Upload trainer_state.json
Browse files- trainer_state.json +274 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,274 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.935691318327976,
|
| 5 |
+
"global_step": 5580,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.46,
|
| 12 |
+
"learning_rate": 0.0002,
|
| 13 |
+
"loss": 1.2054,
|
| 14 |
+
"step": 128
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.91,
|
| 18 |
+
"learning_rate": 0.0002,
|
| 19 |
+
"loss": 1.0285,
|
| 20 |
+
"step": 256
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.37,
|
| 24 |
+
"learning_rate": 0.0002,
|
| 25 |
+
"loss": 0.976,
|
| 26 |
+
"step": 384
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.83,
|
| 30 |
+
"learning_rate": 0.0002,
|
| 31 |
+
"loss": 0.9643,
|
| 32 |
+
"step": 512
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.29,
|
| 36 |
+
"learning_rate": 0.0002,
|
| 37 |
+
"loss": 0.9371,
|
| 38 |
+
"step": 640
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.74,
|
| 42 |
+
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 0.9214,
|
| 44 |
+
"step": 768
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.2,
|
| 48 |
+
"learning_rate": 0.0002,
|
| 49 |
+
"loss": 0.9016,
|
| 50 |
+
"step": 896
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.66,
|
| 54 |
+
"learning_rate": 0.0002,
|
| 55 |
+
"loss": 0.8862,
|
| 56 |
+
"step": 1024
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 4.12,
|
| 60 |
+
"learning_rate": 0.0002,
|
| 61 |
+
"loss": 0.8792,
|
| 62 |
+
"step": 1152
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.57,
|
| 66 |
+
"learning_rate": 0.0002,
|
| 67 |
+
"loss": 0.864,
|
| 68 |
+
"step": 1280
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 5.03,
|
| 72 |
+
"learning_rate": 0.0002,
|
| 73 |
+
"loss": 0.8588,
|
| 74 |
+
"step": 1408
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 5.49,
|
| 78 |
+
"learning_rate": 0.0002,
|
| 79 |
+
"loss": 0.8374,
|
| 80 |
+
"step": 1536
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.94,
|
| 84 |
+
"learning_rate": 0.0002,
|
| 85 |
+
"loss": 0.8414,
|
| 86 |
+
"step": 1664
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 6.4,
|
| 90 |
+
"learning_rate": 0.0002,
|
| 91 |
+
"loss": 0.8152,
|
| 92 |
+
"step": 1792
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 6.86,
|
| 96 |
+
"learning_rate": 0.0002,
|
| 97 |
+
"loss": 0.8216,
|
| 98 |
+
"step": 1920
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 7.32,
|
| 102 |
+
"learning_rate": 0.0002,
|
| 103 |
+
"loss": 0.8031,
|
| 104 |
+
"step": 2048
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 7.77,
|
| 108 |
+
"learning_rate": 0.0002,
|
| 109 |
+
"loss": 0.8031,
|
| 110 |
+
"step": 2176
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 8.23,
|
| 114 |
+
"learning_rate": 0.0002,
|
| 115 |
+
"loss": 0.7887,
|
| 116 |
+
"step": 2304
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 8.69,
|
| 120 |
+
"learning_rate": 0.0002,
|
| 121 |
+
"loss": 0.7847,
|
| 122 |
+
"step": 2432
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 9.15,
|
| 126 |
+
"learning_rate": 0.0002,
|
| 127 |
+
"loss": 0.7746,
|
| 128 |
+
"step": 2560
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 9.6,
|
| 132 |
+
"learning_rate": 0.0002,
|
| 133 |
+
"loss": 0.764,
|
| 134 |
+
"step": 2688
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 10.06,
|
| 138 |
+
"learning_rate": 0.0002,
|
| 139 |
+
"loss": 0.7652,
|
| 140 |
+
"step": 2816
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 10.52,
|
| 144 |
+
"learning_rate": 0.0002,
|
| 145 |
+
"loss": 0.7425,
|
| 146 |
+
"step": 2944
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 10.98,
|
| 150 |
+
"learning_rate": 0.0002,
|
| 151 |
+
"loss": 0.7584,
|
| 152 |
+
"step": 3072
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 11.43,
|
| 156 |
+
"learning_rate": 0.0002,
|
| 157 |
+
"loss": 0.7302,
|
| 158 |
+
"step": 3200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 11.89,
|
| 162 |
+
"learning_rate": 0.0002,
|
| 163 |
+
"loss": 0.7374,
|
| 164 |
+
"step": 3328
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 12.35,
|
| 168 |
+
"learning_rate": 0.0002,
|
| 169 |
+
"loss": 0.7187,
|
| 170 |
+
"step": 3456
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 12.8,
|
| 174 |
+
"learning_rate": 0.0002,
|
| 175 |
+
"loss": 0.7222,
|
| 176 |
+
"step": 3584
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 13.26,
|
| 180 |
+
"learning_rate": 0.0002,
|
| 181 |
+
"loss": 0.7096,
|
| 182 |
+
"step": 3712
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 13.72,
|
| 186 |
+
"learning_rate": 0.0002,
|
| 187 |
+
"loss": 0.7057,
|
| 188 |
+
"step": 3840
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 14.18,
|
| 192 |
+
"learning_rate": 0.0002,
|
| 193 |
+
"loss": 0.7003,
|
| 194 |
+
"step": 3968
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 14.63,
|
| 198 |
+
"learning_rate": 0.0002,
|
| 199 |
+
"loss": 0.6879,
|
| 200 |
+
"step": 4096
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 15.09,
|
| 204 |
+
"learning_rate": 0.0002,
|
| 205 |
+
"loss": 0.6928,
|
| 206 |
+
"step": 4224
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 15.55,
|
| 210 |
+
"learning_rate": 0.0002,
|
| 211 |
+
"loss": 0.6749,
|
| 212 |
+
"step": 4352
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 16.01,
|
| 216 |
+
"learning_rate": 0.0002,
|
| 217 |
+
"loss": 0.6857,
|
| 218 |
+
"step": 4480
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 16.46,
|
| 222 |
+
"learning_rate": 0.0002,
|
| 223 |
+
"loss": 0.6545,
|
| 224 |
+
"step": 4608
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 16.92,
|
| 228 |
+
"learning_rate": 0.0002,
|
| 229 |
+
"loss": 0.6724,
|
| 230 |
+
"step": 4736
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 17.38,
|
| 234 |
+
"learning_rate": 0.0002,
|
| 235 |
+
"loss": 0.6434,
|
| 236 |
+
"step": 4864
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 17.83,
|
| 240 |
+
"learning_rate": 0.0002,
|
| 241 |
+
"loss": 0.6581,
|
| 242 |
+
"step": 4992
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 18.29,
|
| 246 |
+
"learning_rate": 0.0002,
|
| 247 |
+
"loss": 0.644,
|
| 248 |
+
"step": 5120
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 18.75,
|
| 252 |
+
"learning_rate": 0.0002,
|
| 253 |
+
"loss": 0.6392,
|
| 254 |
+
"step": 5248
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 19.21,
|
| 258 |
+
"learning_rate": 0.0002,
|
| 259 |
+
"loss": 0.6346,
|
| 260 |
+
"step": 5376
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 19.66,
|
| 264 |
+
"learning_rate": 0.0002,
|
| 265 |
+
"loss": 0.6286,
|
| 266 |
+
"step": 5504
|
| 267 |
+
}
|
| 268 |
+
],
|
| 269 |
+
"max_steps": 5580,
|
| 270 |
+
"num_train_epochs": 20,
|
| 271 |
+
"total_flos": 3.311540504834867e+16,
|
| 272 |
+
"trial_name": null,
|
| 273 |
+
"trial_params": null
|
| 274 |
+
}
|