Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +795 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,795 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 5000,
|
| 6 |
+
"global_step": 6315,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"learning_rate": 1.9841646872525734e-05,
|
| 14 |
+
"loss": 0.6678,
|
| 15 |
+
"step": 50
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.05,
|
| 19 |
+
"learning_rate": 1.9683293745051467e-05,
|
| 20 |
+
"loss": 0.471,
|
| 21 |
+
"step": 100
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.07,
|
| 25 |
+
"learning_rate": 1.95249406175772e-05,
|
| 26 |
+
"loss": 0.4307,
|
| 27 |
+
"step": 150
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.1,
|
| 31 |
+
"learning_rate": 1.936658749010293e-05,
|
| 32 |
+
"loss": 0.43,
|
| 33 |
+
"step": 200
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.12,
|
| 37 |
+
"learning_rate": 1.9208234362628665e-05,
|
| 38 |
+
"loss": 0.4466,
|
| 39 |
+
"step": 250
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.14,
|
| 43 |
+
"learning_rate": 1.9049881235154395e-05,
|
| 44 |
+
"loss": 0.3861,
|
| 45 |
+
"step": 300
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.17,
|
| 49 |
+
"learning_rate": 1.8891528107680128e-05,
|
| 50 |
+
"loss": 0.3462,
|
| 51 |
+
"step": 350
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.19,
|
| 55 |
+
"learning_rate": 1.873317498020586e-05,
|
| 56 |
+
"loss": 0.3841,
|
| 57 |
+
"step": 400
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.21,
|
| 61 |
+
"learning_rate": 1.8574821852731593e-05,
|
| 62 |
+
"loss": 0.3402,
|
| 63 |
+
"step": 450
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.24,
|
| 67 |
+
"learning_rate": 1.8416468725257326e-05,
|
| 68 |
+
"loss": 0.3153,
|
| 69 |
+
"step": 500
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.26,
|
| 73 |
+
"learning_rate": 1.825811559778306e-05,
|
| 74 |
+
"loss": 0.3443,
|
| 75 |
+
"step": 550
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.29,
|
| 79 |
+
"learning_rate": 1.809976247030879e-05,
|
| 80 |
+
"loss": 0.3269,
|
| 81 |
+
"step": 600
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.31,
|
| 85 |
+
"learning_rate": 1.794140934283452e-05,
|
| 86 |
+
"loss": 0.2908,
|
| 87 |
+
"step": 650
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.33,
|
| 91 |
+
"learning_rate": 1.7783056215360253e-05,
|
| 92 |
+
"loss": 0.3549,
|
| 93 |
+
"step": 700
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.36,
|
| 97 |
+
"learning_rate": 1.762470308788599e-05,
|
| 98 |
+
"loss": 0.31,
|
| 99 |
+
"step": 750
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.38,
|
| 103 |
+
"learning_rate": 1.746634996041172e-05,
|
| 104 |
+
"loss": 0.2744,
|
| 105 |
+
"step": 800
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.4,
|
| 109 |
+
"learning_rate": 1.730799683293745e-05,
|
| 110 |
+
"loss": 0.3012,
|
| 111 |
+
"step": 850
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.43,
|
| 115 |
+
"learning_rate": 1.7149643705463184e-05,
|
| 116 |
+
"loss": 0.2842,
|
| 117 |
+
"step": 900
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.45,
|
| 121 |
+
"learning_rate": 1.6991290577988917e-05,
|
| 122 |
+
"loss": 0.3061,
|
| 123 |
+
"step": 950
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.48,
|
| 127 |
+
"learning_rate": 1.6832937450514647e-05,
|
| 128 |
+
"loss": 0.2939,
|
| 129 |
+
"step": 1000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.5,
|
| 133 |
+
"learning_rate": 1.6674584323040383e-05,
|
| 134 |
+
"loss": 0.3018,
|
| 135 |
+
"step": 1050
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.52,
|
| 139 |
+
"learning_rate": 1.6516231195566115e-05,
|
| 140 |
+
"loss": 0.2574,
|
| 141 |
+
"step": 1100
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.55,
|
| 145 |
+
"learning_rate": 1.6357878068091845e-05,
|
| 146 |
+
"loss": 0.2736,
|
| 147 |
+
"step": 1150
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.57,
|
| 151 |
+
"learning_rate": 1.6199524940617578e-05,
|
| 152 |
+
"loss": 0.257,
|
| 153 |
+
"step": 1200
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.59,
|
| 157 |
+
"learning_rate": 1.604117181314331e-05,
|
| 158 |
+
"loss": 0.2912,
|
| 159 |
+
"step": 1250
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.62,
|
| 163 |
+
"learning_rate": 1.5882818685669043e-05,
|
| 164 |
+
"loss": 0.2788,
|
| 165 |
+
"step": 1300
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.64,
|
| 169 |
+
"learning_rate": 1.5724465558194776e-05,
|
| 170 |
+
"loss": 0.2736,
|
| 171 |
+
"step": 1350
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.67,
|
| 175 |
+
"learning_rate": 1.556611243072051e-05,
|
| 176 |
+
"loss": 0.2993,
|
| 177 |
+
"step": 1400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.69,
|
| 181 |
+
"learning_rate": 1.540775930324624e-05,
|
| 182 |
+
"loss": 0.2691,
|
| 183 |
+
"step": 1450
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.71,
|
| 187 |
+
"learning_rate": 1.5249406175771972e-05,
|
| 188 |
+
"loss": 0.2556,
|
| 189 |
+
"step": 1500
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.74,
|
| 193 |
+
"learning_rate": 1.5091053048297705e-05,
|
| 194 |
+
"loss": 0.2422,
|
| 195 |
+
"step": 1550
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.76,
|
| 199 |
+
"learning_rate": 1.4932699920823436e-05,
|
| 200 |
+
"loss": 0.2496,
|
| 201 |
+
"step": 1600
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.78,
|
| 205 |
+
"learning_rate": 1.477434679334917e-05,
|
| 206 |
+
"loss": 0.229,
|
| 207 |
+
"step": 1650
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.81,
|
| 211 |
+
"learning_rate": 1.4615993665874902e-05,
|
| 212 |
+
"loss": 0.2654,
|
| 213 |
+
"step": 1700
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.83,
|
| 217 |
+
"learning_rate": 1.4457640538400634e-05,
|
| 218 |
+
"loss": 0.2526,
|
| 219 |
+
"step": 1750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.86,
|
| 223 |
+
"learning_rate": 1.4299287410926367e-05,
|
| 224 |
+
"loss": 0.2702,
|
| 225 |
+
"step": 1800
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.88,
|
| 229 |
+
"learning_rate": 1.4140934283452098e-05,
|
| 230 |
+
"loss": 0.255,
|
| 231 |
+
"step": 1850
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.9,
|
| 235 |
+
"learning_rate": 1.3982581155977833e-05,
|
| 236 |
+
"loss": 0.2589,
|
| 237 |
+
"step": 1900
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.93,
|
| 241 |
+
"learning_rate": 1.3824228028503564e-05,
|
| 242 |
+
"loss": 0.2768,
|
| 243 |
+
"step": 1950
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.95,
|
| 247 |
+
"learning_rate": 1.3665874901029297e-05,
|
| 248 |
+
"loss": 0.2335,
|
| 249 |
+
"step": 2000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.97,
|
| 253 |
+
"learning_rate": 1.3507521773555028e-05,
|
| 254 |
+
"loss": 0.2363,
|
| 255 |
+
"step": 2050
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.0,
|
| 259 |
+
"learning_rate": 1.3349168646080762e-05,
|
| 260 |
+
"loss": 0.2501,
|
| 261 |
+
"step": 2100
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.02,
|
| 265 |
+
"learning_rate": 1.3190815518606495e-05,
|
| 266 |
+
"loss": 0.1981,
|
| 267 |
+
"step": 2150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.05,
|
| 271 |
+
"learning_rate": 1.3032462391132226e-05,
|
| 272 |
+
"loss": 0.2133,
|
| 273 |
+
"step": 2200
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.07,
|
| 277 |
+
"learning_rate": 1.2874109263657959e-05,
|
| 278 |
+
"loss": 0.2157,
|
| 279 |
+
"step": 2250
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.09,
|
| 283 |
+
"learning_rate": 1.271575613618369e-05,
|
| 284 |
+
"loss": 0.1893,
|
| 285 |
+
"step": 2300
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.12,
|
| 289 |
+
"learning_rate": 1.2557403008709424e-05,
|
| 290 |
+
"loss": 0.2082,
|
| 291 |
+
"step": 2350
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.14,
|
| 295 |
+
"learning_rate": 1.2399049881235155e-05,
|
| 296 |
+
"loss": 0.1993,
|
| 297 |
+
"step": 2400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.16,
|
| 301 |
+
"learning_rate": 1.2240696753760888e-05,
|
| 302 |
+
"loss": 0.2231,
|
| 303 |
+
"step": 2450
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.19,
|
| 307 |
+
"learning_rate": 1.208234362628662e-05,
|
| 308 |
+
"loss": 0.1788,
|
| 309 |
+
"step": 2500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.21,
|
| 313 |
+
"learning_rate": 1.1923990498812352e-05,
|
| 314 |
+
"loss": 0.1943,
|
| 315 |
+
"step": 2550
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.24,
|
| 319 |
+
"learning_rate": 1.1765637371338086e-05,
|
| 320 |
+
"loss": 0.2029,
|
| 321 |
+
"step": 2600
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.26,
|
| 325 |
+
"learning_rate": 1.1607284243863817e-05,
|
| 326 |
+
"loss": 0.2271,
|
| 327 |
+
"step": 2650
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.28,
|
| 331 |
+
"learning_rate": 1.144893111638955e-05,
|
| 332 |
+
"loss": 0.2009,
|
| 333 |
+
"step": 2700
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.31,
|
| 337 |
+
"learning_rate": 1.1290577988915281e-05,
|
| 338 |
+
"loss": 0.1885,
|
| 339 |
+
"step": 2750
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.33,
|
| 343 |
+
"learning_rate": 1.1132224861441014e-05,
|
| 344 |
+
"loss": 0.2061,
|
| 345 |
+
"step": 2800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.35,
|
| 349 |
+
"learning_rate": 1.0973871733966748e-05,
|
| 350 |
+
"loss": 0.2003,
|
| 351 |
+
"step": 2850
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.38,
|
| 355 |
+
"learning_rate": 1.081551860649248e-05,
|
| 356 |
+
"loss": 0.1973,
|
| 357 |
+
"step": 2900
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.4,
|
| 361 |
+
"learning_rate": 1.0657165479018212e-05,
|
| 362 |
+
"loss": 0.1891,
|
| 363 |
+
"step": 2950
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.43,
|
| 367 |
+
"learning_rate": 1.0498812351543943e-05,
|
| 368 |
+
"loss": 0.2071,
|
| 369 |
+
"step": 3000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.45,
|
| 373 |
+
"learning_rate": 1.0340459224069676e-05,
|
| 374 |
+
"loss": 0.2005,
|
| 375 |
+
"step": 3050
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.47,
|
| 379 |
+
"learning_rate": 1.0182106096595407e-05,
|
| 380 |
+
"loss": 0.1954,
|
| 381 |
+
"step": 3100
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.5,
|
| 385 |
+
"learning_rate": 1.0023752969121141e-05,
|
| 386 |
+
"loss": 0.1913,
|
| 387 |
+
"step": 3150
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.52,
|
| 391 |
+
"learning_rate": 9.865399841646872e-06,
|
| 392 |
+
"loss": 0.2119,
|
| 393 |
+
"step": 3200
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.54,
|
| 397 |
+
"learning_rate": 9.707046714172605e-06,
|
| 398 |
+
"loss": 0.2257,
|
| 399 |
+
"step": 3250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.57,
|
| 403 |
+
"learning_rate": 9.548693586698338e-06,
|
| 404 |
+
"loss": 0.1877,
|
| 405 |
+
"step": 3300
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.59,
|
| 409 |
+
"learning_rate": 9.39034045922407e-06,
|
| 410 |
+
"loss": 0.1965,
|
| 411 |
+
"step": 3350
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.62,
|
| 415 |
+
"learning_rate": 9.231987331749803e-06,
|
| 416 |
+
"loss": 0.1556,
|
| 417 |
+
"step": 3400
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.64,
|
| 421 |
+
"learning_rate": 9.073634204275536e-06,
|
| 422 |
+
"loss": 0.1936,
|
| 423 |
+
"step": 3450
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.66,
|
| 427 |
+
"learning_rate": 8.915281076801267e-06,
|
| 428 |
+
"loss": 0.1704,
|
| 429 |
+
"step": 3500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.69,
|
| 433 |
+
"learning_rate": 8.756927949327e-06,
|
| 434 |
+
"loss": 0.1977,
|
| 435 |
+
"step": 3550
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.71,
|
| 439 |
+
"learning_rate": 8.598574821852733e-06,
|
| 440 |
+
"loss": 0.1724,
|
| 441 |
+
"step": 3600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.73,
|
| 445 |
+
"learning_rate": 8.440221694378464e-06,
|
| 446 |
+
"loss": 0.1683,
|
| 447 |
+
"step": 3650
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.76,
|
| 451 |
+
"learning_rate": 8.281868566904197e-06,
|
| 452 |
+
"loss": 0.1724,
|
| 453 |
+
"step": 3700
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.78,
|
| 457 |
+
"learning_rate": 8.12351543942993e-06,
|
| 458 |
+
"loss": 0.1884,
|
| 459 |
+
"step": 3750
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.81,
|
| 463 |
+
"learning_rate": 7.965162311955662e-06,
|
| 464 |
+
"loss": 0.1831,
|
| 465 |
+
"step": 3800
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.83,
|
| 469 |
+
"learning_rate": 7.806809184481395e-06,
|
| 470 |
+
"loss": 0.2017,
|
| 471 |
+
"step": 3850
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.85,
|
| 475 |
+
"learning_rate": 7.648456057007126e-06,
|
| 476 |
+
"loss": 0.175,
|
| 477 |
+
"step": 3900
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.88,
|
| 481 |
+
"learning_rate": 7.490102929532859e-06,
|
| 482 |
+
"loss": 0.1704,
|
| 483 |
+
"step": 3950
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.9,
|
| 487 |
+
"learning_rate": 7.331749802058591e-06,
|
| 488 |
+
"loss": 0.1768,
|
| 489 |
+
"step": 4000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.92,
|
| 493 |
+
"learning_rate": 7.173396674584323e-06,
|
| 494 |
+
"loss": 0.1704,
|
| 495 |
+
"step": 4050
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.95,
|
| 499 |
+
"learning_rate": 7.015043547110056e-06,
|
| 500 |
+
"loss": 0.1782,
|
| 501 |
+
"step": 4100
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.97,
|
| 505 |
+
"learning_rate": 6.856690419635789e-06,
|
| 506 |
+
"loss": 0.1606,
|
| 507 |
+
"step": 4150
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.0,
|
| 511 |
+
"learning_rate": 6.698337292161521e-06,
|
| 512 |
+
"loss": 0.1752,
|
| 513 |
+
"step": 4200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.02,
|
| 517 |
+
"learning_rate": 6.539984164687253e-06,
|
| 518 |
+
"loss": 0.1277,
|
| 519 |
+
"step": 4250
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.04,
|
| 523 |
+
"learning_rate": 6.3816310372129854e-06,
|
| 524 |
+
"loss": 0.1287,
|
| 525 |
+
"step": 4300
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 2.07,
|
| 529 |
+
"learning_rate": 6.223277909738717e-06,
|
| 530 |
+
"loss": 0.1571,
|
| 531 |
+
"step": 4350
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.09,
|
| 535 |
+
"learning_rate": 6.06492478226445e-06,
|
| 536 |
+
"loss": 0.134,
|
| 537 |
+
"step": 4400
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.11,
|
| 541 |
+
"learning_rate": 5.906571654790183e-06,
|
| 542 |
+
"loss": 0.1333,
|
| 543 |
+
"step": 4450
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.14,
|
| 547 |
+
"learning_rate": 5.748218527315916e-06,
|
| 548 |
+
"loss": 0.1224,
|
| 549 |
+
"step": 4500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.16,
|
| 553 |
+
"learning_rate": 5.5898653998416475e-06,
|
| 554 |
+
"loss": 0.1948,
|
| 555 |
+
"step": 4550
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.19,
|
| 559 |
+
"learning_rate": 5.431512272367379e-06,
|
| 560 |
+
"loss": 0.142,
|
| 561 |
+
"step": 4600
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.21,
|
| 565 |
+
"learning_rate": 5.273159144893112e-06,
|
| 566 |
+
"loss": 0.1536,
|
| 567 |
+
"step": 4650
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 2.23,
|
| 571 |
+
"learning_rate": 5.114806017418844e-06,
|
| 572 |
+
"loss": 0.1283,
|
| 573 |
+
"step": 4700
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.26,
|
| 577 |
+
"learning_rate": 4.956452889944577e-06,
|
| 578 |
+
"loss": 0.1333,
|
| 579 |
+
"step": 4750
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.28,
|
| 583 |
+
"learning_rate": 4.798099762470309e-06,
|
| 584 |
+
"loss": 0.1642,
|
| 585 |
+
"step": 4800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.3,
|
| 589 |
+
"learning_rate": 4.6397466349960415e-06,
|
| 590 |
+
"loss": 0.1254,
|
| 591 |
+
"step": 4850
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 2.33,
|
| 595 |
+
"learning_rate": 4.481393507521774e-06,
|
| 596 |
+
"loss": 0.1602,
|
| 597 |
+
"step": 4900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.35,
|
| 601 |
+
"learning_rate": 4.323040380047506e-06,
|
| 602 |
+
"loss": 0.1618,
|
| 603 |
+
"step": 4950
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 2.38,
|
| 607 |
+
"learning_rate": 4.164687252573238e-06,
|
| 608 |
+
"loss": 0.1187,
|
| 609 |
+
"step": 5000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.38,
|
| 613 |
+
"eval_accuracy": 0.9036697247706422,
|
| 614 |
+
"eval_loss": 0.3797234296798706,
|
| 615 |
+
"eval_runtime": 2.8294,
|
| 616 |
+
"eval_samples_per_second": 308.195,
|
| 617 |
+
"eval_steps_per_second": 19.439,
|
| 618 |
+
"step": 5000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.4,
|
| 622 |
+
"learning_rate": 4.006334125098972e-06,
|
| 623 |
+
"loss": 0.1415,
|
| 624 |
+
"step": 5050
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.42,
|
| 628 |
+
"learning_rate": 3.8479809976247036e-06,
|
| 629 |
+
"loss": 0.1441,
|
| 630 |
+
"step": 5100
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 2.45,
|
| 634 |
+
"learning_rate": 3.6896278701504355e-06,
|
| 635 |
+
"loss": 0.1371,
|
| 636 |
+
"step": 5150
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.47,
|
| 640 |
+
"learning_rate": 3.531274742676168e-06,
|
| 641 |
+
"loss": 0.1262,
|
| 642 |
+
"step": 5200
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.49,
|
| 646 |
+
"learning_rate": 3.3729216152019006e-06,
|
| 647 |
+
"loss": 0.1703,
|
| 648 |
+
"step": 5250
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.52,
|
| 652 |
+
"learning_rate": 3.214568487727633e-06,
|
| 653 |
+
"loss": 0.1291,
|
| 654 |
+
"step": 5300
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 2.54,
|
| 658 |
+
"learning_rate": 3.0562153602533652e-06,
|
| 659 |
+
"loss": 0.1243,
|
| 660 |
+
"step": 5350
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.57,
|
| 664 |
+
"learning_rate": 2.897862232779098e-06,
|
| 665 |
+
"loss": 0.1251,
|
| 666 |
+
"step": 5400
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 2.59,
|
| 670 |
+
"learning_rate": 2.73950910530483e-06,
|
| 671 |
+
"loss": 0.1394,
|
| 672 |
+
"step": 5450
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.61,
|
| 676 |
+
"learning_rate": 2.581155977830562e-06,
|
| 677 |
+
"loss": 0.1505,
|
| 678 |
+
"step": 5500
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.64,
|
| 682 |
+
"learning_rate": 2.422802850356295e-06,
|
| 683 |
+
"loss": 0.1017,
|
| 684 |
+
"step": 5550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.66,
|
| 688 |
+
"learning_rate": 2.264449722882027e-06,
|
| 689 |
+
"loss": 0.1364,
|
| 690 |
+
"step": 5600
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.68,
|
| 694 |
+
"learning_rate": 2.1060965954077596e-06,
|
| 695 |
+
"loss": 0.1263,
|
| 696 |
+
"step": 5650
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 2.71,
|
| 700 |
+
"learning_rate": 1.947743467933492e-06,
|
| 701 |
+
"loss": 0.1267,
|
| 702 |
+
"step": 5700
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.73,
|
| 706 |
+
"learning_rate": 1.789390340459224e-06,
|
| 707 |
+
"loss": 0.1446,
|
| 708 |
+
"step": 5750
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 2.76,
|
| 712 |
+
"learning_rate": 1.6310372129849566e-06,
|
| 713 |
+
"loss": 0.141,
|
| 714 |
+
"step": 5800
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.78,
|
| 718 |
+
"learning_rate": 1.472684085510689e-06,
|
| 719 |
+
"loss": 0.1411,
|
| 720 |
+
"step": 5850
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.8,
|
| 724 |
+
"learning_rate": 1.3143309580364213e-06,
|
| 725 |
+
"loss": 0.1267,
|
| 726 |
+
"step": 5900
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.83,
|
| 730 |
+
"learning_rate": 1.1559778305621536e-06,
|
| 731 |
+
"loss": 0.1328,
|
| 732 |
+
"step": 5950
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 2.85,
|
| 736 |
+
"learning_rate": 9.976247030878861e-07,
|
| 737 |
+
"loss": 0.1163,
|
| 738 |
+
"step": 6000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.87,
|
| 742 |
+
"learning_rate": 8.392715756136184e-07,
|
| 743 |
+
"loss": 0.1369,
|
| 744 |
+
"step": 6050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.9,
|
| 748 |
+
"learning_rate": 6.809184481393508e-07,
|
| 749 |
+
"loss": 0.1245,
|
| 750 |
+
"step": 6100
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.92,
|
| 754 |
+
"learning_rate": 5.225653206650831e-07,
|
| 755 |
+
"loss": 0.1488,
|
| 756 |
+
"step": 6150
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 2.95,
|
| 760 |
+
"learning_rate": 3.642121931908155e-07,
|
| 761 |
+
"loss": 0.1204,
|
| 762 |
+
"step": 6200
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.97,
|
| 766 |
+
"learning_rate": 2.0585906571654793e-07,
|
| 767 |
+
"loss": 0.15,
|
| 768 |
+
"step": 6250
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.99,
|
| 772 |
+
"learning_rate": 4.750593824228029e-08,
|
| 773 |
+
"loss": 0.1052,
|
| 774 |
+
"step": 6300
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 3.0,
|
| 778 |
+
"step": 6315,
|
| 779 |
+
"total_flos": 1.960123773756162e+16,
|
| 780 |
+
"train_loss": 0.2133099720578107,
|
| 781 |
+
"train_runtime": 1796.8324,
|
| 782 |
+
"train_samples_per_second": 112.446,
|
| 783 |
+
"train_steps_per_second": 3.515
|
| 784 |
+
}
|
| 785 |
+
],
|
| 786 |
+
"logging_steps": 50,
|
| 787 |
+
"max_steps": 6315,
|
| 788 |
+
"num_input_tokens_seen": 0,
|
| 789 |
+
"num_train_epochs": 3,
|
| 790 |
+
"save_steps": 5000,
|
| 791 |
+
"total_flos": 1.960123773756162e+16,
|
| 792 |
+
"train_batch_size": 32,
|
| 793 |
+
"trial_name": null,
|
| 794 |
+
"trial_params": null
|
| 795 |
+
}
|