Training in progress, epoch 1, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1324830880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40b4a1e4ec540bc656d73ff8878ee7be835cf08cb3173699ec17509946916f58
|
| 3 |
size 1324830880
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2649896094
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8ad656b279f1fb52687fc536d24a1765d0c84180daa22965672e107fd6870e2
|
| 3 |
size 2649896094
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4148cdc128d12b940ce70f5dee553010f8414faa77e9b07518cb0e0d0dbbdc85
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:032817f2c288f82ea29ad5b7447cbdb475d0523f1714c8c4435e81aa12e3c82b
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1226 +1,327 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./opt_trained/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.023391812865497075,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 9.
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.04678362573099415,
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 9.
|
| 22 |
-
"loss": 3.
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.07017543859649122,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate": 9.
|
| 29 |
-
"loss":
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.0935672514619883,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 9.
|
| 36 |
-
"loss": 2.
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.11695906432748537,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 9.
|
| 43 |
-
"loss": 2.
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.14035087719298245,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 9.
|
| 50 |
-
"loss": 2.
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.16374269005847952,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 9.
|
| 57 |
-
"loss": 2.
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.1871345029239766,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate": 9.
|
| 64 |
-
"loss":
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.21052631578947367,
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate": 9.
|
| 71 |
-
"loss":
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.23391812865497075,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate": 9.
|
| 78 |
-
"loss": 2.
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.2573099415204678,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate": 9.
|
| 85 |
-
"loss": 2.
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.2807017543859649,
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate": 9.
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.30409356725146197,
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate": 9.
|
| 99 |
-
"loss": 2.
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.32748538011695905,
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate": 9.
|
| 106 |
-
"loss": 2.
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.3508771929824561,
|
| 111 |
-
"grad_norm": 5.
|
| 112 |
-
"learning_rate": 9.
|
| 113 |
-
"loss": 2.
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.3742690058479532,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate": 9.
|
| 120 |
-
"loss": 2.
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.39766081871345027,
|
| 125 |
-
"grad_norm":
|
| 126 |
-
"learning_rate": 9.
|
| 127 |
-
"loss": 2.
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.42105263157894735,
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 2.
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.4444444444444444,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss": 2.
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.4678362573099415,
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 2.
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.49122807017543857,
|
| 153 |
-
"grad_norm": 3.
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss": 2.
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.5146198830409356,
|
| 160 |
-
"grad_norm": 2.
|
| 161 |
-
"learning_rate":
|
| 162 |
-
"loss": 2.
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.5380116959064327,
|
| 167 |
-
"grad_norm": 2.
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 2.
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.5614035087719298,
|
| 174 |
-
"grad_norm": 3.
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 2.
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.5847953216374269,
|
| 181 |
-
"grad_norm":
|
| 182 |
-
"learning_rate":
|
| 183 |
-
"loss": 2.
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.6081871345029239,
|
| 188 |
-
"grad_norm": 3.
|
| 189 |
-
"learning_rate":
|
| 190 |
-
"loss": 2.
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.631578947368421,
|
| 195 |
-
"grad_norm":
|
| 196 |
-
"learning_rate":
|
| 197 |
-
"loss": 2.
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.6549707602339181,
|
| 202 |
-
"grad_norm":
|
| 203 |
-
"learning_rate":
|
| 204 |
-
"loss": 2.
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.6783625730994152,
|
| 209 |
-
"grad_norm":
|
| 210 |
-
"learning_rate":
|
| 211 |
-
"loss": 2.
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.7017543859649122,
|
| 216 |
-
"grad_norm":
|
| 217 |
-
"learning_rate":
|
| 218 |
-
"loss": 2.
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.7251461988304093,
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate":
|
| 225 |
-
"loss": 2.
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.7485380116959064,
|
| 230 |
-
"grad_norm": 2.
|
| 231 |
-
"learning_rate":
|
| 232 |
-
"loss": 2.
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.7719298245614035,
|
| 237 |
-
"grad_norm":
|
| 238 |
-
"learning_rate":
|
| 239 |
-
"loss": 2.
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.7953216374269005,
|
| 244 |
-
"grad_norm":
|
| 245 |
-
"learning_rate":
|
| 246 |
-
"loss": 2.
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.8187134502923976,
|
| 251 |
-
"grad_norm":
|
| 252 |
-
"learning_rate":
|
| 253 |
-
"loss": 2.
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.8421052631578947,
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 2.
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.8654970760233918,
|
| 265 |
-
"grad_norm":
|
| 266 |
-
"learning_rate":
|
| 267 |
-
"loss": 2.
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.8888888888888888,
|
| 272 |
-
"grad_norm": 2.
|
| 273 |
-
"learning_rate":
|
| 274 |
-
"loss": 2.
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.9122807017543859,
|
| 279 |
-
"grad_norm": 3.
|
| 280 |
-
"learning_rate":
|
| 281 |
-
"loss": 2.
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.935672514619883,
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate":
|
| 288 |
-
"loss": 2.
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.9590643274853801,
|
| 293 |
-
"grad_norm":
|
| 294 |
-
"learning_rate":
|
| 295 |
-
"loss": 2.
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.9824561403508771,
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 2.
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 1.0,
|
| 307 |
-
"grad_norm":
|
| 308 |
-
"learning_rate":
|
| 309 |
"loss": 1.9135,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 1.0,
|
| 314 |
-
"eval_loss": 0.
|
| 315 |
-
"eval_runtime": 3.
|
| 316 |
-
"eval_samples_per_second":
|
| 317 |
-
"eval_steps_per_second":
|
| 318 |
"step": 43
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 1.023391812865497,
|
| 322 |
-
"grad_norm": 5.024760723114014,
|
| 323 |
-
"learning_rate": 7.440476190476191e-05,
|
| 324 |
-
"loss": 2.5518,
|
| 325 |
-
"step": 44
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 1.0467836257309941,
|
| 329 |
-
"grad_norm": 3.8648533821105957,
|
| 330 |
-
"learning_rate": 7.380952380952382e-05,
|
| 331 |
-
"loss": 2.5287,
|
| 332 |
-
"step": 45
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 1.0701754385964912,
|
| 336 |
-
"grad_norm": 2.0299911499023438,
|
| 337 |
-
"learning_rate": 7.321428571428571e-05,
|
| 338 |
-
"loss": 2.5216,
|
| 339 |
-
"step": 46
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.0935672514619883,
|
| 343 |
-
"grad_norm": 3.747516632080078,
|
| 344 |
-
"learning_rate": 7.261904761904762e-05,
|
| 345 |
-
"loss": 2.6046,
|
| 346 |
-
"step": 47
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 1.1169590643274854,
|
| 350 |
-
"grad_norm": 3.7166852951049805,
|
| 351 |
-
"learning_rate": 7.202380952380953e-05,
|
| 352 |
-
"loss": 2.5714,
|
| 353 |
-
"step": 48
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 1.1403508771929824,
|
| 357 |
-
"grad_norm": 5.546866416931152,
|
| 358 |
-
"learning_rate": 7.142857142857143e-05,
|
| 359 |
-
"loss": 2.555,
|
| 360 |
-
"step": 49
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 1.1637426900584795,
|
| 364 |
-
"grad_norm": 3.976203441619873,
|
| 365 |
-
"learning_rate": 7.083333333333334e-05,
|
| 366 |
-
"loss": 2.5272,
|
| 367 |
-
"step": 50
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 1.1871345029239766,
|
| 371 |
-
"grad_norm": 2.14218807220459,
|
| 372 |
-
"learning_rate": 7.023809523809524e-05,
|
| 373 |
-
"loss": 2.5428,
|
| 374 |
-
"step": 51
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 1.2105263157894737,
|
| 378 |
-
"grad_norm": 3.9988062381744385,
|
| 379 |
-
"learning_rate": 6.964285714285715e-05,
|
| 380 |
-
"loss": 2.5357,
|
| 381 |
-
"step": 52
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.2339181286549707,
|
| 385 |
-
"grad_norm": 3.975252151489258,
|
| 386 |
-
"learning_rate": 6.904761904761905e-05,
|
| 387 |
-
"loss": 2.6062,
|
| 388 |
-
"step": 53
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 1.2573099415204678,
|
| 392 |
-
"grad_norm": 2.3471555709838867,
|
| 393 |
-
"learning_rate": 6.845238095238096e-05,
|
| 394 |
-
"loss": 2.5316,
|
| 395 |
-
"step": 54
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 1.280701754385965,
|
| 399 |
-
"grad_norm": 1.675955057144165,
|
| 400 |
-
"learning_rate": 6.785714285714286e-05,
|
| 401 |
-
"loss": 2.5567,
|
| 402 |
-
"step": 55
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 1.304093567251462,
|
| 406 |
-
"grad_norm": 1.435637354850769,
|
| 407 |
-
"learning_rate": 6.726190476190477e-05,
|
| 408 |
-
"loss": 2.493,
|
| 409 |
-
"step": 56
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 1.327485380116959,
|
| 413 |
-
"grad_norm": 1.7799092531204224,
|
| 414 |
-
"learning_rate": 6.666666666666667e-05,
|
| 415 |
-
"loss": 2.4916,
|
| 416 |
-
"step": 57
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 1.3508771929824561,
|
| 420 |
-
"grad_norm": 2.0183417797088623,
|
| 421 |
-
"learning_rate": 6.607142857142857e-05,
|
| 422 |
-
"loss": 2.523,
|
| 423 |
-
"step": 58
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 1.3742690058479532,
|
| 427 |
-
"grad_norm": 2.379734992980957,
|
| 428 |
-
"learning_rate": 6.547619047619048e-05,
|
| 429 |
-
"loss": 2.5916,
|
| 430 |
-
"step": 59
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 1.3976608187134503,
|
| 434 |
-
"grad_norm": 4.267402172088623,
|
| 435 |
-
"learning_rate": 6.488095238095238e-05,
|
| 436 |
-
"loss": 2.5291,
|
| 437 |
-
"step": 60
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 1.4210526315789473,
|
| 441 |
-
"grad_norm": 5.298277378082275,
|
| 442 |
-
"learning_rate": 6.428571428571429e-05,
|
| 443 |
-
"loss": 2.5419,
|
| 444 |
-
"step": 61
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 1.4444444444444444,
|
| 448 |
-
"grad_norm": 2.4158201217651367,
|
| 449 |
-
"learning_rate": 6.369047619047619e-05,
|
| 450 |
-
"loss": 2.565,
|
| 451 |
-
"step": 62
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 1.4678362573099415,
|
| 455 |
-
"grad_norm": 3.903258800506592,
|
| 456 |
-
"learning_rate": 6.30952380952381e-05,
|
| 457 |
-
"loss": 2.5234,
|
| 458 |
-
"step": 63
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 1.4912280701754386,
|
| 462 |
-
"grad_norm": 1.7788355350494385,
|
| 463 |
-
"learning_rate": 6.25e-05,
|
| 464 |
-
"loss": 2.6098,
|
| 465 |
-
"step": 64
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 1.5146198830409356,
|
| 469 |
-
"grad_norm": 4.656391143798828,
|
| 470 |
-
"learning_rate": 6.19047619047619e-05,
|
| 471 |
-
"loss": 2.5448,
|
| 472 |
-
"step": 65
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 1.5380116959064327,
|
| 476 |
-
"grad_norm": 3.0788934230804443,
|
| 477 |
-
"learning_rate": 6.130952380952381e-05,
|
| 478 |
-
"loss": 2.5008,
|
| 479 |
-
"step": 66
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 1.5614035087719298,
|
| 483 |
-
"grad_norm": 1.855888843536377,
|
| 484 |
-
"learning_rate": 6.0714285714285715e-05,
|
| 485 |
-
"loss": 2.5913,
|
| 486 |
-
"step": 67
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 1.5847953216374269,
|
| 490 |
-
"grad_norm": 2.3010807037353516,
|
| 491 |
-
"learning_rate": 6.011904761904762e-05,
|
| 492 |
-
"loss": 2.6243,
|
| 493 |
-
"step": 68
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 1.608187134502924,
|
| 497 |
-
"grad_norm": 3.488280773162842,
|
| 498 |
-
"learning_rate": 5.9523809523809524e-05,
|
| 499 |
-
"loss": 2.5247,
|
| 500 |
-
"step": 69
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 1.631578947368421,
|
| 504 |
-
"grad_norm": 2.508605480194092,
|
| 505 |
-
"learning_rate": 5.8928571428571435e-05,
|
| 506 |
-
"loss": 2.4904,
|
| 507 |
-
"step": 70
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.654970760233918,
|
| 511 |
-
"grad_norm": 2.490065336227417,
|
| 512 |
-
"learning_rate": 5.833333333333334e-05,
|
| 513 |
-
"loss": 2.5192,
|
| 514 |
-
"step": 71
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 1.6783625730994152,
|
| 518 |
-
"grad_norm": 2.5170626640319824,
|
| 519 |
-
"learning_rate": 5.773809523809524e-05,
|
| 520 |
-
"loss": 2.4772,
|
| 521 |
-
"step": 72
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 1.7017543859649122,
|
| 525 |
-
"grad_norm": 2.740570306777954,
|
| 526 |
-
"learning_rate": 5.714285714285714e-05,
|
| 527 |
-
"loss": 2.5289,
|
| 528 |
-
"step": 73
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 1.7251461988304093,
|
| 532 |
-
"grad_norm": 3.663625717163086,
|
| 533 |
-
"learning_rate": 5.6547619047619046e-05,
|
| 534 |
-
"loss": 2.4998,
|
| 535 |
-
"step": 74
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 1.7485380116959064,
|
| 539 |
-
"grad_norm": 3.0818302631378174,
|
| 540 |
-
"learning_rate": 5.595238095238096e-05,
|
| 541 |
-
"loss": 2.5162,
|
| 542 |
-
"step": 75
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 1.7719298245614035,
|
| 546 |
-
"grad_norm": 4.646051406860352,
|
| 547 |
-
"learning_rate": 5.535714285714286e-05,
|
| 548 |
-
"loss": 2.4664,
|
| 549 |
-
"step": 76
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 1.7953216374269005,
|
| 553 |
-
"grad_norm": 5.245360374450684,
|
| 554 |
-
"learning_rate": 5.4761904761904766e-05,
|
| 555 |
-
"loss": 2.5579,
|
| 556 |
-
"step": 77
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 1.8187134502923976,
|
| 560 |
-
"grad_norm": 2.5500850677490234,
|
| 561 |
-
"learning_rate": 5.4166666666666664e-05,
|
| 562 |
-
"loss": 2.5464,
|
| 563 |
-
"step": 78
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 1.8421052631578947,
|
| 567 |
-
"grad_norm": 3.6748623847961426,
|
| 568 |
-
"learning_rate": 5.3571428571428575e-05,
|
| 569 |
-
"loss": 2.533,
|
| 570 |
-
"step": 79
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 1.8654970760233918,
|
| 574 |
-
"grad_norm": 3.894373893737793,
|
| 575 |
-
"learning_rate": 5.297619047619048e-05,
|
| 576 |
-
"loss": 2.5526,
|
| 577 |
-
"step": 80
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 1.8888888888888888,
|
| 581 |
-
"grad_norm": 2.844043731689453,
|
| 582 |
-
"learning_rate": 5.2380952380952384e-05,
|
| 583 |
-
"loss": 2.5099,
|
| 584 |
-
"step": 81
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 1.912280701754386,
|
| 588 |
-
"grad_norm": 3.2156522274017334,
|
| 589 |
-
"learning_rate": 5.1785714285714296e-05,
|
| 590 |
-
"loss": 2.4995,
|
| 591 |
-
"step": 82
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 1.935672514619883,
|
| 595 |
-
"grad_norm": 3.4957337379455566,
|
| 596 |
-
"learning_rate": 5.119047619047619e-05,
|
| 597 |
-
"loss": 2.4958,
|
| 598 |
-
"step": 83
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 1.95906432748538,
|
| 602 |
-
"grad_norm": 2.481201648712158,
|
| 603 |
-
"learning_rate": 5.05952380952381e-05,
|
| 604 |
-
"loss": 2.5521,
|
| 605 |
-
"step": 84
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 1.9824561403508771,
|
| 609 |
-
"grad_norm": 1.6848102807998657,
|
| 610 |
-
"learning_rate": 5e-05,
|
| 611 |
-
"loss": 2.4841,
|
| 612 |
-
"step": 85
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 2.0,
|
| 616 |
-
"grad_norm": 1.674892783164978,
|
| 617 |
-
"learning_rate": 4.940476190476191e-05,
|
| 618 |
-
"loss": 1.931,
|
| 619 |
-
"step": 86
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 2.0,
|
| 623 |
-
"eval_loss": 0.31422847509384155,
|
| 624 |
-
"eval_runtime": 3.4783,
|
| 625 |
-
"eval_samples_per_second": 62.1,
|
| 626 |
-
"eval_steps_per_second": 15.525,
|
| 627 |
-
"step": 86
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 2.023391812865497,
|
| 631 |
-
"grad_norm": 2.143749952316284,
|
| 632 |
-
"learning_rate": 4.880952380952381e-05,
|
| 633 |
-
"loss": 2.5044,
|
| 634 |
-
"step": 87
|
| 635 |
-
},
|
| 636 |
-
{
|
| 637 |
-
"epoch": 2.046783625730994,
|
| 638 |
-
"grad_norm": 1.8259425163269043,
|
| 639 |
-
"learning_rate": 4.8214285714285716e-05,
|
| 640 |
-
"loss": 2.5145,
|
| 641 |
-
"step": 88
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch": 2.0701754385964914,
|
| 645 |
-
"grad_norm": 2.2776169776916504,
|
| 646 |
-
"learning_rate": 4.761904761904762e-05,
|
| 647 |
-
"loss": 2.5045,
|
| 648 |
-
"step": 89
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 2.0935672514619883,
|
| 652 |
-
"grad_norm": 2.4573814868927,
|
| 653 |
-
"learning_rate": 4.7023809523809525e-05,
|
| 654 |
-
"loss": 2.4763,
|
| 655 |
-
"step": 90
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 2.116959064327485,
|
| 659 |
-
"grad_norm": 1.7515789270401,
|
| 660 |
-
"learning_rate": 4.642857142857143e-05,
|
| 661 |
-
"loss": 2.4691,
|
| 662 |
-
"step": 91
|
| 663 |
-
},
|
| 664 |
-
{
|
| 665 |
-
"epoch": 2.1403508771929824,
|
| 666 |
-
"grad_norm": 3.7100000381469727,
|
| 667 |
-
"learning_rate": 4.5833333333333334e-05,
|
| 668 |
-
"loss": 2.5116,
|
| 669 |
-
"step": 92
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 2.1637426900584797,
|
| 673 |
-
"grad_norm": 4.203453540802002,
|
| 674 |
-
"learning_rate": 4.523809523809524e-05,
|
| 675 |
-
"loss": 2.4578,
|
| 676 |
-
"step": 93
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 2.1871345029239766,
|
| 680 |
-
"grad_norm": 2.1631312370300293,
|
| 681 |
-
"learning_rate": 4.464285714285715e-05,
|
| 682 |
-
"loss": 2.4542,
|
| 683 |
-
"step": 94
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 2.2105263157894735,
|
| 687 |
-
"grad_norm": 2.008855104446411,
|
| 688 |
-
"learning_rate": 4.404761904761905e-05,
|
| 689 |
-
"loss": 2.4875,
|
| 690 |
-
"step": 95
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 2.2339181286549707,
|
| 694 |
-
"grad_norm": 2.7923460006713867,
|
| 695 |
-
"learning_rate": 4.345238095238096e-05,
|
| 696 |
-
"loss": 2.4972,
|
| 697 |
-
"step": 96
|
| 698 |
-
},
|
| 699 |
-
{
|
| 700 |
-
"epoch": 2.257309941520468,
|
| 701 |
-
"grad_norm": 1.9210076332092285,
|
| 702 |
-
"learning_rate": 4.2857142857142856e-05,
|
| 703 |
-
"loss": 2.5156,
|
| 704 |
-
"step": 97
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 2.280701754385965,
|
| 708 |
-
"grad_norm": 2.0055480003356934,
|
| 709 |
-
"learning_rate": 4.226190476190476e-05,
|
| 710 |
-
"loss": 2.4456,
|
| 711 |
-
"step": 98
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 2.3040935672514617,
|
| 715 |
-
"grad_norm": 1.6126084327697754,
|
| 716 |
-
"learning_rate": 4.166666666666667e-05,
|
| 717 |
-
"loss": 2.4566,
|
| 718 |
-
"step": 99
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 2.327485380116959,
|
| 722 |
-
"grad_norm": 2.9103384017944336,
|
| 723 |
-
"learning_rate": 4.107142857142857e-05,
|
| 724 |
-
"loss": 2.4987,
|
| 725 |
-
"step": 100
|
| 726 |
-
},
|
| 727 |
-
{
|
| 728 |
-
"epoch": 2.3508771929824563,
|
| 729 |
-
"grad_norm": 2.317208766937256,
|
| 730 |
-
"learning_rate": 4.047619047619048e-05,
|
| 731 |
-
"loss": 2.4523,
|
| 732 |
-
"step": 101
|
| 733 |
-
},
|
| 734 |
-
{
|
| 735 |
-
"epoch": 2.374269005847953,
|
| 736 |
-
"grad_norm": 1.8047789335250854,
|
| 737 |
-
"learning_rate": 3.9880952380952386e-05,
|
| 738 |
-
"loss": 2.4627,
|
| 739 |
-
"step": 102
|
| 740 |
-
},
|
| 741 |
-
{
|
| 742 |
-
"epoch": 2.39766081871345,
|
| 743 |
-
"grad_norm": 2.57161545753479,
|
| 744 |
-
"learning_rate": 3.928571428571429e-05,
|
| 745 |
-
"loss": 2.4551,
|
| 746 |
-
"step": 103
|
| 747 |
-
},
|
| 748 |
-
{
|
| 749 |
-
"epoch": 2.4210526315789473,
|
| 750 |
-
"grad_norm": 1.9964399337768555,
|
| 751 |
-
"learning_rate": 3.8690476190476195e-05,
|
| 752 |
-
"loss": 2.452,
|
| 753 |
-
"step": 104
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 2.4444444444444446,
|
| 757 |
-
"grad_norm": 1.4267529249191284,
|
| 758 |
-
"learning_rate": 3.809523809523809e-05,
|
| 759 |
-
"loss": 2.4888,
|
| 760 |
-
"step": 105
|
| 761 |
-
},
|
| 762 |
-
{
|
| 763 |
-
"epoch": 2.4678362573099415,
|
| 764 |
-
"grad_norm": 3.7493860721588135,
|
| 765 |
-
"learning_rate": 3.7500000000000003e-05,
|
| 766 |
-
"loss": 2.5197,
|
| 767 |
-
"step": 106
|
| 768 |
-
},
|
| 769 |
-
{
|
| 770 |
-
"epoch": 2.4912280701754383,
|
| 771 |
-
"grad_norm": 2.891245126724243,
|
| 772 |
-
"learning_rate": 3.690476190476191e-05,
|
| 773 |
-
"loss": 2.4384,
|
| 774 |
-
"step": 107
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 2.5146198830409356,
|
| 778 |
-
"grad_norm": 2.032031774520874,
|
| 779 |
-
"learning_rate": 3.630952380952381e-05,
|
| 780 |
-
"loss": 2.4911,
|
| 781 |
-
"step": 108
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 2.538011695906433,
|
| 785 |
-
"grad_norm": 1.5395787954330444,
|
| 786 |
-
"learning_rate": 3.571428571428572e-05,
|
| 787 |
-
"loss": 2.4612,
|
| 788 |
-
"step": 109
|
| 789 |
-
},
|
| 790 |
-
{
|
| 791 |
-
"epoch": 2.56140350877193,
|
| 792 |
-
"grad_norm": 2.0416135787963867,
|
| 793 |
-
"learning_rate": 3.511904761904762e-05,
|
| 794 |
-
"loss": 2.5042,
|
| 795 |
-
"step": 110
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 2.5847953216374266,
|
| 799 |
-
"grad_norm": 1.8836208581924438,
|
| 800 |
-
"learning_rate": 3.4523809523809526e-05,
|
| 801 |
-
"loss": 2.455,
|
| 802 |
-
"step": 111
|
| 803 |
-
},
|
| 804 |
-
{
|
| 805 |
-
"epoch": 2.608187134502924,
|
| 806 |
-
"grad_norm": 2.659785747528076,
|
| 807 |
-
"learning_rate": 3.392857142857143e-05,
|
| 808 |
-
"loss": 2.4787,
|
| 809 |
-
"step": 112
|
| 810 |
-
},
|
| 811 |
-
{
|
| 812 |
-
"epoch": 2.6315789473684212,
|
| 813 |
-
"grad_norm": 1.6700447797775269,
|
| 814 |
-
"learning_rate": 3.3333333333333335e-05,
|
| 815 |
-
"loss": 2.4873,
|
| 816 |
-
"step": 113
|
| 817 |
-
},
|
| 818 |
-
{
|
| 819 |
-
"epoch": 2.654970760233918,
|
| 820 |
-
"grad_norm": 2.6808204650878906,
|
| 821 |
-
"learning_rate": 3.273809523809524e-05,
|
| 822 |
-
"loss": 2.5067,
|
| 823 |
-
"step": 114
|
| 824 |
-
},
|
| 825 |
-
{
|
| 826 |
-
"epoch": 2.678362573099415,
|
| 827 |
-
"grad_norm": 2.243744373321533,
|
| 828 |
-
"learning_rate": 3.2142857142857144e-05,
|
| 829 |
-
"loss": 2.4924,
|
| 830 |
-
"step": 115
|
| 831 |
-
},
|
| 832 |
-
{
|
| 833 |
-
"epoch": 2.7017543859649122,
|
| 834 |
-
"grad_norm": 2.5171408653259277,
|
| 835 |
-
"learning_rate": 3.154761904761905e-05,
|
| 836 |
-
"loss": 2.4808,
|
| 837 |
-
"step": 116
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 2.7251461988304095,
|
| 841 |
-
"grad_norm": 3.511664628982544,
|
| 842 |
-
"learning_rate": 3.095238095238095e-05,
|
| 843 |
-
"loss": 2.4996,
|
| 844 |
-
"step": 117
|
| 845 |
-
},
|
| 846 |
-
{
|
| 847 |
-
"epoch": 2.7485380116959064,
|
| 848 |
-
"grad_norm": 2.554797649383545,
|
| 849 |
-
"learning_rate": 3.0357142857142857e-05,
|
| 850 |
-
"loss": 2.4623,
|
| 851 |
-
"step": 118
|
| 852 |
-
},
|
| 853 |
-
{
|
| 854 |
-
"epoch": 2.7719298245614032,
|
| 855 |
-
"grad_norm": 2.256316900253296,
|
| 856 |
-
"learning_rate": 2.9761904761904762e-05,
|
| 857 |
-
"loss": 2.4806,
|
| 858 |
-
"step": 119
|
| 859 |
-
},
|
| 860 |
-
{
|
| 861 |
-
"epoch": 2.7953216374269005,
|
| 862 |
-
"grad_norm": 2.518251895904541,
|
| 863 |
-
"learning_rate": 2.916666666666667e-05,
|
| 864 |
-
"loss": 2.4336,
|
| 865 |
-
"step": 120
|
| 866 |
-
},
|
| 867 |
-
{
|
| 868 |
-
"epoch": 2.818713450292398,
|
| 869 |
-
"grad_norm": 2.2938737869262695,
|
| 870 |
-
"learning_rate": 2.857142857142857e-05,
|
| 871 |
-
"loss": 2.5024,
|
| 872 |
-
"step": 121
|
| 873 |
-
},
|
| 874 |
-
{
|
| 875 |
-
"epoch": 2.8421052631578947,
|
| 876 |
-
"grad_norm": 2.4636147022247314,
|
| 877 |
-
"learning_rate": 2.797619047619048e-05,
|
| 878 |
-
"loss": 2.5213,
|
| 879 |
-
"step": 122
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 2.8654970760233915,
|
| 883 |
-
"grad_norm": 2.4576895236968994,
|
| 884 |
-
"learning_rate": 2.7380952380952383e-05,
|
| 885 |
-
"loss": 2.4407,
|
| 886 |
-
"step": 123
|
| 887 |
-
},
|
| 888 |
-
{
|
| 889 |
-
"epoch": 2.888888888888889,
|
| 890 |
-
"grad_norm": 2.4760570526123047,
|
| 891 |
-
"learning_rate": 2.6785714285714288e-05,
|
| 892 |
-
"loss": 2.5222,
|
| 893 |
-
"step": 124
|
| 894 |
-
},
|
| 895 |
-
{
|
| 896 |
-
"epoch": 2.912280701754386,
|
| 897 |
-
"grad_norm": 1.9041963815689087,
|
| 898 |
-
"learning_rate": 2.6190476190476192e-05,
|
| 899 |
-
"loss": 2.497,
|
| 900 |
-
"step": 125
|
| 901 |
-
},
|
| 902 |
-
{
|
| 903 |
-
"epoch": 2.935672514619883,
|
| 904 |
-
"grad_norm": 2.592594623565674,
|
| 905 |
-
"learning_rate": 2.5595238095238093e-05,
|
| 906 |
-
"loss": 2.4658,
|
| 907 |
-
"step": 126
|
| 908 |
-
},
|
| 909 |
-
{
|
| 910 |
-
"epoch": 2.95906432748538,
|
| 911 |
-
"grad_norm": 1.9708219766616821,
|
| 912 |
-
"learning_rate": 2.5e-05,
|
| 913 |
-
"loss": 2.469,
|
| 914 |
-
"step": 127
|
| 915 |
-
},
|
| 916 |
-
{
|
| 917 |
-
"epoch": 2.982456140350877,
|
| 918 |
-
"grad_norm": 2.22748064994812,
|
| 919 |
-
"learning_rate": 2.4404761904761906e-05,
|
| 920 |
-
"loss": 2.4454,
|
| 921 |
-
"step": 128
|
| 922 |
-
},
|
| 923 |
-
{
|
| 924 |
-
"epoch": 3.0,
|
| 925 |
-
"grad_norm": 1.8958027362823486,
|
| 926 |
-
"learning_rate": 2.380952380952381e-05,
|
| 927 |
-
"loss": 1.8965,
|
| 928 |
-
"step": 129
|
| 929 |
-
},
|
| 930 |
-
{
|
| 931 |
-
"epoch": 3.0,
|
| 932 |
-
"eval_loss": 0.3103340268135071,
|
| 933 |
-
"eval_runtime": 3.4912,
|
| 934 |
-
"eval_samples_per_second": 61.869,
|
| 935 |
-
"eval_steps_per_second": 15.467,
|
| 936 |
-
"step": 129
|
| 937 |
-
},
|
| 938 |
-
{
|
| 939 |
-
"epoch": 3.023391812865497,
|
| 940 |
-
"grad_norm": 1.7813217639923096,
|
| 941 |
-
"learning_rate": 2.3214285714285715e-05,
|
| 942 |
-
"loss": 2.4409,
|
| 943 |
-
"step": 130
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 3.046783625730994,
|
| 947 |
-
"grad_norm": 1.6289114952087402,
|
| 948 |
-
"learning_rate": 2.261904761904762e-05,
|
| 949 |
-
"loss": 2.4332,
|
| 950 |
-
"step": 131
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 3.0701754385964914,
|
| 954 |
-
"grad_norm": 1.4746161699295044,
|
| 955 |
-
"learning_rate": 2.2023809523809524e-05,
|
| 956 |
-
"loss": 2.4365,
|
| 957 |
-
"step": 132
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 3.0935672514619883,
|
| 961 |
-
"grad_norm": 2.3188133239746094,
|
| 962 |
-
"learning_rate": 2.1428571428571428e-05,
|
| 963 |
-
"loss": 2.4408,
|
| 964 |
-
"step": 133
|
| 965 |
-
},
|
| 966 |
-
{
|
| 967 |
-
"epoch": 3.116959064327485,
|
| 968 |
-
"grad_norm": 1.423759937286377,
|
| 969 |
-
"learning_rate": 2.0833333333333336e-05,
|
| 970 |
-
"loss": 2.3871,
|
| 971 |
-
"step": 134
|
| 972 |
-
},
|
| 973 |
-
{
|
| 974 |
-
"epoch": 3.1403508771929824,
|
| 975 |
-
"grad_norm": 2.078610420227051,
|
| 976 |
-
"learning_rate": 2.023809523809524e-05,
|
| 977 |
-
"loss": 2.4912,
|
| 978 |
-
"step": 135
|
| 979 |
-
},
|
| 980 |
-
{
|
| 981 |
-
"epoch": 3.1637426900584797,
|
| 982 |
-
"grad_norm": 1.9567757844924927,
|
| 983 |
-
"learning_rate": 1.9642857142857145e-05,
|
| 984 |
-
"loss": 2.4318,
|
| 985 |
-
"step": 136
|
| 986 |
-
},
|
| 987 |
-
{
|
| 988 |
-
"epoch": 3.1871345029239766,
|
| 989 |
-
"grad_norm": 2.138343095779419,
|
| 990 |
-
"learning_rate": 1.9047619047619046e-05,
|
| 991 |
-
"loss": 2.4311,
|
| 992 |
-
"step": 137
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 3.2105263157894735,
|
| 996 |
-
"grad_norm": 1.9541492462158203,
|
| 997 |
-
"learning_rate": 1.8452380952380954e-05,
|
| 998 |
-
"loss": 2.4532,
|
| 999 |
-
"step": 138
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 3.2339181286549707,
|
| 1003 |
-
"grad_norm": 1.905840277671814,
|
| 1004 |
-
"learning_rate": 1.785714285714286e-05,
|
| 1005 |
-
"loss": 2.4833,
|
| 1006 |
-
"step": 139
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 3.257309941520468,
|
| 1010 |
-
"grad_norm": 1.5300649404525757,
|
| 1011 |
-
"learning_rate": 1.7261904761904763e-05,
|
| 1012 |
-
"loss": 2.4629,
|
| 1013 |
-
"step": 140
|
| 1014 |
-
},
|
| 1015 |
-
{
|
| 1016 |
-
"epoch": 3.280701754385965,
|
| 1017 |
-
"grad_norm": 1.8787633180618286,
|
| 1018 |
-
"learning_rate": 1.6666666666666667e-05,
|
| 1019 |
-
"loss": 2.4423,
|
| 1020 |
-
"step": 141
|
| 1021 |
-
},
|
| 1022 |
-
{
|
| 1023 |
-
"epoch": 3.3040935672514617,
|
| 1024 |
-
"grad_norm": 1.9487767219543457,
|
| 1025 |
-
"learning_rate": 1.6071428571428572e-05,
|
| 1026 |
-
"loss": 2.4558,
|
| 1027 |
-
"step": 142
|
| 1028 |
-
},
|
| 1029 |
-
{
|
| 1030 |
-
"epoch": 3.327485380116959,
|
| 1031 |
-
"grad_norm": 1.8227450847625732,
|
| 1032 |
-
"learning_rate": 1.5476190476190476e-05,
|
| 1033 |
-
"loss": 2.4575,
|
| 1034 |
-
"step": 143
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 3.3508771929824563,
|
| 1038 |
-
"grad_norm": 1.9631016254425049,
|
| 1039 |
-
"learning_rate": 1.4880952380952381e-05,
|
| 1040 |
-
"loss": 2.4492,
|
| 1041 |
-
"step": 144
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 3.374269005847953,
|
| 1045 |
-
"grad_norm": 1.6142218112945557,
|
| 1046 |
-
"learning_rate": 1.4285714285714285e-05,
|
| 1047 |
-
"loss": 2.4313,
|
| 1048 |
-
"step": 145
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 3.39766081871345,
|
| 1052 |
-
"grad_norm": 1.4569997787475586,
|
| 1053 |
-
"learning_rate": 1.3690476190476192e-05,
|
| 1054 |
-
"loss": 2.3885,
|
| 1055 |
-
"step": 146
|
| 1056 |
-
},
|
| 1057 |
-
{
|
| 1058 |
-
"epoch": 3.4210526315789473,
|
| 1059 |
-
"grad_norm": 2.126253843307495,
|
| 1060 |
-
"learning_rate": 1.3095238095238096e-05,
|
| 1061 |
-
"loss": 2.4366,
|
| 1062 |
-
"step": 147
|
| 1063 |
-
},
|
| 1064 |
-
{
|
| 1065 |
-
"epoch": 3.4444444444444446,
|
| 1066 |
-
"grad_norm": 1.3649730682373047,
|
| 1067 |
-
"learning_rate": 1.25e-05,
|
| 1068 |
-
"loss": 2.4278,
|
| 1069 |
-
"step": 148
|
| 1070 |
-
},
|
| 1071 |
-
{
|
| 1072 |
-
"epoch": 3.4678362573099415,
|
| 1073 |
-
"grad_norm": 1.5861579179763794,
|
| 1074 |
-
"learning_rate": 1.1904761904761905e-05,
|
| 1075 |
-
"loss": 2.4695,
|
| 1076 |
-
"step": 149
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 3.4912280701754383,
|
| 1080 |
-
"grad_norm": 1.743173360824585,
|
| 1081 |
-
"learning_rate": 1.130952380952381e-05,
|
| 1082 |
-
"loss": 2.4422,
|
| 1083 |
-
"step": 150
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 3.5146198830409356,
|
| 1087 |
-
"grad_norm": 1.6894521713256836,
|
| 1088 |
-
"learning_rate": 1.0714285714285714e-05,
|
| 1089 |
-
"loss": 2.4343,
|
| 1090 |
-
"step": 151
|
| 1091 |
-
},
|
| 1092 |
-
{
|
| 1093 |
-
"epoch": 3.538011695906433,
|
| 1094 |
-
"grad_norm": 1.4037004709243774,
|
| 1095 |
-
"learning_rate": 1.011904761904762e-05,
|
| 1096 |
-
"loss": 2.4197,
|
| 1097 |
-
"step": 152
|
| 1098 |
-
},
|
| 1099 |
-
{
|
| 1100 |
-
"epoch": 3.56140350877193,
|
| 1101 |
-
"grad_norm": 2.512463092803955,
|
| 1102 |
-
"learning_rate": 9.523809523809523e-06,
|
| 1103 |
-
"loss": 2.4515,
|
| 1104 |
-
"step": 153
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 3.5847953216374266,
|
| 1108 |
-
"grad_norm": 1.968013048171997,
|
| 1109 |
-
"learning_rate": 8.92857142857143e-06,
|
| 1110 |
-
"loss": 2.4602,
|
| 1111 |
-
"step": 154
|
| 1112 |
-
},
|
| 1113 |
-
{
|
| 1114 |
-
"epoch": 3.608187134502924,
|
| 1115 |
-
"grad_norm": 1.521202802658081,
|
| 1116 |
-
"learning_rate": 8.333333333333334e-06,
|
| 1117 |
-
"loss": 2.4167,
|
| 1118 |
-
"step": 155
|
| 1119 |
-
},
|
| 1120 |
-
{
|
| 1121 |
-
"epoch": 3.6315789473684212,
|
| 1122 |
-
"grad_norm": 1.4753526449203491,
|
| 1123 |
-
"learning_rate": 7.738095238095238e-06,
|
| 1124 |
-
"loss": 2.4613,
|
| 1125 |
-
"step": 156
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 3.654970760233918,
|
| 1129 |
-
"grad_norm": 2.0607075691223145,
|
| 1130 |
-
"learning_rate": 7.142857142857143e-06,
|
| 1131 |
-
"loss": 2.4299,
|
| 1132 |
-
"step": 157
|
| 1133 |
-
},
|
| 1134 |
-
{
|
| 1135 |
-
"epoch": 3.678362573099415,
|
| 1136 |
-
"grad_norm": 1.3630319833755493,
|
| 1137 |
-
"learning_rate": 6.547619047619048e-06,
|
| 1138 |
-
"loss": 2.4417,
|
| 1139 |
-
"step": 158
|
| 1140 |
-
},
|
| 1141 |
-
{
|
| 1142 |
-
"epoch": 3.7017543859649122,
|
| 1143 |
-
"grad_norm": 1.2382704019546509,
|
| 1144 |
-
"learning_rate": 5.9523809523809525e-06,
|
| 1145 |
-
"loss": 2.43,
|
| 1146 |
-
"step": 159
|
| 1147 |
-
},
|
| 1148 |
-
{
|
| 1149 |
-
"epoch": 3.7251461988304095,
|
| 1150 |
-
"grad_norm": 1.4374663829803467,
|
| 1151 |
-
"learning_rate": 5.357142857142857e-06,
|
| 1152 |
-
"loss": 2.4523,
|
| 1153 |
-
"step": 160
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 3.7485380116959064,
|
| 1157 |
-
"grad_norm": 1.6586753129959106,
|
| 1158 |
-
"learning_rate": 4.7619047619047615e-06,
|
| 1159 |
-
"loss": 2.4755,
|
| 1160 |
-
"step": 161
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 3.7719298245614032,
|
| 1164 |
-
"grad_norm": 1.7251605987548828,
|
| 1165 |
-
"learning_rate": 4.166666666666667e-06,
|
| 1166 |
-
"loss": 2.5089,
|
| 1167 |
-
"step": 162
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 3.7953216374269005,
|
| 1171 |
-
"grad_norm": 1.5588628053665161,
|
| 1172 |
-
"learning_rate": 3.5714285714285714e-06,
|
| 1173 |
-
"loss": 2.4195,
|
| 1174 |
-
"step": 163
|
| 1175 |
-
},
|
| 1176 |
-
{
|
| 1177 |
-
"epoch": 3.818713450292398,
|
| 1178 |
-
"grad_norm": 1.722469449043274,
|
| 1179 |
-
"learning_rate": 2.9761904761904763e-06,
|
| 1180 |
-
"loss": 2.4104,
|
| 1181 |
-
"step": 164
|
| 1182 |
-
},
|
| 1183 |
-
{
|
| 1184 |
-
"epoch": 3.8421052631578947,
|
| 1185 |
-
"grad_norm": 1.9132815599441528,
|
| 1186 |
-
"learning_rate": 2.3809523809523808e-06,
|
| 1187 |
-
"loss": 2.4492,
|
| 1188 |
-
"step": 165
|
| 1189 |
-
},
|
| 1190 |
-
{
|
| 1191 |
-
"epoch": 3.8654970760233915,
|
| 1192 |
-
"grad_norm": 2.082366704940796,
|
| 1193 |
-
"learning_rate": 1.7857142857142857e-06,
|
| 1194 |
-
"loss": 2.4535,
|
| 1195 |
-
"step": 166
|
| 1196 |
-
},
|
| 1197 |
-
{
|
| 1198 |
-
"epoch": 3.888888888888889,
|
| 1199 |
-
"grad_norm": 1.7362799644470215,
|
| 1200 |
-
"learning_rate": 1.1904761904761904e-06,
|
| 1201 |
-
"loss": 2.4908,
|
| 1202 |
-
"step": 167
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 3.912280701754386,
|
| 1206 |
-
"grad_norm": 1.6625697612762451,
|
| 1207 |
-
"learning_rate": 5.952380952380952e-07,
|
| 1208 |
-
"loss": 2.4552,
|
| 1209 |
-
"step": 168
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 3.912280701754386,
|
| 1213 |
-
"eval_loss": 0.3090362250804901,
|
| 1214 |
-
"eval_runtime": 3.7281,
|
| 1215 |
-
"eval_samples_per_second": 57.938,
|
| 1216 |
-
"eval_steps_per_second": 14.484,
|
| 1217 |
-
"step": 168
|
| 1218 |
}
|
| 1219 |
],
|
| 1220 |
"logging_steps": 1,
|
| 1221 |
-
"max_steps":
|
| 1222 |
"num_input_tokens_seen": 0,
|
| 1223 |
-
"num_train_epochs":
|
| 1224 |
"save_steps": 500,
|
| 1225 |
"stateful_callbacks": {
|
| 1226 |
"TrainerControl": {
|
|
@@ -1229,12 +330,12 @@
|
|
| 1229 |
"should_evaluate": false,
|
| 1230 |
"should_log": false,
|
| 1231 |
"should_save": true,
|
| 1232 |
-
"should_training_stop":
|
| 1233 |
},
|
| 1234 |
"attributes": {}
|
| 1235 |
}
|
| 1236 |
},
|
| 1237 |
-
"total_flos":
|
| 1238 |
"train_batch_size": 12,
|
| 1239 |
"trial_name": null,
|
| 1240 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.31598106026649475,
|
| 3 |
+
"best_model_checkpoint": "./opt_trained/checkpoint-43",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 43,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.023391812865497075,
|
| 13 |
+
"grad_norm": 1.4262398481369019,
|
| 14 |
+
"learning_rate": 9.995238095238095e-05,
|
| 15 |
+
"loss": 2.4722,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.04678362573099415,
|
| 20 |
+
"grad_norm": 12.206507682800293,
|
| 21 |
+
"learning_rate": 9.990476190476191e-05,
|
| 22 |
+
"loss": 3.176,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.07017543859649122,
|
| 27 |
+
"grad_norm": 11.578840255737305,
|
| 28 |
+
"learning_rate": 9.985714285714287e-05,
|
| 29 |
+
"loss": 2.8396,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.0935672514619883,
|
| 34 |
+
"grad_norm": 9.648161888122559,
|
| 35 |
+
"learning_rate": 9.980952380952382e-05,
|
| 36 |
+
"loss": 2.7336,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.11695906432748537,
|
| 41 |
+
"grad_norm": 3.8782691955566406,
|
| 42 |
+
"learning_rate": 9.976190476190477e-05,
|
| 43 |
+
"loss": 2.6955,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.14035087719298245,
|
| 48 |
+
"grad_norm": 12.644487380981445,
|
| 49 |
+
"learning_rate": 9.971428571428571e-05,
|
| 50 |
+
"loss": 2.6405,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.16374269005847952,
|
| 55 |
+
"grad_norm": 11.057122230529785,
|
| 56 |
+
"learning_rate": 9.966666666666667e-05,
|
| 57 |
+
"loss": 2.7113,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.1871345029239766,
|
| 62 |
+
"grad_norm": 4.860190391540527,
|
| 63 |
+
"learning_rate": 9.961904761904762e-05,
|
| 64 |
+
"loss": 2.7076,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.21052631578947367,
|
| 69 |
+
"grad_norm": 4.317215442657471,
|
| 70 |
+
"learning_rate": 9.957142857142858e-05,
|
| 71 |
+
"loss": 2.6377,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.23391812865497075,
|
| 76 |
+
"grad_norm": 3.3068416118621826,
|
| 77 |
+
"learning_rate": 9.952380952380953e-05,
|
| 78 |
+
"loss": 2.5995,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.2573099415204678,
|
| 83 |
+
"grad_norm": 1.2752724885940552,
|
| 84 |
+
"learning_rate": 9.947619047619048e-05,
|
| 85 |
+
"loss": 2.6285,
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.2807017543859649,
|
| 90 |
+
"grad_norm": 8.849737167358398,
|
| 91 |
+
"learning_rate": 9.942857142857144e-05,
|
| 92 |
+
"loss": 2.6217,
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.30409356725146197,
|
| 97 |
+
"grad_norm": 5.594025611877441,
|
| 98 |
+
"learning_rate": 9.938095238095238e-05,
|
| 99 |
+
"loss": 2.6265,
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.32748538011695905,
|
| 104 |
+
"grad_norm": 3.581617593765259,
|
| 105 |
+
"learning_rate": 9.933333333333334e-05,
|
| 106 |
+
"loss": 2.5984,
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.3508771929824561,
|
| 111 |
+
"grad_norm": 5.33600378036499,
|
| 112 |
+
"learning_rate": 9.92857142857143e-05,
|
| 113 |
+
"loss": 2.6071,
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.3742690058479532,
|
| 118 |
+
"grad_norm": 4.4274983406066895,
|
| 119 |
+
"learning_rate": 9.923809523809524e-05,
|
| 120 |
+
"loss": 2.5908,
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.39766081871345027,
|
| 125 |
+
"grad_norm": 4.5507307052612305,
|
| 126 |
+
"learning_rate": 9.91904761904762e-05,
|
| 127 |
+
"loss": 2.5616,
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.42105263157894735,
|
| 132 |
+
"grad_norm": 3.189161777496338,
|
| 133 |
+
"learning_rate": 9.914285714285715e-05,
|
| 134 |
+
"loss": 2.5784,
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.4444444444444444,
|
| 139 |
+
"grad_norm": 2.706615924835205,
|
| 140 |
+
"learning_rate": 9.909523809523809e-05,
|
| 141 |
+
"loss": 2.5413,
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.4678362573099415,
|
| 146 |
+
"grad_norm": 2.146662712097168,
|
| 147 |
+
"learning_rate": 9.904761904761905e-05,
|
| 148 |
+
"loss": 2.6019,
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.49122807017543857,
|
| 153 |
+
"grad_norm": 3.2252964973449707,
|
| 154 |
+
"learning_rate": 9.900000000000001e-05,
|
| 155 |
+
"loss": 2.5914,
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.5146198830409356,
|
| 160 |
+
"grad_norm": 2.8859879970550537,
|
| 161 |
+
"learning_rate": 9.895238095238095e-05,
|
| 162 |
+
"loss": 2.6133,
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.5380116959064327,
|
| 167 |
+
"grad_norm": 2.5647897720336914,
|
| 168 |
+
"learning_rate": 9.890476190476191e-05,
|
| 169 |
+
"loss": 2.5425,
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.5614035087719298,
|
| 174 |
+
"grad_norm": 3.0347073078155518,
|
| 175 |
+
"learning_rate": 9.885714285714286e-05,
|
| 176 |
+
"loss": 2.5732,
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.5847953216374269,
|
| 181 |
+
"grad_norm": 1.8412858247756958,
|
| 182 |
+
"learning_rate": 9.880952380952381e-05,
|
| 183 |
+
"loss": 2.5776,
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.6081871345029239,
|
| 188 |
+
"grad_norm": 3.1820366382598877,
|
| 189 |
+
"learning_rate": 9.876190476190477e-05,
|
| 190 |
+
"loss": 2.5566,
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.631578947368421,
|
| 195 |
+
"grad_norm": 2.4613687992095947,
|
| 196 |
+
"learning_rate": 9.871428571428572e-05,
|
| 197 |
+
"loss": 2.5113,
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.6549707602339181,
|
| 202 |
+
"grad_norm": 5.942374229431152,
|
| 203 |
+
"learning_rate": 9.866666666666668e-05,
|
| 204 |
+
"loss": 2.575,
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.6783625730994152,
|
| 209 |
+
"grad_norm": 7.427689075469971,
|
| 210 |
+
"learning_rate": 9.861904761904762e-05,
|
| 211 |
+
"loss": 2.5692,
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.7017543859649122,
|
| 216 |
+
"grad_norm": 3.163085699081421,
|
| 217 |
+
"learning_rate": 9.857142857142858e-05,
|
| 218 |
+
"loss": 2.5192,
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.7251461988304093,
|
| 223 |
+
"grad_norm": 3.421778678894043,
|
| 224 |
+
"learning_rate": 9.852380952380952e-05,
|
| 225 |
+
"loss": 2.5533,
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.7485380116959064,
|
| 230 |
+
"grad_norm": 2.486320734024048,
|
| 231 |
+
"learning_rate": 9.847619047619048e-05,
|
| 232 |
+
"loss": 2.515,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.7719298245614035,
|
| 237 |
+
"grad_norm": 3.2825722694396973,
|
| 238 |
+
"learning_rate": 9.842857142857144e-05,
|
| 239 |
+
"loss": 2.4994,
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.7953216374269005,
|
| 244 |
+
"grad_norm": 3.5643672943115234,
|
| 245 |
+
"learning_rate": 9.838095238095238e-05,
|
| 246 |
+
"loss": 2.5327,
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.8187134502923976,
|
| 251 |
+
"grad_norm": 3.3127200603485107,
|
| 252 |
+
"learning_rate": 9.833333333333333e-05,
|
| 253 |
+
"loss": 2.6251,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.8421052631578947,
|
| 258 |
+
"grad_norm": 3.187095880508423,
|
| 259 |
+
"learning_rate": 9.828571428571429e-05,
|
| 260 |
+
"loss": 2.5233,
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.8654970760233918,
|
| 265 |
+
"grad_norm": 3.3743860721588135,
|
| 266 |
+
"learning_rate": 9.823809523809525e-05,
|
| 267 |
+
"loss": 2.4999,
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.8888888888888888,
|
| 272 |
+
"grad_norm": 2.3684120178222656,
|
| 273 |
+
"learning_rate": 9.81904761904762e-05,
|
| 274 |
+
"loss": 2.5302,
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.9122807017543859,
|
| 279 |
+
"grad_norm": 3.7091619968414307,
|
| 280 |
+
"learning_rate": 9.814285714285715e-05,
|
| 281 |
+
"loss": 2.5003,
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.935672514619883,
|
| 286 |
+
"grad_norm": 4.230418682098389,
|
| 287 |
+
"learning_rate": 9.80952380952381e-05,
|
| 288 |
+
"loss": 2.5379,
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.9590643274853801,
|
| 293 |
+
"grad_norm": 3.5879616737365723,
|
| 294 |
+
"learning_rate": 9.804761904761905e-05,
|
| 295 |
+
"loss": 2.5652,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.9824561403508771,
|
| 300 |
+
"grad_norm": 2.621013879776001,
|
| 301 |
+
"learning_rate": 9.8e-05,
|
| 302 |
+
"loss": 2.5704,
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 1.0,
|
| 307 |
+
"grad_norm": 1.3536667823791504,
|
| 308 |
+
"learning_rate": 9.795238095238097e-05,
|
| 309 |
"loss": 1.9135,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 1.0,
|
| 314 |
+
"eval_loss": 0.31598106026649475,
|
| 315 |
+
"eval_runtime": 3.4639,
|
| 316 |
+
"eval_samples_per_second": 62.358,
|
| 317 |
+
"eval_steps_per_second": 15.59,
|
| 318 |
"step": 43
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 319 |
}
|
| 320 |
],
|
| 321 |
"logging_steps": 1,
|
| 322 |
+
"max_steps": 2100,
|
| 323 |
"num_input_tokens_seen": 0,
|
| 324 |
+
"num_train_epochs": 50,
|
| 325 |
"save_steps": 500,
|
| 326 |
"stateful_callbacks": {
|
| 327 |
"TrainerControl": {
|
|
|
|
| 330 |
"should_evaluate": false,
|
| 331 |
"should_log": false,
|
| 332 |
"should_save": true,
|
| 333 |
+
"should_training_stop": false
|
| 334 |
},
|
| 335 |
"attributes": {}
|
| 336 |
}
|
| 337 |
},
|
| 338 |
+
"total_flos": 2151327228493824.0,
|
| 339 |
"train_batch_size": 12,
|
| 340 |
"trial_name": null,
|
| 341 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01ee0c77a84c59226be8580dfbe625c82d72bc4456ab0e2ca0a40baf78f82e26
|
| 3 |
size 5368
|