Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +791 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,791 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.935185185185185,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1073,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.09259259259259259,
|
| 13 |
+
"grad_norm": 4.0609588623046875,
|
| 14 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 15 |
+
"loss": 0.9653,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.18518518518518517,
|
| 20 |
+
"grad_norm": 3.2480485439300537,
|
| 21 |
+
"learning_rate": 7.407407407407407e-05,
|
| 22 |
+
"loss": 0.4377,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.2777777777777778,
|
| 27 |
+
"grad_norm": 1.2286500930786133,
|
| 28 |
+
"learning_rate": 0.00011111111111111112,
|
| 29 |
+
"loss": 0.2765,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.37037037037037035,
|
| 34 |
+
"grad_norm": 1.0655648708343506,
|
| 35 |
+
"learning_rate": 0.00014814814814814815,
|
| 36 |
+
"loss": 0.2268,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.46296296296296297,
|
| 41 |
+
"grad_norm": 0.9578745365142822,
|
| 42 |
+
"learning_rate": 0.0001851851851851852,
|
| 43 |
+
"loss": 0.1874,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.5555555555555556,
|
| 48 |
+
"grad_norm": 1.0437668561935425,
|
| 49 |
+
"learning_rate": 0.00019998289151715887,
|
| 50 |
+
"loss": 0.1747,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.6481481481481481,
|
| 55 |
+
"grad_norm": 1.0130860805511475,
|
| 56 |
+
"learning_rate": 0.00019987836087650596,
|
| 57 |
+
"loss": 0.1637,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.7407407407407407,
|
| 62 |
+
"grad_norm": 0.8844163417816162,
|
| 63 |
+
"learning_rate": 0.0001996789035357663,
|
| 64 |
+
"loss": 0.1493,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.8333333333333334,
|
| 69 |
+
"grad_norm": 0.7460007667541504,
|
| 70 |
+
"learning_rate": 0.00019938470906380134,
|
| 71 |
+
"loss": 0.1391,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.9259259259259259,
|
| 76 |
+
"grad_norm": 0.5707225203514099,
|
| 77 |
+
"learning_rate": 0.0001989960570698294,
|
| 78 |
+
"loss": 0.1181,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.0185185185185186,
|
| 83 |
+
"grad_norm": 0.5503486394882202,
|
| 84 |
+
"learning_rate": 0.00019851331693767843,
|
| 85 |
+
"loss": 0.1248,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.1111111111111112,
|
| 90 |
+
"grad_norm": 0.5087465643882751,
|
| 91 |
+
"learning_rate": 0.00019793694747471515,
|
| 92 |
+
"loss": 0.1145,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.2037037037037037,
|
| 97 |
+
"grad_norm": 0.6456828713417053,
|
| 98 |
+
"learning_rate": 0.00019726749647578387,
|
| 99 |
+
"loss": 0.1065,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.2962962962962963,
|
| 104 |
+
"grad_norm": 2.2336721420288086,
|
| 105 |
+
"learning_rate": 0.00019650560020256975,
|
| 106 |
+
"loss": 0.1151,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.3888888888888888,
|
| 111 |
+
"grad_norm": 2.1042354106903076,
|
| 112 |
+
"learning_rate": 0.00019565198277888085,
|
| 113 |
+
"loss": 0.1096,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.4814814814814814,
|
| 118 |
+
"grad_norm": 3.438588857650757,
|
| 119 |
+
"learning_rate": 0.0001947074555024243,
|
| 120 |
+
"loss": 0.1086,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.574074074074074,
|
| 125 |
+
"grad_norm": 0.5174661874771118,
|
| 126 |
+
"learning_rate": 0.0001936729160737301,
|
| 127 |
+
"loss": 0.1047,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.6666666666666665,
|
| 132 |
+
"grad_norm": 0.5018674731254578,
|
| 133 |
+
"learning_rate": 0.00019254934774295588,
|
| 134 |
+
"loss": 0.0907,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.7592592592592593,
|
| 139 |
+
"grad_norm": 0.47620534896850586,
|
| 140 |
+
"learning_rate": 0.00019133781837538322,
|
| 141 |
+
"loss": 0.0939,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.8518518518518519,
|
| 146 |
+
"grad_norm": 0.4984002709388733,
|
| 147 |
+
"learning_rate": 0.00019003947943649387,
|
| 148 |
+
"loss": 0.09,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.9444444444444444,
|
| 153 |
+
"grad_norm": 0.5507371425628662,
|
| 154 |
+
"learning_rate": 0.00018865556489759038,
|
| 155 |
+
"loss": 0.0874,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.037037037037037,
|
| 160 |
+
"grad_norm": 0.5186899304389954,
|
| 161 |
+
"learning_rate": 0.0001871873900630011,
|
| 162 |
+
"loss": 0.0891,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.1296296296296298,
|
| 167 |
+
"grad_norm": 0.32078972458839417,
|
| 168 |
+
"learning_rate": 0.00018563635031998497,
|
| 169 |
+
"loss": 0.0816,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.2222222222222223,
|
| 174 |
+
"grad_norm": 0.35380446910858154,
|
| 175 |
+
"learning_rate": 0.00018400391981252277,
|
| 176 |
+
"loss": 0.0902,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.314814814814815,
|
| 181 |
+
"grad_norm": 0.3749984800815582,
|
| 182 |
+
"learning_rate": 0.00018229165004025712,
|
| 183 |
+
"loss": 0.0773,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.4074074074074074,
|
| 188 |
+
"grad_norm": 0.9008009433746338,
|
| 189 |
+
"learning_rate": 0.000180501168383911,
|
| 190 |
+
"loss": 0.0853,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 2.5,
|
| 195 |
+
"grad_norm": 0.6653043627738953,
|
| 196 |
+
"learning_rate": 0.00017863417655858776,
|
| 197 |
+
"loss": 0.0796,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 2.5925925925925926,
|
| 202 |
+
"grad_norm": 0.5696312189102173,
|
| 203 |
+
"learning_rate": 0.00017669244899642182,
|
| 204 |
+
"loss": 0.0773,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 2.685185185185185,
|
| 209 |
+
"grad_norm": 0.49613088369369507,
|
| 210 |
+
"learning_rate": 0.00017467783116011722,
|
| 211 |
+
"loss": 0.0755,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.7777777777777777,
|
| 216 |
+
"grad_norm": 0.37407752871513367,
|
| 217 |
+
"learning_rate": 0.00017259223778897748,
|
| 218 |
+
"loss": 0.0795,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 2.8703703703703702,
|
| 223 |
+
"grad_norm": 0.19830532371997833,
|
| 224 |
+
"learning_rate": 0.00017043765107909296,
|
| 225 |
+
"loss": 0.0793,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 2.962962962962963,
|
| 230 |
+
"grad_norm": 0.34773755073547363,
|
| 231 |
+
"learning_rate": 0.0001682161187994161,
|
| 232 |
+
"loss": 0.0642,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.0555555555555554,
|
| 237 |
+
"grad_norm": 0.4124683439731598,
|
| 238 |
+
"learning_rate": 0.00016592975234551438,
|
| 239 |
+
"loss": 0.0751,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.148148148148148,
|
| 244 |
+
"grad_norm": 0.321125328540802,
|
| 245 |
+
"learning_rate": 0.0001635807247328514,
|
| 246 |
+
"loss": 0.0788,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 3.240740740740741,
|
| 251 |
+
"grad_norm": 0.34814882278442383,
|
| 252 |
+
"learning_rate": 0.00016117126853150264,
|
| 253 |
+
"loss": 0.073,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 3.3333333333333335,
|
| 258 |
+
"grad_norm": 0.461489737033844,
|
| 259 |
+
"learning_rate": 0.00015870367374426936,
|
| 260 |
+
"loss": 0.0788,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 3.425925925925926,
|
| 265 |
+
"grad_norm": 0.32594698667526245,
|
| 266 |
+
"learning_rate": 0.00015618028563020727,
|
| 267 |
+
"loss": 0.065,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 3.5185185185185186,
|
| 272 |
+
"grad_norm": 0.311172753572464,
|
| 273 |
+
"learning_rate": 0.00015360350247563774,
|
| 274 |
+
"loss": 0.0611,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 3.611111111111111,
|
| 279 |
+
"grad_norm": 0.4604165852069855,
|
| 280 |
+
"learning_rate": 0.00015097577331476164,
|
| 281 |
+
"loss": 0.063,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 3.7037037037037037,
|
| 286 |
+
"grad_norm": 0.42682307958602905,
|
| 287 |
+
"learning_rate": 0.00014829959560204035,
|
| 288 |
+
"loss": 0.0614,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 3.7962962962962963,
|
| 293 |
+
"grad_norm": 0.39028409123420715,
|
| 294 |
+
"learning_rate": 0.00014557751283855735,
|
| 295 |
+
"loss": 0.0651,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 3.888888888888889,
|
| 300 |
+
"grad_norm": 0.3820561170578003,
|
| 301 |
+
"learning_rate": 0.0001428121121546163,
|
| 302 |
+
"loss": 0.0589,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 3.9814814814814814,
|
| 307 |
+
"grad_norm": 0.31287339329719543,
|
| 308 |
+
"learning_rate": 0.00014000602185087254,
|
| 309 |
+
"loss": 0.0633,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 4.074074074074074,
|
| 314 |
+
"grad_norm": 0.42882245779037476,
|
| 315 |
+
"learning_rate": 0.00013716190890033543,
|
| 316 |
+
"loss": 0.0606,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 4.166666666666667,
|
| 321 |
+
"grad_norm": 0.3459970951080322,
|
| 322 |
+
"learning_rate": 0.00013428247641361572,
|
| 323 |
+
"loss": 0.0532,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 4.2592592592592595,
|
| 328 |
+
"grad_norm": 0.3979005217552185,
|
| 329 |
+
"learning_rate": 0.00013137046106982683,
|
| 330 |
+
"loss": 0.0547,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 4.351851851851852,
|
| 335 |
+
"grad_norm": 0.22608621418476105,
|
| 336 |
+
"learning_rate": 0.00012842863051558168,
|
| 337 |
+
"loss": 0.0507,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 4.444444444444445,
|
| 342 |
+
"grad_norm": 0.3078368604183197,
|
| 343 |
+
"learning_rate": 0.00012545978073455754,
|
| 344 |
+
"loss": 0.0614,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 4.537037037037037,
|
| 349 |
+
"grad_norm": 0.3212282657623291,
|
| 350 |
+
"learning_rate": 0.0001224667333901287,
|
| 351 |
+
"loss": 0.0584,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 4.62962962962963,
|
| 356 |
+
"grad_norm": 0.3688996136188507,
|
| 357 |
+
"learning_rate": 0.0001194523331435924,
|
| 358 |
+
"loss": 0.0604,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 4.722222222222222,
|
| 363 |
+
"grad_norm": 0.2440503090620041,
|
| 364 |
+
"learning_rate": 0.00011641944495053689,
|
| 365 |
+
"loss": 0.0536,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 4.814814814814815,
|
| 370 |
+
"grad_norm": 0.29373106360435486,
|
| 371 |
+
"learning_rate": 0.00011337095133792168,
|
| 372 |
+
"loss": 0.0517,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 4.907407407407407,
|
| 377 |
+
"grad_norm": 0.28415507078170776,
|
| 378 |
+
"learning_rate": 0.00011030974966445735,
|
| 379 |
+
"loss": 0.0483,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 5.0,
|
| 384 |
+
"grad_norm": 0.6739519238471985,
|
| 385 |
+
"learning_rate": 0.00010723874936688885,
|
| 386 |
+
"loss": 0.0597,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 5.092592592592593,
|
| 391 |
+
"grad_norm": 0.3474346101284027,
|
| 392 |
+
"learning_rate": 0.00010416086919479975,
|
| 393 |
+
"loss": 0.0468,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 5.185185185185185,
|
| 398 |
+
"grad_norm": 0.3007705807685852,
|
| 399 |
+
"learning_rate": 0.00010107903443656521,
|
| 400 |
+
"loss": 0.0546,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 5.277777777777778,
|
| 405 |
+
"grad_norm": 0.35039541125297546,
|
| 406 |
+
"learning_rate": 9.799617413909027e-05,
|
| 407 |
+
"loss": 0.0548,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 5.37037037037037,
|
| 412 |
+
"grad_norm": 0.2516862154006958,
|
| 413 |
+
"learning_rate": 9.491521832397619e-05,
|
| 414 |
+
"loss": 0.0542,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 5.462962962962963,
|
| 419 |
+
"grad_norm": 0.3877411484718323,
|
| 420 |
+
"learning_rate": 9.183909520276032e-05,
|
| 421 |
+
"loss": 0.0481,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 5.555555555555555,
|
| 426 |
+
"grad_norm": 0.3625725507736206,
|
| 427 |
+
"learning_rate": 8.877072839387602e-05,
|
| 428 |
+
"loss": 0.0474,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 5.648148148148148,
|
| 433 |
+
"grad_norm": 0.2881794273853302,
|
| 434 |
+
"learning_rate": 8.57130341439784e-05,
|
| 435 |
+
"loss": 0.0513,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 5.7407407407407405,
|
| 440 |
+
"grad_norm": 0.2732185125350952,
|
| 441 |
+
"learning_rate": 8.266891855627611e-05,
|
| 442 |
+
"loss": 0.0445,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 5.833333333333333,
|
| 447 |
+
"grad_norm": 0.34640687704086304,
|
| 448 |
+
"learning_rate": 7.964127482850372e-05,
|
| 449 |
+
"loss": 0.0449,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 5.925925925925926,
|
| 454 |
+
"grad_norm": 0.3285965919494629,
|
| 455 |
+
"learning_rate": 7.663298050315993e-05,
|
| 456 |
+
"loss": 0.0484,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 6.018518518518518,
|
| 461 |
+
"grad_norm": 0.28891757130622864,
|
| 462 |
+
"learning_rate": 7.364689473262452e-05,
|
| 463 |
+
"loss": 0.0521,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 6.111111111111111,
|
| 468 |
+
"grad_norm": 0.34271323680877686,
|
| 469 |
+
"learning_rate": 7.068585556175412e-05,
|
| 470 |
+
"loss": 0.0463,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 6.203703703703703,
|
| 475 |
+
"grad_norm": 0.2897927463054657,
|
| 476 |
+
"learning_rate": 6.77526772305388e-05,
|
| 477 |
+
"loss": 0.0441,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 6.296296296296296,
|
| 482 |
+
"grad_norm": 0.20959234237670898,
|
| 483 |
+
"learning_rate": 6.485014749938336e-05,
|
| 484 |
+
"loss": 0.0379,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 6.388888888888889,
|
| 489 |
+
"grad_norm": 0.26397034525871277,
|
| 490 |
+
"learning_rate": 6.198102499955545e-05,
|
| 491 |
+
"loss": 0.0408,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 6.481481481481482,
|
| 496 |
+
"grad_norm": 0.28571462631225586,
|
| 497 |
+
"learning_rate": 5.914803661131895e-05,
|
| 498 |
+
"loss": 0.0377,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 6.574074074074074,
|
| 503 |
+
"grad_norm": 0.17670756578445435,
|
| 504 |
+
"learning_rate": 5.6353874872243716e-05,
|
| 505 |
+
"loss": 0.04,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 6.666666666666667,
|
| 510 |
+
"grad_norm": 0.2153652012348175,
|
| 511 |
+
"learning_rate": 5.360119541815587e-05,
|
| 512 |
+
"loss": 0.0438,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 6.7592592592592595,
|
| 517 |
+
"grad_norm": 0.20676399767398834,
|
| 518 |
+
"learning_rate": 5.089261445916018e-05,
|
| 519 |
+
"loss": 0.0433,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 6.851851851851852,
|
| 524 |
+
"grad_norm": 0.2535192370414734,
|
| 525 |
+
"learning_rate": 4.8230706293133656e-05,
|
| 526 |
+
"loss": 0.0429,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 6.944444444444445,
|
| 531 |
+
"grad_norm": 0.21110829710960388,
|
| 532 |
+
"learning_rate": 4.561800085905312e-05,
|
| 533 |
+
"loss": 0.0381,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.037037037037037,
|
| 538 |
+
"grad_norm": 0.24194058775901794,
|
| 539 |
+
"learning_rate": 4.305698133248319e-05,
|
| 540 |
+
"loss": 0.0456,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.12962962962963,
|
| 545 |
+
"grad_norm": 0.21028666198253632,
|
| 546 |
+
"learning_rate": 4.055008176550869e-05,
|
| 547 |
+
"loss": 0.0385,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.222222222222222,
|
| 552 |
+
"grad_norm": 0.2054344266653061,
|
| 553 |
+
"learning_rate": 3.8099684773355196e-05,
|
| 554 |
+
"loss": 0.0409,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.314814814814815,
|
| 559 |
+
"grad_norm": 0.23474320769309998,
|
| 560 |
+
"learning_rate": 3.570811926989689e-05,
|
| 561 |
+
"loss": 0.0343,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.407407407407407,
|
| 566 |
+
"grad_norm": 0.1761438399553299,
|
| 567 |
+
"learning_rate": 3.337765825420277e-05,
|
| 568 |
+
"loss": 0.0346,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.5,
|
| 573 |
+
"grad_norm": 0.1905253529548645,
|
| 574 |
+
"learning_rate": 3.111051665022584e-05,
|
| 575 |
+
"loss": 0.0318,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 7.592592592592593,
|
| 580 |
+
"grad_norm": 0.32174453139305115,
|
| 581 |
+
"learning_rate": 2.890884920168856e-05,
|
| 582 |
+
"loss": 0.031,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 7.685185185185185,
|
| 587 |
+
"grad_norm": 0.18767257034778595,
|
| 588 |
+
"learning_rate": 2.6774748424164207e-05,
|
| 589 |
+
"loss": 0.0391,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 7.777777777777778,
|
| 594 |
+
"grad_norm": 0.22317418456077576,
|
| 595 |
+
"learning_rate": 2.471024261630215e-05,
|
| 596 |
+
"loss": 0.0323,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 7.87037037037037,
|
| 601 |
+
"grad_norm": 0.25181177258491516,
|
| 602 |
+
"learning_rate": 2.2717293932085827e-05,
|
| 603 |
+
"loss": 0.039,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 7.962962962962963,
|
| 608 |
+
"grad_norm": 0.1966754049062729,
|
| 609 |
+
"learning_rate": 2.0797796515956237e-05,
|
| 610 |
+
"loss": 0.0326,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 8.055555555555555,
|
| 615 |
+
"grad_norm": 0.1419260948896408,
|
| 616 |
+
"learning_rate": 1.895357470257345e-05,
|
| 617 |
+
"loss": 0.037,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 8.148148148148149,
|
| 622 |
+
"grad_norm": 0.16165828704833984,
|
| 623 |
+
"learning_rate": 1.7186381282926833e-05,
|
| 624 |
+
"loss": 0.0313,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 8.24074074074074,
|
| 629 |
+
"grad_norm": 0.17430561780929565,
|
| 630 |
+
"learning_rate": 1.549789583844192e-05,
|
| 631 |
+
"loss": 0.0357,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 8.333333333333334,
|
| 636 |
+
"grad_norm": 0.2748219668865204,
|
| 637 |
+
"learning_rate": 1.3889723144667477e-05,
|
| 638 |
+
"loss": 0.0378,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 8.425925925925926,
|
| 643 |
+
"grad_norm": 0.27772700786590576,
|
| 644 |
+
"learning_rate": 1.2363391646059886e-05,
|
| 645 |
+
"loss": 0.0354,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 8.518518518518519,
|
| 650 |
+
"grad_norm": 0.2717374563217163,
|
| 651 |
+
"learning_rate": 1.0920352003314105e-05,
|
| 652 |
+
"loss": 0.032,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 8.61111111111111,
|
| 657 |
+
"grad_norm": 0.217156782746315,
|
| 658 |
+
"learning_rate": 9.561975714622295e-06,
|
| 659 |
+
"loss": 0.03,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 8.703703703703704,
|
| 664 |
+
"grad_norm": 0.2336784154176712,
|
| 665 |
+
"learning_rate": 8.289553812170326e-06,
|
| 666 |
+
"loss": 0.0334,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 8.796296296296296,
|
| 671 |
+
"grad_norm": 0.20166803896427155,
|
| 672 |
+
"learning_rate": 7.104295635110936e-06,
|
| 673 |
+
"loss": 0.0298,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 8.88888888888889,
|
| 678 |
+
"grad_norm": 0.16439595818519592,
|
| 679 |
+
"learning_rate": 6.00732768017982e-06,
|
| 680 |
+
"loss": 0.0334,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 8.981481481481481,
|
| 685 |
+
"grad_norm": 0.18842309713363647,
|
| 686 |
+
"learning_rate": 4.999692531047307e-06,
|
| 687 |
+
"loss": 0.0344,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 9.074074074074074,
|
| 692 |
+
"grad_norm": 0.18367870151996613,
|
| 693 |
+
"learning_rate": 4.082347867422553e-06,
|
| 694 |
+
"loss": 0.0337,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 9.166666666666666,
|
| 699 |
+
"grad_norm": 0.2120891809463501,
|
| 700 |
+
"learning_rate": 3.2561655548527636e-06,
|
| 701 |
+
"loss": 0.0309,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 9.25925925925926,
|
| 706 |
+
"grad_norm": 0.27212539315223694,
|
| 707 |
+
"learning_rate": 2.521930816082063e-06,
|
| 708 |
+
"loss": 0.0375,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 9.351851851851851,
|
| 713 |
+
"grad_norm": 0.16872258484363556,
|
| 714 |
+
"learning_rate": 1.8803414847575417e-06,
|
| 715 |
+
"loss": 0.0255,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 9.444444444444445,
|
| 720 |
+
"grad_norm": 0.1719067543745041,
|
| 721 |
+
"learning_rate": 1.3320073421921986e-06,
|
| 722 |
+
"loss": 0.0336,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 9.537037037037036,
|
| 727 |
+
"grad_norm": 0.16031451523303986,
|
| 728 |
+
"learning_rate": 8.77449537814623e-07,
|
| 729 |
+
"loss": 0.0324,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 9.62962962962963,
|
| 734 |
+
"grad_norm": 0.15033039450645447,
|
| 735 |
+
"learning_rate": 5.171000938565484e-07,
|
| 736 |
+
"loss": 0.0284,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 9.722222222222221,
|
| 741 |
+
"grad_norm": 0.1727823168039322,
|
| 742 |
+
"learning_rate": 2.513014947489678e-07,
|
| 743 |
+
"loss": 0.0263,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 9.814814814814815,
|
| 748 |
+
"grad_norm": 0.14446181058883667,
|
| 749 |
+
"learning_rate": 8.030636161701078e-08,
|
| 750 |
+
"loss": 0.029,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 9.907407407407408,
|
| 755 |
+
"grad_norm": 0.3018522262573242,
|
| 756 |
+
"learning_rate": 4.27721218301258e-09,
|
| 757 |
+
"loss": 0.0307,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 9.935185185185185,
|
| 762 |
+
"step": 1073,
|
| 763 |
+
"total_flos": 1.5130009512588355e+17,
|
| 764 |
+
"train_loss": 0.07705365996269678,
|
| 765 |
+
"train_runtime": 1230.2164,
|
| 766 |
+
"train_samples_per_second": 55.821,
|
| 767 |
+
"train_steps_per_second": 0.872
|
| 768 |
+
}
|
| 769 |
+
],
|
| 770 |
+
"logging_steps": 10,
|
| 771 |
+
"max_steps": 1073,
|
| 772 |
+
"num_input_tokens_seen": 0,
|
| 773 |
+
"num_train_epochs": 10,
|
| 774 |
+
"save_steps": 10000,
|
| 775 |
+
"stateful_callbacks": {
|
| 776 |
+
"TrainerControl": {
|
| 777 |
+
"args": {
|
| 778 |
+
"should_epoch_stop": false,
|
| 779 |
+
"should_evaluate": false,
|
| 780 |
+
"should_log": false,
|
| 781 |
+
"should_save": true,
|
| 782 |
+
"should_training_stop": true
|
| 783 |
+
},
|
| 784 |
+
"attributes": {}
|
| 785 |
+
}
|
| 786 |
+
},
|
| 787 |
+
"total_flos": 1.5130009512588355e+17,
|
| 788 |
+
"train_batch_size": 64,
|
| 789 |
+
"trial_name": null,
|
| 790 |
+
"trial_params": null
|
| 791 |
+
}
|