FuLinNa_GLM4-9B-Chat / trainer_state.json
ZhuangXialie's picture
Upload trainer_state.json with huggingface_hub
24e479d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 15.0,
"eval_steps": 500,
"global_step": 840,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.017857142857142856,
"learning_rate": 7.142857142857143e-07,
"loss": 6.1665,
"step": 1
},
{
"epoch": 0.03571428571428571,
"learning_rate": 1.4285714285714286e-06,
"loss": 6.3965,
"step": 2
},
{
"epoch": 0.05357142857142857,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.5581,
"step": 3
},
{
"epoch": 0.07142857142857142,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.5117,
"step": 4
},
{
"epoch": 0.08928571428571429,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.5869,
"step": 5
},
{
"epoch": 0.10714285714285714,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.3418,
"step": 6
},
{
"epoch": 0.125,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.208,
"step": 7
},
{
"epoch": 0.14285714285714285,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.3003,
"step": 8
},
{
"epoch": 0.16071428571428573,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.4839,
"step": 9
},
{
"epoch": 0.17857142857142858,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.1602,
"step": 10
},
{
"epoch": 0.19642857142857142,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.3042,
"step": 11
},
{
"epoch": 0.21428571428571427,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.0332,
"step": 12
},
{
"epoch": 0.23214285714285715,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.5386,
"step": 13
},
{
"epoch": 0.25,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.2471,
"step": 14
},
{
"epoch": 0.26785714285714285,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.5361,
"step": 15
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.9404,
"step": 16
},
{
"epoch": 0.30357142857142855,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.4058,
"step": 17
},
{
"epoch": 0.32142857142857145,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.2627,
"step": 18
},
{
"epoch": 0.3392857142857143,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.272,
"step": 19
},
{
"epoch": 0.35714285714285715,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.4814,
"step": 20
},
{
"epoch": 0.375,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.2671,
"step": 21
},
{
"epoch": 0.39285714285714285,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.1455,
"step": 22
},
{
"epoch": 0.4107142857142857,
"grad_norm": 0.0,
"learning_rate": 2.1428571428571427e-06,
"loss": 6.1328,
"step": 23
},
{
"epoch": 0.42857142857142855,
"grad_norm": 230.8024139404297,
"learning_rate": 2.8571428571428573e-06,
"loss": 6.2754,
"step": 24
},
{
"epoch": 0.44642857142857145,
"grad_norm": 230.8024139404297,
"learning_rate": 3.5714285714285714e-06,
"loss": 4.6094,
"step": 25
},
{
"epoch": 0.4642857142857143,
"grad_norm": 230.8024139404297,
"learning_rate": 4.2857142857142855e-06,
"loss": 4.4473,
"step": 26
},
{
"epoch": 0.48214285714285715,
"grad_norm": 230.8024139404297,
"learning_rate": 4.9999999999999996e-06,
"loss": 4.6074,
"step": 27
},
{
"epoch": 0.5,
"grad_norm": 104.01466369628906,
"learning_rate": 5.7142857142857145e-06,
"loss": 4.8125,
"step": 28
},
{
"epoch": 0.5178571428571429,
"grad_norm": 104.01466369628906,
"learning_rate": 6.428571428571429e-06,
"loss": 3.7014,
"step": 29
},
{
"epoch": 0.5357142857142857,
"grad_norm": 104.01466369628906,
"learning_rate": 7.142857142857143e-06,
"loss": 3.8691,
"step": 30
},
{
"epoch": 0.5535714285714286,
"grad_norm": 104.01466369628906,
"learning_rate": 7.857142857142858e-06,
"loss": 3.6411,
"step": 31
},
{
"epoch": 0.5714285714285714,
"grad_norm": 38.60957336425781,
"learning_rate": 8.571428571428571e-06,
"loss": 3.7668,
"step": 32
},
{
"epoch": 0.5892857142857143,
"grad_norm": 38.60957336425781,
"learning_rate": 9.285714285714286e-06,
"loss": 3.115,
"step": 33
},
{
"epoch": 0.6071428571428571,
"grad_norm": 38.60957336425781,
"learning_rate": 9.999999999999999e-06,
"loss": 3.1592,
"step": 34
},
{
"epoch": 0.625,
"grad_norm": 38.60957336425781,
"learning_rate": 1.0714285714285714e-05,
"loss": 3.2053,
"step": 35
},
{
"epoch": 0.6428571428571429,
"grad_norm": 29.317277908325195,
"learning_rate": 1.1428571428571429e-05,
"loss": 3.1006,
"step": 36
},
{
"epoch": 0.6607142857142857,
"grad_norm": 29.317277908325195,
"learning_rate": 1.2142857142857144e-05,
"loss": 2.7161,
"step": 37
},
{
"epoch": 0.6785714285714286,
"grad_norm": 29.317277908325195,
"learning_rate": 1.2857142857142857e-05,
"loss": 2.6982,
"step": 38
},
{
"epoch": 0.6964285714285714,
"grad_norm": 29.317277908325195,
"learning_rate": 1.3571428571428572e-05,
"loss": 2.6333,
"step": 39
},
{
"epoch": 0.7142857142857143,
"grad_norm": 24.417295455932617,
"learning_rate": 1.4285714285714285e-05,
"loss": 2.7019,
"step": 40
},
{
"epoch": 0.7321428571428571,
"grad_norm": 24.417295455932617,
"learning_rate": 1.5e-05,
"loss": 2.3599,
"step": 41
},
{
"epoch": 0.75,
"grad_norm": 24.417295455932617,
"learning_rate": 1.5714285714285715e-05,
"loss": 2.3931,
"step": 42
},
{
"epoch": 0.7678571428571429,
"grad_norm": 24.417295455932617,
"learning_rate": 1.6428571428571432e-05,
"loss": 2.2354,
"step": 43
},
{
"epoch": 0.7857142857142857,
"grad_norm": 22.96832847595215,
"learning_rate": 1.7142857142857142e-05,
"loss": 2.5569,
"step": 44
},
{
"epoch": 0.8035714285714286,
"grad_norm": 22.96832847595215,
"learning_rate": 1.785714285714286e-05,
"loss": 2.1937,
"step": 45
},
{
"epoch": 0.8214285714285714,
"grad_norm": 22.96832847595215,
"learning_rate": 1.8571428571428572e-05,
"loss": 2.199,
"step": 46
},
{
"epoch": 0.8392857142857143,
"grad_norm": 22.96832847595215,
"learning_rate": 1.928571428571429e-05,
"loss": 2.2279,
"step": 47
},
{
"epoch": 0.8571428571428571,
"grad_norm": 18.068286895751953,
"learning_rate": 1.9999999999999998e-05,
"loss": 2.2932,
"step": 48
},
{
"epoch": 0.875,
"grad_norm": 18.068286895751953,
"learning_rate": 2.0714285714285715e-05,
"loss": 2.1881,
"step": 49
},
{
"epoch": 0.8928571428571429,
"grad_norm": 18.068286895751953,
"learning_rate": 2.1428571428571428e-05,
"loss": 2.1971,
"step": 50
},
{
"epoch": 0.9107142857142857,
"grad_norm": 18.068286895751953,
"learning_rate": 2.2142857142857145e-05,
"loss": 2.0692,
"step": 51
},
{
"epoch": 0.9285714285714286,
"grad_norm": 19.364484786987305,
"learning_rate": 2.2857142857142858e-05,
"loss": 2.1223,
"step": 52
},
{
"epoch": 0.9464285714285714,
"grad_norm": 19.364484786987305,
"learning_rate": 2.357142857142857e-05,
"loss": 2.0865,
"step": 53
},
{
"epoch": 0.9642857142857143,
"grad_norm": 19.364484786987305,
"learning_rate": 2.4285714285714288e-05,
"loss": 2.1046,
"step": 54
},
{
"epoch": 0.9821428571428571,
"grad_norm": 19.364484786987305,
"learning_rate": 2.5e-05,
"loss": 2.0803,
"step": 55
},
{
"epoch": 1.0,
"grad_norm": 15.066437721252441,
"learning_rate": 2.5714285714285714e-05,
"loss": 2.0692,
"step": 56
},
{
"epoch": 1.0178571428571428,
"grad_norm": 15.066437721252441,
"learning_rate": 2.6428571428571428e-05,
"loss": 2.0048,
"step": 57
},
{
"epoch": 1.0357142857142858,
"grad_norm": 15.066437721252441,
"learning_rate": 2.7142857142857144e-05,
"loss": 1.9008,
"step": 58
},
{
"epoch": 1.0535714285714286,
"grad_norm": 15.066437721252441,
"learning_rate": 2.7857142857142858e-05,
"loss": 1.8903,
"step": 59
},
{
"epoch": 1.0714285714285714,
"grad_norm": 13.632275581359863,
"learning_rate": 2.857142857142857e-05,
"loss": 1.9506,
"step": 60
},
{
"epoch": 1.0892857142857142,
"grad_norm": 13.632275581359863,
"learning_rate": 2.9285714285714284e-05,
"loss": 1.7871,
"step": 61
},
{
"epoch": 1.1071428571428572,
"grad_norm": 13.632275581359863,
"learning_rate": 3e-05,
"loss": 1.7432,
"step": 62
},
{
"epoch": 1.125,
"grad_norm": 13.632275581359863,
"learning_rate": 2.9962406015037596e-05,
"loss": 1.7252,
"step": 63
},
{
"epoch": 1.1428571428571428,
"grad_norm": 14.228506088256836,
"learning_rate": 2.9924812030075187e-05,
"loss": 1.8796,
"step": 64
},
{
"epoch": 1.1607142857142858,
"grad_norm": 14.228506088256836,
"learning_rate": 2.988721804511278e-05,
"loss": 1.8615,
"step": 65
},
{
"epoch": 1.1785714285714286,
"grad_norm": 14.228506088256836,
"learning_rate": 2.9849624060150376e-05,
"loss": 1.7247,
"step": 66
},
{
"epoch": 1.1964285714285714,
"grad_norm": 14.228506088256836,
"learning_rate": 2.981203007518797e-05,
"loss": 1.7633,
"step": 67
},
{
"epoch": 1.2142857142857142,
"grad_norm": 15.551071166992188,
"learning_rate": 2.9774436090225562e-05,
"loss": 1.873,
"step": 68
},
{
"epoch": 1.2321428571428572,
"grad_norm": 15.551071166992188,
"learning_rate": 2.9736842105263157e-05,
"loss": 1.8282,
"step": 69
},
{
"epoch": 1.25,
"grad_norm": 15.551071166992188,
"learning_rate": 2.9699248120300752e-05,
"loss": 1.6896,
"step": 70
},
{
"epoch": 1.2678571428571428,
"grad_norm": 15.551071166992188,
"learning_rate": 2.9661654135338347e-05,
"loss": 1.7615,
"step": 71
},
{
"epoch": 1.2857142857142856,
"grad_norm": 14.377410888671875,
"learning_rate": 2.9624060150375938e-05,
"loss": 1.7761,
"step": 72
},
{
"epoch": 1.3035714285714286,
"grad_norm": 14.377410888671875,
"learning_rate": 2.9586466165413533e-05,
"loss": 1.7222,
"step": 73
},
{
"epoch": 1.3214285714285714,
"grad_norm": 14.377410888671875,
"learning_rate": 2.9548872180451128e-05,
"loss": 1.8105,
"step": 74
},
{
"epoch": 1.3392857142857144,
"grad_norm": 14.377410888671875,
"learning_rate": 2.9511278195488722e-05,
"loss": 1.7357,
"step": 75
},
{
"epoch": 1.3571428571428572,
"grad_norm": 15.64229965209961,
"learning_rate": 2.9473684210526314e-05,
"loss": 1.6945,
"step": 76
},
{
"epoch": 1.375,
"grad_norm": 15.64229965209961,
"learning_rate": 2.943609022556391e-05,
"loss": 1.7681,
"step": 77
},
{
"epoch": 1.3928571428571428,
"grad_norm": 15.64229965209961,
"learning_rate": 2.9398496240601503e-05,
"loss": 1.6145,
"step": 78
},
{
"epoch": 1.4107142857142856,
"grad_norm": 15.64229965209961,
"learning_rate": 2.9360902255639098e-05,
"loss": 1.7286,
"step": 79
},
{
"epoch": 1.4285714285714286,
"grad_norm": 15.191073417663574,
"learning_rate": 2.932330827067669e-05,
"loss": 1.5326,
"step": 80
},
{
"epoch": 1.4464285714285714,
"grad_norm": 15.191073417663574,
"learning_rate": 2.9285714285714284e-05,
"loss": 1.7954,
"step": 81
},
{
"epoch": 1.4642857142857144,
"grad_norm": 15.191073417663574,
"learning_rate": 2.924812030075188e-05,
"loss": 1.6355,
"step": 82
},
{
"epoch": 1.4821428571428572,
"grad_norm": 15.191073417663574,
"learning_rate": 2.9210526315789474e-05,
"loss": 1.666,
"step": 83
},
{
"epoch": 1.5,
"grad_norm": 15.714180946350098,
"learning_rate": 2.917293233082707e-05,
"loss": 1.5962,
"step": 84
},
{
"epoch": 1.5178571428571428,
"grad_norm": 15.714180946350098,
"learning_rate": 2.9135338345864663e-05,
"loss": 1.7747,
"step": 85
},
{
"epoch": 1.5357142857142856,
"grad_norm": 15.714180946350098,
"learning_rate": 2.9097744360902258e-05,
"loss": 1.6663,
"step": 86
},
{
"epoch": 1.5535714285714286,
"grad_norm": 15.714180946350098,
"learning_rate": 2.9060150375939853e-05,
"loss": 1.7379,
"step": 87
},
{
"epoch": 1.5714285714285714,
"grad_norm": 14.792555809020996,
"learning_rate": 2.9022556390977444e-05,
"loss": 1.6414,
"step": 88
},
{
"epoch": 1.5892857142857144,
"grad_norm": 14.792555809020996,
"learning_rate": 2.898496240601504e-05,
"loss": 1.6715,
"step": 89
},
{
"epoch": 1.6071428571428572,
"grad_norm": 14.792555809020996,
"learning_rate": 2.8947368421052634e-05,
"loss": 1.7252,
"step": 90
},
{
"epoch": 1.625,
"grad_norm": 14.792555809020996,
"learning_rate": 2.8909774436090228e-05,
"loss": 1.4945,
"step": 91
},
{
"epoch": 1.6428571428571428,
"grad_norm": 14.602505683898926,
"learning_rate": 2.887218045112782e-05,
"loss": 1.819,
"step": 92
},
{
"epoch": 1.6607142857142856,
"grad_norm": 14.602505683898926,
"learning_rate": 2.8834586466165414e-05,
"loss": 1.5228,
"step": 93
},
{
"epoch": 1.6785714285714286,
"grad_norm": 14.602505683898926,
"learning_rate": 2.879699248120301e-05,
"loss": 1.6512,
"step": 94
},
{
"epoch": 1.6964285714285714,
"grad_norm": 14.602505683898926,
"learning_rate": 2.8759398496240604e-05,
"loss": 1.6498,
"step": 95
},
{
"epoch": 1.7142857142857144,
"grad_norm": 14.357579231262207,
"learning_rate": 2.8721804511278195e-05,
"loss": 1.7876,
"step": 96
},
{
"epoch": 1.7321428571428572,
"grad_norm": 14.357579231262207,
"learning_rate": 2.868421052631579e-05,
"loss": 1.6372,
"step": 97
},
{
"epoch": 1.75,
"grad_norm": 14.357579231262207,
"learning_rate": 2.8646616541353385e-05,
"loss": 1.5226,
"step": 98
},
{
"epoch": 1.7678571428571428,
"grad_norm": 14.357579231262207,
"learning_rate": 2.860902255639098e-05,
"loss": 1.6523,
"step": 99
},
{
"epoch": 1.7857142857142856,
"grad_norm": 12.28958797454834,
"learning_rate": 2.857142857142857e-05,
"loss": 1.4866,
"step": 100
},
{
"epoch": 1.8035714285714286,
"grad_norm": 12.28958797454834,
"learning_rate": 2.8533834586466166e-05,
"loss": 1.4967,
"step": 101
},
{
"epoch": 1.8214285714285714,
"grad_norm": 12.28958797454834,
"learning_rate": 2.849624060150376e-05,
"loss": 1.6534,
"step": 102
},
{
"epoch": 1.8392857142857144,
"grad_norm": 12.28958797454834,
"learning_rate": 2.8458646616541355e-05,
"loss": 1.4913,
"step": 103
},
{
"epoch": 1.8571428571428572,
"grad_norm": 11.686190605163574,
"learning_rate": 2.8421052631578946e-05,
"loss": 1.5909,
"step": 104
},
{
"epoch": 1.875,
"grad_norm": 11.686190605163574,
"learning_rate": 2.838345864661654e-05,
"loss": 1.6018,
"step": 105
},
{
"epoch": 1.8928571428571428,
"grad_norm": 11.686190605163574,
"learning_rate": 2.8345864661654136e-05,
"loss": 1.4518,
"step": 106
},
{
"epoch": 1.9107142857142856,
"grad_norm": 11.686190605163574,
"learning_rate": 2.830827067669173e-05,
"loss": 1.7177,
"step": 107
},
{
"epoch": 1.9285714285714286,
"grad_norm": 13.209571838378906,
"learning_rate": 2.8270676691729322e-05,
"loss": 1.5298,
"step": 108
},
{
"epoch": 1.9464285714285714,
"grad_norm": 13.209571838378906,
"learning_rate": 2.8233082706766917e-05,
"loss": 1.5328,
"step": 109
},
{
"epoch": 1.9642857142857144,
"grad_norm": 13.209571838378906,
"learning_rate": 2.819548872180451e-05,
"loss": 1.5262,
"step": 110
},
{
"epoch": 1.9821428571428572,
"grad_norm": 13.209571838378906,
"learning_rate": 2.8157894736842106e-05,
"loss": 1.5762,
"step": 111
},
{
"epoch": 2.0,
"grad_norm": 10.644370079040527,
"learning_rate": 2.8120300751879698e-05,
"loss": 1.4681,
"step": 112
},
{
"epoch": 2.017857142857143,
"grad_norm": 10.644370079040527,
"learning_rate": 2.8082706766917292e-05,
"loss": 1.3092,
"step": 113
},
{
"epoch": 2.0357142857142856,
"grad_norm": 10.644370079040527,
"learning_rate": 2.8045112781954887e-05,
"loss": 1.3781,
"step": 114
},
{
"epoch": 2.0535714285714284,
"grad_norm": 10.644370079040527,
"learning_rate": 2.8007518796992482e-05,
"loss": 1.328,
"step": 115
},
{
"epoch": 2.0714285714285716,
"grad_norm": 10.72980785369873,
"learning_rate": 2.7969924812030073e-05,
"loss": 1.4338,
"step": 116
},
{
"epoch": 2.0892857142857144,
"grad_norm": 10.72980785369873,
"learning_rate": 2.7932330827067668e-05,
"loss": 1.2646,
"step": 117
},
{
"epoch": 2.107142857142857,
"grad_norm": 10.72980785369873,
"learning_rate": 2.7894736842105263e-05,
"loss": 1.4899,
"step": 118
},
{
"epoch": 2.125,
"grad_norm": 10.72980785369873,
"learning_rate": 2.7857142857142858e-05,
"loss": 1.2913,
"step": 119
},
{
"epoch": 2.142857142857143,
"grad_norm": 12.673820495605469,
"learning_rate": 2.781954887218045e-05,
"loss": 1.3636,
"step": 120
},
{
"epoch": 2.1607142857142856,
"grad_norm": 12.673820495605469,
"learning_rate": 2.7781954887218044e-05,
"loss": 1.272,
"step": 121
},
{
"epoch": 2.1785714285714284,
"grad_norm": 12.673820495605469,
"learning_rate": 2.774436090225564e-05,
"loss": 1.2552,
"step": 122
},
{
"epoch": 2.1964285714285716,
"grad_norm": 12.673820495605469,
"learning_rate": 2.7706766917293233e-05,
"loss": 1.1824,
"step": 123
},
{
"epoch": 2.2142857142857144,
"grad_norm": 11.79961109161377,
"learning_rate": 2.7669172932330825e-05,
"loss": 1.276,
"step": 124
},
{
"epoch": 2.232142857142857,
"grad_norm": 11.79961109161377,
"learning_rate": 2.7631578947368423e-05,
"loss": 1.2714,
"step": 125
},
{
"epoch": 2.25,
"grad_norm": 11.79961109161377,
"learning_rate": 2.7593984962406017e-05,
"loss": 1.288,
"step": 126
},
{
"epoch": 2.267857142857143,
"grad_norm": 11.79961109161377,
"learning_rate": 2.7556390977443612e-05,
"loss": 1.3359,
"step": 127
},
{
"epoch": 2.2857142857142856,
"grad_norm": 14.634297370910645,
"learning_rate": 2.7518796992481204e-05,
"loss": 1.2911,
"step": 128
},
{
"epoch": 2.3035714285714284,
"grad_norm": 14.634297370910645,
"learning_rate": 2.74812030075188e-05,
"loss": 1.1898,
"step": 129
},
{
"epoch": 2.3214285714285716,
"grad_norm": 14.634297370910645,
"learning_rate": 2.7443609022556393e-05,
"loss": 1.2327,
"step": 130
},
{
"epoch": 2.3392857142857144,
"grad_norm": 14.634297370910645,
"learning_rate": 2.7406015037593988e-05,
"loss": 1.405,
"step": 131
},
{
"epoch": 2.357142857142857,
"grad_norm": 16.606494903564453,
"learning_rate": 2.736842105263158e-05,
"loss": 1.2003,
"step": 132
},
{
"epoch": 2.375,
"grad_norm": 16.606494903564453,
"learning_rate": 2.7330827067669174e-05,
"loss": 1.4715,
"step": 133
},
{
"epoch": 2.392857142857143,
"grad_norm": 16.606494903564453,
"learning_rate": 2.729323308270677e-05,
"loss": 1.3771,
"step": 134
},
{
"epoch": 2.4107142857142856,
"grad_norm": 16.606494903564453,
"learning_rate": 2.7255639097744363e-05,
"loss": 1.165,
"step": 135
},
{
"epoch": 2.4285714285714284,
"grad_norm": 15.750870704650879,
"learning_rate": 2.7218045112781955e-05,
"loss": 1.3275,
"step": 136
},
{
"epoch": 2.4464285714285716,
"grad_norm": 15.750870704650879,
"learning_rate": 2.718045112781955e-05,
"loss": 1.3169,
"step": 137
},
{
"epoch": 2.4642857142857144,
"grad_norm": 15.750870704650879,
"learning_rate": 2.7142857142857144e-05,
"loss": 1.2426,
"step": 138
},
{
"epoch": 2.482142857142857,
"grad_norm": 15.750870704650879,
"learning_rate": 2.710526315789474e-05,
"loss": 1.2549,
"step": 139
},
{
"epoch": 2.5,
"grad_norm": 16.149423599243164,
"learning_rate": 2.706766917293233e-05,
"loss": 1.3336,
"step": 140
},
{
"epoch": 2.517857142857143,
"grad_norm": 16.149423599243164,
"learning_rate": 2.7030075187969925e-05,
"loss": 1.2404,
"step": 141
},
{
"epoch": 2.5357142857142856,
"grad_norm": 16.149423599243164,
"learning_rate": 2.699248120300752e-05,
"loss": 1.2969,
"step": 142
},
{
"epoch": 2.553571428571429,
"grad_norm": 16.149423599243164,
"learning_rate": 2.6954887218045115e-05,
"loss": 1.1973,
"step": 143
},
{
"epoch": 2.571428571428571,
"grad_norm": 13.522272109985352,
"learning_rate": 2.6917293233082706e-05,
"loss": 1.1382,
"step": 144
},
{
"epoch": 2.5892857142857144,
"grad_norm": 13.522272109985352,
"learning_rate": 2.68796992481203e-05,
"loss": 1.2349,
"step": 145
},
{
"epoch": 2.607142857142857,
"grad_norm": 13.522272109985352,
"learning_rate": 2.6842105263157896e-05,
"loss": 1.1815,
"step": 146
},
{
"epoch": 2.625,
"grad_norm": 13.522272109985352,
"learning_rate": 2.680451127819549e-05,
"loss": 1.0928,
"step": 147
},
{
"epoch": 2.642857142857143,
"grad_norm": 11.987059593200684,
"learning_rate": 2.6766917293233082e-05,
"loss": 1.3141,
"step": 148
},
{
"epoch": 2.6607142857142856,
"grad_norm": 11.987059593200684,
"learning_rate": 2.6729323308270676e-05,
"loss": 1.2795,
"step": 149
},
{
"epoch": 2.678571428571429,
"grad_norm": 11.987059593200684,
"learning_rate": 2.669172932330827e-05,
"loss": 1.0408,
"step": 150
},
{
"epoch": 2.696428571428571,
"grad_norm": 11.987059593200684,
"learning_rate": 2.6654135338345866e-05,
"loss": 1.2148,
"step": 151
},
{
"epoch": 2.7142857142857144,
"grad_norm": 11.788473129272461,
"learning_rate": 2.6616541353383457e-05,
"loss": 1.2111,
"step": 152
},
{
"epoch": 2.732142857142857,
"grad_norm": 11.788473129272461,
"learning_rate": 2.6578947368421052e-05,
"loss": 1.2054,
"step": 153
},
{
"epoch": 2.75,
"grad_norm": 11.788473129272461,
"learning_rate": 2.6541353383458647e-05,
"loss": 1.2268,
"step": 154
},
{
"epoch": 2.767857142857143,
"grad_norm": 11.788473129272461,
"learning_rate": 2.650375939849624e-05,
"loss": 1.1021,
"step": 155
},
{
"epoch": 2.7857142857142856,
"grad_norm": 11.825960159301758,
"learning_rate": 2.6466165413533833e-05,
"loss": 1.2378,
"step": 156
},
{
"epoch": 2.803571428571429,
"grad_norm": 11.825960159301758,
"learning_rate": 2.6428571428571428e-05,
"loss": 1.097,
"step": 157
},
{
"epoch": 2.821428571428571,
"grad_norm": 11.825960159301758,
"learning_rate": 2.6390977443609022e-05,
"loss": 1.2128,
"step": 158
},
{
"epoch": 2.8392857142857144,
"grad_norm": 11.825960159301758,
"learning_rate": 2.6353383458646617e-05,
"loss": 1.1729,
"step": 159
},
{
"epoch": 2.857142857142857,
"grad_norm": 12.994108200073242,
"learning_rate": 2.631578947368421e-05,
"loss": 1.3422,
"step": 160
},
{
"epoch": 2.875,
"grad_norm": 12.994108200073242,
"learning_rate": 2.6278195488721803e-05,
"loss": 1.1797,
"step": 161
},
{
"epoch": 2.892857142857143,
"grad_norm": 12.994108200073242,
"learning_rate": 2.6240601503759398e-05,
"loss": 1.205,
"step": 162
},
{
"epoch": 2.9107142857142856,
"grad_norm": 12.994108200073242,
"learning_rate": 2.6203007518796993e-05,
"loss": 1.1354,
"step": 163
},
{
"epoch": 2.928571428571429,
"grad_norm": 13.220701217651367,
"learning_rate": 2.6165413533834584e-05,
"loss": 1.1998,
"step": 164
},
{
"epoch": 2.946428571428571,
"grad_norm": 13.220701217651367,
"learning_rate": 2.612781954887218e-05,
"loss": 1.0995,
"step": 165
},
{
"epoch": 2.9642857142857144,
"grad_norm": 13.220701217651367,
"learning_rate": 2.6090225563909774e-05,
"loss": 1.1673,
"step": 166
},
{
"epoch": 2.982142857142857,
"grad_norm": 13.220701217651367,
"learning_rate": 2.605263157894737e-05,
"loss": 1.0815,
"step": 167
},
{
"epoch": 3.0,
"grad_norm": 12.115779876708984,
"learning_rate": 2.6015037593984963e-05,
"loss": 1.1843,
"step": 168
},
{
"epoch": 3.017857142857143,
"grad_norm": 12.115779876708984,
"learning_rate": 2.5977443609022558e-05,
"loss": 0.926,
"step": 169
},
{
"epoch": 3.0357142857142856,
"grad_norm": 12.115779876708984,
"learning_rate": 2.5939849624060153e-05,
"loss": 1.0279,
"step": 170
},
{
"epoch": 3.0535714285714284,
"grad_norm": 12.115779876708984,
"learning_rate": 2.5902255639097747e-05,
"loss": 0.9919,
"step": 171
},
{
"epoch": 3.0714285714285716,
"grad_norm": 10.763752937316895,
"learning_rate": 2.586466165413534e-05,
"loss": 0.957,
"step": 172
},
{
"epoch": 3.0892857142857144,
"grad_norm": 10.763752937316895,
"learning_rate": 2.5827067669172934e-05,
"loss": 0.8973,
"step": 173
},
{
"epoch": 3.107142857142857,
"grad_norm": 10.763752937316895,
"learning_rate": 2.578947368421053e-05,
"loss": 0.975,
"step": 174
},
{
"epoch": 3.125,
"grad_norm": 10.763752937316895,
"learning_rate": 2.5751879699248123e-05,
"loss": 0.9436,
"step": 175
},
{
"epoch": 3.142857142857143,
"grad_norm": 9.575577735900879,
"learning_rate": 2.5714285714285714e-05,
"loss": 0.9521,
"step": 176
},
{
"epoch": 3.1607142857142856,
"grad_norm": 9.575577735900879,
"learning_rate": 2.567669172932331e-05,
"loss": 0.9607,
"step": 177
},
{
"epoch": 3.1785714285714284,
"grad_norm": 9.575577735900879,
"learning_rate": 2.5639097744360904e-05,
"loss": 0.9691,
"step": 178
},
{
"epoch": 3.1964285714285716,
"grad_norm": 9.575577735900879,
"learning_rate": 2.56015037593985e-05,
"loss": 0.9874,
"step": 179
},
{
"epoch": 3.2142857142857144,
"grad_norm": 15.285778999328613,
"learning_rate": 2.556390977443609e-05,
"loss": 0.8865,
"step": 180
},
{
"epoch": 3.232142857142857,
"grad_norm": 15.285778999328613,
"learning_rate": 2.5526315789473685e-05,
"loss": 0.8138,
"step": 181
},
{
"epoch": 3.25,
"grad_norm": 15.285778999328613,
"learning_rate": 2.548872180451128e-05,
"loss": 0.9224,
"step": 182
},
{
"epoch": 3.267857142857143,
"grad_norm": 15.285778999328613,
"learning_rate": 2.5451127819548874e-05,
"loss": 0.924,
"step": 183
},
{
"epoch": 3.2857142857142856,
"grad_norm": 13.550517082214355,
"learning_rate": 2.5413533834586466e-05,
"loss": 0.9641,
"step": 184
},
{
"epoch": 3.3035714285714284,
"grad_norm": 13.550517082214355,
"learning_rate": 2.537593984962406e-05,
"loss": 0.9249,
"step": 185
},
{
"epoch": 3.3214285714285716,
"grad_norm": 13.550517082214355,
"learning_rate": 2.5338345864661655e-05,
"loss": 0.8151,
"step": 186
},
{
"epoch": 3.3392857142857144,
"grad_norm": 13.550517082214355,
"learning_rate": 2.530075187969925e-05,
"loss": 0.9991,
"step": 187
},
{
"epoch": 3.357142857142857,
"grad_norm": 13.550517082214355,
"learning_rate": 2.530075187969925e-05,
"loss": 0.9329,
"step": 188
},
{
"epoch": 3.375,
"grad_norm": 13.550517082214355,
"learning_rate": 2.530075187969925e-05,
"loss": 0.8963,
"step": 189
},
{
"epoch": 3.392857142857143,
"grad_norm": 13.550517082214355,
"learning_rate": 2.530075187969925e-05,
"loss": 0.8763,
"step": 190
},
{
"epoch": 3.4107142857142856,
"grad_norm": 13.550517082214355,
"learning_rate": 2.530075187969925e-05,
"loss": 0.895,
"step": 191
},
{
"epoch": 3.4285714285714284,
"grad_norm": 12.928659439086914,
"learning_rate": 2.526315789473684e-05,
"loss": 0.9208,
"step": 192
},
{
"epoch": 3.4464285714285716,
"grad_norm": 12.928659439086914,
"learning_rate": 2.5225563909774436e-05,
"loss": 1.0038,
"step": 193
},
{
"epoch": 3.4642857142857144,
"grad_norm": 12.928659439086914,
"learning_rate": 2.518796992481203e-05,
"loss": 1.0118,
"step": 194
},
{
"epoch": 3.482142857142857,
"grad_norm": 12.928659439086914,
"learning_rate": 2.5150375939849626e-05,
"loss": 1.0811,
"step": 195
},
{
"epoch": 3.5,
"grad_norm": 51.4168586730957,
"learning_rate": 2.5112781954887217e-05,
"loss": 0.9854,
"step": 196
},
{
"epoch": 3.517857142857143,
"grad_norm": 51.4168586730957,
"learning_rate": 2.507518796992481e-05,
"loss": 1.0046,
"step": 197
},
{
"epoch": 3.5357142857142856,
"grad_norm": 51.4168586730957,
"learning_rate": 2.5037593984962406e-05,
"loss": 1.142,
"step": 198
},
{
"epoch": 3.553571428571429,
"grad_norm": 51.4168586730957,
"learning_rate": 2.5e-05,
"loss": 1.144,
"step": 199
},
{
"epoch": 3.571428571428571,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4962406015037593e-05,
"loss": 0.9363,
"step": 200
},
{
"epoch": 3.5892857142857144,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4924812030075187e-05,
"loss": 1.0331,
"step": 201
},
{
"epoch": 3.607142857142857,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4887218045112782e-05,
"loss": 0.9746,
"step": 202
},
{
"epoch": 3.625,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4849624060150377e-05,
"loss": 0.8302,
"step": 203
},
{
"epoch": 3.642857142857143,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4849624060150377e-05,
"loss": 1.1033,
"step": 204
},
{
"epoch": 3.6607142857142856,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4849624060150377e-05,
"loss": 1.1885,
"step": 205
},
{
"epoch": 3.678571428571429,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4849624060150377e-05,
"loss": 1.0773,
"step": 206
},
{
"epoch": 3.696428571428571,
"grad_norm": 23.4835147857666,
"learning_rate": 2.4849624060150377e-05,
"loss": 0.9265,
"step": 207
},
{
"epoch": 3.7142857142857144,
"grad_norm": 34.35456085205078,
"learning_rate": 2.4812030075187968e-05,
"loss": 1.0683,
"step": 208
},
{
"epoch": 3.732142857142857,
"grad_norm": 34.35456085205078,
"learning_rate": 2.4774436090225563e-05,
"loss": 0.9968,
"step": 209
},
{
"epoch": 3.75,
"grad_norm": 34.35456085205078,
"learning_rate": 2.4736842105263158e-05,
"loss": 0.9985,
"step": 210
},
{
"epoch": 3.767857142857143,
"grad_norm": 34.35456085205078,
"learning_rate": 2.4699248120300752e-05,
"loss": 1.1341,
"step": 211
},
{
"epoch": 3.7857142857142856,
"grad_norm": 16.934173583984375,
"learning_rate": 2.4661654135338344e-05,
"loss": 1.1264,
"step": 212
},
{
"epoch": 3.803571428571429,
"grad_norm": 16.934173583984375,
"learning_rate": 2.462406015037594e-05,
"loss": 1.0889,
"step": 213
},
{
"epoch": 3.821428571428571,
"grad_norm": 16.934173583984375,
"learning_rate": 2.4586466165413533e-05,
"loss": 0.883,
"step": 214
},
{
"epoch": 3.8392857142857144,
"grad_norm": 16.934173583984375,
"learning_rate": 2.4548872180451128e-05,
"loss": 1.0402,
"step": 215
},
{
"epoch": 3.857142857142857,
"grad_norm": 32.026092529296875,
"learning_rate": 2.451127819548872e-05,
"loss": 1.1568,
"step": 216
},
{
"epoch": 3.875,
"grad_norm": 32.026092529296875,
"learning_rate": 2.4473684210526318e-05,
"loss": 0.852,
"step": 217
},
{
"epoch": 3.892857142857143,
"grad_norm": 32.026092529296875,
"learning_rate": 2.4436090225563912e-05,
"loss": 1.0053,
"step": 218
},
{
"epoch": 3.9107142857142856,
"grad_norm": 32.026092529296875,
"learning_rate": 2.4398496240601507e-05,
"loss": 0.8912,
"step": 219
},
{
"epoch": 3.928571428571429,
"grad_norm": 16.007280349731445,
"learning_rate": 2.43609022556391e-05,
"loss": 0.9969,
"step": 220
},
{
"epoch": 3.946428571428571,
"grad_norm": 16.007280349731445,
"learning_rate": 2.4323308270676693e-05,
"loss": 0.8044,
"step": 221
},
{
"epoch": 3.9642857142857144,
"grad_norm": 16.007280349731445,
"learning_rate": 2.4285714285714288e-05,
"loss": 0.9554,
"step": 222
},
{
"epoch": 3.982142857142857,
"grad_norm": 16.007280349731445,
"learning_rate": 2.4248120300751883e-05,
"loss": 0.8043,
"step": 223
},
{
"epoch": 4.0,
"grad_norm": 36.69541931152344,
"learning_rate": 2.4210526315789474e-05,
"loss": 1.0418,
"step": 224
},
{
"epoch": 4.017857142857143,
"grad_norm": 36.69541931152344,
"learning_rate": 2.417293233082707e-05,
"loss": 0.7138,
"step": 225
},
{
"epoch": 4.035714285714286,
"grad_norm": 36.69541931152344,
"learning_rate": 2.4135338345864664e-05,
"loss": 0.8305,
"step": 226
},
{
"epoch": 4.053571428571429,
"grad_norm": 36.69541931152344,
"learning_rate": 2.409774436090226e-05,
"loss": 0.7274,
"step": 227
},
{
"epoch": 4.071428571428571,
"grad_norm": 12.90493392944336,
"learning_rate": 2.406015037593985e-05,
"loss": 0.8296,
"step": 228
},
{
"epoch": 4.089285714285714,
"grad_norm": 12.90493392944336,
"learning_rate": 2.4022556390977444e-05,
"loss": 0.7682,
"step": 229
},
{
"epoch": 4.107142857142857,
"grad_norm": 12.90493392944336,
"learning_rate": 2.398496240601504e-05,
"loss": 0.8076,
"step": 230
},
{
"epoch": 4.125,
"grad_norm": 12.90493392944336,
"learning_rate": 2.3947368421052634e-05,
"loss": 0.7304,
"step": 231
},
{
"epoch": 4.142857142857143,
"grad_norm": 26.516355514526367,
"learning_rate": 2.3909774436090225e-05,
"loss": 0.7007,
"step": 232
},
{
"epoch": 4.160714285714286,
"grad_norm": 26.516355514526367,
"learning_rate": 2.387218045112782e-05,
"loss": 0.7606,
"step": 233
},
{
"epoch": 4.178571428571429,
"grad_norm": 26.516355514526367,
"learning_rate": 2.3834586466165415e-05,
"loss": 0.7977,
"step": 234
},
{
"epoch": 4.196428571428571,
"grad_norm": 26.516355514526367,
"learning_rate": 2.379699248120301e-05,
"loss": 0.7859,
"step": 235
},
{
"epoch": 4.214285714285714,
"grad_norm": 14.370207786560059,
"learning_rate": 2.37593984962406e-05,
"loss": 0.8422,
"step": 236
},
{
"epoch": 4.232142857142857,
"grad_norm": 14.370207786560059,
"learning_rate": 2.3721804511278196e-05,
"loss": 0.7304,
"step": 237
},
{
"epoch": 4.25,
"grad_norm": 14.370207786560059,
"learning_rate": 2.368421052631579e-05,
"loss": 0.8508,
"step": 238
},
{
"epoch": 4.267857142857143,
"grad_norm": 14.370207786560059,
"learning_rate": 2.3646616541353385e-05,
"loss": 0.8441,
"step": 239
},
{
"epoch": 4.285714285714286,
"grad_norm": 16.65935707092285,
"learning_rate": 2.3609022556390977e-05,
"loss": 0.7378,
"step": 240
},
{
"epoch": 4.303571428571429,
"grad_norm": 16.65935707092285,
"learning_rate": 2.357142857142857e-05,
"loss": 0.73,
"step": 241
},
{
"epoch": 4.321428571428571,
"grad_norm": 16.65935707092285,
"learning_rate": 2.3533834586466166e-05,
"loss": 0.6546,
"step": 242
},
{
"epoch": 4.339285714285714,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7855,
"step": 243
},
{
"epoch": 4.357142857142857,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7172,
"step": 244
},
{
"epoch": 4.375,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7075,
"step": 245
},
{
"epoch": 4.392857142857143,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7603,
"step": 246
},
{
"epoch": 4.410714285714286,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7388,
"step": 247
},
{
"epoch": 4.428571428571429,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7009,
"step": 248
},
{
"epoch": 4.446428571428571,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7865,
"step": 249
},
{
"epoch": 4.464285714285714,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.764,
"step": 250
},
{
"epoch": 4.482142857142857,
"grad_norm": 16.65935707092285,
"learning_rate": 2.349624060150376e-05,
"loss": 0.7313,
"step": 251
},
{
"epoch": 4.5,
"grad_norm": 23.61573028564453,
"learning_rate": 2.3458646616541352e-05,
"loss": 0.7571,
"step": 252
},
{
"epoch": 4.517857142857143,
"grad_norm": 23.61573028564453,
"learning_rate": 2.3421052631578947e-05,
"loss": 0.7071,
"step": 253
},
{
"epoch": 4.535714285714286,
"grad_norm": 23.61573028564453,
"learning_rate": 2.338345864661654e-05,
"loss": 0.8513,
"step": 254
},
{
"epoch": 4.553571428571429,
"grad_norm": 23.61573028564453,
"learning_rate": 2.3345864661654136e-05,
"loss": 0.7141,
"step": 255
},
{
"epoch": 4.571428571428571,
"grad_norm": 27.054393768310547,
"learning_rate": 2.3308270676691728e-05,
"loss": 0.8198,
"step": 256
},
{
"epoch": 4.589285714285714,
"grad_norm": 27.054393768310547,
"learning_rate": 2.3270676691729323e-05,
"loss": 0.7735,
"step": 257
},
{
"epoch": 4.607142857142857,
"grad_norm": 27.054393768310547,
"learning_rate": 2.3233082706766917e-05,
"loss": 0.788,
"step": 258
},
{
"epoch": 4.625,
"grad_norm": 27.054393768310547,
"learning_rate": 2.3195488721804512e-05,
"loss": 0.8389,
"step": 259
},
{
"epoch": 4.642857142857143,
"grad_norm": 15.842607498168945,
"learning_rate": 2.3157894736842103e-05,
"loss": 0.685,
"step": 260
},
{
"epoch": 4.660714285714286,
"grad_norm": 15.842607498168945,
"learning_rate": 2.3120300751879698e-05,
"loss": 0.762,
"step": 261
},
{
"epoch": 4.678571428571429,
"grad_norm": 15.842607498168945,
"learning_rate": 2.3082706766917293e-05,
"loss": 0.7803,
"step": 262
},
{
"epoch": 4.696428571428571,
"grad_norm": 15.842607498168945,
"learning_rate": 2.3045112781954888e-05,
"loss": 0.7324,
"step": 263
},
{
"epoch": 4.714285714285714,
"grad_norm": 17.179346084594727,
"learning_rate": 2.300751879699248e-05,
"loss": 0.8053,
"step": 264
},
{
"epoch": 4.732142857142857,
"grad_norm": 17.179346084594727,
"learning_rate": 2.2969924812030074e-05,
"loss": 0.6846,
"step": 265
},
{
"epoch": 4.75,
"grad_norm": 17.179346084594727,
"learning_rate": 2.293233082706767e-05,
"loss": 0.8033,
"step": 266
},
{
"epoch": 4.767857142857143,
"grad_norm": 17.179346084594727,
"learning_rate": 2.2894736842105263e-05,
"loss": 0.8026,
"step": 267
},
{
"epoch": 4.785714285714286,
"grad_norm": 18.811100006103516,
"learning_rate": 2.2857142857142858e-05,
"loss": 0.8382,
"step": 268
},
{
"epoch": 4.803571428571429,
"grad_norm": 18.811100006103516,
"learning_rate": 2.2819548872180453e-05,
"loss": 0.7838,
"step": 269
},
{
"epoch": 4.821428571428571,
"grad_norm": 18.811100006103516,
"learning_rate": 2.2781954887218048e-05,
"loss": 0.8593,
"step": 270
},
{
"epoch": 4.839285714285714,
"grad_norm": 18.811100006103516,
"learning_rate": 2.2744360902255642e-05,
"loss": 0.7522,
"step": 271
},
{
"epoch": 4.857142857142857,
"grad_norm": 16.06136131286621,
"learning_rate": 2.2706766917293234e-05,
"loss": 0.7045,
"step": 272
},
{
"epoch": 4.875,
"grad_norm": 16.06136131286621,
"learning_rate": 2.266917293233083e-05,
"loss": 0.8126,
"step": 273
},
{
"epoch": 4.892857142857143,
"grad_norm": 16.06136131286621,
"learning_rate": 2.2631578947368423e-05,
"loss": 0.7216,
"step": 274
},
{
"epoch": 4.910714285714286,
"grad_norm": 16.06136131286621,
"learning_rate": 2.2593984962406018e-05,
"loss": 0.8261,
"step": 275
},
{
"epoch": 4.928571428571429,
"grad_norm": 15.82519817352295,
"learning_rate": 2.255639097744361e-05,
"loss": 0.7559,
"step": 276
},
{
"epoch": 4.946428571428571,
"grad_norm": 15.82519817352295,
"learning_rate": 2.2518796992481204e-05,
"loss": 0.727,
"step": 277
},
{
"epoch": 4.964285714285714,
"grad_norm": 15.82519817352295,
"learning_rate": 2.24812030075188e-05,
"loss": 0.765,
"step": 278
},
{
"epoch": 4.982142857142857,
"grad_norm": 15.82519817352295,
"learning_rate": 2.2443609022556394e-05,
"loss": 0.857,
"step": 279
},
{
"epoch": 5.0,
"grad_norm": 148.74029541015625,
"learning_rate": 2.2406015037593985e-05,
"loss": 0.6433,
"step": 280
},
{
"epoch": 5.017857142857143,
"grad_norm": 148.74029541015625,
"learning_rate": 2.236842105263158e-05,
"loss": 0.5615,
"step": 281
},
{
"epoch": 5.035714285714286,
"grad_norm": 148.74029541015625,
"learning_rate": 2.2330827067669174e-05,
"loss": 0.7077,
"step": 282
},
{
"epoch": 5.053571428571429,
"grad_norm": 148.74029541015625,
"learning_rate": 2.229323308270677e-05,
"loss": 0.8196,
"step": 283
},
{
"epoch": 5.071428571428571,
"grad_norm": 20.725858688354492,
"learning_rate": 2.225563909774436e-05,
"loss": 0.572,
"step": 284
},
{
"epoch": 5.089285714285714,
"grad_norm": 20.725858688354492,
"learning_rate": 2.2218045112781955e-05,
"loss": 0.6329,
"step": 285
},
{
"epoch": 5.107142857142857,
"grad_norm": 20.725858688354492,
"learning_rate": 2.218045112781955e-05,
"loss": 0.7605,
"step": 286
},
{
"epoch": 5.125,
"grad_norm": 20.725858688354492,
"learning_rate": 2.2142857142857145e-05,
"loss": 0.6296,
"step": 287
},
{
"epoch": 5.142857142857143,
"grad_norm": 64.4181900024414,
"learning_rate": 2.2105263157894736e-05,
"loss": 0.8021,
"step": 288
},
{
"epoch": 5.160714285714286,
"grad_norm": 64.4181900024414,
"learning_rate": 2.206766917293233e-05,
"loss": 0.64,
"step": 289
},
{
"epoch": 5.178571428571429,
"grad_norm": 64.4181900024414,
"learning_rate": 2.2030075187969926e-05,
"loss": 0.8771,
"step": 290
},
{
"epoch": 5.196428571428571,
"grad_norm": 64.4181900024414,
"learning_rate": 2.199248120300752e-05,
"loss": 0.6418,
"step": 291
},
{
"epoch": 5.214285714285714,
"grad_norm": 19.20701026916504,
"learning_rate": 2.1954887218045112e-05,
"loss": 0.6417,
"step": 292
},
{
"epoch": 5.232142857142857,
"grad_norm": 19.20701026916504,
"learning_rate": 2.1917293233082707e-05,
"loss": 0.6035,
"step": 293
},
{
"epoch": 5.25,
"grad_norm": 19.20701026916504,
"learning_rate": 2.18796992481203e-05,
"loss": 0.6975,
"step": 294
},
{
"epoch": 5.267857142857143,
"grad_norm": 19.20701026916504,
"learning_rate": 2.1842105263157896e-05,
"loss": 0.6474,
"step": 295
},
{
"epoch": 5.285714285714286,
"grad_norm": 13.516231536865234,
"learning_rate": 2.1804511278195487e-05,
"loss": 0.6649,
"step": 296
},
{
"epoch": 5.303571428571429,
"grad_norm": 13.516231536865234,
"learning_rate": 2.1766917293233082e-05,
"loss": 0.6144,
"step": 297
},
{
"epoch": 5.321428571428571,
"grad_norm": 13.516231536865234,
"learning_rate": 2.1729323308270677e-05,
"loss": 0.5895,
"step": 298
},
{
"epoch": 5.339285714285714,
"grad_norm": 13.516231536865234,
"learning_rate": 2.169172932330827e-05,
"loss": 0.6227,
"step": 299
},
{
"epoch": 5.357142857142857,
"grad_norm": 13.004302978515625,
"learning_rate": 2.1654135338345863e-05,
"loss": 0.6965,
"step": 300
},
{
"epoch": 5.375,
"grad_norm": 13.004302978515625,
"learning_rate": 2.1616541353383458e-05,
"loss": 0.6849,
"step": 301
},
{
"epoch": 5.392857142857143,
"grad_norm": 13.004302978515625,
"learning_rate": 2.1578947368421053e-05,
"loss": 0.5881,
"step": 302
},
{
"epoch": 5.410714285714286,
"grad_norm": 13.004302978515625,
"learning_rate": 2.1541353383458647e-05,
"loss": 0.5755,
"step": 303
},
{
"epoch": 5.428571428571429,
"grad_norm": 13.709508895874023,
"learning_rate": 2.150375939849624e-05,
"loss": 0.5752,
"step": 304
},
{
"epoch": 5.446428571428571,
"grad_norm": 13.709508895874023,
"learning_rate": 2.1466165413533833e-05,
"loss": 0.6799,
"step": 305
},
{
"epoch": 5.464285714285714,
"grad_norm": 13.709508895874023,
"learning_rate": 2.1428571428571428e-05,
"loss": 0.561,
"step": 306
},
{
"epoch": 5.482142857142857,
"grad_norm": 13.709508895874023,
"learning_rate": 2.1390977443609023e-05,
"loss": 0.6116,
"step": 307
},
{
"epoch": 5.5,
"grad_norm": 15.477508544921875,
"learning_rate": 2.1353383458646614e-05,
"loss": 0.6184,
"step": 308
},
{
"epoch": 5.517857142857143,
"grad_norm": 15.477508544921875,
"learning_rate": 2.1315789473684212e-05,
"loss": 0.6287,
"step": 309
},
{
"epoch": 5.535714285714286,
"grad_norm": 15.477508544921875,
"learning_rate": 2.1278195488721807e-05,
"loss": 0.571,
"step": 310
},
{
"epoch": 5.553571428571429,
"grad_norm": 15.477508544921875,
"learning_rate": 2.1240601503759402e-05,
"loss": 0.5509,
"step": 311
},
{
"epoch": 5.571428571428571,
"grad_norm": 19.958415985107422,
"learning_rate": 2.1203007518796993e-05,
"loss": 0.7161,
"step": 312
},
{
"epoch": 5.589285714285714,
"grad_norm": 19.958415985107422,
"learning_rate": 2.1165413533834588e-05,
"loss": 0.5362,
"step": 313
},
{
"epoch": 5.607142857142857,
"grad_norm": 19.958415985107422,
"learning_rate": 2.1127819548872183e-05,
"loss": 0.5581,
"step": 314
},
{
"epoch": 5.625,
"grad_norm": 19.958415985107422,
"learning_rate": 2.1090225563909778e-05,
"loss": 0.6408,
"step": 315
},
{
"epoch": 5.642857142857143,
"grad_norm": 26.75920295715332,
"learning_rate": 2.105263157894737e-05,
"loss": 0.7009,
"step": 316
},
{
"epoch": 5.660714285714286,
"grad_norm": 26.75920295715332,
"learning_rate": 2.1015037593984964e-05,
"loss": 0.6003,
"step": 317
},
{
"epoch": 5.678571428571429,
"grad_norm": 26.75920295715332,
"learning_rate": 2.097744360902256e-05,
"loss": 0.6715,
"step": 318
},
{
"epoch": 5.696428571428571,
"grad_norm": 26.75920295715332,
"learning_rate": 2.0939849624060153e-05,
"loss": 0.6436,
"step": 319
},
{
"epoch": 5.714285714285714,
"grad_norm": 131.8643035888672,
"learning_rate": 2.0902255639097745e-05,
"loss": 0.5557,
"step": 320
},
{
"epoch": 5.732142857142857,
"grad_norm": 131.8643035888672,
"learning_rate": 2.086466165413534e-05,
"loss": 0.6455,
"step": 321
},
{
"epoch": 5.75,
"grad_norm": 131.8643035888672,
"learning_rate": 2.0827067669172934e-05,
"loss": 0.5079,
"step": 322
},
{
"epoch": 5.767857142857143,
"grad_norm": 131.8643035888672,
"learning_rate": 2.078947368421053e-05,
"loss": 0.7021,
"step": 323
},
{
"epoch": 5.785714285714286,
"grad_norm": 65.64579010009766,
"learning_rate": 2.075187969924812e-05,
"loss": 0.5972,
"step": 324
},
{
"epoch": 5.803571428571429,
"grad_norm": 65.64579010009766,
"learning_rate": 2.0714285714285715e-05,
"loss": 0.5571,
"step": 325
},
{
"epoch": 5.821428571428571,
"grad_norm": 65.64579010009766,
"learning_rate": 2.067669172932331e-05,
"loss": 0.637,
"step": 326
},
{
"epoch": 5.839285714285714,
"grad_norm": 65.64579010009766,
"learning_rate": 2.0639097744360904e-05,
"loss": 0.5418,
"step": 327
},
{
"epoch": 5.857142857142857,
"grad_norm": 16.304061889648438,
"learning_rate": 2.0601503759398496e-05,
"loss": 0.6792,
"step": 328
},
{
"epoch": 5.875,
"grad_norm": 16.304061889648438,
"learning_rate": 2.056390977443609e-05,
"loss": 0.6426,
"step": 329
},
{
"epoch": 5.892857142857143,
"grad_norm": 16.304061889648438,
"learning_rate": 2.0526315789473685e-05,
"loss": 0.6114,
"step": 330
},
{
"epoch": 5.910714285714286,
"grad_norm": 16.304061889648438,
"learning_rate": 2.048872180451128e-05,
"loss": 0.6818,
"step": 331
},
{
"epoch": 5.928571428571429,
"grad_norm": 19.96442222595215,
"learning_rate": 2.045112781954887e-05,
"loss": 0.751,
"step": 332
},
{
"epoch": 5.946428571428571,
"grad_norm": 19.96442222595215,
"learning_rate": 2.0413533834586466e-05,
"loss": 0.5235,
"step": 333
},
{
"epoch": 5.964285714285714,
"grad_norm": 19.96442222595215,
"learning_rate": 2.037593984962406e-05,
"loss": 0.5584,
"step": 334
},
{
"epoch": 5.982142857142857,
"grad_norm": 19.96442222595215,
"learning_rate": 2.0338345864661656e-05,
"loss": 0.6518,
"step": 335
},
{
"epoch": 6.0,
"grad_norm": 13.476043701171875,
"learning_rate": 2.0300751879699247e-05,
"loss": 0.528,
"step": 336
},
{
"epoch": 6.017857142857143,
"grad_norm": 13.476043701171875,
"learning_rate": 2.0263157894736842e-05,
"loss": 0.4961,
"step": 337
},
{
"epoch": 6.035714285714286,
"grad_norm": 13.476043701171875,
"learning_rate": 2.0225563909774437e-05,
"loss": 0.5934,
"step": 338
},
{
"epoch": 6.053571428571429,
"grad_norm": 13.476043701171875,
"learning_rate": 2.018796992481203e-05,
"loss": 0.5111,
"step": 339
},
{
"epoch": 6.071428571428571,
"grad_norm": 13.2622652053833,
"learning_rate": 2.0150375939849623e-05,
"loss": 0.4773,
"step": 340
},
{
"epoch": 6.089285714285714,
"grad_norm": 13.2622652053833,
"learning_rate": 2.0112781954887217e-05,
"loss": 0.4694,
"step": 341
},
{
"epoch": 6.107142857142857,
"grad_norm": 13.2622652053833,
"learning_rate": 2.0075187969924812e-05,
"loss": 0.4366,
"step": 342
},
{
"epoch": 6.125,
"grad_norm": 13.2622652053833,
"learning_rate": 2.0037593984962407e-05,
"loss": 0.4481,
"step": 343
},
{
"epoch": 6.142857142857143,
"grad_norm": 11.744352340698242,
"learning_rate": 1.9999999999999998e-05,
"loss": 0.4071,
"step": 344
},
{
"epoch": 6.160714285714286,
"grad_norm": 11.744352340698242,
"learning_rate": 1.9962406015037593e-05,
"loss": 0.47,
"step": 345
},
{
"epoch": 6.178571428571429,
"grad_norm": 11.744352340698242,
"learning_rate": 1.9924812030075188e-05,
"loss": 0.4522,
"step": 346
},
{
"epoch": 6.196428571428571,
"grad_norm": 11.744352340698242,
"learning_rate": 1.9887218045112783e-05,
"loss": 0.4716,
"step": 347
},
{
"epoch": 6.214285714285714,
"grad_norm": 12.050751686096191,
"learning_rate": 1.9849624060150374e-05,
"loss": 0.3732,
"step": 348
},
{
"epoch": 6.232142857142857,
"grad_norm": 12.050751686096191,
"learning_rate": 1.981203007518797e-05,
"loss": 0.5253,
"step": 349
},
{
"epoch": 6.25,
"grad_norm": 12.050751686096191,
"learning_rate": 1.9774436090225563e-05,
"loss": 0.5366,
"step": 350
},
{
"epoch": 6.267857142857143,
"grad_norm": 12.050751686096191,
"learning_rate": 1.9736842105263158e-05,
"loss": 0.4863,
"step": 351
},
{
"epoch": 6.285714285714286,
"grad_norm": 15.379858016967773,
"learning_rate": 1.9699248120300753e-05,
"loss": 0.4549,
"step": 352
},
{
"epoch": 6.303571428571429,
"grad_norm": 15.379858016967773,
"learning_rate": 1.9661654135338348e-05,
"loss": 0.4957,
"step": 353
},
{
"epoch": 6.321428571428571,
"grad_norm": 15.379858016967773,
"learning_rate": 1.9624060150375942e-05,
"loss": 0.4436,
"step": 354
},
{
"epoch": 6.339285714285714,
"grad_norm": 15.379858016967773,
"learning_rate": 1.9586466165413537e-05,
"loss": 0.4853,
"step": 355
},
{
"epoch": 6.357142857142857,
"grad_norm": 13.114079475402832,
"learning_rate": 1.954887218045113e-05,
"loss": 0.4858,
"step": 356
},
{
"epoch": 6.375,
"grad_norm": 13.114079475402832,
"learning_rate": 1.9511278195488723e-05,
"loss": 0.492,
"step": 357
},
{
"epoch": 6.392857142857143,
"grad_norm": 13.114079475402832,
"learning_rate": 1.9473684210526318e-05,
"loss": 0.41,
"step": 358
},
{
"epoch": 6.410714285714286,
"grad_norm": 13.114079475402832,
"learning_rate": 1.9436090225563913e-05,
"loss": 0.4983,
"step": 359
},
{
"epoch": 6.428571428571429,
"grad_norm": 88.1438980102539,
"learning_rate": 1.9398496240601504e-05,
"loss": 0.4149,
"step": 360
},
{
"epoch": 6.446428571428571,
"grad_norm": 88.1438980102539,
"learning_rate": 1.93609022556391e-05,
"loss": 0.4906,
"step": 361
},
{
"epoch": 6.464285714285714,
"grad_norm": 88.1438980102539,
"learning_rate": 1.9323308270676694e-05,
"loss": 0.5251,
"step": 362
},
{
"epoch": 6.482142857142857,
"grad_norm": 88.1438980102539,
"learning_rate": 1.928571428571429e-05,
"loss": 0.5786,
"step": 363
},
{
"epoch": 6.5,
"grad_norm": 21.487791061401367,
"learning_rate": 1.924812030075188e-05,
"loss": 0.4006,
"step": 364
},
{
"epoch": 6.517857142857143,
"grad_norm": 21.487791061401367,
"learning_rate": 1.9210526315789474e-05,
"loss": 0.4504,
"step": 365
},
{
"epoch": 6.535714285714286,
"grad_norm": 21.487791061401367,
"learning_rate": 1.917293233082707e-05,
"loss": 0.4529,
"step": 366
},
{
"epoch": 6.553571428571429,
"grad_norm": 21.487791061401367,
"learning_rate": 1.9135338345864664e-05,
"loss": 0.5762,
"step": 367
},
{
"epoch": 6.571428571428571,
"grad_norm": 17.192222595214844,
"learning_rate": 1.9097744360902255e-05,
"loss": 0.5141,
"step": 368
},
{
"epoch": 6.589285714285714,
"grad_norm": 17.192222595214844,
"learning_rate": 1.906015037593985e-05,
"loss": 0.4267,
"step": 369
},
{
"epoch": 6.607142857142857,
"grad_norm": 17.192222595214844,
"learning_rate": 1.9022556390977445e-05,
"loss": 0.5604,
"step": 370
},
{
"epoch": 6.625,
"grad_norm": 17.192222595214844,
"learning_rate": 1.898496240601504e-05,
"loss": 0.4435,
"step": 371
},
{
"epoch": 6.642857142857143,
"grad_norm": 15.903327941894531,
"learning_rate": 1.894736842105263e-05,
"loss": 0.4905,
"step": 372
},
{
"epoch": 6.660714285714286,
"grad_norm": 15.903327941894531,
"learning_rate": 1.8909774436090226e-05,
"loss": 0.541,
"step": 373
},
{
"epoch": 6.678571428571429,
"grad_norm": 15.903327941894531,
"learning_rate": 1.887218045112782e-05,
"loss": 0.5425,
"step": 374
},
{
"epoch": 6.696428571428571,
"grad_norm": 15.903327941894531,
"learning_rate": 1.8834586466165415e-05,
"loss": 0.6469,
"step": 375
},
{
"epoch": 6.714285714285714,
"grad_norm": 105.3304672241211,
"learning_rate": 1.8796992481203007e-05,
"loss": 0.5695,
"step": 376
},
{
"epoch": 6.732142857142857,
"grad_norm": 105.3304672241211,
"learning_rate": 1.87593984962406e-05,
"loss": 0.5143,
"step": 377
},
{
"epoch": 6.75,
"grad_norm": 105.3304672241211,
"learning_rate": 1.8721804511278196e-05,
"loss": 0.4614,
"step": 378
},
{
"epoch": 6.767857142857143,
"grad_norm": 105.3304672241211,
"learning_rate": 1.868421052631579e-05,
"loss": 0.5282,
"step": 379
},
{
"epoch": 6.785714285714286,
"grad_norm": 14.260865211486816,
"learning_rate": 1.8646616541353382e-05,
"loss": 0.4917,
"step": 380
},
{
"epoch": 6.803571428571429,
"grad_norm": 14.260865211486816,
"learning_rate": 1.8609022556390977e-05,
"loss": 0.4277,
"step": 381
},
{
"epoch": 6.821428571428571,
"grad_norm": 14.260865211486816,
"learning_rate": 1.8571428571428572e-05,
"loss": 0.512,
"step": 382
},
{
"epoch": 6.839285714285714,
"grad_norm": 14.260865211486816,
"learning_rate": 1.8533834586466166e-05,
"loss": 0.4897,
"step": 383
},
{
"epoch": 6.857142857142857,
"grad_norm": 17.528396606445312,
"learning_rate": 1.8496240601503758e-05,
"loss": 0.5596,
"step": 384
},
{
"epoch": 6.875,
"grad_norm": 17.528396606445312,
"learning_rate": 1.8458646616541353e-05,
"loss": 0.4921,
"step": 385
},
{
"epoch": 6.892857142857143,
"grad_norm": 17.528396606445312,
"learning_rate": 1.8421052631578947e-05,
"loss": 0.491,
"step": 386
},
{
"epoch": 6.910714285714286,
"grad_norm": 17.528396606445312,
"learning_rate": 1.8383458646616542e-05,
"loss": 0.5002,
"step": 387
},
{
"epoch": 6.928571428571429,
"grad_norm": 14.27133846282959,
"learning_rate": 1.8345864661654133e-05,
"loss": 0.6862,
"step": 388
},
{
"epoch": 6.946428571428571,
"grad_norm": 14.27133846282959,
"learning_rate": 1.8308270676691728e-05,
"loss": 0.6156,
"step": 389
},
{
"epoch": 6.964285714285714,
"grad_norm": 14.27133846282959,
"learning_rate": 1.8270676691729323e-05,
"loss": 0.4589,
"step": 390
},
{
"epoch": 6.982142857142857,
"grad_norm": 14.27133846282959,
"learning_rate": 1.8233082706766918e-05,
"loss": 0.5865,
"step": 391
},
{
"epoch": 7.0,
"grad_norm": 23.04795265197754,
"learning_rate": 1.819548872180451e-05,
"loss": 0.5523,
"step": 392
},
{
"epoch": 7.017857142857143,
"grad_norm": 23.04795265197754,
"learning_rate": 1.8157894736842107e-05,
"loss": 0.4733,
"step": 393
},
{
"epoch": 7.035714285714286,
"grad_norm": 23.04795265197754,
"learning_rate": 1.8120300751879702e-05,
"loss": 0.4734,
"step": 394
},
{
"epoch": 7.053571428571429,
"grad_norm": 23.04795265197754,
"learning_rate": 1.8082706766917297e-05,
"loss": 0.4098,
"step": 395
},
{
"epoch": 7.071428571428571,
"grad_norm": 16.342947006225586,
"learning_rate": 1.8045112781954888e-05,
"loss": 0.4926,
"step": 396
},
{
"epoch": 7.089285714285714,
"grad_norm": 16.342947006225586,
"learning_rate": 1.8007518796992483e-05,
"loss": 0.4574,
"step": 397
},
{
"epoch": 7.107142857142857,
"grad_norm": 16.342947006225586,
"learning_rate": 1.7969924812030078e-05,
"loss": 0.4756,
"step": 398
},
{
"epoch": 7.125,
"grad_norm": 16.342947006225586,
"learning_rate": 1.7932330827067672e-05,
"loss": 0.3955,
"step": 399
},
{
"epoch": 7.142857142857143,
"grad_norm": 13.775762557983398,
"learning_rate": 1.7894736842105264e-05,
"loss": 0.4146,
"step": 400
},
{
"epoch": 7.160714285714286,
"grad_norm": 13.775762557983398,
"learning_rate": 1.785714285714286e-05,
"loss": 0.4167,
"step": 401
},
{
"epoch": 7.178571428571429,
"grad_norm": 13.775762557983398,
"learning_rate": 1.7819548872180453e-05,
"loss": 0.3745,
"step": 402
},
{
"epoch": 7.196428571428571,
"grad_norm": 13.775762557983398,
"learning_rate": 1.7781954887218048e-05,
"loss": 0.3773,
"step": 403
},
{
"epoch": 7.214285714285714,
"grad_norm": 12.232085227966309,
"learning_rate": 1.774436090225564e-05,
"loss": 0.3974,
"step": 404
},
{
"epoch": 7.232142857142857,
"grad_norm": 12.232085227966309,
"learning_rate": 1.7706766917293234e-05,
"loss": 0.3684,
"step": 405
},
{
"epoch": 7.25,
"grad_norm": 12.232085227966309,
"learning_rate": 1.766917293233083e-05,
"loss": 0.3348,
"step": 406
},
{
"epoch": 7.267857142857143,
"grad_norm": 12.232085227966309,
"learning_rate": 1.7631578947368424e-05,
"loss": 0.4156,
"step": 407
},
{
"epoch": 7.285714285714286,
"grad_norm": 13.244461059570312,
"learning_rate": 1.7593984962406015e-05,
"loss": 0.3479,
"step": 408
},
{
"epoch": 7.303571428571429,
"grad_norm": 13.244461059570312,
"learning_rate": 1.755639097744361e-05,
"loss": 0.3941,
"step": 409
},
{
"epoch": 7.321428571428571,
"grad_norm": 13.244461059570312,
"learning_rate": 1.7518796992481204e-05,
"loss": 0.4444,
"step": 410
},
{
"epoch": 7.339285714285714,
"grad_norm": 13.244461059570312,
"learning_rate": 1.74812030075188e-05,
"loss": 0.4387,
"step": 411
},
{
"epoch": 7.357142857142857,
"grad_norm": 20.053211212158203,
"learning_rate": 1.744360902255639e-05,
"loss": 0.3831,
"step": 412
},
{
"epoch": 7.375,
"grad_norm": 20.053211212158203,
"learning_rate": 1.7406015037593985e-05,
"loss": 0.3901,
"step": 413
},
{
"epoch": 7.392857142857143,
"grad_norm": 20.053211212158203,
"learning_rate": 1.736842105263158e-05,
"loss": 0.4138,
"step": 414
},
{
"epoch": 7.410714285714286,
"grad_norm": 20.053211212158203,
"learning_rate": 1.7330827067669175e-05,
"loss": 0.3726,
"step": 415
},
{
"epoch": 7.428571428571429,
"grad_norm": 16.389062881469727,
"learning_rate": 1.7293233082706766e-05,
"loss": 0.3764,
"step": 416
},
{
"epoch": 7.446428571428571,
"grad_norm": 16.389062881469727,
"learning_rate": 1.725563909774436e-05,
"loss": 0.3651,
"step": 417
},
{
"epoch": 7.464285714285714,
"grad_norm": 16.389062881469727,
"learning_rate": 1.7218045112781956e-05,
"loss": 0.3952,
"step": 418
},
{
"epoch": 7.482142857142857,
"grad_norm": 16.389062881469727,
"learning_rate": 1.718045112781955e-05,
"loss": 0.4158,
"step": 419
},
{
"epoch": 7.5,
"grad_norm": 20.468734741210938,
"learning_rate": 1.7142857142857142e-05,
"loss": 0.3389,
"step": 420
},
{
"epoch": 7.517857142857143,
"grad_norm": 20.468734741210938,
"learning_rate": 1.7105263157894737e-05,
"loss": 0.3317,
"step": 421
},
{
"epoch": 7.535714285714286,
"grad_norm": 20.468734741210938,
"learning_rate": 1.706766917293233e-05,
"loss": 0.4567,
"step": 422
},
{
"epoch": 7.553571428571429,
"grad_norm": 20.468734741210938,
"learning_rate": 1.7030075187969926e-05,
"loss": 0.4025,
"step": 423
},
{
"epoch": 7.571428571428571,
"grad_norm": 15.636091232299805,
"learning_rate": 1.6992481203007517e-05,
"loss": 0.3463,
"step": 424
},
{
"epoch": 7.589285714285714,
"grad_norm": 15.636091232299805,
"learning_rate": 1.6954887218045112e-05,
"loss": 0.3588,
"step": 425
},
{
"epoch": 7.607142857142857,
"grad_norm": 15.636091232299805,
"learning_rate": 1.6917293233082707e-05,
"loss": 0.3925,
"step": 426
},
{
"epoch": 7.625,
"grad_norm": 15.636091232299805,
"learning_rate": 1.6879699248120302e-05,
"loss": 0.4384,
"step": 427
},
{
"epoch": 7.642857142857143,
"grad_norm": 15.85193920135498,
"learning_rate": 1.6842105263157893e-05,
"loss": 0.4041,
"step": 428
},
{
"epoch": 7.660714285714286,
"grad_norm": 15.85193920135498,
"learning_rate": 1.6804511278195488e-05,
"loss": 0.4091,
"step": 429
},
{
"epoch": 7.678571428571429,
"grad_norm": 15.85193920135498,
"learning_rate": 1.6766917293233083e-05,
"loss": 0.4424,
"step": 430
},
{
"epoch": 7.696428571428571,
"grad_norm": 15.85193920135498,
"learning_rate": 1.6729323308270677e-05,
"loss": 0.4385,
"step": 431
},
{
"epoch": 7.714285714285714,
"grad_norm": 15.357510566711426,
"learning_rate": 1.669172932330827e-05,
"loss": 0.4408,
"step": 432
},
{
"epoch": 7.732142857142857,
"grad_norm": 15.357510566711426,
"learning_rate": 1.6654135338345863e-05,
"loss": 0.343,
"step": 433
},
{
"epoch": 7.75,
"grad_norm": 15.357510566711426,
"learning_rate": 1.6616541353383458e-05,
"loss": 0.4187,
"step": 434
},
{
"epoch": 7.767857142857143,
"grad_norm": 15.357510566711426,
"learning_rate": 1.6578947368421053e-05,
"loss": 0.3525,
"step": 435
},
{
"epoch": 7.785714285714286,
"grad_norm": 14.422098159790039,
"learning_rate": 1.6541353383458648e-05,
"loss": 0.4261,
"step": 436
},
{
"epoch": 7.803571428571429,
"grad_norm": 14.422098159790039,
"learning_rate": 1.6503759398496242e-05,
"loss": 0.4448,
"step": 437
},
{
"epoch": 7.821428571428571,
"grad_norm": 14.422098159790039,
"learning_rate": 1.6466165413533837e-05,
"loss": 0.4053,
"step": 438
},
{
"epoch": 7.839285714285714,
"grad_norm": 14.422098159790039,
"learning_rate": 1.6428571428571432e-05,
"loss": 0.4036,
"step": 439
},
{
"epoch": 7.857142857142857,
"grad_norm": 17.375715255737305,
"learning_rate": 1.6390977443609023e-05,
"loss": 0.3913,
"step": 440
},
{
"epoch": 7.875,
"grad_norm": 17.375715255737305,
"learning_rate": 1.6353383458646618e-05,
"loss": 0.4118,
"step": 441
},
{
"epoch": 7.892857142857143,
"grad_norm": 17.375715255737305,
"learning_rate": 1.6315789473684213e-05,
"loss": 0.3686,
"step": 442
},
{
"epoch": 7.910714285714286,
"grad_norm": 17.375715255737305,
"learning_rate": 1.6278195488721808e-05,
"loss": 0.362,
"step": 443
},
{
"epoch": 7.928571428571429,
"grad_norm": 14.241915702819824,
"learning_rate": 1.62406015037594e-05,
"loss": 0.3846,
"step": 444
},
{
"epoch": 7.946428571428571,
"grad_norm": 14.241915702819824,
"learning_rate": 1.6203007518796994e-05,
"loss": 0.364,
"step": 445
},
{
"epoch": 7.964285714285714,
"grad_norm": 14.241915702819824,
"learning_rate": 1.616541353383459e-05,
"loss": 0.3958,
"step": 446
},
{
"epoch": 7.982142857142857,
"grad_norm": 14.241915702819824,
"learning_rate": 1.6127819548872183e-05,
"loss": 0.418,
"step": 447
},
{
"epoch": 8.0,
"grad_norm": 74.3487548828125,
"learning_rate": 1.6090225563909775e-05,
"loss": 0.4073,
"step": 448
},
{
"epoch": 8.017857142857142,
"grad_norm": 74.3487548828125,
"learning_rate": 1.605263157894737e-05,
"loss": 0.3958,
"step": 449
},
{
"epoch": 8.035714285714286,
"grad_norm": 74.3487548828125,
"learning_rate": 1.6015037593984964e-05,
"loss": 0.3208,
"step": 450
},
{
"epoch": 8.053571428571429,
"grad_norm": 74.3487548828125,
"learning_rate": 1.597744360902256e-05,
"loss": 0.3101,
"step": 451
},
{
"epoch": 8.071428571428571,
"grad_norm": 14.820595741271973,
"learning_rate": 1.593984962406015e-05,
"loss": 0.3648,
"step": 452
},
{
"epoch": 8.089285714285714,
"grad_norm": 14.820595741271973,
"learning_rate": 1.5902255639097745e-05,
"loss": 0.3431,
"step": 453
},
{
"epoch": 8.107142857142858,
"grad_norm": 14.820595741271973,
"learning_rate": 1.586466165413534e-05,
"loss": 0.3969,
"step": 454
},
{
"epoch": 8.125,
"grad_norm": 14.820595741271973,
"learning_rate": 1.5827067669172934e-05,
"loss": 0.3015,
"step": 455
},
{
"epoch": 8.142857142857142,
"grad_norm": 14.003570556640625,
"learning_rate": 1.5789473684210526e-05,
"loss": 0.3364,
"step": 456
},
{
"epoch": 8.160714285714286,
"grad_norm": 14.003570556640625,
"learning_rate": 1.575187969924812e-05,
"loss": 0.3881,
"step": 457
},
{
"epoch": 8.178571428571429,
"grad_norm": 14.003570556640625,
"learning_rate": 1.5714285714285715e-05,
"loss": 0.3818,
"step": 458
},
{
"epoch": 8.196428571428571,
"grad_norm": 14.003570556640625,
"learning_rate": 1.567669172932331e-05,
"loss": 0.3441,
"step": 459
},
{
"epoch": 8.214285714285714,
"grad_norm": 14.003570556640625,
"learning_rate": 1.567669172932331e-05,
"loss": 0.3389,
"step": 460
},
{
"epoch": 8.232142857142858,
"grad_norm": 14.003570556640625,
"learning_rate": 1.567669172932331e-05,
"loss": 0.367,
"step": 461
},
{
"epoch": 8.25,
"grad_norm": 14.003570556640625,
"learning_rate": 1.567669172932331e-05,
"loss": 0.3949,
"step": 462
},
{
"epoch": 8.267857142857142,
"grad_norm": 14.003570556640625,
"learning_rate": 1.567669172932331e-05,
"loss": 0.3828,
"step": 463
},
{
"epoch": 8.285714285714286,
"grad_norm": 34.45131301879883,
"learning_rate": 1.56390977443609e-05,
"loss": 0.3212,
"step": 464
},
{
"epoch": 8.303571428571429,
"grad_norm": 34.45131301879883,
"learning_rate": 1.5601503759398496e-05,
"loss": 0.315,
"step": 465
},
{
"epoch": 8.321428571428571,
"grad_norm": 34.45131301879883,
"learning_rate": 1.556390977443609e-05,
"loss": 0.3802,
"step": 466
},
{
"epoch": 8.339285714285714,
"grad_norm": 34.45131301879883,
"learning_rate": 1.5526315789473686e-05,
"loss": 0.4092,
"step": 467
},
{
"epoch": 8.357142857142858,
"grad_norm": 84.08926391601562,
"learning_rate": 1.5488721804511277e-05,
"loss": 0.391,
"step": 468
},
{
"epoch": 8.375,
"grad_norm": 84.08926391601562,
"learning_rate": 1.5451127819548872e-05,
"loss": 0.3006,
"step": 469
},
{
"epoch": 8.392857142857142,
"grad_norm": 84.08926391601562,
"learning_rate": 1.5413533834586467e-05,
"loss": 0.3889,
"step": 470
},
{
"epoch": 8.410714285714286,
"grad_norm": 84.08926391601562,
"learning_rate": 1.537593984962406e-05,
"loss": 0.4272,
"step": 471
},
{
"epoch": 8.428571428571429,
"grad_norm": 20.935237884521484,
"learning_rate": 1.5338345864661653e-05,
"loss": 0.3553,
"step": 472
},
{
"epoch": 8.446428571428571,
"grad_norm": 20.935237884521484,
"learning_rate": 1.5300751879699247e-05,
"loss": 0.3851,
"step": 473
},
{
"epoch": 8.464285714285714,
"grad_norm": 20.935237884521484,
"learning_rate": 1.5263157894736842e-05,
"loss": 0.3777,
"step": 474
},
{
"epoch": 8.482142857142858,
"grad_norm": 20.935237884521484,
"learning_rate": 1.5225563909774439e-05,
"loss": 0.4261,
"step": 475
},
{
"epoch": 8.5,
"grad_norm": 144.63853454589844,
"learning_rate": 1.518796992481203e-05,
"loss": 0.4572,
"step": 476
},
{
"epoch": 8.517857142857142,
"grad_norm": 144.63853454589844,
"learning_rate": 1.5150375939849625e-05,
"loss": 0.37,
"step": 477
},
{
"epoch": 8.535714285714286,
"grad_norm": 144.63853454589844,
"learning_rate": 1.511278195488722e-05,
"loss": 0.37,
"step": 478
},
{
"epoch": 8.553571428571429,
"grad_norm": 144.63853454589844,
"learning_rate": 1.5075187969924814e-05,
"loss": 0.4446,
"step": 479
},
{
"epoch": 8.571428571428571,
"grad_norm": 20.895305633544922,
"learning_rate": 1.5037593984962406e-05,
"loss": 0.3549,
"step": 480
},
{
"epoch": 8.589285714285714,
"grad_norm": 20.895305633544922,
"learning_rate": 1.5e-05,
"loss": 0.3891,
"step": 481
},
{
"epoch": 8.607142857142858,
"grad_norm": 20.895305633544922,
"learning_rate": 1.4962406015037593e-05,
"loss": 0.4171,
"step": 482
},
{
"epoch": 8.625,
"grad_norm": 20.895305633544922,
"learning_rate": 1.4924812030075188e-05,
"loss": 0.4291,
"step": 483
},
{
"epoch": 8.642857142857142,
"grad_norm": 78.05601501464844,
"learning_rate": 1.4887218045112781e-05,
"loss": 0.4723,
"step": 484
},
{
"epoch": 8.660714285714286,
"grad_norm": 78.05601501464844,
"learning_rate": 1.4849624060150376e-05,
"loss": 0.3286,
"step": 485
},
{
"epoch": 8.678571428571429,
"grad_norm": 78.05601501464844,
"learning_rate": 1.4812030075187969e-05,
"loss": 0.3919,
"step": 486
},
{
"epoch": 8.696428571428571,
"grad_norm": 78.05601501464844,
"learning_rate": 1.4774436090225564e-05,
"loss": 0.4245,
"step": 487
},
{
"epoch": 8.714285714285714,
"grad_norm": 110.17274475097656,
"learning_rate": 1.4736842105263157e-05,
"loss": 0.3833,
"step": 488
},
{
"epoch": 8.732142857142858,
"grad_norm": 110.17274475097656,
"learning_rate": 1.4699248120300752e-05,
"loss": 0.3771,
"step": 489
},
{
"epoch": 8.75,
"grad_norm": 110.17274475097656,
"learning_rate": 1.4661654135338345e-05,
"loss": 0.2863,
"step": 490
},
{
"epoch": 8.767857142857142,
"grad_norm": 110.17274475097656,
"learning_rate": 1.462406015037594e-05,
"loss": 0.3917,
"step": 491
},
{
"epoch": 8.785714285714286,
"grad_norm": 28.010690689086914,
"learning_rate": 1.4586466165413534e-05,
"loss": 0.3884,
"step": 492
},
{
"epoch": 8.803571428571429,
"grad_norm": 28.010690689086914,
"learning_rate": 1.4548872180451129e-05,
"loss": 0.4437,
"step": 493
},
{
"epoch": 8.821428571428571,
"grad_norm": 28.010690689086914,
"learning_rate": 1.4511278195488722e-05,
"loss": 0.408,
"step": 494
},
{
"epoch": 8.839285714285714,
"grad_norm": 28.010690689086914,
"learning_rate": 1.4473684210526317e-05,
"loss": 0.3535,
"step": 495
},
{
"epoch": 8.857142857142858,
"grad_norm": 18.033123016357422,
"learning_rate": 1.443609022556391e-05,
"loss": 0.4159,
"step": 496
},
{
"epoch": 8.875,
"grad_norm": 18.033123016357422,
"learning_rate": 1.4398496240601505e-05,
"loss": 0.346,
"step": 497
},
{
"epoch": 8.892857142857142,
"grad_norm": 18.033123016357422,
"learning_rate": 1.4360902255639098e-05,
"loss": 0.3646,
"step": 498
},
{
"epoch": 8.910714285714286,
"grad_norm": 18.033123016357422,
"learning_rate": 1.4323308270676692e-05,
"loss": 0.2897,
"step": 499
},
{
"epoch": 8.928571428571429,
"grad_norm": 16.059865951538086,
"learning_rate": 1.4285714285714285e-05,
"loss": 0.363,
"step": 500
},
{
"epoch": 8.946428571428571,
"grad_norm": 16.059865951538086,
"learning_rate": 1.424812030075188e-05,
"loss": 0.4248,
"step": 501
},
{
"epoch": 8.964285714285714,
"grad_norm": 16.059865951538086,
"learning_rate": 1.4210526315789473e-05,
"loss": 0.4242,
"step": 502
},
{
"epoch": 8.982142857142858,
"grad_norm": 16.059865951538086,
"learning_rate": 1.4172932330827068e-05,
"loss": 0.421,
"step": 503
},
{
"epoch": 9.0,
"grad_norm": 16.365734100341797,
"learning_rate": 1.4135338345864661e-05,
"loss": 0.3539,
"step": 504
},
{
"epoch": 9.017857142857142,
"grad_norm": 16.365734100341797,
"learning_rate": 1.4097744360902256e-05,
"loss": 0.3184,
"step": 505
},
{
"epoch": 9.035714285714286,
"grad_norm": 16.365734100341797,
"learning_rate": 1.4060150375939849e-05,
"loss": 0.2725,
"step": 506
},
{
"epoch": 9.053571428571429,
"grad_norm": 16.365734100341797,
"learning_rate": 1.4022556390977444e-05,
"loss": 0.3417,
"step": 507
},
{
"epoch": 9.071428571428571,
"grad_norm": 13.366304397583008,
"learning_rate": 1.3984962406015037e-05,
"loss": 0.2846,
"step": 508
},
{
"epoch": 9.089285714285714,
"grad_norm": 13.366304397583008,
"learning_rate": 1.3947368421052631e-05,
"loss": 0.2303,
"step": 509
},
{
"epoch": 9.107142857142858,
"grad_norm": 13.366304397583008,
"learning_rate": 1.3909774436090224e-05,
"loss": 0.3145,
"step": 510
},
{
"epoch": 9.125,
"grad_norm": 13.366304397583008,
"learning_rate": 1.387218045112782e-05,
"loss": 0.2973,
"step": 511
},
{
"epoch": 9.142857142857142,
"grad_norm": 11.945900917053223,
"learning_rate": 1.3834586466165412e-05,
"loss": 0.268,
"step": 512
},
{
"epoch": 9.160714285714286,
"grad_norm": 11.945900917053223,
"learning_rate": 1.3796992481203009e-05,
"loss": 0.3456,
"step": 513
},
{
"epoch": 9.178571428571429,
"grad_norm": 11.945900917053223,
"learning_rate": 1.3759398496240602e-05,
"loss": 0.252,
"step": 514
},
{
"epoch": 9.196428571428571,
"grad_norm": 11.945900917053223,
"learning_rate": 1.3721804511278197e-05,
"loss": 0.3288,
"step": 515
},
{
"epoch": 9.214285714285714,
"grad_norm": 11.11719036102295,
"learning_rate": 1.368421052631579e-05,
"loss": 0.2966,
"step": 516
},
{
"epoch": 9.232142857142858,
"grad_norm": 11.11719036102295,
"learning_rate": 1.3646616541353384e-05,
"loss": 0.2566,
"step": 517
},
{
"epoch": 9.25,
"grad_norm": 11.11719036102295,
"learning_rate": 1.3609022556390977e-05,
"loss": 0.3144,
"step": 518
},
{
"epoch": 9.267857142857142,
"grad_norm": 11.11719036102295,
"learning_rate": 1.3571428571428572e-05,
"loss": 0.274,
"step": 519
},
{
"epoch": 9.285714285714286,
"grad_norm": 31.70884132385254,
"learning_rate": 1.3533834586466165e-05,
"loss": 0.3115,
"step": 520
},
{
"epoch": 9.303571428571429,
"grad_norm": 31.70884132385254,
"learning_rate": 1.349624060150376e-05,
"loss": 0.28,
"step": 521
},
{
"epoch": 9.321428571428571,
"grad_norm": 31.70884132385254,
"learning_rate": 1.3458646616541353e-05,
"loss": 0.2901,
"step": 522
},
{
"epoch": 9.339285714285714,
"grad_norm": 31.70884132385254,
"learning_rate": 1.3421052631578948e-05,
"loss": 0.2487,
"step": 523
},
{
"epoch": 9.357142857142858,
"grad_norm": 12.3983793258667,
"learning_rate": 1.3383458646616541e-05,
"loss": 0.2812,
"step": 524
},
{
"epoch": 9.375,
"grad_norm": 12.3983793258667,
"learning_rate": 1.3345864661654136e-05,
"loss": 0.3141,
"step": 525
},
{
"epoch": 9.392857142857142,
"grad_norm": 12.3983793258667,
"learning_rate": 1.3308270676691729e-05,
"loss": 0.3222,
"step": 526
},
{
"epoch": 9.410714285714286,
"grad_norm": 12.3983793258667,
"learning_rate": 1.3270676691729323e-05,
"loss": 0.3082,
"step": 527
},
{
"epoch": 9.428571428571429,
"grad_norm": 15.644429206848145,
"learning_rate": 1.3233082706766916e-05,
"loss": 0.2636,
"step": 528
},
{
"epoch": 9.446428571428571,
"grad_norm": 15.644429206848145,
"learning_rate": 1.3195488721804511e-05,
"loss": 0.3269,
"step": 529
},
{
"epoch": 9.464285714285714,
"grad_norm": 15.644429206848145,
"learning_rate": 1.3157894736842104e-05,
"loss": 0.2514,
"step": 530
},
{
"epoch": 9.482142857142858,
"grad_norm": 15.644429206848145,
"learning_rate": 1.3120300751879699e-05,
"loss": 0.2844,
"step": 531
},
{
"epoch": 9.5,
"grad_norm": 17.27918815612793,
"learning_rate": 1.3082706766917292e-05,
"loss": 0.3076,
"step": 532
},
{
"epoch": 9.517857142857142,
"grad_norm": 17.27918815612793,
"learning_rate": 1.3045112781954887e-05,
"loss": 0.3231,
"step": 533
},
{
"epoch": 9.535714285714286,
"grad_norm": 17.27918815612793,
"learning_rate": 1.3007518796992482e-05,
"loss": 0.2596,
"step": 534
},
{
"epoch": 9.553571428571429,
"grad_norm": 17.27918815612793,
"learning_rate": 1.2969924812030076e-05,
"loss": 0.3386,
"step": 535
},
{
"epoch": 9.571428571428571,
"grad_norm": 17.236318588256836,
"learning_rate": 1.293233082706767e-05,
"loss": 0.2981,
"step": 536
},
{
"epoch": 9.589285714285714,
"grad_norm": 17.236318588256836,
"learning_rate": 1.2894736842105264e-05,
"loss": 0.2367,
"step": 537
},
{
"epoch": 9.607142857142858,
"grad_norm": 17.236318588256836,
"learning_rate": 1.2857142857142857e-05,
"loss": 0.3058,
"step": 538
},
{
"epoch": 9.625,
"grad_norm": 17.236318588256836,
"learning_rate": 1.2819548872180452e-05,
"loss": 0.2686,
"step": 539
},
{
"epoch": 9.642857142857142,
"grad_norm": 13.27581787109375,
"learning_rate": 1.2781954887218045e-05,
"loss": 0.3181,
"step": 540
},
{
"epoch": 9.660714285714286,
"grad_norm": 13.27581787109375,
"learning_rate": 1.274436090225564e-05,
"loss": 0.3266,
"step": 541
},
{
"epoch": 9.678571428571429,
"grad_norm": 13.27581787109375,
"learning_rate": 1.2706766917293233e-05,
"loss": 0.3302,
"step": 542
},
{
"epoch": 9.696428571428571,
"grad_norm": 13.27581787109375,
"learning_rate": 1.2669172932330828e-05,
"loss": 0.2802,
"step": 543
},
{
"epoch": 9.714285714285714,
"grad_norm": 16.214496612548828,
"learning_rate": 1.263157894736842e-05,
"loss": 0.295,
"step": 544
},
{
"epoch": 9.732142857142858,
"grad_norm": 16.214496612548828,
"learning_rate": 1.2593984962406015e-05,
"loss": 0.2859,
"step": 545
},
{
"epoch": 9.75,
"grad_norm": 16.214496612548828,
"learning_rate": 1.2556390977443608e-05,
"loss": 0.3203,
"step": 546
},
{
"epoch": 9.767857142857142,
"grad_norm": 16.214496612548828,
"learning_rate": 1.2518796992481203e-05,
"loss": 0.2894,
"step": 547
},
{
"epoch": 9.785714285714286,
"grad_norm": 15.310676574707031,
"learning_rate": 1.2481203007518796e-05,
"loss": 0.2915,
"step": 548
},
{
"epoch": 9.803571428571429,
"grad_norm": 15.310676574707031,
"learning_rate": 1.2443609022556391e-05,
"loss": 0.2659,
"step": 549
},
{
"epoch": 9.821428571428571,
"grad_norm": 15.310676574707031,
"learning_rate": 1.2406015037593984e-05,
"loss": 0.276,
"step": 550
},
{
"epoch": 9.839285714285714,
"grad_norm": 15.310676574707031,
"learning_rate": 1.2368421052631579e-05,
"loss": 0.3742,
"step": 551
},
{
"epoch": 9.857142857142858,
"grad_norm": 19.767837524414062,
"learning_rate": 1.2330827067669172e-05,
"loss": 0.3844,
"step": 552
},
{
"epoch": 9.875,
"grad_norm": 19.767837524414062,
"learning_rate": 1.2293233082706767e-05,
"loss": 0.314,
"step": 553
},
{
"epoch": 9.892857142857142,
"grad_norm": 19.767837524414062,
"learning_rate": 1.225563909774436e-05,
"loss": 0.3383,
"step": 554
},
{
"epoch": 9.910714285714286,
"grad_norm": 19.767837524414062,
"learning_rate": 1.2218045112781956e-05,
"loss": 0.2629,
"step": 555
},
{
"epoch": 9.928571428571429,
"grad_norm": 28.232803344726562,
"learning_rate": 1.218045112781955e-05,
"loss": 0.3078,
"step": 556
},
{
"epoch": 9.946428571428571,
"grad_norm": 28.232803344726562,
"learning_rate": 1.2142857142857144e-05,
"loss": 0.2649,
"step": 557
},
{
"epoch": 9.964285714285714,
"grad_norm": 28.232803344726562,
"learning_rate": 1.2105263157894737e-05,
"loss": 0.2651,
"step": 558
},
{
"epoch": 9.982142857142858,
"grad_norm": 28.232803344726562,
"learning_rate": 1.2067669172932332e-05,
"loss": 0.233,
"step": 559
},
{
"epoch": 10.0,
"grad_norm": 10.35251235961914,
"learning_rate": 1.2030075187969925e-05,
"loss": 0.2915,
"step": 560
},
{
"epoch": 10.017857142857142,
"grad_norm": 10.35251235961914,
"learning_rate": 1.199248120300752e-05,
"loss": 0.2209,
"step": 561
},
{
"epoch": 10.035714285714286,
"grad_norm": 10.35251235961914,
"learning_rate": 1.1954887218045113e-05,
"loss": 0.268,
"step": 562
},
{
"epoch": 10.053571428571429,
"grad_norm": 10.35251235961914,
"learning_rate": 1.1917293233082707e-05,
"loss": 0.2213,
"step": 563
},
{
"epoch": 10.071428571428571,
"grad_norm": 9.80018424987793,
"learning_rate": 1.18796992481203e-05,
"loss": 0.2546,
"step": 564
},
{
"epoch": 10.089285714285714,
"grad_norm": 9.80018424987793,
"learning_rate": 1.1842105263157895e-05,
"loss": 0.2101,
"step": 565
},
{
"epoch": 10.107142857142858,
"grad_norm": 9.80018424987793,
"learning_rate": 1.1804511278195488e-05,
"loss": 0.1867,
"step": 566
},
{
"epoch": 10.125,
"grad_norm": 9.80018424987793,
"learning_rate": 1.1766917293233083e-05,
"loss": 0.2667,
"step": 567
},
{
"epoch": 10.142857142857142,
"grad_norm": 10.689797401428223,
"learning_rate": 1.1729323308270676e-05,
"loss": 0.1916,
"step": 568
},
{
"epoch": 10.160714285714286,
"grad_norm": 10.689797401428223,
"learning_rate": 1.169172932330827e-05,
"loss": 0.2905,
"step": 569
},
{
"epoch": 10.178571428571429,
"grad_norm": 10.689797401428223,
"learning_rate": 1.1654135338345864e-05,
"loss": 0.2146,
"step": 570
},
{
"epoch": 10.196428571428571,
"grad_norm": 10.689797401428223,
"learning_rate": 1.1616541353383459e-05,
"loss": 0.2162,
"step": 571
},
{
"epoch": 10.214285714285714,
"grad_norm": 10.927838325500488,
"learning_rate": 1.1578947368421052e-05,
"loss": 0.1946,
"step": 572
},
{
"epoch": 10.232142857142858,
"grad_norm": 10.927838325500488,
"learning_rate": 1.1541353383458646e-05,
"loss": 0.2238,
"step": 573
},
{
"epoch": 10.25,
"grad_norm": 10.927838325500488,
"learning_rate": 1.150375939849624e-05,
"loss": 0.2283,
"step": 574
},
{
"epoch": 10.267857142857142,
"grad_norm": 10.927838325500488,
"learning_rate": 1.1466165413533834e-05,
"loss": 0.2407,
"step": 575
},
{
"epoch": 10.285714285714286,
"grad_norm": 10.374297142028809,
"learning_rate": 1.1428571428571429e-05,
"loss": 0.1796,
"step": 576
},
{
"epoch": 10.303571428571429,
"grad_norm": 10.374297142028809,
"learning_rate": 1.1390977443609024e-05,
"loss": 0.228,
"step": 577
},
{
"epoch": 10.321428571428571,
"grad_norm": 10.374297142028809,
"learning_rate": 1.1353383458646617e-05,
"loss": 0.2779,
"step": 578
},
{
"epoch": 10.339285714285714,
"grad_norm": 10.374297142028809,
"learning_rate": 1.1315789473684212e-05,
"loss": 0.224,
"step": 579
},
{
"epoch": 10.357142857142858,
"grad_norm": 12.753782272338867,
"learning_rate": 1.1278195488721805e-05,
"loss": 0.2138,
"step": 580
},
{
"epoch": 10.375,
"grad_norm": 12.753782272338867,
"learning_rate": 1.12406015037594e-05,
"loss": 0.2141,
"step": 581
},
{
"epoch": 10.392857142857142,
"grad_norm": 12.753782272338867,
"learning_rate": 1.1203007518796992e-05,
"loss": 0.2194,
"step": 582
},
{
"epoch": 10.410714285714286,
"grad_norm": 12.753782272338867,
"learning_rate": 1.1165413533834587e-05,
"loss": 0.1991,
"step": 583
},
{
"epoch": 10.428571428571429,
"grad_norm": 21.148046493530273,
"learning_rate": 1.112781954887218e-05,
"loss": 0.2708,
"step": 584
},
{
"epoch": 10.446428571428571,
"grad_norm": 21.148046493530273,
"learning_rate": 1.1090225563909775e-05,
"loss": 0.2496,
"step": 585
},
{
"epoch": 10.464285714285714,
"grad_norm": 21.148046493530273,
"learning_rate": 1.1052631578947368e-05,
"loss": 0.1911,
"step": 586
},
{
"epoch": 10.482142857142858,
"grad_norm": 21.148046493530273,
"learning_rate": 1.1015037593984963e-05,
"loss": 0.2473,
"step": 587
},
{
"epoch": 10.5,
"grad_norm": 12.04593276977539,
"learning_rate": 1.0977443609022556e-05,
"loss": 0.2556,
"step": 588
},
{
"epoch": 10.517857142857142,
"grad_norm": 12.04593276977539,
"learning_rate": 1.093984962406015e-05,
"loss": 0.2423,
"step": 589
},
{
"epoch": 10.535714285714286,
"grad_norm": 12.04593276977539,
"learning_rate": 1.0902255639097744e-05,
"loss": 0.1698,
"step": 590
},
{
"epoch": 10.553571428571429,
"grad_norm": 12.04593276977539,
"learning_rate": 1.0864661654135338e-05,
"loss": 0.2162,
"step": 591
},
{
"epoch": 10.571428571428571,
"grad_norm": 14.653116226196289,
"learning_rate": 1.0827067669172932e-05,
"loss": 0.2297,
"step": 592
},
{
"epoch": 10.589285714285714,
"grad_norm": 14.653116226196289,
"learning_rate": 1.0789473684210526e-05,
"loss": 0.1786,
"step": 593
},
{
"epoch": 10.607142857142858,
"grad_norm": 14.653116226196289,
"learning_rate": 1.075187969924812e-05,
"loss": 0.2218,
"step": 594
},
{
"epoch": 10.625,
"grad_norm": 14.653116226196289,
"learning_rate": 1.0714285714285714e-05,
"loss": 0.2042,
"step": 595
},
{
"epoch": 10.642857142857142,
"grad_norm": 13.415528297424316,
"learning_rate": 1.0676691729323307e-05,
"loss": 0.2565,
"step": 596
},
{
"epoch": 10.660714285714286,
"grad_norm": 13.415528297424316,
"learning_rate": 1.0639097744360904e-05,
"loss": 0.2454,
"step": 597
},
{
"epoch": 10.678571428571429,
"grad_norm": 13.415528297424316,
"learning_rate": 1.0601503759398497e-05,
"loss": 0.2245,
"step": 598
},
{
"epoch": 10.696428571428571,
"grad_norm": 13.415528297424316,
"learning_rate": 1.0563909774436091e-05,
"loss": 0.188,
"step": 599
},
{
"epoch": 10.714285714285714,
"grad_norm": 11.686877250671387,
"learning_rate": 1.0526315789473684e-05,
"loss": 0.2103,
"step": 600
},
{
"epoch": 10.732142857142858,
"grad_norm": 11.686877250671387,
"learning_rate": 1.048872180451128e-05,
"loss": 0.2552,
"step": 601
},
{
"epoch": 10.75,
"grad_norm": 11.686877250671387,
"learning_rate": 1.0451127819548872e-05,
"loss": 0.2208,
"step": 602
},
{
"epoch": 10.767857142857142,
"grad_norm": 11.686877250671387,
"learning_rate": 1.0413533834586467e-05,
"loss": 0.2197,
"step": 603
},
{
"epoch": 10.785714285714286,
"grad_norm": 15.676922798156738,
"learning_rate": 1.037593984962406e-05,
"loss": 0.2041,
"step": 604
},
{
"epoch": 10.803571428571429,
"grad_norm": 15.676922798156738,
"learning_rate": 1.0338345864661655e-05,
"loss": 0.2056,
"step": 605
},
{
"epoch": 10.821428571428571,
"grad_norm": 15.676922798156738,
"learning_rate": 1.0300751879699248e-05,
"loss": 0.2556,
"step": 606
},
{
"epoch": 10.839285714285714,
"grad_norm": 15.676922798156738,
"learning_rate": 1.0263157894736843e-05,
"loss": 0.2674,
"step": 607
},
{
"epoch": 10.857142857142858,
"grad_norm": 13.185648918151855,
"learning_rate": 1.0225563909774436e-05,
"loss": 0.2076,
"step": 608
},
{
"epoch": 10.875,
"grad_norm": 13.185648918151855,
"learning_rate": 1.018796992481203e-05,
"loss": 0.2189,
"step": 609
},
{
"epoch": 10.892857142857142,
"grad_norm": 13.185648918151855,
"learning_rate": 1.0150375939849624e-05,
"loss": 0.1981,
"step": 610
},
{
"epoch": 10.910714285714286,
"grad_norm": 13.185648918151855,
"learning_rate": 1.0112781954887218e-05,
"loss": 0.2155,
"step": 611
},
{
"epoch": 10.928571428571429,
"grad_norm": 14.26292896270752,
"learning_rate": 1.0075187969924811e-05,
"loss": 0.2292,
"step": 612
},
{
"epoch": 10.946428571428571,
"grad_norm": 14.26292896270752,
"learning_rate": 1.0037593984962406e-05,
"loss": 0.2382,
"step": 613
},
{
"epoch": 10.964285714285714,
"grad_norm": 14.26292896270752,
"learning_rate": 9.999999999999999e-06,
"loss": 0.2027,
"step": 614
},
{
"epoch": 10.982142857142858,
"grad_norm": 14.26292896270752,
"learning_rate": 9.962406015037594e-06,
"loss": 0.2087,
"step": 615
},
{
"epoch": 11.0,
"grad_norm": 12.144270896911621,
"learning_rate": 9.924812030075187e-06,
"loss": 0.2356,
"step": 616
},
{
"epoch": 11.017857142857142,
"grad_norm": 12.144270896911621,
"learning_rate": 9.887218045112782e-06,
"loss": 0.1954,
"step": 617
},
{
"epoch": 11.035714285714286,
"grad_norm": 12.144270896911621,
"learning_rate": 9.849624060150376e-06,
"loss": 0.1647,
"step": 618
},
{
"epoch": 11.053571428571429,
"grad_norm": 12.144270896911621,
"learning_rate": 9.812030075187971e-06,
"loss": 0.1658,
"step": 619
},
{
"epoch": 11.071428571428571,
"grad_norm": 44.392921447753906,
"learning_rate": 9.774436090225564e-06,
"loss": 0.1965,
"step": 620
},
{
"epoch": 11.089285714285714,
"grad_norm": 44.392921447753906,
"learning_rate": 9.736842105263159e-06,
"loss": 0.1646,
"step": 621
},
{
"epoch": 11.107142857142858,
"grad_norm": 44.392921447753906,
"learning_rate": 9.699248120300752e-06,
"loss": 0.2028,
"step": 622
},
{
"epoch": 11.125,
"grad_norm": 44.392921447753906,
"learning_rate": 9.661654135338347e-06,
"loss": 0.1933,
"step": 623
},
{
"epoch": 11.142857142857142,
"grad_norm": 234.00765991210938,
"learning_rate": 9.62406015037594e-06,
"loss": 0.1635,
"step": 624
},
{
"epoch": 11.160714285714286,
"grad_norm": 234.00765991210938,
"learning_rate": 9.586466165413535e-06,
"loss": 0.1916,
"step": 625
},
{
"epoch": 11.178571428571429,
"grad_norm": 234.00765991210938,
"learning_rate": 9.548872180451128e-06,
"loss": 0.1579,
"step": 626
},
{
"epoch": 11.196428571428571,
"grad_norm": 234.00765991210938,
"learning_rate": 9.511278195488722e-06,
"loss": 0.1893,
"step": 627
},
{
"epoch": 11.214285714285714,
"grad_norm": 16.09220314025879,
"learning_rate": 9.473684210526315e-06,
"loss": 0.1885,
"step": 628
},
{
"epoch": 11.232142857142858,
"grad_norm": 16.09220314025879,
"learning_rate": 9.43609022556391e-06,
"loss": 0.1928,
"step": 629
},
{
"epoch": 11.25,
"grad_norm": 16.09220314025879,
"learning_rate": 9.398496240601503e-06,
"loss": 0.2045,
"step": 630
},
{
"epoch": 11.267857142857142,
"grad_norm": 16.09220314025879,
"learning_rate": 9.360902255639098e-06,
"loss": 0.2016,
"step": 631
},
{
"epoch": 11.285714285714286,
"grad_norm": 18.606782913208008,
"learning_rate": 9.323308270676691e-06,
"loss": 0.1913,
"step": 632
},
{
"epoch": 11.303571428571429,
"grad_norm": 18.606782913208008,
"learning_rate": 9.285714285714286e-06,
"loss": 0.1771,
"step": 633
},
{
"epoch": 11.321428571428571,
"grad_norm": 18.606782913208008,
"learning_rate": 9.248120300751879e-06,
"loss": 0.1754,
"step": 634
},
{
"epoch": 11.339285714285714,
"grad_norm": 18.606782913208008,
"learning_rate": 9.210526315789474e-06,
"loss": 0.1687,
"step": 635
},
{
"epoch": 11.357142857142858,
"grad_norm": 7.952782154083252,
"learning_rate": 9.172932330827067e-06,
"loss": 0.1816,
"step": 636
},
{
"epoch": 11.375,
"grad_norm": 7.952782154083252,
"learning_rate": 9.135338345864661e-06,
"loss": 0.1575,
"step": 637
},
{
"epoch": 11.392857142857142,
"grad_norm": 7.952782154083252,
"learning_rate": 9.097744360902255e-06,
"loss": 0.1841,
"step": 638
},
{
"epoch": 11.410714285714286,
"grad_norm": 7.952782154083252,
"learning_rate": 9.060150375939851e-06,
"loss": 0.2001,
"step": 639
},
{
"epoch": 11.428571428571429,
"grad_norm": 10.40957260131836,
"learning_rate": 9.022556390977444e-06,
"loss": 0.1515,
"step": 640
},
{
"epoch": 11.446428571428571,
"grad_norm": 10.40957260131836,
"learning_rate": 8.984962406015039e-06,
"loss": 0.1743,
"step": 641
},
{
"epoch": 11.464285714285714,
"grad_norm": 10.40957260131836,
"learning_rate": 8.947368421052632e-06,
"loss": 0.1741,
"step": 642
},
{
"epoch": 11.482142857142858,
"grad_norm": 10.40957260131836,
"learning_rate": 8.909774436090227e-06,
"loss": 0.1859,
"step": 643
},
{
"epoch": 11.5,
"grad_norm": 188.40196228027344,
"learning_rate": 8.87218045112782e-06,
"loss": 0.1494,
"step": 644
},
{
"epoch": 11.517857142857142,
"grad_norm": 188.40196228027344,
"learning_rate": 8.834586466165414e-06,
"loss": 0.1649,
"step": 645
},
{
"epoch": 11.535714285714286,
"grad_norm": 188.40196228027344,
"learning_rate": 8.796992481203007e-06,
"loss": 0.191,
"step": 646
},
{
"epoch": 11.553571428571429,
"grad_norm": 188.40196228027344,
"learning_rate": 8.759398496240602e-06,
"loss": 0.1807,
"step": 647
},
{
"epoch": 11.571428571428571,
"grad_norm": 52.5207405090332,
"learning_rate": 8.721804511278195e-06,
"loss": 0.1628,
"step": 648
},
{
"epoch": 11.589285714285714,
"grad_norm": 52.5207405090332,
"learning_rate": 8.68421052631579e-06,
"loss": 0.1634,
"step": 649
},
{
"epoch": 11.607142857142858,
"grad_norm": 52.5207405090332,
"learning_rate": 8.646616541353383e-06,
"loss": 0.1692,
"step": 650
},
{
"epoch": 11.625,
"grad_norm": 52.5207405090332,
"learning_rate": 8.609022556390978e-06,
"loss": 0.2182,
"step": 651
},
{
"epoch": 11.642857142857142,
"grad_norm": 10.022271156311035,
"learning_rate": 8.571428571428571e-06,
"loss": 0.1702,
"step": 652
},
{
"epoch": 11.660714285714286,
"grad_norm": 10.022271156311035,
"learning_rate": 8.533834586466166e-06,
"loss": 0.1597,
"step": 653
},
{
"epoch": 11.678571428571429,
"grad_norm": 10.022271156311035,
"learning_rate": 8.496240601503759e-06,
"loss": 0.2204,
"step": 654
},
{
"epoch": 11.696428571428571,
"grad_norm": 10.022271156311035,
"learning_rate": 8.458646616541353e-06,
"loss": 0.1739,
"step": 655
},
{
"epoch": 11.714285714285714,
"grad_norm": 11.935576438903809,
"learning_rate": 8.421052631578947e-06,
"loss": 0.1827,
"step": 656
},
{
"epoch": 11.732142857142858,
"grad_norm": 11.935576438903809,
"learning_rate": 8.383458646616541e-06,
"loss": 0.2052,
"step": 657
},
{
"epoch": 11.75,
"grad_norm": 11.935576438903809,
"learning_rate": 8.345864661654134e-06,
"loss": 0.1798,
"step": 658
},
{
"epoch": 11.767857142857142,
"grad_norm": 11.935576438903809,
"learning_rate": 8.308270676691729e-06,
"loss": 0.1735,
"step": 659
},
{
"epoch": 11.785714285714286,
"grad_norm": 11.527193069458008,
"learning_rate": 8.270676691729324e-06,
"loss": 0.1643,
"step": 660
},
{
"epoch": 11.803571428571429,
"grad_norm": 11.527193069458008,
"learning_rate": 8.233082706766919e-06,
"loss": 0.1622,
"step": 661
},
{
"epoch": 11.821428571428571,
"grad_norm": 11.527193069458008,
"learning_rate": 8.195488721804512e-06,
"loss": 0.1754,
"step": 662
},
{
"epoch": 11.839285714285714,
"grad_norm": 11.527193069458008,
"learning_rate": 8.157894736842106e-06,
"loss": 0.1603,
"step": 663
},
{
"epoch": 11.857142857142858,
"grad_norm": 16.521615982055664,
"learning_rate": 8.1203007518797e-06,
"loss": 0.2075,
"step": 664
},
{
"epoch": 11.875,
"grad_norm": 16.521615982055664,
"learning_rate": 8.082706766917294e-06,
"loss": 0.1332,
"step": 665
},
{
"epoch": 11.892857142857142,
"grad_norm": 16.521615982055664,
"learning_rate": 8.045112781954887e-06,
"loss": 0.1897,
"step": 666
},
{
"epoch": 11.910714285714286,
"grad_norm": 16.521615982055664,
"learning_rate": 8.007518796992482e-06,
"loss": 0.1577,
"step": 667
},
{
"epoch": 11.928571428571429,
"grad_norm": 13.228716850280762,
"learning_rate": 7.969924812030075e-06,
"loss": 0.1892,
"step": 668
},
{
"epoch": 11.946428571428571,
"grad_norm": 13.228716850280762,
"learning_rate": 7.93233082706767e-06,
"loss": 0.1249,
"step": 669
},
{
"epoch": 11.964285714285714,
"grad_norm": 13.228716850280762,
"learning_rate": 7.894736842105263e-06,
"loss": 0.1955,
"step": 670
},
{
"epoch": 11.982142857142858,
"grad_norm": 13.228716850280762,
"learning_rate": 7.857142857142858e-06,
"loss": 0.1721,
"step": 671
},
{
"epoch": 12.0,
"grad_norm": 12.504434585571289,
"learning_rate": 7.81954887218045e-06,
"loss": 0.1626,
"step": 672
},
{
"epoch": 12.017857142857142,
"grad_norm": 12.504434585571289,
"learning_rate": 7.781954887218045e-06,
"loss": 0.1537,
"step": 673
},
{
"epoch": 12.035714285714286,
"grad_norm": 12.504434585571289,
"learning_rate": 7.744360902255639e-06,
"loss": 0.1682,
"step": 674
},
{
"epoch": 12.053571428571429,
"grad_norm": 12.504434585571289,
"learning_rate": 7.706766917293233e-06,
"loss": 0.1476,
"step": 675
},
{
"epoch": 12.071428571428571,
"grad_norm": 7.761070251464844,
"learning_rate": 7.669172932330826e-06,
"loss": 0.1587,
"step": 676
},
{
"epoch": 12.089285714285714,
"grad_norm": 7.761070251464844,
"learning_rate": 7.631578947368421e-06,
"loss": 0.1647,
"step": 677
},
{
"epoch": 12.107142857142858,
"grad_norm": 7.761070251464844,
"learning_rate": 7.593984962406015e-06,
"loss": 0.144,
"step": 678
},
{
"epoch": 12.125,
"grad_norm": 7.761070251464844,
"learning_rate": 7.55639097744361e-06,
"loss": 0.1525,
"step": 679
},
{
"epoch": 12.142857142857142,
"grad_norm": 8.14785099029541,
"learning_rate": 7.518796992481203e-06,
"loss": 0.1711,
"step": 680
},
{
"epoch": 12.160714285714286,
"grad_norm": 8.14785099029541,
"learning_rate": 7.481203007518797e-06,
"loss": 0.1503,
"step": 681
},
{
"epoch": 12.178571428571429,
"grad_norm": 8.14785099029541,
"learning_rate": 7.443609022556391e-06,
"loss": 0.143,
"step": 682
},
{
"epoch": 12.196428571428571,
"grad_norm": 8.14785099029541,
"learning_rate": 7.4060150375939845e-06,
"loss": 0.1627,
"step": 683
},
{
"epoch": 12.214285714285714,
"grad_norm": 8.028116226196289,
"learning_rate": 7.3684210526315784e-06,
"loss": 0.1305,
"step": 684
},
{
"epoch": 12.232142857142858,
"grad_norm": 8.028116226196289,
"learning_rate": 7.330827067669172e-06,
"loss": 0.1211,
"step": 685
},
{
"epoch": 12.25,
"grad_norm": 8.028116226196289,
"learning_rate": 7.293233082706767e-06,
"loss": 0.1931,
"step": 686
},
{
"epoch": 12.267857142857142,
"grad_norm": 8.028116226196289,
"learning_rate": 7.255639097744361e-06,
"loss": 0.1243,
"step": 687
},
{
"epoch": 12.285714285714286,
"grad_norm": 10.175227165222168,
"learning_rate": 7.218045112781955e-06,
"loss": 0.1329,
"step": 688
},
{
"epoch": 12.303571428571429,
"grad_norm": 10.175227165222168,
"learning_rate": 7.180451127819549e-06,
"loss": 0.1555,
"step": 689
},
{
"epoch": 12.321428571428571,
"grad_norm": 10.175227165222168,
"learning_rate": 7.142857142857143e-06,
"loss": 0.1274,
"step": 690
},
{
"epoch": 12.339285714285714,
"grad_norm": 10.175227165222168,
"learning_rate": 7.105263157894737e-06,
"loss": 0.1172,
"step": 691
},
{
"epoch": 12.357142857142858,
"grad_norm": 10.643868446350098,
"learning_rate": 7.0676691729323305e-06,
"loss": 0.1423,
"step": 692
},
{
"epoch": 12.375,
"grad_norm": 10.643868446350098,
"learning_rate": 7.0300751879699244e-06,
"loss": 0.1271,
"step": 693
},
{
"epoch": 12.392857142857142,
"grad_norm": 10.643868446350098,
"learning_rate": 6.992481203007518e-06,
"loss": 0.1431,
"step": 694
},
{
"epoch": 12.410714285714286,
"grad_norm": 10.643868446350098,
"learning_rate": 6.954887218045112e-06,
"loss": 0.1311,
"step": 695
},
{
"epoch": 12.428571428571429,
"grad_norm": 9.100255012512207,
"learning_rate": 6.917293233082706e-06,
"loss": 0.0794,
"step": 696
},
{
"epoch": 12.446428571428571,
"grad_norm": 9.100255012512207,
"learning_rate": 6.879699248120301e-06,
"loss": 0.1843,
"step": 697
},
{
"epoch": 12.464285714285714,
"grad_norm": 9.100255012512207,
"learning_rate": 6.842105263157895e-06,
"loss": 0.2053,
"step": 698
},
{
"epoch": 12.482142857142858,
"grad_norm": 9.100255012512207,
"learning_rate": 6.804511278195489e-06,
"loss": 0.1439,
"step": 699
},
{
"epoch": 12.5,
"grad_norm": 11.272979736328125,
"learning_rate": 6.766917293233083e-06,
"loss": 0.1663,
"step": 700
},
{
"epoch": 12.517857142857142,
"grad_norm": 11.272979736328125,
"learning_rate": 6.7293233082706765e-06,
"loss": 0.1325,
"step": 701
},
{
"epoch": 12.535714285714286,
"grad_norm": 11.272979736328125,
"learning_rate": 6.6917293233082704e-06,
"loss": 0.1302,
"step": 702
},
{
"epoch": 12.553571428571429,
"grad_norm": 11.272979736328125,
"learning_rate": 6.654135338345864e-06,
"loss": 0.1316,
"step": 703
},
{
"epoch": 12.571428571428571,
"grad_norm": 9.36953067779541,
"learning_rate": 6.616541353383458e-06,
"loss": 0.1444,
"step": 704
},
{
"epoch": 12.589285714285714,
"grad_norm": 9.36953067779541,
"learning_rate": 6.578947368421052e-06,
"loss": 0.1322,
"step": 705
},
{
"epoch": 12.607142857142858,
"grad_norm": 9.36953067779541,
"learning_rate": 6.541353383458646e-06,
"loss": 0.1504,
"step": 706
},
{
"epoch": 12.625,
"grad_norm": 9.36953067779541,
"learning_rate": 6.503759398496241e-06,
"loss": 0.1145,
"step": 707
},
{
"epoch": 12.642857142857142,
"grad_norm": 20.24629020690918,
"learning_rate": 6.466165413533835e-06,
"loss": 0.1746,
"step": 708
},
{
"epoch": 12.660714285714286,
"grad_norm": 20.24629020690918,
"learning_rate": 6.428571428571429e-06,
"loss": 0.1365,
"step": 709
},
{
"epoch": 12.678571428571429,
"grad_norm": 20.24629020690918,
"learning_rate": 6.3909774436090225e-06,
"loss": 0.1366,
"step": 710
},
{
"epoch": 12.696428571428571,
"grad_norm": 20.24629020690918,
"learning_rate": 6.3533834586466164e-06,
"loss": 0.1359,
"step": 711
},
{
"epoch": 12.714285714285714,
"grad_norm": 90.94606018066406,
"learning_rate": 6.31578947368421e-06,
"loss": 0.1788,
"step": 712
},
{
"epoch": 12.732142857142858,
"grad_norm": 90.94606018066406,
"learning_rate": 6.278195488721804e-06,
"loss": 0.1676,
"step": 713
},
{
"epoch": 12.75,
"grad_norm": 90.94606018066406,
"learning_rate": 6.240601503759398e-06,
"loss": 0.1329,
"step": 714
},
{
"epoch": 12.767857142857142,
"grad_norm": 90.94606018066406,
"learning_rate": 6.203007518796992e-06,
"loss": 0.207,
"step": 715
},
{
"epoch": 12.785714285714286,
"grad_norm": 17.4215087890625,
"learning_rate": 6.165413533834586e-06,
"loss": 0.1412,
"step": 716
},
{
"epoch": 12.803571428571429,
"grad_norm": 17.4215087890625,
"learning_rate": 6.12781954887218e-06,
"loss": 0.1341,
"step": 717
},
{
"epoch": 12.821428571428571,
"grad_norm": 17.4215087890625,
"learning_rate": 6.090225563909775e-06,
"loss": 0.1616,
"step": 718
},
{
"epoch": 12.839285714285714,
"grad_norm": 17.4215087890625,
"learning_rate": 6.0526315789473685e-06,
"loss": 0.165,
"step": 719
},
{
"epoch": 12.857142857142858,
"grad_norm": 9.475699424743652,
"learning_rate": 6.015037593984962e-06,
"loss": 0.1632,
"step": 720
},
{
"epoch": 12.875,
"grad_norm": 9.475699424743652,
"learning_rate": 5.977443609022556e-06,
"loss": 0.1351,
"step": 721
},
{
"epoch": 12.892857142857142,
"grad_norm": 9.475699424743652,
"learning_rate": 5.93984962406015e-06,
"loss": 0.143,
"step": 722
},
{
"epoch": 12.910714285714286,
"grad_norm": 9.475699424743652,
"learning_rate": 5.902255639097744e-06,
"loss": 0.1378,
"step": 723
},
{
"epoch": 12.928571428571429,
"grad_norm": 8.660880088806152,
"learning_rate": 5.864661654135338e-06,
"loss": 0.1521,
"step": 724
},
{
"epoch": 12.946428571428571,
"grad_norm": 8.660880088806152,
"learning_rate": 5.827067669172932e-06,
"loss": 0.1602,
"step": 725
},
{
"epoch": 12.964285714285714,
"grad_norm": 8.660880088806152,
"learning_rate": 5.789473684210526e-06,
"loss": 0.1753,
"step": 726
},
{
"epoch": 12.982142857142858,
"grad_norm": 8.660880088806152,
"learning_rate": 5.75187969924812e-06,
"loss": 0.1231,
"step": 727
},
{
"epoch": 13.0,
"grad_norm": 12.282486915588379,
"learning_rate": 5.7142857142857145e-06,
"loss": 0.1537,
"step": 728
},
{
"epoch": 13.017857142857142,
"grad_norm": 12.282486915588379,
"learning_rate": 5.676691729323308e-06,
"loss": 0.1233,
"step": 729
},
{
"epoch": 13.035714285714286,
"grad_norm": 12.282486915588379,
"learning_rate": 5.639097744360902e-06,
"loss": 0.093,
"step": 730
},
{
"epoch": 13.053571428571429,
"grad_norm": 12.282486915588379,
"learning_rate": 5.601503759398496e-06,
"loss": 0.1009,
"step": 731
},
{
"epoch": 13.071428571428571,
"grad_norm": 8.800283432006836,
"learning_rate": 5.56390977443609e-06,
"loss": 0.1383,
"step": 732
},
{
"epoch": 13.089285714285714,
"grad_norm": 8.800283432006836,
"learning_rate": 5.526315789473684e-06,
"loss": 0.1129,
"step": 733
},
{
"epoch": 13.107142857142858,
"grad_norm": 8.800283432006836,
"learning_rate": 5.488721804511278e-06,
"loss": 0.0979,
"step": 734
},
{
"epoch": 13.125,
"grad_norm": 8.800283432006836,
"learning_rate": 5.451127819548872e-06,
"loss": 0.1158,
"step": 735
},
{
"epoch": 13.142857142857142,
"grad_norm": 7.84540319442749,
"learning_rate": 5.413533834586466e-06,
"loss": 0.1014,
"step": 736
},
{
"epoch": 13.160714285714286,
"grad_norm": 7.84540319442749,
"learning_rate": 5.37593984962406e-06,
"loss": 0.1181,
"step": 737
},
{
"epoch": 13.178571428571429,
"grad_norm": 7.84540319442749,
"learning_rate": 5.3383458646616536e-06,
"loss": 0.1273,
"step": 738
},
{
"epoch": 13.196428571428571,
"grad_norm": 7.84540319442749,
"learning_rate": 5.300751879699248e-06,
"loss": 0.1039,
"step": 739
},
{
"epoch": 13.214285714285714,
"grad_norm": 18.80352210998535,
"learning_rate": 5.263157894736842e-06,
"loss": 0.1427,
"step": 740
},
{
"epoch": 13.232142857142858,
"grad_norm": 18.80352210998535,
"learning_rate": 5.225563909774436e-06,
"loss": 0.1245,
"step": 741
},
{
"epoch": 13.25,
"grad_norm": 18.80352210998535,
"learning_rate": 5.18796992481203e-06,
"loss": 0.1345,
"step": 742
},
{
"epoch": 13.267857142857142,
"grad_norm": 18.80352210998535,
"learning_rate": 5.150375939849624e-06,
"loss": 0.1038,
"step": 743
},
{
"epoch": 13.285714285714286,
"grad_norm": 8.5668306350708,
"learning_rate": 5.112781954887218e-06,
"loss": 0.1285,
"step": 744
},
{
"epoch": 13.303571428571429,
"grad_norm": 8.5668306350708,
"learning_rate": 5.075187969924812e-06,
"loss": 0.0958,
"step": 745
},
{
"epoch": 13.321428571428571,
"grad_norm": 8.5668306350708,
"learning_rate": 5.037593984962406e-06,
"loss": 0.1313,
"step": 746
},
{
"epoch": 13.339285714285714,
"grad_norm": 8.5668306350708,
"learning_rate": 4.9999999999999996e-06,
"loss": 0.1349,
"step": 747
},
{
"epoch": 13.357142857142858,
"grad_norm": 11.687544822692871,
"learning_rate": 4.9624060150375935e-06,
"loss": 0.1037,
"step": 748
},
{
"epoch": 13.375,
"grad_norm": 11.687544822692871,
"learning_rate": 4.924812030075188e-06,
"loss": 0.1182,
"step": 749
},
{
"epoch": 13.392857142857142,
"grad_norm": 11.687544822692871,
"learning_rate": 4.887218045112782e-06,
"loss": 0.1554,
"step": 750
},
{
"epoch": 13.410714285714286,
"grad_norm": 11.687544822692871,
"learning_rate": 4.849624060150376e-06,
"loss": 0.1296,
"step": 751
},
{
"epoch": 13.428571428571429,
"grad_norm": 10.747458457946777,
"learning_rate": 4.81203007518797e-06,
"loss": 0.1593,
"step": 752
},
{
"epoch": 13.446428571428571,
"grad_norm": 10.747458457946777,
"learning_rate": 4.774436090225564e-06,
"loss": 0.1088,
"step": 753
},
{
"epoch": 13.464285714285714,
"grad_norm": 10.747458457946777,
"learning_rate": 4.736842105263158e-06,
"loss": 0.1116,
"step": 754
},
{
"epoch": 13.482142857142858,
"grad_norm": 10.747458457946777,
"learning_rate": 4.699248120300752e-06,
"loss": 0.1416,
"step": 755
},
{
"epoch": 13.5,
"grad_norm": 10.563179016113281,
"learning_rate": 4.6616541353383456e-06,
"loss": 0.1121,
"step": 756
},
{
"epoch": 13.517857142857142,
"grad_norm": 10.563179016113281,
"learning_rate": 4.6240601503759395e-06,
"loss": 0.1185,
"step": 757
},
{
"epoch": 13.535714285714286,
"grad_norm": 10.563179016113281,
"learning_rate": 4.586466165413533e-06,
"loss": 0.1119,
"step": 758
},
{
"epoch": 13.553571428571429,
"grad_norm": 10.563179016113281,
"learning_rate": 4.548872180451127e-06,
"loss": 0.0942,
"step": 759
},
{
"epoch": 13.571428571428571,
"grad_norm": 9.400467872619629,
"learning_rate": 4.511278195488722e-06,
"loss": 0.1783,
"step": 760
},
{
"epoch": 13.589285714285714,
"grad_norm": 9.400467872619629,
"learning_rate": 4.473684210526316e-06,
"loss": 0.1191,
"step": 761
},
{
"epoch": 13.607142857142858,
"grad_norm": 9.400467872619629,
"learning_rate": 4.43609022556391e-06,
"loss": 0.147,
"step": 762
},
{
"epoch": 13.625,
"grad_norm": 9.400467872619629,
"learning_rate": 4.398496240601504e-06,
"loss": 0.1812,
"step": 763
},
{
"epoch": 13.642857142857142,
"grad_norm": 11.890040397644043,
"learning_rate": 4.360902255639098e-06,
"loss": 0.098,
"step": 764
},
{
"epoch": 13.660714285714286,
"grad_norm": 11.890040397644043,
"learning_rate": 4.3233082706766916e-06,
"loss": 0.1357,
"step": 765
},
{
"epoch": 13.678571428571429,
"grad_norm": 11.890040397644043,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.1062,
"step": 766
},
{
"epoch": 13.696428571428571,
"grad_norm": 11.890040397644043,
"learning_rate": 4.248120300751879e-06,
"loss": 0.1168,
"step": 767
},
{
"epoch": 13.714285714285714,
"grad_norm": 63.02323532104492,
"learning_rate": 4.210526315789473e-06,
"loss": 0.1257,
"step": 768
},
{
"epoch": 13.732142857142858,
"grad_norm": 63.02323532104492,
"learning_rate": 4.172932330827067e-06,
"loss": 0.113,
"step": 769
},
{
"epoch": 13.75,
"grad_norm": 63.02323532104492,
"learning_rate": 4.135338345864662e-06,
"loss": 0.1284,
"step": 770
},
{
"epoch": 13.767857142857142,
"grad_norm": 63.02323532104492,
"learning_rate": 4.097744360902256e-06,
"loss": 0.1134,
"step": 771
},
{
"epoch": 13.785714285714286,
"grad_norm": 135.0414276123047,
"learning_rate": 4.06015037593985e-06,
"loss": 0.1289,
"step": 772
},
{
"epoch": 13.803571428571429,
"grad_norm": 135.0414276123047,
"learning_rate": 4.022556390977444e-06,
"loss": 0.1274,
"step": 773
},
{
"epoch": 13.821428571428571,
"grad_norm": 135.0414276123047,
"learning_rate": 3.9849624060150376e-06,
"loss": 0.1403,
"step": 774
},
{
"epoch": 13.839285714285714,
"grad_norm": 135.0414276123047,
"learning_rate": 3.9473684210526315e-06,
"loss": 0.1275,
"step": 775
},
{
"epoch": 13.857142857142858,
"grad_norm": 9.245959281921387,
"learning_rate": 3.909774436090225e-06,
"loss": 0.145,
"step": 776
},
{
"epoch": 13.875,
"grad_norm": 9.245959281921387,
"learning_rate": 3.872180451127819e-06,
"loss": 0.1408,
"step": 777
},
{
"epoch": 13.892857142857142,
"grad_norm": 9.245959281921387,
"learning_rate": 3.834586466165413e-06,
"loss": 0.1199,
"step": 778
},
{
"epoch": 13.910714285714286,
"grad_norm": 9.245959281921387,
"learning_rate": 3.7969924812030075e-06,
"loss": 0.1696,
"step": 779
},
{
"epoch": 13.928571428571429,
"grad_norm": 14.7429838180542,
"learning_rate": 3.7593984962406014e-06,
"loss": 0.1293,
"step": 780
},
{
"epoch": 13.946428571428571,
"grad_norm": 14.7429838180542,
"learning_rate": 3.7218045112781953e-06,
"loss": 0.1171,
"step": 781
},
{
"epoch": 13.964285714285714,
"grad_norm": 14.7429838180542,
"learning_rate": 3.6842105263157892e-06,
"loss": 0.1374,
"step": 782
},
{
"epoch": 13.982142857142858,
"grad_norm": 14.7429838180542,
"learning_rate": 3.6466165413533835e-06,
"loss": 0.1101,
"step": 783
},
{
"epoch": 14.0,
"grad_norm": 16.484983444213867,
"learning_rate": 3.6090225563909775e-06,
"loss": 0.112,
"step": 784
},
{
"epoch": 14.017857142857142,
"grad_norm": 16.484983444213867,
"learning_rate": 3.5714285714285714e-06,
"loss": 0.0927,
"step": 785
},
{
"epoch": 14.035714285714286,
"grad_norm": 16.484983444213867,
"learning_rate": 3.5338345864661653e-06,
"loss": 0.123,
"step": 786
},
{
"epoch": 14.053571428571429,
"grad_norm": 16.484983444213867,
"learning_rate": 3.496240601503759e-06,
"loss": 0.0987,
"step": 787
},
{
"epoch": 14.071428571428571,
"grad_norm": 180.909912109375,
"learning_rate": 3.458646616541353e-06,
"loss": 0.1222,
"step": 788
},
{
"epoch": 14.089285714285714,
"grad_norm": 180.909912109375,
"learning_rate": 3.4210526315789474e-06,
"loss": 0.1117,
"step": 789
},
{
"epoch": 14.107142857142858,
"grad_norm": 180.909912109375,
"learning_rate": 3.3834586466165413e-06,
"loss": 0.118,
"step": 790
},
{
"epoch": 14.125,
"grad_norm": 180.909912109375,
"learning_rate": 3.3458646616541352e-06,
"loss": 0.1298,
"step": 791
},
{
"epoch": 14.142857142857142,
"grad_norm": 49.04324722290039,
"learning_rate": 3.308270676691729e-06,
"loss": 0.097,
"step": 792
},
{
"epoch": 14.160714285714286,
"grad_norm": 49.04324722290039,
"learning_rate": 3.270676691729323e-06,
"loss": 0.1227,
"step": 793
},
{
"epoch": 14.178571428571429,
"grad_norm": 49.04324722290039,
"learning_rate": 3.2330827067669174e-06,
"loss": 0.0954,
"step": 794
},
{
"epoch": 14.196428571428571,
"grad_norm": 49.04324722290039,
"learning_rate": 3.1954887218045113e-06,
"loss": 0.083,
"step": 795
},
{
"epoch": 14.214285714285714,
"grad_norm": 8.683980941772461,
"learning_rate": 3.157894736842105e-06,
"loss": 0.0824,
"step": 796
},
{
"epoch": 14.232142857142858,
"grad_norm": 8.683980941772461,
"learning_rate": 3.120300751879699e-06,
"loss": 0.1248,
"step": 797
},
{
"epoch": 14.25,
"grad_norm": 8.683980941772461,
"learning_rate": 3.082706766917293e-06,
"loss": 0.0967,
"step": 798
},
{
"epoch": 14.267857142857142,
"grad_norm": 8.683980941772461,
"learning_rate": 3.0451127819548873e-06,
"loss": 0.099,
"step": 799
},
{
"epoch": 14.285714285714286,
"grad_norm": 8.626225471496582,
"learning_rate": 3.007518796992481e-06,
"loss": 0.1285,
"step": 800
},
{
"epoch": 14.303571428571429,
"grad_norm": 8.626225471496582,
"learning_rate": 2.969924812030075e-06,
"loss": 0.1077,
"step": 801
},
{
"epoch": 14.321428571428571,
"grad_norm": 8.626225471496582,
"learning_rate": 2.932330827067669e-06,
"loss": 0.1203,
"step": 802
},
{
"epoch": 14.339285714285714,
"grad_norm": 8.626225471496582,
"learning_rate": 2.894736842105263e-06,
"loss": 0.0806,
"step": 803
},
{
"epoch": 14.357142857142858,
"grad_norm": 13.891680717468262,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.0968,
"step": 804
},
{
"epoch": 14.375,
"grad_norm": 13.891680717468262,
"learning_rate": 2.819548872180451e-06,
"loss": 0.0964,
"step": 805
},
{
"epoch": 14.392857142857142,
"grad_norm": 13.891680717468262,
"learning_rate": 2.781954887218045e-06,
"loss": 0.119,
"step": 806
},
{
"epoch": 14.410714285714286,
"grad_norm": 13.891680717468262,
"learning_rate": 2.744360902255639e-06,
"loss": 0.1079,
"step": 807
},
{
"epoch": 14.428571428571429,
"grad_norm": 9.656736373901367,
"learning_rate": 2.706766917293233e-06,
"loss": 0.139,
"step": 808
},
{
"epoch": 14.446428571428571,
"grad_norm": 9.656736373901367,
"learning_rate": 2.6691729323308268e-06,
"loss": 0.1294,
"step": 809
},
{
"epoch": 14.464285714285714,
"grad_norm": 9.656736373901367,
"learning_rate": 2.631578947368421e-06,
"loss": 0.1195,
"step": 810
},
{
"epoch": 14.482142857142858,
"grad_norm": 9.656736373901367,
"learning_rate": 2.593984962406015e-06,
"loss": 0.1103,
"step": 811
},
{
"epoch": 14.5,
"grad_norm": 13.92542552947998,
"learning_rate": 2.556390977443609e-06,
"loss": 0.091,
"step": 812
},
{
"epoch": 14.517857142857142,
"grad_norm": 13.92542552947998,
"learning_rate": 2.518796992481203e-06,
"loss": 0.0996,
"step": 813
},
{
"epoch": 14.535714285714286,
"grad_norm": 13.92542552947998,
"learning_rate": 2.4812030075187967e-06,
"loss": 0.1196,
"step": 814
},
{
"epoch": 14.553571428571429,
"grad_norm": 13.92542552947998,
"learning_rate": 2.443609022556391e-06,
"loss": 0.1125,
"step": 815
},
{
"epoch": 14.571428571428571,
"grad_norm": 9.114486694335938,
"learning_rate": 2.406015037593985e-06,
"loss": 0.117,
"step": 816
},
{
"epoch": 14.589285714285714,
"grad_norm": 9.114486694335938,
"learning_rate": 2.368421052631579e-06,
"loss": 0.1029,
"step": 817
},
{
"epoch": 14.607142857142858,
"grad_norm": 9.114486694335938,
"learning_rate": 2.3308270676691728e-06,
"loss": 0.0963,
"step": 818
},
{
"epoch": 14.625,
"grad_norm": 9.114486694335938,
"learning_rate": 2.2932330827067667e-06,
"loss": 0.1191,
"step": 819
},
{
"epoch": 14.642857142857142,
"grad_norm": 6.985568046569824,
"learning_rate": 2.255639097744361e-06,
"loss": 0.0824,
"step": 820
},
{
"epoch": 14.660714285714286,
"grad_norm": 6.985568046569824,
"learning_rate": 2.218045112781955e-06,
"loss": 0.1007,
"step": 821
},
{
"epoch": 14.678571428571429,
"grad_norm": 6.985568046569824,
"learning_rate": 2.180451127819549e-06,
"loss": 0.0988,
"step": 822
},
{
"epoch": 14.696428571428571,
"grad_norm": 6.985568046569824,
"learning_rate": 2.1428571428571427e-06,
"loss": 0.1135,
"step": 823
},
{
"epoch": 14.714285714285714,
"grad_norm": 7.672319412231445,
"learning_rate": 2.1052631578947366e-06,
"loss": 0.1125,
"step": 824
},
{
"epoch": 14.732142857142858,
"grad_norm": 7.672319412231445,
"learning_rate": 2.067669172932331e-06,
"loss": 0.119,
"step": 825
},
{
"epoch": 14.75,
"grad_norm": 7.672319412231445,
"learning_rate": 2.030075187969925e-06,
"loss": 0.1089,
"step": 826
},
{
"epoch": 14.767857142857142,
"grad_norm": 7.672319412231445,
"learning_rate": 1.9924812030075188e-06,
"loss": 0.1061,
"step": 827
},
{
"epoch": 14.785714285714286,
"grad_norm": 5.853728294372559,
"learning_rate": 1.9548872180451127e-06,
"loss": 0.0846,
"step": 828
},
{
"epoch": 14.803571428571429,
"grad_norm": 5.853728294372559,
"learning_rate": 1.9172932330827066e-06,
"loss": 0.0833,
"step": 829
},
{
"epoch": 14.821428571428571,
"grad_norm": 5.853728294372559,
"learning_rate": 1.8796992481203007e-06,
"loss": 0.0817,
"step": 830
},
{
"epoch": 14.839285714285714,
"grad_norm": 5.853728294372559,
"learning_rate": 1.8421052631578946e-06,
"loss": 0.0775,
"step": 831
},
{
"epoch": 14.857142857142858,
"grad_norm": 8.482511520385742,
"learning_rate": 1.8045112781954887e-06,
"loss": 0.1402,
"step": 832
},
{
"epoch": 14.875,
"grad_norm": 8.482511520385742,
"learning_rate": 1.7669172932330826e-06,
"loss": 0.1064,
"step": 833
},
{
"epoch": 14.892857142857142,
"grad_norm": 8.482511520385742,
"learning_rate": 1.7293233082706765e-06,
"loss": 0.0926,
"step": 834
},
{
"epoch": 14.910714285714286,
"grad_norm": 8.482511520385742,
"learning_rate": 1.6917293233082707e-06,
"loss": 0.0972,
"step": 835
},
{
"epoch": 14.928571428571429,
"grad_norm": 5.480091094970703,
"learning_rate": 1.6541353383458646e-06,
"loss": 0.0951,
"step": 836
},
{
"epoch": 14.946428571428571,
"grad_norm": 5.480091094970703,
"learning_rate": 1.6165413533834587e-06,
"loss": 0.1108,
"step": 837
},
{
"epoch": 14.964285714285714,
"grad_norm": 5.480091094970703,
"learning_rate": 1.5789473684210526e-06,
"loss": 0.1028,
"step": 838
},
{
"epoch": 14.982142857142858,
"grad_norm": 5.480091094970703,
"learning_rate": 1.5413533834586465e-06,
"loss": 0.1665,
"step": 839
},
{
"epoch": 15.0,
"grad_norm": 6.169451713562012,
"learning_rate": 1.5037593984962406e-06,
"loss": 0.0851,
"step": 840
},
{
"epoch": 15.0,
"step": 840,
"total_flos": 3.057064373701837e+17,
"train_loss": 0.8023447536286854,
"train_runtime": 2633.5405,
"train_samples_per_second": 10.156,
"train_steps_per_second": 0.319
}
],
"logging_steps": 1.0,
"max_steps": 840,
"num_input_tokens_seen": 0,
"num_train_epochs": 15,
"save_steps": 400,
"total_flos": 3.057064373701837e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}