Upload 8 files
Browse files- optimizer.pt +2 -2
- pytorch_model.bin +1 -1
- rng_state.pth +0 -0
- scheduler.pt +0 -0
- trainer_state.json +13 -571
- training_args.bin +0 -0
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96d1ed6dbda3bbbeecb2dcd41b152da54ba605c70cfdee703aa510c1cb7626ef
|
| 3 |
+
size 686884869
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 347183353
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8681a6843eaa8b33be91a8e49f1f781a9dc65fcb60237810f5e7f1dab110e3f1
|
| 3 |
size 347183353
|
rng_state.pth
CHANGED
|
Binary files a/rng_state.pth and b/rng_state.pth differ
|
|
|
scheduler.pt
CHANGED
|
Binary files a/scheduler.pt and b/scheduler.pt differ
|
|
|
trainer_state.json
CHANGED
|
@@ -1,598 +1,40 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
-
{
|
| 11 |
-
"epoch": 0.23,
|
| 12 |
-
"learning_rate": 4.92375724306191e-06,
|
| 13 |
-
"loss": 4.617,
|
| 14 |
-
"step": 1000
|
| 15 |
-
},
|
| 16 |
-
{
|
| 17 |
-
"epoch": 0.46,
|
| 18 |
-
"learning_rate": 4.8475144861238184e-06,
|
| 19 |
-
"loss": 4.3463,
|
| 20 |
-
"step": 2000
|
| 21 |
-
},
|
| 22 |
-
{
|
| 23 |
-
"epoch": 0.69,
|
| 24 |
-
"learning_rate": 4.771271729185728e-06,
|
| 25 |
-
"loss": 4.0219,
|
| 26 |
-
"step": 3000
|
| 27 |
-
},
|
| 28 |
-
{
|
| 29 |
-
"epoch": 0.91,
|
| 30 |
-
"learning_rate": 4.6950289722476364e-06,
|
| 31 |
-
"loss": 3.7277,
|
| 32 |
-
"step": 4000
|
| 33 |
-
},
|
| 34 |
-
{
|
| 35 |
-
"epoch": 1.14,
|
| 36 |
-
"learning_rate": 4.618786215309546e-06,
|
| 37 |
-
"loss": 3.4665,
|
| 38 |
-
"step": 5000
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 1.37,
|
| 42 |
-
"learning_rate": 4.542543458371455e-06,
|
| 43 |
-
"loss": 3.2523,
|
| 44 |
-
"step": 6000
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 1.6,
|
| 48 |
-
"learning_rate": 4.466300701433364e-06,
|
| 49 |
-
"loss": 3.1099,
|
| 50 |
-
"step": 7000
|
| 51 |
-
},
|
| 52 |
-
{
|
| 53 |
-
"epoch": 1.83,
|
| 54 |
-
"learning_rate": 4.390057944495273e-06,
|
| 55 |
-
"loss": 2.9365,
|
| 56 |
-
"step": 8000
|
| 57 |
-
},
|
| 58 |
-
{
|
| 59 |
-
"epoch": 2.06,
|
| 60 |
-
"learning_rate": 4.313815187557183e-06,
|
| 61 |
-
"loss": 2.7993,
|
| 62 |
-
"step": 9000
|
| 63 |
-
},
|
| 64 |
{
|
| 65 |
"epoch": 2.29,
|
| 66 |
-
"learning_rate":
|
| 67 |
-
"loss": 2.
|
| 68 |
"step": 10000
|
| 69 |
},
|
| 70 |
-
{
|
| 71 |
-
"epoch": 2.52,
|
| 72 |
-
"learning_rate": 4.161329673681001e-06,
|
| 73 |
-
"loss": 2.5923,
|
| 74 |
-
"step": 11000
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"epoch": 2.74,
|
| 78 |
-
"learning_rate": 4.085086916742909e-06,
|
| 79 |
-
"loss": 2.5037,
|
| 80 |
-
"step": 12000
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 2.97,
|
| 84 |
-
"learning_rate": 4.008844159804819e-06,
|
| 85 |
-
"loss": 2.4135,
|
| 86 |
-
"step": 13000
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 3.2,
|
| 90 |
-
"learning_rate": 3.932601402866728e-06,
|
| 91 |
-
"loss": 2.3354,
|
| 92 |
-
"step": 14000
|
| 93 |
-
},
|
| 94 |
-
{
|
| 95 |
-
"epoch": 3.43,
|
| 96 |
-
"learning_rate": 3.856358645928638e-06,
|
| 97 |
-
"loss": 2.2609,
|
| 98 |
-
"step": 15000
|
| 99 |
-
},
|
| 100 |
-
{
|
| 101 |
-
"epoch": 3.66,
|
| 102 |
-
"learning_rate": 3.780115888990546e-06,
|
| 103 |
-
"loss": 2.2187,
|
| 104 |
-
"step": 16000
|
| 105 |
-
},
|
| 106 |
-
{
|
| 107 |
-
"epoch": 3.89,
|
| 108 |
-
"learning_rate": 3.7038731320524552e-06,
|
| 109 |
-
"loss": 2.1399,
|
| 110 |
-
"step": 17000
|
| 111 |
-
},
|
| 112 |
-
{
|
| 113 |
-
"epoch": 4.12,
|
| 114 |
-
"learning_rate": 3.6276303751143642e-06,
|
| 115 |
-
"loss": 2.0927,
|
| 116 |
-
"step": 18000
|
| 117 |
-
},
|
| 118 |
-
{
|
| 119 |
-
"epoch": 4.35,
|
| 120 |
-
"learning_rate": 3.5513876181762737e-06,
|
| 121 |
-
"loss": 2.0264,
|
| 122 |
-
"step": 19000
|
| 123 |
-
},
|
| 124 |
{
|
| 125 |
"epoch": 4.57,
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 20000
|
| 129 |
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 4.8,
|
| 132 |
-
"learning_rate": 3.398902104300092e-06,
|
| 133 |
-
"loss": 1.9323,
|
| 134 |
-
"step": 21000
|
| 135 |
-
},
|
| 136 |
-
{
|
| 137 |
-
"epoch": 5.03,
|
| 138 |
-
"learning_rate": 3.322659347362001e-06,
|
| 139 |
-
"loss": 1.9327,
|
| 140 |
-
"step": 22000
|
| 141 |
-
},
|
| 142 |
-
{
|
| 143 |
-
"epoch": 5.26,
|
| 144 |
-
"learning_rate": 3.2464165904239097e-06,
|
| 145 |
-
"loss": 1.8627,
|
| 146 |
-
"step": 23000
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 5.49,
|
| 150 |
-
"learning_rate": 3.170173833485819e-06,
|
| 151 |
-
"loss": 1.8122,
|
| 152 |
-
"step": 24000
|
| 153 |
-
},
|
| 154 |
-
{
|
| 155 |
-
"epoch": 5.72,
|
| 156 |
-
"learning_rate": 3.093931076547728e-06,
|
| 157 |
-
"loss": 1.8073,
|
| 158 |
-
"step": 25000
|
| 159 |
-
},
|
| 160 |
-
{
|
| 161 |
-
"epoch": 5.95,
|
| 162 |
-
"learning_rate": 3.0176883196096376e-06,
|
| 163 |
-
"loss": 1.7809,
|
| 164 |
-
"step": 26000
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 6.18,
|
| 168 |
-
"learning_rate": 2.9414455626715466e-06,
|
| 169 |
-
"loss": 1.7327,
|
| 170 |
-
"step": 27000
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 6.4,
|
| 174 |
-
"learning_rate": 2.8652028057334556e-06,
|
| 175 |
-
"loss": 1.6954,
|
| 176 |
-
"step": 28000
|
| 177 |
-
},
|
| 178 |
-
{
|
| 179 |
-
"epoch": 6.63,
|
| 180 |
-
"learning_rate": 2.7889600487953646e-06,
|
| 181 |
-
"loss": 1.6836,
|
| 182 |
-
"step": 29000
|
| 183 |
-
},
|
| 184 |
{
|
| 185 |
"epoch": 6.86,
|
| 186 |
-
"learning_rate":
|
| 187 |
-
"loss":
|
| 188 |
"step": 30000
|
| 189 |
},
|
| 190 |
-
{
|
| 191 |
-
"epoch": 7.09,
|
| 192 |
-
"learning_rate": 2.6364745349191826e-06,
|
| 193 |
-
"loss": 1.6564,
|
| 194 |
-
"step": 31000
|
| 195 |
-
},
|
| 196 |
-
{
|
| 197 |
-
"epoch": 7.32,
|
| 198 |
-
"learning_rate": 2.560231777981092e-06,
|
| 199 |
-
"loss": 1.6141,
|
| 200 |
-
"step": 32000
|
| 201 |
-
},
|
| 202 |
-
{
|
| 203 |
-
"epoch": 7.55,
|
| 204 |
-
"learning_rate": 2.483989021043001e-06,
|
| 205 |
-
"loss": 1.5769,
|
| 206 |
-
"step": 33000
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 7.78,
|
| 210 |
-
"learning_rate": 2.40774626410491e-06,
|
| 211 |
-
"loss": 1.6009,
|
| 212 |
-
"step": 34000
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 8.01,
|
| 216 |
-
"learning_rate": 2.3315035071668195e-06,
|
| 217 |
-
"loss": 1.581,
|
| 218 |
-
"step": 35000
|
| 219 |
-
},
|
| 220 |
-
{
|
| 221 |
-
"epoch": 8.23,
|
| 222 |
-
"learning_rate": 2.2552607502287285e-06,
|
| 223 |
-
"loss": 1.5363,
|
| 224 |
-
"step": 36000
|
| 225 |
-
},
|
| 226 |
-
{
|
| 227 |
-
"epoch": 8.46,
|
| 228 |
-
"learning_rate": 2.1790179932906375e-06,
|
| 229 |
-
"loss": 1.5425,
|
| 230 |
-
"step": 37000
|
| 231 |
-
},
|
| 232 |
-
{
|
| 233 |
-
"epoch": 8.69,
|
| 234 |
-
"learning_rate": 2.1027752363525465e-06,
|
| 235 |
-
"loss": 1.5053,
|
| 236 |
-
"step": 38000
|
| 237 |
-
},
|
| 238 |
-
{
|
| 239 |
-
"epoch": 8.92,
|
| 240 |
-
"learning_rate": 2.026532479414456e-06,
|
| 241 |
-
"loss": 1.4986,
|
| 242 |
-
"step": 39000
|
| 243 |
-
},
|
| 244 |
{
|
| 245 |
"epoch": 9.15,
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss":
|
| 248 |
"step": 40000
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 9.38,
|
| 252 |
-
"learning_rate": 1.874046965538274e-06,
|
| 253 |
-
"loss": 1.4829,
|
| 254 |
-
"step": 41000
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 9.61,
|
| 258 |
-
"learning_rate": 1.7978042086001832e-06,
|
| 259 |
-
"loss": 1.4634,
|
| 260 |
-
"step": 42000
|
| 261 |
-
},
|
| 262 |
-
{
|
| 263 |
-
"epoch": 9.84,
|
| 264 |
-
"learning_rate": 1.7215614516620924e-06,
|
| 265 |
-
"loss": 1.4559,
|
| 266 |
-
"step": 43000
|
| 267 |
-
},
|
| 268 |
-
{
|
| 269 |
-
"epoch": 10.06,
|
| 270 |
-
"learning_rate": 1.6453186947240012e-06,
|
| 271 |
-
"loss": 1.4388,
|
| 272 |
-
"step": 44000
|
| 273 |
-
},
|
| 274 |
-
{
|
| 275 |
-
"epoch": 10.29,
|
| 276 |
-
"learning_rate": 1.5690759377859104e-06,
|
| 277 |
-
"loss": 1.4299,
|
| 278 |
-
"step": 45000
|
| 279 |
-
},
|
| 280 |
-
{
|
| 281 |
-
"epoch": 10.52,
|
| 282 |
-
"learning_rate": 1.4928331808478196e-06,
|
| 283 |
-
"loss": 1.4328,
|
| 284 |
-
"step": 46000
|
| 285 |
-
},
|
| 286 |
-
{
|
| 287 |
-
"epoch": 10.75,
|
| 288 |
-
"learning_rate": 1.4165904239097286e-06,
|
| 289 |
-
"loss": 1.423,
|
| 290 |
-
"step": 47000
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 10.98,
|
| 294 |
-
"learning_rate": 1.3403476669716379e-06,
|
| 295 |
-
"loss": 1.3973,
|
| 296 |
-
"step": 48000
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 11.21,
|
| 300 |
-
"learning_rate": 1.264104910033547e-06,
|
| 301 |
-
"loss": 1.4033,
|
| 302 |
-
"step": 49000
|
| 303 |
-
},
|
| 304 |
-
{
|
| 305 |
-
"epoch": 11.44,
|
| 306 |
-
"learning_rate": 1.187862153095456e-06,
|
| 307 |
-
"loss": 1.3921,
|
| 308 |
-
"step": 50000
|
| 309 |
-
},
|
| 310 |
-
{
|
| 311 |
-
"epoch": 11.67,
|
| 312 |
-
"learning_rate": 1.111619396157365e-06,
|
| 313 |
-
"loss": 1.385,
|
| 314 |
-
"step": 51000
|
| 315 |
-
},
|
| 316 |
-
{
|
| 317 |
-
"epoch": 11.89,
|
| 318 |
-
"learning_rate": 1.035376639219274e-06,
|
| 319 |
-
"loss": 1.3807,
|
| 320 |
-
"step": 52000
|
| 321 |
-
},
|
| 322 |
-
{
|
| 323 |
-
"epoch": 12.12,
|
| 324 |
-
"learning_rate": 9.591338822811833e-07,
|
| 325 |
-
"loss": 1.3734,
|
| 326 |
-
"step": 53000
|
| 327 |
-
},
|
| 328 |
-
{
|
| 329 |
-
"epoch": 12.35,
|
| 330 |
-
"learning_rate": 8.828911253430924e-07,
|
| 331 |
-
"loss": 1.3459,
|
| 332 |
-
"step": 54000
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 12.58,
|
| 336 |
-
"learning_rate": 8.066483684050017e-07,
|
| 337 |
-
"loss": 1.3675,
|
| 338 |
-
"step": 55000
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 12.81,
|
| 342 |
-
"learning_rate": 7.304056114669108e-07,
|
| 343 |
-
"loss": 1.3436,
|
| 344 |
-
"step": 56000
|
| 345 |
-
},
|
| 346 |
-
{
|
| 347 |
-
"epoch": 13.04,
|
| 348 |
-
"learning_rate": 6.541628545288198e-07,
|
| 349 |
-
"loss": 1.3673,
|
| 350 |
-
"step": 57000
|
| 351 |
-
},
|
| 352 |
-
{
|
| 353 |
-
"epoch": 13.27,
|
| 354 |
-
"learning_rate": 5.779200975907289e-07,
|
| 355 |
-
"loss": 1.3417,
|
| 356 |
-
"step": 58000
|
| 357 |
-
},
|
| 358 |
-
{
|
| 359 |
-
"epoch": 13.49,
|
| 360 |
-
"learning_rate": 2.750838670326319e-06,
|
| 361 |
-
"loss": 1.3706,
|
| 362 |
-
"step": 59000
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 13.72,
|
| 366 |
-
"learning_rate": 2.7127172918572736e-06,
|
| 367 |
-
"loss": 1.3743,
|
| 368 |
-
"step": 60000
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 13.95,
|
| 372 |
-
"learning_rate": 2.6745959133882283e-06,
|
| 373 |
-
"loss": 1.3504,
|
| 374 |
-
"step": 61000
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 14.18,
|
| 378 |
-
"learning_rate": 2.6364745349191826e-06,
|
| 379 |
-
"loss": 1.3362,
|
| 380 |
-
"step": 62000
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 14.41,
|
| 384 |
-
"learning_rate": 2.5983531564501373e-06,
|
| 385 |
-
"loss": 1.3308,
|
| 386 |
-
"step": 63000
|
| 387 |
-
},
|
| 388 |
-
{
|
| 389 |
-
"epoch": 14.64,
|
| 390 |
-
"learning_rate": 2.560231777981092e-06,
|
| 391 |
-
"loss": 1.3162,
|
| 392 |
-
"step": 64000
|
| 393 |
-
},
|
| 394 |
-
{
|
| 395 |
-
"epoch": 14.87,
|
| 396 |
-
"learning_rate": 2.5221103995120468e-06,
|
| 397 |
-
"loss": 1.2831,
|
| 398 |
-
"step": 65000
|
| 399 |
-
},
|
| 400 |
-
{
|
| 401 |
-
"epoch": 15.1,
|
| 402 |
-
"learning_rate": 2.483989021043001e-06,
|
| 403 |
-
"loss": 1.2969,
|
| 404 |
-
"step": 66000
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 15.32,
|
| 408 |
-
"learning_rate": 2.4458676425739553e-06,
|
| 409 |
-
"loss": 1.2785,
|
| 410 |
-
"step": 67000
|
| 411 |
-
},
|
| 412 |
-
{
|
| 413 |
-
"epoch": 15.55,
|
| 414 |
-
"learning_rate": 2.40774626410491e-06,
|
| 415 |
-
"loss": 1.2769,
|
| 416 |
-
"step": 68000
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 15.78,
|
| 420 |
-
"learning_rate": 2.3696248856358648e-06,
|
| 421 |
-
"loss": 1.2799,
|
| 422 |
-
"step": 69000
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 16.01,
|
| 426 |
-
"learning_rate": 2.3315035071668195e-06,
|
| 427 |
-
"loss": 1.2572,
|
| 428 |
-
"step": 70000
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 16.24,
|
| 432 |
-
"learning_rate": 2.2933821286977738e-06,
|
| 433 |
-
"loss": 1.2495,
|
| 434 |
-
"step": 71000
|
| 435 |
-
},
|
| 436 |
-
{
|
| 437 |
-
"epoch": 16.47,
|
| 438 |
-
"learning_rate": 2.2552607502287285e-06,
|
| 439 |
-
"loss": 1.2338,
|
| 440 |
-
"step": 72000
|
| 441 |
-
},
|
| 442 |
-
{
|
| 443 |
-
"epoch": 16.7,
|
| 444 |
-
"learning_rate": 2.217139371759683e-06,
|
| 445 |
-
"loss": 1.2572,
|
| 446 |
-
"step": 73000
|
| 447 |
-
},
|
| 448 |
-
{
|
| 449 |
-
"epoch": 16.93,
|
| 450 |
-
"learning_rate": 2.1790179932906375e-06,
|
| 451 |
-
"loss": 1.2191,
|
| 452 |
-
"step": 74000
|
| 453 |
-
},
|
| 454 |
-
{
|
| 455 |
-
"epoch": 17.15,
|
| 456 |
-
"learning_rate": 2.1408966148215922e-06,
|
| 457 |
-
"loss": 1.2256,
|
| 458 |
-
"step": 75000
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 17.38,
|
| 462 |
-
"learning_rate": 2.1027752363525465e-06,
|
| 463 |
-
"loss": 1.2336,
|
| 464 |
-
"step": 76000
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 17.61,
|
| 468 |
-
"learning_rate": 2.0646538578835012e-06,
|
| 469 |
-
"loss": 1.2135,
|
| 470 |
-
"step": 77000
|
| 471 |
-
},
|
| 472 |
-
{
|
| 473 |
-
"epoch": 17.84,
|
| 474 |
-
"learning_rate": 2.026532479414456e-06,
|
| 475 |
-
"loss": 1.1806,
|
| 476 |
-
"step": 78000
|
| 477 |
-
},
|
| 478 |
-
{
|
| 479 |
-
"epoch": 18.07,
|
| 480 |
-
"learning_rate": 1.9884111009454107e-06,
|
| 481 |
-
"loss": 1.1971,
|
| 482 |
-
"step": 79000
|
| 483 |
-
},
|
| 484 |
-
{
|
| 485 |
-
"epoch": 18.3,
|
| 486 |
-
"learning_rate": 1.950289722476365e-06,
|
| 487 |
-
"loss": 1.1823,
|
| 488 |
-
"step": 80000
|
| 489 |
-
},
|
| 490 |
-
{
|
| 491 |
-
"epoch": 18.53,
|
| 492 |
-
"learning_rate": 1.9121683440073192e-06,
|
| 493 |
-
"loss": 1.1678,
|
| 494 |
-
"step": 81000
|
| 495 |
-
},
|
| 496 |
-
{
|
| 497 |
-
"epoch": 18.76,
|
| 498 |
-
"learning_rate": 1.874046965538274e-06,
|
| 499 |
-
"loss": 1.1587,
|
| 500 |
-
"step": 82000
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 18.98,
|
| 504 |
-
"learning_rate": 1.8359255870692287e-06,
|
| 505 |
-
"loss": 1.179,
|
| 506 |
-
"step": 83000
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 19.21,
|
| 510 |
-
"learning_rate": 1.7978042086001832e-06,
|
| 511 |
-
"loss": 1.1547,
|
| 512 |
-
"step": 84000
|
| 513 |
-
},
|
| 514 |
-
{
|
| 515 |
-
"epoch": 19.44,
|
| 516 |
-
"learning_rate": 1.7596828301311377e-06,
|
| 517 |
-
"loss": 1.1558,
|
| 518 |
-
"step": 85000
|
| 519 |
-
},
|
| 520 |
-
{
|
| 521 |
-
"epoch": 19.67,
|
| 522 |
-
"learning_rate": 1.7215614516620924e-06,
|
| 523 |
-
"loss": 1.1634,
|
| 524 |
-
"step": 86000
|
| 525 |
-
},
|
| 526 |
-
{
|
| 527 |
-
"epoch": 19.9,
|
| 528 |
-
"learning_rate": 1.6834400731930467e-06,
|
| 529 |
-
"loss": 1.1421,
|
| 530 |
-
"step": 87000
|
| 531 |
-
},
|
| 532 |
-
{
|
| 533 |
-
"epoch": 20.13,
|
| 534 |
-
"learning_rate": 1.6453186947240012e-06,
|
| 535 |
-
"loss": 1.1497,
|
| 536 |
-
"step": 88000
|
| 537 |
-
},
|
| 538 |
-
{
|
| 539 |
-
"epoch": 20.36,
|
| 540 |
-
"learning_rate": 1.607197316254956e-06,
|
| 541 |
-
"loss": 1.1239,
|
| 542 |
-
"step": 89000
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 20.59,
|
| 546 |
-
"learning_rate": 1.5690759377859104e-06,
|
| 547 |
-
"loss": 1.1538,
|
| 548 |
-
"step": 90000
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 20.81,
|
| 552 |
-
"learning_rate": 1.5309545593168651e-06,
|
| 553 |
-
"loss": 1.1357,
|
| 554 |
-
"step": 91000
|
| 555 |
-
},
|
| 556 |
-
{
|
| 557 |
-
"epoch": 21.04,
|
| 558 |
-
"learning_rate": 1.4928331808478196e-06,
|
| 559 |
-
"loss": 1.1193,
|
| 560 |
-
"step": 92000
|
| 561 |
-
},
|
| 562 |
-
{
|
| 563 |
-
"epoch": 21.27,
|
| 564 |
-
"learning_rate": 1.454711802378774e-06,
|
| 565 |
-
"loss": 1.1302,
|
| 566 |
-
"step": 93000
|
| 567 |
-
},
|
| 568 |
-
{
|
| 569 |
-
"epoch": 21.5,
|
| 570 |
-
"learning_rate": 1.4165904239097286e-06,
|
| 571 |
-
"loss": 1.0902,
|
| 572 |
-
"step": 94000
|
| 573 |
-
},
|
| 574 |
-
{
|
| 575 |
-
"epoch": 21.73,
|
| 576 |
-
"learning_rate": 1.3784690454406831e-06,
|
| 577 |
-
"loss": 1.1061,
|
| 578 |
-
"step": 95000
|
| 579 |
-
},
|
| 580 |
-
{
|
| 581 |
-
"epoch": 21.96,
|
| 582 |
-
"learning_rate": 1.3403476669716379e-06,
|
| 583 |
-
"loss": 1.1025,
|
| 584 |
-
"step": 96000
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 22.19,
|
| 588 |
-
"learning_rate": 1.3022262885025924e-06,
|
| 589 |
-
"loss": 1.1007,
|
| 590 |
-
"step": 97000
|
| 591 |
}
|
| 592 |
],
|
| 593 |
-
"max_steps":
|
| 594 |
-
"num_train_epochs":
|
| 595 |
-
"total_flos":
|
| 596 |
"trial_name": null,
|
| 597 |
"trial_params": null
|
| 598 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.149130832570906,
|
| 5 |
+
"global_step": 40000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
{
|
| 11 |
"epoch": 2.29,
|
| 12 |
+
"learning_rate": 9.42817932296432e-06,
|
| 13 |
+
"loss": 2.4808,
|
| 14 |
"step": 10000
|
| 15 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
{
|
| 17 |
"epoch": 4.57,
|
| 18 |
+
"learning_rate": 8.856358645928637e-06,
|
| 19 |
+
"loss": 1.2166,
|
| 20 |
"step": 20000
|
| 21 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
{
|
| 23 |
"epoch": 6.86,
|
| 24 |
+
"learning_rate": 8.284537968892956e-06,
|
| 25 |
+
"loss": 0.9323,
|
| 26 |
"step": 30000
|
| 27 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
{
|
| 29 |
"epoch": 9.15,
|
| 30 |
+
"learning_rate": 7.712717291857275e-06,
|
| 31 |
+
"loss": 0.7901,
|
| 32 |
"step": 40000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
}
|
| 34 |
],
|
| 35 |
+
"max_steps": 174880,
|
| 36 |
+
"num_train_epochs": 40,
|
| 37 |
+
"total_flos": 9.92204617219769e+19,
|
| 38 |
"trial_name": null,
|
| 39 |
"trial_params": null
|
| 40 |
}
|
training_args.bin
CHANGED
|
Binary files a/training_args.bin and b/training_args.bin differ
|
|
|