0349,
       "step": 163
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9967009429637575e-05,
-      "loss": 0.0413,
+      "loss": 0.036,
       "step": 164
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.996597081300263e-05,
-      "loss": 0.0321,
+      "loss": 0.0289,
       "step": 165
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.996491611834165e-05,
-      "loss": 0.0581,
+      "loss": 0.0308,
       "step": 166
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.996384534678769e-05,
-      "loss": 0.0432,
+      "loss": 0.0305,
       "step": 167
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9962758499491093e-05,
-      "loss": 0.0337,
+      "loss": 0.0314,
       "step": 168
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.9961655577619466e-05,
-      "loss": 0.04,
+      "loss": 0.0335,
       "step": 169
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.996053658235768e-05,
-      "loss": 0.0582,
+      "loss": 0.0466,
       "step": 170
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.995940151490788e-05,
-      "loss": 0.034,
+      "loss": 0.0363,
       "step": 171
     },
     {
       "epoch": 0.05,
       "learning_rate": 2.995825037648947e-05,
-      "loss": 0.0374,
+      "loss": 0.0358,
       "step": 172
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.995708316833913e-05,
-      "loss": 0.0383,
+      "loss": 0.0453,
       "step": 173
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.99558998917108e-05,
-      "loss": 0.0329,
+      "loss": 0.0344,
       "step": 174
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9954700547875675e-05,
-      "loss": 0.0296,
+      "loss": 0.0267,
       "step": 175
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.995348513812222e-05,
-      "loss": 0.025,
+      "loss": 0.0248,
       "step": 176
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.995225366375616e-05,
-      "loss": 0.038,
+      "loss": 0.0423,
       "step": 177
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9951006126100466e-05,
-      "loss": 0.0183,
+      "loss": 0.0172,
       "step": 178
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9949742526495382e-05,
-      "loss": 0.0243,
+      "loss": 0.0275,
       "step": 179
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.99484628662984e-05,
-      "loss": 0.0314,
+      "loss": 0.0256,
       "step": 180
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.994716714688426e-05,
-      "loss": 0.0416,
+      "loss": 0.0415,
       "step": 181
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.994585536964497e-05,
-      "loss": 0.0354,
+      "loss": 0.0389,
       "step": 182
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9944527535989765e-05,
-      "loss": 0.0348,
+      "loss": 0.0401,
       "step": 183
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.994318364734516e-05,
-      "loss": 0.0303,
+      "loss": 0.0417,
       "step": 184
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9941823705154893e-05,
-      "loss": 0.0376,
+      "loss": 0.0497,
       "step": 185
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9940447710879957e-05,
-      "loss": 0.0301,
+      "loss": 0.0305,
       "step": 186
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9939055665998594e-05,
-      "loss": 0.0261,
+      "loss": 0.0241,
       "step": 187
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9937647572006287e-05,
-      "loss": 0.0314,
+      "loss": 0.0202,
       "step": 188
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9936223430415756e-05,
-      "loss": 0.0208,
+      "loss": 0.0264,
       "step": 189
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9934783242756963e-05,
-      "loss": 0.0325,
+      "loss": 0.0337,
       "step": 190
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9933327010577115e-05,
-      "loss": 0.031,
+      "loss": 0.0415,
       "step": 191
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9931854735440647e-05,
-      "loss": 0.0364,
+      "loss": 0.0406,
       "step": 192
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9930366418929225e-05,
-      "loss": 0.0241,
+      "loss": 0.0346,
       "step": 193
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9928862062641773e-05,
-      "loss": 0.0159,
+      "loss": 0.0243,
       "step": 194
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.992734166819442e-05,
-      "loss": 0.0412,
+      "loss": 0.032,
       "step": 195
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9925805237220526e-05,
-      "loss": 0.0299,
+      "loss": 0.0321,
       "step": 196
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.99242527713707e-05,
-      "loss": 0.0298,
+      "loss": 0.0301,
       "step": 197
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9922684272312758e-05,
-      "loss": 0.0312,
+      "loss": 0.0379,
       "step": 198
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.992109974173175e-05,
-      "loss": 0.0247,
+      "loss": 0.0242,
       "step": 199
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9919499181329948e-05,
-      "loss": 0.0404,
+      "loss": 0.0337,
       "step": 200
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.9917882592826838e-05,
-      "loss": 0.0324,
+      "loss": 0.022,
       "step": 201
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.991624997795913e-05,
-      "loss": 0.0273,
+      "loss": 0.031,
       "step": 202
     },
     {
       "epoch": 0.06,
       "learning_rate": 2.991460133848076e-05,
-      "loss": 0.0294,
+      "loss": 0.0229,
       "step": 203
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9912936676162856e-05,
-      "loss": 0.0319,
+      "loss": 0.0255,
       "step": 204
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9911255992793788e-05,
-      "loss": 0.0331,
+      "loss": 0.0339,
       "step": 205
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9909559290179112e-05,
-      "loss": 0.0251,
+      "loss": 0.0272,
       "step": 206
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.990784657014162e-05,
-      "loss": 0.0266,
+      "loss": 0.0297,
       "step": 207
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9906117834521283e-05,
-      "loss": 0.0342,
+      "loss": 0.033,
       "step": 208
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9904373085175293e-05,
-      "loss": 0.0394,
+      "loss": 0.0328,
       "step": 209
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.990261232397805e-05,
-      "loss": 0.0522,
+      "loss": 0.0357,
       "step": 210
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9900835552821145e-05,
-      "loss": 0.039,
+      "loss": 0.0307,
       "step": 211
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9899042773613376e-05,
-      "loss": 0.0291,
+      "loss": 0.0347,
       "step": 212
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9897233988280735e-05,
-      "loss": 0.0299,
+      "loss": 0.0289,
       "step": 213
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.989540919876641e-05,
-      "loss": 0.0232,
+      "loss": 0.0224,
       "step": 214
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.989356840703078e-05,
-      "loss": 0.0246,
+      "loss": 0.0231,
       "step": 215
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9891711615051417e-05,
-      "loss": 0.0282,
+      "loss": 0.0207,
       "step": 216
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.988983882482309e-05,
-      "loss": 0.0335,
+      "loss": 0.0437,
       "step": 217
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9887950038357747e-05,
-      "loss": 0.0315,
+      "loss": 0.0295,
       "step": 218
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9886045257684514e-05,
-      "loss": 0.0673,
+      "loss": 0.0549,
       "step": 219
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9884124484849716e-05,
-      "loss": 0.0171,
+      "loss": 0.0135,
       "step": 220
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.988218772191684e-05,
-      "loss": 0.0242,
+      "loss": 0.0252,
       "step": 221
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.988023497096657e-05,
-      "loss": 0.0148,
+      "loss": 0.0204,
       "step": 222
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.987826623409676e-05,
-      "loss": 0.0182,
+      "loss": 0.0202,
       "step": 223
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9876281513422426e-05,
-      "loss": 0.0249,
+      "loss": 0.0227,
       "step": 224
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9874280811075767e-05,
-      "loss": 0.0452,
+      "loss": 0.0178,
       "step": 225
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9872264129206155e-05,
-      "loss": 0.0172,
+      "loss": 0.0222,
       "step": 226
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9870231469980112e-05,
-      "loss": 0.0296,
+      "loss": 0.0278,
       "step": 227
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9868182835581344e-05,
-      "loss": 0.0288,
+      "loss": 0.0362,
       "step": 228
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9866118228210704e-05,
-      "loss": 0.0312,
+      "loss": 0.0455,
       "step": 229
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.986403765008622e-05,
-      "loss": 0.0235,
+      "loss": 0.0291,
       "step": 230
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9861941103443057e-05,
-      "loss": 0.0274,
+      "loss": 0.0258,
       "step": 231
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9859828590533554e-05,
-      "loss": 0.028,
+      "loss": 0.0206,
       "step": 232
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.98577001136272e-05,
-      "loss": 0.032,
+      "loss": 0.0354,
       "step": 233
     },
     {
       "epoch": 0.07,
       "learning_rate": 2.9855555675010618e-05,
-      "loss": 0.0339,
+      "loss": 0.0358,
       "step": 234
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9853395276987598e-05,
-      "loss": 0.0301,
+      "loss": 0.0186,
       "step": 235
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9851218921879064e-05,
-      "loss": 0.0351,
+      "loss": 0.0388,
       "step": 236
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9849026612023093e-05,
-      "loss": 0.0246,
+      "loss": 0.0176,
       "step": 237
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.984681834977489e-05,
-      "loss": 0.0215,
+      "loss": 0.0172,
       "step": 238
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9844594137506803e-05,
-      "loss": 0.035,
+      "loss": 0.0294,
       "step": 239
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9842353977608316e-05,
-      "loss": 0.0197,
+      "loss": 0.0321,
       "step": 240
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9840097872486048e-05,
-      "loss": 0.0222,
+      "loss": 0.0187,
       "step": 241
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9837825824563744e-05,
-      "loss": 0.0192,
+      "loss": 0.0195,
       "step": 242
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9835537836282277e-05,
-      "loss": 0.0189,
+      "loss": 0.0166,
       "step": 243
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9833233910099647e-05,
-      "loss": 0.0296,
+      "loss": 0.0283,
       "step": 244
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9830914048490977e-05,
-      "loss": 0.0222,
+      "loss": 0.0284,
       "step": 245
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9828578253948498e-05,
-      "loss": 0.0104,
+      "loss": 0.0158,
       "step": 246
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.982622652898158e-05,
-      "loss": 0.0239,
+      "loss": 0.0315,
       "step": 247
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.982385887611668e-05,
-      "loss": 0.0167,
+      "loss": 0.0121,
       "step": 248
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9821475297897395e-05,
-      "loss": 0.0143,
+      "loss": 0.0182,
       "step": 249
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.981907579688441e-05,
-      "loss": 0.0158,
+      "loss": 0.0227,
       "step": 250
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9816660375655517e-05,
-      "loss": 0.0352,
+      "loss": 0.0497,
       "step": 251
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9814229036805624e-05,
-      "loss": 0.0179,
+      "loss": 0.0186,
       "step": 252
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9811781782946732e-05,
-      "loss": 0.0276,
+      "loss": 0.0299,
       "step": 253
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9809318616707933e-05,
-      "loss": 0.0148,
+      "loss": 0.017,
       "step": 254
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9806839540735426e-05,
-      "loss": 0.0119,
+      "loss": 0.0139,
       "step": 255
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9804344557692492e-05,
-      "loss": 0.0194,
+      "loss": 0.0208,
       "step": 256
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9801833670259516e-05,
-      "loss": 0.0399,
+      "loss": 0.0302,
       "step": 257
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9799306881133942e-05,
-      "loss": 0.0643,
+      "loss": 0.0457,
       "step": 258
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9796764193030325e-05,
-      "loss": 0.0269,
+      "loss": 0.0312,
       "step": 259
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9794205608680285e-05,
-      "loss": 0.0381,
+      "loss": 0.0328,
       "step": 260
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9791631130832525e-05,
-      "loss": 0.0184,
+      "loss": 0.0291,
       "step": 261
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9789040762252818e-05,
-      "loss": 0.0105,
+      "loss": 0.0171,
       "step": 262
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.978643450572402e-05,
-      "loss": 0.0327,
+      "loss": 0.0384,
       "step": 263
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.978381236404603e-05,
-      "loss": 0.0607,
+      "loss": 0.0136,
       "step": 264
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9781174340035845e-05,
-      "loss": 0.0203,
+      "loss": 0.025,
       "step": 265
     },
     {
       "epoch": 0.08,
       "learning_rate": 2.9778520436527504e-05,
-      "loss": 0.0147,
+      "loss": 0.0163,
       "step": 266
     },
     {
@@ -1621,61 +1621,61 @@
     {
       "epoch": 0.09,
       "learning_rate": 2.9773165002437824e-05,
-      "loss": 0.0197,
+      "loss": 0.016,
       "step": 268
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9770463477609862e-05,
-      "loss": 0.0243,
+      "loss": 0.0195,
       "step": 269
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.976774608479048e-05,
-      "loss": 0.0142,
+      "loss": 0.0145,
       "step": 270
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.976501282689899e-05,
-      "loss": 0.0225,
+      "loss": 0.0196,
       "step": 271
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9762263706871746e-05,
-      "loss": 0.0156,
+      "loss": 0.0087,
       "step": 272
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.975949872766215e-05,
-      "loss": 0.0308,
+      "loss": 0.0213,
       "step": 273
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.975671789224063e-05,
-      "loss": 0.0255,
+      "loss": 0.0301,
       "step": 274
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9753921203594646e-05,
-      "loss": 0.0151,
+      "loss": 0.008,
       "step": 275
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9751108664728705e-05,
-      "loss": 0.015,
+      "loss": 0.0115,
       "step": 276
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9748280278664332e-05,
-      "loss": 0.0423,
+      "loss": 0.0288,
       "step": 277
     },
     {
@@ -1687,301 +1687,301 @@
     {
       "epoch": 0.09,
       "learning_rate": 2.9742575977111513e-05,
-      "loss": 0.026,
+      "loss": 0.0207,
       "step": 279
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9739700067751224e-05,
-      "loss": 0.0166,
+      "loss": 0.0132,
       "step": 280
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.973680832344883e-05,
-      "loss": 0.0566,
+      "loss": 0.0368,
       "step": 281
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9733900747310935e-05,
-      "loss": 0.0295,
+      "loss": 0.0173,
       "step": 282
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9730977342461168e-05,
-      "loss": 0.0444,
+      "loss": 0.0315,
       "step": 283
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.972803811204016e-05,
-      "loss": 0.0211,
+      "loss": 0.0176,
       "step": 284
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9725083059205544e-05,
-      "loss": 0.0189,
+      "loss": 0.0151,
       "step": 285
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.972211218713195e-05,
-      "loss": 0.03,
+      "loss": 0.0287,
       "step": 286
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9719125499010995e-05,
-      "loss": 0.0435,
+      "loss": 0.0446,
       "step": 287
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9716122998051304e-05,
-      "loss": 0.025,
+      "loss": 0.0296,
       "step": 288
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.971310468747848e-05,
-      "loss": 0.0349,
+      "loss": 0.0354,
       "step": 289
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.97100705705351e-05,
-      "loss": 0.0243,
+      "loss": 0.0291,
       "step": 290
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9707020650480744e-05,
-      "loss": 0.0421,
+      "loss": 0.0287,
       "step": 291
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9703954930591958e-05,
-      "loss": 0.0177,
+      "loss": 0.014,
       "step": 292
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9700873414162257e-05,
-      "loss": 0.0231,
+      "loss": 0.0144,
       "step": 293
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.969777610450213e-05,
-      "loss": 0.0305,
+      "loss": 0.029,
       "step": 294
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9694663004939043e-05,
-      "loss": 0.0276,
+      "loss": 0.0399,
       "step": 295
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9691534118817406e-05,
-      "loss": 0.0193,
+      "loss": 0.0232,
       "step": 296
     },
     {
       "epoch": 0.09,
       "learning_rate": 2.9688389449498605e-05,
-      "loss": 0.0192,
+      "loss": 0.0171,
       "step": 297
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9685229000360976e-05,
-      "loss": 0.0144,
+      "loss": 0.0158,
       "step": 298
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.96820527747998e-05,
-      "loss": 0.0172,
+      "loss": 0.0184,
       "step": 299
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9678860776227327e-05,
-      "loss": 0.0116,
+      "loss": 0.0202,
       "step": 300
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9675653008072725e-05,
-      "loss": 0.0165,
+      "loss": 0.0221,
       "step": 301
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9672429473782123e-05,
-      "loss": 0.0247,
+      "loss": 0.025,
       "step": 302
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9669190176818585e-05,
-      "loss": 0.0458,
+      "loss": 0.044,
       "step": 303
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9665935120662102e-05,
-      "loss": 0.0223,
+      "loss": 0.0203,
       "step": 304
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.96626643088096e-05,
-      "loss": 0.0159,
+      "loss": 0.0196,
       "step": 305
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.965937774477493e-05,
-      "loss": 0.0162,
+      "loss": 0.0228,
       "step": 306
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9656075432088868e-05,
-      "loss": 0.016,
+      "loss": 0.0225,
       "step": 307
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9652757374299103e-05,
-      "loss": 0.0143,
+      "loss": 0.0138,
       "step": 308
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9649423574970244e-05,
-      "loss": 0.0457,
+      "loss": 0.0498,
       "step": 309
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9646074037683818e-05,
-      "loss": 0.0225,
+      "loss": 0.0244,
       "step": 310
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.964270876603824e-05,
-      "loss": 0.0302,
+      "loss": 0.0249,
       "step": 311
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.963932776364885e-05,
-      "loss": 0.0255,
+      "loss": 0.0154,
       "step": 312
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.963593103414787e-05,
-      "loss": 0.0224,
+      "loss": 0.0217,
       "step": 313
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.963251858118443e-05,
-      "loss": 0.0226,
+      "loss": 0.0217,
       "step": 314
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9629090408424542e-05,
-      "loss": 0.0304,
+      "loss": 0.0164,
       "step": 315
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9625646519551117e-05,
-      "loss": 0.029,
+      "loss": 0.0254,
       "step": 316
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9622186918263947e-05,
-      "loss": 0.0179,
+      "loss": 0.0244,
       "step": 317
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.961871160827969e-05,
-      "loss": 0.0253,
+      "loss": 0.0484,
       "step": 318
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.96152205933319e-05,
-      "loss": 0.0106,
+      "loss": 0.0123,
       "step": 319
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9611713877170995e-05,
-      "loss": 0.023,
+      "loss": 0.0163,
       "step": 320
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.960819146356425e-05,
-      "loss": 0.0167,
+      "loss": 0.0166,
       "step": 321
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9604653356295823e-05,
-      "loss": 0.0303,
+      "loss": 0.0351,
       "step": 322
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9601099559166714e-05,
-      "loss": 0.0204,
+      "loss": 0.0193,
       "step": 323
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.9597530075994798e-05,
-      "loss": 0.0166,
+      "loss": 0.0154,
       "step": 324
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.959394491061478e-05,
-      "loss": 0.0278,
+      "loss": 0.0265,
       "step": 325
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.959034406687823e-05,
-      "loss": 0.0129,
+      "loss": 0.0149,
       "step": 326
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.958672754865355e-05,
-      "loss": 0.0117,
+      "loss": 0.0154,
       "step": 327
     },
     {
       "epoch": 0.1,
       "learning_rate": 2.958309535982599e-05,
-      "loss": 0.033,
+      "loss": 0.0233,
       "step": 328
     },
     {
@@ -1993,1213 +1993,1213 @@
     {
       "epoch": 0.11,
       "learning_rate": 2.9575783985987377e-05,
-      "loss": 0.0182,
+      "loss": 0.0165,
       "step": 330
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.957210480883097e-05,
-      "loss": 0.0168,
+      "loss": 0.0197,
       "step": 331
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9568409976780977e-05,
-      "loss": 0.0146,
+      "loss": 0.0135,
       "step": 332
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9564699493806773e-05,
-      "loss": 0.0174,
+      "loss": 0.0197,
       "step": 333
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.956097336389455e-05,
-      "loss": 0.0275,
+      "loss": 0.028,
       "step": 334
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9557231591047302e-05,
-      "loss": 0.0256,
+      "loss": 0.0195,
       "step": 335
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.955347417928485e-05,
-      "loss": 0.0201,
+      "loss": 0.0238,
       "step": 336
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9549701132643796e-05,
-      "loss": 0.0219,
+      "loss": 0.0254,
       "step": 337
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9545912455177546e-05,
-      "loss": 0.0126,
+      "loss": 0.0119,
       "step": 338
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9542108150956296e-05,
-      "loss": 0.0214,
+      "loss": 0.0176,
       "step": 339
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.953828822406703e-05,
-      "loss": 0.0318,
+      "loss": 0.0144,
       "step": 340
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9534452678613522e-05,
-      "loss": 0.0164,
+      "loss": 0.0097,
       "step": 341
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9530601518716317e-05,
-      "loss": 0.033,
+      "loss": 0.0245,
       "step": 342
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.952673474851274e-05,
-      "loss": 0.0262,
+      "loss": 0.0207,
       "step": 343
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9522852372156875e-05,
-      "loss": 0.0066,
+      "loss": 0.0084,
       "step": 344
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9518954393819587e-05,
-      "loss": 0.017,
+      "loss": 0.0195,
       "step": 345
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9515040817688504e-05,
-      "loss": 0.0161,
+      "loss": 0.0189,
       "step": 346
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9511111647967987e-05,
-      "loss": 0.023,
+      "loss": 0.0297,
       "step": 347
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9507166888879175e-05,
-      "loss": 0.0146,
+      "loss": 0.0153,
       "step": 348
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.950320654465994e-05,
-      "loss": 0.0222,
+      "loss": 0.0125,
       "step": 349
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9499230619564908e-05,
-      "loss": 0.0227,
+      "loss": 0.0207,
       "step": 350
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9495239117865433e-05,
-      "loss": 0.0193,
+      "loss": 0.0239,
       "step": 351
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9491232043849605e-05,
-      "loss": 0.0201,
+      "loss": 0.0242,
       "step": 352
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.948720940182225e-05,
-      "loss": 0.0164,
+      "loss": 0.0145,
       "step": 353
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9483171196104915e-05,
-      "loss": 0.0312,
+      "loss": 0.0393,
       "step": 354
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9479117431035867e-05,
-      "loss": 0.0164,
+      "loss": 0.0172,
       "step": 355
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9475048110970087e-05,
-      "loss": 0.018,
+      "loss": 0.0226,
       "step": 356
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.947096324027927e-05,
-      "loss": 0.0202,
+      "loss": 0.0174,
       "step": 357
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9466862823351814e-05,
-      "loss": 0.0167,
+      "loss": 0.0177,
       "step": 358
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.946274686459282e-05,
-      "loss": 0.016,
+      "loss": 0.023,
       "step": 359
     },
     {
       "epoch": 0.11,
       "learning_rate": 2.9458615368424082e-05,
-      "loss": 0.0278,
+      "loss": 0.0323,
       "step": 360
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9454468339284095e-05,
-      "loss": 0.0166,
+      "loss": 0.0238,
       "step": 361
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9450305781628033e-05,
-      "loss": 0.0171,
+      "loss": 0.0186,
       "step": 362
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9446127699927752e-05,
-      "loss": 0.0171,
+      "loss": 0.0169,
       "step": 363
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9441934098671796e-05,
-      "loss": 0.0122,
+      "loss": 0.0147,
       "step": 364
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.943772498236537e-05,
-      "loss": 0.0386,
+      "loss": 0.0349,
       "step": 365
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9433500355530352e-05,
-      "loss": 0.0223,
+      "loss": 0.0208,
       "step": 366
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.942926022270528e-05,
-      "loss": 0.0271,
+      "loss": 0.0174,
       "step": 367
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9425004588445353e-05,
-      "loss": 0.0113,
+      "loss": 0.0105,
       "step": 368
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9420733457322427e-05,
-      "loss": 0.0129,
+      "loss": 0.0144,
       "step": 369
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9416446833924995e-05,
-      "loss": 0.0389,
+      "loss": 0.0246,
       "step": 370
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9412144722858205e-05,
-      "loss": 0.0182,
+      "loss": 0.018,
       "step": 371
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9407827128743838e-05,
-      "loss": 0.0447,
+      "loss": 0.0373,
       "step": 372
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9403494056220312e-05,
-      "loss": 0.0295,
+      "loss": 0.0207,
       "step": 373
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9399145509942665e-05,
-      "loss": 0.0112,
+      "loss": 0.0127,
       "step": 374
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9394781494582568e-05,
-      "loss": 0.0186,
+      "loss": 0.0151,
       "step": 375
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.939040201482831e-05,
-      "loss": 0.0116,
+      "loss": 0.0124,
       "step": 376
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9386007075384786e-05,
-      "loss": 0.0233,
+      "loss": 0.0184,
       "step": 377
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9381596680973506e-05,
-      "loss": 0.0071,
+      "loss": 0.0192,
       "step": 378
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.937717083633258e-05,
-      "loss": 0.0209,
+      "loss": 0.0383,
       "step": 379
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9372729546216718e-05,
-      "loss": 0.0357,
+      "loss": 0.0283,
       "step": 380
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9368272815397224e-05,
-      "loss": 0.0235,
+      "loss": 0.0258,
       "step": 381
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9363800648661987e-05,
-      "loss": 0.0191,
+      "loss": 0.0203,
       "step": 382
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9359313050815486e-05,
-      "loss": 0.0191,
+      "loss": 0.02,
       "step": 383
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9354810026678766e-05,
-      "loss": 0.0222,
+      "loss": 0.0273,
       "step": 384
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.935029158108945e-05,
-      "loss": 0.0173,
+      "loss": 0.0226,
       "step": 385
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9345757718901736e-05,
-      "loss": 0.0184,
+      "loss": 0.0422,
       "step": 386
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9341208444986375e-05,
-      "loss": 0.0195,
+      "loss": 0.0187,
       "step": 387
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9336643764230676e-05,
-      "loss": 0.0154,
+      "loss": 0.0132,
       "step": 388
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.93320636815385e-05,
-      "loss": 0.0195,
+      "loss": 0.0175,
       "step": 389
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.9327468201830258e-05,
-      "loss": 0.0225,
+      "loss": 0.0291,
       "step": 390
     },
     {
       "epoch": 0.12,
       "learning_rate": 2.93228573300429e-05,
-      "loss": 0.0175,
+      "loss": 0.0225,
       "step": 391
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9318231071129905e-05,
-      "loss": 0.0157,
+      "loss": 0.0171,
       "step": 392
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9313589430061297e-05,
-      "loss": 0.0107,
+      "loss": 0.0177,
       "step": 393
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9308932411823615e-05,
-      "loss": 0.0179,
+      "loss": 0.0127,
       "step": 394
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9304260021419912e-05,
-      "loss": 0.0186,
+      "loss": 0.0238,
       "step": 395
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9299572263869777e-05,
-      "loss": 0.01,
+      "loss": 0.0117,
       "step": 396
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9294869144209282e-05,
-      "loss": 0.0135,
+      "loss": 0.0158,
       "step": 397
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9290150667491022e-05,
-      "loss": 0.023,
+      "loss": 0.0222,
       "step": 398
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9285416838784076e-05,
-      "loss": 0.0162,
+      "loss": 0.019,
       "step": 399
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9280667663174026e-05,
-      "loss": 0.0113,
+      "loss": 0.008,
       "step": 400
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9275903145762942e-05,
-      "loss": 0.0239,
+      "loss": 0.0261,
       "step": 401
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.927112329166937e-05,
-      "loss": 0.0236,
+      "loss": 0.0457,
       "step": 402
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.926632810602833e-05,
-      "loss": 0.0143,
+      "loss": 0.0169,
       "step": 403
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9261517593991315e-05,
-      "loss": 0.0097,
+      "loss": 0.015,
       "step": 404
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9256691760726296e-05,
-      "loss": 0.0248,
+      "loss": 0.0222,
       "step": 405
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.925185061141768e-05,
-      "loss": 0.014,
+      "loss": 0.0111,
       "step": 406
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.924699415126635e-05,
-      "loss": 0.0189,
+      "loss": 0.0179,
       "step": 407
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.924212238548962e-05,
-      "loss": 0.0217,
+      "loss": 0.0224,
       "step": 408
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9237235319321264e-05,
-      "loss": 0.0261,
+      "loss": 0.0294,
       "step": 409
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9232332958011477e-05,
-      "loss": 0.0269,
+      "loss": 0.0225,
       "step": 410
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.92274153068269e-05,
-      "loss": 0.0228,
+      "loss": 0.0207,
       "step": 411
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9222482371050586e-05,
-      "loss": 0.0251,
+      "loss": 0.0222,
       "step": 412
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.921753415598202e-05,
-      "loss": 0.0162,
+      "loss": 0.0169,
       "step": 413
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9212570666937097e-05,
-      "loss": 0.0132,
+      "loss": 0.031,
       "step": 414
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9207591909248117e-05,
-      "loss": 0.0196,
+      "loss": 0.0249,
       "step": 415
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9202597888263793e-05,
-      "loss": 0.017,
+      "loss": 0.0137,
       "step": 416
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9197588609349224e-05,
-      "loss": 0.0134,
+      "loss": 0.0128,
       "step": 417
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.919256407788591e-05,
-      "loss": 0.0115,
+      "loss": 0.0189,
       "step": 418
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9187524299271736e-05,
-      "loss": 0.0098,
+      "loss": 0.009,
       "step": 419
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.918246927892096e-05,
-      "loss": 0.031,
+      "loss": 0.0196,
       "step": 420
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9177399022264214e-05,
-      "loss": 0.0106,
+      "loss": 0.0113,
       "step": 421
     },
     {
       "epoch": 0.13,
       "learning_rate": 2.9172313534748512e-05,
-      "loss": 0.0095,
+      "loss": 0.0129,
       "step": 422
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.916721282183722e-05,
-      "loss": 0.0139,
+      "loss": 0.0156,
       "step": 423
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9162096889010056e-05,
-      "loss": 0.0209,
+      "loss": 0.0185,
       "step": 424
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.91569657417631e-05,
-      "loss": 0.0134,
+      "loss": 0.0131,
       "step": 425
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9151819385608774e-05,
-      "loss": 0.013,
+      "loss": 0.0209,
       "step": 426
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9146657826075833e-05,
-      "loss": 0.0145,
+      "loss": 0.0131,
       "step": 427
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9141481068709375e-05,
-      "loss": 0.0177,
+      "loss": 0.0164,
       "step": 428
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.913628911907081e-05,
-      "loss": 0.0077,
+      "loss": 0.0145,
       "step": 429
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.913108198273789e-05,
-      "loss": 0.0166,
+      "loss": 0.0154,
       "step": 430
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.912585966530466e-05,
-      "loss": 0.0289,
+      "loss": 0.0119,
       "step": 431
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9120622172381485e-05,
-      "loss": 0.0125,
+      "loss": 0.0104,
       "step": 432
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9115369509595046e-05,
-      "loss": 0.0127,
+      "loss": 0.0068,
       "step": 433
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9110101682588294e-05,
-      "loss": 0.0208,
+      "loss": 0.028,
       "step": 434
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9104818697020494e-05,
-      "loss": 0.0051,
+      "loss": 0.0086,
       "step": 435
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.909952055856718e-05,
-      "loss": 0.0125,
+      "loss": 0.0153,
       "step": 436
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9094207272920174e-05,
-      "loss": 0.0151,
+      "loss": 0.0205,
       "step": 437
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9088878845787573e-05,
-      "loss": 0.0102,
+      "loss": 0.0157,
       "step": 438
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9083535282893733e-05,
-      "loss": 0.0255,
+      "loss": 0.0228,
       "step": 439
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9078176589979268e-05,
-      "loss": 0.0203,
+      "loss": 0.0149,
       "step": 440
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9072802772801056e-05,
-      "loss": 0.0205,
+      "loss": 0.0173,
       "step": 441
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9067413837132228e-05,
-      "loss": 0.014,
+      "loss": 0.0139,
       "step": 442
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.906200978876213e-05,
-      "loss": 0.0126,
+      "loss": 0.0113,
       "step": 443
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9056590633496377e-05,
-      "loss": 0.0136,
+      "loss": 0.0102,
       "step": 444
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9051156377156782e-05,
-      "loss": 0.0143,
+      "loss": 0.0072,
       "step": 445
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9045707025581413e-05,
-      "loss": 0.0123,
+      "loss": 0.0157,
       "step": 446
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.904024258462453e-05,
-      "loss": 0.0176,
+      "loss": 0.0172,
       "step": 447
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9034763060156616e-05,
-      "loss": 0.0052,
+      "loss": 0.0058,
       "step": 448
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.902926845806435e-05,
-      "loss": 0.0088,
+      "loss": 0.009,
       "step": 449
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.902375878425062e-05,
-      "loss": 0.0175,
+      "loss": 0.009,
       "step": 450
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.90182340446345e-05,
-      "loss": 0.0181,
+      "loss": 0.0185,
       "step": 451
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.9012694245151238e-05,
-      "loss": 0.0307,
+      "loss": 0.0236,
       "step": 452
     },
     {
       "epoch": 0.14,
       "learning_rate": 2.900713939175228e-05,
-      "loss": 0.0203,
+      "loss": 0.0166,
       "step": 453
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.9001569490405234e-05,
-      "loss": 0.0178,
+      "loss": 0.0212,
       "step": 454
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8995984547093876e-05,
-      "loss": 0.0062,
+      "loss": 0.0061,
       "step": 455
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8990384567818144e-05,
-      "loss": 0.0202,
+      "loss": 0.0222,
       "step": 456
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8984769558594128e-05,
-      "loss": 0.0114,
+      "loss": 0.0131,
       "step": 457
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8979139525454058e-05,
-      "loss": 0.0152,
+      "loss": 0.0161,
       "step": 458
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8973494474446308e-05,
-      "loss": 0.0197,
+      "loss": 0.0145,
       "step": 459
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.89678344116354e-05,
-      "loss": 0.0057,
+      "loss": 0.0085,
       "step": 460
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.896215934310196e-05,
-      "loss": 0.0188,
+      "loss": 0.0205,
       "step": 461
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8956469274942753e-05,
-      "loss": 0.0157,
+      "loss": 0.0083,
       "step": 462
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8950764213270647e-05,
-      "loss": 0.0159,
+      "loss": 0.0116,
       "step": 463
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.894504416421462e-05,
-      "loss": 0.0336,
+      "loss": 0.034,
       "step": 464
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8939309133919762e-05,
-      "loss": 0.0158,
+      "loss": 0.0209,
       "step": 465
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8933559128547236e-05,
-      "loss": 0.011,
+      "loss": 0.0205,
       "step": 466
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8927794154274313e-05,
-      "loss": 0.0174,
+      "loss": 0.0167,
       "step": 467
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8922014217294333e-05,
-      "loss": 0.0249,
+      "loss": 0.0326,
       "step": 468
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8916219323816716e-05,
-      "loss": 0.0119,
+      "loss": 0.0073,
       "step": 469
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8910409480066943e-05,
-      "loss": 0.0128,
+      "loss": 0.0192,
       "step": 470
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.890458469228657e-05,
-      "loss": 0.0078,
+      "loss": 0.0079,
       "step": 471
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8898744966733196e-05,
-      "loss": 0.0122,
+      "loss": 0.0097,
       "step": 472
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8892890309680475e-05,
-      "loss": 0.0193,
+      "loss": 0.0091,
       "step": 473
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.888702072741809e-05,
-      "loss": 0.013,
+      "loss": 0.0112,
       "step": 474
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.888113622625177e-05,
-      "loss": 0.0118,
+      "loss": 0.0088,
       "step": 475
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8875236812503266e-05,
-      "loss": 0.0192,
+      "loss": 0.013,
       "step": 476
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8869322492510348e-05,
-      "loss": 0.011,
+      "loss": 0.0137,
       "step": 477
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.886339327262681e-05,
-      "loss": 0.0167,
+      "loss": 0.0164,
       "step": 478
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8857449159222448e-05,
-      "loss": 0.013,
+      "loss": 0.0152,
       "step": 479
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8851490158683046e-05,
-      "loss": 0.01,
+      "loss": 0.0109,
       "step": 480
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.88455162774104e-05,
-      "loss": 0.0113,
+      "loss": 0.0132,
       "step": 481
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.883952752182228e-05,
-      "loss": 0.0334,
+      "loss": 0.0222,
       "step": 482
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.8833523898352438e-05,
-      "loss": 0.0246,
+      "loss": 0.0159,
       "step": 483
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.88275054134506e-05,
-      "loss": 0.0086,
+      "loss": 0.0154,
       "step": 484
     },
     {
       "epoch": 0.15,
       "learning_rate": 2.882147207358247e-05,
-      "loss": 0.0216,
+      "loss": 0.0214,
       "step": 485
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8815423885229683e-05,
-      "loss": 0.02,
+      "loss": 0.0159,
       "step": 486
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8809360854889848e-05,
-      "loss": 0.0172,
+      "loss": 0.0199,
       "step": 487
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8803282989076513e-05,
-      "loss": 0.015,
+      "loss": 0.0126,
       "step": 488
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8797190294319164e-05,
-      "loss": 0.015,
+      "loss": 0.0128,
       "step": 489
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8791082777163216e-05,
-      "loss": 0.0063,
+      "loss": 0.005,
       "step": 490
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8784960444170004e-05,
-      "loss": 0.0221,
+      "loss": 0.0202,
       "step": 491
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.877882330191679e-05,
-      "loss": 0.0114,
+      "loss": 0.0127,
       "step": 492
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8772671356996744e-05,
-      "loss": 0.0234,
+      "loss": 0.0361,
       "step": 493
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8766504616018925e-05,
-      "loss": 0.0191,
+      "loss": 0.0122,
       "step": 494
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8760323085608305e-05,
-      "loss": 0.0091,
+      "loss": 0.0086,
       "step": 495
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.875412677240573e-05,
-      "loss": 0.0292,
+      "loss": 0.0221,
       "step": 496
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.874791568306794e-05,
-      "loss": 0.0136,
+      "loss": 0.0061,
       "step": 497
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.874168982426754e-05,
-      "loss": 0.013,
+      "loss": 0.0103,
       "step": 498
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8735449202693002e-05,
-      "loss": 0.0137,
+      "loss": 0.015,
       "step": 499
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.872919382504867e-05,
-      "loss": 0.0141,
+      "loss": 0.0191,
       "step": 500
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.872292369805472e-05,
-      "loss": 0.0172,
+      "loss": 0.0231,
       "step": 501
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8716638828447197e-05,
-      "loss": 0.0166,
+      "loss": 0.0142,
       "step": 502
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8710339222977956e-05,
-      "loss": 0.0108,
+      "loss": 0.0071,
       "step": 503
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8704024888414712e-05,
-      "loss": 0.0161,
+      "loss": 0.021,
       "step": 504
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.869769583154098e-05,
-      "loss": 0.0136,
+      "loss": 0.0191,
       "step": 505
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.869135205915611e-05,
-      "loss": 0.011,
+      "loss": 0.0154,
       "step": 506
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8684993578075248e-05,
-      "loss": 0.0127,
+      "loss": 0.0101,
       "step": 507
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8678620395129338e-05,
-      "loss": 0.0146,
+      "loss": 0.0103,
       "step": 508
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8672232517165138e-05,
-      "loss": 0.0225,
+      "loss": 0.0168,
       "step": 509
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8665829951045176e-05,
-      "loss": 0.0114,
+      "loss": 0.0117,
       "step": 510
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8659412703647767e-05,
-      "loss": 0.0249,
+      "loss": 0.0167,
       "step": 511
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8652980781866993e-05,
-      "loss": 0.0144,
+      "loss": 0.0089,
       "step": 512
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8646534192612704e-05,
-      "loss": 0.0113,
+      "loss": 0.0108,
       "step": 513
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8640072942810507e-05,
-      "loss": 0.0174,
+      "loss": 0.0133,
       "step": 514
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.863359703940176e-05,
-      "loss": 0.0081,
+      "loss": 0.0104,
       "step": 515
     },
     {
       "epoch": 0.16,
       "learning_rate": 2.8627106489343557e-05,
-      "loss": 0.0302,
+      "loss": 0.0212,
       "step": 516
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8620601299608738e-05,
-      "loss": 0.0193,
+      "loss": 0.0167,
       "step": 517
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8614081477185866e-05,
-      "loss": 0.0102,
+      "loss": 0.0133,
       "step": 518
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8607547029079216e-05,
-      "loss": 0.0068,
+      "loss": 0.0074,
       "step": 519
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8600997962308786e-05,
-      "loss": 0.0109,
+      "loss": 0.0123,
       "step": 520
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8594434283910277e-05,
-      "loss": 0.014,
+      "loss": 0.0182,
       "step": 521
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8587856000935078e-05,
-      "loss": 0.018,
+      "loss": 0.0162,
       "step": 522
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8581263120450286e-05,
-      "loss": 0.0128,
+      "loss": 0.0125,
       "step": 523
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.857465564953866e-05,
-      "loss": 0.0199,
+      "loss": 0.0143,
       "step": 524
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8568033595298643e-05,
-      "loss": 0.0143,
+      "loss": 0.01,
       "step": 525
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8561396964844354e-05,
-      "loss": 0.0192,
+      "loss": 0.0161,
       "step": 526
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8554745765305553e-05,
-      "loss": 0.0132,
+      "loss": 0.0095,
       "step": 527
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8548080003827668e-05,
-      "loss": 0.0213,
+      "loss": 0.0211,
       "step": 528
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8541399687571753e-05,
-      "loss": 0.0138,
+      "loss": 0.0147,
       "step": 529
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8534704823714524e-05,
-      "loss": 0.0187,
+      "loss": 0.0229,
       "step": 530
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8527995419448298e-05,
-      "loss": 0.0137,
+      "loss": 0.0088,
       "step": 531
     },
     {
@@ -3211,727 +3211,727 @@
     {
       "epoch": 0.17,
       "learning_rate": 2.8514533018536286e-05,
-      "loss": 0.0059,
+      "loss": 0.0124,
       "step": 533
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8507780036353235e-05,
-      "loss": 0.0126,
+      "loss": 0.0082,
       "step": 534
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8501012542686637e-05,
-      "loss": 0.0115,
+      "loss": 0.0086,
       "step": 535
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8494230544806853e-05,
-      "loss": 0.0118,
+      "loss": 0.013,
       "step": 536
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8487434049999824e-05,
-      "loss": 0.0112,
+      "loss": 0.0082,
       "step": 537
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.848062306556705e-05,
-      "loss": 0.0091,
+      "loss": 0.0098,
       "step": 538
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.847379759882563e-05,
-      "loss": 0.0151,
+      "loss": 0.0121,
       "step": 539
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.846695765710818e-05,
-      "loss": 0.0211,
+      "loss": 0.0244,
       "step": 540
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8460103247762904e-05,
-      "loss": 0.011,
+      "loss": 0.0098,
       "step": 541
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.845323437815352e-05,
-      "loss": 0.0119,
+      "loss": 0.0112,
       "step": 542
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8446351055659303e-05,
-      "loss": 0.0119,
+      "loss": 0.0083,
       "step": 543
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8439453287675034e-05,
-      "loss": 0.0155,
+      "loss": 0.0107,
       "step": 544
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.8432541081611036e-05,
-      "loss": 0.0174,
+      "loss": 0.0242,
       "step": 545
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.842561444489312e-05,
-      "loss": 0.0325,
+      "loss": 0.0244,
       "step": 546
     },
     {
       "epoch": 0.17,
       "learning_rate": 2.841867338496262e-05,
-      "loss": 0.0076,
+      "loss": 0.0058,
       "step": 547
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8411717909276348e-05,
-      "loss": 0.0147,
+      "loss": 0.0165,
       "step": 548
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8404748025306614e-05,
-      "loss": 0.0185,
+      "loss": 0.0169,
       "step": 549
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.83977637405412e-05,
-      "loss": 0.018,
+      "loss": 0.0154,
       "step": 550
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8390765062483364e-05,
-      "loss": 0.0121,
+      "loss": 0.0094,
       "step": 551
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8383751998651828e-05,
-      "loss": 0.0148,
+      "loss": 0.0189,
       "step": 552
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.837672455658076e-05,
-      "loss": 0.0096,
+      "loss": 0.0104,
       "step": 553
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.836968274381978e-05,
-      "loss": 0.0466,
+      "loss": 0.053,
       "step": 554
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8362626567933958e-05,
-      "loss": 0.0218,
+      "loss": 0.0136,
       "step": 555
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.835555603650377e-05,
-      "loss": 0.025,
+      "loss": 0.0148,
       "step": 556
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8348471157125132e-05,
-      "loss": 0.0253,
+      "loss": 0.0193,
       "step": 557
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.834137193740937e-05,
-      "loss": 0.0301,
+      "loss": 0.0151,
       "step": 558
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.833425838498322e-05,
-      "loss": 0.0182,
+      "loss": 0.021,
       "step": 559
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.83271305074888e-05,
-      "loss": 0.0093,
+      "loss": 0.0072,
       "step": 560
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.831998831258364e-05,
-      "loss": 0.0092,
+      "loss": 0.0108,
       "step": 561
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8312831807940634e-05,
-      "loss": 0.0127,
+      "loss": 0.0168,
       "step": 562
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.830566100124805e-05,
-      "loss": 0.0158,
+      "loss": 0.0181,
       "step": 563
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8298475900209542e-05,
-      "loss": 0.0208,
+      "loss": 0.0085,
       "step": 564
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8291276512544084e-05,
-      "loss": 0.0065,
+      "loss": 0.0089,
       "step": 565
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8284062845986033e-05,
-      "loss": 0.0122,
+      "loss": 0.0115,
       "step": 566
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8276834908285064e-05,
-      "loss": 0.0122,
+      "loss": 0.0067,
       "step": 567
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8269592707206198e-05,
-      "loss": 0.0156,
+      "loss": 0.0172,
       "step": 568
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.826233625052976e-05,
-      "loss": 0.0137,
+      "loss": 0.0189,
       "step": 569
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.825506554605141e-05,
-      "loss": 0.0112,
+      "loss": 0.0101,
       "step": 570
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8247780601582114e-05,
-      "loss": 0.0105,
+      "loss": 0.0109,
       "step": 571
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8240481424948115e-05,
-      "loss": 0.0165,
+      "loss": 0.0116,
       "step": 572
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8233168023990963e-05,
-      "loss": 0.0155,
+      "loss": 0.0147,
       "step": 573
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8225840406567486e-05,
-      "loss": 0.0188,
+      "loss": 0.0121,
       "step": 574
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.821849858054979e-05,
-      "loss": 0.0262,
+      "loss": 0.0184,
       "step": 575
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.821114255382523e-05,
-      "loss": 0.0166,
+      "loss": 0.0179,
       "step": 576
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8203772334296436e-05,
-      "loss": 0.0074,
+      "loss": 0.0064,
       "step": 577
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.819638792988127e-05,
-      "loss": 0.0223,
+      "loss": 0.0119,
       "step": 578
     },
     {
       "epoch": 0.18,
       "learning_rate": 2.8188989348512842e-05,
-      "loss": 0.0219,
+      "loss": 0.0249,
       "step": 579
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8181576598139484e-05,
-      "loss": 0.013,
+      "loss": 0.0128,
       "step": 580
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8174149686724758e-05,
-      "loss": 0.0112,
+      "loss": 0.0114,
       "step": 581
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8166708622247443e-05,
-      "loss": 0.0107,
+      "loss": 0.0091,
       "step": 582
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8159253412701507e-05,
-      "loss": 0.0159,
+      "loss": 0.0168,
       "step": 583
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8151784066096126e-05,
-      "loss": 0.0153,
+      "loss": 0.027,
       "step": 584
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.814430059045566e-05,
-      "loss": 0.017,
+      "loss": 0.0091,
       "step": 585
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8136802993819654e-05,
-      "loss": 0.0213,
+      "loss": 0.0113,
       "step": 586
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8129291284242814e-05,
-      "loss": 0.0116,
+      "loss": 0.0109,
       "step": 587
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8121765469795006e-05,
-      "loss": 0.0203,
+      "loss": 0.0189,
       "step": 588
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8114225558561265e-05,
-      "loss": 0.0133,
+      "loss": 0.0125,
       "step": 589
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8106671558641753e-05,
-      "loss": 0.0178,
+      "loss": 0.0124,
       "step": 590
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8099103478151777e-05,
-      "loss": 0.0063,
+      "loss": 0.0071,
       "step": 591
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.809152132522177e-05,
-      "loss": 0.0059,
+      "loss": 0.0063,
       "step": 592
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8083925107997276e-05,
-      "loss": 0.0122,
+      "loss": 0.0098,
       "step": 593
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8076314834638954e-05,
-      "loss": 0.0141,
+      "loss": 0.016,
       "step": 594
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8068690513322566e-05,
-      "loss": 0.0086,
+      "loss": 0.0084,
       "step": 595
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.806105215223897e-05,
-      "loss": 0.0175,
+      "loss": 0.007,
       "step": 596
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8053399759594084e-05,
-      "loss": 0.0126,
+      "loss": 0.0143,
       "step": 597
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.804573334360893e-05,
-      "loss": 0.0237,
+      "loss": 0.0219,
       "step": 598
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.803805291251957e-05,
-      "loss": 0.0091,
+      "loss": 0.0079,
       "step": 599
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.803035847457714e-05,
-      "loss": 0.0137,
+      "loss": 0.0179,
       "step": 600
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.802265003804782e-05,
-      "loss": 0.0301,
+      "loss": 0.0359,
       "step": 601
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.8014927611212824e-05,
-      "loss": 0.0145,
+      "loss": 0.0094,
       "step": 602
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.80071912023684e-05,
-      "loss": 0.0164,
+      "loss": 0.0127,
       "step": 603
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.7999440819825813e-05,
-      "loss": 0.0155,
+      "loss": 0.0174,
       "step": 604
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.799167647191134e-05,
-      "loss": 0.0191,
+      "loss": 0.0113,
       "step": 605
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.798389816696627e-05,
-      "loss": 0.0108,
+      "loss": 0.0169,
       "step": 606
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.7976105913346874e-05,
-      "loss": 0.0075,
+      "loss": 0.012,
       "step": 607
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.7968299719424417e-05,
-      "loss": 0.0109,
+      "loss": 0.01,
       "step": 608
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.796047959358513e-05,
-      "loss": 0.0083,
+      "loss": 0.0112,
       "step": 609
     },
     {
       "epoch": 0.19,
       "learning_rate": 2.7952645544230226e-05,
-      "loss": 0.0108,
+      "loss": 0.0121,
       "step": 610
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7944797579775864e-05,
-      "loss": 0.0187,
+      "loss": 0.017,
       "step": 611
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.793693570865316e-05,
-      "loss": 0.022,
+      "loss": 0.0183,
       "step": 612
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.792905993930816e-05,
-      "loss": 0.0155,
+      "loss": 0.0154,
       "step": 613
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7921170280201847e-05,
-      "loss": 0.0117,
+      "loss": 0.0102,
       "step": 614
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7913266739810136e-05,
-      "loss": 0.0101,
+      "loss": 0.0113,
       "step": 615
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7905349326623832e-05,
-      "loss": 0.0293,
+      "loss": 0.0249,
       "step": 616
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7897418049148664e-05,
-      "loss": 0.011,
+      "loss": 0.0191,
       "step": 617
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.788947291590524e-05,
-      "loss": 0.0128,
+      "loss": 0.0142,
       "step": 618
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7881513935429075e-05,
-      "loss": 0.0161,
+      "loss": 0.0171,
       "step": 619
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7873541116270534e-05,
-      "loss": 0.0087,
+      "loss": 0.0082,
       "step": 620
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7865554466994865e-05,
-      "loss": 0.0155,
+      "loss": 0.0165,
       "step": 621
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7857553996182165e-05,
-      "loss": 0.0131,
+      "loss": 0.0159,
       "step": 622
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7849539712427396e-05,
-      "loss": 0.0069,
+      "loss": 0.0059,
       "step": 623
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7841511624340333e-05,
-      "loss": 0.0151,
+      "loss": 0.0269,
       "step": 624
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.783346974054561e-05,
-      "loss": 0.0192,
+      "loss": 0.0198,
       "step": 625
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.782541406968266e-05,
-      "loss": 0.0076,
+      "loss": 0.0074,
       "step": 626
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7817344620405734e-05,
-      "loss": 0.012,
+      "loss": 0.0094,
       "step": 627
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7809261401383894e-05,
-      "loss": 0.0086,
+      "loss": 0.008,
       "step": 628
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7801164421300977e-05,
-      "loss": 0.0131,
+      "loss": 0.0152,
       "step": 629
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.779305368885562e-05,
-      "loss": 0.0109,
+      "loss": 0.0172,
       "step": 630
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7784929212761227e-05,
-      "loss": 0.0152,
+      "loss": 0.0181,
       "step": 631
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7776791001745972e-05,
-      "loss": 0.0155,
+      "loss": 0.0163,
       "step": 632
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7768639064552778e-05,
-      "loss": 0.0135,
+      "loss": 0.0118,
       "step": 633
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.776047340993931e-05,
-      "loss": 0.0088,
+      "loss": 0.0123,
       "step": 634
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.775229404667799e-05,
-      "loss": 0.0176,
+      "loss": 0.0179,
       "step": 635
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.774410098355594e-05,
-      "loss": 0.0167,
+      "loss": 0.0123,
       "step": 636
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7735894229375018e-05,
-      "loss": 0.0085,
+      "loss": 0.0114,
       "step": 637
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7727673792951793e-05,
-      "loss": 0.0117,
+      "loss": 0.0167,
       "step": 638
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7719439683117514e-05,
-      "loss": 0.0063,
+      "loss": 0.0055,
       "step": 639
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7711191908718137e-05,
-      "loss": 0.0101,
+      "loss": 0.0047,
       "step": 640
     },
     {
       "epoch": 0.2,
       "learning_rate": 2.7702930478614297e-05,
-      "loss": 0.0158,
+      "loss": 0.02,
       "step": 641
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7694655401681283e-05,
-      "loss": 0.0093,
+      "loss": 0.0101,
       "step": 642
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.768636668680907e-05,
-      "loss": 0.0073,
+      "loss": 0.0111,
       "step": 643
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7678064342902258e-05,
-      "loss": 0.0082,
+      "loss": 0.0105,
       "step": 644
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7669748378880112e-05,
-      "loss": 0.012,
+      "loss": 0.0124,
       "step": 645
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.766141880367651e-05,
-      "loss": 0.0159,
+      "loss": 0.0164,
       "step": 646
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7653075626239965e-05,
-      "loss": 0.0144,
+      "loss": 0.0148,
       "step": 647
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.76447188555336e-05,
-      "loss": 0.0152,
+      "loss": 0.0154,
       "step": 648
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7636348500535143e-05,
-      "loss": 0.0108,
+      "loss": 0.0093,
       "step": 649
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7627964570236907e-05,
-      "loss": 0.0161,
+      "loss": 0.0197,
       "step": 650
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7619567073645793e-05,
-      "loss": 0.0092,
+      "loss": 0.0102,
       "step": 651
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7611156019783292e-05,
-      "loss": 0.012,
+      "loss": 0.0144,
       "step": 652
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.760273141768543e-05,
-      "loss": 0.0191,
+      "loss": 0.0113,
       "step": 653
     },
     {
@@ -3943,331 +3943,331 @@
     {
       "epoch": 0.21,
       "learning_rate": 2.7585841605000582e-05,
-      "loss": 0.0239,
+      "loss": 0.0122,
       "step": 655
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7577376412558417e-05,
-      "loss": 0.0072,
+      "loss": 0.0075,
       "step": 656
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7568897708170516e-05,
-      "loss": 0.0157,
+      "loss": 0.0144,
       "step": 657
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.75604055009456e-05,
-      "loss": 0.0131,
+      "loss": 0.011,
       "step": 658
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.75518998000069e-05,
-      "loss": 0.016,
+      "loss": 0.0104,
       "step": 659
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7543380614492134e-05,
-      "loss": 0.0123,
+      "loss": 0.0085,
       "step": 660
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.753484795355351e-05,
-      "loss": 0.0124,
+      "loss": 0.0102,
       "step": 661
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7526301826357714e-05,
-      "loss": 0.0046,
+      "loss": 0.0043,
       "step": 662
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.75177422420859e-05,
-      "loss": 0.0095,
+      "loss": 0.0083,
       "step": 663
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7509169209933673e-05,
-      "loss": 0.0224,
+      "loss": 0.0212,
       "step": 664
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7500582739111097e-05,
-      "loss": 0.0166,
+      "loss": 0.0162,
       "step": 665
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7491982838842655e-05,
-      "loss": 0.0132,
+      "loss": 0.0121,
       "step": 666
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7483369518367274e-05,
-      "loss": 0.0071,
+      "loss": 0.0077,
       "step": 667
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7474742786938292e-05,
-      "loss": 0.0093,
+      "loss": 0.008,
       "step": 668
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7466102653823455e-05,
-      "loss": 0.0068,
+      "loss": 0.0066,
       "step": 669
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7457449128304905e-05,
-      "loss": 0.0073,
+      "loss": 0.038,
       "step": 670
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7448782219679172e-05,
-      "loss": 0.0051,
+      "loss": 0.0046,
       "step": 671
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.7440101937257167e-05,
-      "loss": 0.0156,
+      "loss": 0.01,
       "step": 672
     },
     {
       "epoch": 0.21,
       "learning_rate": 2.743140829036417e-05,
-      "loss": 0.0131,
+      "loss": 0.0149,
       "step": 673
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7422701288339804e-05,
-      "loss": 0.0185,
+      "loss": 0.0146,
       "step": 674
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.741398094053805e-05,
-      "loss": 0.009,
+      "loss": 0.0104,
       "step": 675
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.740524725632724e-05,
-      "loss": 0.011,
+      "loss": 0.0073,
       "step": 676
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7396500245090008e-05,
-      "loss": 0.0061,
+      "loss": 0.0055,
       "step": 677
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.738773991622332e-05,
-      "loss": 0.0119,
+      "loss": 0.0107,
       "step": 678
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7378966279138448e-05,
-      "loss": 0.0084,
+      "loss": 0.0093,
       "step": 679
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7370179343260952e-05,
-      "loss": 0.0083,
+      "loss": 0.009,
       "step": 680
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.73613791180307e-05,
-      "loss": 0.0156,
+      "loss": 0.0154,
       "step": 681
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7352565612901806e-05,
-      "loss": 0.0084,
+      "loss": 0.0103,
       "step": 682
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7343738837342685e-05,
-      "loss": 0.0129,
+      "loss": 0.0104,
       "step": 683
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.733489880083598e-05,
-      "loss": 0.012,
+      "loss": 0.011,
       "step": 684
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7326045512878594e-05,
-      "loss": 0.0172,
+      "loss": 0.0213,
       "step": 685
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7317178982981664e-05,
-      "loss": 0.005,
+      "loss": 0.0034,
       "step": 686
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7308299220670553e-05,
-      "loss": 0.0094,
+      "loss": 0.009,
       "step": 687
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7299406235484844e-05,
-      "loss": 0.0153,
+      "loss": 0.0198,
       "step": 688
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.729050003697831e-05,
-      "loss": 0.0077,
+      "loss": 0.0049,
       "step": 689
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.728158063471894e-05,
-      "loss": 0.0146,
+      "loss": 0.0111,
       "step": 690
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7272648038288893e-05,
-      "loss": 0.0108,
+      "loss": 0.0186,
       "step": 691
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7263702257284503e-05,
-      "loss": 0.0071,
+      "loss": 0.0083,
       "step": 692
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.725474330131628e-05,
-      "loss": 0.0148,
+      "loss": 0.0128,
       "step": 693
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7245771180008877e-05,
-      "loss": 0.0458,
+      "loss": 0.024,
       "step": 694
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7236785903001095e-05,
-      "loss": 0.0126,
+      "loss": 0.0196,
       "step": 695
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.722778747994587e-05,
-      "loss": 0.025,
+      "loss": 0.0248,
       "step": 696
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.721877592051025e-05,
-      "loss": 0.0062,
+      "loss": 0.0052,
       "step": 697
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7209751234375417e-05,
-      "loss": 0.0127,
+      "loss": 0.0195,
       "step": 698
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7200713431236624e-05,
-      "loss": 0.0099,
+      "loss": 0.0091,
       "step": 699
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7191662520803248e-05,
-      "loss": 0.0118,
+      "loss": 0.0076,
       "step": 700
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7182598512798728e-05,
-      "loss": 0.0126,
+      "loss": 0.0081,
       "step": 701
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7173521416960575e-05,
-      "loss": 0.0069,
+      "loss": 0.0066,
       "step": 702
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.7164431243040368e-05,
-      "loss": 0.0099,
+      "loss": 0.0109,
       "step": 703
     },
     {
       "epoch": 0.22,
       "learning_rate": 2.715532800080373e-05,
-      "loss": 0.0129,
+      "loss": 0.0141,
       "step": 704
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7146211700030323e-05,
-      "loss": 0.0082,
+      "loss": 0.0077,
       "step": 705
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7137082350513844e-05,
-      "loss": 0.0247,
+      "loss": 0.017,
       "step": 706
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7127939962062002e-05,
-      "loss": 0.0227,
+      "loss": 0.0173,
       "step": 707
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.711878454449652e-05,
-      "loss": 0.0114,
+      "loss": 0.012,
       "step": 708
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.710961610765311e-05,
-      "loss": 0.0095,
+      "loss": 0.0123,
       "step": 709
     },
     {
@@ -4279,507 +4279,507 @@
     {
       "epoch": 0.23,
       "learning_rate": 2.7091240215545282e-05,
-      "loss": 0.005,
+      "loss": 0.0064,
       "step": 711
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7082032780022202e-05,
-      "loss": 0.0059,
+      "loss": 0.005,
       "step": 712
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7072812364703817e-05,
-      "loss": 0.0178,
+      "loss": 0.0215,
       "step": 713
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7063578979495672e-05,
-      "loss": 0.0095,
+      "loss": 0.0071,
       "step": 714
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.705433263431725e-05,
-      "loss": 0.0147,
+      "loss": 0.0123,
       "step": 715
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7045073339101945e-05,
-      "loss": 0.0203,
+      "loss": 0.0152,
       "step": 716
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.703580110379707e-05,
-      "loss": 0.0088,
+      "loss": 0.008,
       "step": 717
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7026515938363845e-05,
-      "loss": 0.0279,
+      "loss": 0.0178,
       "step": 718
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.7017217852777368e-05,
-      "loss": 0.0217,
+      "loss": 0.0142,
       "step": 719
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.700790685702663e-05,
-      "loss": 0.011,
+      "loss": 0.0125,
       "step": 720
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6998582961114485e-05,
-      "loss": 0.0122,
+      "loss": 0.0143,
       "step": 721
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6989246175057646e-05,
-      "loss": 0.0187,
+      "loss": 0.0132,
       "step": 722
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.697989650888667e-05,
-      "loss": 0.0131,
+      "loss": 0.0129,
       "step": 723
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6970533972645957e-05,
-      "loss": 0.0066,
+      "loss": 0.006,
       "step": 724
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6961158576393737e-05,
-      "loss": 0.0195,
+      "loss": 0.0254,
       "step": 725
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.695177033020205e-05,
-      "loss": 0.0117,
+      "loss": 0.0151,
       "step": 726
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6942369244156732e-05,
-      "loss": 0.0179,
+      "loss": 0.0212,
       "step": 727
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6932955328357433e-05,
-      "loss": 0.0087,
+      "loss": 0.0102,
       "step": 728
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.692352859291757e-05,
-      "loss": 0.0142,
+      "loss": 0.0129,
       "step": 729
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6914089047964342e-05,
-      "loss": 0.0136,
+      "loss": 0.0082,
       "step": 730
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.69046367036387e-05,
-      "loss": 0.0181,
+      "loss": 0.0212,
       "step": 731
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.689517157009535e-05,
-      "loss": 0.0166,
+      "loss": 0.0207,
       "step": 732
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6885693657502742e-05,
-      "loss": 0.0128,
+      "loss": 0.0109,
       "step": 733
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.687620297604305e-05,
-      "loss": 0.02,
+      "loss": 0.0191,
       "step": 734
     },
     {
       "epoch": 0.23,
       "learning_rate": 2.6866699535912164e-05,
-      "loss": 0.0177,
+      "loss": 0.0208,
       "step": 735
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.685718334731968e-05,
-      "loss": 0.014,
+      "loss": 0.0111,
       "step": 736
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6847654420488895e-05,
-      "loss": 0.0084,
+      "loss": 0.0125,
       "step": 737
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6838112765656794e-05,
-      "loss": 0.0105,
+      "loss": 0.0094,
       "step": 738
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6828558393074022e-05,
-      "loss": 0.0154,
+      "loss": 0.0069,
       "step": 739
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.68189913130049e-05,
-      "loss": 0.0042,
+      "loss": 0.0052,
       "step": 740
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6809411535727388e-05,
-      "loss": 0.0083,
+      "loss": 0.0082,
       "step": 741
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6799819071533104e-05,
-      "loss": 0.0125,
+      "loss": 0.0104,
       "step": 742
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.679021393072729e-05,
-      "loss": 0.0105,
+      "loss": 0.0119,
       "step": 743
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6780596123628783e-05,
-      "loss": 0.0118,
+      "loss": 0.0143,
       "step": 744
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.677096566057006e-05,
-      "loss": 0.0077,
+      "loss": 0.0071,
       "step": 745
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6761322551897182e-05,
-      "loss": 0.0143,
+      "loss": 0.0124,
       "step": 746
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.675166680796979e-05,
-      "loss": 0.0135,
+      "loss": 0.012,
       "step": 747
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6741998439161105e-05,
-      "loss": 0.0122,
+      "loss": 0.0132,
       "step": 748
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6732317455857907e-05,
-      "loss": 0.0138,
+      "loss": 0.0078,
       "step": 749
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.672262386846054e-05,
-      "loss": 0.0125,
+      "loss": 0.0088,
       "step": 750
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6712917687382872e-05,
-      "loss": 0.0074,
+      "loss": 0.0042,
       "step": 751
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6703198923052304e-05,
-      "loss": 0.0149,
+      "loss": 0.0108,
       "step": 752
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6693467585909767e-05,
-      "loss": 0.0096,
+      "loss": 0.0084,
       "step": 753
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6683723686409686e-05,
-      "loss": 0.0088,
+      "loss": 0.0074,
       "step": 754
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6673967235019986e-05,
-      "loss": 0.0171,
+      "loss": 0.0157,
       "step": 755
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6664198242222077e-05,
-      "loss": 0.0552,
+      "loss": 0.0261,
       "step": 756
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6654416718510846e-05,
-      "loss": 0.0101,
+      "loss": 0.0099,
       "step": 757
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6644622674394635e-05,
-      "loss": 0.0089,
+      "loss": 0.0097,
       "step": 758
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6634816120395242e-05,
-      "loss": 0.0148,
+      "loss": 0.0151,
       "step": 759
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6624997067047908e-05,
-      "loss": 0.0113,
+      "loss": 0.0106,
       "step": 760
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6615165524901286e-05,
-      "loss": 0.0091,
+      "loss": 0.0086,
       "step": 761
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6605321504517464e-05,
-      "loss": 0.0113,
+      "loss": 0.0111,
       "step": 762
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.659546501647193e-05,
-      "loss": 0.0068,
+      "loss": 0.0091,
       "step": 763
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6585596071353555e-05,
-      "loss": 0.0187,
+      "loss": 0.0105,
       "step": 764
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6575714679764613e-05,
-      "loss": 0.0077,
+      "loss": 0.0058,
       "step": 765
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6565820852320727e-05,
-      "loss": 0.0168,
+      "loss": 0.0197,
       "step": 766
     },
     {
       "epoch": 0.24,
       "learning_rate": 2.6555914599650903e-05,
-      "loss": 0.0338,
+      "loss": 0.0277,
       "step": 767
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6545995932397466e-05,
-      "loss": 0.0112,
+      "loss": 0.012,
       "step": 768
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6536064861216117e-05,
-      "loss": 0.0118,
+      "loss": 0.0146,
       "step": 769
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6526121396775848e-05,
-      "loss": 0.0082,
+      "loss": 0.0067,
       "step": 770
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.651616554975898e-05,
-      "loss": 0.0088,
+      "loss": 0.009,
       "step": 771
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6506197330861138e-05,
-      "loss": 0.0061,
+      "loss": 0.0064,
       "step": 772
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6496216750791237e-05,
-      "loss": 0.0179,
+      "loss": 0.0127,
       "step": 773
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.648622382027146e-05,
-      "loss": 0.0052,
+      "loss": 0.0044,
       "step": 774
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.647621855003728e-05,
-      "loss": 0.0088,
+      "loss": 0.0077,
       "step": 775
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.646620095083741e-05,
-      "loss": 0.0203,
+      "loss": 0.0184,
       "step": 776
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6456171033433807e-05,
-      "loss": 0.0113,
+      "loss": 0.0192,
       "step": 777
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6446128808601677e-05,
-      "loss": 0.0066,
+      "loss": 0.0061,
       "step": 778
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6436074287129432e-05,
-      "loss": 0.0082,
+      "loss": 0.0084,
       "step": 779
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6426007479818703e-05,
-      "loss": 0.0225,
+      "loss": 0.0136,
       "step": 780
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6415928397484314e-05,
-      "loss": 0.0123,
+      "loss": 0.0102,
       "step": 781
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6405837050954285e-05,
-      "loss": 0.0101,
+      "loss": 0.0076,
       "step": 782
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6395733451069804e-05,
-      "loss": 0.0111,
+      "loss": 0.0107,
       "step": 783
     },
     {
       "epoch": 0.25,
-      "eval_loss": 0.014288174919784069,
-      "eval_runtime": 28.1407,
-      "eval_samples_per_second": 187.38,
-      "eval_steps_per_second": 5.863,
+      "eval_loss": 0.013721341267228127,
+      "eval_runtime": 28.4521,
+      "eval_samples_per_second": 185.329,
+      "eval_steps_per_second": 11.598,
       "step": 783
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.638561760868522e-05,
-      "loss": 0.0074,
+      "loss": 0.0058,
       "step": 784
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.637548953466805e-05,
-      "loss": 0.0172,
+      "loss": 0.0105,
       "step": 785
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.636534923989893e-05,
-      "loss": 0.0066,
+      "loss": 0.0162,
       "step": 786
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6355196735271636e-05,
-      "loss": 0.0108,
+      "loss": 0.0125,
       "step": 787
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6345032031693063e-05,
-      "loss": 0.0093,
+      "loss": 0.0069,
       "step": 788
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6334855140083208e-05,
-      "loss": 0.0078,
+      "loss": 0.0129,
       "step": 789
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6324666071375166e-05,
-      "loss": 0.0169,
+      "loss": 0.0152,
       "step": 790
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6314464836515107e-05,
-      "loss": 0.0134,
+      "loss": 0.0136,
       "step": 791
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.630425144646228e-05,
-      "loss": 0.0113,
+      "loss": 0.0145,
       "step": 792
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6294025912188976e-05,
-      "loss": 0.0102,
+      "loss": 0.0127,
       "step": 793
     },
     {
@@ -4791,1147 +4791,1147 @@
     {
       "epoch": 0.25,
       "learning_rate": 2.627353845493539e-05,
-      "loss": 0.0193,
+      "loss": 0.0176,
       "step": 795
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6263276553964895e-05,
-      "loss": 0.0247,
+      "loss": 0.0143,
       "step": 796
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.6253002552793484e-05,
-      "loss": 0.0141,
+      "loss": 0.014,
       "step": 797
     },
     {
       "epoch": 0.25,
       "learning_rate": 2.624271646245858e-05,
-      "loss": 0.008,
+      "loss": 0.0096,
       "step": 798
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.623241829401058e-05,
-      "loss": 0.0055,
+      "loss": 0.0052,
       "step": 799
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6222108058512866e-05,
-      "loss": 0.0133,
+      "loss": 0.0176,
       "step": 800
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.621178576704178e-05,
-      "loss": 0.0175,
+      "loss": 0.0271,
       "step": 801
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6201451430686617e-05,
-      "loss": 0.0083,
+      "loss": 0.0064,
       "step": 802
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.619110506054962e-05,
-      "loss": 0.0191,
+      "loss": 0.0186,
       "step": 803
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6180746667745946e-05,
-      "loss": 0.0119,
+      "loss": 0.0103,
       "step": 804
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6170376263403676e-05,
-      "loss": 0.0058,
+      "loss": 0.0053,
       "step": 805
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.615999385866379e-05,
-      "loss": 0.0229,
+      "loss": 0.0249,
       "step": 806
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6149599464680166e-05,
-      "loss": 0.0096,
+      "loss": 0.0109,
       "step": 807
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.613919309261956e-05,
-      "loss": 0.0176,
+      "loss": 0.0228,
       "step": 808
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6128774753661594e-05,
-      "loss": 0.0277,
+      "loss": 0.0269,
       "step": 809
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6118344458998745e-05,
-      "loss": 0.0112,
+      "loss": 0.0077,
       "step": 810
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6107902219836346e-05,
-      "loss": 0.0064,
+      "loss": 0.0056,
       "step": 811
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6097448047392544e-05,
-      "loss": 0.0074,
+      "loss": 0.0047,
       "step": 812
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6086981952898324e-05,
-      "loss": 0.0128,
+      "loss": 0.013,
       "step": 813
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6076503947597457e-05,
-      "loss": 0.0169,
+      "loss": 0.0136,
       "step": 814
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6066014042746533e-05,
-      "loss": 0.0092,
+      "loss": 0.0085,
       "step": 815
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6055512249614915e-05,
-      "loss": 0.0197,
+      "loss": 0.014,
       "step": 816
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.604499857948473e-05,
-      "loss": 0.0095,
+      "loss": 0.0083,
       "step": 817
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.603447304365088e-05,
-      "loss": 0.0069,
+      "loss": 0.0075,
       "step": 818
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6023935653421e-05,
-      "loss": 0.024,
+      "loss": 0.0122,
       "step": 819
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6013386420115476e-05,
-      "loss": 0.0122,
+      "loss": 0.0113,
       "step": 820
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.6002825355067402e-05,
-      "loss": 0.0146,
+      "loss": 0.019,
       "step": 821
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.599225246962259e-05,
-      "loss": 0.0051,
+      "loss": 0.02,
       "step": 822
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.5981667775139554e-05,
-      "loss": 0.0094,
+      "loss": 0.0089,
       "step": 823
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.597107128298949e-05,
-      "loss": 0.0091,
+      "loss": 0.0125,
       "step": 824
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.5960463004556257e-05,
-      "loss": 0.0146,
+      "loss": 0.0123,
       "step": 825
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.5949842951236402e-05,
-      "loss": 0.0056,
+      "loss": 0.0058,
       "step": 826
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.5939211134439105e-05,
-      "loss": 0.016,
+      "loss": 0.0192,
       "step": 827
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.592856756558618e-05,
-      "loss": 0.0095,
+      "loss": 0.0081,
       "step": 828
     },
     {
       "epoch": 0.26,
       "learning_rate": 2.5917912256112074e-05,
-      "loss": 0.0166,
+      "loss": 0.0178,
       "step": 829
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5907245217463852e-05,
-      "loss": 0.008,
+      "loss": 0.0117,
       "step": 830
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5896566461101168e-05,
-      "loss": 0.0271,
+      "loss": 0.02,
       "step": 831
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.588587599849627e-05,
-      "loss": 0.0119,
+      "loss": 0.0141,
       "step": 832
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5875173841133985e-05,
-      "loss": 0.0107,
+      "loss": 0.0085,
       "step": 833
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5864460000511693e-05,
-      "loss": 0.0059,
+      "loss": 0.0067,
       "step": 834
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5853734488139337e-05,
-      "loss": 0.0123,
+      "loss": 0.0174,
       "step": 835
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5842997315539395e-05,
-      "loss": 0.0084,
+      "loss": 0.0068,
       "step": 836
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5832248494246873e-05,
-      "loss": 0.0148,
+      "loss": 0.0196,
       "step": 837
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.582148803580929e-05,
-      "loss": 0.0093,
+      "loss": 0.0106,
       "step": 838
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5810715951786665e-05,
-      "loss": 0.0171,
+      "loss": 0.0115,
       "step": 839
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5799932253751506e-05,
-      "loss": 0.0124,
+      "loss": 0.0114,
       "step": 840
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.57891369532888e-05,
-      "loss": 0.0095,
+      "loss": 0.0087,
       "step": 841
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5778330061996007e-05,
-      "loss": 0.02,
+      "loss": 0.0133,
       "step": 842
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5767511591483016e-05,
-      "loss": 0.0099,
+      "loss": 0.0071,
       "step": 843
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.575668155337219e-05,
-      "loss": 0.0141,
+      "loss": 0.0122,
       "step": 844
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5745839959298277e-05,
-      "loss": 0.0098,
+      "loss": 0.0043,
       "step": 845
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5734986820908478e-05,
-      "loss": 0.0101,
+      "loss": 0.0091,
       "step": 846
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.572412214986238e-05,
-      "loss": 0.0094,
+      "loss": 0.0125,
       "step": 847
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5713245957831948e-05,
-      "loss": 0.0082,
+      "loss": 0.0128,
       "step": 848
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5702358256501544e-05,
-      "loss": 0.0124,
+      "loss": 0.0117,
       "step": 849
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5691459057567887e-05,
-      "loss": 0.0066,
+      "loss": 0.0059,
       "step": 850
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5680548372740047e-05,
-      "loss": 0.0093,
+      "loss": 0.013,
       "step": 851
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5669626213739427e-05,
-      "loss": 0.0083,
+      "loss": 0.006,
       "step": 852
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5658692592299772e-05,
-      "loss": 0.0069,
+      "loss": 0.0074,
       "step": 853
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5647747520167135e-05,
-      "loss": 0.0062,
+      "loss": 0.0082,
       "step": 854
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5636791009099855e-05,
-      "loss": 0.0108,
+      "loss": 0.0111,
       "step": 855
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.562582307086858e-05,
-      "loss": 0.0074,
+      "loss": 0.0065,
       "step": 856
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5614843717256233e-05,
-      "loss": 0.0104,
+      "loss": 0.0101,
       "step": 857
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5603852960057984e-05,
-      "loss": 0.0055,
+      "loss": 0.0074,
       "step": 858
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5592850811081274e-05,
-      "loss": 0.0068,
+      "loss": 0.0081,
       "step": 859
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.558183728214577e-05,
-      "loss": 0.0104,
+      "loss": 0.0099,
       "step": 860
     },
     {
       "epoch": 0.27,
       "learning_rate": 2.5570812385083365e-05,
-      "loss": 0.0094,
+      "loss": 0.0063,
       "step": 861
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5559776131738167e-05,
-      "loss": 0.0078,
+      "loss": 0.012,
       "step": 862
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5548728533966495e-05,
-      "loss": 0.0084,
+      "loss": 0.0088,
       "step": 863
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5537669603636834e-05,
-      "loss": 0.0118,
+      "loss": 0.0127,
       "step": 864
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.552659935262986e-05,
-      "loss": 0.0138,
+      "loss": 0.0155,
       "step": 865
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5515517792838397e-05,
-      "loss": 0.0069,
+      "loss": 0.0049,
       "step": 866
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.550442493616744e-05,
-      "loss": 0.0195,
+      "loss": 0.019,
       "step": 867
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5493320794534096e-05,
-      "loss": 0.0149,
+      "loss": 0.0192,
       "step": 868
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.548220537986761e-05,
-      "loss": 0.0181,
+      "loss": 0.0164,
       "step": 869
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5471078704109335e-05,
-      "loss": 0.0063,
+      "loss": 0.0054,
       "step": 870
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5459940779212726e-05,
-      "loss": 0.0112,
+      "loss": 0.0108,
       "step": 871
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5448791617143304e-05,
-      "loss": 0.0156,
+      "loss": 0.0184,
       "step": 872
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.543763122987869e-05,
-      "loss": 0.0118,
+      "loss": 0.0158,
       "step": 873
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5426459629408546e-05,
-      "loss": 0.0059,
+      "loss": 0.0067,
       "step": 874
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5415276827734577e-05,
-      "loss": 0.0126,
+      "loss": 0.0113,
       "step": 875
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.540408283687054e-05,
-      "loss": 0.0057,
+      "loss": 0.0051,
       "step": 876
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.53928776688422e-05,
-      "loss": 0.0091,
+      "loss": 0.0104,
       "step": 877
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5381661335687324e-05,
-      "loss": 0.0097,
+      "loss": 0.0078,
       "step": 878
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5370433849455686e-05,
-      "loss": 0.0085,
+      "loss": 0.0067,
       "step": 879
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5359195222209044e-05,
-      "loss": 0.0076,
+      "loss": 0.0057,
       "step": 880
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5347945466021107e-05,
-      "loss": 0.0106,
+      "loss": 0.015,
       "step": 881
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5336684592977558e-05,
-      "loss": 0.0109,
+      "loss": 0.0064,
       "step": 882
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5325412615176017e-05,
-      "loss": 0.0077,
+      "loss": 0.0087,
       "step": 883
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.531412954472602e-05,
-      "loss": 0.0026,
+      "loss": 0.0046,
       "step": 884
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.530283539374905e-05,
-      "loss": 0.0084,
+      "loss": 0.0098,
       "step": 885
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5291530174378465e-05,
-      "loss": 0.0138,
+      "loss": 0.0149,
       "step": 886
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.528021389875953e-05,
-      "loss": 0.0139,
+      "loss": 0.0156,
       "step": 887
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5268886579049378e-05,
-      "loss": 0.0119,
+      "loss": 0.0161,
       "step": 888
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.525754822741702e-05,
-      "loss": 0.0097,
+      "loss": 0.0116,
       "step": 889
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5246198856043304e-05,
-      "loss": 0.0235,
+      "loss": 0.0327,
       "step": 890
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5234838477120928e-05,
-      "loss": 0.0081,
+      "loss": 0.0085,
       "step": 891
     },
     {
       "epoch": 0.28,
       "learning_rate": 2.5223467102854404e-05,
-      "loss": 0.0078,
+      "loss": 0.0074,
       "step": 892
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5212084745460068e-05,
-      "loss": 0.0117,
+      "loss": 0.0108,
       "step": 893
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.520069141716605e-05,
-      "loss": 0.0058,
+      "loss": 0.0068,
       "step": 894
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5189287130212265e-05,
-      "loss": 0.018,
+      "loss": 0.0151,
       "step": 895
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.51778718968504e-05,
-      "loss": 0.0094,
+      "loss": 0.0138,
       "step": 896
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.516644572934391e-05,
-      "loss": 0.0103,
+      "loss": 0.0102,
       "step": 897
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5155008639967982e-05,
-      "loss": 0.0097,
+      "loss": 0.0141,
       "step": 898
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5143560641009552e-05,
-      "loss": 0.0179,
+      "loss": 0.0217,
       "step": 899
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5132101744767267e-05,
-      "loss": 0.0124,
+      "loss": 0.014,
       "step": 900
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.512063196355148e-05,
-      "loss": 0.0096,
+      "loss": 0.0116,
       "step": 901
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5109151309684247e-05,
-      "loss": 0.0148,
+      "loss": 0.0111,
       "step": 902
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5097659795499297e-05,
-      "loss": 0.0104,
+      "loss": 0.0088,
       "step": 903
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5086157433342028e-05,
-      "loss": 0.011,
+      "loss": 0.0085,
       "step": 904
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5074644235569487e-05,
-      "loss": 0.0054,
+      "loss": 0.0075,
       "step": 905
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.506312021455037e-05,
-      "loss": 0.0114,
+      "loss": 0.0103,
       "step": 906
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5051585382665e-05,
-      "loss": 0.0113,
+      "loss": 0.0104,
       "step": 907
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.504003975230531e-05,
-      "loss": 0.01,
+      "loss": 0.0101,
       "step": 908
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5028483335874828e-05,
-      "loss": 0.0077,
+      "loss": 0.0117,
       "step": 909
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.501691614578869e-05,
-      "loss": 0.0062,
+      "loss": 0.0044,
       "step": 910
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.5005338194473577e-05,
-      "loss": 0.0142,
+      "loss": 0.0126,
       "step": 911
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4993749494367762e-05,
-      "loss": 0.0034,
+      "loss": 0.0059,
       "step": 912
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4982150057921044e-05,
-      "loss": 0.0136,
+      "loss": 0.0138,
       "step": 913
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4970539897594757e-05,
-      "loss": 0.0156,
+      "loss": 0.0081,
       "step": 914
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4958919025861764e-05,
-      "loss": 0.0041,
+      "loss": 0.0087,
       "step": 915
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.494728745520644e-05,
-      "loss": 0.0165,
+      "loss": 0.0134,
       "step": 916
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.493564519812464e-05,
-      "loss": 0.0151,
+      "loss": 0.0083,
       "step": 917
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4923992267123707e-05,
-      "loss": 0.0084,
+      "loss": 0.0101,
       "step": 918
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4912328674722446e-05,
-      "loss": 0.0133,
+      "loss": 0.0148,
       "step": 919
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4900654433451128e-05,
-      "loss": 0.0081,
+      "loss": 0.0088,
       "step": 920
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4888969555851447e-05,
-      "loss": 0.011,
+      "loss": 0.0097,
       "step": 921
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4877274054476536e-05,
-      "loss": 0.0067,
+      "loss": 0.0071,
       "step": 922
     },
     {
       "epoch": 0.29,
       "learning_rate": 2.4865567941890938e-05,
-      "loss": 0.0117,
+      "loss": 0.0148,
       "step": 923
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4853851230670594e-05,
-      "loss": 0.0122,
+      "loss": 0.016,
       "step": 924
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.484212393340283e-05,
-      "loss": 0.0176,
+      "loss": 0.0137,
       "step": 925
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4830386062686358e-05,
-      "loss": 0.0095,
+      "loss": 0.01,
       "step": 926
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.481863763113122e-05,
-      "loss": 0.0097,
+      "loss": 0.017,
       "step": 927
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4806878651358834e-05,
-      "loss": 0.0158,
+      "loss": 0.0147,
       "step": 928
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4795109136001938e-05,
-      "loss": 0.0096,
+      "loss": 0.0087,
       "step": 929
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4783329097704583e-05,
-      "loss": 0.0078,
+      "loss": 0.0092,
       "step": 930
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.477153854912213e-05,
-      "loss": 0.0181,
+      "loss": 0.0134,
       "step": 931
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.475973750292124e-05,
-      "loss": 0.007,
+      "loss": 0.0077,
       "step": 932
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.474792597177983e-05,
-      "loss": 0.0054,
+      "loss": 0.0056,
       "step": 933
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.47361039683871e-05,
-      "loss": 0.0084,
+      "loss": 0.0063,
       "step": 934
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4724271505443496e-05,
-      "loss": 0.0121,
+      "loss": 0.0157,
       "step": 935
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4712428595660698e-05,
-      "loss": 0.007,
+      "loss": 0.0061,
       "step": 936
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.470057525176161e-05,
-      "loss": 0.0105,
+      "loss": 0.009,
       "step": 937
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4688711486480342e-05,
-      "loss": 0.0074,
+      "loss": 0.0088,
       "step": 938
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4676837312562207e-05,
-      "loss": 0.0065,
+      "loss": 0.0067,
       "step": 939
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4664952742763697e-05,
-      "loss": 0.0107,
+      "loss": 0.0116,
       "step": 940
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4653057789852465e-05,
-      "loss": 0.008,
+      "loss": 0.0077,
       "step": 941
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4641152466607336e-05,
-      "loss": 0.0088,
+      "loss": 0.0064,
       "step": 942
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4629236785818262e-05,
-      "loss": 0.0056,
+      "loss": 0.0087,
       "step": 943
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.461731076028632e-05,
-      "loss": 0.0286,
+      "loss": 0.02,
       "step": 944
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.460537440282371e-05,
-      "loss": 0.014,
+      "loss": 0.0117,
       "step": 945
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4593427726253724e-05,
-      "loss": 0.0073,
+      "loss": 0.007,
       "step": 946
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4581470743410755e-05,
-      "loss": 0.0122,
+      "loss": 0.0061,
       "step": 947
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4569503467140236e-05,
-      "loss": 0.0138,
+      "loss": 0.0089,
       "step": 948
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4557525910298697e-05,
-      "loss": 0.0117,
+      "loss": 0.0142,
       "step": 949
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4545538085753685e-05,
-      "loss": 0.0132,
+      "loss": 0.0108,
       "step": 950
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4533540006383793e-05,
-      "loss": 0.0091,
+      "loss": 0.0071,
       "step": 951
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4521531685078622e-05,
-      "loss": 0.008,
+      "loss": 0.0072,
       "step": 952
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4509513134738773e-05,
-      "loss": 0.0138,
+      "loss": 0.0153,
       "step": 953
     },
     {
       "epoch": 0.3,
       "learning_rate": 2.4497484368275844e-05,
-      "loss": 0.0053,
+      "loss": 0.0045,
       "step": 954
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4485445398612415e-05,
-      "loss": 0.0135,
+      "loss": 0.0107,
       "step": 955
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.447339623868201e-05,
-      "loss": 0.0233,
+      "loss": 0.0109,
       "step": 956
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4461336901429106e-05,
-      "loss": 0.0195,
+      "loss": 0.0136,
       "step": 957
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.444926739980913e-05,
-      "loss": 0.0142,
+      "loss": 0.0115,
       "step": 958
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4437187746788395e-05,
-      "loss": 0.0067,
+      "loss": 0.0083,
       "step": 959
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4425097955344158e-05,
-      "loss": 0.0093,
+      "loss": 0.0111,
       "step": 960
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.441299803846454e-05,
-      "loss": 0.0095,
+      "loss": 0.0088,
       "step": 961
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.440088800914855e-05,
-      "loss": 0.0138,
+      "loss": 0.0131,
       "step": 962
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4388767880406068e-05,
-      "loss": 0.0103,
+      "loss": 0.0098,
       "step": 963
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4376637665257804e-05,
-      "loss": 0.0099,
+      "loss": 0.0068,
       "step": 964
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4364497376735324e-05,
-      "loss": 0.0066,
+      "loss": 0.0085,
       "step": 965
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4352347027881003e-05,
-      "loss": 0.0099,
+      "loss": 0.0069,
       "step": 966
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4340186631748034e-05,
-      "loss": 0.0088,
+      "loss": 0.0099,
       "step": 967
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4328016201400395e-05,
-      "loss": 0.008,
+      "loss": 0.0056,
       "step": 968
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.431583574991284e-05,
-      "loss": 0.0095,
+      "loss": 0.008,
       "step": 969
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4303645290370913e-05,
-      "loss": 0.0062,
+      "loss": 0.0067,
       "step": 970
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4291444835870882e-05,
-      "loss": 0.0118,
+      "loss": 0.0156,
       "step": 971
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4279234399519765e-05,
-      "loss": 0.0144,
+      "loss": 0.0145,
       "step": 972
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.42670139944353e-05,
-      "loss": 0.006,
+      "loss": 0.0063,
       "step": 973
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4254783633745942e-05,
-      "loss": 0.0071,
+      "loss": 0.0085,
       "step": 974
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4242543330590828e-05,
-      "loss": 0.0103,
+      "loss": 0.0084,
       "step": 975
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4230293098119787e-05,
-      "loss": 0.0132,
+      "loss": 0.0094,
       "step": 976
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4218032949493317e-05,
-      "loss": 0.0079,
+      "loss": 0.0065,
       "step": 977
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.420576289788256e-05,
-      "loss": 0.0101,
+      "loss": 0.0114,
       "step": 978
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4193482956469305e-05,
-      "loss": 0.0101,
+      "loss": 0.0076,
       "step": 979
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4181193138445952e-05,
-      "loss": 0.0075,
+      "loss": 0.0058,
       "step": 980
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4168893457015528e-05,
-      "loss": 0.0084,
+      "loss": 0.0076,
       "step": 981
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4156583925391652e-05,
-      "loss": 0.0103,
+      "loss": 0.0075,
       "step": 982
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.414426455679852e-05,
-      "loss": 0.0051,
+      "loss": 0.0059,
       "step": 983
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.4131935364470895e-05,
-      "loss": 0.004,
+      "loss": 0.0131,
       "step": 984
     },
     {
       "epoch": 0.31,
       "learning_rate": 2.41195963616541e-05,
-      "loss": 0.0059,
+      "loss": 0.0065,
       "step": 985
     },
     {
@@ -5943,229 +5943,229 @@
     {
       "epoch": 0.32,
       "learning_rate": 2.4094888977586968e-05,
-      "loss": 0.0137,
+      "loss": 0.0156,
       "step": 987
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.408252062287991e-05,
-      "loss": 0.0129,
+      "loss": 0.0132,
       "step": 988
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.4070142510770212e-05,
-      "loss": 0.0113,
+      "loss": 0.0118,
       "step": 989
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.4057754654555755e-05,
-      "loss": 0.0086,
+      "loss": 0.0083,
       "step": 990
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.4045357067544875e-05,
-      "loss": 0.0207,
+      "loss": 0.0143,
       "step": 991
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.403294976305638e-05,
-      "loss": 0.0097,
+      "loss": 0.0119,
       "step": 992
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.4020532754419492e-05,
-      "loss": 0.0145,
+      "loss": 0.0116,
       "step": 993
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.4008106054973883e-05,
-      "loss": 0.0158,
+      "loss": 0.0104,
       "step": 994
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3995669678069627e-05,
-      "loss": 0.0132,
+      "loss": 0.0107,
       "step": 995
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.398322363706719e-05,
-      "loss": 0.0125,
+      "loss": 0.0098,
       "step": 996
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3970767945337417e-05,
-      "loss": 0.0088,
+      "loss": 0.0084,
       "step": 997
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.395830261626153e-05,
-      "loss": 0.0062,
+      "loss": 0.0068,
       "step": 998
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.394582766323111e-05,
-      "loss": 0.0109,
+      "loss": 0.0141,
       "step": 999
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.393334309964806e-05,
-      "loss": 0.0107,
+      "loss": 0.0106,
       "step": 1000
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3920848938924625e-05,
-      "loss": 0.0155,
+      "loss": 0.0091,
       "step": 1001
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3908345194483345e-05,
-      "loss": 0.0096,
+      "loss": 0.0126,
       "step": 1002
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.389583187975706e-05,
-      "loss": 0.013,
+      "loss": 0.0089,
       "step": 1003
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3883309008188898e-05,
-      "loss": 0.0086,
+      "loss": 0.0094,
       "step": 1004
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.387077659323225e-05,
-      "loss": 0.006,
+      "loss": 0.0077,
       "step": 1005
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.385823464835076e-05,
-      "loss": 0.0093,
+      "loss": 0.0054,
       "step": 1006
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3845683187018303e-05,
-      "loss": 0.0181,
+      "loss": 0.0102,
       "step": 1007
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3833122222718993e-05,
-      "loss": 0.0124,
+      "loss": 0.0128,
       "step": 1008
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.382055176894714e-05,
-      "loss": 0.0068,
+      "loss": 0.0063,
       "step": 1009
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3807971839207258e-05,
-      "loss": 0.0294,
+      "loss": 0.0367,
       "step": 1010
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3795382447014025e-05,
-      "loss": 0.0168,
+      "loss": 0.0173,
       "step": 1011
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.37827836058923e-05,
-      "loss": 0.0084,
+      "loss": 0.0142,
       "step": 1012
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.377017532937709e-05,
-      "loss": 0.0126,
+      "loss": 0.0095,
       "step": 1013
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.375755763101354e-05,
-      "loss": 0.0111,
+      "loss": 0.0167,
       "step": 1014
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.374493052435691e-05,
-      "loss": 0.0065,
+      "loss": 0.0072,
       "step": 1015
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.373229402297257e-05,
-      "loss": 0.0071,
+      "loss": 0.0048,
       "step": 1016
     },
     {
       "epoch": 0.32,
       "learning_rate": 2.3719648140435986e-05,
-      "loss": 0.0101,
+      "loss": 0.0074,
       "step": 1017
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.37069928903327e-05,
-      "loss": 0.0132,
+      "loss": 0.012,
       "step": 1018
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.369432828625832e-05,
-      "loss": 0.0087,
+      "loss": 0.0079,
       "step": 1019
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3681654341818494e-05,
-      "loss": 0.0291,
+      "loss": 0.0124,
       "step": 1020
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.366897107062892e-05,
-      "loss": 0.0081,
+      "loss": 0.0115,
       "step": 1021
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3656278486315302e-05,
-      "loss": 0.0084,
+      "loss": 0.0111,
       "step": 1022
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.364357660251335e-05,
-      "loss": 0.0096,
+      "loss": 0.0073,
       "step": 1023
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3630865432868776e-05,
-      "loss": 0.0312,
+      "loss": 0.025,
       "step": 1024
     },
     {
@@ -6177,19 +6177,19 @@
     {
       "epoch": 0.33,
       "learning_rate": 2.3605415290684442e-05,
-      "loss": 0.0098,
+      "loss": 0.0109,
       "step": 1026
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.359267634548591e-05,
-      "loss": 0.0066,
+      "loss": 0.0059,
       "step": 1027
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3579928169127188e-05,
-      "loss": 0.0106,
+      "loss": 0.0079,
       "step": 1028
     },
     {
@@ -6201,151 +6201,151 @@
     {
       "epoch": 0.33,
       "learning_rate": 2.3554404177720808e-05,
-      "loss": 0.006,
+      "loss": 0.0068,
       "step": 1030
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.354162839009372e-05,
-      "loss": 0.0097,
+      "loss": 0.0064,
       "step": 1031
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.352884342614754e-05,
-      "loss": 0.0116,
+      "loss": 0.0115,
       "step": 1032
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.351604929961723e-05,
-      "loss": 0.0111,
+      "loss": 0.0105,
       "step": 1033
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3503246024247583e-05,
-      "loss": 0.0087,
+      "loss": 0.0088,
       "step": 1034
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3490433613793245e-05,
-      "loss": 0.0116,
+      "loss": 0.0108,
       "step": 1035
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.347761208201865e-05,
-      "loss": 0.0088,
+      "loss": 0.0127,
       "step": 1036
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3464781442698037e-05,
-      "loss": 0.008,
+      "loss": 0.0067,
       "step": 1037
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3451941709615442e-05,
-      "loss": 0.0088,
+      "loss": 0.0092,
       "step": 1038
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3439092896564658e-05,
-      "loss": 0.0114,
+      "loss": 0.01,
       "step": 1039
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.342623501734924e-05,
-      "loss": 0.0124,
+      "loss": 0.0125,
       "step": 1040
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3413368085782473e-05,
-      "loss": 0.0079,
+      "loss": 0.0053,
       "step": 1041
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.340049211568738e-05,
-      "loss": 0.0094,
+      "loss": 0.0087,
       "step": 1042
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3387607120896683e-05,
-      "loss": 0.0094,
+      "loss": 0.0101,
       "step": 1043
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3374713115252806e-05,
-      "loss": 0.0192,
+      "loss": 0.0203,
       "step": 1044
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.336181011260785e-05,
-      "loss": 0.0127,
+      "loss": 0.0126,
       "step": 1045
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3348898126823584e-05,
-      "loss": 0.0092,
+      "loss": 0.0054,
       "step": 1046
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3335977171771426e-05,
-      "loss": 0.0055,
+      "loss": 0.0076,
       "step": 1047
     },
     {
       "epoch": 0.33,
       "learning_rate": 2.3323047261332438e-05,
-      "loss": 0.0201,
+      "loss": 0.0249,
       "step": 1048
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3310108409397276e-05,
-      "loss": 0.0138,
+      "loss": 0.0134,
       "step": 1049
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.329716062986624e-05,
-      "loss": 0.0151,
+      "loss": 0.0127,
       "step": 1050
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3284203936649194e-05,
-      "loss": 0.0086,
+      "loss": 0.0071,
       "step": 1051
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3271238343665582e-05,
-      "loss": 0.0111,
+      "loss": 0.0084,
       "step": 1052
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.325826386484442e-05,
-      "loss": 0.0082,
+      "loss": 0.0067,
       "step": 1053
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.324528051412426e-05,
-      "loss": 0.0051,
+      "loss": 0.0055,
       "step": 1054
     },
     {
@@ -6357,109 +6357,109 @@
     {
       "epoch": 0.34,
       "learning_rate": 2.3219287252788805e-05,
-      "loss": 0.0101,
+      "loss": 0.0072,
       "step": 1056
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3206277370098218e-05,
-      "loss": 0.0087,
+      "loss": 0.0112,
       "step": 1057
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3193258671358018e-05,
-      "loss": 0.0046,
+      "loss": 0.0053,
       "step": 1058
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.318023117055426e-05,
-      "loss": 0.0069,
+      "loss": 0.0072,
       "step": 1059
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3167194881682457e-05,
-      "loss": 0.0077,
+      "loss": 0.005,
       "step": 1060
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3154149818747583e-05,
-      "loss": 0.0231,
+      "loss": 0.0151,
       "step": 1061
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.314109599576401e-05,
-      "loss": 0.0128,
+      "loss": 0.0111,
       "step": 1062
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3128033426755534e-05,
-      "loss": 0.0014,
+      "loss": 0.0023,
       "step": 1063
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3114962125755347e-05,
-      "loss": 0.0106,
+      "loss": 0.0115,
       "step": 1064
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.310188210680602e-05,
-      "loss": 0.0105,
+      "loss": 0.0106,
       "step": 1065
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3088793383959492e-05,
-      "loss": 0.0071,
+      "loss": 0.0067,
       "step": 1066
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3075695971277054e-05,
-      "loss": 0.0091,
+      "loss": 0.0122,
       "step": 1067
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.306258988282932e-05,
-      "loss": 0.0088,
+      "loss": 0.009,
       "step": 1068
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3049475132696244e-05,
-      "loss": 0.0126,
+      "loss": 0.0111,
       "step": 1069
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.303635173496707e-05,
-      "loss": 0.0046,
+      "loss": 0.0053,
       "step": 1070
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.302321970374034e-05,
-      "loss": 0.0098,
+      "loss": 0.0087,
       "step": 1071
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.3010079053123864e-05,
-      "loss": 0.0123,
+      "loss": 0.0079,
       "step": 1072
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.2996929797234724e-05,
-      "loss": 0.0042,
+      "loss": 0.0057,
       "step": 1073
     },
     {
@@ -6471,7 +6471,7 @@
     {
       "epoch": 0.34,
       "learning_rate": 2.297060552615294e-05,
-      "loss": 0.012,
+      "loss": 0.0127,
       "step": 1075
     },
     {
@@ -6483,79 +6483,79 @@
     {
       "epoch": 0.34,
       "learning_rate": 2.2944247003616205e-05,
-      "loss": 0.0133,
+      "loss": 0.0123,
       "step": 1077
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.2931054933442867e-05,
-      "loss": 0.0077,
+      "loss": 0.0081,
       "step": 1078
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.291785434289292e-05,
-      "loss": 0.0139,
+      "loss": 0.0117,
       "step": 1079
     },
     {
       "epoch": 0.34,
       "learning_rate": 2.2904645246147823e-05,
-      "loss": 0.0146,
+      "loss": 0.0228,
       "step": 1080
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2891427657398184e-05,
-      "loss": 0.0148,
+      "loss": 0.0156,
       "step": 1081
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2878201590843736e-05,
-      "loss": 0.0102,
+      "loss": 0.0095,
       "step": 1082
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.286496706069331e-05,
-      "loss": 0.0078,
+      "loss": 0.0066,
       "step": 1083
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2851724081164844e-05,
-      "loss": 0.0068,
+      "loss": 0.0074,
       "step": 1084
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2838472666485342e-05,
-      "loss": 0.0079,
+      "loss": 0.0064,
       "step": 1085
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.282521283089087e-05,
-      "loss": 0.0045,
+      "loss": 0.0039,
       "step": 1086
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2811944588626537e-05,
-      "loss": 0.0055,
+      "loss": 0.0058,
       "step": 1087
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.279866795394651e-05,
-      "loss": 0.0069,
+      "loss": 0.0088,
       "step": 1088
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2785382941113928e-05,
-      "loss": 0.0101,
+      "loss": 0.0086,
       "step": 1089
     },
     {
@@ -6567,397 +6567,397 @@
     {
       "epoch": 0.35,
       "learning_rate": 2.2758787838088785e-05,
-      "loss": 0.0045,
+      "loss": 0.0042,
       "step": 1091
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2745477776467484e-05,
-      "loss": 0.0162,
+      "loss": 0.016,
       "step": 1092
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2732159393836145e-05,
-      "loss": 0.0151,
+      "loss": 0.0112,
       "step": 1093
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2718832704502782e-05,
-      "loss": 0.0063,
+      "loss": 0.0119,
       "step": 1094
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2705497722784327e-05,
-      "loss": 0.0084,
+      "loss": 0.0064,
       "step": 1095
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2692154463006627e-05,
-      "loss": 0.0056,
+      "loss": 0.004,
       "step": 1096
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2678802939504424e-05,
-      "loss": 0.008,
+      "loss": 0.0083,
       "step": 1097
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.266544316662133e-05,
-      "loss": 0.0138,
+      "loss": 0.0099,
       "step": 1098
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2652075158709828e-05,
-      "loss": 0.0073,
+      "loss": 0.0153,
       "step": 1099
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.263869893013124e-05,
-      "loss": 0.0057,
+      "loss": 0.0056,
       "step": 1100
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2625314495255726e-05,
-      "loss": 0.0214,
+      "loss": 0.0143,
       "step": 1101
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2611921868462254e-05,
-      "loss": 0.0088,
+      "loss": 0.0126,
       "step": 1102
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2598521064138603e-05,
-      "loss": 0.0069,
+      "loss": 0.0098,
       "step": 1103
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2585112096681328e-05,
-      "loss": 0.0042,
+      "loss": 0.0067,
       "step": 1104
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.257169498049576e-05,
-      "loss": 0.0087,
+      "loss": 0.0083,
       "step": 1105
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2558269729995985e-05,
-      "loss": 0.0062,
+      "loss": 0.0085,
       "step": 1106
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2544836359604822e-05,
-      "loss": 0.0034,
+      "loss": 0.0088,
       "step": 1107
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2531394883753814e-05,
-      "loss": 0.01,
+      "loss": 0.0124,
       "step": 1108
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.251794531688321e-05,
-      "loss": 0.005,
+      "loss": 0.0091,
       "step": 1109
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.2504487673441963e-05,
-      "loss": 0.0104,
+      "loss": 0.0085,
       "step": 1110
     },
     {
       "epoch": 0.35,
       "learning_rate": 2.249102196788769e-05,
-      "loss": 0.0078,
+      "loss": 0.0056,
       "step": 1111
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2477548214686678e-05,
-      "loss": 0.0076,
+      "loss": 0.0068,
       "step": 1112
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2464066428313846e-05,
-      "loss": 0.0069,
+      "loss": 0.0047,
       "step": 1113
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2450576623252757e-05,
-      "loss": 0.0138,
+      "loss": 0.0091,
       "step": 1114
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.243707881399559e-05,
-      "loss": 0.0048,
+      "loss": 0.0051,
       "step": 1115
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2423573015043115e-05,
-      "loss": 0.0057,
+      "loss": 0.006,
       "step": 1116
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.241005924090468e-05,
-      "loss": 0.0108,
+      "loss": 0.0098,
       "step": 1117
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.239653750609821e-05,
-      "loss": 0.0112,
+      "loss": 0.0101,
       "step": 1118
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2383007825150188e-05,
-      "loss": 0.011,
+      "loss": 0.0073,
       "step": 1119
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2369470212595618e-05,
-      "loss": 0.0075,
+      "loss": 0.0079,
       "step": 1120
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2355924682978037e-05,
-      "loss": 0.0128,
+      "loss": 0.0119,
       "step": 1121
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.234237125084948e-05,
-      "loss": 0.0084,
+      "loss": 0.0075,
       "step": 1122
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.232880993077048e-05,
-      "loss": 0.013,
+      "loss": 0.0119,
       "step": 1123
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.231524073731003e-05,
-      "loss": 0.0066,
+      "loss": 0.01,
       "step": 1124
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2301663685045596e-05,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 1125
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.228807878856308e-05,
-      "loss": 0.0066,
+      "loss": 0.0072,
       "step": 1126
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2274486062456815e-05,
-      "loss": 0.0083,
+      "loss": 0.0046,
       "step": 1127
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2260885521329542e-05,
-      "loss": 0.011,
+      "loss": 0.0151,
       "step": 1128
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2247277179792392e-05,
-      "loss": 0.0159,
+      "loss": 0.0204,
       "step": 1129
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2233661052464892e-05,
-      "loss": 0.012,
+      "loss": 0.0156,
       "step": 1130
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2220037153974915e-05,
-      "loss": 0.0018,
+      "loss": 0.0017,
       "step": 1131
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2206405498958698e-05,
-      "loss": 0.0172,
+      "loss": 0.0152,
       "step": 1132
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2192766102060798e-05,
-      "loss": 0.0062,
+      "loss": 0.0061,
       "step": 1133
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2179118977934098e-05,
-      "loss": 0.0087,
+      "loss": 0.0088,
       "step": 1134
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.216546414123978e-05,
-      "loss": 0.0113,
+      "loss": 0.0128,
       "step": 1135
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2151801606647314e-05,
-      "loss": 0.006,
+      "loss": 0.0043,
       "step": 1136
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2138131388834434e-05,
-      "loss": 0.0135,
+      "loss": 0.0116,
       "step": 1137
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2124453502487134e-05,
-      "loss": 0.0124,
+      "loss": 0.0097,
       "step": 1138
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2110767962299635e-05,
-      "loss": 0.0126,
+      "loss": 0.0122,
       "step": 1139
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2097074782974394e-05,
-      "loss": 0.0099,
+      "loss": 0.0085,
       "step": 1140
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2083373979222077e-05,
-      "loss": 0.0057,
+      "loss": 0.0071,
       "step": 1141
     },
     {
       "epoch": 0.36,
       "learning_rate": 2.2069665565761533e-05,
-      "loss": 0.0109,
+      "loss": 0.0186,
       "step": 1142
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.2055949557319775e-05,
-      "loss": 0.0118,
+      "loss": 0.0122,
       "step": 1143
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.2042225968631997e-05,
-      "loss": 0.0124,
+      "loss": 0.0081,
       "step": 1144
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.202849481444153e-05,
-      "loss": 0.0075,
+      "loss": 0.007,
       "step": 1145
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.2014756109499823e-05,
-      "loss": 0.0079,
+      "loss": 0.0062,
       "step": 1146
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.2001009868566442e-05,
-      "loss": 0.0051,
+      "loss": 0.0065,
       "step": 1147
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1987256106409058e-05,
-      "loss": 0.006,
+      "loss": 0.0055,
       "step": 1148
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1973494837803408e-05,
-      "loss": 0.0119,
+      "loss": 0.0161,
       "step": 1149
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1959726077533307e-05,
-      "loss": 0.0053,
+      "loss": 0.0101,
       "step": 1150
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1945949840390603e-05,
-      "loss": 0.0064,
+      "loss": 0.008,
       "step": 1151
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1932166141175188e-05,
-      "loss": 0.0115,
+      "loss": 0.0105,
       "step": 1152
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1918374994694963e-05,
-      "loss": 0.0092,
+      "loss": 0.0077,
       "step": 1153
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1904576415765846e-05,
-      "loss": 0.0093,
+      "loss": 0.0098,
       "step": 1154
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1890770419211712e-05,
-      "loss": 0.0112,
+      "loss": 0.0096,
       "step": 1155
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.187695701986443e-05,
-      "loss": 0.0081,
+      "loss": 0.012,
       "step": 1156
     },
     {
@@ -6969,121 +6969,121 @@
     {
       "epoch": 0.37,
       "learning_rate": 2.1849308072157595e-05,
-      "loss": 0.0122,
+      "loss": 0.0142,
       "step": 1158
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1835472553501463e-05,
-      "loss": 0.01,
+      "loss": 0.0101,
       "step": 1159
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1821629691458986e-05,
-      "loss": 0.0079,
+      "loss": 0.0077,
       "step": 1160
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.180777950090163e-05,
-      "loss": 0.0113,
+      "loss": 0.0109,
       "step": 1161
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.179392199670873e-05,
-      "loss": 0.007,
+      "loss": 0.0078,
       "step": 1162
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1780057193767478e-05,
-      "loss": 0.0093,
+      "loss": 0.0076,
       "step": 1163
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1766185106972913e-05,
-      "loss": 0.0156,
+      "loss": 0.0215,
       "step": 1164
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.17523057512279e-05,
-      "loss": 0.007,
+      "loss": 0.0076,
       "step": 1165
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1738419141443095e-05,
-      "loss": 0.0114,
+      "loss": 0.0115,
       "step": 1166
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1724525292536967e-05,
-      "loss": 0.0078,
+      "loss": 0.0088,
       "step": 1167
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.171062421943576e-05,
-      "loss": 0.0042,
+      "loss": 0.0041,
       "step": 1168
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.169671593707347e-05,
-      "loss": 0.0069,
+      "loss": 0.0068,
       "step": 1169
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.168280046039184e-05,
-      "loss": 0.0047,
+      "loss": 0.0068,
       "step": 1170
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1668877804340354e-05,
-      "loss": 0.0046,
+      "loss": 0.0053,
       "step": 1171
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.165494798387619e-05,
-      "loss": 0.0063,
+      "loss": 0.0076,
       "step": 1172
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1641011013964234e-05,
-      "loss": 0.0086,
+      "loss": 0.0093,
       "step": 1173
     },
     {
       "epoch": 0.37,
       "learning_rate": 2.1627066909577052e-05,
-      "loss": 0.0083,
+      "loss": 0.009,
       "step": 1174
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1613115685694868e-05,
-      "loss": 0.0129,
+      "loss": 0.0086,
       "step": 1175
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1599157357305572e-05,
-      "loss": 0.0104,
+      "loss": 0.0101,
       "step": 1176
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.158519193940466e-05,
-      "loss": 0.0085,
+      "loss": 0.0067,
       "step": 1177
     },
     {
@@ -7095,13 +7095,13 @@
     {
       "epoch": 0.38,
       "learning_rate": 2.155723989508812e-05,
-      "loss": 0.0107,
+      "loss": 0.0071,
       "step": 1179
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1543253298701526e-05,
-      "loss": 0.0068,
+      "loss": 0.0077,
       "step": 1180
     },
     {
@@ -7113,43 +7113,43 @@
     {
       "epoch": 0.38,
       "learning_rate": 2.1515259032601067e-05,
-      "loss": 0.0089,
+      "loss": 0.0118,
       "step": 1182
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.15012513929616e-05,
-      "loss": 0.0083,
+      "loss": 0.0079,
       "step": 1183
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1487236768991443e-05,
-      "loss": 0.017,
+      "loss": 0.0144,
       "step": 1184
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.147321517574659e-05,
-      "loss": 0.0075,
+      "loss": 0.0107,
       "step": 1185
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1459186628290522e-05,
-      "loss": 0.0062,
+      "loss": 0.0067,
       "step": 1186
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1445151141694174e-05,
-      "loss": 0.0073,
+      "loss": 0.0119,
       "step": 1187
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1431108731035953e-05,
-      "loss": 0.0094,
+      "loss": 0.0093,
       "step": 1188
     },
     {
@@ -7161,31 +7161,31 @@
     {
       "epoch": 0.38,
       "learning_rate": 2.1403003197884687e-05,
-      "loss": 0.0074,
+      "loss": 0.0096,
       "step": 1190
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.138894010558557e-05,
-      "loss": 0.0077,
+      "loss": 0.0097,
       "step": 1191
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.137487014961241e-05,
-      "loss": 0.0191,
+      "loss": 0.0184,
       "step": 1192
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.136079334508064e-05,
-      "loss": 0.0072,
+      "loss": 0.007,
       "step": 1193
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1346709707113053e-05,
-      "loss": 0.0079,
+      "loss": 0.0074,
       "step": 1194
     },
     {
@@ -7197,337 +7197,337 @@
     {
       "epoch": 0.38,
       "learning_rate": 2.1318521991398265e-05,
-      "loss": 0.0109,
+      "loss": 0.0088,
       "step": 1196
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.130441794393328e-05,
-      "loss": 0.0236,
+      "loss": 0.0094,
       "step": 1197
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.129030712359689e-05,
-      "loss": 0.0038,
+      "loss": 0.0033,
       "step": 1198
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1276189545548424e-05,
-      "loss": 0.0068,
+      "loss": 0.007,
       "step": 1199
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1262065224954467e-05,
-      "loss": 0.0114,
+      "loss": 0.0088,
       "step": 1200
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.124793417698886e-05,
-      "loss": 0.0071,
+      "loss": 0.0081,
       "step": 1201
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.123379641683267e-05,
-      "loss": 0.0124,
+      "loss": 0.0114,
       "step": 1202
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1219651959674172e-05,
-      "loss": 0.0083,
+      "loss": 0.01,
       "step": 1203
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1205500820708832e-05,
-      "loss": 0.0038,
+      "loss": 0.0036,
       "step": 1204
     },
     {
       "epoch": 0.38,
       "learning_rate": 2.1191343015139297e-05,
-      "loss": 0.0186,
+      "loss": 0.0184,
       "step": 1205
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.117717855817538e-05,
-      "loss": 0.0153,
+      "loss": 0.0071,
       "step": 1206
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1163007465034037e-05,
-      "loss": 0.0112,
+      "loss": 0.0107,
       "step": 1207
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1148829750939342e-05,
-      "loss": 0.0073,
+      "loss": 0.006,
       "step": 1208
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1134645431122507e-05,
-      "loss": 0.0161,
+      "loss": 0.0114,
       "step": 1209
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1120454520821812e-05,
-      "loss": 0.0081,
+      "loss": 0.0141,
       "step": 1210
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1106257035282646e-05,
-      "loss": 0.0057,
+      "loss": 0.006,
       "step": 1211
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1092052989757436e-05,
-      "loss": 0.0072,
+      "loss": 0.0067,
       "step": 1212
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1077842399505673e-05,
-      "loss": 0.0028,
+      "loss": 0.0025,
       "step": 1213
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1063625279793872e-05,
-      "loss": 0.0118,
+      "loss": 0.0079,
       "step": 1214
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1049401645895556e-05,
-      "loss": 0.0084,
+      "loss": 0.007,
       "step": 1215
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.103517151309127e-05,
-      "loss": 0.0053,
+      "loss": 0.0069,
       "step": 1216
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.1020934896668516e-05,
-      "loss": 0.0092,
+      "loss": 0.0138,
       "step": 1217
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.100669181192177e-05,
-      "loss": 0.0086,
+      "loss": 0.0085,
       "step": 1218
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.099244227415246e-05,
-      "loss": 0.0169,
+      "loss": 0.0171,
       "step": 1219
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0978186298668944e-05,
-      "loss": 0.0072,
+      "loss": 0.0051,
       "step": 1220
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0963923900786492e-05,
-      "loss": 0.0154,
+      "loss": 0.0133,
       "step": 1221
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0949655095827288e-05,
-      "loss": 0.0106,
+      "loss": 0.0112,
       "step": 1222
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0935379899120373e-05,
-      "loss": 0.0097,
+      "loss": 0.0103,
       "step": 1223
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0921098326001683e-05,
-      "loss": 0.0107,
+      "loss": 0.0127,
       "step": 1224
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0906810391813987e-05,
-      "loss": 0.0054,
+      "loss": 0.0094,
       "step": 1225
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0892516111906897e-05,
-      "loss": 0.0077,
+      "loss": 0.0076,
       "step": 1226
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0878215501636832e-05,
-      "loss": 0.0053,
+      "loss": 0.0052,
       "step": 1227
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0863908576367015e-05,
-      "loss": 0.0109,
+      "loss": 0.01,
       "step": 1228
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.084959535146747e-05,
-      "loss": 0.0089,
+      "loss": 0.0088,
       "step": 1229
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.083527584231496e-05,
-      "loss": 0.009,
+      "loss": 0.0097,
       "step": 1230
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0820950064293022e-05,
-      "loss": 0.0097,
+      "loss": 0.0113,
       "step": 1231
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0806618032791916e-05,
-      "loss": 0.0093,
+      "loss": 0.0102,
       "step": 1232
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.079227976320862e-05,
-      "loss": 0.0269,
+      "loss": 0.0416,
       "step": 1233
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0777935270946824e-05,
-      "loss": 0.0082,
+      "loss": 0.0067,
       "step": 1234
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.0763584571416888e-05,
-      "loss": 0.0186,
+      "loss": 0.0135,
       "step": 1235
     },
     {
       "epoch": 0.39,
       "learning_rate": 2.074922768003586e-05,
-      "loss": 0.0108,
+      "loss": 0.0074,
       "step": 1236
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0734864612227417e-05,
-      "loss": 0.0146,
+      "loss": 0.0148,
       "step": 1237
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0720495383421887e-05,
-      "loss": 0.0118,
+      "loss": 0.0078,
       "step": 1238
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.070612000905622e-05,
-      "loss": 0.0128,
+      "loss": 0.0216,
       "step": 1239
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.069173850457395e-05,
-      "loss": 0.0075,
+      "loss": 0.0067,
       "step": 1240
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.067735088542521e-05,
-      "loss": 0.0084,
+      "loss": 0.0069,
       "step": 1241
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0662957167066706e-05,
-      "loss": 0.0058,
+      "loss": 0.0059,
       "step": 1242
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.064855736496169e-05,
-      "loss": 0.0071,
+      "loss": 0.0095,
       "step": 1243
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.063415149457994e-05,
-      "loss": 0.0152,
+      "loss": 0.0101,
       "step": 1244
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.061973957139777e-05,
-      "loss": 0.0083,
+      "loss": 0.0091,
       "step": 1245
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0605321610897988e-05,
-      "loss": 0.0149,
+      "loss": 0.012,
       "step": 1246
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0590897628569894e-05,
-      "loss": 0.007,
+      "loss": 0.0077,
       "step": 1247
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.057646763990925e-05,
-      "loss": 0.0084,
+      "loss": 0.0092,
       "step": 1248
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.056203166041827e-05,
-      "loss": 0.0062,
+      "loss": 0.0066,
       "step": 1249
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0547589705605604e-05,
-      "loss": 0.0099,
+      "loss": 0.0148,
       "step": 1250
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0533141790986333e-05,
-      "loss": 0.0113,
+      "loss": 0.0083,
       "step": 1251
     },
     {
@@ -7539,157 +7539,157 @@
     {
       "epoch": 0.4,
       "learning_rate": 2.050422814442026e-05,
-      "loss": 0.0025,
+      "loss": 0.0051,
       "step": 1253
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0489762443535543e-05,
-      "loss": 0.0146,
+      "loss": 0.0119,
       "step": 1254
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0475290844968372e-05,
-      "loss": 0.0105,
+      "loss": 0.0149,
       "step": 1255
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0460813364265666e-05,
-      "loss": 0.0172,
+      "loss": 0.0174,
       "step": 1256
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.044633001698066e-05,
-      "loss": 0.0121,
+      "loss": 0.0075,
       "step": 1257
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0431840818672894e-05,
-      "loss": 0.012,
+      "loss": 0.009,
       "step": 1258
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0417345784908195e-05,
-      "loss": 0.0065,
+      "loss": 0.0069,
       "step": 1259
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0402844931258662e-05,
-      "loss": 0.0091,
+      "loss": 0.0102,
       "step": 1260
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0388338273302638e-05,
-      "loss": 0.0049,
+      "loss": 0.005,
       "step": 1261
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.037382582662471e-05,
-      "loss": 0.0116,
+      "loss": 0.0124,
       "step": 1262
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0359307606815683e-05,
-      "loss": 0.0118,
+      "loss": 0.0078,
       "step": 1263
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.034478362947256e-05,
-      "loss": 0.0122,
+      "loss": 0.0117,
       "step": 1264
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0330253910198527e-05,
-      "loss": 0.0097,
+      "loss": 0.0089,
       "step": 1265
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0315718464602947e-05,
-      "loss": 0.0116,
+      "loss": 0.0102,
       "step": 1266
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.0301177308301323e-05,
-      "loss": 0.0097,
+      "loss": 0.0124,
       "step": 1267
     },
     {
       "epoch": 0.4,
       "learning_rate": 2.028663045691531e-05,
-      "loss": 0.0072,
+      "loss": 0.0086,
       "step": 1268
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0272077926072667e-05,
-      "loss": 0.0108,
+      "loss": 0.0096,
       "step": 1269
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0257519731407264e-05,
-      "loss": 0.0109,
+      "loss": 0.0119,
       "step": 1270
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0242955888559042e-05,
-      "loss": 0.0112,
+      "loss": 0.0147,
       "step": 1271
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0228386413174025e-05,
-      "loss": 0.0054,
+      "loss": 0.0103,
       "step": 1272
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0213811320904283e-05,
-      "loss": 0.0081,
+      "loss": 0.0076,
       "step": 1273
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0199230627407912e-05,
-      "loss": 0.0069,
+      "loss": 0.0078,
       "step": 1274
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.018464434834904e-05,
-      "loss": 0.0062,
+      "loss": 0.0059,
       "step": 1275
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0170052499397785e-05,
-      "loss": 0.0117,
+      "loss": 0.0135,
       "step": 1276
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0155455096230248e-05,
-      "loss": 0.0084,
+      "loss": 0.0117,
       "step": 1277
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.014085215452851e-05,
-      "loss": 0.0078,
+      "loss": 0.0076,
       "step": 1278
     },
     {
@@ -7701,205 +7701,205 @@
     {
       "epoch": 0.41,
       "learning_rate": 2.011162971828043e-05,
-      "loss": 0.0039,
+      "loss": 0.0036,
       "step": 1280
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0097010255127914e-05,
-      "loss": 0.0031,
+      "loss": 0.0045,
       "step": 1281
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0082385316228816e-05,
-      "loss": 0.013,
+      "loss": 0.0111,
       "step": 1282
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0067754917294783e-05,
-      "loss": 0.0048,
+      "loss": 0.004,
       "step": 1283
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0053119074043336e-05,
-      "loss": 0.0073,
+      "loss": 0.0055,
       "step": 1284
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.003847780219784e-05,
-      "loss": 0.0114,
+      "loss": 0.0101,
       "step": 1285
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0023831117487496e-05,
-      "loss": 0.0067,
+      "loss": 0.0072,
       "step": 1286
     },
     {
       "epoch": 0.41,
       "learning_rate": 2.0009179035647325e-05,
-      "loss": 0.0111,
+      "loss": 0.0112,
       "step": 1287
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9994521572418137e-05,
-      "loss": 0.0066,
+      "loss": 0.0075,
       "step": 1288
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9979858743546524e-05,
-      "loss": 0.0082,
+      "loss": 0.0092,
       "step": 1289
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9965190564784848e-05,
-      "loss": 0.0125,
+      "loss": 0.0136,
       "step": 1290
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.995051705189122e-05,
-      "loss": 0.0102,
+      "loss": 0.0136,
       "step": 1291
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9935838220629464e-05,
-      "loss": 0.0102,
+      "loss": 0.0146,
       "step": 1292
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.992115408676914e-05,
-      "loss": 0.0143,
+      "loss": 0.0152,
       "step": 1293
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9906464666085488e-05,
-      "loss": 0.0123,
+      "loss": 0.0133,
       "step": 1294
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.989176997435944e-05,
-      "loss": 0.0077,
+      "loss": 0.0079,
       "step": 1295
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9877070027377593e-05,
-      "loss": 0.0128,
+      "loss": 0.0119,
       "step": 1296
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.986236484093217e-05,
-      "loss": 0.008,
+      "loss": 0.0065,
       "step": 1297
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9847654430821036e-05,
-      "loss": 0.0118,
+      "loss": 0.0115,
       "step": 1298
     },
     {
       "epoch": 0.41,
       "learning_rate": 1.9832938812847668e-05,
-      "loss": 0.0098,
+      "loss": 0.0085,
       "step": 1299
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9818218002821144e-05,
-      "loss": 0.0056,
+      "loss": 0.0059,
       "step": 1300
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9803492016556105e-05,
-      "loss": 0.0085,
+      "loss": 0.0068,
       "step": 1301
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.978876086987276e-05,
-      "loss": 0.009,
+      "loss": 0.0092,
       "step": 1302
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9774024578596866e-05,
-      "loss": 0.011,
+      "loss": 0.0085,
       "step": 1303
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.97592831585597e-05,
-      "loss": 0.0157,
+      "loss": 0.0114,
       "step": 1304
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9744536625598056e-05,
-      "loss": 0.0055,
+      "loss": 0.0047,
       "step": 1305
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9729784995554213e-05,
-      "loss": 0.0122,
+      "loss": 0.0132,
       "step": 1306
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9715028284275926e-05,
-      "loss": 0.0061,
+      "loss": 0.0046,
       "step": 1307
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.970026650761642e-05,
-      "loss": 0.0081,
+      "loss": 0.008,
       "step": 1308
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.968549968143435e-05,
-      "loss": 0.0054,
+      "loss": 0.0038,
       "step": 1309
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9670727821593802e-05,
-      "loss": 0.0123,
+      "loss": 0.013,
       "step": 1310
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9655950943964267e-05,
-      "loss": 0.0181,
+      "loss": 0.0161,
       "step": 1311
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.964116906442062e-05,
-      "loss": 0.0056,
+      "loss": 0.0062,
       "step": 1312
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.962638219884313e-05,
-      "loss": 0.011,
+      "loss": 0.0084,
       "step": 1313
     },
     {
@@ -7911,205 +7911,205 @@
     {
       "epoch": 0.42,
       "learning_rate": 1.9596793573134384e-05,
-      "loss": 0.0074,
+      "loss": 0.0077,
       "step": 1315
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.958199184479036e-05,
-      "loss": 0.0131,
+      "loss": 0.0152,
       "step": 1316
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9567185193986896e-05,
-      "loss": 0.0114,
+      "loss": 0.0099,
       "step": 1317
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9552373636630873e-05,
-      "loss": 0.0092,
+      "loss": 0.0116,
       "step": 1318
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9537557188634424e-05,
-      "loss": 0.0078,
+      "loss": 0.008,
       "step": 1319
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.952273586591494e-05,
-      "loss": 0.0117,
+      "loss": 0.0119,
       "step": 1320
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9507909684395054e-05,
-      "loss": 0.0084,
+      "loss": 0.008,
       "step": 1321
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.949307866000261e-05,
-      "loss": 0.0081,
+      "loss": 0.0089,
       "step": 1322
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.947824280867067e-05,
-      "loss": 0.0093,
+      "loss": 0.0137,
       "step": 1323
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9463402146337467e-05,
-      "loss": 0.0049,
+      "loss": 0.0052,
       "step": 1324
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9448556688946403e-05,
-      "loss": 0.0109,
+      "loss": 0.0085,
       "step": 1325
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.943370645244604e-05,
-      "loss": 0.0059,
+      "loss": 0.0062,
       "step": 1326
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9418851452790075e-05,
-      "loss": 0.0102,
+      "loss": 0.0071,
       "step": 1327
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.940399170593731e-05,
-      "loss": 0.0069,
+      "loss": 0.008,
       "step": 1328
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.9389127227851654e-05,
-      "loss": 0.0173,
+      "loss": 0.0195,
       "step": 1329
     },
     {
       "epoch": 0.42,
       "learning_rate": 1.937425803450211e-05,
-      "loss": 0.0107,
+      "loss": 0.0108,
       "step": 1330
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9359384141862717e-05,
-      "loss": 0.007,
+      "loss": 0.0103,
       "step": 1331
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9344505565912593e-05,
-      "loss": 0.0128,
+      "loss": 0.0099,
       "step": 1332
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9329622322635872e-05,
-      "loss": 0.0063,
+      "loss": 0.0066,
       "step": 1333
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9314734428021703e-05,
-      "loss": 0.0109,
+      "loss": 0.0086,
       "step": 1334
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.929984189806424e-05,
-      "loss": 0.0054,
+      "loss": 0.0052,
       "step": 1335
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.928494474876261e-05,
-      "loss": 0.0125,
+      "loss": 0.0137,
       "step": 1336
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.92700429961209e-05,
-      "loss": 0.0068,
+      "loss": 0.0083,
       "step": 1337
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.925513665614815e-05,
-      "loss": 0.0105,
+      "loss": 0.0082,
       "step": 1338
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9240225744858318e-05,
-      "loss": 0.0101,
+      "loss": 0.0105,
       "step": 1339
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.922531027827029e-05,
-      "loss": 0.0094,
+      "loss": 0.0153,
       "step": 1340
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9210390272407826e-05,
-      "loss": 0.0097,
+      "loss": 0.0079,
       "step": 1341
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9195465743299574e-05,
-      "loss": 0.046,
+      "loss": 0.0398,
       "step": 1342
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9180536706979043e-05,
-      "loss": 0.009,
+      "loss": 0.0124,
       "step": 1343
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9165603179484574e-05,
-      "loss": 0.0089,
+      "loss": 0.0145,
       "step": 1344
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9150665176859345e-05,
-      "loss": 0.0119,
+      "loss": 0.0102,
       "step": 1345
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.913572271515133e-05,
-      "loss": 0.0059,
+      "loss": 0.0037,
       "step": 1346
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9120775810413297e-05,
-      "loss": 0.0124,
+      "loss": 0.0135,
       "step": 1347
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.91058244787028e-05,
-      "loss": 0.0042,
+      "loss": 0.0044,
       "step": 1348
     },
     {
@@ -8121,13 +8121,13 @@
     {
       "epoch": 0.43,
       "learning_rate": 1.9075908598618326e-05,
-      "loss": 0.0071,
+      "loss": 0.0096,
       "step": 1350
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.9060944082383142e-05,
-      "loss": 0.0123,
+      "loss": 0.0095,
       "step": 1351
     },
     {
@@ -8139,241 +8139,241 @@
     {
       "epoch": 0.43,
       "learning_rate": 1.9031001977909205e-05,
-      "loss": 0.0079,
+      "loss": 0.0201,
       "step": 1353
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.901602442183742e-05,
-      "loss": 0.0113,
+      "loss": 0.0121,
       "step": 1354
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.900104255132816e-05,
-      "loss": 0.0037,
+      "loss": 0.0032,
       "step": 1355
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.8986056382476535e-05,
-      "loss": 0.0101,
+      "loss": 0.0182,
       "step": 1356
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.897106593138228e-05,
-      "loss": 0.0059,
+      "loss": 0.0057,
       "step": 1357
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.8956071214149716e-05,
-      "loss": 0.0117,
+      "loss": 0.0135,
       "step": 1358
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.8941072246887747e-05,
-      "loss": 0.0082,
+      "loss": 0.0088,
       "step": 1359
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.8926069045709845e-05,
-      "loss": 0.0116,
+      "loss": 0.0102,
       "step": 1360
     },
     {
       "epoch": 0.43,
       "learning_rate": 1.8911061626734043e-05,
-      "loss": 0.0065,
+      "loss": 0.0064,
       "step": 1361
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8896050006082886e-05,
-      "loss": 0.0079,
+      "loss": 0.0076,
       "step": 1362
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.888103419988345e-05,
-      "loss": 0.0105,
+      "loss": 0.0091,
       "step": 1363
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.886601422426729e-05,
-      "loss": 0.0047,
+      "loss": 0.0066,
       "step": 1364
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8850990095370457e-05,
-      "loss": 0.0083,
+      "loss": 0.0076,
       "step": 1365
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8835961829333452e-05,
-      "loss": 0.0083,
+      "loss": 0.009,
       "step": 1366
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.882092944230123e-05,
-      "loss": 0.0054,
+      "loss": 0.005,
       "step": 1367
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8805892950423165e-05,
-      "loss": 0.008,
+      "loss": 0.0079,
       "step": 1368
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8790852369853042e-05,
-      "loss": 0.0112,
+      "loss": 0.0101,
       "step": 1369
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.877580771674905e-05,
-      "loss": 0.0062,
+      "loss": 0.0059,
       "step": 1370
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8760759007273732e-05,
-      "loss": 0.0086,
+      "loss": 0.0115,
       "step": 1371
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8745706257594007e-05,
-      "loss": 0.0058,
+      "loss": 0.0064,
       "step": 1372
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.873064948388113e-05,
-      "loss": 0.0101,
+      "loss": 0.0114,
       "step": 1373
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8715588702310675e-05,
-      "loss": 0.0055,
+      "loss": 0.0056,
       "step": 1374
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.870052392906252e-05,
-      "loss": 0.0088,
+      "loss": 0.009,
       "step": 1375
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8685455180320842e-05,
-      "loss": 0.0041,
+      "loss": 0.0044,
       "step": 1376
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8670382472274076e-05,
-      "loss": 0.0064,
+      "loss": 0.0068,
       "step": 1377
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8655305821114916e-05,
-      "loss": 0.0125,
+      "loss": 0.0058,
       "step": 1378
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8640225243040295e-05,
-      "loss": 0.0056,
+      "loss": 0.0078,
       "step": 1379
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8625140754251366e-05,
-      "loss": 0.0036,
+      "loss": 0.0045,
       "step": 1380
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8610052370953473e-05,
-      "loss": 0.0022,
+      "loss": 0.0018,
       "step": 1381
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8594960109356153e-05,
-      "loss": 0.0109,
+      "loss": 0.0082,
       "step": 1382
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.857986398567311e-05,
-      "loss": 0.0121,
+      "loss": 0.0108,
       "step": 1383
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.856476401612219e-05,
-      "loss": 0.0064,
+      "loss": 0.0053,
       "step": 1384
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8549660216925375e-05,
-      "loss": 0.0055,
+      "loss": 0.0056,
       "step": 1385
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8534552604308767e-05,
-      "loss": 0.0117,
+      "loss": 0.0102,
       "step": 1386
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.851944119450255e-05,
-      "loss": 0.0052,
+      "loss": 0.0033,
       "step": 1387
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8504326003740995e-05,
-      "loss": 0.0104,
+      "loss": 0.0121,
       "step": 1388
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8489207048262448e-05,
-      "loss": 0.0146,
+      "loss": 0.012,
       "step": 1389
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8474084344309274e-05,
-      "loss": 0.0091,
+      "loss": 0.0078,
       "step": 1390
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8458957908127885e-05,
-      "loss": 0.0096,
+      "loss": 0.0111,
       "step": 1391
     },
     {
       "epoch": 0.44,
       "learning_rate": 1.8443827755968694e-05,
-      "loss": 0.0124,
+      "loss": 0.0058,
       "step": 1392
     },
     {
@@ -8385,595 +8385,595 @@
     {
       "epoch": 0.45,
       "learning_rate": 1.8413556368738513e-05,
-      "loss": 0.0044,
+      "loss": 0.0057,
       "step": 1394
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.839841516618824e-05,
-      "loss": 0.0063,
+      "loss": 0.0089,
       "step": 1395
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8383270312701574e-05,
-      "loss": 0.0065,
+      "loss": 0.0064,
       "step": 1396
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8368121824548714e-05,
-      "loss": 0.0045,
+      "loss": 0.0044,
       "step": 1397
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8352969718003765e-05,
-      "loss": 0.013,
+      "loss": 0.015,
       "step": 1398
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8337814009344716e-05,
-      "loss": 0.0085,
+      "loss": 0.0095,
       "step": 1399
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8322654714853436e-05,
-      "loss": 0.0076,
+      "loss": 0.0057,
       "step": 1400
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.830749185081564e-05,
-      "loss": 0.0062,
+      "loss": 0.0055,
       "step": 1401
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.829232543352087e-05,
-      "loss": 0.0057,
+      "loss": 0.0059,
       "step": 1402
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8277155479262496e-05,
-      "loss": 0.0074,
+      "loss": 0.0085,
       "step": 1403
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.826198200433768e-05,
-      "loss": 0.0104,
+      "loss": 0.0098,
       "step": 1404
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8246805025047388e-05,
-      "loss": 0.0081,
+      "loss": 0.0099,
       "step": 1405
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8231624557696314e-05,
-      "loss": 0.0077,
+      "loss": 0.007,
       "step": 1406
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8216440618592932e-05,
-      "loss": 0.0146,
+      "loss": 0.0147,
       "step": 1407
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8201253224049423e-05,
-      "loss": 0.0097,
+      "loss": 0.0129,
       "step": 1408
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8186062390381698e-05,
-      "loss": 0.0065,
+      "loss": 0.0077,
       "step": 1409
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8170868133909348e-05,
-      "loss": 0.0046,
+      "loss": 0.0084,
       "step": 1410
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8155670470955653e-05,
-      "loss": 0.0069,
+      "loss": 0.0088,
       "step": 1411
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8140469417847545e-05,
-      "loss": 0.0063,
+      "loss": 0.0055,
       "step": 1412
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8125264990915597e-05,
-      "loss": 0.0086,
+      "loss": 0.0106,
       "step": 1413
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8110057206494012e-05,
-      "loss": 0.0025,
+      "loss": 0.0021,
       "step": 1414
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8094846080920602e-05,
-      "loss": 0.0059,
+      "loss": 0.007,
       "step": 1415
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8079631630536757e-05,
-      "loss": 0.0033,
+      "loss": 0.0027,
       "step": 1416
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8064413871687452e-05,
-      "loss": 0.0058,
+      "loss": 0.0062,
       "step": 1417
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8049192820721205e-05,
-      "loss": 0.0127,
+      "loss": 0.0125,
       "step": 1418
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8033968493990082e-05,
-      "loss": 0.0117,
+      "loss": 0.0141,
       "step": 1419
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.801874090784966e-05,
-      "loss": 0.0104,
+      "loss": 0.0075,
       "step": 1420
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.8003510078659015e-05,
-      "loss": 0.0087,
+      "loss": 0.0083,
       "step": 1421
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.7988276022780717e-05,
-      "loss": 0.0061,
+      "loss": 0.0066,
       "step": 1422
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.79730387565808e-05,
-      "loss": 0.0088,
+      "loss": 0.0083,
       "step": 1423
     },
     {
       "epoch": 0.45,
       "learning_rate": 1.7957798296428735e-05,
-      "loss": 0.0046,
+      "loss": 0.0039,
       "step": 1424
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7942554658697446e-05,
-      "loss": 0.0062,
+      "loss": 0.0061,
       "step": 1425
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.792730785976324e-05,
-      "loss": 0.0081,
+      "loss": 0.0103,
       "step": 1426
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7912057916005857e-05,
-      "loss": 0.0069,
+      "loss": 0.0081,
       "step": 1427
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7896804843808393e-05,
-      "loss": 0.0066,
+      "loss": 0.0076,
       "step": 1428
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.78815486595573e-05,
-      "loss": 0.0051,
+      "loss": 0.0067,
       "step": 1429
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7866289379642388e-05,
-      "loss": 0.006,
+      "loss": 0.0068,
       "step": 1430
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.785102702045678e-05,
-      "loss": 0.0093,
+      "loss": 0.0066,
       "step": 1431
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.783576159839692e-05,
-      "loss": 0.0129,
+      "loss": 0.0124,
       "step": 1432
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7820493129862542e-05,
-      "loss": 0.0128,
+      "loss": 0.0113,
       "step": 1433
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7805221631256638e-05,
-      "loss": 0.0072,
+      "loss": 0.0094,
       "step": 1434
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.778994711898546e-05,
-      "loss": 0.0176,
+      "loss": 0.0212,
       "step": 1435
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7774669609458515e-05,
-      "loss": 0.0055,
+      "loss": 0.0073,
       "step": 1436
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.775938911908851e-05,
-      "loss": 0.005,
+      "loss": 0.0054,
       "step": 1437
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.774410566429136e-05,
-      "loss": 0.0036,
+      "loss": 0.0031,
       "step": 1438
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7728819261486166e-05,
-      "loss": 0.0061,
+      "loss": 0.0062,
       "step": 1439
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7713529927095198e-05,
-      "loss": 0.0031,
+      "loss": 0.0036,
       "step": 1440
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.769823767754387e-05,
-      "loss": 0.0107,
+      "loss": 0.0078,
       "step": 1441
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.768294252926074e-05,
-      "loss": 0.0083,
+      "loss": 0.0079,
       "step": 1442
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7667644498677465e-05,
-      "loss": 0.0147,
+      "loss": 0.0154,
       "step": 1443
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.76523436022288e-05,
-      "loss": 0.0095,
+      "loss": 0.0081,
       "step": 1444
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7637039856352587e-05,
-      "loss": 0.0054,
+      "loss": 0.0063,
       "step": 1445
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.762173327748973e-05,
-      "loss": 0.0052,
+      "loss": 0.0057,
       "step": 1446
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7606423882084168e-05,
-      "loss": 0.0076,
+      "loss": 0.0078,
       "step": 1447
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7591111686582866e-05,
-      "loss": 0.0055,
+      "loss": 0.0065,
       "step": 1448
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.757579670743581e-05,
-      "loss": 0.013,
+      "loss": 0.0105,
       "step": 1449
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7560478961095965e-05,
-      "loss": 0.0088,
+      "loss": 0.0089,
       "step": 1450
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7545158464019273e-05,
-      "loss": 0.0059,
+      "loss": 0.0068,
       "step": 1451
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.752983523266462e-05,
-      "loss": 0.0113,
+      "loss": 0.0105,
       "step": 1452
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7514509283493848e-05,
-      "loss": 0.0046,
+      "loss": 0.0055,
       "step": 1453
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7499180632971708e-05,
-      "loss": 0.0075,
+      "loss": 0.006,
       "step": 1454
     },
     {
       "epoch": 0.46,
       "learning_rate": 1.7483849297565853e-05,
-      "loss": 0.0104,
+      "loss": 0.0093,
       "step": 1455
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7468515293746826e-05,
-      "loss": 0.0112,
+      "loss": 0.0078,
       "step": 1456
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7453178637988032e-05,
-      "loss": 0.0077,
+      "loss": 0.0074,
       "step": 1457
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.743783934676572e-05,
-      "loss": 0.022,
+      "loss": 0.0232,
       "step": 1458
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.742249743655899e-05,
-      "loss": 0.0073,
+      "loss": 0.0066,
       "step": 1459
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7407152923849725e-05,
-      "loss": 0.01,
+      "loss": 0.0087,
       "step": 1460
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.739180582512263e-05,
-      "loss": 0.0058,
+      "loss": 0.0054,
       "step": 1461
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7376456156865174e-05,
-      "loss": 0.0091,
+      "loss": 0.0088,
       "step": 1462
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7361103935567596e-05,
-      "loss": 0.0101,
+      "loss": 0.0113,
       "step": 1463
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7345749177722874e-05,
-      "loss": 0.0114,
+      "loss": 0.0118,
       "step": 1464
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7330391899826706e-05,
-      "loss": 0.0066,
+      "loss": 0.0077,
       "step": 1465
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7315032118377503e-05,
-      "loss": 0.0053,
+      "loss": 0.005,
       "step": 1466
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7299669849876367e-05,
-      "loss": 0.0124,
+      "loss": 0.0057,
       "step": 1467
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7284305110827065e-05,
-      "loss": 0.0051,
+      "loss": 0.0055,
       "step": 1468
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7268937917736025e-05,
-      "loss": 0.006,
+      "loss": 0.0057,
       "step": 1469
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7253568287112303e-05,
-      "loss": 0.0084,
+      "loss": 0.0108,
       "step": 1470
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7238196235467583e-05,
-      "loss": 0.0065,
+      "loss": 0.0064,
       "step": 1471
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7222821779316147e-05,
-      "loss": 0.0049,
+      "loss": 0.0041,
       "step": 1472
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7207444935174858e-05,
-      "loss": 0.0027,
+      "loss": 0.0083,
       "step": 1473
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7192065719563144e-05,
-      "loss": 0.0067,
+      "loss": 0.006,
       "step": 1474
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.717668414900298e-05,
-      "loss": 0.0094,
+      "loss": 0.0087,
       "step": 1475
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.716130024001888e-05,
-      "loss": 0.0099,
+      "loss": 0.0098,
       "step": 1476
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.714591400913786e-05,
-      "loss": 0.0066,
+      "loss": 0.0059,
       "step": 1477
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7130525472889434e-05,
-      "loss": 0.0074,
+      "loss": 0.0045,
       "step": 1478
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7115134647805585e-05,
-      "loss": 0.0149,
+      "loss": 0.01,
       "step": 1479
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7099741550420776e-05,
-      "loss": 0.0131,
+      "loss": 0.0111,
       "step": 1480
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7084346197271894e-05,
-      "loss": 0.0078,
+      "loss": 0.0079,
       "step": 1481
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.706894860489825e-05,
-      "loss": 0.0064,
+      "loss": 0.0071,
       "step": 1482
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7053548789841557e-05,
-      "loss": 0.0062,
+      "loss": 0.0052,
       "step": 1483
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.703814676864593e-05,
-      "loss": 0.0101,
+      "loss": 0.0058,
       "step": 1484
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.702274255785785e-05,
-      "loss": 0.0072,
+      "loss": 0.0102,
       "step": 1485
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.7007336174026135e-05,
-      "loss": 0.0068,
+      "loss": 0.0128,
       "step": 1486
     },
     {
       "epoch": 0.47,
       "learning_rate": 1.699192763370196e-05,
-      "loss": 0.0052,
+      "loss": 0.0053,
       "step": 1487
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6976516953438798e-05,
-      "loss": 0.011,
+      "loss": 0.0119,
       "step": 1488
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6961104149792437e-05,
-      "loss": 0.0081,
+      "loss": 0.0079,
       "step": 1489
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6945689239320936e-05,
-      "loss": 0.0093,
+      "loss": 0.0133,
       "step": 1490
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6930272238584613e-05,
-      "loss": 0.0093,
+      "loss": 0.0112,
       "step": 1491
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.691485316414604e-05,
-      "loss": 0.0069,
+      "loss": 0.0076,
       "step": 1492
     },
     {
@@ -8985,115 +8985,115 @@
     {
       "epoch": 0.48,
       "learning_rate": 1.688400886042356e-05,
-      "loss": 0.014,
+      "loss": 0.0154,
       "step": 1494
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6868583664275855e-05,
-      "loss": 0.0089,
+      "loss": 0.0092,
       "step": 1495
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6853156460698282e-05,
-      "loss": 0.0115,
+      "loss": 0.0095,
       "step": 1496
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6837727266264377e-05,
-      "loss": 0.0039,
+      "loss": 0.0048,
       "step": 1497
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6822296097549802e-05,
-      "loss": 0.011,
+      "loss": 0.0094,
       "step": 1498
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.680686297113235e-05,
-      "loss": 0.0052,
+      "loss": 0.0058,
       "step": 1499
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6791427903591916e-05,
-      "loss": 0.0084,
+      "loss": 0.0069,
       "step": 1500
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6775990911510473e-05,
-      "loss": 0.01,
+      "loss": 0.011,
       "step": 1501
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6760552011472077e-05,
-      "loss": 0.0091,
+      "loss": 0.0097,
       "step": 1502
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.674511122006283e-05,
-      "loss": 0.0062,
+      "loss": 0.0069,
       "step": 1503
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6729668553870845e-05,
-      "loss": 0.0093,
+      "loss": 0.0083,
       "step": 1504
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.671422402948627e-05,
-      "loss": 0.0046,
+      "loss": 0.0062,
       "step": 1505
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.669877766350124e-05,
-      "loss": 0.0178,
+      "loss": 0.0151,
       "step": 1506
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6683329472509885e-05,
-      "loss": 0.0069,
+      "loss": 0.0066,
       "step": 1507
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6667879473108277e-05,
-      "loss": 0.007,
+      "loss": 0.0068,
       "step": 1508
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6652427681894432e-05,
-      "loss": 0.0086,
+      "loss": 0.0132,
       "step": 1509
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.66369741154683e-05,
-      "loss": 0.0062,
+      "loss": 0.0073,
       "step": 1510
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6621518790431728e-05,
-      "loss": 0.0029,
+      "loss": 0.003,
       "step": 1511
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6606061723388465e-05,
-      "loss": 0.007,
+      "loss": 0.009,
       "step": 1512
     },
     {
@@ -9105,25 +9105,25 @@
     {
       "epoch": 0.48,
       "learning_rate": 1.6575142429706165e-05,
-      "loss": 0.0062,
+      "loss": 0.01,
       "step": 1514
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.655968023628389e-05,
-      "loss": 0.0033,
+      "loss": 0.0064,
       "step": 1515
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.6544216367288435e-05,
-      "loss": 0.0121,
+      "loss": 0.0116,
       "step": 1516
     },
     {
       "epoch": 0.48,
       "learning_rate": 1.652875083933271e-05,
-      "loss": 0.0038,
+      "loss": 0.0044,
       "step": 1517
     },
     {
@@ -9135,393 +9135,393 @@
     {
       "epoch": 0.49,
       "learning_rate": 1.6497814873001027e-05,
-      "loss": 0.0043,
+      "loss": 0.0047,
       "step": 1519
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6482344467859753e-05,
-      "loss": 0.0117,
+      "loss": 0.011,
       "step": 1520
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6466872470227535e-05,
-      "loss": 0.0057,
+      "loss": 0.0085,
       "step": 1521
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6451398896726033e-05,
-      "loss": 0.0058,
+      "loss": 0.0108,
       "step": 1522
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6435923763978584e-05,
-      "loss": 0.0053,
+      "loss": 0.0095,
       "step": 1523
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.642044708861022e-05,
-      "loss": 0.011,
+      "loss": 0.0106,
       "step": 1524
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.640496888724761e-05,
-      "loss": 0.0109,
+      "loss": 0.0102,
       "step": 1525
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6389489176519084e-05,
-      "loss": 0.0065,
+      "loss": 0.0074,
       "step": 1526
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.637400797305457e-05,
-      "loss": 0.0068,
+      "loss": 0.0061,
       "step": 1527
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.635852529348561e-05,
-      "loss": 0.0056,
+      "loss": 0.0057,
       "step": 1528
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6343041154445343e-05,
-      "loss": 0.0091,
+      "loss": 0.0106,
       "step": 1529
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6327555572568465e-05,
-      "loss": 0.0044,
+      "loss": 0.0051,
       "step": 1530
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.631206856449122e-05,
-      "loss": 0.0101,
+      "loss": 0.0098,
       "step": 1531
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6296580146851384e-05,
-      "loss": 0.0104,
+      "loss": 0.0086,
       "step": 1532
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6281090336288257e-05,
-      "loss": 0.0053,
+      "loss": 0.0063,
       "step": 1533
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.626559914944262e-05,
-      "loss": 0.0077,
+      "loss": 0.0049,
       "step": 1534
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6250106602956753e-05,
-      "loss": 0.0082,
+      "loss": 0.008,
       "step": 1535
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6234612713474383e-05,
-      "loss": 0.0077,
+      "loss": 0.0058,
       "step": 1536
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.621911749764067e-05,
-      "loss": 0.0094,
+      "loss": 0.0108,
       "step": 1537
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6203620972102224e-05,
-      "loss": 0.0074,
+      "loss": 0.0071,
       "step": 1538
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6188123153507044e-05,
-      "loss": 0.0093,
+      "loss": 0.0099,
       "step": 1539
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6172624058504526e-05,
-      "loss": 0.0119,
+      "loss": 0.0105,
       "step": 1540
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6157123703745433e-05,
-      "loss": 0.0064,
+      "loss": 0.0058,
       "step": 1541
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.614162210588188e-05,
-      "loss": 0.01,
+      "loss": 0.0082,
       "step": 1542
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6126119281567324e-05,
-      "loss": 0.0065,
+      "loss": 0.0056,
       "step": 1543
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.611061524745653e-05,
-      "loss": 0.0083,
+      "loss": 0.0087,
       "step": 1544
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.609511002020557e-05,
-      "loss": 0.0071,
+      "loss": 0.0054,
       "step": 1545
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.60796036164718e-05,
-      "loss": 0.0026,
+      "loss": 0.0027,
       "step": 1546
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.606409605291384e-05,
-      "loss": 0.0057,
+      "loss": 0.0052,
       "step": 1547
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6048587346191543e-05,
-      "loss": 0.0075,
+      "loss": 0.0084,
       "step": 1548
     },
     {
       "epoch": 0.49,
       "learning_rate": 1.6033077512966003e-05,
-      "loss": 0.0055,
+      "loss": 0.0069,
       "step": 1549
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.601756656989952e-05,
-      "loss": 0.0058,
+      "loss": 0.0055,
       "step": 1550
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.6002054533655583e-05,
-      "loss": 0.0114,
+      "loss": 0.0142,
       "step": 1551
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5986541420898873e-05,
-      "loss": 0.0048,
+      "loss": 0.0047,
       "step": 1552
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5971027248295197e-05,
-      "loss": 0.0071,
+      "loss": 0.0087,
       "step": 1553
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5955512032511526e-05,
-      "loss": 0.0084,
+      "loss": 0.0061,
       "step": 1554
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5939995790215942e-05,
-      "loss": 0.0086,
+      "loss": 0.0068,
       "step": 1555
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5924478538077632e-05,
-      "loss": 0.0067,
+      "loss": 0.0062,
       "step": 1556
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5908960292766862e-05,
-      "loss": 0.0066,
+      "loss": 0.0092,
       "step": 1557
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5893441070954976e-05,
-      "loss": 0.0089,
+      "loss": 0.0075,
       "step": 1558
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.587792088931435e-05,
-      "loss": 0.0057,
+      "loss": 0.004,
       "step": 1559
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.586239976451841e-05,
-      "loss": 0.0099,
+      "loss": 0.0074,
       "step": 1560
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5846877713241592e-05,
-      "loss": 0.0149,
+      "loss": 0.0137,
       "step": 1561
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.583135475215931e-05,
-      "loss": 0.0087,
+      "loss": 0.0084,
       "step": 1562
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5815830897947976e-05,
-      "loss": 0.011,
+      "loss": 0.0111,
       "step": 1563
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5800306167284953e-05,
-      "loss": 0.0086,
+      "loss": 0.008,
       "step": 1564
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5784780576848543e-05,
-      "loss": 0.0098,
+      "loss": 0.0093,
       "step": 1565
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5769254143317977e-05,
-      "loss": 0.0121,
+      "loss": 0.013,
       "step": 1566
     },
     {
       "epoch": 0.5,
-      "eval_loss": 0.01018073782324791,
-      "eval_runtime": 28.108,
-      "eval_samples_per_second": 187.598,
-      "eval_steps_per_second": 5.87,
+      "eval_loss": 0.009752409532666206,
+      "eval_runtime": 28.4298,
+      "eval_samples_per_second": 185.474,
+      "eval_steps_per_second": 11.608,
       "step": 1566
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5753726883373387e-05,
-      "loss": 0.0048,
+      "loss": 0.0059,
       "step": 1567
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.57381988136958e-05,
-      "loss": 0.0064,
+      "loss": 0.0066,
       "step": 1568
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.57226699509671e-05,
-      "loss": 0.0116,
+      "loss": 0.0118,
       "step": 1569
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5707140311870037e-05,
-      "loss": 0.0047,
+      "loss": 0.0034,
       "step": 1570
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5691609913088187e-05,
-      "loss": 0.01,
+      "loss": 0.0113,
       "step": 1571
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.567607877130594e-05,
-      "loss": 0.0117,
+      "loss": 0.0088,
       "step": 1572
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5660546903208494e-05,
-      "loss": 0.0128,
+      "loss": 0.0146,
       "step": 1573
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5645014325481824e-05,
-      "loss": 0.0054,
+      "loss": 0.0072,
       "step": 1574
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.562948105481266e-05,
-      "loss": 0.0076,
+      "loss": 0.005,
       "step": 1575
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5613947107888475e-05,
-      "loss": 0.0173,
+      "loss": 0.022,
       "step": 1576
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.559841250139749e-05,
-      "loss": 0.0077,
+      "loss": 0.0071,
       "step": 1577
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5582877252028616e-05,
-      "loss": 0.009,
+      "loss": 0.0093,
       "step": 1578
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.556734137647145e-05,
-      "loss": 0.0071,
+      "loss": 0.0063,
       "step": 1579
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5551804891416274e-05,
-      "loss": 0.0044,
+      "loss": 0.0041,
       "step": 1580
     },
     {
       "epoch": 0.5,
       "learning_rate": 1.5536267813554023e-05,
-      "loss": 0.0085,
+      "loss": 0.0097,
       "step": 1581
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5520730159576265e-05,
-      "loss": 0.0065,
+      "loss": 0.0075,
       "step": 1582
     },
     {
@@ -9533,307 +9533,307 @@
     {
       "epoch": 0.51,
       "learning_rate": 1.5489653190043578e-05,
-      "loss": 0.0046,
+      "loss": 0.0045,
       "step": 1584
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5474113907874806e-05,
-      "loss": 0.0079,
+      "loss": 0.0067,
       "step": 1585
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5458574116362825e-05,
-      "loss": 0.0157,
+      "loss": 0.0179,
       "step": 1586
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5443033832202106e-05,
-      "loss": 0.0098,
+      "loss": 0.0075,
       "step": 1587
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5427493072087662e-05,
-      "loss": 0.0051,
+      "loss": 0.004,
       "step": 1588
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.541195185271503e-05,
-      "loss": 0.0073,
+      "loss": 0.0075,
       "step": 1589
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5396410190780217e-05,
-      "loss": 0.0037,
+      "loss": 0.0036,
       "step": 1590
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.538086810297972e-05,
-      "loss": 0.0096,
+      "loss": 0.0052,
       "step": 1591
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5365325606010496e-05,
-      "loss": 0.0121,
+      "loss": 0.0117,
       "step": 1592
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5349782716569935e-05,
-      "loss": 0.0078,
+      "loss": 0.0079,
       "step": 1593
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5334239451355844e-05,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 1594
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5318695827066445e-05,
-      "loss": 0.0038,
+      "loss": 0.0036,
       "step": 1595
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.530315186040034e-05,
-      "loss": 0.009,
+      "loss": 0.0098,
       "step": 1596
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5287607568056498e-05,
-      "loss": 0.0073,
+      "loss": 0.0076,
       "step": 1597
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5272062966734232e-05,
-      "loss": 0.0044,
+      "loss": 0.0047,
       "step": 1598
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5256518073133209e-05,
-      "loss": 0.0134,
+      "loss": 0.0115,
       "step": 1599
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5240972903953384e-05,
-      "loss": 0.0087,
+      "loss": 0.0076,
       "step": 1600
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5225427475895025e-05,
-      "loss": 0.0065,
+      "loss": 0.0062,
       "step": 1601
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.520988180565867e-05,
-      "loss": 0.0142,
+      "loss": 0.0122,
       "step": 1602
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5194335909945112e-05,
-      "loss": 0.0066,
+      "loss": 0.0092,
       "step": 1603
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5178789805455408e-05,
-      "loss": 0.0038,
+      "loss": 0.0025,
       "step": 1604
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5163243508890813e-05,
-      "loss": 0.0053,
+      "loss": 0.0051,
       "step": 1605
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5147697036952808e-05,
-      "loss": 0.0115,
+      "loss": 0.0109,
       "step": 1606
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5132150406343045e-05,
-      "loss": 0.0091,
+      "loss": 0.0107,
       "step": 1607
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5116603633763372e-05,
-      "loss": 0.0031,
+      "loss": 0.0035,
       "step": 1608
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5101056735915758e-05,
-      "loss": 0.0093,
+      "loss": 0.0075,
       "step": 1609
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5085509729502328e-05,
-      "loss": 0.0059,
+      "loss": 0.0073,
       "step": 1610
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5069962631225321e-05,
-      "loss": 0.0038,
+      "loss": 0.0035,
       "step": 1611
     },
     {
       "epoch": 0.51,
       "learning_rate": 1.5054415457787065e-05,
-      "loss": 0.0109,
+      "loss": 0.0089,
       "step": 1612
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.5038868225889979e-05,
-      "loss": 0.0111,
+      "loss": 0.0108,
       "step": 1613
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.5023320952236542e-05,
-      "loss": 0.0102,
+      "loss": 0.0097,
       "step": 1614
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.5007773653529277e-05,
-      "loss": 0.0094,
+      "loss": 0.01,
       "step": 1615
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4992226346470727e-05,
-      "loss": 0.004,
+      "loss": 0.0049,
       "step": 1616
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4976679047763464e-05,
-      "loss": 0.0113,
+      "loss": 0.0122,
       "step": 1617
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4961131774110024e-05,
-      "loss": 0.0103,
+      "loss": 0.0123,
       "step": 1618
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4945584542212936e-05,
-      "loss": 0.0078,
+      "loss": 0.0075,
       "step": 1619
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4930037368774686e-05,
-      "loss": 0.0082,
+      "loss": 0.0083,
       "step": 1620
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4914490270497673e-05,
-      "loss": 0.0048,
+      "loss": 0.0043,
       "step": 1621
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4898943264084248e-05,
-      "loss": 0.0115,
+      "loss": 0.0161,
       "step": 1622
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4883396366236632e-05,
-      "loss": 0.0108,
+      "loss": 0.0103,
       "step": 1623
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4867849593656954e-05,
-      "loss": 0.0068,
+      "loss": 0.0089,
       "step": 1624
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.48523029630472e-05,
-      "loss": 0.0041,
+      "loss": 0.0038,
       "step": 1625
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4836756491109189e-05,
-      "loss": 0.0095,
+      "loss": 0.0064,
       "step": 1626
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4821210194544593e-05,
-      "loss": 0.0059,
+      "loss": 0.0062,
       "step": 1627
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.480566409005489e-05,
-      "loss": 0.0092,
+      "loss": 0.009,
       "step": 1628
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4790118194341334e-05,
-      "loss": 0.0054,
+      "loss": 0.0061,
       "step": 1629
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.477457252410498e-05,
-      "loss": 0.0021,
+      "loss": 0.0027,
       "step": 1630
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4759027096046617e-05,
-      "loss": 0.0061,
+      "loss": 0.0066,
       "step": 1631
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.474348192686679e-05,
-      "loss": 0.011,
+      "loss": 0.0083,
       "step": 1632
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4727937033265768e-05,
-      "loss": 0.0074,
+      "loss": 0.0078,
       "step": 1633
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4712392431943506e-05,
-      "loss": 0.0039,
+      "loss": 0.0033,
       "step": 1634
     },
     {
@@ -9845,139 +9845,139 @@
     {
       "epoch": 0.52,
       "learning_rate": 1.4681304172933557e-05,
-      "loss": 0.003,
+      "loss": 0.0033,
       "step": 1636
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4665760548644158e-05,
-      "loss": 0.0156,
+      "loss": 0.0143,
       "step": 1637
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.465021728343007e-05,
-      "loss": 0.0055,
+      "loss": 0.006,
       "step": 1638
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4634674393989504e-05,
-      "loss": 0.0067,
+      "loss": 0.0066,
       "step": 1639
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4619131897020278e-05,
-      "loss": 0.0082,
+      "loss": 0.0063,
       "step": 1640
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4603589809219789e-05,
-      "loss": 0.0039,
+      "loss": 0.0036,
       "step": 1641
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4588048147284973e-05,
-      "loss": 0.0098,
+      "loss": 0.0102,
       "step": 1642
     },
     {
       "epoch": 0.52,
       "learning_rate": 1.4572506927912342e-05,
-      "loss": 0.012,
+      "loss": 0.0109,
       "step": 1643
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4556966167797898e-05,
-      "loss": 0.0044,
+      "loss": 0.0045,
       "step": 1644
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4541425883637176e-05,
-      "loss": 0.0053,
+      "loss": 0.0064,
       "step": 1645
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4525886092125195e-05,
-      "loss": 0.0097,
+      "loss": 0.0091,
       "step": 1646
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4510346809956426e-05,
-      "loss": 0.0094,
+      "loss": 0.0048,
       "step": 1647
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4494808053824821e-05,
-      "loss": 0.0036,
+      "loss": 0.003,
       "step": 1648
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.447926984042374e-05,
-      "loss": 0.0082,
+      "loss": 0.0086,
       "step": 1649
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.446373218644598e-05,
-      "loss": 0.009,
+      "loss": 0.006,
       "step": 1650
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4448195108583728e-05,
-      "loss": 0.0099,
+      "loss": 0.0167,
       "step": 1651
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4432658623528554e-05,
-      "loss": 0.0122,
+      "loss": 0.0111,
       "step": 1652
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4417122747971388e-05,
-      "loss": 0.0102,
+      "loss": 0.0116,
       "step": 1653
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.440158749860251e-05,
-      "loss": 0.0109,
+      "loss": 0.0133,
       "step": 1654
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4386052892111522e-05,
-      "loss": 0.0054,
+      "loss": 0.01,
       "step": 1655
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4370518945187349e-05,
-      "loss": 0.0129,
+      "loss": 0.0115,
       "step": 1656
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.435498567451818e-05,
-      "loss": 0.0046,
+      "loss": 0.0055,
       "step": 1657
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4339453096791505e-05,
-      "loss": 0.0071,
+      "loss": 0.008,
       "step": 1658
     },
     {
@@ -9989,97 +9989,97 @@
     {
       "epoch": 0.53,
       "learning_rate": 1.4308390086911819e-05,
-      "loss": 0.0066,
+      "loss": 0.0049,
       "step": 1660
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.429285968812997e-05,
-      "loss": 0.0121,
+      "loss": 0.0117,
       "step": 1661
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4277330049032904e-05,
-      "loss": 0.0057,
+      "loss": 0.0063,
       "step": 1662
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4261801186304205e-05,
-      "loss": 0.0076,
+      "loss": 0.0071,
       "step": 1663
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4246273116626615e-05,
-      "loss": 0.005,
+      "loss": 0.0053,
       "step": 1664
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4230745856682027e-05,
-      "loss": 0.0119,
+      "loss": 0.0141,
       "step": 1665
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.421521942315146e-05,
-      "loss": 0.0042,
+      "loss": 0.004,
       "step": 1666
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4199693832715051e-05,
-      "loss": 0.0083,
+      "loss": 0.0087,
       "step": 1667
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4184169102052023e-05,
-      "loss": 0.0052,
+      "loss": 0.0065,
       "step": 1668
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4168645247840695e-05,
-      "loss": 0.0065,
+      "loss": 0.006,
       "step": 1669
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4153122286758414e-05,
-      "loss": 0.005,
+      "loss": 0.0054,
       "step": 1670
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4137600235481588e-05,
-      "loss": 0.0072,
+      "loss": 0.0071,
       "step": 1671
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4122079110685655e-05,
-      "loss": 0.0047,
+      "loss": 0.005,
       "step": 1672
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4106558929045028e-05,
-      "loss": 0.0054,
+      "loss": 0.0052,
       "step": 1673
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4091039707233144e-05,
-      "loss": 0.007,
+      "loss": 0.0075,
       "step": 1674
     },
     {
       "epoch": 0.53,
       "learning_rate": 1.4075521461922372e-05,
-      "loss": 0.0022,
+      "loss": 0.0021,
       "step": 1675
     },
     {
@@ -10091,145 +10091,145 @@
     {
       "epoch": 0.54,
       "learning_rate": 1.4044487967488475e-05,
-      "loss": 0.0112,
+      "loss": 0.0107,
       "step": 1677
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.4028972751704806e-05,
-      "loss": 0.0052,
+      "loss": 0.0041,
       "step": 1678
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.4013458579101131e-05,
-      "loss": 0.006,
+      "loss": 0.0059,
       "step": 1679
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3997945466344416e-05,
-      "loss": 0.0094,
+      "loss": 0.0096,
       "step": 1680
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3982433430100481e-05,
-      "loss": 0.0043,
+      "loss": 0.0024,
       "step": 1681
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3966922487034004e-05,
-      "loss": 0.007,
+      "loss": 0.0073,
       "step": 1682
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.395141265380846e-05,
-      "loss": 0.0075,
+      "loss": 0.0078,
       "step": 1683
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.393590394708616e-05,
-      "loss": 0.0078,
+      "loss": 0.0071,
       "step": 1684
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.39203963835282e-05,
-      "loss": 0.0088,
+      "loss": 0.0107,
       "step": 1685
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3904889979794428e-05,
-      "loss": 0.0036,
+      "loss": 0.0069,
       "step": 1686
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3889384752543477e-05,
-      "loss": 0.0053,
+      "loss": 0.0048,
       "step": 1687
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3873880718432681e-05,
-      "loss": 0.0036,
+      "loss": 0.003,
       "step": 1688
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.385837789411812e-05,
-      "loss": 0.0183,
+      "loss": 0.0152,
       "step": 1689
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3842876296254571e-05,
-      "loss": 0.0036,
+      "loss": 0.0034,
       "step": 1690
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3827375941495475e-05,
-      "loss": 0.011,
+      "loss": 0.0104,
       "step": 1691
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3811876846492955e-05,
-      "loss": 0.0023,
+      "loss": 0.0025,
       "step": 1692
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3796379027897777e-05,
-      "loss": 0.0138,
+      "loss": 0.0113,
       "step": 1693
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.378088250235933e-05,
-      "loss": 0.0054,
+      "loss": 0.005,
       "step": 1694
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3765387286525625e-05,
-      "loss": 0.009,
+      "loss": 0.0098,
       "step": 1695
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3749893397043248e-05,
-      "loss": 0.0046,
+      "loss": 0.0044,
       "step": 1696
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3734400850557384e-05,
-      "loss": 0.0086,
+      "loss": 0.0091,
       "step": 1697
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3718909663711749e-05,
-      "loss": 0.0023,
+      "loss": 0.0024,
       "step": 1698
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3703419853148617e-05,
-      "loss": 0.009,
+      "loss": 0.0095,
       "step": 1699
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3687931435508788e-05,
-      "loss": 0.008,
+      "loss": 0.0078,
       "step": 1700
     },
     {
@@ -10241,79 +10241,79 @@
     {
       "epoch": 0.54,
       "learning_rate": 1.3656958845554653e-05,
-      "loss": 0.007,
+      "loss": 0.0075,
       "step": 1702
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3641474706514393e-05,
-      "loss": 0.0075,
+      "loss": 0.0069,
       "step": 1703
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3625992026945434e-05,
-      "loss": 0.0081,
+      "loss": 0.0054,
       "step": 1704
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3610510823480924e-05,
-      "loss": 0.0077,
+      "loss": 0.0066,
       "step": 1705
     },
     {
       "epoch": 0.54,
       "learning_rate": 1.3595031112752394e-05,
-      "loss": 0.0094,
+      "loss": 0.009,
       "step": 1706
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3579552911389782e-05,
-      "loss": 0.0051,
+      "loss": 0.0042,
       "step": 1707
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.356407623602142e-05,
-      "loss": 0.0127,
+      "loss": 0.0141,
       "step": 1708
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3548601103273969e-05,
-      "loss": 0.0044,
+      "loss": 0.0043,
       "step": 1709
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3533127529772469e-05,
-      "loss": 0.0071,
+      "loss": 0.0075,
       "step": 1710
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3517655532140251e-05,
-      "loss": 0.0053,
+      "loss": 0.0055,
       "step": 1711
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3502185126998973e-05,
-      "loss": 0.0057,
+      "loss": 0.0064,
       "step": 1712
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3486716330968586e-05,
-      "loss": 0.0088,
+      "loss": 0.0085,
       "step": 1713
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3471249160667291e-05,
-      "loss": 0.0078,
+      "loss": 0.0129,
       "step": 1714
     },
     {
@@ -10325,67 +10325,67 @@
     {
       "epoch": 0.55,
       "learning_rate": 1.3440319763716108e-05,
-      "loss": 0.0072,
+      "loss": 0.0078,
       "step": 1716
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3424857570293837e-05,
-      "loss": 0.0049,
+      "loss": 0.0053,
       "step": 1717
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3409397069055885e-05,
-      "loss": 0.0051,
+      "loss": 0.0049,
       "step": 1718
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3393938276611538e-05,
-      "loss": 0.0077,
+      "loss": 0.0054,
       "step": 1719
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3378481209568276e-05,
-      "loss": 0.0025,
+      "loss": 0.0032,
       "step": 1720
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3363025884531705e-05,
-      "loss": 0.0109,
+      "loss": 0.0088,
       "step": 1721
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3347572318105572e-05,
-      "loss": 0.0035,
+      "loss": 0.0047,
       "step": 1722
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3332120526891729e-05,
-      "loss": 0.0027,
+      "loss": 0.0035,
       "step": 1723
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3316670527490116e-05,
-      "loss": 0.0081,
+      "loss": 0.0069,
       "step": 1724
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3301222336498755e-05,
-      "loss": 0.0055,
+      "loss": 0.0061,
       "step": 1725
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3285775970513738e-05,
-      "loss": 0.0065,
+      "loss": 0.006,
       "step": 1726
     },
     {
@@ -10397,31 +10397,31 @@
     {
       "epoch": 0.55,
       "learning_rate": 1.3254888779937178e-05,
-      "loss": 0.0096,
+      "loss": 0.0094,
       "step": 1728
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3239447988527922e-05,
-      "loss": 0.005,
+      "loss": 0.0048,
       "step": 1729
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3224009088489526e-05,
-      "loss": 0.0117,
+      "loss": 0.0048,
       "step": 1730
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3208572096408092e-05,
-      "loss": 0.0085,
+      "loss": 0.0084,
       "step": 1731
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3193137028867654e-05,
-      "loss": 0.0076,
+      "loss": 0.0068,
       "step": 1732
     },
     {
@@ -10433,223 +10433,223 @@
     {
       "epoch": 0.55,
       "learning_rate": 1.3162272733735627e-05,
-      "loss": 0.0094,
+      "loss": 0.0105,
       "step": 1734
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3146843539301719e-05,
-      "loss": 0.009,
+      "loss": 0.0089,
       "step": 1735
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3131416335724148e-05,
-      "loss": 0.0042,
+      "loss": 0.0038,
       "step": 1736
     },
     {
       "epoch": 0.55,
       "learning_rate": 1.3115991139576444e-05,
-      "loss": 0.0072,
+      "loss": 0.0083,
       "step": 1737
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.310056796742998e-05,
-      "loss": 0.0095,
+      "loss": 0.0083,
       "step": 1738
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.3085146835853963e-05,
-      "loss": 0.0126,
+      "loss": 0.0106,
       "step": 1739
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.3069727761415389e-05,
-      "loss": 0.0093,
+      "loss": 0.0088,
       "step": 1740
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.3054310760679072e-05,
-      "loss": 0.0044,
+      "loss": 0.0074,
       "step": 1741
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.3038895850207565e-05,
-      "loss": 0.0061,
+      "loss": 0.0067,
       "step": 1742
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.3023483046561198e-05,
-      "loss": 0.0059,
+      "loss": 0.0074,
       "step": 1743
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.3008072366298045e-05,
-      "loss": 0.0076,
+      "loss": 0.0064,
       "step": 1744
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2992663825973866e-05,
-      "loss": 0.005,
+      "loss": 0.0048,
       "step": 1745
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2977257442142155e-05,
-      "loss": 0.0067,
+      "loss": 0.0075,
       "step": 1746
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2961853231354071e-05,
-      "loss": 0.0107,
+      "loss": 0.0139,
       "step": 1747
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2946451210158447e-05,
-      "loss": 0.0101,
+      "loss": 0.0134,
       "step": 1748
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2931051395101755e-05,
-      "loss": 0.0088,
+      "loss": 0.0087,
       "step": 1749
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2915653802728108e-05,
-      "loss": 0.0046,
+      "loss": 0.0103,
       "step": 1750
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.290025844957922e-05,
-      "loss": 0.0074,
+      "loss": 0.0066,
       "step": 1751
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2884865352194414e-05,
-      "loss": 0.0125,
+      "loss": 0.0104,
       "step": 1752
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.286947452711057e-05,
-      "loss": 0.0054,
+      "loss": 0.006,
       "step": 1753
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2854085990862145e-05,
-      "loss": 0.0118,
+      "loss": 0.0104,
       "step": 1754
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2838699759981123e-05,
-      "loss": 0.0045,
+      "loss": 0.0036,
       "step": 1755
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.282331585099702e-05,
-      "loss": 0.0065,
+      "loss": 0.0045,
       "step": 1756
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2807934280436862e-05,
-      "loss": 0.0303,
+      "loss": 0.023,
       "step": 1757
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2792555064825145e-05,
-      "loss": 0.0029,
+      "loss": 0.0026,
       "step": 1758
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2777178220683857e-05,
-      "loss": 0.0073,
+      "loss": 0.0083,
       "step": 1759
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2761803764532418e-05,
-      "loss": 0.0114,
+      "loss": 0.0165,
       "step": 1760
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2746431712887701e-05,
-      "loss": 0.0064,
+      "loss": 0.0055,
       "step": 1761
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.273106208226398e-05,
-      "loss": 0.006,
+      "loss": 0.01,
       "step": 1762
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.271569488917294e-05,
-      "loss": 0.0093,
+      "loss": 0.0119,
       "step": 1763
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2700330150123633e-05,
-      "loss": 0.0098,
+      "loss": 0.0103,
       "step": 1764
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2684967881622498e-05,
-      "loss": 0.005,
+      "loss": 0.0055,
       "step": 1765
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2669608100173293e-05,
-      "loss": 0.0064,
+      "loss": 0.0074,
       "step": 1766
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.265425082227713e-05,
-      "loss": 0.0087,
+      "loss": 0.0101,
       "step": 1767
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2638896064432405e-05,
-      "loss": 0.0083,
+      "loss": 0.0081,
       "step": 1768
     },
     {
       "epoch": 0.56,
       "learning_rate": 1.2623543843134825e-05,
-      "loss": 0.0057,
+      "loss": 0.0065,
       "step": 1769
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2608194174877376e-05,
-      "loss": 0.0063,
+      "loss": 0.0061,
       "step": 1770
     },
     {
@@ -10661,199 +10661,199 @@
     {
       "epoch": 0.57,
       "learning_rate": 1.2577502563441019e-05,
-      "loss": 0.0091,
+      "loss": 0.0102,
       "step": 1772
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2562160653234278e-05,
-      "loss": 0.0105,
+      "loss": 0.0117,
       "step": 1773
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2546821362011969e-05,
-      "loss": 0.0109,
+      "loss": 0.01,
       "step": 1774
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2531484706253173e-05,
-      "loss": 0.0077,
+      "loss": 0.005,
       "step": 1775
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2516150702434147e-05,
-      "loss": 0.0119,
+      "loss": 0.0108,
       "step": 1776
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2500819367028291e-05,
-      "loss": 0.0156,
+      "loss": 0.009,
       "step": 1777
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2485490716506156e-05,
-      "loss": 0.0043,
+      "loss": 0.004,
       "step": 1778
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2470164767335381e-05,
-      "loss": 0.0102,
+      "loss": 0.0092,
       "step": 1779
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2454841535980735e-05,
-      "loss": 0.0062,
+      "loss": 0.0077,
       "step": 1780
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2439521038904036e-05,
-      "loss": 0.0059,
+      "loss": 0.0156,
       "step": 1781
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2424203292564188e-05,
-      "loss": 0.0018,
+      "loss": 0.0021,
       "step": 1782
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2408888313417135e-05,
-      "loss": 0.0071,
+      "loss": 0.0089,
       "step": 1783
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2393576117915835e-05,
-      "loss": 0.0075,
+      "loss": 0.0091,
       "step": 1784
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2378266722510276e-05,
-      "loss": 0.006,
+      "loss": 0.0067,
       "step": 1785
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2362960143647415e-05,
-      "loss": 0.0074,
+      "loss": 0.0068,
       "step": 1786
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2347656397771202e-05,
-      "loss": 0.0082,
+      "loss": 0.0072,
       "step": 1787
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2332355501322541e-05,
-      "loss": 0.0078,
+      "loss": 0.0071,
       "step": 1788
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2317057470739262e-05,
-      "loss": 0.0048,
+      "loss": 0.0047,
       "step": 1789
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2301762322456126e-05,
-      "loss": 0.0083,
+      "loss": 0.0078,
       "step": 1790
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2286470072904805e-05,
-      "loss": 0.0153,
+      "loss": 0.006,
       "step": 1791
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2271180738513834e-05,
-      "loss": 0.0056,
+      "loss": 0.0058,
       "step": 1792
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2255894335708647e-05,
-      "loss": 0.0071,
+      "loss": 0.0074,
       "step": 1793
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2240610880911494e-05,
-      "loss": 0.0084,
+      "loss": 0.0113,
       "step": 1794
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2225330390541483e-05,
-      "loss": 0.0075,
+      "loss": 0.007,
       "step": 1795
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2210052881014543e-05,
-      "loss": 0.0036,
+      "loss": 0.0073,
       "step": 1796
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2194778368743366e-05,
-      "loss": 0.003,
+      "loss": 0.0031,
       "step": 1797
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2179506870137464e-05,
-      "loss": 0.0138,
+      "loss": 0.0148,
       "step": 1798
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2164238401603078e-05,
-      "loss": 0.0051,
+      "loss": 0.0039,
       "step": 1799
     },
     {
       "epoch": 0.57,
       "learning_rate": 1.2148972979543219e-05,
-      "loss": 0.004,
+      "loss": 0.0041,
       "step": 1800
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.213371062035762e-05,
-      "loss": 0.0152,
+      "loss": 0.012,
       "step": 1801
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.2118451340442703e-05,
-      "loss": 0.0104,
+      "loss": 0.0113,
       "step": 1802
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.2103195156191613e-05,
-      "loss": 0.0046,
+      "loss": 0.0058,
       "step": 1803
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.2087942083994145e-05,
-      "loss": 0.0065,
+      "loss": 0.0043,
       "step": 1804
     },
     {
@@ -10865,259 +10865,259 @@
     {
       "epoch": 0.58,
       "learning_rate": 1.2057445341302564e-05,
-      "loss": 0.0065,
+      "loss": 0.0052,
       "step": 1806
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.2042201703571267e-05,
-      "loss": 0.0049,
+      "loss": 0.0046,
       "step": 1807
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.2026961243419202e-05,
-      "loss": 0.0128,
+      "loss": 0.012,
       "step": 1808
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.2011723977219286e-05,
-      "loss": 0.0088,
+      "loss": 0.0083,
       "step": 1809
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1996489921340985e-05,
-      "loss": 0.0134,
+      "loss": 0.0119,
       "step": 1810
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1981259092150347e-05,
-      "loss": 0.0069,
+      "loss": 0.0097,
       "step": 1811
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.196603150600992e-05,
-      "loss": 0.0086,
+      "loss": 0.0087,
       "step": 1812
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1950807179278794e-05,
-      "loss": 0.0104,
+      "loss": 0.0091,
       "step": 1813
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1935586128312554e-05,
-      "loss": 0.0035,
+      "loss": 0.0036,
       "step": 1814
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1920368369463245e-05,
-      "loss": 0.0048,
+      "loss": 0.008,
       "step": 1815
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1905153919079404e-05,
-      "loss": 0.0022,
+      "loss": 0.0024,
       "step": 1816
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.188994279350599e-05,
-      "loss": 0.0069,
+      "loss": 0.0075,
       "step": 1817
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1874735009084407e-05,
-      "loss": 0.0071,
+      "loss": 0.0072,
       "step": 1818
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1859530582152464e-05,
-      "loss": 0.0098,
+      "loss": 0.0072,
       "step": 1819
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1844329529044352e-05,
-      "loss": 0.0097,
+      "loss": 0.0103,
       "step": 1820
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1829131866090651e-05,
-      "loss": 0.0081,
+      "loss": 0.0071,
       "step": 1821
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1813937609618308e-05,
-      "loss": 0.0036,
+      "loss": 0.004,
       "step": 1822
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1798746775950578e-05,
-      "loss": 0.0069,
+      "loss": 0.0062,
       "step": 1823
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1783559381407074e-05,
-      "loss": 0.0084,
+      "loss": 0.0055,
       "step": 1824
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1768375442303687e-05,
-      "loss": 0.0078,
+      "loss": 0.0101,
       "step": 1825
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1753194974952615e-05,
-      "loss": 0.0124,
+      "loss": 0.0139,
       "step": 1826
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.173801799566232e-05,
-      "loss": 0.0091,
+      "loss": 0.0072,
       "step": 1827
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1722844520737509e-05,
-      "loss": 0.0108,
+      "loss": 0.0106,
       "step": 1828
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1707674566479139e-05,
-      "loss": 0.008,
+      "loss": 0.0087,
       "step": 1829
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1692508149184366e-05,
-      "loss": 0.0066,
+      "loss": 0.0067,
       "step": 1830
     },
     {
       "epoch": 0.58,
       "learning_rate": 1.1677345285146566e-05,
-      "loss": 0.0089,
+      "loss": 0.0056,
       "step": 1831
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1662185990655285e-05,
-      "loss": 0.0053,
+      "loss": 0.005,
       "step": 1832
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.164703028199624e-05,
-      "loss": 0.006,
+      "loss": 0.0061,
       "step": 1833
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1631878175451285e-05,
-      "loss": 0.0025,
+      "loss": 0.0027,
       "step": 1834
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.161672968729843e-05,
-      "loss": 0.0079,
+      "loss": 0.008,
       "step": 1835
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.160158483381176e-05,
-      "loss": 0.0074,
+      "loss": 0.0078,
       "step": 1836
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1586443631261493e-05,
-      "loss": 0.0128,
+      "loss": 0.0086,
       "step": 1837
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1571306095913895e-05,
-      "loss": 0.0058,
+      "loss": 0.0063,
       "step": 1838
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1556172244031306e-05,
-      "loss": 0.008,
+      "loss": 0.0075,
       "step": 1839
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1541042091872119e-05,
-      "loss": 0.0144,
+      "loss": 0.0111,
       "step": 1840
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1525915655690727e-05,
-      "loss": 0.0037,
+      "loss": 0.0036,
       "step": 1841
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.151079295173756e-05,
-      "loss": 0.0034,
+      "loss": 0.0037,
       "step": 1842
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1495673996259006e-05,
-      "loss": 0.0059,
+      "loss": 0.0055,
       "step": 1843
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1480558805497455e-05,
-      "loss": 0.0108,
+      "loss": 0.0099,
       "step": 1844
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1465447395691237e-05,
-      "loss": 0.0079,
+      "loss": 0.0112,
       "step": 1845
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1450339783074626e-05,
-      "loss": 0.005,
+      "loss": 0.0049,
       "step": 1846
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1435235983877812e-05,
-      "loss": 0.006,
+      "loss": 0.0059,
       "step": 1847
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1420136014326893e-05,
-      "loss": 0.0064,
+      "loss": 0.0075,
       "step": 1848
     },
     {
@@ -11129,571 +11129,571 @@
     {
       "epoch": 0.59,
       "learning_rate": 1.138994762904653e-05,
-      "loss": 0.0138,
+      "loss": 0.0107,
       "step": 1850
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1374859245748637e-05,
-      "loss": 0.0118,
+      "loss": 0.0082,
       "step": 1851
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1359774756959704e-05,
-      "loss": 0.0062,
+      "loss": 0.0061,
       "step": 1852
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1344694178885088e-05,
-      "loss": 0.0061,
+      "loss": 0.0062,
       "step": 1853
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1329617527725926e-05,
-      "loss": 0.0095,
+      "loss": 0.0117,
       "step": 1854
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1314544819679165e-05,
-      "loss": 0.0092,
+      "loss": 0.0071,
       "step": 1855
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.129947607093748e-05,
-      "loss": 0.0038,
+      "loss": 0.0037,
       "step": 1856
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1284411297689326e-05,
-      "loss": 0.0093,
+      "loss": 0.007,
       "step": 1857
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1269350516118871e-05,
-      "loss": 0.0034,
+      "loss": 0.0035,
       "step": 1858
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1254293742405992e-05,
-      "loss": 0.0041,
+      "loss": 0.0044,
       "step": 1859
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.123924099272627e-05,
-      "loss": 0.0206,
+      "loss": 0.0197,
       "step": 1860
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1224192283250955e-05,
-      "loss": 0.0033,
+      "loss": 0.0041,
       "step": 1861
     },
     {
       "epoch": 0.59,
       "learning_rate": 1.1209147630146955e-05,
-      "loss": 0.0151,
+      "loss": 0.0164,
       "step": 1862
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.119410704957684e-05,
-      "loss": 0.0049,
+      "loss": 0.0047,
       "step": 1863
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1179070557698772e-05,
-      "loss": 0.0054,
+      "loss": 0.0048,
       "step": 1864
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1164038170666547e-05,
-      "loss": 0.0103,
+      "loss": 0.0092,
       "step": 1865
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1149009904629549e-05,
-      "loss": 0.0099,
+      "loss": 0.0087,
       "step": 1866
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.113398577573271e-05,
-      "loss": 0.0075,
+      "loss": 0.0078,
       "step": 1867
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1118965800116555e-05,
-      "loss": 0.0049,
+      "loss": 0.0042,
       "step": 1868
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1103949993917115e-05,
-      "loss": 0.0085,
+      "loss": 0.013,
       "step": 1869
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1088938373265956e-05,
-      "loss": 0.0057,
+      "loss": 0.0073,
       "step": 1870
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1073930954290157e-05,
-      "loss": 0.0108,
+      "loss": 0.0106,
       "step": 1871
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.105892775311226e-05,
-      "loss": 0.0069,
+      "loss": 0.0108,
       "step": 1872
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.104392878585029e-05,
-      "loss": 0.011,
+      "loss": 0.0094,
       "step": 1873
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1028934068617722e-05,
-      "loss": 0.0043,
+      "loss": 0.0044,
       "step": 1874
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.1013943617523462e-05,
-      "loss": 0.0061,
+      "loss": 0.0043,
       "step": 1875
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0998957448671848e-05,
-      "loss": 0.0074,
+      "loss": 0.0067,
       "step": 1876
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0983975578162586e-05,
-      "loss": 0.0054,
+      "loss": 0.0059,
       "step": 1877
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0968998022090798e-05,
-      "loss": 0.0078,
+      "loss": 0.0081,
       "step": 1878
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0954024796546947e-05,
-      "loss": 0.0147,
+      "loss": 0.0148,
       "step": 1879
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0939055917616855e-05,
-      "loss": 0.0126,
+      "loss": 0.0132,
       "step": 1880
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.092409140138168e-05,
-      "loss": 0.005,
+      "loss": 0.0069,
       "step": 1881
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0909131263917874e-05,
-      "loss": 0.0066,
+      "loss": 0.0063,
       "step": 1882
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.08941755212972e-05,
-      "loss": 0.0026,
+      "loss": 0.0025,
       "step": 1883
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0879224189586705e-05,
-      "loss": 0.0052,
+      "loss": 0.0037,
       "step": 1884
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0864277284848672e-05,
-      "loss": 0.0054,
+      "loss": 0.0058,
       "step": 1885
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0849334823140661e-05,
-      "loss": 0.0062,
+      "loss": 0.006,
       "step": 1886
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.083439682051543e-05,
-      "loss": 0.0061,
+      "loss": 0.0062,
       "step": 1887
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0819463293020957e-05,
-      "loss": 0.0107,
+      "loss": 0.0109,
       "step": 1888
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0804534256700429e-05,
-      "loss": 0.0051,
+      "loss": 0.0044,
       "step": 1889
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0789609727592175e-05,
-      "loss": 0.0073,
+      "loss": 0.0059,
       "step": 1890
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0774689721729715e-05,
-      "loss": 0.0056,
+      "loss": 0.0067,
       "step": 1891
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0759774255141681e-05,
-      "loss": 0.0058,
+      "loss": 0.0064,
       "step": 1892
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0744863343851853e-05,
-      "loss": 0.0049,
+      "loss": 0.0059,
       "step": 1893
     },
     {
       "epoch": 0.6,
       "learning_rate": 1.0729957003879104e-05,
-      "loss": 0.0065,
+      "loss": 0.0073,
       "step": 1894
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0715055251237393e-05,
-      "loss": 0.0107,
+      "loss": 0.0087,
       "step": 1895
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.070015810193576e-05,
-      "loss": 0.0042,
+      "loss": 0.0045,
       "step": 1896
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0685265571978302e-05,
-      "loss": 0.003,
+      "loss": 0.0034,
       "step": 1897
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.067037767736413e-05,
-      "loss": 0.0036,
+      "loss": 0.0047,
       "step": 1898
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0655494434087414e-05,
-      "loss": 0.0077,
+      "loss": 0.0084,
       "step": 1899
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0640615858137289e-05,
-      "loss": 0.0093,
+      "loss": 0.0091,
       "step": 1900
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0625741965497897e-05,
-      "loss": 0.0103,
+      "loss": 0.0102,
       "step": 1901
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.061087277214835e-05,
-      "loss": 0.0056,
+      "loss": 0.0073,
       "step": 1902
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0596008294062692e-05,
-      "loss": 0.0064,
+      "loss": 0.007,
       "step": 1903
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.058114854720993e-05,
-      "loss": 0.0047,
+      "loss": 0.0048,
       "step": 1904
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.056629354755396e-05,
-      "loss": 0.0041,
+      "loss": 0.0047,
       "step": 1905
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0551443311053596e-05,
-      "loss": 0.008,
+      "loss": 0.0085,
       "step": 1906
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0536597853662539e-05,
-      "loss": 0.006,
+      "loss": 0.0059,
       "step": 1907
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.052175719132933e-05,
-      "loss": 0.0152,
+      "loss": 0.0162,
       "step": 1908
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0506921339997387e-05,
-      "loss": 0.0119,
+      "loss": 0.0096,
       "step": 1909
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0492090315604952e-05,
-      "loss": 0.0039,
+      "loss": 0.0043,
       "step": 1910
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0477264134085063e-05,
-      "loss": 0.0095,
+      "loss": 0.0107,
       "step": 1911
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0462442811365582e-05,
-      "loss": 0.0106,
+      "loss": 0.0105,
       "step": 1912
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.044762636336913e-05,
-      "loss": 0.0084,
+      "loss": 0.0104,
       "step": 1913
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0432814806013104e-05,
-      "loss": 0.004,
+      "loss": 0.0034,
       "step": 1914
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0418008155209646e-05,
-      "loss": 0.0156,
+      "loss": 0.0132,
       "step": 1915
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0403206426865619e-05,
-      "loss": 0.0072,
+      "loss": 0.0084,
       "step": 1916
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0388409636882606e-05,
-      "loss": 0.0042,
+      "loss": 0.0036,
       "step": 1917
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0373617801156872e-05,
-      "loss": 0.0071,
+      "loss": 0.0045,
       "step": 1918
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0358830935579378e-05,
-      "loss": 0.0027,
+      "loss": 0.0033,
       "step": 1919
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0344049056035741e-05,
-      "loss": 0.0098,
+      "loss": 0.0088,
       "step": 1920
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.03292721784062e-05,
-      "loss": 0.009,
+      "loss": 0.0096,
       "step": 1921
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0314500318565653e-05,
-      "loss": 0.0058,
+      "loss": 0.0051,
       "step": 1922
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0299733492383582e-05,
-      "loss": 0.0029,
+      "loss": 0.0033,
       "step": 1923
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0284971715724073e-05,
-      "loss": 0.0068,
+      "loss": 0.0072,
       "step": 1924
     },
     {
       "epoch": 0.61,
       "learning_rate": 1.0270215004445793e-05,
-      "loss": 0.0053,
+      "loss": 0.0061,
       "step": 1925
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0255463374401945e-05,
-      "loss": 0.0061,
+      "loss": 0.0067,
       "step": 1926
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0240716841440301e-05,
-      "loss": 0.0153,
+      "loss": 0.0163,
       "step": 1927
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0225975421403135e-05,
-      "loss": 0.0053,
+      "loss": 0.0056,
       "step": 1928
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.021123913012724e-05,
-      "loss": 0.0072,
+      "loss": 0.0067,
       "step": 1929
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0196507983443899e-05,
-      "loss": 0.0061,
+      "loss": 0.0052,
       "step": 1930
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0181781997178859e-05,
-      "loss": 0.0068,
+      "loss": 0.0069,
       "step": 1931
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.016706118715233e-05,
-      "loss": 0.0036,
+      "loss": 0.0048,
       "step": 1932
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.015234556917897e-05,
-      "loss": 0.0083,
+      "loss": 0.0095,
       "step": 1933
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0137635159067834e-05,
-      "loss": 0.0095,
+      "loss": 0.0065,
       "step": 1934
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0122929972622414e-05,
-      "loss": 0.0087,
+      "loss": 0.0074,
       "step": 1935
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.010823002564056e-05,
-      "loss": 0.0078,
+      "loss": 0.0094,
       "step": 1936
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0093535333914513e-05,
-      "loss": 0.006,
+      "loss": 0.0058,
       "step": 1937
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0078845913230868e-05,
-      "loss": 0.0066,
+      "loss": 0.0081,
       "step": 1938
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.006416177937054e-05,
-      "loss": 0.0134,
+      "loss": 0.0081,
       "step": 1939
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0049482948108787e-05,
-      "loss": 0.0168,
+      "loss": 0.016,
       "step": 1940
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0034809435215154e-05,
-      "loss": 0.0092,
+      "loss": 0.0085,
       "step": 1941
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0020141256453477e-05,
-      "loss": 0.0129,
+      "loss": 0.0121,
       "step": 1942
     },
     {
       "epoch": 0.62,
       "learning_rate": 1.0005478427581865e-05,
-      "loss": 0.0038,
+      "loss": 0.0039,
       "step": 1943
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.990820964352678e-06,
-      "loss": 0.004,
+      "loss": 0.005,
       "step": 1944
     },
     {
@@ -11705,79 +11705,79 @@
     {
       "epoch": 0.62,
       "learning_rate": 9.961522197802166e-06,
-      "loss": 0.0054,
+      "loss": 0.0044,
       "step": 1946
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.946880925956668e-06,
-      "loss": 0.0246,
+      "loss": 0.0071,
       "step": 1947
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.932245082705223e-06,
-      "loss": 0.0114,
+      "loss": 0.01,
       "step": 1948
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.917614683771187e-06,
-      "loss": 0.0038,
+      "loss": 0.0036,
       "step": 1949
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.902989744872083e-06,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 1950
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.888370281719573e-06,
-      "loss": 0.0068,
+      "loss": 0.0061,
       "step": 1951
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.873756310019417e-06,
-      "loss": 0.0123,
+      "loss": 0.0117,
       "step": 1952
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.859147845471495e-06,
-      "loss": 0.0041,
+      "loss": 0.0044,
       "step": 1953
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.84454490376975e-06,
-      "loss": 0.0115,
+      "loss": 0.0156,
       "step": 1954
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.829947500602217e-06,
-      "loss": 0.0082,
+      "loss": 0.0106,
       "step": 1955
     },
     {
       "epoch": 0.62,
       "learning_rate": 9.815355651650962e-06,
-      "loss": 0.0045,
+      "loss": 0.0044,
       "step": 1956
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.800769372592088e-06,
-      "loss": 0.0098,
+      "loss": 0.009,
       "step": 1957
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.78618867909572e-06,
-      "loss": 0.0041,
+      "loss": 0.0042,
       "step": 1958
     },
     {
@@ -11789,187 +11789,187 @@
     {
       "epoch": 0.63,
       "learning_rate": 9.757044111440957e-06,
-      "loss": 0.0064,
+      "loss": 0.0059,
       "step": 1960
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.742480268592742e-06,
-      "loss": 0.0057,
+      "loss": 0.0055,
       "step": 1961
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.727922073927334e-06,
-      "loss": 0.0057,
+      "loss": 0.006,
       "step": 1962
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.713369543084689e-06,
-      "loss": 0.0068,
+      "loss": 0.0059,
       "step": 1963
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.69882269169868e-06,
-      "loss": 0.0104,
+      "loss": 0.0115,
       "step": 1964
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.684281535397057e-06,
-      "loss": 0.0067,
+      "loss": 0.0058,
       "step": 1965
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.669746089801478e-06,
-      "loss": 0.0043,
+      "loss": 0.004,
       "step": 1966
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.655216370527445e-06,
-      "loss": 0.006,
+      "loss": 0.0051,
       "step": 1967
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.640692393184316e-06,
-      "loss": 0.0055,
+      "loss": 0.006,
       "step": 1968
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.62617417337529e-06,
-      "loss": 0.0083,
+      "loss": 0.0086,
       "step": 1969
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.611661726697365e-06,
-      "loss": 0.0093,
+      "loss": 0.0087,
       "step": 1970
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.59715506874134e-06,
-      "loss": 0.0051,
+      "loss": 0.0061,
       "step": 1971
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.58265421509181e-06,
-      "loss": 0.0045,
+      "loss": 0.0058,
       "step": 1972
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.568159181327108e-06,
-      "loss": 0.0097,
+      "loss": 0.0099,
       "step": 1973
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.553669983019345e-06,
-      "loss": 0.009,
+      "loss": 0.0076,
       "step": 1974
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.539186635734336e-06,
-      "loss": 0.0051,
+      "loss": 0.0048,
       "step": 1975
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.524709155031624e-06,
-      "loss": 0.0057,
+      "loss": 0.0069,
       "step": 1976
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.510237556464458e-06,
-      "loss": 0.0093,
+      "loss": 0.0102,
       "step": 1977
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.495771855579742e-06,
-      "loss": 0.0148,
+      "loss": 0.0149,
       "step": 1978
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.481312067918074e-06,
-      "loss": 0.0115,
+      "loss": 0.0079,
       "step": 1979
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.466858209013668e-06,
-      "loss": 0.0119,
+      "loss": 0.0094,
       "step": 1980
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.452410294394399e-06,
-      "loss": 0.009,
+      "loss": 0.0073,
       "step": 1981
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.43796833958174e-06,
-      "loss": 0.0066,
+      "loss": 0.0101,
       "step": 1982
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.423532360090755e-06,
-      "loss": 0.0065,
+      "loss": 0.0094,
       "step": 1983
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.409102371430107e-06,
-      "loss": 0.0047,
+      "loss": 0.0038,
       "step": 1984
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.394678389102014e-06,
-      "loss": 0.0073,
+      "loss": 0.0076,
       "step": 1985
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.38026042860223e-06,
-      "loss": 0.0063,
+      "loss": 0.0074,
       "step": 1986
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.365848505420066e-06,
-      "loss": 0.0092,
+      "loss": 0.0089,
       "step": 1987
     },
     {
       "epoch": 0.63,
       "learning_rate": 9.351442635038315e-06,
-      "loss": 0.0073,
+      "loss": 0.0083,
       "step": 1988
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.337042832933291e-06,
-      "loss": 0.0045,
+      "loss": 0.0059,
       "step": 1989
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.322649114574791e-06,
-      "loss": 0.004,
+      "loss": 0.0039,
       "step": 1990
     },
     {
@@ -11981,373 +11981,373 @@
     {
       "epoch": 0.64,
       "learning_rate": 9.293879990943788e-06,
-      "loss": 0.0203,
+      "loss": 0.0183,
       "step": 1992
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.279504616578114e-06,
-      "loss": 0.0055,
+      "loss": 0.0062,
       "step": 1993
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.265135387772586e-06,
-      "loss": 0.0064,
+      "loss": 0.0073,
       "step": 1994
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.250772319964148e-06,
-      "loss": 0.0089,
+      "loss": 0.0079,
       "step": 1995
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.236415428583113e-06,
-      "loss": 0.0039,
+      "loss": 0.0029,
       "step": 1996
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.222064729053185e-06,
-      "loss": 0.0128,
+      "loss": 0.012,
       "step": 1997
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.207720236791382e-06,
-      "loss": 0.0072,
+      "loss": 0.0086,
       "step": 1998
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.19338196720809e-06,
-      "loss": 0.0057,
+      "loss": 0.0084,
       "step": 1999
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.179049935706984e-06,
-      "loss": 0.0099,
+      "loss": 0.0121,
       "step": 2000
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.164724157685042e-06,
-      "loss": 0.0073,
+      "loss": 0.0071,
       "step": 2001
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.15040464853253e-06,
-      "loss": 0.0063,
+      "loss": 0.0062,
       "step": 2002
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.136091423632984e-06,
-      "loss": 0.008,
+      "loss": 0.0081,
       "step": 2003
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.121784498363169e-06,
-      "loss": 0.0101,
+      "loss": 0.0123,
       "step": 2004
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.107483888093109e-06,
-      "loss": 0.0144,
+      "loss": 0.0092,
       "step": 2005
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.093189608186015e-06,
-      "loss": 0.0038,
+      "loss": 0.0043,
       "step": 2006
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.078901673998317e-06,
-      "loss": 0.0087,
+      "loss": 0.0048,
       "step": 2007
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.064620100879631e-06,
-      "loss": 0.0047,
+      "loss": 0.0037,
       "step": 2008
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.050344904172718e-06,
-      "loss": 0.0165,
+      "loss": 0.0138,
       "step": 2009
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.036076099213512e-06,
-      "loss": 0.014,
+      "loss": 0.0127,
       "step": 2010
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.021813701331059e-06,
-      "loss": 0.006,
+      "loss": 0.0074,
       "step": 2011
     },
     {
       "epoch": 0.64,
       "learning_rate": 9.007557725847543e-06,
-      "loss": 0.0086,
+      "loss": 0.0096,
       "step": 2012
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.993308188078232e-06,
-      "loss": 0.0092,
+      "loss": 0.0057,
       "step": 2013
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.979065103331487e-06,
-      "loss": 0.0052,
+      "loss": 0.0048,
       "step": 2014
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.964828486908729e-06,
-      "loss": 0.0089,
+      "loss": 0.0091,
       "step": 2015
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.950598354104444e-06,
-      "loss": 0.0048,
+      "loss": 0.0035,
       "step": 2016
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.936374720206133e-06,
-      "loss": 0.0043,
+      "loss": 0.0055,
       "step": 2017
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.922157600494331e-06,
-      "loss": 0.0095,
+      "loss": 0.0054,
       "step": 2018
     },
     {
       "epoch": 0.64,
       "learning_rate": 8.907947010242566e-06,
-      "loss": 0.0054,
+      "loss": 0.0064,
       "step": 2019
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.893742964717354e-06,
-      "loss": 0.0093,
+      "loss": 0.0094,
       "step": 2020
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.879545479178188e-06,
-      "loss": 0.0081,
+      "loss": 0.0096,
       "step": 2021
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.865354568877496e-06,
-      "loss": 0.0034,
+      "loss": 0.0039,
       "step": 2022
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.85117024906066e-06,
-      "loss": 0.0052,
+      "loss": 0.0051,
       "step": 2023
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.836992534965969e-06,
-      "loss": 0.0137,
+      "loss": 0.0101,
       "step": 2024
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.822821441824622e-06,
-      "loss": 0.0039,
+      "loss": 0.005,
       "step": 2025
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.808656984860704e-06,
-      "loss": 0.005,
+      "loss": 0.0054,
       "step": 2026
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.79449917929117e-06,
-      "loss": 0.0097,
+      "loss": 0.0108,
       "step": 2027
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.78034804032583e-06,
-      "loss": 0.0084,
+      "loss": 0.0093,
       "step": 2028
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.766203583167331e-06,
-      "loss": 0.006,
+      "loss": 0.0074,
       "step": 2029
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.752065823011135e-06,
-      "loss": 0.0116,
+      "loss": 0.0115,
       "step": 2030
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.737934775045536e-06,
-      "loss": 0.0057,
+      "loss": 0.0044,
       "step": 2031
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.723810454451579e-06,
-      "loss": 0.0057,
+      "loss": 0.0055,
       "step": 2032
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.70969287640311e-06,
-      "loss": 0.0058,
+      "loss": 0.0064,
       "step": 2033
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.695582056066723e-06,
-      "loss": 0.0063,
+      "loss": 0.0109,
       "step": 2034
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.681478008601741e-06,
-      "loss": 0.0073,
+      "loss": 0.0077,
       "step": 2035
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.667380749160231e-06,
-      "loss": 0.0038,
+      "loss": 0.0035,
       "step": 2036
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.653290292886953e-06,
-      "loss": 0.0083,
+      "loss": 0.0077,
       "step": 2037
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.63920665491936e-06,
-      "loss": 0.0052,
+      "loss": 0.0051,
       "step": 2038
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.625129850387593e-06,
-      "loss": 0.0062,
+      "loss": 0.0054,
       "step": 2039
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.611059894414431e-06,
-      "loss": 0.0028,
+      "loss": 0.0034,
       "step": 2040
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.596996802115315e-06,
-      "loss": 0.0079,
+      "loss": 0.0081,
       "step": 2041
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.582940588598296e-06,
-      "loss": 0.0111,
+      "loss": 0.0092,
       "step": 2042
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.568891268964044e-06,
-      "loss": 0.0115,
+      "loss": 0.0144,
       "step": 2043
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.55484885830583e-06,
-      "loss": 0.0048,
+      "loss": 0.0051,
       "step": 2044
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.540813371709484e-06,
-      "loss": 0.0034,
+      "loss": 0.003,
       "step": 2045
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.52678482425341e-06,
-      "loss": 0.0057,
+      "loss": 0.0062,
       "step": 2046
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.512763231008563e-06,
-      "loss": 0.0109,
+      "loss": 0.0124,
       "step": 2047
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.498748607038406e-06,
-      "loss": 0.0056,
+      "loss": 0.0055,
       "step": 2048
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.48474096739894e-06,
-      "loss": 0.0036,
+      "loss": 0.0054,
       "step": 2049
     },
     {
       "epoch": 0.65,
       "learning_rate": 8.470740327138637e-06,
-      "loss": 0.0074,
+      "loss": 0.0088,
       "step": 2050
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.456746701298473e-06,
-      "loss": 0.0061,
+      "loss": 0.0062,
       "step": 2051
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.442760104911882e-06,
-      "loss": 0.0053,
+      "loss": 0.0052,
       "step": 2052
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.42878055300473e-06,
-      "loss": 0.0091,
+      "loss": 0.0093,
       "step": 2053
     },
     {
@@ -12359,301 +12359,301 @@
     {
       "epoch": 0.66,
       "learning_rate": 8.400842642694429e-06,
-      "loss": 0.0063,
+      "loss": 0.0059,
       "step": 2055
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.386884314305126e-06,
-      "loss": 0.0089,
+      "loss": 0.0083,
       "step": 2056
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.37293309042295e-06,
-      "loss": 0.0033,
+      "loss": 0.0042,
       "step": 2057
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.358988986035765e-06,
-      "loss": 0.0042,
+      "loss": 0.0048,
       "step": 2058
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.345052016123812e-06,
-      "loss": 0.0083,
+      "loss": 0.0091,
       "step": 2059
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.331122195659646e-06,
-      "loss": 0.0061,
+      "loss": 0.0031,
       "step": 2060
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.31719953960816e-06,
-      "loss": 0.0061,
+      "loss": 0.006,
       "step": 2061
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.30328406292653e-06,
-      "loss": 0.0094,
+      "loss": 0.0132,
       "step": 2062
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.289375780564245e-06,
-      "loss": 0.0062,
+      "loss": 0.006,
       "step": 2063
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.275474707463034e-06,
-      "loss": 0.0055,
+      "loss": 0.0062,
       "step": 2064
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.261580858556911e-06,
-      "loss": 0.0029,
+      "loss": 0.0053,
       "step": 2065
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.247694248772105e-06,
-      "loss": 0.0047,
+      "loss": 0.0048,
       "step": 2066
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.23381489302709e-06,
-      "loss": 0.0055,
+      "loss": 0.0073,
       "step": 2067
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.219942806232524e-06,
-      "loss": 0.0073,
+      "loss": 0.0093,
       "step": 2068
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.206078003291273e-06,
-      "loss": 0.0061,
+      "loss": 0.0069,
       "step": 2069
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.192220499098374e-06,
-      "loss": 0.006,
+      "loss": 0.0054,
       "step": 2070
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.178370308541015e-06,
-      "loss": 0.0048,
+      "loss": 0.0046,
       "step": 2071
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.164527446498541e-06,
-      "loss": 0.0025,
+      "loss": 0.0038,
       "step": 2072
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.150691927842408e-06,
-      "loss": 0.0073,
+      "loss": 0.0088,
       "step": 2073
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.136863767436193e-06,
-      "loss": 0.0097,
+      "loss": 0.0062,
       "step": 2074
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.123042980135576e-06,
-      "loss": 0.0036,
+      "loss": 0.0038,
       "step": 2075
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.10922958078829e-06,
-      "loss": 0.0042,
+      "loss": 0.0039,
       "step": 2076
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.095423584234157e-06,
-      "loss": 0.0041,
+      "loss": 0.005,
       "step": 2077
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.081625005305036e-06,
-      "loss": 0.0066,
+      "loss": 0.0073,
       "step": 2078
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.067833858824814e-06,
-      "loss": 0.0061,
+      "loss": 0.0055,
       "step": 2079
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.054050159609401e-06,
-      "loss": 0.0045,
+      "loss": 0.0049,
       "step": 2080
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.040273922466696e-06,
-      "loss": 0.0063,
+      "loss": 0.0058,
       "step": 2081
     },
     {
       "epoch": 0.66,
       "learning_rate": 8.02650516219659e-06,
-      "loss": 0.0037,
+      "loss": 0.0023,
       "step": 2082
     },
     {
       "epoch": 0.67,
       "learning_rate": 8.012743893590943e-06,
-      "loss": 0.0075,
+      "loss": 0.0061,
       "step": 2083
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.998990131433554e-06,
-      "loss": 0.0027,
+      "loss": 0.0026,
       "step": 2084
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.985243890500181e-06,
-      "loss": 0.0025,
+      "loss": 0.0028,
       "step": 2085
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.971505185558471e-06,
-      "loss": 0.0092,
+      "loss": 0.0089,
       "step": 2086
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.957774031368004e-06,
-      "loss": 0.005,
+      "loss": 0.0062,
       "step": 2087
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.944050442680225e-06,
-      "loss": 0.0069,
+      "loss": 0.0091,
       "step": 2088
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.930334434238475e-06,
-      "loss": 0.0068,
+      "loss": 0.0067,
       "step": 2089
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.916626020777924e-06,
-      "loss": 0.0115,
+      "loss": 0.0116,
       "step": 2090
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.90292521702561e-06,
-      "loss": 0.0105,
+      "loss": 0.0112,
       "step": 2091
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.88923203770037e-06,
-      "loss": 0.005,
+      "loss": 0.0051,
       "step": 2092
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.875546497512879e-06,
-      "loss": 0.0048,
+      "loss": 0.0055,
       "step": 2093
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.861868611165572e-06,
-      "loss": 0.0034,
+      "loss": 0.0037,
       "step": 2094
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.848198393352689e-06,
-      "loss": 0.0038,
+      "loss": 0.0035,
       "step": 2095
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.834535858760225e-06,
-      "loss": 0.0101,
+      "loss": 0.0107,
       "step": 2096
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.820881022065902e-06,
-      "loss": 0.009,
+      "loss": 0.0067,
       "step": 2097
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.80723389793921e-06,
-      "loss": 0.0062,
+      "loss": 0.0096,
       "step": 2098
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.793594501041306e-06,
-      "loss": 0.01,
+      "loss": 0.008,
       "step": 2099
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.779962846025086e-06,
-      "loss": 0.0119,
+      "loss": 0.0115,
       "step": 2100
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.766338947535113e-06,
-      "loss": 0.0118,
+      "loss": 0.0068,
       "step": 2101
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.752722820207607e-06,
-      "loss": 0.0084,
+      "loss": 0.009,
       "step": 2102
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.739114478670463e-06,
-      "loss": 0.009,
+      "loss": 0.008,
       "step": 2103
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.725513937543184e-06,
-      "loss": 0.0039,
+      "loss": 0.0042,
       "step": 2104
     },
     {
@@ -12665,61 +12665,61 @@
     {
       "epoch": 0.67,
       "learning_rate": 7.698336314954408e-06,
-      "loss": 0.0085,
+      "loss": 0.0114,
       "step": 2106
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.684759262689971e-06,
-      "loss": 0.0028,
+      "loss": 0.0029,
       "step": 2107
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.67119006922952e-06,
-      "loss": 0.0062,
+      "loss": 0.0063,
       "step": 2108
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.65762874915052e-06,
-      "loss": 0.0065,
+      "loss": 0.0048,
       "step": 2109
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.644075317021962e-06,
-      "loss": 0.0026,
+      "loss": 0.0036,
       "step": 2110
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.630529787404383e-06,
-      "loss": 0.0062,
+      "loss": 0.007,
       "step": 2111
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.616992174849811e-06,
-      "loss": 0.0132,
+      "loss": 0.0112,
       "step": 2112
     },
     {
       "epoch": 0.67,
       "learning_rate": 7.603462493901785e-06,
-      "loss": 0.0077,
+      "loss": 0.0084,
       "step": 2113
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.589940759095322e-06,
-      "loss": 0.0098,
+      "loss": 0.0091,
       "step": 2114
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.576426984956885e-06,
-      "loss": 0.0029,
+      "loss": 0.003,
       "step": 2115
     },
     {
@@ -12731,133 +12731,133 @@
     {
       "epoch": 0.68,
       "learning_rate": 7.549423376747245e-06,
-      "loss": 0.0078,
+      "loss": 0.0088,
       "step": 2117
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.5359335716861586e-06,
-      "loss": 0.0058,
+      "loss": 0.0065,
       "step": 2118
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.522451785313331e-06,
-      "loss": 0.0066,
+      "loss": 0.0076,
       "step": 2119
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.5089780321123136e-06,
-      "loss": 0.0054,
+      "loss": 0.0055,
       "step": 2120
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.49551232655804e-06,
-      "loss": 0.0089,
+      "loss": 0.0082,
       "step": 2121
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.482054683116795e-06,
-      "loss": 0.0071,
+      "loss": 0.009,
       "step": 2122
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.468605116246191e-06,
-      "loss": 0.0058,
+      "loss": 0.007,
       "step": 2123
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.4551636403951845e-06,
-      "loss": 0.0037,
+      "loss": 0.0038,
       "step": 2124
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.441730270004016e-06,
-      "loss": 0.0063,
+      "loss": 0.0066,
       "step": 2125
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.428305019504237e-06,
-      "loss": 0.0098,
+      "loss": 0.0114,
       "step": 2126
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.414887903318674e-06,
-      "loss": 0.005,
+      "loss": 0.0037,
       "step": 2127
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.401478935861399e-06,
-      "loss": 0.0038,
+      "loss": 0.0033,
       "step": 2128
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.388078131537751e-06,
-      "loss": 0.0031,
+      "loss": 0.0033,
       "step": 2129
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.374685504744278e-06,
-      "loss": 0.006,
+      "loss": 0.0059,
       "step": 2130
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.361301069868759e-06,
-      "loss": 0.0074,
+      "loss": 0.0083,
       "step": 2131
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.347924841290174e-06,
-      "loss": 0.0047,
+      "loss": 0.0058,
       "step": 2132
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.33455683337867e-06,
-      "loss": 0.011,
+      "loss": 0.0115,
       "step": 2133
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.321197060495575e-06,
-      "loss": 0.0047,
+      "loss": 0.0052,
       "step": 2134
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.3078455369933746e-06,
-      "loss": 0.008,
+      "loss": 0.0048,
       "step": 2135
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.294502277215674e-06,
-      "loss": 0.0021,
+      "loss": 0.0024,
       "step": 2136
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.281167295497223e-06,
-      "loss": 0.0077,
+      "loss": 0.0132,
       "step": 2137
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.2678406061638545e-06,
-      "loss": 0.0149,
+      "loss": 0.0113,
       "step": 2138
     },
     {
@@ -12869,115 +12869,115 @@
     {
       "epoch": 0.68,
       "learning_rate": 7.241212161911217e-06,
-      "loss": 0.0052,
+      "loss": 0.0063,
       "step": 2140
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.2279104355990265e-06,
-      "loss": 0.0046,
+      "loss": 0.0048,
       "step": 2141
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.214617058886073e-06,
-      "loss": 0.005,
+      "loss": 0.0039,
       "step": 2142
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.201332046053492e-06,
-      "loss": 0.0053,
+      "loss": 0.0065,
       "step": 2143
     },
     {
       "epoch": 0.68,
       "learning_rate": 7.188055411373463e-06,
-      "loss": 0.0046,
+      "loss": 0.0048,
       "step": 2144
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.174787169109135e-06,
-      "loss": 0.0076,
+      "loss": 0.0065,
       "step": 2145
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.161527333514664e-06,
-      "loss": 0.0115,
+      "loss": 0.0099,
       "step": 2146
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.148275918835162e-06,
-      "loss": 0.0059,
+      "loss": 0.0065,
       "step": 2147
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.1350329393066925e-06,
-      "loss": 0.004,
+      "loss": 0.0042,
       "step": 2148
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.121798409156267e-06,
-      "loss": 0.0025,
+      "loss": 0.0027,
       "step": 2149
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.108572342601821e-06,
-      "loss": 0.0078,
+      "loss": 0.0051,
       "step": 2150
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.095354753852181e-06,
-      "loss": 0.0057,
+      "loss": 0.0054,
       "step": 2151
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.082145657107082e-06,
-      "loss": 0.0064,
+      "loss": 0.0065,
       "step": 2152
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.068945066557135e-06,
-      "loss": 0.0065,
+      "loss": 0.0044,
       "step": 2153
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.055752996383797e-06,
-      "loss": 0.0029,
+      "loss": 0.0026,
       "step": 2154
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.0425694607593946e-06,
-      "loss": 0.0126,
+      "loss": 0.0106,
       "step": 2155
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.029394473847063e-06,
-      "loss": 0.0059,
+      "loss": 0.0107,
       "step": 2156
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.016228049800768e-06,
-      "loss": 0.003,
+      "loss": 0.0033,
       "step": 2157
     },
     {
       "epoch": 0.69,
       "learning_rate": 7.003070202765279e-06,
-      "loss": 0.011,
+      "loss": 0.0122,
       "step": 2158
     },
     {
@@ -12989,67 +12989,67 @@
     {
       "epoch": 0.69,
       "learning_rate": 6.976780296259665e-06,
-      "loss": 0.0082,
+      "loss": 0.0095,
       "step": 2160
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.9636482650329316e-06,
-      "loss": 0.0082,
+      "loss": 0.0076,
       "step": 2161
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.950524867303756e-06,
-      "loss": 0.0081,
+      "loss": 0.0078,
       "step": 2162
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.937410117170683e-06,
-      "loss": 0.006,
+      "loss": 0.0066,
       "step": 2163
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.924304028722949e-06,
-      "loss": 0.0062,
+      "loss": 0.0055,
       "step": 2164
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.911206616040505e-06,
-      "loss": 0.0132,
+      "loss": 0.0142,
       "step": 2165
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.89811789319398e-06,
-      "loss": 0.0049,
+      "loss": 0.0063,
       "step": 2166
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.885037874244654e-06,
-      "loss": 0.0144,
+      "loss": 0.0148,
       "step": 2167
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.871966573244469e-06,
-      "loss": 0.0032,
+      "loss": 0.0034,
       "step": 2168
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.858904004235991e-06,
-      "loss": 0.0058,
+      "loss": 0.0065,
       "step": 2169
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.84585018125242e-06,
-      "loss": 0.0028,
+      "loss": 0.0026,
       "step": 2170
     },
     {
@@ -13061,169 +13061,169 @@
     {
       "epoch": 0.69,
       "learning_rate": 6.819768829445746e-06,
-      "loss": 0.0053,
+      "loss": 0.0069,
       "step": 2172
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.806741328641985e-06,
-      "loss": 0.0078,
+      "loss": 0.0049,
       "step": 2173
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.793722629901785e-06,
-      "loss": 0.0063,
+      "loss": 0.0064,
       "step": 2174
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.780712747211196e-06,
-      "loss": 0.0049,
+      "loss": 0.0044,
       "step": 2175
     },
     {
       "epoch": 0.69,
       "learning_rate": 6.767711694546816e-06,
-      "loss": 0.0114,
+      "loss": 0.0125,
       "step": 2176
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.754719485875743e-06,
-      "loss": 0.0067,
+      "loss": 0.0068,
       "step": 2177
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.741736135155584e-06,
-      "loss": 0.0064,
+      "loss": 0.0069,
       "step": 2178
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.728761656334419e-06,
-      "loss": 0.0047,
+      "loss": 0.0048,
       "step": 2179
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.715796063350809e-06,
-      "loss": 0.0044,
+      "loss": 0.0034,
       "step": 2180
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.702839370133763e-06,
-      "loss": 0.0049,
+      "loss": 0.0051,
       "step": 2181
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.689891590602722e-06,
-      "loss": 0.0084,
+      "loss": 0.0089,
       "step": 2182
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.676952738667564e-06,
-      "loss": 0.0059,
+      "loss": 0.0107,
       "step": 2183
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.664022828228573e-06,
-      "loss": 0.0074,
+      "loss": 0.0081,
       "step": 2184
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.651101873176413e-06,
-      "loss": 0.01,
+      "loss": 0.0076,
       "step": 2185
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.638189887392152e-06,
-      "loss": 0.0068,
+      "loss": 0.0059,
       "step": 2186
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.625286884747197e-06,
-      "loss": 0.0079,
+      "loss": 0.0068,
       "step": 2187
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.6123928791033165e-06,
-      "loss": 0.0078,
+      "loss": 0.0077,
       "step": 2188
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.599507884312624e-06,
-      "loss": 0.0082,
+      "loss": 0.0147,
       "step": 2189
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.586631914217527e-06,
-      "loss": 0.0087,
+      "loss": 0.0104,
       "step": 2190
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.573764982650766e-06,
-      "loss": 0.0064,
+      "loss": 0.0076,
       "step": 2191
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.560907103435343e-06,
-      "loss": 0.0093,
+      "loss": 0.0057,
       "step": 2192
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.548058290384556e-06,
-      "loss": 0.0038,
+      "loss": 0.004,
       "step": 2193
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.535218557301965e-06,
-      "loss": 0.0058,
+      "loss": 0.0065,
       "step": 2194
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.522387917981352e-06,
-      "loss": 0.0069,
+      "loss": 0.0051,
       "step": 2195
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.509566386206752e-06,
-      "loss": 0.0056,
+      "loss": 0.005,
       "step": 2196
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.496753975752415e-06,
-      "loss": 0.0034,
+      "loss": 0.0054,
       "step": 2197
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.48395070038277e-06,
-      "loss": 0.0057,
+      "loss": 0.0064,
       "step": 2198
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.471156573852462e-06,
-      "loss": 0.0086,
+      "loss": 0.0075,
       "step": 2199
     },
     {
@@ -13235,133 +13235,133 @@
     {
       "epoch": 0.7,
       "learning_rate": 6.445595822279195e-06,
-      "loss": 0.0111,
+      "loss": 0.013,
       "step": 2201
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.4328292246963e-06,
-      "loss": 0.0039,
+      "loss": 0.0054,
       "step": 2202
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.420071830872816e-06,
-      "loss": 0.0112,
+      "loss": 0.0097,
       "step": 2203
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.407323654514093e-06,
-      "loss": 0.0058,
+      "loss": 0.0061,
       "step": 2204
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.394584709315559e-06,
-      "loss": 0.0069,
+      "loss": 0.0053,
       "step": 2205
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.38185500896274e-06,
-      "loss": 0.0042,
+      "loss": 0.0043,
       "step": 2206
     },
     {
       "epoch": 0.7,
       "learning_rate": 6.369134567131226e-06,
-      "loss": 0.0045,
+      "loss": 0.0062,
       "step": 2207
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.356423397486652e-06,
-      "loss": 0.0061,
+      "loss": 0.0058,
       "step": 2208
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.343721513684701e-06,
-      "loss": 0.0056,
+      "loss": 0.0059,
       "step": 2209
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.331028929371086e-06,
-      "loss": 0.0098,
+      "loss": 0.0106,
       "step": 2210
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.318345658181506e-06,
-      "loss": 0.0107,
+      "loss": 0.0079,
       "step": 2211
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.305671713741686e-06,
-      "loss": 0.0036,
+      "loss": 0.0042,
       "step": 2212
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.293007109667301e-06,
-      "loss": 0.0109,
+      "loss": 0.0091,
       "step": 2213
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.280351859564013e-06,
-      "loss": 0.0054,
+      "loss": 0.0058,
       "step": 2214
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.267705977027432e-06,
-      "loss": 0.0051,
+      "loss": 0.0053,
       "step": 2215
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.255069475643089e-06,
-      "loss": 0.0088,
+      "loss": 0.0071,
       "step": 2216
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.242442368986462e-06,
-      "loss": 0.0064,
+      "loss": 0.0085,
       "step": 2217
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.229824670622908e-06,
-      "loss": 0.0042,
+      "loss": 0.0038,
       "step": 2218
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.217216394107697e-06,
-      "loss": 0.006,
+      "loss": 0.0053,
       "step": 2219
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.2046175529859785e-06,
-      "loss": 0.0121,
+      "loss": 0.0098,
       "step": 2220
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.192028160792746e-06,
-      "loss": 0.0037,
+      "loss": 0.0042,
       "step": 2221
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.1794482310528614e-06,
-      "loss": 0.0045,
+      "loss": 0.0047,
       "step": 2222
     },
     {
@@ -13373,55 +13373,55 @@
     {
       "epoch": 0.71,
       "learning_rate": 6.154316812981694e-06,
-      "loss": 0.0069,
+      "loss": 0.0039,
       "step": 2224
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.141765351649245e-06,
-      "loss": 0.0058,
+      "loss": 0.0085,
       "step": 2225
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.12922340676775e-06,
-      "loss": 0.0098,
+      "loss": 0.0092,
       "step": 2226
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.1166909918111055e-06,
-      "loss": 0.0055,
+      "loss": 0.0059,
       "step": 2227
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.104168120242943e-06,
-      "loss": 0.0151,
+      "loss": 0.014,
       "step": 2228
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.091654805516661e-06,
-      "loss": 0.0117,
+      "loss": 0.0124,
       "step": 2229
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.079151061075382e-06,
-      "loss": 0.0038,
+      "loss": 0.0132,
       "step": 2230
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.06665690035194e-06,
-      "loss": 0.0017,
+      "loss": 0.0021,
       "step": 2231
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.05417233676889e-06,
-      "loss": 0.0036,
+      "loss": 0.0038,
       "step": 2232
     },
     {
@@ -13433,67 +13433,67 @@
     {
       "epoch": 0.71,
       "learning_rate": 6.0292320546625875e-06,
-      "loss": 0.0057,
+      "loss": 0.005,
       "step": 2234
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.016776362932819e-06,
-      "loss": 0.0048,
+      "loss": 0.0042,
       "step": 2235
     },
     {
       "epoch": 0.71,
       "learning_rate": 6.0043303219303755e-06,
-      "loss": 0.0082,
+      "loss": 0.0065,
       "step": 2236
     },
     {
       "epoch": 0.71,
       "learning_rate": 5.9918939450261154e-06,
-      "loss": 0.004,
+      "loss": 0.0045,
       "step": 2237
     },
     {
       "epoch": 0.71,
       "learning_rate": 5.9794672455805095e-06,
-      "loss": 0.0079,
+      "loss": 0.0074,
       "step": 2238
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.967050236943624e-06,
-      "loss": 0.0065,
+      "loss": 0.0059,
       "step": 2239
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.954642932455122e-06,
-      "loss": 0.0045,
+      "loss": 0.0049,
       "step": 2240
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.942245345444248e-06,
-      "loss": 0.0064,
+      "loss": 0.0078,
       "step": 2241
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.929857489229788e-06,
-      "loss": 0.0064,
+      "loss": 0.0056,
       "step": 2242
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.9174793771200954e-06,
-      "loss": 0.0045,
+      "loss": 0.0039,
       "step": 2243
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.9051110224130336e-06,
-      "loss": 0.0071,
+      "loss": 0.0051,
       "step": 2244
     },
     {
@@ -13505,61 +13505,61 @@
     {
       "epoch": 0.72,
       "learning_rate": 5.880403638345901e-06,
-      "loss": 0.0089,
+      "loss": 0.0072,
       "step": 2246
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.868064635529105e-06,
-      "loss": 0.0096,
+      "loss": 0.0101,
       "step": 2247
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.855735443201485e-06,
-      "loss": 0.0077,
+      "loss": 0.0092,
       "step": 2248
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.843416074608348e-06,
-      "loss": 0.01,
+      "loss": 0.0105,
       "step": 2249
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.831106542984469e-06,
-      "loss": 0.0116,
+      "loss": 0.0107,
       "step": 2250
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.8188068615540505e-06,
-      "loss": 0.0039,
+      "loss": 0.0048,
       "step": 2251
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.806517043530697e-06,
-      "loss": 0.0065,
+      "loss": 0.005,
       "step": 2252
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.794237102117442e-06,
-      "loss": 0.0062,
+      "loss": 0.007,
       "step": 2253
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.781967050506683e-06,
-      "loss": 0.0057,
+      "loss": 0.0077,
       "step": 2254
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.769706901880214e-06,
-      "loss": 0.0074,
+      "loss": 0.008,
       "step": 2255
     },
     {
@@ -13571,295 +13571,295 @@
     {
       "epoch": 0.72,
       "learning_rate": 5.745216366254065e-06,
-      "loss": 0.0072,
+      "loss": 0.0068,
       "step": 2257
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.732986005564702e-06,
-      "loss": 0.0078,
+      "loss": 0.0086,
       "step": 2258
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.720765600480241e-06,
-      "loss": 0.0048,
+      "loss": 0.0042,
       "step": 2259
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.70855516412912e-06,
-      "loss": 0.0073,
+      "loss": 0.0055,
       "step": 2260
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.6963547096290915e-06,
-      "loss": 0.0072,
+      "loss": 0.0073,
       "step": 2261
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.68416425008716e-06,
-      "loss": 0.0034,
+      "loss": 0.0032,
       "step": 2262
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.671983798599609e-06,
-      "loss": 0.0077,
+      "loss": 0.0083,
       "step": 2263
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.65981336825197e-06,
-      "loss": 0.0066,
+      "loss": 0.0059,
       "step": 2264
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.647652972118998e-06,
-      "loss": 0.0047,
+      "loss": 0.0049,
       "step": 2265
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.635502623264681e-06,
-      "loss": 0.0071,
+      "loss": 0.0091,
       "step": 2266
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.623362334742198e-06,
-      "loss": 0.0069,
+      "loss": 0.0072,
       "step": 2267
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.611232119593934e-06,
-      "loss": 0.0049,
+      "loss": 0.0046,
       "step": 2268
     },
     {
       "epoch": 0.72,
       "learning_rate": 5.59911199085145e-06,
-      "loss": 0.01,
+      "loss": 0.0087,
       "step": 2269
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.58700196153546e-06,
-      "loss": 0.0069,
+      "loss": 0.0061,
       "step": 2270
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.5749020446558415e-06,
-      "loss": 0.0073,
+      "loss": 0.0064,
       "step": 2271
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.562812253211606e-06,
-      "loss": 0.0055,
+      "loss": 0.0065,
       "step": 2272
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.5507326001908735e-06,
-      "loss": 0.0053,
+      "loss": 0.0056,
       "step": 2273
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.538663098570895e-06,
-      "loss": 0.0041,
+      "loss": 0.003,
       "step": 2274
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.526603761317994e-06,
-      "loss": 0.0085,
+      "loss": 0.0098,
       "step": 2275
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.514554601387585e-06,
-      "loss": 0.0043,
+      "loss": 0.0027,
       "step": 2276
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.5025156317241555e-06,
-      "loss": 0.0106,
+      "loss": 0.0093,
       "step": 2277
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.490486865261229e-06,
-      "loss": 0.0095,
+      "loss": 0.0098,
       "step": 2278
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.478468314921384e-06,
-      "loss": 0.0074,
+      "loss": 0.0089,
       "step": 2279
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.4664599936162064e-06,
-      "loss": 0.0106,
+      "loss": 0.011,
       "step": 2280
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.454461914246316e-06,
-      "loss": 0.0115,
+      "loss": 0.0107,
       "step": 2281
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.442474089701302e-06,
-      "loss": 0.0076,
+      "loss": 0.0081,
       "step": 2282
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.430496532859765e-06,
-      "loss": 0.0068,
+      "loss": 0.0069,
       "step": 2283
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.418529256589249e-06,
-      "loss": 0.0032,
+      "loss": 0.0043,
       "step": 2284
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.406572273746277e-06,
-      "loss": 0.0114,
+      "loss": 0.0101,
       "step": 2285
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.394625597176292e-06,
-      "loss": 0.0085,
+      "loss": 0.0093,
       "step": 2286
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.382689239713684e-06,
-      "loss": 0.0068,
+      "loss": 0.0044,
       "step": 2287
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.37076321418174e-06,
-      "loss": 0.0081,
+      "loss": 0.0086,
       "step": 2288
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.358847533392661e-06,
-      "loss": 0.008,
+      "loss": 0.0083,
       "step": 2289
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.346942210147535e-06,
-      "loss": 0.0037,
+      "loss": 0.0039,
       "step": 2290
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.3350472572363065e-06,
-      "loss": 0.0039,
+      "loss": 0.0046,
       "step": 2291
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.323162687437798e-06,
-      "loss": 0.0101,
+      "loss": 0.01,
       "step": 2292
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.311288513519661e-06,
-      "loss": 0.0085,
+      "loss": 0.0065,
       "step": 2293
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.299424748238392e-06,
-      "loss": 0.0072,
+      "loss": 0.0126,
       "step": 2294
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.287571404339307e-06,
-      "loss": 0.0033,
+      "loss": 0.0037,
       "step": 2295
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.275728494556506e-06,
-      "loss": 0.0075,
+      "loss": 0.0089,
       "step": 2296
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.263896031612904e-06,
-      "loss": 0.0045,
+      "loss": 0.005,
       "step": 2297
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.252074028220174e-06,
-      "loss": 0.0041,
+      "loss": 0.0051,
       "step": 2298
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.2402624970787625e-06,
-      "loss": 0.006,
+      "loss": 0.0059,
       "step": 2299
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.2284614508778714e-06,
-      "loss": 0.0058,
+      "loss": 0.0053,
       "step": 2300
     },
     {
       "epoch": 0.73,
       "learning_rate": 5.216670902295419e-06,
-      "loss": 0.0058,
+      "loss": 0.0061,
       "step": 2301
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.204890863998063e-06,
-      "loss": 0.0048,
+      "loss": 0.0051,
       "step": 2302
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.193121348641167e-06,
-      "loss": 0.0081,
+      "loss": 0.0083,
       "step": 2303
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.1813623688687804e-06,
-      "loss": 0.0077,
+      "loss": 0.0072,
       "step": 2304
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.169613937313649e-06,
-      "loss": 0.0083,
+      "loss": 0.0065,
       "step": 2305
     },
     {
@@ -13871,169 +13871,169 @@
     {
       "epoch": 0.74,
       "learning_rate": 5.146148769329405e-06,
-      "loss": 0.0048,
+      "loss": 0.0047,
       "step": 2307
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.134432058109064e-06,
-      "loss": 0.006,
+      "loss": 0.0045,
       "step": 2308
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.122725945523465e-06,
-      "loss": 0.0096,
+      "loss": 0.0108,
       "step": 2309
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.111030444148556e-06,
-      "loss": 0.0083,
+      "loss": 0.0073,
       "step": 2310
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.099345566548874e-06,
-      "loss": 0.0082,
+      "loss": 0.0088,
       "step": 2311
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.087671325277556e-06,
-      "loss": 0.0128,
+      "loss": 0.0088,
       "step": 2312
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.0760077328763e-06,
-      "loss": 0.0072,
+      "loss": 0.0067,
       "step": 2313
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.064354801875364e-06,
-      "loss": 0.0049,
+      "loss": 0.0052,
       "step": 2314
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.05271254479356e-06,
-      "loss": 0.0053,
+      "loss": 0.0051,
       "step": 2315
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.041080974138237e-06,
-      "loss": 0.0144,
+      "loss": 0.0138,
       "step": 2316
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.029460102405248e-06,
-      "loss": 0.0065,
+      "loss": 0.0067,
       "step": 2317
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.017849942078964e-06,
-      "loss": 0.0076,
+      "loss": 0.0068,
       "step": 2318
     },
     {
       "epoch": 0.74,
       "learning_rate": 5.0062505056322425e-06,
-      "loss": 0.0147,
+      "loss": 0.0146,
       "step": 2319
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.994661805526423e-06,
-      "loss": 0.003,
+      "loss": 0.0036,
       "step": 2320
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.983083854211317e-06,
-      "loss": 0.0089,
+      "loss": 0.0079,
       "step": 2321
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.971516664125174e-06,
-      "loss": 0.0058,
+      "loss": 0.0051,
       "step": 2322
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.959960247694698e-06,
-      "loss": 0.0064,
+      "loss": 0.0066,
       "step": 2323
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.9484146173350014e-06,
-      "loss": 0.008,
+      "loss": 0.0077,
       "step": 2324
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.936879785449629e-06,
-      "loss": 0.0054,
+      "loss": 0.0053,
       "step": 2325
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.925355764430516e-06,
-      "loss": 0.0052,
+      "loss": 0.0061,
       "step": 2326
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.913842566657974e-06,
-      "loss": 0.0053,
+      "loss": 0.0049,
       "step": 2327
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.902340204500705e-06,
-      "loss": 0.0082,
+      "loss": 0.0079,
       "step": 2328
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.890848690315752e-06,
-      "loss": 0.0124,
+      "loss": 0.0091,
       "step": 2329
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.879368036448517e-06,
-      "loss": 0.0088,
+      "loss": 0.01,
       "step": 2330
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.8678982552327356e-06,
-      "loss": 0.0043,
+      "loss": 0.005,
       "step": 2331
     },
     {
       "epoch": 0.74,
       "learning_rate": 4.856439358990448e-06,
-      "loss": 0.0075,
+      "loss": 0.0078,
       "step": 2332
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.844991360032016e-06,
-      "loss": 0.0081,
+      "loss": 0.0084,
       "step": 2333
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.833554270656093e-06,
-      "loss": 0.0057,
+      "loss": 0.0044,
       "step": 2334
     },
     {
@@ -14045,73 +14045,73 @@
     {
       "epoch": 0.75,
       "learning_rate": 4.8107128697877385e-06,
-      "loss": 0.0091,
+      "loss": 0.0078,
       "step": 2336
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.79930858283395e-06,
-      "loss": 0.008,
+      "loss": 0.0081,
       "step": 2337
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.787915254539933e-06,
-      "loss": 0.0055,
+      "loss": 0.0056,
       "step": 2338
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.776532897145597e-06,
-      "loss": 0.0104,
+      "loss": 0.0083,
       "step": 2339
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.765161522879077e-06,
-      "loss": 0.005,
+      "loss": 0.0053,
       "step": 2340
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.753801143956697e-06,
-      "loss": 0.007,
+      "loss": 0.0076,
       "step": 2341
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.742451772582984e-06,
-      "loss": 0.0055,
+      "loss": 0.0081,
       "step": 2342
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.731113420950624e-06,
-      "loss": 0.0037,
+      "loss": 0.004,
       "step": 2343
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.719786101240477e-06,
-      "loss": 0.0041,
+      "loss": 0.0043,
       "step": 2344
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.708469825621538e-06,
-      "loss": 0.0069,
+      "loss": 0.0061,
       "step": 2345
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.697164606250952e-06,
-      "loss": 0.0061,
+      "loss": 0.0058,
       "step": 2346
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.6858704552739825e-06,
-      "loss": 0.0072,
+      "loss": 0.0058,
       "step": 2347
     },
     {
@@ -14123,231 +14123,231 @@
     {
       "epoch": 0.75,
       "learning_rate": 4.663315407022446e-06,
-      "loss": 0.0047,
+      "loss": 0.0077,
       "step": 2349
     },
     {
       "epoch": 0.75,
-      "eval_loss": 0.008719976991415024,
-      "eval_runtime": 28.1817,
-      "eval_samples_per_second": 187.108,
-      "eval_steps_per_second": 5.855,
+      "eval_loss": 0.008244536817073822,
+      "eval_runtime": 28.4604,
+      "eval_samples_per_second": 185.275,
+      "eval_steps_per_second": 11.595,
       "step": 2349
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.652054533978895e-06,
-      "loss": 0.0049,
+      "loss": 0.0054,
       "step": 2350
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.640804777790957e-06,
-      "loss": 0.0046,
+      "loss": 0.0042,
       "step": 2351
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.629566150544315e-06,
-      "loss": 0.0093,
+      "loss": 0.007,
       "step": 2352
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.618338664312678e-06,
-      "loss": 0.0054,
+      "loss": 0.007,
       "step": 2353
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.607122331157806e-06,
-      "loss": 0.0047,
+      "loss": 0.0044,
       "step": 2354
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.5959171631294605e-06,
-      "loss": 0.0084,
+      "loss": 0.008,
       "step": 2355
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.584723172265421e-06,
-      "loss": 0.0049,
+      "loss": 0.0059,
       "step": 2356
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.5735403705914594e-06,
-      "loss": 0.0071,
+      "loss": 0.0066,
       "step": 2357
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.562368770121308e-06,
-      "loss": 0.0078,
+      "loss": 0.0055,
       "step": 2358
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.551208382856692e-06,
-      "loss": 0.0083,
+      "loss": 0.0084,
       "step": 2359
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.5400592207872754e-06,
-      "loss": 0.0072,
+      "loss": 0.0069,
       "step": 2360
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.528921295890659e-06,
-      "loss": 0.0044,
+      "loss": 0.0047,
       "step": 2361
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.517794620132389e-06,
-      "loss": 0.0096,
+      "loss": 0.0081,
       "step": 2362
     },
     {
       "epoch": 0.75,
       "learning_rate": 4.506679205465903e-06,
-      "loss": 0.0067,
+      "loss": 0.0065,
       "step": 2363
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.495575063832563e-06,
-      "loss": 0.0031,
+      "loss": 0.0056,
       "step": 2364
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.4844822071616015e-06,
-      "loss": 0.0069,
+      "loss": 0.0075,
       "step": 2365
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.4734006473701455e-06,
-      "loss": 0.013,
+      "loss": 0.0179,
       "step": 2366
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.462330396363166e-06,
-      "loss": 0.008,
+      "loss": 0.0075,
       "step": 2367
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.451271466033508e-06,
-      "loss": 0.0125,
+      "loss": 0.0116,
       "step": 2368
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.440223868261831e-06,
-      "loss": 0.0095,
+      "loss": 0.0096,
       "step": 2369
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.42918761491664e-06,
-      "loss": 0.0043,
+      "loss": 0.0066,
       "step": 2370
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.418162717854234e-06,
-      "loss": 0.0043,
+      "loss": 0.0038,
       "step": 2371
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.4071491889187316e-06,
-      "loss": 0.0119,
+      "loss": 0.0109,
       "step": 2372
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.3961470399420194e-06,
-      "loss": 0.008,
+      "loss": 0.0086,
       "step": 2373
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.3851562827437716e-06,
-      "loss": 0.0058,
+      "loss": 0.0061,
       "step": 2374
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.374176929131424e-06,
-      "loss": 0.0076,
+      "loss": 0.0078,
       "step": 2375
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.36320899090015e-06,
-      "loss": 0.005,
+      "loss": 0.0052,
       "step": 2376
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.35225247983287e-06,
-      "loss": 0.0071,
+      "loss": 0.0066,
       "step": 2377
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.34130740770023e-06,
-      "loss": 0.0066,
+      "loss": 0.0071,
       "step": 2378
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.330373786260573e-06,
-      "loss": 0.007,
+      "loss": 0.0064,
       "step": 2379
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.3194516272599585e-06,
-      "loss": 0.0081,
+      "loss": 0.0069,
       "step": 2380
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.308540942432115e-06,
-      "loss": 0.0067,
+      "loss": 0.0074,
       "step": 2381
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.2976417434984565e-06,
-      "loss": 0.0039,
+      "loss": 0.0041,
       "step": 2382
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.2867540421680566e-06,
-      "loss": 0.0063,
+      "loss": 0.0064,
       "step": 2383
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.275877850137624e-06,
-      "loss": 0.0064,
+      "loss": 0.0066,
       "step": 2384
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.265013179091525e-06,
-      "loss": 0.0053,
+      "loss": 0.005,
       "step": 2385
     },
     {
@@ -14359,19 +14359,19 @@
     {
       "epoch": 0.76,
       "learning_rate": 4.243318446627813e-06,
-      "loss": 0.006,
+      "loss": 0.0051,
       "step": 2387
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.2324884085169815e-06,
-      "loss": 0.0025,
+      "loss": 0.0024,
       "step": 2388
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.221669938003995e-06,
-      "loss": 0.0039,
+      "loss": 0.0048,
       "step": 2389
     },
     {
@@ -14383,13 +14383,13 @@
     {
       "epoch": 0.76,
       "learning_rate": 4.200067746248495e-06,
-      "loss": 0.0027,
+      "loss": 0.0025,
       "step": 2391
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.189284048213336e-06,
-      "loss": 0.0048,
+      "loss": 0.0057,
       "step": 2392
     },
     {
@@ -14401,319 +14401,319 @@
     {
       "epoch": 0.76,
       "learning_rate": 4.16775150575313e-06,
-      "loss": 0.008,
+      "loss": 0.0075,
       "step": 2394
     },
     {
       "epoch": 0.76,
       "learning_rate": 4.157002684460606e-06,
-      "loss": 0.012,
+      "loss": 0.0121,
       "step": 2395
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.146265511860669e-06,
-      "loss": 0.0079,
+      "loss": 0.0081,
       "step": 2396
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.135539999488312e-06,
-      "loss": 0.0099,
+      "loss": 0.0083,
       "step": 2397
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.124826158866024e-06,
-      "loss": 0.0032,
+      "loss": 0.003,
       "step": 2398
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.114124001503733e-06,
-      "loss": 0.0036,
+      "loss": 0.0034,
       "step": 2399
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.103433538898834e-06,
-      "loss": 0.0046,
+      "loss": 0.0044,
       "step": 2400
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.092754782536151e-06,
-      "loss": 0.0071,
+      "loss": 0.0064,
       "step": 2401
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.0820877438879255e-06,
-      "loss": 0.0079,
+      "loss": 0.0082,
       "step": 2402
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.071432434413826e-06,
-      "loss": 0.0044,
+      "loss": 0.0042,
       "step": 2403
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.060788865560899e-06,
-      "loss": 0.0072,
+      "loss": 0.0074,
       "step": 2404
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.0501570487635975e-06,
-      "loss": 0.0048,
+      "loss": 0.0043,
       "step": 2405
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.039536995443745e-06,
-      "loss": 0.0114,
+      "loss": 0.0151,
       "step": 2406
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.028928717010515e-06,
-      "loss": 0.0057,
+      "loss": 0.0041,
       "step": 2407
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.018332224860446e-06,
-      "loss": 0.0045,
+      "loss": 0.0048,
       "step": 2408
     },
     {
       "epoch": 0.77,
       "learning_rate": 4.00774753037741e-06,
-      "loss": 0.0043,
+      "loss": 0.0034,
       "step": 2409
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.997174644932599e-06,
-      "loss": 0.008,
+      "loss": 0.0074,
       "step": 2410
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.986613579884528e-06,
-      "loss": 0.0067,
+      "loss": 0.01,
       "step": 2411
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.976064346579e-06,
-      "loss": 0.0024,
+      "loss": 0.0028,
       "step": 2412
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.965526956349123e-06,
-      "loss": 0.0062,
+      "loss": 0.0071,
       "step": 2413
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.955001420515273e-06,
-      "loss": 0.008,
+      "loss": 0.0071,
       "step": 2414
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.944487750385088e-06,
-      "loss": 0.0077,
+      "loss": 0.0061,
       "step": 2415
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.933985957253469e-06,
-      "loss": 0.0047,
+      "loss": 0.0059,
       "step": 2416
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.923496052402542e-06,
-      "loss": 0.0063,
+      "loss": 0.0057,
       "step": 2417
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.913018047101675e-06,
-      "loss": 0.0055,
+      "loss": 0.0052,
       "step": 2418
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.902551952607455e-06,
-      "loss": 0.0069,
+      "loss": 0.0077,
       "step": 2419
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.892097780163652e-06,
-      "loss": 0.0105,
+      "loss": 0.0097,
       "step": 2420
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.8816555410012545e-06,
-      "loss": 0.0105,
+      "loss": 0.0124,
       "step": 2421
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.871225246338408e-06,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 2422
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.860806907380444e-06,
-      "loss": 0.0071,
+      "loss": 0.0058,
       "step": 2423
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.850400535319835e-06,
-      "loss": 0.0087,
+      "loss": 0.0097,
       "step": 2424
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.840006141336215e-06,
-      "loss": 0.0062,
+      "loss": 0.008,
       "step": 2425
     },
     {
       "epoch": 0.77,
       "learning_rate": 3.8296237365963276e-06,
-      "loss": 0.0045,
+      "loss": 0.0043,
       "step": 2426
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.819253332254059e-06,
-      "loss": 0.0187,
+      "loss": 0.012,
       "step": 2427
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.808894939450382e-06,
-      "loss": 0.0046,
+      "loss": 0.0044,
       "step": 2428
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7985485693133844e-06,
-      "loss": 0.0059,
+      "loss": 0.0065,
       "step": 2429
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.788214232958224e-06,
-      "loss": 0.0082,
+      "loss": 0.0083,
       "step": 2430
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7778919414871377e-06,
-      "loss": 0.0074,
+      "loss": 0.0055,
       "step": 2431
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7675817059894257e-06,
-      "loss": 0.0081,
+      "loss": 0.0085,
       "step": 2432
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7572835375414234e-06,
-      "loss": 0.0129,
+      "loss": 0.013,
       "step": 2433
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.746997447206514e-06,
-      "loss": 0.0112,
+      "loss": 0.0094,
       "step": 2434
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.736723446035108e-06,
-      "loss": 0.0085,
+      "loss": 0.0081,
       "step": 2435
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7264615450646116e-06,
-      "loss": 0.0065,
+      "loss": 0.0051,
       "step": 2436
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7162117553194513e-06,
-      "loss": 0.0055,
+      "loss": 0.0064,
       "step": 2437
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.7059740878110255e-06,
-      "loss": 0.0083,
+      "loss": 0.0086,
       "step": 2438
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.6957485535377226e-06,
-      "loss": 0.0069,
+      "loss": 0.008,
       "step": 2439
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.6855351634848928e-06,
-      "loss": 0.0103,
+      "loss": 0.0092,
       "step": 2440
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.6753339286248334e-06,
-      "loss": 0.0076,
+      "loss": 0.007,
       "step": 2441
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.665144859916793e-06,
-      "loss": 0.0055,
+      "loss": 0.0067,
       "step": 2442
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.6549679683069393e-06,
-      "loss": 0.006,
+      "loss": 0.0052,
       "step": 2443
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.6448032647283647e-06,
-      "loss": 0.0038,
+      "loss": 0.0039,
       "step": 2444
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.634650760101076e-06,
-      "loss": 0.0029,
+      "loss": 0.0041,
       "step": 2445
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.6245104653319517e-06,
-      "loss": 0.004,
+      "loss": 0.0073,
       "step": 2446
     },
     {
@@ -14725,229 +14725,229 @@
     {
       "epoch": 0.78,
       "learning_rate": 3.6042665489301956e-06,
-      "loss": 0.0075,
+      "loss": 0.0084,
       "step": 2448
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.5941629490457173e-06,
-      "loss": 0.0055,
+      "loss": 0.007,
       "step": 2449
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.584071602515685e-06,
-      "loss": 0.0023,
+      "loss": 0.0029,
       "step": 2450
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.573992520181302e-06,
-      "loss": 0.0068,
+      "loss": 0.0069,
       "step": 2451
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.5639257128705694e-06,
-      "loss": 0.0057,
+      "loss": 0.007,
       "step": 2452
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.5538711913983284e-06,
-      "loss": 0.0109,
+      "loss": 0.0083,
       "step": 2453
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.5438289665661955e-06,
-      "loss": 0.0112,
+      "loss": 0.0115,
       "step": 2454
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.533799049162597e-06,
-      "loss": 0.0086,
+      "loss": 0.0089,
       "step": 2455
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.523781449962724e-06,
-      "loss": 0.0049,
+      "loss": 0.0051,
       "step": 2456
     },
     {
       "epoch": 0.78,
       "learning_rate": 3.5137761797285396e-06,
-      "loss": 0.0076,
+      "loss": 0.0066,
       "step": 2457
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.5037832492087686e-06,
-      "loss": 0.0065,
+      "loss": 0.0076,
       "step": 2458
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.4938026691388636e-06,
-      "loss": 0.0037,
+      "loss": 0.0033,
       "step": 2459
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.483834450241023e-06,
-      "loss": 0.0083,
+      "loss": 0.0082,
       "step": 2460
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.4738786032241553e-06,
-      "loss": 0.0079,
+      "loss": 0.0068,
       "step": 2461
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.463935138783885e-06,
-      "loss": 0.0073,
+      "loss": 0.0062,
       "step": 2462
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.4540040676025343e-06,
-      "loss": 0.0063,
+      "loss": 0.0057,
       "step": 2463
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.444085400349103e-06,
-      "loss": 0.0071,
+      "loss": 0.0051,
       "step": 2464
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.434179147679273e-06,
-      "loss": 0.0075,
+      "loss": 0.0074,
       "step": 2465
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.4242853202353922e-06,
-      "loss": 0.0046,
+      "loss": 0.005,
       "step": 2466
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.4144039286464447e-06,
-      "loss": 0.009,
+      "loss": 0.0093,
       "step": 2467
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.4045349835280744e-06,
-      "loss": 0.0055,
+      "loss": 0.0059,
       "step": 2468
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3946784954825355e-06,
-      "loss": 0.0119,
+      "loss": 0.0138,
       "step": 2469
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3848344750987115e-06,
-      "loss": 0.0053,
+      "loss": 0.0046,
       "step": 2470
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.375002932952094e-06,
-      "loss": 0.0054,
+      "loss": 0.0066,
       "step": 2471
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3651838796047553e-06,
-      "loss": 0.0102,
+      "loss": 0.0076,
       "step": 2472
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.355377325605366e-06,
-      "loss": 0.0099,
+      "loss": 0.0062,
       "step": 2473
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3455832814891548e-06,
-      "loss": 0.0108,
+      "loss": 0.0079,
       "step": 2474
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3358017577779225e-06,
-      "loss": 0.0058,
+      "loss": 0.0071,
       "step": 2475
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.326032764980018e-06,
-      "loss": 0.0077,
+      "loss": 0.0081,
       "step": 2476
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3162763135903196e-06,
-      "loss": 0.0053,
+      "loss": 0.0096,
       "step": 2477
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.3065324140902358e-06,
-      "loss": 0.0059,
+      "loss": 0.0052,
       "step": 2478
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.2968010769477e-06,
-      "loss": 0.0051,
+      "loss": 0.006,
       "step": 2479
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.2870823126171317e-06,
-      "loss": 0.007,
+      "loss": 0.0064,
       "step": 2480
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.2773761315394635e-06,
-      "loss": 0.0124,
+      "loss": 0.0132,
       "step": 2481
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.267682544142092e-06,
-      "loss": 0.0057,
+      "loss": 0.0065,
       "step": 2482
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.2580015608388956e-06,
-      "loss": 0.0042,
+      "loss": 0.0039,
       "step": 2483
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.248333192030215e-06,
-      "loss": 0.0064,
+      "loss": 0.0057,
       "step": 2484
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.2386774481028213e-06,
-      "loss": 0.009,
+      "loss": 0.0092,
       "step": 2485
     },
     {
@@ -14959,103 +14959,103 @@
     {
       "epoch": 0.79,
       "learning_rate": 3.2194038763712204e-06,
-      "loss": 0.0095,
+      "loss": 0.0079,
       "step": 2487
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.209786069272715e-06,
-      "loss": 0.0048,
+      "loss": 0.0041,
       "step": 2488
     },
     {
       "epoch": 0.79,
       "learning_rate": 3.2001809284668957e-06,
-      "loss": 0.0043,
+      "loss": 0.0075,
       "step": 2489
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.190588464272611e-06,
-      "loss": 0.0037,
+      "loss": 0.0032,
       "step": 2490
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.1810086869951062e-06,
-      "loss": 0.0047,
+      "loss": 0.0054,
       "step": 2491
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.1714416069259813e-06,
-      "loss": 0.005,
+      "loss": 0.0051,
       "step": 2492
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.1618872343432074e-06,
-      "loss": 0.0034,
+      "loss": 0.0035,
       "step": 2493
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.1523455795111066e-06,
-      "loss": 0.0029,
+      "loss": 0.003,
       "step": 2494
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.142816652680322e-06,
-      "loss": 0.0049,
+      "loss": 0.0057,
       "step": 2495
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.1333004640878383e-06,
-      "loss": 0.0089,
+      "loss": 0.0083,
       "step": 2496
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.123797023956954e-06,
-      "loss": 0.0118,
+      "loss": 0.0111,
       "step": 2497
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.114306342497258e-06,
-      "loss": 0.0057,
+      "loss": 0.0063,
       "step": 2498
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.104828429904652e-06,
-      "loss": 0.0062,
+      "loss": 0.0059,
       "step": 2499
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.0953632963613014e-06,
-      "loss": 0.0079,
+      "loss": 0.0067,
       "step": 2500
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.085910952035658e-06,
-      "loss": 0.0048,
+      "loss": 0.0032,
       "step": 2501
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.0764714070824305e-06,
-      "loss": 0.0065,
+      "loss": 0.0071,
       "step": 2502
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.067044671642567e-06,
-      "loss": 0.006,
+      "loss": 0.0067,
       "step": 2503
     },
     {
@@ -15067,163 +15067,163 @@
     {
       "epoch": 0.8,
       "learning_rate": 3.048229669797954e-06,
-      "loss": 0.0115,
+      "loss": 0.0091,
       "step": 2505
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.0388414236062656e-06,
-      "loss": 0.0028,
+      "loss": 0.0031,
       "step": 2506
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.029466027354043e-06,
-      "loss": 0.006,
+      "loss": 0.0074,
       "step": 2507
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.0201034911133344e-06,
-      "loss": 0.0073,
+      "loss": 0.0078,
       "step": 2508
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.010753824942358e-06,
-      "loss": 0.0092,
+      "loss": 0.011,
       "step": 2509
     },
     {
       "epoch": 0.8,
       "learning_rate": 3.0014170388855195e-06,
-      "loss": 0.0055,
+      "loss": 0.0068,
       "step": 2510
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.9920931429733713e-06,
-      "loss": 0.0065,
+      "loss": 0.0059,
       "step": 2511
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.982782147222635e-06,
-      "loss": 0.0069,
+      "loss": 0.0055,
       "step": 2512
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.9734840616361587e-06,
-      "loss": 0.0081,
+      "loss": 0.0075,
       "step": 2513
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.964198896202931e-06,
-      "loss": 0.008,
+      "loss": 0.0075,
       "step": 2514
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.9549266608980595e-06,
-      "loss": 0.0094,
+      "loss": 0.0086,
       "step": 2515
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.945667365682753e-06,
-      "loss": 0.005,
+      "loss": 0.0044,
       "step": 2516
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.9364210205043314e-06,
-      "loss": 0.0043,
+      "loss": 0.0038,
       "step": 2517
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.927187635296186e-06,
-      "loss": 0.0061,
+      "loss": 0.0067,
       "step": 2518
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.9179672199777984e-06,
-      "loss": 0.0047,
+      "loss": 0.0046,
       "step": 2519
     },
     {
       "epoch": 0.8,
       "learning_rate": 2.9087597844547176e-06,
-      "loss": 0.0086,
+      "loss": 0.0081,
       "step": 2520
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.8995653386185327e-06,
-      "loss": 0.0059,
+      "loss": 0.0062,
       "step": 2521
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.890383892346898e-06,
-      "loss": 0.0048,
+      "loss": 0.0052,
       "step": 2522
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.8812154555034843e-06,
-      "loss": 0.0053,
+      "loss": 0.0059,
       "step": 2523
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.8720600379379964e-06,
-      "loss": 0.0081,
+      "loss": 0.0065,
       "step": 2524
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.862917649486158e-06,
-      "loss": 0.0097,
+      "loss": 0.0069,
       "step": 2525
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.8537882999696762e-06,
-      "loss": 0.009,
+      "loss": 0.0101,
       "step": 2526
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.8446719991962705e-06,
-      "loss": 0.0094,
+      "loss": 0.0116,
       "step": 2527
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.835568756959635e-06,
-      "loss": 0.0076,
+      "loss": 0.0062,
       "step": 2528
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.826478583039425e-06,
-      "loss": 0.0042,
+      "loss": 0.0036,
       "step": 2529
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.817401487201276e-06,
-      "loss": 0.0077,
+      "loss": 0.0079,
       "step": 2530
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.808337479196752e-06,
-      "loss": 0.0053,
+      "loss": 0.0061,
       "step": 2531
     },
     {
@@ -15235,19 +15235,19 @@
     {
       "epoch": 0.81,
       "learning_rate": 2.7902487656245874e-06,
-      "loss": 0.004,
+      "loss": 0.0045,
       "step": 2533
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.78122407948975e-06,
-      "loss": 0.0066,
+      "loss": 0.0053,
       "step": 2534
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.772212520054131e-06,
-      "loss": 0.0098,
+      "loss": 0.0114,
       "step": 2535
     },
     {
@@ -15259,289 +15259,289 @@
     {
       "epoch": 0.81,
       "learning_rate": 2.7542288199911233e-06,
-      "loss": 0.0048,
+      "loss": 0.0045,
       "step": 2537
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.7452566986837233e-06,
-      "loss": 0.0053,
+      "loss": 0.0045,
       "step": 2538
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.7362977427154993e-06,
-      "loss": 0.0065,
+      "loss": 0.0073,
       "step": 2539
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.727351961711112e-06,
-      "loss": 0.008,
+      "loss": 0.0074,
       "step": 2540
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.7184193652810654e-06,
-      "loss": 0.003,
+      "loss": 0.0039,
       "step": 2541
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.709499963021691e-06,
-      "loss": 0.0084,
+      "loss": 0.008,
       "step": 2542
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.7005937645151628e-06,
-      "loss": 0.0087,
+      "loss": 0.0083,
       "step": 2543
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.6917007793294475e-06,
-      "loss": 0.0049,
+      "loss": 0.0046,
       "step": 2544
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.682821017018337e-06,
-      "loss": 0.0039,
+      "loss": 0.0038,
       "step": 2545
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.6739544871214094e-06,
-      "loss": 0.0025,
+      "loss": 0.0027,
       "step": 2546
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.665101199164023e-06,
-      "loss": 0.0088,
+      "loss": 0.0079,
       "step": 2547
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.6562611626573207e-06,
-      "loss": 0.0038,
+      "loss": 0.0049,
       "step": 2548
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.647434387098195e-06,
-      "loss": 0.0044,
+      "loss": 0.0045,
       "step": 2549
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.6386208819693037e-06,
-      "loss": 0.0055,
+      "loss": 0.0052,
       "step": 2550
     },
     {
       "epoch": 0.81,
       "learning_rate": 2.62982065673905e-06,
-      "loss": 0.005,
+      "loss": 0.0071,
       "step": 2551
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.621033720861556e-06,
-      "loss": 0.01,
+      "loss": 0.0087,
       "step": 2552
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.612260083776684e-06,
-      "loss": 0.0053,
+      "loss": 0.0046,
       "step": 2553
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.603499754909995e-06,
-      "loss": 0.0055,
+      "loss": 0.0051,
       "step": 2554
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.5947527436727596e-06,
-      "loss": 0.0047,
+      "loss": 0.0063,
       "step": 2555
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.5860190594619486e-06,
-      "loss": 0.0148,
+      "loss": 0.015,
       "step": 2556
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.577298711660198e-06,
-      "loss": 0.0069,
+      "loss": 0.0062,
       "step": 2557
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.5685917096358326e-06,
-      "loss": 0.0029,
+      "loss": 0.003,
       "step": 2558
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.5598980627428324e-06,
-      "loss": 0.0035,
+      "loss": 0.0029,
       "step": 2559
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.5512177803208293e-06,
-      "loss": 0.0076,
+      "loss": 0.0088,
       "step": 2560
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.542550871695096e-06,
-      "loss": 0.0024,
+      "loss": 0.003,
       "step": 2561
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.533897346176549e-06,
-      "loss": 0.0053,
+      "loss": 0.0043,
       "step": 2562
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.5252572130617113e-06,
-      "loss": 0.0161,
+      "loss": 0.0112,
       "step": 2563
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.516630481632731e-06,
-      "loss": 0.0071,
+      "loss": 0.0058,
       "step": 2564
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.508017161157349e-06,
-      "loss": 0.0047,
+      "loss": 0.0042,
       "step": 2565
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.4994172608889103e-06,
-      "loss": 0.0076,
+      "loss": 0.0066,
       "step": 2566
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.490830790066329e-06,
-      "loss": 0.0081,
+      "loss": 0.0074,
       "step": 2567
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.482257757914101e-06,
-      "loss": 0.0038,
+      "loss": 0.0041,
       "step": 2568
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.4736981736422887e-06,
-      "loss": 0.0062,
+      "loss": 0.005,
       "step": 2569
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.4651520464464914e-06,
-      "loss": 0.0034,
+      "loss": 0.0054,
       "step": 2570
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.4566193855078655e-06,
-      "loss": 0.0042,
+      "loss": 0.0043,
       "step": 2571
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.448100199993101e-06,
-      "loss": 0.0053,
+      "loss": 0.0045,
       "step": 2572
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.4395944990543994e-06,
-      "loss": 0.005,
+      "loss": 0.0049,
       "step": 2573
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.4311022918294874e-06,
-      "loss": 0.0114,
+      "loss": 0.0102,
       "step": 2574
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.422623587441587e-06,
-      "loss": 0.0039,
+      "loss": 0.0103,
       "step": 2575
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.414158394999419e-06,
-      "loss": 0.0089,
+      "loss": 0.0083,
       "step": 2576
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.40570672359719e-06,
-      "loss": 0.0056,
+      "loss": 0.0055,
       "step": 2577
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.397268582314573e-06,
-      "loss": 0.0125,
+      "loss": 0.0141,
       "step": 2578
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.3888439802167157e-06,
-      "loss": 0.0065,
+      "loss": 0.0068,
       "step": 2579
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.3804329263542066e-06,
-      "loss": 0.0103,
+      "loss": 0.0111,
       "step": 2580
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.3720354297630954e-06,
-      "loss": 0.0088,
+      "loss": 0.0048,
       "step": 2581
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.3636514994648613e-06,
-      "loss": 0.0053,
+      "loss": 0.0047,
       "step": 2582
     },
     {
       "epoch": 0.82,
       "learning_rate": 2.3552811444663996e-06,
-      "loss": 0.0081,
+      "loss": 0.0078,
       "step": 2583
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.3469243737600326e-06,
-      "loss": 0.0046,
+      "loss": 0.005,
       "step": 2584
     },
     {
@@ -15553,25 +15553,25 @@
     {
       "epoch": 0.83,
       "learning_rate": 2.3302516211198883e-06,
-      "loss": 0.0027,
+      "loss": 0.0033,
       "step": 2586
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.3219356570977428e-06,
-      "loss": 0.0062,
+      "loss": 0.006,
       "step": 2587
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.3136333131909316e-06,
-      "loss": 0.006,
+      "loss": 0.0058,
       "step": 2588
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.3053445983187177e-06,
-      "loss": 0.0097,
+      "loss": 0.0098,
       "step": 2589
     },
     {
@@ -15583,205 +15583,205 @@
     {
       "epoch": 0.83,
       "learning_rate": 2.2888080912818644e-06,
-      "loss": 0.0087,
+      "loss": 0.0089,
       "step": 2591
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.280560316882492e-06,
-      "loss": 0.0051,
+      "loss": 0.005,
       "step": 2592
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.2723262070482132e-06,
-      "loss": 0.007,
+      "loss": 0.0077,
       "step": 2593
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.2641057706249824e-06,
-      "loss": 0.0049,
+      "loss": 0.0038,
       "step": 2594
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.2558990164440656e-06,
-      "loss": 0.0064,
+      "loss": 0.0098,
       "step": 2595
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.247705953322014e-06,
-      "loss": 0.0064,
+      "loss": 0.0063,
       "step": 2596
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.2395265900606916e-06,
-      "loss": 0.0056,
+      "loss": 0.008,
       "step": 2597
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.231360935447226e-06,
-      "loss": 0.0069,
+      "loss": 0.0071,
       "step": 2598
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.223208998254028e-06,
-      "loss": 0.0096,
+      "loss": 0.0091,
       "step": 2599
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.2150707872387733e-06,
-      "loss": 0.0038,
+      "loss": 0.0039,
       "step": 2600
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.2069463111443805e-06,
-      "loss": 0.0074,
+      "loss": 0.0044,
       "step": 2601
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1988355786990234e-06,
-      "loss": 0.0062,
+      "loss": 0.0063,
       "step": 2602
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.19073859861611e-06,
-      "loss": 0.0057,
+      "loss": 0.0048,
       "step": 2603
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1826553795942655e-06,
-      "loss": 0.0092,
+      "loss": 0.0121,
       "step": 2604
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1745859303173438e-06,
-      "loss": 0.0041,
+      "loss": 0.0048,
       "step": 2605
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1665302594543918e-06,
-      "loss": 0.0075,
+      "loss": 0.0067,
       "step": 2606
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.158488375659665e-06,
-      "loss": 0.0091,
+      "loss": 0.0082,
       "step": 2607
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1504602875726083e-06,
-      "loss": 0.0162,
+      "loss": 0.0114,
       "step": 2608
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1424460038178333e-06,
-      "loss": 0.0049,
+      "loss": 0.0067,
       "step": 2609
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1344455330051387e-06,
-      "loss": 0.0124,
+      "loss": 0.0086,
       "step": 2610
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1264588837294672e-06,
-      "loss": 0.0029,
+      "loss": 0.0024,
       "step": 2611
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.1184860645709243e-06,
-      "loss": 0.0084,
+      "loss": 0.006,
       "step": 2612
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.110527084094756e-06,
-      "loss": 0.0095,
+      "loss": 0.0084,
       "step": 2613
     },
     {
       "epoch": 0.83,
       "learning_rate": 2.102581950851335e-06,
-      "loss": 0.0079,
+      "loss": 0.0081,
       "step": 2614
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.0946506733761683e-06,
-      "loss": 0.0074,
+      "loss": 0.0078,
       "step": 2615
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.0867332601898635e-06,
-      "loss": 0.0095,
+      "loss": 0.0105,
       "step": 2616
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.078829719798152e-06,
-      "loss": 0.0091,
+      "loss": 0.0064,
       "step": 2617
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.07094006069184e-06,
-      "loss": 0.0072,
+      "loss": 0.007,
       "step": 2618
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.063064291346844e-06,
-      "loss": 0.004,
+      "loss": 0.0041,
       "step": 2619
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.055202420224136e-06,
-      "loss": 0.0043,
+      "loss": 0.0046,
       "step": 2620
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.0473544557697765e-06,
-      "loss": 0.0079,
+      "loss": 0.0064,
       "step": 2621
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.0395204064148698e-06,
-      "loss": 0.0087,
+      "loss": 0.0082,
       "step": 2622
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.0317002805755887e-06,
-      "loss": 0.004,
+      "loss": 0.0041,
       "step": 2623
     },
     {
       "epoch": 0.84,
       "learning_rate": 2.023894086653129e-06,
-      "loss": 0.0039,
+      "loss": 0.004,
       "step": 2624
     },
     {
@@ -15793,7 +15793,7 @@
     {
       "epoch": 0.84,
       "learning_rate": 2.008323528088661e-06,
-      "loss": 0.0094,
+      "loss": 0.0099,
       "step": 2626
     },
     {
@@ -15805,43 +15805,43 @@
     {
       "epoch": 0.84,
       "learning_rate": 1.992808797631604e-06,
-      "loss": 0.0044,
+      "loss": 0.0042,
       "step": 2628
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.9850723887871757e-06,
-      "loss": 0.0055,
+      "loss": 0.0059,
       "step": 2629
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.9773499619521794e-06,
-      "loss": 0.0052,
+      "loss": 0.0055,
       "step": 2630
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.969641525422862e-06,
-      "loss": 0.0113,
+      "loss": 0.0134,
       "step": 2631
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.961947087480433e-06,
-      "loss": 0.0022,
+      "loss": 0.0023,
       "step": 2632
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.9542666563910745e-06,
-      "loss": 0.0082,
+      "loss": 0.0069,
       "step": 2633
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.946600240405918e-06,
-      "loss": 0.0104,
+      "loss": 0.0105,
       "step": 2634
     },
     {
@@ -15853,97 +15853,97 @@
     {
       "epoch": 0.84,
       "learning_rate": 1.9313094866774327e-06,
-      "loss": 0.0121,
+      "loss": 0.0124,
       "step": 2636
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.9236851653610454e-06,
-      "loss": 0.0025,
+      "loss": 0.0027,
       "step": 2637
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.916074892002725e-06,
-      "loss": 0.0076,
+      "loss": 0.0071,
       "step": 2638
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.908478674778233e-06,
-      "loss": 0.0078,
+      "loss": 0.0086,
       "step": 2639
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.9008965218482216e-06,
-      "loss": 0.0047,
+      "loss": 0.0046,
       "step": 2640
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.893328441358247e-06,
-      "loss": 0.003,
+      "loss": 0.0036,
       "step": 2641
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.885774441438735e-06,
-      "loss": 0.0034,
+      "loss": 0.0028,
       "step": 2642
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.8782345302049947e-06,
-      "loss": 0.0042,
+      "loss": 0.0167,
       "step": 2643
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.8707087157571895e-06,
-      "loss": 0.0114,
+      "loss": 0.0092,
       "step": 2644
     },
     {
       "epoch": 0.84,
       "learning_rate": 1.8631970061803504e-06,
-      "loss": 0.0044,
+      "loss": 0.0053,
       "step": 2645
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.8556994095443408e-06,
-      "loss": 0.0077,
+      "loss": 0.0073,
       "step": 2646
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.848215933903879e-06,
-      "loss": 0.0051,
+      "loss": 0.0062,
       "step": 2647
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.8407465872984975e-06,
-      "loss": 0.0082,
+      "loss": 0.0065,
       "step": 2648
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.8332913777525623e-06,
-      "loss": 0.0018,
+      "loss": 0.0023,
       "step": 2649
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.8258503132752436e-06,
-      "loss": 0.0041,
+      "loss": 0.0054,
       "step": 2650
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.8184234018605185e-06,
-      "loss": 0.008,
+      "loss": 0.0074,
       "step": 2651
     },
     {
@@ -15961,175 +15961,175 @@
     {
       "epoch": 0.85,
       "learning_rate": 1.7962276657035665e-06,
-      "loss": 0.0128,
+      "loss": 0.011,
       "step": 2654
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7888574461747687e-06,
-      "loss": 0.0075,
+      "loss": 0.0077,
       "step": 2655
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7815014194502111e-06,
-      "loss": 0.0111,
+      "loss": 0.0107,
       "step": 2656
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7741595934325138e-06,
-      "loss": 0.004,
+      "loss": 0.0033,
       "step": 2657
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.766831976009038e-06,
-      "loss": 0.0117,
+      "loss": 0.0109,
       "step": 2658
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7595185750518844e-06,
-      "loss": 0.0055,
+      "loss": 0.0069,
       "step": 2659
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7522193984178886e-06,
-      "loss": 0.0061,
+      "loss": 0.0071,
       "step": 2660
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7449344539485845e-06,
-      "loss": 0.0057,
+      "loss": 0.0068,
       "step": 2661
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7376637494702403e-06,
-      "loss": 0.0076,
+      "loss": 0.0087,
       "step": 2662
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.730407292793804e-06,
-      "loss": 0.003,
+      "loss": 0.0031,
       "step": 2663
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.723165091714934e-06,
-      "loss": 0.0141,
+      "loss": 0.0111,
       "step": 2664
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7159371540139674e-06,
-      "loss": 0.0023,
+      "loss": 0.0026,
       "step": 2665
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7087234874559149e-06,
-      "loss": 0.005,
+      "loss": 0.0043,
       "step": 2666
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.7015240997904624e-06,
-      "loss": 0.0034,
+      "loss": 0.0032,
       "step": 2667
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.694338998751948e-06,
-      "loss": 0.0072,
+      "loss": 0.0078,
       "step": 2668
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6871681920593674e-06,
-      "loss": 0.0062,
+      "loss": 0.0059,
       "step": 2669
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6800116874163613e-06,
-      "loss": 0.0042,
+      "loss": 0.005,
       "step": 2670
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6728694925111981e-06,
-      "loss": 0.0222,
+      "loss": 0.0241,
       "step": 2671
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6657416150167836e-06,
-      "loss": 0.0093,
+      "loss": 0.0087,
       "step": 2672
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6586280625906308e-06,
-      "loss": 0.0064,
+      "loss": 0.0055,
       "step": 2673
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6515288428748698e-06,
-      "loss": 0.0045,
+      "loss": 0.0043,
       "step": 2674
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6444439634962349e-06,
-      "loss": 0.0089,
+      "loss": 0.0099,
       "step": 2675
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.637373432066046e-06,
-      "loss": 0.0134,
+      "loss": 0.0079,
       "step": 2676
     },
     {
       "epoch": 0.85,
       "learning_rate": 1.6303172561802188e-06,
-      "loss": 0.0063,
+      "loss": 0.0061,
       "step": 2677
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.6232754434192442e-06,
-      "loss": 0.0104,
+      "loss": 0.0118,
       "step": 2678
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.6162480013481761e-06,
-      "loss": 0.0029,
+      "loss": 0.0028,
       "step": 2679
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.6092349375166403e-06,
-      "loss": 0.0042,
+      "loss": 0.0044,
       "step": 2680
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.6022362594588037e-06,
-      "loss": 0.0098,
+      "loss": 0.0078,
       "step": 2681
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5952519746933897e-06,
-      "loss": 0.0102,
+      "loss": 0.0109,
       "step": 2682
     },
     {
@@ -16141,79 +16141,79 @@
     {
       "epoch": 0.86,
       "learning_rate": 1.5813266150373818e-06,
-      "loss": 0.0056,
+      "loss": 0.0039,
       "step": 2684
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5743855551068809e-06,
-      "loss": 0.0117,
+      "loss": 0.0109,
       "step": 2685
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5674589183889659e-06,
-      "loss": 0.0053,
+      "loss": 0.0106,
       "step": 2686
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5605467123249646e-06,
-      "loss": 0.006,
+      "loss": 0.0059,
       "step": 2687
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5536489443407016e-06,
-      "loss": 0.008,
+      "loss": 0.01,
       "step": 2688
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.546765621846481e-06,
-      "loss": 0.0069,
+      "loss": 0.0067,
       "step": 2689
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5398967522370978e-06,
-      "loss": 0.0054,
+      "loss": 0.0049,
       "step": 2690
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5330423428918205e-06,
-      "loss": 0.0061,
+      "loss": 0.006,
       "step": 2691
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5262024011743742e-06,
-      "loss": 0.0084,
+      "loss": 0.0066,
       "step": 2692
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5193769344329488e-06,
-      "loss": 0.0055,
+      "loss": 0.0056,
       "step": 2693
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.512565950000181e-06,
-      "loss": 0.0057,
+      "loss": 0.0058,
       "step": 2694
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.5057694551931473e-06,
-      "loss": 0.0054,
+      "loss": 0.0057,
       "step": 2695
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4989874573133654e-06,
-      "loss": 0.0052,
+      "loss": 0.0051,
       "step": 2696
     },
     {
@@ -16225,43 +16225,43 @@
     {
       "epoch": 0.86,
       "learning_rate": 1.4854669814637145e-06,
-      "loss": 0.0101,
+      "loss": 0.007,
       "step": 2698
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.478728518018968e-06,
-      "loss": 0.0043,
+      "loss": 0.0032,
       "step": 2699
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4720045805517024e-06,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 2700
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.465295176285476e-06,
-      "loss": 0.0099,
+      "loss": 0.0072,
       "step": 2701
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4586003124282465e-06,
-      "loss": 0.0051,
+      "loss": 0.0046,
       "step": 2702
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4519199961723372e-06,
-      "loss": 0.0043,
+      "loss": 0.0063,
       "step": 2703
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4452542346944492e-06,
-      "loss": 0.0087,
+      "loss": 0.0062,
       "step": 2704
     },
     {
@@ -16273,25 +16273,25 @@
     {
       "epoch": 0.86,
       "learning_rate": 1.4319664047013593e-06,
-      "loss": 0.0051,
+      "loss": 0.0053,
       "step": 2706
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4253443504613434e-06,
-      "loss": 0.0033,
+      "loss": 0.0044,
       "step": 2707
     },
     {
       "epoch": 0.86,
       "learning_rate": 1.4187368795497175e-06,
-      "loss": 0.0059,
+      "loss": 0.0275,
       "step": 2708
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.4121439990649232e-06,
-      "loss": 0.0044,
+      "loss": 0.0045,
       "step": 2709
     },
     {
@@ -16303,67 +16303,67 @@
     {
       "epoch": 0.87,
       "learning_rate": 1.3990020376912165e-06,
-      "loss": 0.0062,
+      "loss": 0.0051,
       "step": 2711
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3924529709207874e-06,
-      "loss": 0.008,
+      "loss": 0.0074,
       "step": 2712
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3859185228141359e-06,
-      "loss": 0.0172,
+      "loss": 0.0139,
       "step": 2713
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3793987003912634e-06,
-      "loss": 0.0059,
+      "loss": 0.0061,
       "step": 2714
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.372893510656444e-06,
-      "loss": 0.0041,
+      "loss": 0.004,
       "step": 2715
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3664029605982453e-06,
-      "loss": 0.0135,
+      "loss": 0.011,
       "step": 2716
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3599270571894951e-06,
-      "loss": 0.0075,
+      "loss": 0.0072,
       "step": 2717
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3534658073872963e-06,
-      "loss": 0.0046,
+      "loss": 0.0044,
       "step": 2718
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3470192181330081e-06,
-      "loss": 0.0086,
+      "loss": 0.0076,
       "step": 2719
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3405872963522325e-06,
-      "loss": 0.0092,
+      "loss": 0.0106,
       "step": 2720
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.334170048954822e-06,
-      "loss": 0.0057,
+      "loss": 0.007,
       "step": 2721
     },
     {
@@ -16381,115 +16381,115 @@
     {
       "epoch": 0.87,
       "learning_rate": 1.3150064219247571e-06,
-      "loss": 0.0051,
+      "loss": 0.004,
       "step": 2724
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3086479408438912e-06,
-      "loss": 0.008,
+      "loss": 0.0079,
       "step": 2725
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.3023041684590214e-06,
-      "loss": 0.0068,
+      "loss": 0.0067,
       "step": 2726
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2959751115852896e-06,
-      "loss": 0.0114,
+      "loss": 0.0094,
       "step": 2727
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2896607770220453e-06,
-      "loss": 0.0046,
+      "loss": 0.0044,
       "step": 2728
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.283361171552807e-06,
-      "loss": 0.0152,
+      "loss": 0.0127,
       "step": 2729
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2770763019452813e-06,
-      "loss": 0.0036,
+      "loss": 0.0041,
       "step": 2730
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2708061749513323e-06,
-      "loss": 0.0059,
+      "loss": 0.006,
       "step": 2731
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2645507973069997e-06,
-      "loss": 0.0048,
+      "loss": 0.0053,
       "step": 2732
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2583101757324633e-06,
-      "loss": 0.0067,
+      "loss": 0.0059,
       "step": 2733
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2520843169320616e-06,
-      "loss": 0.0058,
+      "loss": 0.0053,
       "step": 2734
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.245873227594273e-06,
-      "loss": 0.0029,
+      "loss": 0.003,
       "step": 2735
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.239676914391698e-06,
-      "loss": 0.0096,
+      "loss": 0.0097,
       "step": 2736
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2334953839810787e-06,
-      "loss": 0.0037,
+      "loss": 0.0033,
       "step": 2737
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.2273286430032593e-06,
-      "loss": 0.0066,
+      "loss": 0.0077,
       "step": 2738
     },
     {
       "epoch": 0.87,
       "learning_rate": 1.221176698083209e-06,
-      "loss": 0.0096,
+      "loss": 0.0088,
       "step": 2739
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.2150395558299977e-06,
-      "loss": 0.0088,
+      "loss": 0.0087,
       "step": 2740
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.2089172228367862e-06,
-      "loss": 0.0159,
+      "loss": 0.017,
       "step": 2741
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.2028097056808384e-06,
-      "loss": 0.0042,
+      "loss": 0.0044,
       "step": 2742
     },
     {
@@ -16501,43 +16501,43 @@
     {
       "epoch": 0.88,
       "learning_rate": 1.1906391451101512e-06,
-      "loss": 0.0051,
+      "loss": 0.0043,
       "step": 2744
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1845761147703193e-06,
-      "loss": 0.0105,
+      "loss": 0.0099,
       "step": 2745
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1785279264175319e-06,
-      "loss": 0.0052,
+      "loss": 0.0048,
       "step": 2746
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.172494586549398e-06,
-      "loss": 0.0091,
+      "loss": 0.0089,
       "step": 2747
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1664761016475645e-06,
-      "loss": 0.0076,
+      "loss": 0.0063,
       "step": 2748
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1604724781777226e-06,
-      "loss": 0.0055,
+      "loss": 0.006,
       "step": 2749
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1544837225896033e-06,
-      "loss": 0.0116,
+      "loss": 0.0112,
       "step": 2750
     },
     {
@@ -16549,253 +16549,253 @@
     {
       "epoch": 0.88,
       "learning_rate": 1.142550840777552e-06,
-      "loss": 0.0074,
+      "loss": 0.0075,
       "step": 2752
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.136606727373189e-06,
-      "loss": 0.0047,
+      "loss": 0.0064,
       "step": 2753
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1306775074896508e-06,
-      "loss": 0.0075,
+      "loss": 0.0076,
       "step": 2754
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1247631874967374e-06,
-      "loss": 0.0107,
+      "loss": 0.01,
       "step": 2755
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1188637737482344e-06,
-      "loss": 0.007,
+      "loss": 0.0073,
       "step": 2756
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1129792725819137e-06,
-      "loss": 0.0036,
+      "loss": 0.0051,
       "step": 2757
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1071096903195287e-06,
-      "loss": 0.012,
+      "loss": 0.0115,
       "step": 2758
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.1012550332668037e-06,
-      "loss": 0.0034,
+      "loss": 0.003,
       "step": 2759
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0954153077134322e-06,
-      "loss": 0.006,
+      "loss": 0.0058,
       "step": 2760
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0895905199330592e-06,
-      "loss": 0.0049,
+      "loss": 0.0038,
       "step": 2761
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0837806761832891e-06,
-      "loss": 0.0073,
+      "loss": 0.0064,
       "step": 2762
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.077985782705671e-06,
-      "loss": 0.0049,
+      "loss": 0.005,
       "step": 2763
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.07220584572569e-06,
-      "loss": 0.004,
+      "loss": 0.0043,
       "step": 2764
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.066440871452764e-06,
-      "loss": 0.0074,
+      "loss": 0.0068,
       "step": 2765
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0606908660802407e-06,
-      "loss": 0.0099,
+      "loss": 0.0122,
       "step": 2766
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0549558357853766e-06,
-      "loss": 0.0094,
+      "loss": 0.0075,
       "step": 2767
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0492357867293522e-06,
-      "loss": 0.0085,
+      "loss": 0.01,
       "step": 2768
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.043530725057245e-06,
-      "loss": 0.0055,
+      "loss": 0.0059,
       "step": 2769
     },
     {
       "epoch": 0.88,
       "learning_rate": 1.0378406568980376e-06,
-      "loss": 0.006,
+      "loss": 0.0051,
       "step": 2770
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.0321655883645998e-06,
-      "loss": 0.0085,
+      "loss": 0.01,
       "step": 2771
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.0265055255536898e-06,
-      "loss": 0.0055,
+      "loss": 0.0056,
       "step": 2772
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.0208604745459454e-06,
-      "loss": 0.0159,
+      "loss": 0.0123,
       "step": 2773
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.0152304414058756e-06,
-      "loss": 0.0061,
+      "loss": 0.0076,
       "step": 2774
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.0096154321818551e-06,
-      "loss": 0.0026,
+      "loss": 0.0031,
       "step": 2775
     },
     {
       "epoch": 0.89,
       "learning_rate": 1.004015452906124e-06,
-      "loss": 0.006,
+      "loss": 0.0069,
       "step": 2776
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.984305095947648e-07,
-      "loss": 0.0032,
+      "loss": 0.0029,
       "step": 2777
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.928606082477214e-07,
-      "loss": 0.0041,
+      "loss": 0.0039,
       "step": 2778
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.873057548487623e-07,
-      "loss": 0.0061,
+      "loss": 0.0062,
       "step": 2779
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.817659553655023e-07,
-      "loss": 0.0101,
+      "loss": 0.0074,
       "step": 2780
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.76241215749379e-07,
-      "loss": 0.0039,
+      "loss": 0.0044,
       "step": 2781
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.707315419356477e-07,
-      "loss": 0.0113,
+      "loss": 0.0121,
       "step": 2782
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.65236939843384e-07,
-      "loss": 0.0092,
+      "loss": 0.0072,
       "step": 2783
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.59757415375468e-07,
-      "loss": 0.0058,
+      "loss": 0.0064,
       "step": 2784
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.542929744185885e-07,
-      "loss": 0.0104,
+      "loss": 0.0119,
       "step": 2785
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.488436228432184e-07,
-      "loss": 0.0051,
+      "loss": 0.0054,
       "step": 2786
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.43409366503628e-07,
-      "loss": 0.0152,
+      "loss": 0.0146,
       "step": 2787
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.379902112378708e-07,
-      "loss": 0.0069,
+      "loss": 0.0066,
       "step": 2788
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.325861628677779e-07,
-      "loss": 0.0104,
+      "loss": 0.0092,
       "step": 2789
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.271972271989438e-07,
-      "loss": 0.0094,
+      "loss": 0.0098,
       "step": 2790
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.218234100207357e-07,
-      "loss": 0.005,
+      "loss": 0.0046,
       "step": 2791
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.164647171062723e-07,
-      "loss": 0.0052,
+      "loss": 0.0051,
       "step": 2792
     },
     {
       "epoch": 0.89,
       "learning_rate": 9.111211542124287e-07,
-      "loss": 0.0061,
+      "loss": 0.008,
       "step": 2793
     },
     {
@@ -16807,169 +16807,169 @@
     {
       "epoch": 0.89,
       "learning_rate": 9.004794414328221e-07,
-      "loss": 0.0036,
+      "loss": 0.0047,
       "step": 2795
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.951813029795075e-07,
-      "loss": 0.0058,
+      "loss": 0.0064,
       "step": 2796
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.898983174117075e-07,
-      "loss": 0.0062,
+      "loss": 0.007,
       "step": 2797
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.846304904049557e-07,
-      "loss": 0.0068,
+      "loss": 0.0049,
       "step": 2798
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.793778276185149e-07,
-      "loss": 0.0048,
+      "loss": 0.0034,
       "step": 2799
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.74140334695343e-07,
-      "loss": 0.0068,
+      "loss": 0.0069,
       "step": 2800
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.689180172621142e-07,
-      "loss": 0.0117,
+      "loss": 0.0126,
       "step": 2801
     },
     {
       "epoch": 0.89,
       "learning_rate": 8.637108809291904e-07,
-      "loss": 0.0082,
+      "loss": 0.0092,
       "step": 2802
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.585189312906272e-07,
-      "loss": 0.0043,
+      "loss": 0.0035,
       "step": 2803
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.533421739241676e-07,
-      "loss": 0.0076,
+      "loss": 0.0056,
       "step": 2804
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.481806143912263e-07,
-      "loss": 0.0057,
+      "loss": 0.0058,
       "step": 2805
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.430342582368994e-07,
-      "loss": 0.0039,
+      "loss": 0.0069,
       "step": 2806
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.379031109899459e-07,
-      "loss": 0.0045,
+      "loss": 0.0048,
       "step": 2807
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.327871781627816e-07,
-      "loss": 0.0094,
+      "loss": 0.0111,
       "step": 2808
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.276864652514882e-07,
-      "loss": 0.0127,
+      "loss": 0.0142,
       "step": 2809
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.226009777357857e-07,
-      "loss": 0.0085,
+      "loss": 0.0088,
       "step": 2810
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.175307210790439e-07,
-      "loss": 0.0109,
+      "loss": 0.0102,
       "step": 2811
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.124757007282657e-07,
-      "loss": 0.005,
+      "loss": 0.0068,
       "step": 2812
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.074359221140903e-07,
-      "loss": 0.0047,
+      "loss": 0.0032,
       "step": 2813
     },
     {
       "epoch": 0.9,
       "learning_rate": 8.024113906507752e-07,
-      "loss": 0.0069,
+      "loss": 0.0054,
       "step": 2814
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.974021117362091e-07,
-      "loss": 0.0057,
+      "loss": 0.0064,
       "step": 2815
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.924080907518838e-07,
-      "loss": 0.0096,
+      "loss": 0.0095,
       "step": 2816
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.874293330629078e-07,
-      "loss": 0.0043,
+      "loss": 0.0042,
       "step": 2817
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.824658440179827e-07,
-      "loss": 0.0047,
+      "loss": 0.004,
       "step": 2818
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.775176289494146e-07,
-      "loss": 0.0062,
+      "loss": 0.0067,
       "step": 2819
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.725846931731029e-07,
-      "loss": 0.0064,
+      "loss": 0.0061,
       "step": 2820
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.676670419885234e-07,
-      "loss": 0.002,
+      "loss": 0.0024,
       "step": 2821
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.627646806787386e-07,
-      "loss": 0.0051,
+      "loss": 0.0052,
       "step": 2822
     },
     {
@@ -16981,97 +16981,97 @@
     {
       "epoch": 0.9,
       "learning_rate": 7.530058487336511e-07,
-      "loss": 0.0081,
+      "loss": 0.0069,
       "step": 2824
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.481493885823204e-07,
-      "loss": 0.0077,
+      "loss": 0.0089,
       "step": 2825
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.433082392737067e-07,
-      "loss": 0.0055,
+      "loss": 0.0049,
       "step": 2826
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.384824060086831e-07,
-      "loss": 0.0056,
+      "loss": 0.0052,
       "step": 2827
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.336718939716724e-07,
-      "loss": 0.0097,
+      "loss": 0.0083,
       "step": 2828
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.288767083306308e-07,
-      "loss": 0.012,
+      "loss": 0.0122,
       "step": 2829
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.240968542370574e-07,
-      "loss": 0.0067,
+      "loss": 0.0065,
       "step": 2830
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.19332336825973e-07,
-      "loss": 0.0117,
+      "loss": 0.0107,
       "step": 2831
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.145831612159248e-07,
-      "loss": 0.0062,
+      "loss": 0.0077,
       "step": 2832
     },
     {
       "epoch": 0.9,
       "learning_rate": 7.098493325089817e-07,
-      "loss": 0.0119,
+      "loss": 0.0098,
       "step": 2833
     },
     {
       "epoch": 0.91,
       "learning_rate": 7.051308557907187e-07,
-      "loss": 0.0073,
+      "loss": 0.0076,
       "step": 2834
     },
     {
       "epoch": 0.91,
       "learning_rate": 7.004277361302264e-07,
-      "loss": 0.0065,
+      "loss": 0.0062,
       "step": 2835
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.957399785800861e-07,
-      "loss": 0.0074,
+      "loss": 0.0078,
       "step": 2836
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.910675881763861e-07,
-      "loss": 0.0029,
+      "loss": 0.003,
       "step": 2837
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.86410569938703e-07,
-      "loss": 0.0067,
+      "loss": 0.0086,
       "step": 2838
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.817689288700962e-07,
-      "loss": 0.0064,
+      "loss": 0.0054,
       "step": 2839
     },
     {
@@ -17083,43 +17083,43 @@
     {
       "epoch": 0.91,
       "learning_rate": 6.725317981697448e-07,
-      "loss": 0.0068,
+      "loss": 0.0054,
       "step": 2841
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.679363184615006e-07,
-      "loss": 0.0076,
+      "loss": 0.003,
       "step": 2842
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.633562357693268e-07,
-      "loss": 0.0067,
+      "loss": 0.0073,
       "step": 2843
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.587915550136259e-07,
-      "loss": 0.0095,
+      "loss": 0.0059,
       "step": 2844
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.542422810982634e-07,
-      "loss": 0.007,
+      "loss": 0.0066,
       "step": 2845
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.497084189105501e-07,
-      "loss": 0.0059,
+      "loss": 0.0054,
       "step": 2846
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.451899733212363e-07,
-      "loss": 0.0084,
+      "loss": 0.0061,
       "step": 2847
     },
     {
@@ -17131,79 +17131,79 @@
     {
       "epoch": 0.91,
       "learning_rate": 6.361993513380121e-07,
-      "loss": 0.0089,
+      "loss": 0.0092,
       "step": 2849
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.317271846027772e-07,
-      "loss": 0.0077,
+      "loss": 0.0082,
       "step": 2850
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.272704537832846e-07,
-      "loss": 0.0092,
+      "loss": 0.0082,
       "step": 2851
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.228291636674227e-07,
-      "loss": 0.0051,
+      "loss": 0.0044,
       "step": 2852
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.184033190264982e-07,
-      "loss": 0.0028,
+      "loss": 0.003,
       "step": 2853
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.139929246152159e-07,
-      "loss": 0.0057,
+      "loss": 0.0052,
       "step": 2854
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.09597985171691e-07,
-      "loss": 0.0036,
+      "loss": 0.004,
       "step": 2855
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.052185054174314e-07,
-      "loss": 0.0083,
+      "loss": 0.009,
       "step": 2856
     },
     {
       "epoch": 0.91,
       "learning_rate": 6.008544900573354e-07,
-      "loss": 0.0033,
+      "loss": 0.0043,
       "step": 2857
     },
     {
       "epoch": 0.91,
       "learning_rate": 5.965059437796894e-07,
-      "loss": 0.0113,
+      "loss": 0.0104,
       "step": 2858
     },
     {
       "epoch": 0.91,
       "learning_rate": 5.921728712561614e-07,
-      "loss": 0.0141,
+      "loss": 0.0123,
       "step": 2859
     },
     {
       "epoch": 0.91,
       "learning_rate": 5.878552771417945e-07,
-      "loss": 0.0502,
+      "loss": 0.0094,
       "step": 2860
     },
     {
       "epoch": 0.91,
       "learning_rate": 5.835531660750048e-07,
-      "loss": 0.0055,
+      "loss": 0.005,
       "step": 2861
     },
     {
@@ -17215,223 +17215,223 @@
     {
       "epoch": 0.91,
       "learning_rate": 5.749954115546447e-07,
-      "loss": 0.0114,
+      "loss": 0.01,
       "step": 2863
     },
     {
       "epoch": 0.91,
       "learning_rate": 5.707397772947198e-07,
-      "loss": 0.0051,
+      "loss": 0.0059,
       "step": 2864
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.664996444696475e-07,
-      "loss": 0.0051,
+      "loss": 0.0052,
       "step": 2865
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.622750176346286e-07,
-      "loss": 0.0046,
+      "loss": 0.0053,
       "step": 2866
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.58065901328203e-07,
-      "loss": 0.0069,
+      "loss": 0.0067,
       "step": 2867
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.538723000722467e-07,
-      "loss": 0.0064,
+      "loss": 0.0065,
       "step": 2868
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.496942183719711e-07,
-      "loss": 0.0032,
+      "loss": 0.0028,
       "step": 2869
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.455316607159072e-07,
-      "loss": 0.0084,
+      "loss": 0.0076,
       "step": 2870
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.413846315759197e-07,
-      "loss": 0.0078,
+      "loss": 0.0079,
       "step": 2871
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.372531354071847e-07,
-      "loss": 0.0068,
+      "loss": 0.0064,
       "step": 2872
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.331371766481885e-07,
-      "loss": 0.0035,
+      "loss": 0.0039,
       "step": 2873
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.290367597207324e-07,
-      "loss": 0.0074,
+      "loss": 0.0047,
       "step": 2874
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.249518890299143e-07,
-      "loss": 0.0044,
+      "loss": 0.0043,
       "step": 2875
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.208825689641322e-07,
-      "loss": 0.0106,
+      "loss": 0.0096,
       "step": 2876
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.168288038950841e-07,
-      "loss": 0.0032,
+      "loss": 0.0031,
       "step": 2877
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.127905981777497e-07,
-      "loss": 0.0059,
+      "loss": 0.0035,
       "step": 2878
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.087679561503972e-07,
-      "loss": 0.0036,
+      "loss": 0.0033,
       "step": 2879
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.047608821345695e-07,
-      "loss": 0.0088,
+      "loss": 0.0083,
       "step": 2880
     },
     {
       "epoch": 0.92,
       "learning_rate": 5.007693804350932e-07,
-      "loss": 0.0058,
+      "loss": 0.005,
       "step": 2881
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.967934553400599e-07,
-      "loss": 0.0059,
+      "loss": 0.007,
       "step": 2882
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.928331111208262e-07,
-      "loss": 0.0097,
+      "loss": 0.0098,
       "step": 2883
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.888883520320137e-07,
-      "loss": 0.0055,
+      "loss": 0.0062,
       "step": 2884
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.849591823114991e-07,
-      "loss": 0.0025,
+      "loss": 0.0029,
       "step": 2885
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.810456061804108e-07,
-      "loss": 0.0036,
+      "loss": 0.0054,
       "step": 2886
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.771476278431258e-07,
-      "loss": 0.0056,
+      "loss": 0.0053,
       "step": 2887
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.732652514872643e-07,
-      "loss": 0.0077,
+      "loss": 0.0072,
       "step": 2888
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.6939848128368325e-07,
-      "loss": 0.0049,
+      "loss": 0.0057,
       "step": 2889
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.6554732138647806e-07,
-      "loss": 0.0085,
+      "loss": 0.0058,
       "step": 2890
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.617117759329675e-07,
-      "loss": 0.0084,
+      "loss": 0.0079,
       "step": 2891
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.5789184904370565e-07,
-      "loss": 0.0067,
+      "loss": 0.0065,
       "step": 2892
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.5408754482245464e-07,
-      "loss": 0.0063,
+      "loss": 0.0048,
       "step": 2893
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.502988673562053e-07,
-      "loss": 0.0085,
+      "loss": 0.009,
       "step": 2894
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.465258207151501e-07,
-      "loss": 0.0059,
+      "loss": 0.0056,
       "step": 2895
     },
     {
       "epoch": 0.92,
       "learning_rate": 4.4276840895269674e-07,
-      "loss": 0.006,
+      "loss": 0.0057,
       "step": 2896
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.390266361054529e-07,
-      "loss": 0.0101,
+      "loss": 0.0093,
       "step": 2897
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.353005061932297e-07,
-      "loss": 0.0101,
+      "loss": 0.0107,
       "step": 2898
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.3159002321902175e-07,
-      "loss": 0.0049,
+      "loss": 0.006,
       "step": 2899
     },
     {
@@ -17443,43 +17443,43 @@
     {
       "epoch": 0.93,
       "learning_rate": 4.2421601401262557e-07,
-      "loss": 0.0071,
+      "loss": 0.0091,
       "step": 2901
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.2055249570237374e-07,
-      "loss": 0.0043,
+      "loss": 0.0041,
       "step": 2902
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.169046401740134e-07,
-      "loss": 0.0066,
+      "loss": 0.0062,
       "step": 2903
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.132724513464514e-07,
-      "loss": 0.0052,
+      "loss": 0.0051,
       "step": 2904
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.0965593312177474e-07,
-      "loss": 0.0037,
+      "loss": 0.0032,
       "step": 2905
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.0605508938522216e-07,
-      "loss": 0.0124,
+      "loss": 0.0121,
       "step": 2906
     },
     {
       "epoch": 0.93,
       "learning_rate": 4.0246992400520435e-07,
-      "loss": 0.0061,
+      "loss": 0.0063,
       "step": 2907
     },
     {
@@ -17491,109 +17491,109 @@
     {
       "epoch": 0.93,
       "learning_rate": 3.9534664370418084e-07,
-      "loss": 0.007,
+      "loss": 0.0085,
       "step": 2909
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.9180853643575207e-07,
-      "loss": 0.0059,
+      "loss": 0.0048,
       "step": 2910
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.8828612282900857e-07,
-      "loss": 0.0053,
+      "loss": 0.0049,
       "step": 2911
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.847794066680982e-07,
-      "loss": 0.0104,
+      "loss": 0.008,
       "step": 2912
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.812883917203092e-07,
-      "loss": 0.0039,
+      "loss": 0.0057,
       "step": 2913
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.7781308173605477e-07,
-      "loss": 0.011,
+      "loss": 0.0103,
       "step": 2914
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.7435348044888187e-07,
-      "loss": 0.0076,
+      "loss": 0.007,
       "step": 2915
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.70909591575459e-07,
-      "loss": 0.0064,
+      "loss": 0.0066,
       "step": 2916
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.674814188155734e-07,
-      "loss": 0.0123,
+      "loss": 0.0114,
       "step": 2917
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.64068965852134e-07,
-      "loss": 0.0043,
+      "loss": 0.0045,
       "step": 2918
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.606722363511533e-07,
-      "loss": 0.0064,
+      "loss": 0.0051,
       "step": 2919
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.572912339617623e-07,
-      "loss": 0.0089,
+      "loss": 0.0085,
       "step": 2920
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.5392596231618555e-07,
-      "loss": 0.0085,
+      "loss": 0.0097,
       "step": 2921
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.5057642502975606e-07,
-      "loss": 0.0056,
+      "loss": 0.0053,
       "step": 2922
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.4724262570090036e-07,
-      "loss": 0.006,
+      "loss": 0.0064,
       "step": 2923
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.439245679111369e-07,
-      "loss": 0.0092,
+      "loss": 0.0118,
       "step": 2924
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.406222552250726e-07,
-      "loss": 0.0096,
+      "loss": 0.012,
       "step": 2925
     },
     {
       "epoch": 0.93,
       "learning_rate": 3.3733569119040296e-07,
-      "loss": 0.0083,
+      "loss": 0.007,
       "step": 2926
     },
     {
@@ -17605,151 +17605,151 @@
     {
       "epoch": 0.94,
       "learning_rate": 3.308098231814188e-07,
-      "loss": 0.0044,
+      "loss": 0.0042,
       "step": 2928
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.2757052621787796e-07,
-      "loss": 0.0039,
+      "loss": 0.0031,
       "step": 2929
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.2434699192727736e-07,
-      "loss": 0.004,
+      "loss": 0.005,
       "step": 2930
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.2113922377267845e-07,
-      "loss": 0.0095,
+      "loss": 0.0083,
       "step": 2931
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.179472252001997e-07,
-      "loss": 0.0083,
+      "loss": 0.0089,
       "step": 2932
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.1477099963902635e-07,
-      "loss": 0.0092,
+      "loss": 0.0089,
       "step": 2933
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.1161055050139555e-07,
-      "loss": 0.0075,
+      "loss": 0.0167,
       "step": 2934
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.084658811825947e-07,
-      "loss": 0.0058,
+      "loss": 0.0044,
       "step": 2935
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.053369950609597e-07,
-      "loss": 0.0047,
+      "loss": 0.0057,
       "step": 2936
     },
     {
       "epoch": 0.94,
       "learning_rate": 3.0222389549786845e-07,
-      "loss": 0.0083,
+      "loss": 0.008,
       "step": 2937
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.99126585837744e-07,
-      "loss": 0.0062,
+      "loss": 0.0057,
       "step": 2938
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.96045069408043e-07,
-      "loss": 0.0024,
+      "loss": 0.0019,
       "step": 2939
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.9297934951925564e-07,
-      "loss": 0.0048,
+      "loss": 0.0051,
       "step": 2940
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.8992942946490075e-07,
-      "loss": 0.0029,
+      "loss": 0.0063,
       "step": 2941
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.86895312521524e-07,
-      "loss": 0.007,
+      "loss": 0.0069,
       "step": 2942
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.838770019486964e-07,
-      "loss": 0.0037,
+      "loss": 0.0034,
       "step": 2943
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.808745009890057e-07,
-      "loss": 0.0104,
+      "loss": 0.0115,
       "step": 2944
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.778878128680534e-07,
-      "loss": 0.0081,
+      "loss": 0.0067,
       "step": 2945
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.749169407944563e-07,
-      "loss": 0.0056,
+      "loss": 0.006,
       "step": 2946
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.7196188795983955e-07,
-      "loss": 0.004,
+      "loss": 0.0037,
       "step": 2947
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.6902265753883213e-07,
-      "loss": 0.0071,
+      "loss": 0.0062,
       "step": 2948
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.6609925268906656e-07,
-      "loss": 0.0107,
+      "loss": 0.0099,
       "step": 2949
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.631916765511722e-07,
-      "loss": 0.0058,
+      "loss": 0.0065,
       "step": 2950
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.6029993224877367e-07,
-      "loss": 0.0064,
+      "loss": 0.0068,
       "step": 2951
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.574240228884911e-07,
-      "loss": 0.0082,
+      "loss": 0.0072,
       "step": 2952
     },
     {
@@ -17761,145 +17761,145 @@
     {
       "epoch": 0.94,
       "learning_rate": 2.5171972133566976e-07,
-      "loss": 0.0067,
+      "loss": 0.0062,
       "step": 2954
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.4889133527129573e-07,
-      "loss": 0.0066,
+      "loss": 0.0061,
       "step": 2955
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.4607879640535515e-07,
-      "loss": 0.0037,
+      "loss": 0.0043,
       "step": 2956
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.432821077593744e-07,
-      "loss": 0.0052,
+      "loss": 0.0049,
       "step": 2957
     },
     {
       "epoch": 0.94,
       "learning_rate": 2.4050127233785193e-07,
-      "loss": 0.0093,
+      "loss": 0.0091,
       "step": 2958
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.3773629312825284e-07,
-      "loss": 0.0084,
+      "loss": 0.007,
       "step": 2959
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.3498717310101114e-07,
-      "loss": 0.0051,
+      "loss": 0.0054,
       "step": 2960
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.32253915209521e-07,
-      "loss": 0.009,
+      "loss": 0.0075,
       "step": 2961
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.295365223901402e-07,
-      "loss": 0.0069,
+      "loss": 0.0068,
       "step": 2962
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.2683499756217362e-07,
-      "loss": 0.004,
+      "loss": 0.0043,
       "step": 2963
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.241493436278863e-07,
-      "loss": 0.0072,
+      "loss": 0.007,
       "step": 2964
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.2147956347249366e-07,
-      "loss": 0.0066,
+      "loss": 0.0051,
       "step": 2965
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.1882565996415305e-07,
-      "loss": 0.0089,
+      "loss": 0.0093,
       "step": 2966
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.1618763595397052e-07,
-      "loss": 0.0058,
+      "loss": 0.0048,
       "step": 2967
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.1356549427598571e-07,
-      "loss": 0.007,
+      "loss": 0.0063,
       "step": 2968
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.1095923774718197e-07,
-      "loss": 0.0074,
+      "loss": 0.0067,
       "step": 2969
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.083688691674762e-07,
-      "loss": 0.0058,
+      "loss": 0.0051,
       "step": 2970
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.0579439131971577e-07,
-      "loss": 0.0075,
+      "loss": 0.0045,
       "step": 2971
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.0323580696967659e-07,
-      "loss": 0.0076,
+      "loss": 0.0069,
       "step": 2972
     },
     {
       "epoch": 0.95,
       "learning_rate": 2.0069311886605822e-07,
-      "loss": 0.0055,
+      "loss": 0.0051,
       "step": 2973
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.9816632974048566e-07,
-      "loss": 0.0054,
+      "loss": 0.0064,
       "step": 2974
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.9565544230750576e-07,
-      "loss": 0.0067,
+      "loss": 0.0071,
       "step": 2975
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.9316045926457414e-07,
-      "loss": 0.0081,
+      "loss": 0.0075,
       "step": 2976
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.9068138329206842e-07,
-      "loss": 0.0048,
+      "loss": 0.0045,
       "step": 2977
     },
     {
@@ -17911,13 +17911,13 @@
     {
       "epoch": 0.95,
       "learning_rate": 1.857709631943766e-07,
-      "loss": 0.0048,
+      "loss": 0.0053,
       "step": 2979
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.8333962434448402e-07,
-      "loss": 0.0087,
+      "loss": 0.0065,
       "step": 2980
     },
     {
@@ -17929,43 +17929,43 @@
     {
       "epoch": 0.95,
       "learning_rate": 1.7852470210260708e-07,
-      "loss": 0.0096,
+      "loss": 0.0086,
       "step": 2982
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.7614112388331993e-07,
-      "loss": 0.0039,
+      "loss": 0.0062,
       "step": 2983
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.73773471018423e-07,
-      "loss": 0.005,
+      "loss": 0.006,
       "step": 2984
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.714217460515033e-07,
-      "loss": 0.003,
+      "loss": 0.0025,
       "step": 2985
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.690859515090265e-07,
-      "loss": 0.0059,
+      "loss": 0.006,
       "step": 2986
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.667660899003537e-07,
-      "loss": 0.0044,
+      "loss": 0.0039,
       "step": 2987
     },
     {
       "epoch": 0.95,
       "learning_rate": 1.6446216371772293e-07,
-      "loss": 0.0093,
+      "loss": 0.0129,
       "step": 2988
     },
     {
@@ -17977,25 +17977,25 @@
     {
       "epoch": 0.95,
       "learning_rate": 1.5990212751395173e-07,
-      "loss": 0.0074,
+      "loss": 0.006,
       "step": 2990
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.576460223916859e-07,
-      "loss": 0.0071,
+      "loss": 0.0049,
       "step": 2991
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.5540586249319977e-07,
-      "loss": 0.006,
+      "loss": 0.0061,
       "step": 2992
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.5318165022511322e-07,
-      "loss": 0.0052,
+      "loss": 0.0068,
       "step": 2993
     },
     {
@@ -18007,31 +18007,31 @@
     {
       "epoch": 0.96,
       "learning_rate": 1.4878107812093368e-07,
-      "loss": 0.0057,
+      "loss": 0.0066,
       "step": 2995
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.4660472301240235e-07,
-      "loss": 0.0066,
+      "loss": 0.0061,
       "step": 2996
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.4444432498938232e-07,
-      "loss": 0.0045,
+      "loss": 0.0035,
       "step": 2997
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.4229988637280367e-07,
-      "loss": 0.0085,
+      "loss": 0.0086,
       "step": 2998
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.4017140946644357e-07,
-      "loss": 0.0029,
+      "loss": 0.0036,
       "step": 2999
     },
     {
@@ -18043,19 +18043,19 @@
     {
       "epoch": 0.96,
       "learning_rate": 1.35962349913783e-07,
-      "loss": 0.0054,
+      "loss": 0.0061,
       "step": 3001
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.338817717892954e-07,
-      "loss": 0.0036,
+      "loss": 0.0042,
       "step": 3002
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.318171644186572e-07,
-      "loss": 0.0102,
+      "loss": 0.014,
       "step": 3003
     },
     {
@@ -18067,13 +18067,13 @@
     {
       "epoch": 0.96,
       "learning_rate": 1.2773587079384751e-07,
-      "loss": 0.0078,
+      "loss": 0.0094,
       "step": 3005
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.2571918892423207e-07,
-      "loss": 0.0066,
+      "loss": 0.0082,
       "step": 3006
     },
     {
@@ -18085,97 +18085,97 @@
     {
       "epoch": 0.96,
       "learning_rate": 1.2173376590324136e-07,
-      "loss": 0.0067,
+      "loss": 0.006,
       "step": 3008
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.197650290334279e-07,
-      "loss": 0.0068,
+      "loss": 0.0064,
       "step": 3009
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.1781227808315919e-07,
-      "loss": 0.0054,
+      "loss": 0.0057,
       "step": 3010
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.1587551515028716e-07,
-      "loss": 0.006,
+      "loss": 0.0053,
       "step": 3011
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.1395474231548742e-07,
-      "loss": 0.008,
+      "loss": 0.0058,
       "step": 3012
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.1204996164225433e-07,
-      "loss": 0.007,
+      "loss": 0.0067,
       "step": 3013
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.1016117517690938e-07,
-      "loss": 0.0103,
+      "loss": 0.0112,
       "step": 3014
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.0828838494858107e-07,
-      "loss": 0.0078,
+      "loss": 0.0084,
       "step": 3015
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.0643159296922334e-07,
-      "loss": 0.0066,
+      "loss": 0.0109,
       "step": 3016
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.0459080123359221e-07,
-      "loss": 0.0117,
+      "loss": 0.0108,
       "step": 3017
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.027660117192658e-07,
-      "loss": 0.0098,
+      "loss": 0.0109,
       "step": 3018
     },
     {
       "epoch": 0.96,
       "learning_rate": 1.009572263866243e-07,
-      "loss": 0.0107,
+      "loss": 0.0088,
       "step": 3019
     },
     {
       "epoch": 0.96,
       "learning_rate": 9.916444717885497e-08,
-      "loss": 0.0115,
+      "loss": 0.0136,
       "step": 3020
     },
     {
       "epoch": 0.96,
       "learning_rate": 9.738767602195053e-08,
-      "loss": 0.0069,
+      "loss": 0.0078,
       "step": 3021
     },
     {
       "epoch": 0.97,
       "learning_rate": 9.562691482470743e-08,
-      "loss": 0.0125,
+      "loss": 0.0099,
       "step": 3022
     },
     {
       "epoch": 0.97,
       "learning_rate": 9.38821654787192e-08,
-      "loss": 0.0073,
+      "loss": 0.0077,
       "step": 3023
     },
     {
@@ -18187,61 +18187,61 @@
     {
       "epoch": 0.97,
       "learning_rate": 9.044070982088537e-08,
-      "loss": 0.0071,
+      "loss": 0.007,
       "step": 3025
     },
     {
       "epoch": 0.97,
       "learning_rate": 8.874400720621401e-08,
-      "loss": 0.0087,
+      "loss": 0.011,
       "step": 3026
     },
     {
       "epoch": 0.97,
       "learning_rate": 8.706332383714432e-08,
-      "loss": 0.0098,
+      "loss": 0.0086,
       "step": 3027
     },
     {
       "epoch": 0.97,
       "learning_rate": 8.539866151924369e-08,
-      "loss": 0.0046,
+      "loss": 0.0041,
       "step": 3028
     },
     {
       "epoch": 0.97,
       "learning_rate": 8.37500220408699e-08,
-      "loss": 0.0102,
+      "loss": 0.0099,
       "step": 3029
     },
     {
       "epoch": 0.97,
       "learning_rate": 8.211740717316452e-08,
-      "loss": 0.0055,
+      "loss": 0.0057,
       "step": 3030
     },
     {
       "epoch": 0.97,
       "learning_rate": 8.050081867005455e-08,
-      "loss": 0.007,
+      "loss": 0.0071,
       "step": 3031
     },
     {
       "epoch": 0.97,
       "learning_rate": 7.89002582682491e-08,
-      "loss": 0.0134,
+      "loss": 0.013,
       "step": 3032
     },
     {
       "epoch": 0.97,
       "learning_rate": 7.731572768724105e-08,
-      "loss": 0.0124,
+      "loss": 0.012,
       "step": 3033
     },
     {
       "epoch": 0.97,
       "learning_rate": 7.574722862930206e-08,
-      "loss": 0.0068,
+      "loss": 0.0118,
       "step": 3034
     },
     {
@@ -18253,19 +18253,19 @@
     {
       "epoch": 0.97,
       "learning_rate": 7.265833180558512e-08,
-      "loss": 0.0048,
+      "loss": 0.0054,
       "step": 3036
     },
     {
       "epoch": 0.97,
       "learning_rate": 7.113793735822771e-08,
-      "loss": 0.0053,
+      "loss": 0.0067,
       "step": 3037
     },
     {
       "epoch": 0.97,
       "learning_rate": 6.96335810707721e-08,
-      "loss": 0.0054,
+      "loss": 0.0042,
       "step": 3038
     },
     {
@@ -18277,19 +18277,19 @@
     {
       "epoch": 0.97,
       "learning_rate": 6.667298942288736e-08,
-      "loss": 0.0087,
+      "loss": 0.0117,
       "step": 3040
     },
     {
       "epoch": 0.97,
       "learning_rate": 6.521675724303732e-08,
-      "loss": 0.0092,
+      "loss": 0.009,
       "step": 3041
     },
     {
       "epoch": 0.97,
       "learning_rate": 6.377656958424571e-08,
-      "loss": 0.0084,
+      "loss": 0.0077,
       "step": 3042
     },
     {
@@ -18301,43 +18301,43 @@
     {
       "epoch": 0.97,
       "learning_rate": 6.094433400140598e-08,
-      "loss": 0.0059,
+      "loss": 0.0048,
       "step": 3044
     },
     {
       "epoch": 0.97,
       "learning_rate": 5.955228912004396e-08,
-      "loss": 0.0089,
+      "loss": 0.0083,
       "step": 3045
     },
     {
       "epoch": 0.97,
       "learning_rate": 5.817629484511089e-08,
-      "loss": 0.0095,
+      "loss": 0.011,
       "step": 3046
     },
     {
       "epoch": 0.97,
       "learning_rate": 5.6816352654844305e-08,
-      "loss": 0.0052,
+      "loss": 0.0055,
       "step": 3047
     },
     {
       "epoch": 0.97,
       "learning_rate": 5.547246401023554e-08,
-      "loss": 0.0096,
+      "loss": 0.0077,
       "step": 3048
     },
     {
       "epoch": 0.97,
       "learning_rate": 5.4144630355034694e-08,
-      "loss": 0.0025,
+      "loss": 0.0026,
       "step": 3049
     },
     {
       "epoch": 0.97,
       "learning_rate": 5.283285311574071e-08,
-      "loss": 0.0059,
+      "loss": 0.0049,
       "step": 3050
     },
     {
@@ -18349,43 +18349,43 @@
     {
       "epoch": 0.97,
       "learning_rate": 5.025747350461629e-08,
-      "loss": 0.0131,
+      "loss": 0.0135,
       "step": 3052
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.8993873899532696e-08,
-      "loss": 0.0036,
+      "loss": 0.0039,
       "step": 3053
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.7746336243839614e-08,
-      "loss": 0.0044,
+      "loss": 0.0048,
       "step": 3054
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.651486187777831e-08,
-      "loss": 0.004,
+      "loss": 0.0069,
       "step": 3055
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.529945212432385e-08,
-      "loss": 0.0068,
+      "loss": 0.0065,
       "step": 3056
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.410010828920008e-08,
-      "loss": 0.0095,
+      "loss": 0.0079,
       "step": 3057
     },
     {
       "epoch": 0.98,
       "learning_rate": 4.291683166086968e-08,
-      "loss": 0.0062,
+      "loss": 0.006,
       "step": 3058
     },
     {
@@ -18397,31 +18397,31 @@
     {
       "epoch": 0.98,
       "learning_rate": 4.0598485092123717e-08,
-      "loss": 0.0068,
+      "loss": 0.0078,
       "step": 3060
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.9463417642320865e-08,
-      "loss": 0.0077,
+      "loss": 0.0072,
       "step": 3061
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.834442238053515e-08,
-      "loss": 0.0077,
+      "loss": 0.0074,
       "step": 3062
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.7241500508904955e-08,
-      "loss": 0.0069,
+      "loss": 0.0067,
       "step": 3063
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.615465321230915e-08,
-      "loss": 0.0063,
+      "loss": 0.0061,
       "step": 3064
     },
     {
@@ -18433,301 +18433,301 @@
     {
       "epoch": 0.98,
       "learning_rate": 3.402918699736857e-08,
-      "loss": 0.0082,
+      "loss": 0.0084,
       "step": 3066
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.299057036242559e-08,
-      "loss": 0.01,
+      "loss": 0.0098,
       "step": 3067
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.196803286931393e-08,
-      "loss": 0.0089,
+      "loss": 0.0088,
       "step": 3068
     },
     {
       "epoch": 0.98,
       "learning_rate": 3.0961575616549886e-08,
-      "loss": 0.0071,
+      "loss": 0.0078,
       "step": 3069
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.997119968537687e-08,
-      "loss": 0.0039,
+      "loss": 0.0054,
       "step": 3070
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.899690613976047e-08,
-      "loss": 0.0037,
+      "loss": 0.0034,
       "step": 3071
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.8038696026391753e-08,
-      "loss": 0.0094,
+      "loss": 0.0085,
       "step": 3072
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.7096570374680607e-08,
-      "loss": 0.0086,
+      "loss": 0.0072,
       "step": 3073
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.6170530196757415e-08,
-      "loss": 0.0066,
+      "loss": 0.0069,
       "step": 3074
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.5260576487471376e-08,
-      "loss": 0.0043,
+      "loss": 0.0036,
       "step": 3075
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.436671022439385e-08,
-      "loss": 0.0042,
+      "loss": 0.0044,
       "step": 3076
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.3488932367806694e-08,
-      "loss": 0.0074,
+      "loss": 0.0053,
       "step": 3077
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.2627243860712244e-08,
-      "loss": 0.0132,
+      "loss": 0.014,
       "step": 3078
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.1781645628830006e-08,
-      "loss": 0.0027,
+      "loss": 0.0031,
       "step": 3079
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.0952138580586643e-08,
-      "loss": 0.0089,
+      "loss": 0.0088,
       "step": 3080
     },
     {
       "epoch": 0.98,
       "learning_rate": 2.0138723607130982e-08,
-      "loss": 0.0054,
+      "loss": 0.0041,
       "step": 3081
     },
     {
       "epoch": 0.98,
       "learning_rate": 1.9341401582315676e-08,
-      "loss": 0.0044,
+      "loss": 0.0042,
       "step": 3082
     },
     {
       "epoch": 0.98,
       "learning_rate": 1.856017336271054e-08,
-      "loss": 0.0074,
+      "loss": 0.0076,
       "step": 3083
     },
     {
       "epoch": 0.98,
       "learning_rate": 1.7795039787590896e-08,
-      "loss": 0.0078,
+      "loss": 0.0073,
       "step": 3084
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.7046001678947543e-08,
-      "loss": 0.0055,
+      "loss": 0.0059,
       "step": 3085
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.63130598414718e-08,
-      "loss": 0.0142,
+      "loss": 0.0111,
       "step": 3086
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.5596215062570475e-08,
-      "loss": 0.0058,
+      "loss": 0.0066,
       "step": 3087
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.4895468112354205e-08,
-      "loss": 0.0068,
+      "loss": 0.0064,
       "step": 3088
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.4210819743637471e-08,
-      "loss": 0.0086,
+      "loss": 0.0111,
       "step": 3089
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.3542270691941916e-08,
-      "loss": 0.0125,
+      "loss": 0.0109,
       "step": 3090
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.2889821675494684e-08,
-      "loss": 0.0067,
+      "loss": 0.0042,
       "step": 3091
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.225347339522509e-08,
-      "loss": 0.0025,
+      "loss": 0.0028,
       "step": 3092
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.1633226534764618e-08,
-      "loss": 0.0131,
+      "loss": 0.0098,
       "step": 3093
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.1029081760450255e-08,
-      "loss": 0.0048,
+      "loss": 0.0062,
       "step": 3094
     },
     {
       "epoch": 0.99,
       "learning_rate": 1.0441039721314494e-08,
-      "loss": 0.0065,
+      "loss": 0.0061,
       "step": 3095
     },
     {
       "epoch": 0.99,
       "learning_rate": 9.869101049096996e-09,
-      "loss": 0.0135,
+      "loss": 0.0129,
       "step": 3096
     },
     {
       "epoch": 0.99,
       "learning_rate": 9.313266358234596e-09,
-      "loss": 0.0152,
+      "loss": 0.0118,
       "step": 3097
     },
     {
       "epoch": 0.99,
       "learning_rate": 8.773536245862967e-09,
-      "loss": 0.0026,
+      "loss": 0.0031,
       "step": 3098
     },
     {
       "epoch": 0.99,
       "learning_rate": 8.249911291816625e-09,
-      "loss": 0.0076,
+      "loss": 0.0068,
       "step": 3099
     },
     {
       "epoch": 0.99,
       "learning_rate": 7.742392058630587e-09,
-      "loss": 0.0118,
+      "loss": 0.0127,
       "step": 3100
     },
     {
       "epoch": 0.99,
       "learning_rate": 7.250979091533716e-09,
-      "loss": 0.0063,
+      "loss": 0.0066,
       "step": 3101
     },
     {
       "epoch": 0.99,
       "learning_rate": 6.775672918453713e-09,
-      "loss": 0.0094,
+      "loss": 0.0099,
       "step": 3102
     },
     {
       "epoch": 0.99,
       "learning_rate": 6.316474050015452e-09,
-      "loss": 0.005,
+      "loss": 0.0057,
       "step": 3103
     },
     {
       "epoch": 0.99,
       "learning_rate": 5.8733829795393215e-09,
-      "loss": 0.0106,
+      "loss": 0.011,
       "step": 3104
     },
     {
       "epoch": 0.99,
       "learning_rate": 5.446400183039546e-09,
-      "loss": 0.0049,
+      "loss": 0.0056,
       "step": 3105
     },
     {
       "epoch": 0.99,
       "learning_rate": 5.0355261192258635e-09,
-      "loss": 0.006,
+      "loss": 0.0061,
       "step": 3106
     },
     {
       "epoch": 0.99,
       "learning_rate": 4.640761229503521e-09,
-      "loss": 0.0088,
+      "loss": 0.0084,
       "step": 3107
     },
     {
       "epoch": 0.99,
       "learning_rate": 4.2621059379716055e-09,
-      "loss": 0.0029,
+      "loss": 0.0025,
       "step": 3108
     },
     {
       "epoch": 0.99,
       "learning_rate": 3.899560651418055e-09,
-      "loss": 0.0046,
+      "loss": 0.0042,
       "step": 3109
     },
     {
       "epoch": 0.99,
       "learning_rate": 3.5531257593313103e-09,
-      "loss": 0.0153,
+      "loss": 0.0157,
       "step": 3110
     },
     {
       "epoch": 0.99,
       "learning_rate": 3.2228016338853307e-09,
-      "loss": 0.0067,
+      "loss": 0.0084,
       "step": 3111
     },
     {
       "epoch": 0.99,
       "learning_rate": 2.9085886299512477e-09,
-      "loss": 0.0055,
+      "loss": 0.0052,
       "step": 3112
     },
     {
       "epoch": 0.99,
       "learning_rate": 2.6104870850890417e-09,
-      "loss": 0.0088,
+      "loss": 0.0069,
       "step": 3113
     },
     {
       "epoch": 0.99,
       "learning_rate": 2.3284973195508707e-09,
-      "loss": 0.0026,
+      "loss": 0.0031,
       "step": 3114
     },
     {
       "epoch": 0.99,
       "learning_rate": 2.062619636279406e-09,
-      "loss": 0.0096,
+      "loss": 0.0078,
       "step": 3115
     },
     {
@@ -18739,55 +18739,55 @@
     {
       "epoch": 1.0,
       "learning_rate": 1.579201641764838e-09,
-      "loss": 0.0045,
+      "loss": 0.0047,
       "step": 3117
     },
     {
       "epoch": 1.0,
       "learning_rate": 1.361661849859641e-09,
-      "loss": 0.006,
+      "loss": 0.0054,
       "step": 3118
     },
     {
       "epoch": 1.0,
       "learning_rate": 1.1602351788986276e-09,
-      "loss": 0.0051,
+      "loss": 0.0048,
       "step": 3119
     },
     {
       "epoch": 1.0,
       "learning_rate": 9.749218452753672e-10,
-      "loss": 0.0065,
+      "loss": 0.0057,
       "step": 3120
     },
     {
       "epoch": 1.0,
       "learning_rate": 8.057220480722771e-10,
-      "loss": 0.0051,
+      "loss": 0.0052,
       "step": 3121
     },
     {
       "epoch": 1.0,
       "learning_rate": 6.526359690622873e-10,
-      "loss": 0.0136,
+      "loss": 0.0131,
       "step": 3122
     },
     {
       "epoch": 1.0,
       "learning_rate": 5.156637727071756e-10,
-      "loss": 0.0061,
+      "loss": 0.0059,
       "step": 3123
     },
     {
       "epoch": 1.0,
       "learning_rate": 3.9480560615756665e-10,
-      "loss": 0.0042,
+      "loss": 0.0043,
       "step": 3124
     },
     {
       "epoch": 1.0,
       "learning_rate": 2.900615992496025e-10,
-      "loss": 0.0078,
+      "loss": 0.0082,
       "step": 3125
     },
     {
@@ -18799,31 +18799,31 @@
     {
       "epoch": 1.0,
       "learning_rate": 1.2891649715907062e-10,
-      "loss": 0.006,
+      "loss": 0.0066,
       "step": 3127
     },
     {
       "epoch": 1.0,
       "learning_rate": 7.251557509635465e-11,
-      "loss": 0.011,
+      "loss": 0.0059,
       "step": 3128
     },
     {
       "epoch": 1.0,
       "learning_rate": 3.2229158913321676e-11,
-      "loss": 0.0066,
+      "loss": 0.0099,
       "step": 3129
     },
     {
       "epoch": 1.0,
       "learning_rate": 8.057291892016316e-12,
-      "loss": 0.0066,
+      "loss": 0.0069,
       "step": 3130
     },
     {
       "epoch": 1.0,
       "learning_rate": 0.0,
-      "loss": 0.0085,
+      "loss": 0.0092,
       "step": 3131
     }
   ],
@@ -18832,8 +18832,8 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,
-  "total_flos": 1.978219893526364e+17,
-  "train_batch_size": 32,
+  "total_flos": 1.806664471459922e+17,
+  "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null
 }