Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +817 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,817 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.144796380090498,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 112,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03619909502262444,
|
| 13 |
+
"grad_norm": 2.867363214492798,
|
| 14 |
+
"learning_rate": 3.7037037037037036e-07,
|
| 15 |
+
"loss": 0.7721,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07239819004524888,
|
| 20 |
+
"grad_norm": 2.795196771621704,
|
| 21 |
+
"learning_rate": 7.407407407407407e-07,
|
| 22 |
+
"loss": 0.7534,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1085972850678733,
|
| 27 |
+
"grad_norm": 2.8518760204315186,
|
| 28 |
+
"learning_rate": 1.111111111111111e-06,
|
| 29 |
+
"loss": 0.7531,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.14479638009049775,
|
| 34 |
+
"grad_norm": 2.7276031970977783,
|
| 35 |
+
"learning_rate": 1.4814814814814815e-06,
|
| 36 |
+
"loss": 0.7604,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.18099547511312217,
|
| 41 |
+
"grad_norm": 2.6922106742858887,
|
| 42 |
+
"learning_rate": 1.8518518518518519e-06,
|
| 43 |
+
"loss": 0.7775,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.2171945701357466,
|
| 48 |
+
"grad_norm": 2.8505051136016846,
|
| 49 |
+
"learning_rate": 2.222222222222222e-06,
|
| 50 |
+
"loss": 0.7689,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.25339366515837103,
|
| 55 |
+
"grad_norm": 2.711665153503418,
|
| 56 |
+
"learning_rate": 2.5925925925925925e-06,
|
| 57 |
+
"loss": 0.7687,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.2895927601809955,
|
| 62 |
+
"grad_norm": 2.7716190814971924,
|
| 63 |
+
"learning_rate": 2.962962962962963e-06,
|
| 64 |
+
"loss": 0.769,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3257918552036199,
|
| 69 |
+
"grad_norm": 2.607780933380127,
|
| 70 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 71 |
+
"loss": 0.7674,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.36199095022624433,
|
| 76 |
+
"grad_norm": 2.305697202682495,
|
| 77 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 78 |
+
"loss": 0.7529,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.39819004524886875,
|
| 83 |
+
"grad_norm": 1.5863313674926758,
|
| 84 |
+
"learning_rate": 4.074074074074074e-06,
|
| 85 |
+
"loss": 0.7259,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.4343891402714932,
|
| 90 |
+
"grad_norm": 1.2682443857192993,
|
| 91 |
+
"learning_rate": 4.444444444444444e-06,
|
| 92 |
+
"loss": 0.7206,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.47058823529411764,
|
| 97 |
+
"grad_norm": 1.010533332824707,
|
| 98 |
+
"learning_rate": 4.814814814814815e-06,
|
| 99 |
+
"loss": 0.6921,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.5067873303167421,
|
| 104 |
+
"grad_norm": 0.9136636257171631,
|
| 105 |
+
"learning_rate": 5.185185185185185e-06,
|
| 106 |
+
"loss": 0.6987,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5429864253393665,
|
| 111 |
+
"grad_norm": 1.275439739227295,
|
| 112 |
+
"learning_rate": 5.555555555555557e-06,
|
| 113 |
+
"loss": 0.6997,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.579185520361991,
|
| 118 |
+
"grad_norm": 1.4839826822280884,
|
| 119 |
+
"learning_rate": 5.925925925925926e-06,
|
| 120 |
+
"loss": 0.6773,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6153846153846154,
|
| 125 |
+
"grad_norm": 1.5695866346359253,
|
| 126 |
+
"learning_rate": 6.296296296296297e-06,
|
| 127 |
+
"loss": 0.6582,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.6515837104072398,
|
| 132 |
+
"grad_norm": 1.6517796516418457,
|
| 133 |
+
"learning_rate": 6.666666666666667e-06,
|
| 134 |
+
"loss": 0.6912,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6877828054298643,
|
| 139 |
+
"grad_norm": 1.3464380502700806,
|
| 140 |
+
"learning_rate": 7.0370370370370375e-06,
|
| 141 |
+
"loss": 0.6636,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.7239819004524887,
|
| 146 |
+
"grad_norm": 1.0865730047225952,
|
| 147 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 148 |
+
"loss": 0.6571,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.7601809954751131,
|
| 153 |
+
"grad_norm": 0.8862206339836121,
|
| 154 |
+
"learning_rate": 7.77777777777778e-06,
|
| 155 |
+
"loss": 0.6583,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7963800904977375,
|
| 160 |
+
"grad_norm": 0.9783180952072144,
|
| 161 |
+
"learning_rate": 8.148148148148148e-06,
|
| 162 |
+
"loss": 0.6688,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.832579185520362,
|
| 167 |
+
"grad_norm": 0.7160141468048096,
|
| 168 |
+
"learning_rate": 8.518518518518519e-06,
|
| 169 |
+
"loss": 0.6408,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8687782805429864,
|
| 174 |
+
"grad_norm": 0.6991405487060547,
|
| 175 |
+
"learning_rate": 8.888888888888888e-06,
|
| 176 |
+
"loss": 0.6447,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.9049773755656109,
|
| 181 |
+
"grad_norm": 0.7763178944587708,
|
| 182 |
+
"learning_rate": 9.25925925925926e-06,
|
| 183 |
+
"loss": 0.6512,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.9411764705882353,
|
| 188 |
+
"grad_norm": 0.5905259847640991,
|
| 189 |
+
"learning_rate": 9.62962962962963e-06,
|
| 190 |
+
"loss": 0.6337,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.9773755656108597,
|
| 195 |
+
"grad_norm": 0.7515297532081604,
|
| 196 |
+
"learning_rate": 1e-05,
|
| 197 |
+
"loss": 0.655,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.0361990950226245,
|
| 202 |
+
"grad_norm": 1.260597825050354,
|
| 203 |
+
"learning_rate": 9.999582149277188e-06,
|
| 204 |
+
"loss": 1.2675,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.0723981900452488,
|
| 209 |
+
"grad_norm": 0.5624723434448242,
|
| 210 |
+
"learning_rate": 9.998328666948437e-06,
|
| 211 |
+
"loss": 0.5869,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.1085972850678734,
|
| 216 |
+
"grad_norm": 0.5378224849700928,
|
| 217 |
+
"learning_rate": 9.996239762521152e-06,
|
| 218 |
+
"loss": 0.5919,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.1447963800904977,
|
| 223 |
+
"grad_norm": 0.48451751470565796,
|
| 224 |
+
"learning_rate": 9.993315785135417e-06,
|
| 225 |
+
"loss": 0.5952,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.1809954751131222,
|
| 230 |
+
"grad_norm": 0.4453476071357727,
|
| 231 |
+
"learning_rate": 9.989557223505661e-06,
|
| 232 |
+
"loss": 0.5941,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.2171945701357467,
|
| 237 |
+
"grad_norm": 0.5747421979904175,
|
| 238 |
+
"learning_rate": 9.98496470583896e-06,
|
| 239 |
+
"loss": 0.6023,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.253393665158371,
|
| 244 |
+
"grad_norm": 0.4679132401943207,
|
| 245 |
+
"learning_rate": 9.979538999730047e-06,
|
| 246 |
+
"loss": 0.5768,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.2895927601809956,
|
| 251 |
+
"grad_norm": 0.4853627383708954,
|
| 252 |
+
"learning_rate": 9.973281012033009e-06,
|
| 253 |
+
"loss": 0.5566,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.3257918552036199,
|
| 258 |
+
"grad_norm": 0.44682788848876953,
|
| 259 |
+
"learning_rate": 9.966191788709716e-06,
|
| 260 |
+
"loss": 0.5981,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.3619909502262444,
|
| 265 |
+
"grad_norm": 0.4072805941104889,
|
| 266 |
+
"learning_rate": 9.958272514655006e-06,
|
| 267 |
+
"loss": 0.5664,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.3981900452488687,
|
| 272 |
+
"grad_norm": 0.35199496150016785,
|
| 273 |
+
"learning_rate": 9.949524513498636e-06,
|
| 274 |
+
"loss": 0.5864,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.4343891402714932,
|
| 279 |
+
"grad_norm": 0.420901894569397,
|
| 280 |
+
"learning_rate": 9.939949247384046e-06,
|
| 281 |
+
"loss": 0.5558,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.4705882352941178,
|
| 286 |
+
"grad_norm": 0.39620932936668396,
|
| 287 |
+
"learning_rate": 9.929548316723983e-06,
|
| 288 |
+
"loss": 0.5713,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.506787330316742,
|
| 293 |
+
"grad_norm": 0.36159488558769226,
|
| 294 |
+
"learning_rate": 9.918323459933006e-06,
|
| 295 |
+
"loss": 0.5712,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.5429864253393664,
|
| 300 |
+
"grad_norm": 0.3682953119277954,
|
| 301 |
+
"learning_rate": 9.906276553136924e-06,
|
| 302 |
+
"loss": 0.5543,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.5791855203619911,
|
| 307 |
+
"grad_norm": 0.35126638412475586,
|
| 308 |
+
"learning_rate": 9.893409609859221e-06,
|
| 309 |
+
"loss": 0.5615,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.6153846153846154,
|
| 314 |
+
"grad_norm": 0.38464972376823425,
|
| 315 |
+
"learning_rate": 9.879724780684518e-06,
|
| 316 |
+
"loss": 0.5635,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.6515837104072397,
|
| 321 |
+
"grad_norm": 0.37480273842811584,
|
| 322 |
+
"learning_rate": 9.86522435289912e-06,
|
| 323 |
+
"loss": 0.5618,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.6877828054298643,
|
| 328 |
+
"grad_norm": 0.388458788394928,
|
| 329 |
+
"learning_rate": 9.849910750108718e-06,
|
| 330 |
+
"loss": 0.5635,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.7239819004524888,
|
| 335 |
+
"grad_norm": 0.4044710099697113,
|
| 336 |
+
"learning_rate": 9.833786531833311e-06,
|
| 337 |
+
"loss": 0.5665,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.760180995475113,
|
| 342 |
+
"grad_norm": 0.3484383821487427,
|
| 343 |
+
"learning_rate": 9.816854393079402e-06,
|
| 344 |
+
"loss": 0.5522,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.7963800904977374,
|
| 349 |
+
"grad_norm": 0.33232375979423523,
|
| 350 |
+
"learning_rate": 9.79911716388956e-06,
|
| 351 |
+
"loss": 0.5712,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.8325791855203621,
|
| 356 |
+
"grad_norm": 0.4234112501144409,
|
| 357 |
+
"learning_rate": 9.7805778088694e-06,
|
| 358 |
+
"loss": 0.5496,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.8687782805429864,
|
| 363 |
+
"grad_norm": 0.3586530387401581,
|
| 364 |
+
"learning_rate": 9.761239426692077e-06,
|
| 365 |
+
"loss": 0.5488,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.9049773755656108,
|
| 370 |
+
"grad_norm": 0.34319135546684265,
|
| 371 |
+
"learning_rate": 9.741105249580383e-06,
|
| 372 |
+
"loss": 0.5534,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.9411764705882353,
|
| 377 |
+
"grad_norm": 0.406502366065979,
|
| 378 |
+
"learning_rate": 9.7201786427665e-06,
|
| 379 |
+
"loss": 0.5765,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.9773755656108598,
|
| 384 |
+
"grad_norm": 0.3931889533996582,
|
| 385 |
+
"learning_rate": 9.698463103929542e-06,
|
| 386 |
+
"loss": 0.561,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.0361990950226243,
|
| 391 |
+
"grad_norm": 0.7969871759414673,
|
| 392 |
+
"learning_rate": 9.67596226261095e-06,
|
| 393 |
+
"loss": 0.9762,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.072398190045249,
|
| 398 |
+
"grad_norm": 0.34094685316085815,
|
| 399 |
+
"learning_rate": 9.652679879607843e-06,
|
| 400 |
+
"loss": 0.5055,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.1085972850678734,
|
| 405 |
+
"grad_norm": 0.38012585043907166,
|
| 406 |
+
"learning_rate": 9.628619846344453e-06,
|
| 407 |
+
"loss": 0.4799,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.1447963800904977,
|
| 412 |
+
"grad_norm": 0.3892729580402374,
|
| 413 |
+
"learning_rate": 9.603786184221693e-06,
|
| 414 |
+
"loss": 0.4758,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.180995475113122,
|
| 419 |
+
"grad_norm": 0.47513318061828613,
|
| 420 |
+
"learning_rate": 9.578183043945031e-06,
|
| 421 |
+
"loss": 0.4635,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.2171945701357467,
|
| 426 |
+
"grad_norm": 0.3308075964450836,
|
| 427 |
+
"learning_rate": 9.551814704830734e-06,
|
| 428 |
+
"loss": 0.4698,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.253393665158371,
|
| 433 |
+
"grad_norm": 0.3714766800403595,
|
| 434 |
+
"learning_rate": 9.524685574090627e-06,
|
| 435 |
+
"loss": 0.478,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.2895927601809953,
|
| 440 |
+
"grad_norm": 0.37184762954711914,
|
| 441 |
+
"learning_rate": 9.496800186095466e-06,
|
| 442 |
+
"loss": 0.4802,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.32579185520362,
|
| 447 |
+
"grad_norm": 0.4524374306201935,
|
| 448 |
+
"learning_rate": 9.468163201617063e-06,
|
| 449 |
+
"loss": 0.4848,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.3619909502262444,
|
| 454 |
+
"grad_norm": 0.3513346016407013,
|
| 455 |
+
"learning_rate": 9.438779407049282e-06,
|
| 456 |
+
"loss": 0.4408,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.3981900452488687,
|
| 461 |
+
"grad_norm": 0.3224553167819977,
|
| 462 |
+
"learning_rate": 9.40865371360804e-06,
|
| 463 |
+
"loss": 0.4523,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.4343891402714934,
|
| 468 |
+
"grad_norm": 0.3933435380458832,
|
| 469 |
+
"learning_rate": 9.377791156510456e-06,
|
| 470 |
+
"loss": 0.458,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.4705882352941178,
|
| 475 |
+
"grad_norm": 0.3878864645957947,
|
| 476 |
+
"learning_rate": 9.346196894133239e-06,
|
| 477 |
+
"loss": 0.4732,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.506787330316742,
|
| 482 |
+
"grad_norm": 0.44199082255363464,
|
| 483 |
+
"learning_rate": 9.313876207150544e-06,
|
| 484 |
+
"loss": 0.4412,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.5429864253393664,
|
| 489 |
+
"grad_norm": 0.3837670087814331,
|
| 490 |
+
"learning_rate": 9.280834497651334e-06,
|
| 491 |
+
"loss": 0.4461,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.579185520361991,
|
| 496 |
+
"grad_norm": 0.34122994542121887,
|
| 497 |
+
"learning_rate": 9.247077288236488e-06,
|
| 498 |
+
"loss": 0.4565,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.6153846153846154,
|
| 503 |
+
"grad_norm": 0.344008207321167,
|
| 504 |
+
"learning_rate": 9.212610221095748e-06,
|
| 505 |
+
"loss": 0.4317,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.6515837104072397,
|
| 510 |
+
"grad_norm": 0.3702448606491089,
|
| 511 |
+
"learning_rate": 9.177439057064684e-06,
|
| 512 |
+
"loss": 0.4384,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.6877828054298645,
|
| 517 |
+
"grad_norm": 0.3998362720012665,
|
| 518 |
+
"learning_rate": 9.141569674661816e-06,
|
| 519 |
+
"loss": 0.4646,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.723981900452489,
|
| 524 |
+
"grad_norm": 0.4177471101284027,
|
| 525 |
+
"learning_rate": 9.105008069106093e-06,
|
| 526 |
+
"loss": 0.4555,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.760180995475113,
|
| 531 |
+
"grad_norm": 0.3557385802268982,
|
| 532 |
+
"learning_rate": 9.067760351314838e-06,
|
| 533 |
+
"loss": 0.451,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.7963800904977374,
|
| 538 |
+
"grad_norm": 0.36134663224220276,
|
| 539 |
+
"learning_rate": 9.029832746882372e-06,
|
| 540 |
+
"loss": 0.4339,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.832579185520362,
|
| 545 |
+
"grad_norm": 0.3386276960372925,
|
| 546 |
+
"learning_rate": 8.991231595039464e-06,
|
| 547 |
+
"loss": 0.4272,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.8687782805429864,
|
| 552 |
+
"grad_norm": 0.3196982443332672,
|
| 553 |
+
"learning_rate": 8.951963347593797e-06,
|
| 554 |
+
"loss": 0.4476,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.9049773755656108,
|
| 559 |
+
"grad_norm": 0.3494734764099121,
|
| 560 |
+
"learning_rate": 8.9120345678516e-06,
|
| 561 |
+
"loss": 0.4669,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.9411764705882355,
|
| 566 |
+
"grad_norm": 0.369515597820282,
|
| 567 |
+
"learning_rate": 8.871451929520662e-06,
|
| 568 |
+
"loss": 0.4459,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.97737556561086,
|
| 573 |
+
"grad_norm": 0.3368425965309143,
|
| 574 |
+
"learning_rate": 8.83022221559489e-06,
|
| 575 |
+
"loss": 0.4641,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 3.0361990950226243,
|
| 580 |
+
"grad_norm": 0.778194010257721,
|
| 581 |
+
"learning_rate": 8.78835231722059e-06,
|
| 582 |
+
"loss": 0.8712,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 3.072398190045249,
|
| 587 |
+
"grad_norm": 0.5022688508033752,
|
| 588 |
+
"learning_rate": 8.74584923254468e-06,
|
| 589 |
+
"loss": 0.3798,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.1085972850678734,
|
| 594 |
+
"grad_norm": 0.42150193452835083,
|
| 595 |
+
"learning_rate": 8.702720065545024e-06,
|
| 596 |
+
"loss": 0.3818,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 3.1447963800904977,
|
| 601 |
+
"grad_norm": 0.4261153042316437,
|
| 602 |
+
"learning_rate": 8.658972024843063e-06,
|
| 603 |
+
"loss": 0.3746,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 3.180995475113122,
|
| 608 |
+
"grad_norm": 0.4205586314201355,
|
| 609 |
+
"learning_rate": 8.614612422498965e-06,
|
| 610 |
+
"loss": 0.3523,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.2171945701357467,
|
| 615 |
+
"grad_norm": 0.43494337797164917,
|
| 616 |
+
"learning_rate": 8.569648672789496e-06,
|
| 617 |
+
"loss": 0.3459,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.253393665158371,
|
| 622 |
+
"grad_norm": 0.408480703830719,
|
| 623 |
+
"learning_rate": 8.524088290968781e-06,
|
| 624 |
+
"loss": 0.3362,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.2895927601809953,
|
| 629 |
+
"grad_norm": 0.3939341902732849,
|
| 630 |
+
"learning_rate": 8.477938892012209e-06,
|
| 631 |
+
"loss": 0.3394,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.32579185520362,
|
| 636 |
+
"grad_norm": 0.41535684466362,
|
| 637 |
+
"learning_rate": 8.43120818934367e-06,
|
| 638 |
+
"loss": 0.3434,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.3619909502262444,
|
| 643 |
+
"grad_norm": 0.4341293275356293,
|
| 644 |
+
"learning_rate": 8.38390399354631e-06,
|
| 645 |
+
"loss": 0.3558,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.3981900452488687,
|
| 650 |
+
"grad_norm": 0.3654901683330536,
|
| 651 |
+
"learning_rate": 8.336034211057098e-06,
|
| 652 |
+
"loss": 0.3394,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.4343891402714934,
|
| 657 |
+
"grad_norm": 0.3632822632789612,
|
| 658 |
+
"learning_rate": 8.28760684284532e-06,
|
| 659 |
+
"loss": 0.352,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.4705882352941178,
|
| 664 |
+
"grad_norm": 0.3629053235054016,
|
| 665 |
+
"learning_rate": 8.238629983075296e-06,
|
| 666 |
+
"loss": 0.3314,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.506787330316742,
|
| 671 |
+
"grad_norm": 0.4233010709285736,
|
| 672 |
+
"learning_rate": 8.18911181775353e-06,
|
| 673 |
+
"loss": 0.346,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.5429864253393664,
|
| 678 |
+
"grad_norm": 0.36843806505203247,
|
| 679 |
+
"learning_rate": 8.139060623360494e-06,
|
| 680 |
+
"loss": 0.3608,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.579185520361991,
|
| 685 |
+
"grad_norm": 0.367023229598999,
|
| 686 |
+
"learning_rate": 8.088484765467286e-06,
|
| 687 |
+
"loss": 0.3555,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.6153846153846154,
|
| 692 |
+
"grad_norm": 0.3499699831008911,
|
| 693 |
+
"learning_rate": 8.037392697337418e-06,
|
| 694 |
+
"loss": 0.3274,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.6515837104072397,
|
| 699 |
+
"grad_norm": 0.35652127861976624,
|
| 700 |
+
"learning_rate": 7.985792958513932e-06,
|
| 701 |
+
"loss": 0.3565,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.6877828054298645,
|
| 706 |
+
"grad_norm": 0.38484910130500793,
|
| 707 |
+
"learning_rate": 7.93369417339209e-06,
|
| 708 |
+
"loss": 0.331,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.723981900452489,
|
| 713 |
+
"grad_norm": 0.37475016713142395,
|
| 714 |
+
"learning_rate": 7.881105049777902e-06,
|
| 715 |
+
"loss": 0.3454,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.760180995475113,
|
| 720 |
+
"grad_norm": 0.33727630972862244,
|
| 721 |
+
"learning_rate": 7.828034377432694e-06,
|
| 722 |
+
"loss": 0.3645,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.7963800904977374,
|
| 727 |
+
"grad_norm": 0.3419555723667145,
|
| 728 |
+
"learning_rate": 7.774491026603985e-06,
|
| 729 |
+
"loss": 0.3574,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.832579185520362,
|
| 734 |
+
"grad_norm": 0.5590242147445679,
|
| 735 |
+
"learning_rate": 7.720483946542913e-06,
|
| 736 |
+
"loss": 0.334,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.8687782805429864,
|
| 741 |
+
"grad_norm": 0.36669477820396423,
|
| 742 |
+
"learning_rate": 7.666022164008458e-06,
|
| 743 |
+
"loss": 0.326,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.9049773755656108,
|
| 748 |
+
"grad_norm": 0.3452495038509369,
|
| 749 |
+
"learning_rate": 7.6111147817586925e-06,
|
| 750 |
+
"loss": 0.3542,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.9411764705882355,
|
| 755 |
+
"grad_norm": 0.3405420482158661,
|
| 756 |
+
"learning_rate": 7.5557709770293664e-06,
|
| 757 |
+
"loss": 0.3582,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.97737556561086,
|
| 762 |
+
"grad_norm": 0.3876648545265198,
|
| 763 |
+
"learning_rate": 7.500000000000001e-06,
|
| 764 |
+
"loss": 0.3284,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 4.036199095022624,
|
| 769 |
+
"grad_norm": 0.8777516484260559,
|
| 770 |
+
"learning_rate": 7.443811172247822e-06,
|
| 771 |
+
"loss": 0.5983,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 4.072398190045249,
|
| 776 |
+
"grad_norm": 0.5651780366897583,
|
| 777 |
+
"learning_rate": 7.387213885189746e-06,
|
| 778 |
+
"loss": 0.2736,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 4.108597285067873,
|
| 783 |
+
"grad_norm": 0.4696848392486572,
|
| 784 |
+
"learning_rate": 7.330217598512696e-06,
|
| 785 |
+
"loss": 0.2527,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.144796380090498,
|
| 790 |
+
"grad_norm": 0.41402825713157654,
|
| 791 |
+
"learning_rate": 7.2728318385925035e-06,
|
| 792 |
+
"loss": 0.247,
|
| 793 |
+
"step": 112
|
| 794 |
+
}
|
| 795 |
+
],
|
| 796 |
+
"logging_steps": 1,
|
| 797 |
+
"max_steps": 270,
|
| 798 |
+
"num_input_tokens_seen": 0,
|
| 799 |
+
"num_train_epochs": 10,
|
| 800 |
+
"save_steps": 28,
|
| 801 |
+
"stateful_callbacks": {
|
| 802 |
+
"TrainerControl": {
|
| 803 |
+
"args": {
|
| 804 |
+
"should_epoch_stop": false,
|
| 805 |
+
"should_evaluate": false,
|
| 806 |
+
"should_log": false,
|
| 807 |
+
"should_save": true,
|
| 808 |
+
"should_training_stop": false
|
| 809 |
+
},
|
| 810 |
+
"attributes": {}
|
| 811 |
+
}
|
| 812 |
+
},
|
| 813 |
+
"total_flos": 141386024878080.0,
|
| 814 |
+
"train_batch_size": 1,
|
| 815 |
+
"trial_name": null,
|
| 816 |
+
"trial_params": null
|
| 817 |
+
}
|