--per_device_train_batch_size 4 --learning_rate 3e-5 --num_train_epochs 2 --max_seq_length 512 --doc_stride 250 --max_answer_length 200
Browse files- trainer_state.json +330 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,330 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 25070,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04,
|
| 13 |
+
"learning_rate": 2.9401675309134427e-05,
|
| 14 |
+
"loss": 0.7213,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.08,
|
| 19 |
+
"learning_rate": 2.8803350618268846e-05,
|
| 20 |
+
"loss": 0.463,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.12,
|
| 25 |
+
"learning_rate": 2.8205025927403272e-05,
|
| 26 |
+
"loss": 0.4433,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.16,
|
| 31 |
+
"learning_rate": 2.7606701236537695e-05,
|
| 32 |
+
"loss": 0.3953,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.2,
|
| 37 |
+
"learning_rate": 2.700837654567212e-05,
|
| 38 |
+
"loss": 0.2915,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.24,
|
| 43 |
+
"learning_rate": 2.641005185480654e-05,
|
| 44 |
+
"loss": 0.3485,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.28,
|
| 49 |
+
"learning_rate": 2.5811727163940966e-05,
|
| 50 |
+
"loss": 0.3159,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.32,
|
| 55 |
+
"learning_rate": 2.521340247307539e-05,
|
| 56 |
+
"loss": 0.2976,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.36,
|
| 61 |
+
"learning_rate": 2.4615077782209815e-05,
|
| 62 |
+
"loss": 0.3095,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.4,
|
| 67 |
+
"learning_rate": 2.4016753091344235e-05,
|
| 68 |
+
"loss": 0.2794,
|
| 69 |
+
"step": 5000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.44,
|
| 73 |
+
"learning_rate": 2.341842840047866e-05,
|
| 74 |
+
"loss": 0.2936,
|
| 75 |
+
"step": 5500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.48,
|
| 79 |
+
"learning_rate": 2.2820103709613083e-05,
|
| 80 |
+
"loss": 0.2691,
|
| 81 |
+
"step": 6000
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.52,
|
| 85 |
+
"learning_rate": 2.222177901874751e-05,
|
| 86 |
+
"loss": 0.2911,
|
| 87 |
+
"step": 6500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.56,
|
| 91 |
+
"learning_rate": 2.162345432788193e-05,
|
| 92 |
+
"loss": 0.3109,
|
| 93 |
+
"step": 7000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.6,
|
| 97 |
+
"learning_rate": 2.1025129637016355e-05,
|
| 98 |
+
"loss": 0.3147,
|
| 99 |
+
"step": 7500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.64,
|
| 103 |
+
"learning_rate": 2.042680494615078e-05,
|
| 104 |
+
"loss": 0.2418,
|
| 105 |
+
"step": 8000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.68,
|
| 109 |
+
"learning_rate": 1.98284802552852e-05,
|
| 110 |
+
"loss": 0.2512,
|
| 111 |
+
"step": 8500
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.72,
|
| 115 |
+
"learning_rate": 1.9230155564419626e-05,
|
| 116 |
+
"loss": 0.2615,
|
| 117 |
+
"step": 9000
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.76,
|
| 121 |
+
"learning_rate": 1.863183087355405e-05,
|
| 122 |
+
"loss": 0.2867,
|
| 123 |
+
"step": 9500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.8,
|
| 127 |
+
"learning_rate": 1.8033506182688475e-05,
|
| 128 |
+
"loss": 0.2778,
|
| 129 |
+
"step": 10000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.84,
|
| 133 |
+
"learning_rate": 1.7435181491822894e-05,
|
| 134 |
+
"loss": 0.2556,
|
| 135 |
+
"step": 10500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.88,
|
| 139 |
+
"learning_rate": 1.683685680095732e-05,
|
| 140 |
+
"loss": 0.2778,
|
| 141 |
+
"step": 11000
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.92,
|
| 145 |
+
"learning_rate": 1.6238532110091743e-05,
|
| 146 |
+
"loss": 0.2175,
|
| 147 |
+
"step": 11500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.96,
|
| 151 |
+
"learning_rate": 1.564020741922617e-05,
|
| 152 |
+
"loss": 0.2036,
|
| 153 |
+
"step": 12000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.0,
|
| 157 |
+
"learning_rate": 1.504188272836059e-05,
|
| 158 |
+
"loss": 0.2023,
|
| 159 |
+
"step": 12500
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.04,
|
| 163 |
+
"learning_rate": 1.4443558037495015e-05,
|
| 164 |
+
"loss": 0.1519,
|
| 165 |
+
"step": 13000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.08,
|
| 169 |
+
"learning_rate": 1.3845233346629439e-05,
|
| 170 |
+
"loss": 0.1213,
|
| 171 |
+
"step": 13500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.12,
|
| 175 |
+
"learning_rate": 1.3246908655763862e-05,
|
| 176 |
+
"loss": 0.1656,
|
| 177 |
+
"step": 14000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.16,
|
| 181 |
+
"learning_rate": 1.2648583964898285e-05,
|
| 182 |
+
"loss": 0.1239,
|
| 183 |
+
"step": 14500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 1.2,
|
| 187 |
+
"learning_rate": 1.2050259274032709e-05,
|
| 188 |
+
"loss": 0.1242,
|
| 189 |
+
"step": 15000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.24,
|
| 193 |
+
"learning_rate": 1.1451934583167132e-05,
|
| 194 |
+
"loss": 0.0917,
|
| 195 |
+
"step": 15500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 1.28,
|
| 199 |
+
"learning_rate": 1.0853609892301556e-05,
|
| 200 |
+
"loss": 0.0922,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.32,
|
| 205 |
+
"learning_rate": 1.0255285201435979e-05,
|
| 206 |
+
"loss": 0.1306,
|
| 207 |
+
"step": 16500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.36,
|
| 211 |
+
"learning_rate": 9.656960510570403e-06,
|
| 212 |
+
"loss": 0.0843,
|
| 213 |
+
"step": 17000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.4,
|
| 217 |
+
"learning_rate": 9.058635819704826e-06,
|
| 218 |
+
"loss": 0.1032,
|
| 219 |
+
"step": 17500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.44,
|
| 223 |
+
"learning_rate": 8.46031112883925e-06,
|
| 224 |
+
"loss": 0.1202,
|
| 225 |
+
"step": 18000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.48,
|
| 229 |
+
"learning_rate": 7.861986437973673e-06,
|
| 230 |
+
"loss": 0.1296,
|
| 231 |
+
"step": 18500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.52,
|
| 235 |
+
"learning_rate": 7.263661747108097e-06,
|
| 236 |
+
"loss": 0.1354,
|
| 237 |
+
"step": 19000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 1.56,
|
| 241 |
+
"learning_rate": 6.665337056242521e-06,
|
| 242 |
+
"loss": 0.1039,
|
| 243 |
+
"step": 19500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.6,
|
| 247 |
+
"learning_rate": 6.0670123653769444e-06,
|
| 248 |
+
"loss": 0.0892,
|
| 249 |
+
"step": 20000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.64,
|
| 253 |
+
"learning_rate": 5.468687674511368e-06,
|
| 254 |
+
"loss": 0.096,
|
| 255 |
+
"step": 20500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.68,
|
| 259 |
+
"learning_rate": 4.8703629836457915e-06,
|
| 260 |
+
"loss": 0.1045,
|
| 261 |
+
"step": 21000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.72,
|
| 265 |
+
"learning_rate": 4.272038292780215e-06,
|
| 266 |
+
"loss": 0.0965,
|
| 267 |
+
"step": 21500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.76,
|
| 271 |
+
"learning_rate": 3.673713601914639e-06,
|
| 272 |
+
"loss": 0.0805,
|
| 273 |
+
"step": 22000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.79,
|
| 277 |
+
"learning_rate": 3.0753889110490626e-06,
|
| 278 |
+
"loss": 0.1027,
|
| 279 |
+
"step": 22500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.83,
|
| 283 |
+
"learning_rate": 2.4770642201834866e-06,
|
| 284 |
+
"loss": 0.1287,
|
| 285 |
+
"step": 23000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.87,
|
| 289 |
+
"learning_rate": 1.87873952931791e-06,
|
| 290 |
+
"loss": 0.0844,
|
| 291 |
+
"step": 23500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.91,
|
| 295 |
+
"learning_rate": 1.2804148384523335e-06,
|
| 296 |
+
"loss": 0.0861,
|
| 297 |
+
"step": 24000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.95,
|
| 301 |
+
"learning_rate": 6.820901475867571e-07,
|
| 302 |
+
"loss": 0.1154,
|
| 303 |
+
"step": 24500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.99,
|
| 307 |
+
"learning_rate": 8.37654567211807e-08,
|
| 308 |
+
"loss": 0.0838,
|
| 309 |
+
"step": 25000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.0,
|
| 313 |
+
"step": 25070,
|
| 314 |
+
"total_flos": 9.312702234262733e+16,
|
| 315 |
+
"train_loss": 0.2107836102126935,
|
| 316 |
+
"train_runtime": 30875.5596,
|
| 317 |
+
"train_samples_per_second": 3.248,
|
| 318 |
+
"train_steps_per_second": 0.812
|
| 319 |
+
}
|
| 320 |
+
],
|
| 321 |
+
"logging_steps": 500,
|
| 322 |
+
"max_steps": 25070,
|
| 323 |
+
"num_input_tokens_seen": 0,
|
| 324 |
+
"num_train_epochs": 2,
|
| 325 |
+
"save_steps": 500,
|
| 326 |
+
"total_flos": 9.312702234262733e+16,
|
| 327 |
+
"train_batch_size": 4,
|
| 328 |
+
"trial_name": null,
|
| 329 |
+
"trial_params": null
|
| 330 |
+
}
|