Training in progress, step 100, 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 2254879384
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a31fb4e150696fbd81b685c233a7efae3a01549a539473980c0eb76b0c56dc3d
|
| 3 |
size 2254879384
|
last-checkpoint/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:ad2e709ecb45e0af969d49e7f7cbb19ea0dd7bac74b71ef4d5d82503da0757b6
|
| 3 |
+
size 1130472782
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1d565802a8e26c4e8a31328752b7a7fdc186d9401aa008e65697d0ad8c22e33
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8953
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c33e7e06e97e0f115a37527cc80045ed4276cbd1c8253ac1179212ed437b981
|
| 3 |
size 8953
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,2828 +2,712 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 200,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.001,
|
| 14 |
-
"grad_norm": 6.
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
-
"loss": 11.
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.002,
|
| 21 |
-
"grad_norm": 6.
|
| 22 |
-
"learning_rate": 0.
|
| 23 |
-
"loss": 11.
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.003,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 0.
|
| 30 |
-
"loss":
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.004,
|
| 35 |
-
"grad_norm": 3.
|
| 36 |
-
"learning_rate": 0.
|
| 37 |
-
"loss":
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.005,
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss":
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.006,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate": 0.
|
| 51 |
-
"loss":
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.007,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate": 0.
|
| 58 |
-
"loss":
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.008,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate": 0.
|
| 65 |
-
"loss":
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.009,
|
| 70 |
-
"grad_norm": 1.
|
| 71 |
-
"learning_rate": 0.
|
| 72 |
-
"loss": 8.
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.01,
|
| 77 |
-
"grad_norm":
|
| 78 |
-
"learning_rate": 0.
|
| 79 |
-
"loss": 8.
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.011,
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate": 0.
|
| 86 |
-
"loss": 8.
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.012,
|
| 91 |
-
"grad_norm":
|
| 92 |
-
"learning_rate": 0.
|
| 93 |
-
"loss": 8.
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.013,
|
| 98 |
-
"grad_norm": 1.
|
| 99 |
-
"learning_rate": 0.
|
| 100 |
-
"loss": 8.
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.014,
|
| 105 |
-
"grad_norm":
|
| 106 |
-
"learning_rate": 0.
|
| 107 |
-
"loss": 8.
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.015,
|
| 112 |
-
"grad_norm":
|
| 113 |
-
"learning_rate": 0.
|
| 114 |
-
"loss": 8.
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.016,
|
| 119 |
-
"grad_norm":
|
| 120 |
-
"learning_rate": 0.
|
| 121 |
-
"loss": 8.
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.017,
|
| 126 |
-
"grad_norm": 1.
|
| 127 |
-
"learning_rate": 0.
|
| 128 |
-
"loss": 8.
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.018,
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate": 0.
|
| 135 |
-
"loss": 8.
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.019,
|
| 140 |
-
"grad_norm":
|
| 141 |
-
"learning_rate": 0.
|
| 142 |
-
"loss": 8.
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.02,
|
| 147 |
-
"grad_norm":
|
| 148 |
-
"learning_rate": 0.
|
| 149 |
-
"loss": 8.
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.021,
|
| 154 |
-
"grad_norm":
|
| 155 |
-
"learning_rate": 0.
|
| 156 |
-
"loss": 8.
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.022,
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate": 0.
|
| 163 |
-
"loss": 8.
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.023,
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate": 0.
|
| 170 |
-
"loss": 8.
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.024,
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
-
"learning_rate": 0.
|
| 177 |
-
"loss": 8.
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.025,
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
-
"learning_rate": 0.
|
| 184 |
-
"loss": 8.
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.026,
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate": 0.
|
| 191 |
-
"loss": 8.
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.027,
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
-
"learning_rate": 0.
|
| 198 |
-
"loss": 8.
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.028,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate": 0.
|
| 205 |
-
"loss": 8.
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.029,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate": 0.
|
| 212 |
-
"loss":
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.03,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate": 0.
|
| 219 |
-
"loss":
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.031,
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate": 0.
|
| 226 |
-
"loss":
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.032,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate": 0.
|
| 233 |
-
"loss":
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.033,
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate": 0.
|
| 240 |
-
"loss":
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.034,
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
-
"learning_rate": 0.
|
| 247 |
-
"loss":
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.035,
|
| 252 |
-
"grad_norm":
|
| 253 |
-
"learning_rate": 0.
|
| 254 |
-
"loss":
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.036,
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate": 0.
|
| 261 |
-
"loss":
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.037,
|
| 266 |
-
"grad_norm":
|
| 267 |
-
"learning_rate": 0.
|
| 268 |
-
"loss":
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.038,
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
-
"learning_rate": 0.
|
| 275 |
-
"loss":
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.039,
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
-
"learning_rate": 0.
|
| 282 |
-
"loss":
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.04,
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"learning_rate": 0.
|
| 289 |
-
"loss":
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.041,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate": 0.
|
| 296 |
-
"loss":
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.042,
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
-
"learning_rate": 0.
|
| 303 |
-
"loss":
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.043,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
-
"learning_rate": 0.
|
| 310 |
-
"loss":
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.044,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
-
"learning_rate": 0.
|
| 317 |
-
"loss":
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.045,
|
| 322 |
-
"grad_norm":
|
| 323 |
-
"learning_rate": 0.
|
| 324 |
-
"loss":
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.046,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
-
"learning_rate": 0.
|
| 331 |
-
"loss":
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.047,
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
-
"learning_rate": 0.
|
| 338 |
-
"loss":
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.048,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
-
"learning_rate": 0.
|
| 345 |
-
"loss":
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.049,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
-
"learning_rate": 0.
|
| 352 |
-
"loss":
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.05,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
-
"learning_rate": 0.
|
| 359 |
-
"loss":
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.051,
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
-
"learning_rate": 0.
|
| 366 |
-
"loss":
|
| 367 |
"step": 51
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.052,
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
-
"learning_rate": 0.
|
| 373 |
-
"loss":
|
| 374 |
"step": 52
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.053,
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
-
"learning_rate": 0.
|
| 380 |
-
"loss":
|
| 381 |
"step": 53
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.054,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
-
"learning_rate": 0.
|
| 387 |
-
"loss":
|
| 388 |
"step": 54
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.055,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
-
"learning_rate": 0.
|
| 394 |
-
"loss":
|
| 395 |
"step": 55
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.056,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
-
"learning_rate": 0.
|
| 401 |
-
"loss":
|
| 402 |
"step": 56
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.057,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
-
"learning_rate": 0.
|
| 408 |
-
"loss":
|
| 409 |
"step": 57
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.058,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
-
"learning_rate": 0.
|
| 415 |
-
"loss":
|
| 416 |
"step": 58
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.059,
|
| 420 |
-
"grad_norm":
|
| 421 |
-
"learning_rate": 0.
|
| 422 |
-
"loss":
|
| 423 |
"step": 59
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.06,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
-
"learning_rate": 0.
|
| 429 |
-
"loss":
|
| 430 |
"step": 60
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.061,
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate": 0.
|
| 436 |
-
"loss":
|
| 437 |
"step": 61
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.062,
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate": 0.
|
| 443 |
-
"loss":
|
| 444 |
"step": 62
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.063,
|
| 448 |
-
"grad_norm":
|
| 449 |
-
"learning_rate": 0.
|
| 450 |
-
"loss":
|
| 451 |
"step": 63
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.064,
|
| 455 |
-
"grad_norm":
|
| 456 |
-
"learning_rate": 0.
|
| 457 |
-
"loss":
|
| 458 |
"step": 64
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.065,
|
| 462 |
-
"grad_norm":
|
| 463 |
-
"learning_rate": 0.
|
| 464 |
-
"loss":
|
| 465 |
"step": 65
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.066,
|
| 469 |
-
"grad_norm":
|
| 470 |
-
"learning_rate": 0.
|
| 471 |
-
"loss":
|
| 472 |
"step": 66
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.067,
|
| 476 |
-
"grad_norm":
|
| 477 |
-
"learning_rate": 0.
|
| 478 |
-
"loss":
|
| 479 |
"step": 67
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.068,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
-
"learning_rate": 0.
|
| 485 |
-
"loss":
|
| 486 |
"step": 68
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.069,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
-
"learning_rate": 0.
|
| 492 |
-
"loss":
|
| 493 |
"step": 69
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.07,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
-
"learning_rate": 0.
|
| 499 |
-
"loss":
|
| 500 |
"step": 70
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.071,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
-
"learning_rate": 0.
|
| 506 |
-
"loss":
|
| 507 |
"step": 71
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.072,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
-
"learning_rate": 0.
|
| 513 |
-
"loss":
|
| 514 |
"step": 72
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.073,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
-
"learning_rate": 0.
|
| 520 |
-
"loss":
|
| 521 |
"step": 73
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.074,
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
-
"learning_rate": 0.
|
| 527 |
-
"loss":
|
| 528 |
"step": 74
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.075,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
-
"learning_rate": 0.
|
| 534 |
-
"loss":
|
| 535 |
"step": 75
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.076,
|
| 539 |
-
"grad_norm":
|
| 540 |
-
"learning_rate": 0.
|
| 541 |
-
"loss":
|
| 542 |
"step": 76
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.077,
|
| 546 |
-
"grad_norm":
|
| 547 |
-
"learning_rate": 0.
|
| 548 |
-
"loss":
|
| 549 |
"step": 77
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.078,
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
-
"learning_rate": 0.
|
| 555 |
-
"loss":
|
| 556 |
"step": 78
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.079,
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
-
"learning_rate": 0.
|
| 562 |
-
"loss":
|
| 563 |
"step": 79
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.08,
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
-
"learning_rate": 0.
|
| 569 |
-
"loss":
|
| 570 |
"step": 80
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.081,
|
| 574 |
-
"grad_norm": 0.
|
| 575 |
-
"learning_rate": 0.
|
| 576 |
-
"loss":
|
| 577 |
"step": 81
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.082,
|
| 581 |
-
"grad_norm": 0.
|
| 582 |
-
"learning_rate": 0.
|
| 583 |
-
"loss":
|
| 584 |
"step": 82
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.083,
|
| 588 |
-
"grad_norm": 0.
|
| 589 |
-
"learning_rate": 0.
|
| 590 |
-
"loss":
|
| 591 |
"step": 83
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.084,
|
| 595 |
-
"grad_norm": 0.
|
| 596 |
-
"learning_rate": 0.
|
| 597 |
-
"loss":
|
| 598 |
"step": 84
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.085,
|
| 602 |
-
"grad_norm": 0.
|
| 603 |
-
"learning_rate": 0.
|
| 604 |
-
"loss":
|
| 605 |
"step": 85
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.086,
|
| 609 |
-
"grad_norm": 0.
|
| 610 |
-
"learning_rate": 0.
|
| 611 |
-
"loss":
|
| 612 |
"step": 86
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.087,
|
| 616 |
-
"grad_norm": 0.
|
| 617 |
-
"learning_rate": 0.
|
| 618 |
-
"loss":
|
| 619 |
"step": 87
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.088,
|
| 623 |
-
"grad_norm": 0.
|
| 624 |
-
"learning_rate": 0.
|
| 625 |
-
"loss":
|
| 626 |
"step": 88
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.089,
|
| 630 |
-
"grad_norm": 0.
|
| 631 |
-
"learning_rate": 0.
|
| 632 |
-
"loss":
|
| 633 |
"step": 89
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.09,
|
| 637 |
-
"grad_norm": 0.
|
| 638 |
-
"learning_rate": 0.
|
| 639 |
-
"loss":
|
| 640 |
"step": 90
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.091,
|
| 644 |
-
"grad_norm": 0.
|
| 645 |
-
"learning_rate": 0.
|
| 646 |
-
"loss":
|
| 647 |
"step": 91
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.092,
|
| 651 |
-
"grad_norm": 0.
|
| 652 |
-
"learning_rate": 0.
|
| 653 |
-
"loss":
|
| 654 |
"step": 92
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.093,
|
| 658 |
-
"grad_norm": 0.
|
| 659 |
-
"learning_rate": 0.
|
| 660 |
-
"loss":
|
| 661 |
"step": 93
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.094,
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
-
"learning_rate": 0.
|
| 667 |
-
"loss":
|
| 668 |
"step": 94
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.095,
|
| 672 |
-
"grad_norm": 0.
|
| 673 |
-
"learning_rate": 0.
|
| 674 |
-
"loss":
|
| 675 |
"step": 95
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.096,
|
| 679 |
-
"grad_norm": 0.
|
| 680 |
-
"learning_rate": 0.
|
| 681 |
-
"loss":
|
| 682 |
"step": 96
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.097,
|
| 686 |
-
"grad_norm": 0.
|
| 687 |
-
"learning_rate": 0.
|
| 688 |
-
"loss":
|
| 689 |
"step": 97
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.098,
|
| 693 |
-
"grad_norm": 0.
|
| 694 |
-
"learning_rate": 0.
|
| 695 |
-
"loss":
|
| 696 |
"step": 98
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.099,
|
| 700 |
-
"grad_norm": 0.
|
| 701 |
-
"learning_rate": 0.
|
| 702 |
-
"loss":
|
| 703 |
"step": 99
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.1,
|
| 707 |
-
"grad_norm": 0.
|
| 708 |
-
"learning_rate": 0.
|
| 709 |
-
"loss":
|
| 710 |
"step": 100
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.101,
|
| 714 |
-
"grad_norm": 0.4212554097175598,
|
| 715 |
-
"learning_rate": 0.024,
|
| 716 |
-
"loss": 5.925599575042725,
|
| 717 |
-
"step": 101
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 0.102,
|
| 721 |
-
"grad_norm": 0.43675127625465393,
|
| 722 |
-
"learning_rate": 0.024,
|
| 723 |
-
"loss": 5.880373954772949,
|
| 724 |
-
"step": 102
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 0.103,
|
| 728 |
-
"grad_norm": 0.4089275598526001,
|
| 729 |
-
"learning_rate": 0.024,
|
| 730 |
-
"loss": 5.869070053100586,
|
| 731 |
-
"step": 103
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 0.104,
|
| 735 |
-
"grad_norm": 0.3986571133136749,
|
| 736 |
-
"learning_rate": 0.024,
|
| 737 |
-
"loss": 5.865777969360352,
|
| 738 |
-
"step": 104
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 0.105,
|
| 742 |
-
"grad_norm": 0.3812260627746582,
|
| 743 |
-
"learning_rate": 0.024,
|
| 744 |
-
"loss": 5.828693389892578,
|
| 745 |
-
"step": 105
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 0.106,
|
| 749 |
-
"grad_norm": 0.3968827724456787,
|
| 750 |
-
"learning_rate": 0.024,
|
| 751 |
-
"loss": 5.83615255355835,
|
| 752 |
-
"step": 106
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 0.107,
|
| 756 |
-
"grad_norm": 0.41598859429359436,
|
| 757 |
-
"learning_rate": 0.024,
|
| 758 |
-
"loss": 5.805192470550537,
|
| 759 |
-
"step": 107
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.108,
|
| 763 |
-
"grad_norm": 0.44609910249710083,
|
| 764 |
-
"learning_rate": 0.024,
|
| 765 |
-
"loss": 5.782190799713135,
|
| 766 |
-
"step": 108
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.109,
|
| 770 |
-
"grad_norm": 0.4319760501384735,
|
| 771 |
-
"learning_rate": 0.024,
|
| 772 |
-
"loss": 5.757563591003418,
|
| 773 |
-
"step": 109
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 0.11,
|
| 777 |
-
"grad_norm": 0.37581533193588257,
|
| 778 |
-
"learning_rate": 0.024,
|
| 779 |
-
"loss": 5.757608890533447,
|
| 780 |
-
"step": 110
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 0.111,
|
| 784 |
-
"grad_norm": 0.3572186827659607,
|
| 785 |
-
"learning_rate": 0.024,
|
| 786 |
-
"loss": 5.740207195281982,
|
| 787 |
-
"step": 111
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 0.112,
|
| 791 |
-
"grad_norm": 0.3538333475589752,
|
| 792 |
-
"learning_rate": 0.024,
|
| 793 |
-
"loss": 5.714384078979492,
|
| 794 |
-
"step": 112
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"epoch": 0.113,
|
| 798 |
-
"grad_norm": 0.41134923696517944,
|
| 799 |
-
"learning_rate": 0.024,
|
| 800 |
-
"loss": 5.687575817108154,
|
| 801 |
-
"step": 113
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 0.114,
|
| 805 |
-
"grad_norm": 0.4088628888130188,
|
| 806 |
-
"learning_rate": 0.024,
|
| 807 |
-
"loss": 5.700284481048584,
|
| 808 |
-
"step": 114
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"epoch": 0.115,
|
| 812 |
-
"grad_norm": 0.38709452748298645,
|
| 813 |
-
"learning_rate": 0.024,
|
| 814 |
-
"loss": 5.702281951904297,
|
| 815 |
-
"step": 115
|
| 816 |
-
},
|
| 817 |
-
{
|
| 818 |
-
"epoch": 0.116,
|
| 819 |
-
"grad_norm": 0.4472334682941437,
|
| 820 |
-
"learning_rate": 0.024,
|
| 821 |
-
"loss": 5.692593574523926,
|
| 822 |
-
"step": 116
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 0.117,
|
| 826 |
-
"grad_norm": 0.439623087644577,
|
| 827 |
-
"learning_rate": 0.024,
|
| 828 |
-
"loss": 5.638166427612305,
|
| 829 |
-
"step": 117
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 0.118,
|
| 833 |
-
"grad_norm": 0.44986841082572937,
|
| 834 |
-
"learning_rate": 0.024,
|
| 835 |
-
"loss": 5.64177131652832,
|
| 836 |
-
"step": 118
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 0.119,
|
| 840 |
-
"grad_norm": 0.41504648327827454,
|
| 841 |
-
"learning_rate": 0.024,
|
| 842 |
-
"loss": 5.635551929473877,
|
| 843 |
-
"step": 119
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 0.12,
|
| 847 |
-
"grad_norm": 0.38233494758605957,
|
| 848 |
-
"learning_rate": 0.024,
|
| 849 |
-
"loss": 5.6383867263793945,
|
| 850 |
-
"step": 120
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"epoch": 0.121,
|
| 854 |
-
"grad_norm": 0.40537798404693604,
|
| 855 |
-
"learning_rate": 0.024,
|
| 856 |
-
"loss": 5.63343620300293,
|
| 857 |
-
"step": 121
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 0.122,
|
| 861 |
-
"grad_norm": 0.37236836552619934,
|
| 862 |
-
"learning_rate": 0.024,
|
| 863 |
-
"loss": 5.618805408477783,
|
| 864 |
-
"step": 122
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 0.123,
|
| 868 |
-
"grad_norm": 0.3938913643360138,
|
| 869 |
-
"learning_rate": 0.024,
|
| 870 |
-
"loss": 5.555117607116699,
|
| 871 |
-
"step": 123
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.124,
|
| 875 |
-
"grad_norm": 0.3891066610813141,
|
| 876 |
-
"learning_rate": 0.024,
|
| 877 |
-
"loss": 5.562790393829346,
|
| 878 |
-
"step": 124
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 0.125,
|
| 882 |
-
"grad_norm": 0.3771722614765167,
|
| 883 |
-
"learning_rate": 0.024,
|
| 884 |
-
"loss": 5.549802303314209,
|
| 885 |
-
"step": 125
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 0.126,
|
| 889 |
-
"grad_norm": 0.3803214430809021,
|
| 890 |
-
"learning_rate": 0.024,
|
| 891 |
-
"loss": 5.552084445953369,
|
| 892 |
-
"step": 126
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 0.127,
|
| 896 |
-
"grad_norm": 0.47765660285949707,
|
| 897 |
-
"learning_rate": 0.024,
|
| 898 |
-
"loss": 5.499300956726074,
|
| 899 |
-
"step": 127
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 0.128,
|
| 903 |
-
"grad_norm": 0.4789082109928131,
|
| 904 |
-
"learning_rate": 0.024,
|
| 905 |
-
"loss": 5.517937660217285,
|
| 906 |
-
"step": 128
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 0.129,
|
| 910 |
-
"grad_norm": 0.42897605895996094,
|
| 911 |
-
"learning_rate": 0.024,
|
| 912 |
-
"loss": 5.476377010345459,
|
| 913 |
-
"step": 129
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 0.13,
|
| 917 |
-
"grad_norm": 0.4472588896751404,
|
| 918 |
-
"learning_rate": 0.024,
|
| 919 |
-
"loss": 5.498976707458496,
|
| 920 |
-
"step": 130
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 0.131,
|
| 924 |
-
"grad_norm": 0.45371782779693604,
|
| 925 |
-
"learning_rate": 0.024,
|
| 926 |
-
"loss": 5.4586286544799805,
|
| 927 |
-
"step": 131
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.132,
|
| 931 |
-
"grad_norm": 0.3807578384876251,
|
| 932 |
-
"learning_rate": 0.024,
|
| 933 |
-
"loss": 5.480893611907959,
|
| 934 |
-
"step": 132
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 0.133,
|
| 938 |
-
"grad_norm": 0.39887917041778564,
|
| 939 |
-
"learning_rate": 0.024,
|
| 940 |
-
"loss": 5.41438627243042,
|
| 941 |
-
"step": 133
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 0.134,
|
| 945 |
-
"grad_norm": 0.4278720021247864,
|
| 946 |
-
"learning_rate": 0.024,
|
| 947 |
-
"loss": 5.456109523773193,
|
| 948 |
-
"step": 134
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 0.135,
|
| 952 |
-
"grad_norm": 0.3651004135608673,
|
| 953 |
-
"learning_rate": 0.024,
|
| 954 |
-
"loss": 5.454498767852783,
|
| 955 |
-
"step": 135
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 0.136,
|
| 959 |
-
"grad_norm": 0.3450686037540436,
|
| 960 |
-
"learning_rate": 0.024,
|
| 961 |
-
"loss": 5.410224437713623,
|
| 962 |
-
"step": 136
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 0.137,
|
| 966 |
-
"grad_norm": 0.3696085810661316,
|
| 967 |
-
"learning_rate": 0.024,
|
| 968 |
-
"loss": 5.37379789352417,
|
| 969 |
-
"step": 137
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 0.138,
|
| 973 |
-
"grad_norm": 0.3826468288898468,
|
| 974 |
-
"learning_rate": 0.024,
|
| 975 |
-
"loss": 5.395396709442139,
|
| 976 |
-
"step": 138
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 0.139,
|
| 980 |
-
"grad_norm": 0.33191096782684326,
|
| 981 |
-
"learning_rate": 0.024,
|
| 982 |
-
"loss": 5.357860565185547,
|
| 983 |
-
"step": 139
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 0.14,
|
| 987 |
-
"grad_norm": 0.3096979856491089,
|
| 988 |
-
"learning_rate": 0.024,
|
| 989 |
-
"loss": 5.343928337097168,
|
| 990 |
-
"step": 140
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 0.141,
|
| 994 |
-
"grad_norm": 0.32421982288360596,
|
| 995 |
-
"learning_rate": 0.024,
|
| 996 |
-
"loss": 5.348172187805176,
|
| 997 |
-
"step": 141
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 0.142,
|
| 1001 |
-
"grad_norm": 0.3381337523460388,
|
| 1002 |
-
"learning_rate": 0.024,
|
| 1003 |
-
"loss": 5.329306125640869,
|
| 1004 |
-
"step": 142
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 0.143,
|
| 1008 |
-
"grad_norm": 0.3032286465167999,
|
| 1009 |
-
"learning_rate": 0.024,
|
| 1010 |
-
"loss": 5.370608806610107,
|
| 1011 |
-
"step": 143
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 0.144,
|
| 1015 |
-
"grad_norm": 0.2875230312347412,
|
| 1016 |
-
"learning_rate": 0.024,
|
| 1017 |
-
"loss": 5.315526962280273,
|
| 1018 |
-
"step": 144
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 0.145,
|
| 1022 |
-
"grad_norm": 0.32873865962028503,
|
| 1023 |
-
"learning_rate": 0.024,
|
| 1024 |
-
"loss": 5.271420955657959,
|
| 1025 |
-
"step": 145
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 0.146,
|
| 1029 |
-
"grad_norm": 0.32419320940971375,
|
| 1030 |
-
"learning_rate": 0.024,
|
| 1031 |
-
"loss": 5.3021650314331055,
|
| 1032 |
-
"step": 146
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 0.147,
|
| 1036 |
-
"grad_norm": 0.30146142840385437,
|
| 1037 |
-
"learning_rate": 0.024,
|
| 1038 |
-
"loss": 5.267069339752197,
|
| 1039 |
-
"step": 147
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 0.148,
|
| 1043 |
-
"grad_norm": 0.3197166323661804,
|
| 1044 |
-
"learning_rate": 0.024,
|
| 1045 |
-
"loss": 5.262673377990723,
|
| 1046 |
-
"step": 148
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 0.149,
|
| 1050 |
-
"grad_norm": 0.35318228602409363,
|
| 1051 |
-
"learning_rate": 0.024,
|
| 1052 |
-
"loss": 5.247670650482178,
|
| 1053 |
-
"step": 149
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 0.15,
|
| 1057 |
-
"grad_norm": 0.34273043274879456,
|
| 1058 |
-
"learning_rate": 0.024,
|
| 1059 |
-
"loss": 5.256842136383057,
|
| 1060 |
-
"step": 150
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 0.151,
|
| 1064 |
-
"grad_norm": 0.32589635252952576,
|
| 1065 |
-
"learning_rate": 0.024,
|
| 1066 |
-
"loss": 5.201720714569092,
|
| 1067 |
-
"step": 151
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 0.152,
|
| 1071 |
-
"grad_norm": 0.3270612061023712,
|
| 1072 |
-
"learning_rate": 0.024,
|
| 1073 |
-
"loss": 5.230245113372803,
|
| 1074 |
-
"step": 152
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 0.153,
|
| 1078 |
-
"grad_norm": 0.3230215907096863,
|
| 1079 |
-
"learning_rate": 0.024,
|
| 1080 |
-
"loss": 5.215639114379883,
|
| 1081 |
-
"step": 153
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 0.154,
|
| 1085 |
-
"grad_norm": 0.35351404547691345,
|
| 1086 |
-
"learning_rate": 0.024,
|
| 1087 |
-
"loss": 5.178455829620361,
|
| 1088 |
-
"step": 154
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.155,
|
| 1092 |
-
"grad_norm": 0.3672751486301422,
|
| 1093 |
-
"learning_rate": 0.024,
|
| 1094 |
-
"loss": 5.208954334259033,
|
| 1095 |
-
"step": 155
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 0.156,
|
| 1099 |
-
"grad_norm": 0.3283761739730835,
|
| 1100 |
-
"learning_rate": 0.024,
|
| 1101 |
-
"loss": 5.2582526206970215,
|
| 1102 |
-
"step": 156
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 0.157,
|
| 1106 |
-
"grad_norm": 0.3091425895690918,
|
| 1107 |
-
"learning_rate": 0.024,
|
| 1108 |
-
"loss": 5.183006286621094,
|
| 1109 |
-
"step": 157
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 0.158,
|
| 1113 |
-
"grad_norm": 0.3499504029750824,
|
| 1114 |
-
"learning_rate": 0.024,
|
| 1115 |
-
"loss": 5.178630352020264,
|
| 1116 |
-
"step": 158
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 0.159,
|
| 1120 |
-
"grad_norm": 0.37615883350372314,
|
| 1121 |
-
"learning_rate": 0.024,
|
| 1122 |
-
"loss": 5.164978981018066,
|
| 1123 |
-
"step": 159
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 0.16,
|
| 1127 |
-
"grad_norm": 0.3276405334472656,
|
| 1128 |
-
"learning_rate": 0.024,
|
| 1129 |
-
"loss": 5.173659801483154,
|
| 1130 |
-
"step": 160
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.161,
|
| 1134 |
-
"grad_norm": 0.27133050560951233,
|
| 1135 |
-
"learning_rate": 0.024,
|
| 1136 |
-
"loss": 5.127079486846924,
|
| 1137 |
-
"step": 161
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 0.162,
|
| 1141 |
-
"grad_norm": 0.2708619236946106,
|
| 1142 |
-
"learning_rate": 0.024,
|
| 1143 |
-
"loss": 5.136236190795898,
|
| 1144 |
-
"step": 162
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 0.163,
|
| 1148 |
-
"grad_norm": 0.2757193446159363,
|
| 1149 |
-
"learning_rate": 0.024,
|
| 1150 |
-
"loss": 5.095175743103027,
|
| 1151 |
-
"step": 163
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 0.164,
|
| 1155 |
-
"grad_norm": 0.315839022397995,
|
| 1156 |
-
"learning_rate": 0.024,
|
| 1157 |
-
"loss": 5.1019392013549805,
|
| 1158 |
-
"step": 164
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 0.165,
|
| 1162 |
-
"grad_norm": 0.3292011320590973,
|
| 1163 |
-
"learning_rate": 0.024,
|
| 1164 |
-
"loss": 5.118902683258057,
|
| 1165 |
-
"step": 165
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 0.166,
|
| 1169 |
-
"grad_norm": 0.29490938782691956,
|
| 1170 |
-
"learning_rate": 0.024,
|
| 1171 |
-
"loss": 5.095141887664795,
|
| 1172 |
-
"step": 166
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.167,
|
| 1176 |
-
"grad_norm": 0.27785980701446533,
|
| 1177 |
-
"learning_rate": 0.024,
|
| 1178 |
-
"loss": 5.086520195007324,
|
| 1179 |
-
"step": 167
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.168,
|
| 1183 |
-
"grad_norm": 0.29426130652427673,
|
| 1184 |
-
"learning_rate": 0.024,
|
| 1185 |
-
"loss": 5.080230712890625,
|
| 1186 |
-
"step": 168
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 0.169,
|
| 1190 |
-
"grad_norm": 0.3200998902320862,
|
| 1191 |
-
"learning_rate": 0.024,
|
| 1192 |
-
"loss": 5.094149112701416,
|
| 1193 |
-
"step": 169
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 0.17,
|
| 1197 |
-
"grad_norm": 0.30668869614601135,
|
| 1198 |
-
"learning_rate": 0.024,
|
| 1199 |
-
"loss": 5.042888641357422,
|
| 1200 |
-
"step": 170
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 0.171,
|
| 1204 |
-
"grad_norm": 0.30815911293029785,
|
| 1205 |
-
"learning_rate": 0.024,
|
| 1206 |
-
"loss": 5.0581512451171875,
|
| 1207 |
-
"step": 171
|
| 1208 |
-
},
|
| 1209 |
-
{
|
| 1210 |
-
"epoch": 0.172,
|
| 1211 |
-
"grad_norm": 0.33038681745529175,
|
| 1212 |
-
"learning_rate": 0.024,
|
| 1213 |
-
"loss": 5.0244245529174805,
|
| 1214 |
-
"step": 172
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 0.173,
|
| 1218 |
-
"grad_norm": 0.32660111784935,
|
| 1219 |
-
"learning_rate": 0.024,
|
| 1220 |
-
"loss": 4.989992618560791,
|
| 1221 |
-
"step": 173
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 0.174,
|
| 1225 |
-
"grad_norm": 0.2998931109905243,
|
| 1226 |
-
"learning_rate": 0.024,
|
| 1227 |
-
"loss": 5.049340724945068,
|
| 1228 |
-
"step": 174
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 0.175,
|
| 1232 |
-
"grad_norm": 0.2632315456867218,
|
| 1233 |
-
"learning_rate": 0.024,
|
| 1234 |
-
"loss": 5.025455951690674,
|
| 1235 |
-
"step": 175
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 0.176,
|
| 1239 |
-
"grad_norm": 0.26309406757354736,
|
| 1240 |
-
"learning_rate": 0.024,
|
| 1241 |
-
"loss": 5.0016961097717285,
|
| 1242 |
-
"step": 176
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 0.177,
|
| 1246 |
-
"grad_norm": 0.2989042103290558,
|
| 1247 |
-
"learning_rate": 0.024,
|
| 1248 |
-
"loss": 5.0056610107421875,
|
| 1249 |
-
"step": 177
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 0.178,
|
| 1253 |
-
"grad_norm": 0.30713799595832825,
|
| 1254 |
-
"learning_rate": 0.024,
|
| 1255 |
-
"loss": 5.0450873374938965,
|
| 1256 |
-
"step": 178
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 0.179,
|
| 1260 |
-
"grad_norm": 0.3065016269683838,
|
| 1261 |
-
"learning_rate": 0.024,
|
| 1262 |
-
"loss": 4.956238746643066,
|
| 1263 |
-
"step": 179
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 0.18,
|
| 1267 |
-
"grad_norm": 0.319240927696228,
|
| 1268 |
-
"learning_rate": 0.024,
|
| 1269 |
-
"loss": 4.951620578765869,
|
| 1270 |
-
"step": 180
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"epoch": 0.181,
|
| 1274 |
-
"grad_norm": 0.38088586926460266,
|
| 1275 |
-
"learning_rate": 0.024,
|
| 1276 |
-
"loss": 4.973544597625732,
|
| 1277 |
-
"step": 181
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 0.182,
|
| 1281 |
-
"grad_norm": 0.34409934282302856,
|
| 1282 |
-
"learning_rate": 0.024,
|
| 1283 |
-
"loss": 4.934852600097656,
|
| 1284 |
-
"step": 182
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 0.183,
|
| 1288 |
-
"grad_norm": 0.28856003284454346,
|
| 1289 |
-
"learning_rate": 0.024,
|
| 1290 |
-
"loss": 4.957813262939453,
|
| 1291 |
-
"step": 183
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 0.184,
|
| 1295 |
-
"grad_norm": 0.2784363627433777,
|
| 1296 |
-
"learning_rate": 0.024,
|
| 1297 |
-
"loss": 4.947636127471924,
|
| 1298 |
-
"step": 184
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 0.185,
|
| 1302 |
-
"grad_norm": 0.2657927870750427,
|
| 1303 |
-
"learning_rate": 0.024,
|
| 1304 |
-
"loss": 4.917152404785156,
|
| 1305 |
-
"step": 185
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 0.186,
|
| 1309 |
-
"grad_norm": 0.2621709406375885,
|
| 1310 |
-
"learning_rate": 0.024,
|
| 1311 |
-
"loss": 4.872308254241943,
|
| 1312 |
-
"step": 186
|
| 1313 |
-
},
|
| 1314 |
-
{
|
| 1315 |
-
"epoch": 0.187,
|
| 1316 |
-
"grad_norm": 0.24439243972301483,
|
| 1317 |
-
"learning_rate": 0.024,
|
| 1318 |
-
"loss": 4.862739562988281,
|
| 1319 |
-
"step": 187
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 0.188,
|
| 1323 |
-
"grad_norm": 0.25019505620002747,
|
| 1324 |
-
"learning_rate": 0.024,
|
| 1325 |
-
"loss": 4.921043395996094,
|
| 1326 |
-
"step": 188
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"epoch": 0.189,
|
| 1330 |
-
"grad_norm": 0.2653026878833771,
|
| 1331 |
-
"learning_rate": 0.024,
|
| 1332 |
-
"loss": 4.9358954429626465,
|
| 1333 |
-
"step": 189
|
| 1334 |
-
},
|
| 1335 |
-
{
|
| 1336 |
-
"epoch": 0.19,
|
| 1337 |
-
"grad_norm": 0.3002541661262512,
|
| 1338 |
-
"learning_rate": 0.024,
|
| 1339 |
-
"loss": 4.93791389465332,
|
| 1340 |
-
"step": 190
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 0.191,
|
| 1344 |
-
"grad_norm": 0.30416810512542725,
|
| 1345 |
-
"learning_rate": 0.024,
|
| 1346 |
-
"loss": 4.85439920425415,
|
| 1347 |
-
"step": 191
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 0.192,
|
| 1351 |
-
"grad_norm": 0.3138468265533447,
|
| 1352 |
-
"learning_rate": 0.024,
|
| 1353 |
-
"loss": 4.843986511230469,
|
| 1354 |
-
"step": 192
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"epoch": 0.193,
|
| 1358 |
-
"grad_norm": 0.2839624285697937,
|
| 1359 |
-
"learning_rate": 0.024,
|
| 1360 |
-
"loss": 4.890965461730957,
|
| 1361 |
-
"step": 193
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 0.194,
|
| 1365 |
-
"grad_norm": 0.2886679470539093,
|
| 1366 |
-
"learning_rate": 0.024,
|
| 1367 |
-
"loss": 4.882617950439453,
|
| 1368 |
-
"step": 194
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 0.195,
|
| 1372 |
-
"grad_norm": 0.3105316162109375,
|
| 1373 |
-
"learning_rate": 0.024,
|
| 1374 |
-
"loss": 4.840507984161377,
|
| 1375 |
-
"step": 195
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 0.196,
|
| 1379 |
-
"grad_norm": 0.31287243962287903,
|
| 1380 |
-
"learning_rate": 0.024,
|
| 1381 |
-
"loss": 4.8086161613464355,
|
| 1382 |
-
"step": 196
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 0.197,
|
| 1386 |
-
"grad_norm": 0.30985069274902344,
|
| 1387 |
-
"learning_rate": 0.024,
|
| 1388 |
-
"loss": 4.817875862121582,
|
| 1389 |
-
"step": 197
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 0.198,
|
| 1393 |
-
"grad_norm": 0.2861211597919464,
|
| 1394 |
-
"learning_rate": 0.024,
|
| 1395 |
-
"loss": 4.773270606994629,
|
| 1396 |
-
"step": 198
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 0.199,
|
| 1400 |
-
"grad_norm": 0.2677602171897888,
|
| 1401 |
-
"learning_rate": 0.024,
|
| 1402 |
-
"loss": 4.834056377410889,
|
| 1403 |
-
"step": 199
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 0.2,
|
| 1407 |
-
"grad_norm": 0.26574546098709106,
|
| 1408 |
-
"learning_rate": 0.024,
|
| 1409 |
-
"loss": 4.789336204528809,
|
| 1410 |
-
"step": 200
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 0.2,
|
| 1414 |
-
"eval_loss": 4.760849952697754,
|
| 1415 |
-
"eval_runtime": 72.6568,
|
| 1416 |
-
"eval_samples_per_second": 14.094,
|
| 1417 |
-
"eval_steps_per_second": 3.523,
|
| 1418 |
-
"step": 200
|
| 1419 |
-
},
|
| 1420 |
-
{
|
| 1421 |
-
"epoch": 0.201,
|
| 1422 |
-
"grad_norm": 0.273313969373703,
|
| 1423 |
-
"learning_rate": 0.024,
|
| 1424 |
-
"loss": 4.819768905639648,
|
| 1425 |
-
"step": 201
|
| 1426 |
-
},
|
| 1427 |
-
{
|
| 1428 |
-
"epoch": 0.202,
|
| 1429 |
-
"grad_norm": 0.2709055244922638,
|
| 1430 |
-
"learning_rate": 0.024,
|
| 1431 |
-
"loss": 4.8297600746154785,
|
| 1432 |
-
"step": 202
|
| 1433 |
-
},
|
| 1434 |
-
{
|
| 1435 |
-
"epoch": 0.203,
|
| 1436 |
-
"grad_norm": 0.27067214250564575,
|
| 1437 |
-
"learning_rate": 0.024,
|
| 1438 |
-
"loss": 4.81519889831543,
|
| 1439 |
-
"step": 203
|
| 1440 |
-
},
|
| 1441 |
-
{
|
| 1442 |
-
"epoch": 0.204,
|
| 1443 |
-
"grad_norm": 0.25037306547164917,
|
| 1444 |
-
"learning_rate": 0.024,
|
| 1445 |
-
"loss": 4.762233734130859,
|
| 1446 |
-
"step": 204
|
| 1447 |
-
},
|
| 1448 |
-
{
|
| 1449 |
-
"epoch": 0.205,
|
| 1450 |
-
"grad_norm": 0.23847787082195282,
|
| 1451 |
-
"learning_rate": 0.024,
|
| 1452 |
-
"loss": 4.744489669799805,
|
| 1453 |
-
"step": 205
|
| 1454 |
-
},
|
| 1455 |
-
{
|
| 1456 |
-
"epoch": 0.206,
|
| 1457 |
-
"grad_norm": 0.2536039352416992,
|
| 1458 |
-
"learning_rate": 0.024,
|
| 1459 |
-
"loss": 4.737757205963135,
|
| 1460 |
-
"step": 206
|
| 1461 |
-
},
|
| 1462 |
-
{
|
| 1463 |
-
"epoch": 0.207,
|
| 1464 |
-
"grad_norm": 0.27073943614959717,
|
| 1465 |
-
"learning_rate": 0.024,
|
| 1466 |
-
"loss": 4.729128360748291,
|
| 1467 |
-
"step": 207
|
| 1468 |
-
},
|
| 1469 |
-
{
|
| 1470 |
-
"epoch": 0.208,
|
| 1471 |
-
"grad_norm": 0.2774869203567505,
|
| 1472 |
-
"learning_rate": 0.024,
|
| 1473 |
-
"loss": 4.695976734161377,
|
| 1474 |
-
"step": 208
|
| 1475 |
-
},
|
| 1476 |
-
{
|
| 1477 |
-
"epoch": 0.209,
|
| 1478 |
-
"grad_norm": 0.3016758859157562,
|
| 1479 |
-
"learning_rate": 0.024,
|
| 1480 |
-
"loss": 4.732336521148682,
|
| 1481 |
-
"step": 209
|
| 1482 |
-
},
|
| 1483 |
-
{
|
| 1484 |
-
"epoch": 0.21,
|
| 1485 |
-
"grad_norm": 0.31534770131111145,
|
| 1486 |
-
"learning_rate": 0.024,
|
| 1487 |
-
"loss": 4.722041130065918,
|
| 1488 |
-
"step": 210
|
| 1489 |
-
},
|
| 1490 |
-
{
|
| 1491 |
-
"epoch": 0.211,
|
| 1492 |
-
"grad_norm": 0.28376662731170654,
|
| 1493 |
-
"learning_rate": 0.024,
|
| 1494 |
-
"loss": 4.731420993804932,
|
| 1495 |
-
"step": 211
|
| 1496 |
-
},
|
| 1497 |
-
{
|
| 1498 |
-
"epoch": 0.212,
|
| 1499 |
-
"grad_norm": 0.25542202591896057,
|
| 1500 |
-
"learning_rate": 0.024,
|
| 1501 |
-
"loss": 4.707083225250244,
|
| 1502 |
-
"step": 212
|
| 1503 |
-
},
|
| 1504 |
-
{
|
| 1505 |
-
"epoch": 0.213,
|
| 1506 |
-
"grad_norm": 0.25231537222862244,
|
| 1507 |
-
"learning_rate": 0.024,
|
| 1508 |
-
"loss": 4.6861467361450195,
|
| 1509 |
-
"step": 213
|
| 1510 |
-
},
|
| 1511 |
-
{
|
| 1512 |
-
"epoch": 0.214,
|
| 1513 |
-
"grad_norm": 0.24746422469615936,
|
| 1514 |
-
"learning_rate": 0.024,
|
| 1515 |
-
"loss": 4.684498310089111,
|
| 1516 |
-
"step": 214
|
| 1517 |
-
},
|
| 1518 |
-
{
|
| 1519 |
-
"epoch": 0.215,
|
| 1520 |
-
"grad_norm": 0.26093682646751404,
|
| 1521 |
-
"learning_rate": 0.024,
|
| 1522 |
-
"loss": 4.689898490905762,
|
| 1523 |
-
"step": 215
|
| 1524 |
-
},
|
| 1525 |
-
{
|
| 1526 |
-
"epoch": 0.216,
|
| 1527 |
-
"grad_norm": 0.2589157223701477,
|
| 1528 |
-
"learning_rate": 0.024,
|
| 1529 |
-
"loss": 4.704195499420166,
|
| 1530 |
-
"step": 216
|
| 1531 |
-
},
|
| 1532 |
-
{
|
| 1533 |
-
"epoch": 0.217,
|
| 1534 |
-
"grad_norm": 0.30941376090049744,
|
| 1535 |
-
"learning_rate": 0.024,
|
| 1536 |
-
"loss": 4.737393856048584,
|
| 1537 |
-
"step": 217
|
| 1538 |
-
},
|
| 1539 |
-
{
|
| 1540 |
-
"epoch": 0.218,
|
| 1541 |
-
"grad_norm": 0.31288063526153564,
|
| 1542 |
-
"learning_rate": 0.024,
|
| 1543 |
-
"loss": 4.69362735748291,
|
| 1544 |
-
"step": 218
|
| 1545 |
-
},
|
| 1546 |
-
{
|
| 1547 |
-
"epoch": 0.219,
|
| 1548 |
-
"grad_norm": 0.2729438841342926,
|
| 1549 |
-
"learning_rate": 0.024,
|
| 1550 |
-
"loss": 4.676018714904785,
|
| 1551 |
-
"step": 219
|
| 1552 |
-
},
|
| 1553 |
-
{
|
| 1554 |
-
"epoch": 0.22,
|
| 1555 |
-
"grad_norm": 0.25406497716903687,
|
| 1556 |
-
"learning_rate": 0.024,
|
| 1557 |
-
"loss": 4.667109966278076,
|
| 1558 |
-
"step": 220
|
| 1559 |
-
},
|
| 1560 |
-
{
|
| 1561 |
-
"epoch": 0.221,
|
| 1562 |
-
"grad_norm": 0.2425123155117035,
|
| 1563 |
-
"learning_rate": 0.024,
|
| 1564 |
-
"loss": 4.601018905639648,
|
| 1565 |
-
"step": 221
|
| 1566 |
-
},
|
| 1567 |
-
{
|
| 1568 |
-
"epoch": 0.222,
|
| 1569 |
-
"grad_norm": 0.2454339563846588,
|
| 1570 |
-
"learning_rate": 0.024,
|
| 1571 |
-
"loss": 4.58155632019043,
|
| 1572 |
-
"step": 222
|
| 1573 |
-
},
|
| 1574 |
-
{
|
| 1575 |
-
"epoch": 0.223,
|
| 1576 |
-
"grad_norm": 0.2466762810945511,
|
| 1577 |
-
"learning_rate": 0.024,
|
| 1578 |
-
"loss": 4.6468424797058105,
|
| 1579 |
-
"step": 223
|
| 1580 |
-
},
|
| 1581 |
-
{
|
| 1582 |
-
"epoch": 0.224,
|
| 1583 |
-
"grad_norm": 0.2499387413263321,
|
| 1584 |
-
"learning_rate": 0.024,
|
| 1585 |
-
"loss": 4.654460430145264,
|
| 1586 |
-
"step": 224
|
| 1587 |
-
},
|
| 1588 |
-
{
|
| 1589 |
-
"epoch": 0.225,
|
| 1590 |
-
"grad_norm": 0.2456265240907669,
|
| 1591 |
-
"learning_rate": 0.024,
|
| 1592 |
-
"loss": 4.664284706115723,
|
| 1593 |
-
"step": 225
|
| 1594 |
-
},
|
| 1595 |
-
{
|
| 1596 |
-
"epoch": 0.226,
|
| 1597 |
-
"grad_norm": 0.23937228322029114,
|
| 1598 |
-
"learning_rate": 0.024,
|
| 1599 |
-
"loss": 4.641874313354492,
|
| 1600 |
-
"step": 226
|
| 1601 |
-
},
|
| 1602 |
-
{
|
| 1603 |
-
"epoch": 0.227,
|
| 1604 |
-
"grad_norm": 0.24739213287830353,
|
| 1605 |
-
"learning_rate": 0.024,
|
| 1606 |
-
"loss": 4.60195255279541,
|
| 1607 |
-
"step": 227
|
| 1608 |
-
},
|
| 1609 |
-
{
|
| 1610 |
-
"epoch": 0.228,
|
| 1611 |
-
"grad_norm": 0.24062694609165192,
|
| 1612 |
-
"learning_rate": 0.024,
|
| 1613 |
-
"loss": 4.602538108825684,
|
| 1614 |
-
"step": 228
|
| 1615 |
-
},
|
| 1616 |
-
{
|
| 1617 |
-
"epoch": 0.229,
|
| 1618 |
-
"grad_norm": 0.23135443031787872,
|
| 1619 |
-
"learning_rate": 0.024,
|
| 1620 |
-
"loss": 4.605818748474121,
|
| 1621 |
-
"step": 229
|
| 1622 |
-
},
|
| 1623 |
-
{
|
| 1624 |
-
"epoch": 0.23,
|
| 1625 |
-
"grad_norm": 0.25950688123703003,
|
| 1626 |
-
"learning_rate": 0.024,
|
| 1627 |
-
"loss": 4.559548377990723,
|
| 1628 |
-
"step": 230
|
| 1629 |
-
},
|
| 1630 |
-
{
|
| 1631 |
-
"epoch": 0.231,
|
| 1632 |
-
"grad_norm": 0.3428628146648407,
|
| 1633 |
-
"learning_rate": 0.024,
|
| 1634 |
-
"loss": 4.630019664764404,
|
| 1635 |
-
"step": 231
|
| 1636 |
-
},
|
| 1637 |
-
{
|
| 1638 |
-
"epoch": 0.232,
|
| 1639 |
-
"grad_norm": 0.3382345736026764,
|
| 1640 |
-
"learning_rate": 0.024,
|
| 1641 |
-
"loss": 4.561378002166748,
|
| 1642 |
-
"step": 232
|
| 1643 |
-
},
|
| 1644 |
-
{
|
| 1645 |
-
"epoch": 0.233,
|
| 1646 |
-
"grad_norm": 0.32074934244155884,
|
| 1647 |
-
"learning_rate": 0.024,
|
| 1648 |
-
"loss": 4.608393669128418,
|
| 1649 |
-
"step": 233
|
| 1650 |
-
},
|
| 1651 |
-
{
|
| 1652 |
-
"epoch": 0.234,
|
| 1653 |
-
"grad_norm": 0.31480836868286133,
|
| 1654 |
-
"learning_rate": 0.024,
|
| 1655 |
-
"loss": 4.584751605987549,
|
| 1656 |
-
"step": 234
|
| 1657 |
-
},
|
| 1658 |
-
{
|
| 1659 |
-
"epoch": 0.235,
|
| 1660 |
-
"grad_norm": 0.2775425910949707,
|
| 1661 |
-
"learning_rate": 0.024,
|
| 1662 |
-
"loss": 4.579282283782959,
|
| 1663 |
-
"step": 235
|
| 1664 |
-
},
|
| 1665 |
-
{
|
| 1666 |
-
"epoch": 0.236,
|
| 1667 |
-
"grad_norm": 0.2386392503976822,
|
| 1668 |
-
"learning_rate": 0.024,
|
| 1669 |
-
"loss": 4.549663543701172,
|
| 1670 |
-
"step": 236
|
| 1671 |
-
},
|
| 1672 |
-
{
|
| 1673 |
-
"epoch": 0.237,
|
| 1674 |
-
"grad_norm": 0.2256796509027481,
|
| 1675 |
-
"learning_rate": 0.024,
|
| 1676 |
-
"loss": 4.584214687347412,
|
| 1677 |
-
"step": 237
|
| 1678 |
-
},
|
| 1679 |
-
{
|
| 1680 |
-
"epoch": 0.238,
|
| 1681 |
-
"grad_norm": 0.22827120125293732,
|
| 1682 |
-
"learning_rate": 0.024,
|
| 1683 |
-
"loss": 4.4978485107421875,
|
| 1684 |
-
"step": 238
|
| 1685 |
-
},
|
| 1686 |
-
{
|
| 1687 |
-
"epoch": 0.239,
|
| 1688 |
-
"grad_norm": 0.21113577485084534,
|
| 1689 |
-
"learning_rate": 0.024,
|
| 1690 |
-
"loss": 4.586427211761475,
|
| 1691 |
-
"step": 239
|
| 1692 |
-
},
|
| 1693 |
-
{
|
| 1694 |
-
"epoch": 0.24,
|
| 1695 |
-
"grad_norm": 0.2041008472442627,
|
| 1696 |
-
"learning_rate": 0.024,
|
| 1697 |
-
"loss": 4.53650426864624,
|
| 1698 |
-
"step": 240
|
| 1699 |
-
},
|
| 1700 |
-
{
|
| 1701 |
-
"epoch": 0.241,
|
| 1702 |
-
"grad_norm": 0.21871869266033173,
|
| 1703 |
-
"learning_rate": 0.024,
|
| 1704 |
-
"loss": 4.534363746643066,
|
| 1705 |
-
"step": 241
|
| 1706 |
-
},
|
| 1707 |
-
{
|
| 1708 |
-
"epoch": 0.242,
|
| 1709 |
-
"grad_norm": 0.22946219146251678,
|
| 1710 |
-
"learning_rate": 0.024,
|
| 1711 |
-
"loss": 4.509822368621826,
|
| 1712 |
-
"step": 242
|
| 1713 |
-
},
|
| 1714 |
-
{
|
| 1715 |
-
"epoch": 0.243,
|
| 1716 |
-
"grad_norm": 0.23029088973999023,
|
| 1717 |
-
"learning_rate": 0.024,
|
| 1718 |
-
"loss": 4.540820121765137,
|
| 1719 |
-
"step": 243
|
| 1720 |
-
},
|
| 1721 |
-
{
|
| 1722 |
-
"epoch": 0.244,
|
| 1723 |
-
"grad_norm": 0.22023244202136993,
|
| 1724 |
-
"learning_rate": 0.024,
|
| 1725 |
-
"loss": 4.545165538787842,
|
| 1726 |
-
"step": 244
|
| 1727 |
-
},
|
| 1728 |
-
{
|
| 1729 |
-
"epoch": 0.245,
|
| 1730 |
-
"grad_norm": 0.23225648701190948,
|
| 1731 |
-
"learning_rate": 0.024,
|
| 1732 |
-
"loss": 4.544320106506348,
|
| 1733 |
-
"step": 245
|
| 1734 |
-
},
|
| 1735 |
-
{
|
| 1736 |
-
"epoch": 0.246,
|
| 1737 |
-
"grad_norm": 0.24750667810440063,
|
| 1738 |
-
"learning_rate": 0.024,
|
| 1739 |
-
"loss": 4.4896087646484375,
|
| 1740 |
-
"step": 246
|
| 1741 |
-
},
|
| 1742 |
-
{
|
| 1743 |
-
"epoch": 0.247,
|
| 1744 |
-
"grad_norm": 0.2301473468542099,
|
| 1745 |
-
"learning_rate": 0.024,
|
| 1746 |
-
"loss": 4.504151821136475,
|
| 1747 |
-
"step": 247
|
| 1748 |
-
},
|
| 1749 |
-
{
|
| 1750 |
-
"epoch": 0.248,
|
| 1751 |
-
"grad_norm": 0.22472187876701355,
|
| 1752 |
-
"learning_rate": 0.024,
|
| 1753 |
-
"loss": 4.462813854217529,
|
| 1754 |
-
"step": 248
|
| 1755 |
-
},
|
| 1756 |
-
{
|
| 1757 |
-
"epoch": 0.249,
|
| 1758 |
-
"grad_norm": 0.23330852389335632,
|
| 1759 |
-
"learning_rate": 0.024,
|
| 1760 |
-
"loss": 4.554251194000244,
|
| 1761 |
-
"step": 249
|
| 1762 |
-
},
|
| 1763 |
-
{
|
| 1764 |
-
"epoch": 0.25,
|
| 1765 |
-
"grad_norm": 0.2380797266960144,
|
| 1766 |
-
"learning_rate": 0.024,
|
| 1767 |
-
"loss": 4.470944404602051,
|
| 1768 |
-
"step": 250
|
| 1769 |
-
},
|
| 1770 |
-
{
|
| 1771 |
-
"epoch": 0.251,
|
| 1772 |
-
"grad_norm": 0.23439039289951324,
|
| 1773 |
-
"learning_rate": 0.024,
|
| 1774 |
-
"loss": 4.4591145515441895,
|
| 1775 |
-
"step": 251
|
| 1776 |
-
},
|
| 1777 |
-
{
|
| 1778 |
-
"epoch": 0.252,
|
| 1779 |
-
"grad_norm": 0.2279406636953354,
|
| 1780 |
-
"learning_rate": 0.024,
|
| 1781 |
-
"loss": 4.456786632537842,
|
| 1782 |
-
"step": 252
|
| 1783 |
-
},
|
| 1784 |
-
{
|
| 1785 |
-
"epoch": 0.253,
|
| 1786 |
-
"grad_norm": 0.25927624106407166,
|
| 1787 |
-
"learning_rate": 0.024,
|
| 1788 |
-
"loss": 4.4737324714660645,
|
| 1789 |
-
"step": 253
|
| 1790 |
-
},
|
| 1791 |
-
{
|
| 1792 |
-
"epoch": 0.254,
|
| 1793 |
-
"grad_norm": 0.2689415216445923,
|
| 1794 |
-
"learning_rate": 0.024,
|
| 1795 |
-
"loss": 4.493502616882324,
|
| 1796 |
-
"step": 254
|
| 1797 |
-
},
|
| 1798 |
-
{
|
| 1799 |
-
"epoch": 0.255,
|
| 1800 |
-
"grad_norm": 0.26093730330467224,
|
| 1801 |
-
"learning_rate": 0.024,
|
| 1802 |
-
"loss": 4.460394382476807,
|
| 1803 |
-
"step": 255
|
| 1804 |
-
},
|
| 1805 |
-
{
|
| 1806 |
-
"epoch": 0.256,
|
| 1807 |
-
"grad_norm": 0.26564300060272217,
|
| 1808 |
-
"learning_rate": 0.024,
|
| 1809 |
-
"loss": 4.513102054595947,
|
| 1810 |
-
"step": 256
|
| 1811 |
-
},
|
| 1812 |
-
{
|
| 1813 |
-
"epoch": 0.257,
|
| 1814 |
-
"grad_norm": 0.26798102259635925,
|
| 1815 |
-
"learning_rate": 0.024,
|
| 1816 |
-
"loss": 4.475131034851074,
|
| 1817 |
-
"step": 257
|
| 1818 |
-
},
|
| 1819 |
-
{
|
| 1820 |
-
"epoch": 0.258,
|
| 1821 |
-
"grad_norm": 0.24574024975299835,
|
| 1822 |
-
"learning_rate": 0.024,
|
| 1823 |
-
"loss": 4.493810176849365,
|
| 1824 |
-
"step": 258
|
| 1825 |
-
},
|
| 1826 |
-
{
|
| 1827 |
-
"epoch": 0.259,
|
| 1828 |
-
"grad_norm": 0.23991075158119202,
|
| 1829 |
-
"learning_rate": 0.024,
|
| 1830 |
-
"loss": 4.496291160583496,
|
| 1831 |
-
"step": 259
|
| 1832 |
-
},
|
| 1833 |
-
{
|
| 1834 |
-
"epoch": 0.26,
|
| 1835 |
-
"grad_norm": 0.2346344292163849,
|
| 1836 |
-
"learning_rate": 0.024,
|
| 1837 |
-
"loss": 4.438704013824463,
|
| 1838 |
-
"step": 260
|
| 1839 |
-
},
|
| 1840 |
-
{
|
| 1841 |
-
"epoch": 0.261,
|
| 1842 |
-
"grad_norm": 0.2712085545063019,
|
| 1843 |
-
"learning_rate": 0.024,
|
| 1844 |
-
"loss": 4.532087802886963,
|
| 1845 |
-
"step": 261
|
| 1846 |
-
},
|
| 1847 |
-
{
|
| 1848 |
-
"epoch": 0.262,
|
| 1849 |
-
"grad_norm": 0.2706504464149475,
|
| 1850 |
-
"learning_rate": 0.024,
|
| 1851 |
-
"loss": 4.472548961639404,
|
| 1852 |
-
"step": 262
|
| 1853 |
-
},
|
| 1854 |
-
{
|
| 1855 |
-
"epoch": 0.263,
|
| 1856 |
-
"grad_norm": 0.2372240275144577,
|
| 1857 |
-
"learning_rate": 0.024,
|
| 1858 |
-
"loss": 4.440901279449463,
|
| 1859 |
-
"step": 263
|
| 1860 |
-
},
|
| 1861 |
-
{
|
| 1862 |
-
"epoch": 0.264,
|
| 1863 |
-
"grad_norm": 0.21736063063144684,
|
| 1864 |
-
"learning_rate": 0.024,
|
| 1865 |
-
"loss": 4.435408592224121,
|
| 1866 |
-
"step": 264
|
| 1867 |
-
},
|
| 1868 |
-
{
|
| 1869 |
-
"epoch": 0.265,
|
| 1870 |
-
"grad_norm": 0.2175782173871994,
|
| 1871 |
-
"learning_rate": 0.024,
|
| 1872 |
-
"loss": 4.449498176574707,
|
| 1873 |
-
"step": 265
|
| 1874 |
-
},
|
| 1875 |
-
{
|
| 1876 |
-
"epoch": 0.266,
|
| 1877 |
-
"grad_norm": 0.26380375027656555,
|
| 1878 |
-
"learning_rate": 0.024,
|
| 1879 |
-
"loss": 4.433669567108154,
|
| 1880 |
-
"step": 266
|
| 1881 |
-
},
|
| 1882 |
-
{
|
| 1883 |
-
"epoch": 0.267,
|
| 1884 |
-
"grad_norm": 0.2703624963760376,
|
| 1885 |
-
"learning_rate": 0.024,
|
| 1886 |
-
"loss": 4.4369893074035645,
|
| 1887 |
-
"step": 267
|
| 1888 |
-
},
|
| 1889 |
-
{
|
| 1890 |
-
"epoch": 0.268,
|
| 1891 |
-
"grad_norm": 0.2735675573348999,
|
| 1892 |
-
"learning_rate": 0.024,
|
| 1893 |
-
"loss": 4.412627220153809,
|
| 1894 |
-
"step": 268
|
| 1895 |
-
},
|
| 1896 |
-
{
|
| 1897 |
-
"epoch": 0.269,
|
| 1898 |
-
"grad_norm": 0.25625836849212646,
|
| 1899 |
-
"learning_rate": 0.024,
|
| 1900 |
-
"loss": 4.4429707527160645,
|
| 1901 |
-
"step": 269
|
| 1902 |
-
},
|
| 1903 |
-
{
|
| 1904 |
-
"epoch": 0.27,
|
| 1905 |
-
"grad_norm": 0.2225719541311264,
|
| 1906 |
-
"learning_rate": 0.024,
|
| 1907 |
-
"loss": 4.443222522735596,
|
| 1908 |
-
"step": 270
|
| 1909 |
-
},
|
| 1910 |
-
{
|
| 1911 |
-
"epoch": 0.271,
|
| 1912 |
-
"grad_norm": 0.21454544365406036,
|
| 1913 |
-
"learning_rate": 0.024,
|
| 1914 |
-
"loss": 4.368286609649658,
|
| 1915 |
-
"step": 271
|
| 1916 |
-
},
|
| 1917 |
-
{
|
| 1918 |
-
"epoch": 0.272,
|
| 1919 |
-
"grad_norm": 0.2482341080904007,
|
| 1920 |
-
"learning_rate": 0.024,
|
| 1921 |
-
"loss": 4.426663875579834,
|
| 1922 |
-
"step": 272
|
| 1923 |
-
},
|
| 1924 |
-
{
|
| 1925 |
-
"epoch": 0.273,
|
| 1926 |
-
"grad_norm": 0.2593468129634857,
|
| 1927 |
-
"learning_rate": 0.024,
|
| 1928 |
-
"loss": 4.462672233581543,
|
| 1929 |
-
"step": 273
|
| 1930 |
-
},
|
| 1931 |
-
{
|
| 1932 |
-
"epoch": 0.274,
|
| 1933 |
-
"grad_norm": 0.24938873946666718,
|
| 1934 |
-
"learning_rate": 0.024,
|
| 1935 |
-
"loss": 4.447465896606445,
|
| 1936 |
-
"step": 274
|
| 1937 |
-
},
|
| 1938 |
-
{
|
| 1939 |
-
"epoch": 0.275,
|
| 1940 |
-
"grad_norm": 0.22095774114131927,
|
| 1941 |
-
"learning_rate": 0.024,
|
| 1942 |
-
"loss": 4.4406938552856445,
|
| 1943 |
-
"step": 275
|
| 1944 |
-
},
|
| 1945 |
-
{
|
| 1946 |
-
"epoch": 0.276,
|
| 1947 |
-
"grad_norm": 0.22666075825691223,
|
| 1948 |
-
"learning_rate": 0.024,
|
| 1949 |
-
"loss": 4.47224760055542,
|
| 1950 |
-
"step": 276
|
| 1951 |
-
},
|
| 1952 |
-
{
|
| 1953 |
-
"epoch": 0.277,
|
| 1954 |
-
"grad_norm": 0.2311767041683197,
|
| 1955 |
-
"learning_rate": 0.024,
|
| 1956 |
-
"loss": 4.3759870529174805,
|
| 1957 |
-
"step": 277
|
| 1958 |
-
},
|
| 1959 |
-
{
|
| 1960 |
-
"epoch": 0.278,
|
| 1961 |
-
"grad_norm": 0.20944620668888092,
|
| 1962 |
-
"learning_rate": 0.024,
|
| 1963 |
-
"loss": 4.391687870025635,
|
| 1964 |
-
"step": 278
|
| 1965 |
-
},
|
| 1966 |
-
{
|
| 1967 |
-
"epoch": 0.279,
|
| 1968 |
-
"grad_norm": 0.19864924252033234,
|
| 1969 |
-
"learning_rate": 0.024,
|
| 1970 |
-
"loss": 4.384091854095459,
|
| 1971 |
-
"step": 279
|
| 1972 |
-
},
|
| 1973 |
-
{
|
| 1974 |
-
"epoch": 0.28,
|
| 1975 |
-
"grad_norm": 0.20920057594776154,
|
| 1976 |
-
"learning_rate": 0.024,
|
| 1977 |
-
"loss": 4.4474406242370605,
|
| 1978 |
-
"step": 280
|
| 1979 |
-
},
|
| 1980 |
-
{
|
| 1981 |
-
"epoch": 0.281,
|
| 1982 |
-
"grad_norm": 0.24142234027385712,
|
| 1983 |
-
"learning_rate": 0.024,
|
| 1984 |
-
"loss": 4.398764610290527,
|
| 1985 |
-
"step": 281
|
| 1986 |
-
},
|
| 1987 |
-
{
|
| 1988 |
-
"epoch": 0.282,
|
| 1989 |
-
"grad_norm": 0.2301894724369049,
|
| 1990 |
-
"learning_rate": 0.024,
|
| 1991 |
-
"loss": 4.420306205749512,
|
| 1992 |
-
"step": 282
|
| 1993 |
-
},
|
| 1994 |
-
{
|
| 1995 |
-
"epoch": 0.283,
|
| 1996 |
-
"grad_norm": 0.2104264348745346,
|
| 1997 |
-
"learning_rate": 0.024,
|
| 1998 |
-
"loss": 4.390679836273193,
|
| 1999 |
-
"step": 283
|
| 2000 |
-
},
|
| 2001 |
-
{
|
| 2002 |
-
"epoch": 0.284,
|
| 2003 |
-
"grad_norm": 0.19881607592105865,
|
| 2004 |
-
"learning_rate": 0.024,
|
| 2005 |
-
"loss": 4.327366828918457,
|
| 2006 |
-
"step": 284
|
| 2007 |
-
},
|
| 2008 |
-
{
|
| 2009 |
-
"epoch": 0.285,
|
| 2010 |
-
"grad_norm": 0.21168561279773712,
|
| 2011 |
-
"learning_rate": 0.024,
|
| 2012 |
-
"loss": 4.418213367462158,
|
| 2013 |
-
"step": 285
|
| 2014 |
-
},
|
| 2015 |
-
{
|
| 2016 |
-
"epoch": 0.286,
|
| 2017 |
-
"grad_norm": 0.21123912930488586,
|
| 2018 |
-
"learning_rate": 0.024,
|
| 2019 |
-
"loss": 4.362180709838867,
|
| 2020 |
-
"step": 286
|
| 2021 |
-
},
|
| 2022 |
-
{
|
| 2023 |
-
"epoch": 0.287,
|
| 2024 |
-
"grad_norm": 0.22161200642585754,
|
| 2025 |
-
"learning_rate": 0.024,
|
| 2026 |
-
"loss": 4.3849592208862305,
|
| 2027 |
-
"step": 287
|
| 2028 |
-
},
|
| 2029 |
-
{
|
| 2030 |
-
"epoch": 0.288,
|
| 2031 |
-
"grad_norm": 0.22461362183094025,
|
| 2032 |
-
"learning_rate": 0.024,
|
| 2033 |
-
"loss": 4.396172046661377,
|
| 2034 |
-
"step": 288
|
| 2035 |
-
},
|
| 2036 |
-
{
|
| 2037 |
-
"epoch": 0.289,
|
| 2038 |
-
"grad_norm": 0.2322544902563095,
|
| 2039 |
-
"learning_rate": 0.024,
|
| 2040 |
-
"loss": 4.376404762268066,
|
| 2041 |
-
"step": 289
|
| 2042 |
-
},
|
| 2043 |
-
{
|
| 2044 |
-
"epoch": 0.29,
|
| 2045 |
-
"grad_norm": 0.2289140671491623,
|
| 2046 |
-
"learning_rate": 0.024,
|
| 2047 |
-
"loss": 4.334985733032227,
|
| 2048 |
-
"step": 290
|
| 2049 |
-
},
|
| 2050 |
-
{
|
| 2051 |
-
"epoch": 0.291,
|
| 2052 |
-
"grad_norm": 0.21877148747444153,
|
| 2053 |
-
"learning_rate": 0.024,
|
| 2054 |
-
"loss": 4.351469993591309,
|
| 2055 |
-
"step": 291
|
| 2056 |
-
},
|
| 2057 |
-
{
|
| 2058 |
-
"epoch": 0.292,
|
| 2059 |
-
"grad_norm": 0.224874347448349,
|
| 2060 |
-
"learning_rate": 0.024,
|
| 2061 |
-
"loss": 4.338824272155762,
|
| 2062 |
-
"step": 292
|
| 2063 |
-
},
|
| 2064 |
-
{
|
| 2065 |
-
"epoch": 0.293,
|
| 2066 |
-
"grad_norm": 0.2283969521522522,
|
| 2067 |
-
"learning_rate": 0.024,
|
| 2068 |
-
"loss": 4.321679592132568,
|
| 2069 |
-
"step": 293
|
| 2070 |
-
},
|
| 2071 |
-
{
|
| 2072 |
-
"epoch": 0.294,
|
| 2073 |
-
"grad_norm": 0.21595022082328796,
|
| 2074 |
-
"learning_rate": 0.024,
|
| 2075 |
-
"loss": 4.407853603363037,
|
| 2076 |
-
"step": 294
|
| 2077 |
-
},
|
| 2078 |
-
{
|
| 2079 |
-
"epoch": 0.295,
|
| 2080 |
-
"grad_norm": 0.21735002100467682,
|
| 2081 |
-
"learning_rate": 0.024,
|
| 2082 |
-
"loss": 4.343842029571533,
|
| 2083 |
-
"step": 295
|
| 2084 |
-
},
|
| 2085 |
-
{
|
| 2086 |
-
"epoch": 0.296,
|
| 2087 |
-
"grad_norm": 0.20310106873512268,
|
| 2088 |
-
"learning_rate": 0.024,
|
| 2089 |
-
"loss": 4.315837860107422,
|
| 2090 |
-
"step": 296
|
| 2091 |
-
},
|
| 2092 |
-
{
|
| 2093 |
-
"epoch": 0.297,
|
| 2094 |
-
"grad_norm": 0.18402153253555298,
|
| 2095 |
-
"learning_rate": 0.024,
|
| 2096 |
-
"loss": 4.3025946617126465,
|
| 2097 |
-
"step": 297
|
| 2098 |
-
},
|
| 2099 |
-
{
|
| 2100 |
-
"epoch": 0.298,
|
| 2101 |
-
"grad_norm": 0.21048487722873688,
|
| 2102 |
-
"learning_rate": 0.024,
|
| 2103 |
-
"loss": 4.315160274505615,
|
| 2104 |
-
"step": 298
|
| 2105 |
-
},
|
| 2106 |
-
{
|
| 2107 |
-
"epoch": 0.299,
|
| 2108 |
-
"grad_norm": 0.24944548308849335,
|
| 2109 |
-
"learning_rate": 0.024,
|
| 2110 |
-
"loss": 4.338711261749268,
|
| 2111 |
-
"step": 299
|
| 2112 |
-
},
|
| 2113 |
-
{
|
| 2114 |
-
"epoch": 0.3,
|
| 2115 |
-
"grad_norm": 0.2511313259601593,
|
| 2116 |
-
"learning_rate": 0.024,
|
| 2117 |
-
"loss": 4.31109094619751,
|
| 2118 |
-
"step": 300
|
| 2119 |
-
},
|
| 2120 |
-
{
|
| 2121 |
-
"epoch": 0.301,
|
| 2122 |
-
"grad_norm": 0.22077518701553345,
|
| 2123 |
-
"learning_rate": 0.024,
|
| 2124 |
-
"loss": 4.319232940673828,
|
| 2125 |
-
"step": 301
|
| 2126 |
-
},
|
| 2127 |
-
{
|
| 2128 |
-
"epoch": 0.302,
|
| 2129 |
-
"grad_norm": 0.1910482943058014,
|
| 2130 |
-
"learning_rate": 0.024,
|
| 2131 |
-
"loss": 4.328339099884033,
|
| 2132 |
-
"step": 302
|
| 2133 |
-
},
|
| 2134 |
-
{
|
| 2135 |
-
"epoch": 0.303,
|
| 2136 |
-
"grad_norm": 0.19043000042438507,
|
| 2137 |
-
"learning_rate": 0.024,
|
| 2138 |
-
"loss": 4.32870626449585,
|
| 2139 |
-
"step": 303
|
| 2140 |
-
},
|
| 2141 |
-
{
|
| 2142 |
-
"epoch": 0.304,
|
| 2143 |
-
"grad_norm": 0.20257574319839478,
|
| 2144 |
-
"learning_rate": 0.024,
|
| 2145 |
-
"loss": 4.328325271606445,
|
| 2146 |
-
"step": 304
|
| 2147 |
-
},
|
| 2148 |
-
{
|
| 2149 |
-
"epoch": 0.305,
|
| 2150 |
-
"grad_norm": 0.2169131189584732,
|
| 2151 |
-
"learning_rate": 0.024,
|
| 2152 |
-
"loss": 4.340607166290283,
|
| 2153 |
-
"step": 305
|
| 2154 |
-
},
|
| 2155 |
-
{
|
| 2156 |
-
"epoch": 0.306,
|
| 2157 |
-
"grad_norm": 0.22644677758216858,
|
| 2158 |
-
"learning_rate": 0.024,
|
| 2159 |
-
"loss": 4.3024983406066895,
|
| 2160 |
-
"step": 306
|
| 2161 |
-
},
|
| 2162 |
-
{
|
| 2163 |
-
"epoch": 0.307,
|
| 2164 |
-
"grad_norm": 0.24283051490783691,
|
| 2165 |
-
"learning_rate": 0.024,
|
| 2166 |
-
"loss": 4.28092622756958,
|
| 2167 |
-
"step": 307
|
| 2168 |
-
},
|
| 2169 |
-
{
|
| 2170 |
-
"epoch": 0.308,
|
| 2171 |
-
"grad_norm": 0.23159705102443695,
|
| 2172 |
-
"learning_rate": 0.024,
|
| 2173 |
-
"loss": 4.334567546844482,
|
| 2174 |
-
"step": 308
|
| 2175 |
-
},
|
| 2176 |
-
{
|
| 2177 |
-
"epoch": 0.309,
|
| 2178 |
-
"grad_norm": 0.22213050723075867,
|
| 2179 |
-
"learning_rate": 0.024,
|
| 2180 |
-
"loss": 4.301955223083496,
|
| 2181 |
-
"step": 309
|
| 2182 |
-
},
|
| 2183 |
-
{
|
| 2184 |
-
"epoch": 0.31,
|
| 2185 |
-
"grad_norm": 0.21359163522720337,
|
| 2186 |
-
"learning_rate": 0.024,
|
| 2187 |
-
"loss": 4.302872657775879,
|
| 2188 |
-
"step": 310
|
| 2189 |
-
},
|
| 2190 |
-
{
|
| 2191 |
-
"epoch": 0.311,
|
| 2192 |
-
"grad_norm": 0.19992981851100922,
|
| 2193 |
-
"learning_rate": 0.024,
|
| 2194 |
-
"loss": 4.353500843048096,
|
| 2195 |
-
"step": 311
|
| 2196 |
-
},
|
| 2197 |
-
{
|
| 2198 |
-
"epoch": 0.312,
|
| 2199 |
-
"grad_norm": 0.20680435001850128,
|
| 2200 |
-
"learning_rate": 0.024,
|
| 2201 |
-
"loss": 4.305147647857666,
|
| 2202 |
-
"step": 312
|
| 2203 |
-
},
|
| 2204 |
-
{
|
| 2205 |
-
"epoch": 0.313,
|
| 2206 |
-
"grad_norm": 0.2300131916999817,
|
| 2207 |
-
"learning_rate": 0.024,
|
| 2208 |
-
"loss": 4.278622150421143,
|
| 2209 |
-
"step": 313
|
| 2210 |
-
},
|
| 2211 |
-
{
|
| 2212 |
-
"epoch": 0.314,
|
| 2213 |
-
"grad_norm": 0.2551494538784027,
|
| 2214 |
-
"learning_rate": 0.024,
|
| 2215 |
-
"loss": 4.321536540985107,
|
| 2216 |
-
"step": 314
|
| 2217 |
-
},
|
| 2218 |
-
{
|
| 2219 |
-
"epoch": 0.315,
|
| 2220 |
-
"grad_norm": 0.24081861972808838,
|
| 2221 |
-
"learning_rate": 0.024,
|
| 2222 |
-
"loss": 4.299147129058838,
|
| 2223 |
-
"step": 315
|
| 2224 |
-
},
|
| 2225 |
-
{
|
| 2226 |
-
"epoch": 0.316,
|
| 2227 |
-
"grad_norm": 0.22601237893104553,
|
| 2228 |
-
"learning_rate": 0.024,
|
| 2229 |
-
"loss": 4.2895331382751465,
|
| 2230 |
-
"step": 316
|
| 2231 |
-
},
|
| 2232 |
-
{
|
| 2233 |
-
"epoch": 0.317,
|
| 2234 |
-
"grad_norm": 0.21750295162200928,
|
| 2235 |
-
"learning_rate": 0.024,
|
| 2236 |
-
"loss": 4.3120927810668945,
|
| 2237 |
-
"step": 317
|
| 2238 |
-
},
|
| 2239 |
-
{
|
| 2240 |
-
"epoch": 0.318,
|
| 2241 |
-
"grad_norm": 0.2133520096540451,
|
| 2242 |
-
"learning_rate": 0.024,
|
| 2243 |
-
"loss": 4.288265705108643,
|
| 2244 |
-
"step": 318
|
| 2245 |
-
},
|
| 2246 |
-
{
|
| 2247 |
-
"epoch": 0.319,
|
| 2248 |
-
"grad_norm": 0.22883279621601105,
|
| 2249 |
-
"learning_rate": 0.024,
|
| 2250 |
-
"loss": 4.274940013885498,
|
| 2251 |
-
"step": 319
|
| 2252 |
-
},
|
| 2253 |
-
{
|
| 2254 |
-
"epoch": 0.32,
|
| 2255 |
-
"grad_norm": 0.24649634957313538,
|
| 2256 |
-
"learning_rate": 0.024,
|
| 2257 |
-
"loss": 4.309512615203857,
|
| 2258 |
-
"step": 320
|
| 2259 |
-
},
|
| 2260 |
-
{
|
| 2261 |
-
"epoch": 0.321,
|
| 2262 |
-
"grad_norm": 0.24665279686450958,
|
| 2263 |
-
"learning_rate": 0.024,
|
| 2264 |
-
"loss": 4.295479774475098,
|
| 2265 |
-
"step": 321
|
| 2266 |
-
},
|
| 2267 |
-
{
|
| 2268 |
-
"epoch": 0.322,
|
| 2269 |
-
"grad_norm": 0.23282158374786377,
|
| 2270 |
-
"learning_rate": 0.024,
|
| 2271 |
-
"loss": 4.286541938781738,
|
| 2272 |
-
"step": 322
|
| 2273 |
-
},
|
| 2274 |
-
{
|
| 2275 |
-
"epoch": 0.323,
|
| 2276 |
-
"grad_norm": 0.23521634936332703,
|
| 2277 |
-
"learning_rate": 0.024,
|
| 2278 |
-
"loss": 4.270679950714111,
|
| 2279 |
-
"step": 323
|
| 2280 |
-
},
|
| 2281 |
-
{
|
| 2282 |
-
"epoch": 0.324,
|
| 2283 |
-
"grad_norm": 0.23924297094345093,
|
| 2284 |
-
"learning_rate": 0.024,
|
| 2285 |
-
"loss": 4.291947364807129,
|
| 2286 |
-
"step": 324
|
| 2287 |
-
},
|
| 2288 |
-
{
|
| 2289 |
-
"epoch": 0.325,
|
| 2290 |
-
"grad_norm": 0.22422827780246735,
|
| 2291 |
-
"learning_rate": 0.024,
|
| 2292 |
-
"loss": 4.302128314971924,
|
| 2293 |
-
"step": 325
|
| 2294 |
-
},
|
| 2295 |
-
{
|
| 2296 |
-
"epoch": 0.326,
|
| 2297 |
-
"grad_norm": 0.19356408715248108,
|
| 2298 |
-
"learning_rate": 0.024,
|
| 2299 |
-
"loss": 4.265312671661377,
|
| 2300 |
-
"step": 326
|
| 2301 |
-
},
|
| 2302 |
-
{
|
| 2303 |
-
"epoch": 0.327,
|
| 2304 |
-
"grad_norm": 0.16424980759620667,
|
| 2305 |
-
"learning_rate": 0.024,
|
| 2306 |
-
"loss": 4.270499229431152,
|
| 2307 |
-
"step": 327
|
| 2308 |
-
},
|
| 2309 |
-
{
|
| 2310 |
-
"epoch": 0.328,
|
| 2311 |
-
"grad_norm": 0.18081772327423096,
|
| 2312 |
-
"learning_rate": 0.024,
|
| 2313 |
-
"loss": 4.278689861297607,
|
| 2314 |
-
"step": 328
|
| 2315 |
-
},
|
| 2316 |
-
{
|
| 2317 |
-
"epoch": 0.329,
|
| 2318 |
-
"grad_norm": 0.21151621639728546,
|
| 2319 |
-
"learning_rate": 0.024,
|
| 2320 |
-
"loss": 4.216006755828857,
|
| 2321 |
-
"step": 329
|
| 2322 |
-
},
|
| 2323 |
-
{
|
| 2324 |
-
"epoch": 0.33,
|
| 2325 |
-
"grad_norm": 0.23496125638484955,
|
| 2326 |
-
"learning_rate": 0.024,
|
| 2327 |
-
"loss": 4.263980865478516,
|
| 2328 |
-
"step": 330
|
| 2329 |
-
},
|
| 2330 |
-
{
|
| 2331 |
-
"epoch": 0.331,
|
| 2332 |
-
"grad_norm": 0.24402771890163422,
|
| 2333 |
-
"learning_rate": 0.024,
|
| 2334 |
-
"loss": 4.278521537780762,
|
| 2335 |
-
"step": 331
|
| 2336 |
-
},
|
| 2337 |
-
{
|
| 2338 |
-
"epoch": 0.332,
|
| 2339 |
-
"grad_norm": 0.22897781431674957,
|
| 2340 |
-
"learning_rate": 0.024,
|
| 2341 |
-
"loss": 4.274071216583252,
|
| 2342 |
-
"step": 332
|
| 2343 |
-
},
|
| 2344 |
-
{
|
| 2345 |
-
"epoch": 0.333,
|
| 2346 |
-
"grad_norm": 0.20162004232406616,
|
| 2347 |
-
"learning_rate": 0.024,
|
| 2348 |
-
"loss": 4.216704368591309,
|
| 2349 |
-
"step": 333
|
| 2350 |
-
},
|
| 2351 |
-
{
|
| 2352 |
-
"epoch": 0.334,
|
| 2353 |
-
"grad_norm": 0.18197917938232422,
|
| 2354 |
-
"learning_rate": 0.024,
|
| 2355 |
-
"loss": 4.23854923248291,
|
| 2356 |
-
"step": 334
|
| 2357 |
-
},
|
| 2358 |
-
{
|
| 2359 |
-
"epoch": 0.335,
|
| 2360 |
-
"grad_norm": 0.1665588915348053,
|
| 2361 |
-
"learning_rate": 0.024,
|
| 2362 |
-
"loss": 4.2324934005737305,
|
| 2363 |
-
"step": 335
|
| 2364 |
-
},
|
| 2365 |
-
{
|
| 2366 |
-
"epoch": 0.336,
|
| 2367 |
-
"grad_norm": 0.17334792017936707,
|
| 2368 |
-
"learning_rate": 0.024,
|
| 2369 |
-
"loss": 4.2093119621276855,
|
| 2370 |
-
"step": 336
|
| 2371 |
-
},
|
| 2372 |
-
{
|
| 2373 |
-
"epoch": 0.337,
|
| 2374 |
-
"grad_norm": 0.17798443138599396,
|
| 2375 |
-
"learning_rate": 0.024,
|
| 2376 |
-
"loss": 4.285927772521973,
|
| 2377 |
-
"step": 337
|
| 2378 |
-
},
|
| 2379 |
-
{
|
| 2380 |
-
"epoch": 0.338,
|
| 2381 |
-
"grad_norm": 0.18512047827243805,
|
| 2382 |
-
"learning_rate": 0.024,
|
| 2383 |
-
"loss": 4.225658893585205,
|
| 2384 |
-
"step": 338
|
| 2385 |
-
},
|
| 2386 |
-
{
|
| 2387 |
-
"epoch": 0.339,
|
| 2388 |
-
"grad_norm": 0.2026454657316208,
|
| 2389 |
-
"learning_rate": 0.024,
|
| 2390 |
-
"loss": 4.244827747344971,
|
| 2391 |
-
"step": 339
|
| 2392 |
-
},
|
| 2393 |
-
{
|
| 2394 |
-
"epoch": 0.34,
|
| 2395 |
-
"grad_norm": 0.2169608324766159,
|
| 2396 |
-
"learning_rate": 0.024,
|
| 2397 |
-
"loss": 4.288611888885498,
|
| 2398 |
-
"step": 340
|
| 2399 |
-
},
|
| 2400 |
-
{
|
| 2401 |
-
"epoch": 0.341,
|
| 2402 |
-
"grad_norm": 0.20575760304927826,
|
| 2403 |
-
"learning_rate": 0.024,
|
| 2404 |
-
"loss": 4.259310722351074,
|
| 2405 |
-
"step": 341
|
| 2406 |
-
},
|
| 2407 |
-
{
|
| 2408 |
-
"epoch": 0.342,
|
| 2409 |
-
"grad_norm": 0.19972142577171326,
|
| 2410 |
-
"learning_rate": 0.024,
|
| 2411 |
-
"loss": 4.254790782928467,
|
| 2412 |
-
"step": 342
|
| 2413 |
-
},
|
| 2414 |
-
{
|
| 2415 |
-
"epoch": 0.343,
|
| 2416 |
-
"grad_norm": 0.1894325166940689,
|
| 2417 |
-
"learning_rate": 0.024,
|
| 2418 |
-
"loss": 4.191615104675293,
|
| 2419 |
-
"step": 343
|
| 2420 |
-
},
|
| 2421 |
-
{
|
| 2422 |
-
"epoch": 0.344,
|
| 2423 |
-
"grad_norm": 0.18804578483104706,
|
| 2424 |
-
"learning_rate": 0.024,
|
| 2425 |
-
"loss": 4.204664707183838,
|
| 2426 |
-
"step": 344
|
| 2427 |
-
},
|
| 2428 |
-
{
|
| 2429 |
-
"epoch": 0.345,
|
| 2430 |
-
"grad_norm": 0.19048374891281128,
|
| 2431 |
-
"learning_rate": 0.024,
|
| 2432 |
-
"loss": 4.248066425323486,
|
| 2433 |
-
"step": 345
|
| 2434 |
-
},
|
| 2435 |
-
{
|
| 2436 |
-
"epoch": 0.346,
|
| 2437 |
-
"grad_norm": 0.19222740828990936,
|
| 2438 |
-
"learning_rate": 0.024,
|
| 2439 |
-
"loss": 4.218929290771484,
|
| 2440 |
-
"step": 346
|
| 2441 |
-
},
|
| 2442 |
-
{
|
| 2443 |
-
"epoch": 0.347,
|
| 2444 |
-
"grad_norm": 0.19046323001384735,
|
| 2445 |
-
"learning_rate": 0.024,
|
| 2446 |
-
"loss": 4.2713727951049805,
|
| 2447 |
-
"step": 347
|
| 2448 |
-
},
|
| 2449 |
-
{
|
| 2450 |
-
"epoch": 0.348,
|
| 2451 |
-
"grad_norm": 0.1851787567138672,
|
| 2452 |
-
"learning_rate": 0.024,
|
| 2453 |
-
"loss": 4.231989860534668,
|
| 2454 |
-
"step": 348
|
| 2455 |
-
},
|
| 2456 |
-
{
|
| 2457 |
-
"epoch": 0.349,
|
| 2458 |
-
"grad_norm": 0.19500042498111725,
|
| 2459 |
-
"learning_rate": 0.024,
|
| 2460 |
-
"loss": 4.204928874969482,
|
| 2461 |
-
"step": 349
|
| 2462 |
-
},
|
| 2463 |
-
{
|
| 2464 |
-
"epoch": 0.35,
|
| 2465 |
-
"grad_norm": 0.18231108784675598,
|
| 2466 |
-
"learning_rate": 0.024,
|
| 2467 |
-
"loss": 4.182326316833496,
|
| 2468 |
-
"step": 350
|
| 2469 |
-
},
|
| 2470 |
-
{
|
| 2471 |
-
"epoch": 0.351,
|
| 2472 |
-
"grad_norm": 0.17235255241394043,
|
| 2473 |
-
"learning_rate": 0.024,
|
| 2474 |
-
"loss": 4.203644275665283,
|
| 2475 |
-
"step": 351
|
| 2476 |
-
},
|
| 2477 |
-
{
|
| 2478 |
-
"epoch": 0.352,
|
| 2479 |
-
"grad_norm": 0.20085746049880981,
|
| 2480 |
-
"learning_rate": 0.024,
|
| 2481 |
-
"loss": 4.231534004211426,
|
| 2482 |
-
"step": 352
|
| 2483 |
-
},
|
| 2484 |
-
{
|
| 2485 |
-
"epoch": 0.353,
|
| 2486 |
-
"grad_norm": 0.2105642408132553,
|
| 2487 |
-
"learning_rate": 0.024,
|
| 2488 |
-
"loss": 4.1583662033081055,
|
| 2489 |
-
"step": 353
|
| 2490 |
-
},
|
| 2491 |
-
{
|
| 2492 |
-
"epoch": 0.354,
|
| 2493 |
-
"grad_norm": 0.22092968225479126,
|
| 2494 |
-
"learning_rate": 0.024,
|
| 2495 |
-
"loss": 4.239068984985352,
|
| 2496 |
-
"step": 354
|
| 2497 |
-
},
|
| 2498 |
-
{
|
| 2499 |
-
"epoch": 0.355,
|
| 2500 |
-
"grad_norm": 0.22030363976955414,
|
| 2501 |
-
"learning_rate": 0.024,
|
| 2502 |
-
"loss": 4.18521785736084,
|
| 2503 |
-
"step": 355
|
| 2504 |
-
},
|
| 2505 |
-
{
|
| 2506 |
-
"epoch": 0.356,
|
| 2507 |
-
"grad_norm": 0.18694446980953217,
|
| 2508 |
-
"learning_rate": 0.024,
|
| 2509 |
-
"loss": 4.15192985534668,
|
| 2510 |
-
"step": 356
|
| 2511 |
-
},
|
| 2512 |
-
{
|
| 2513 |
-
"epoch": 0.357,
|
| 2514 |
-
"grad_norm": 0.18254533410072327,
|
| 2515 |
-
"learning_rate": 0.024,
|
| 2516 |
-
"loss": 4.173466682434082,
|
| 2517 |
-
"step": 357
|
| 2518 |
-
},
|
| 2519 |
-
{
|
| 2520 |
-
"epoch": 0.358,
|
| 2521 |
-
"grad_norm": 0.19583667814731598,
|
| 2522 |
-
"learning_rate": 0.024,
|
| 2523 |
-
"loss": 4.242255210876465,
|
| 2524 |
-
"step": 358
|
| 2525 |
-
},
|
| 2526 |
-
{
|
| 2527 |
-
"epoch": 0.359,
|
| 2528 |
-
"grad_norm": 0.20744380354881287,
|
| 2529 |
-
"learning_rate": 0.024,
|
| 2530 |
-
"loss": 4.229676246643066,
|
| 2531 |
-
"step": 359
|
| 2532 |
-
},
|
| 2533 |
-
{
|
| 2534 |
-
"epoch": 0.36,
|
| 2535 |
-
"grad_norm": 0.19830869138240814,
|
| 2536 |
-
"learning_rate": 0.024,
|
| 2537 |
-
"loss": 4.228641033172607,
|
| 2538 |
-
"step": 360
|
| 2539 |
-
},
|
| 2540 |
-
{
|
| 2541 |
-
"epoch": 0.361,
|
| 2542 |
-
"grad_norm": 0.17661122977733612,
|
| 2543 |
-
"learning_rate": 0.024,
|
| 2544 |
-
"loss": 4.205435276031494,
|
| 2545 |
-
"step": 361
|
| 2546 |
-
},
|
| 2547 |
-
{
|
| 2548 |
-
"epoch": 0.362,
|
| 2549 |
-
"grad_norm": 0.1804812103509903,
|
| 2550 |
-
"learning_rate": 0.024,
|
| 2551 |
-
"loss": 4.2216081619262695,
|
| 2552 |
-
"step": 362
|
| 2553 |
-
},
|
| 2554 |
-
{
|
| 2555 |
-
"epoch": 0.363,
|
| 2556 |
-
"grad_norm": 0.2033960223197937,
|
| 2557 |
-
"learning_rate": 0.024,
|
| 2558 |
-
"loss": 4.234243869781494,
|
| 2559 |
-
"step": 363
|
| 2560 |
-
},
|
| 2561 |
-
{
|
| 2562 |
-
"epoch": 0.364,
|
| 2563 |
-
"grad_norm": 0.20846650004386902,
|
| 2564 |
-
"learning_rate": 0.024,
|
| 2565 |
-
"loss": 4.212010383605957,
|
| 2566 |
-
"step": 364
|
| 2567 |
-
},
|
| 2568 |
-
{
|
| 2569 |
-
"epoch": 0.365,
|
| 2570 |
-
"grad_norm": 0.20494279265403748,
|
| 2571 |
-
"learning_rate": 0.024,
|
| 2572 |
-
"loss": 4.222409725189209,
|
| 2573 |
-
"step": 365
|
| 2574 |
-
},
|
| 2575 |
-
{
|
| 2576 |
-
"epoch": 0.366,
|
| 2577 |
-
"grad_norm": 0.18129266798496246,
|
| 2578 |
-
"learning_rate": 0.024,
|
| 2579 |
-
"loss": 4.1648030281066895,
|
| 2580 |
-
"step": 366
|
| 2581 |
-
},
|
| 2582 |
-
{
|
| 2583 |
-
"epoch": 0.367,
|
| 2584 |
-
"grad_norm": 0.1789444088935852,
|
| 2585 |
-
"learning_rate": 0.024,
|
| 2586 |
-
"loss": 4.162021160125732,
|
| 2587 |
-
"step": 367
|
| 2588 |
-
},
|
| 2589 |
-
{
|
| 2590 |
-
"epoch": 0.368,
|
| 2591 |
-
"grad_norm": 0.2089833915233612,
|
| 2592 |
-
"learning_rate": 0.024,
|
| 2593 |
-
"loss": 4.180186748504639,
|
| 2594 |
-
"step": 368
|
| 2595 |
-
},
|
| 2596 |
-
{
|
| 2597 |
-
"epoch": 0.369,
|
| 2598 |
-
"grad_norm": 0.2269478291273117,
|
| 2599 |
-
"learning_rate": 0.024,
|
| 2600 |
-
"loss": 4.221427917480469,
|
| 2601 |
-
"step": 369
|
| 2602 |
-
},
|
| 2603 |
-
{
|
| 2604 |
-
"epoch": 0.37,
|
| 2605 |
-
"grad_norm": 0.23993784189224243,
|
| 2606 |
-
"learning_rate": 0.024,
|
| 2607 |
-
"loss": 4.169924259185791,
|
| 2608 |
-
"step": 370
|
| 2609 |
-
},
|
| 2610 |
-
{
|
| 2611 |
-
"epoch": 0.371,
|
| 2612 |
-
"grad_norm": 0.2263093888759613,
|
| 2613 |
-
"learning_rate": 0.024,
|
| 2614 |
-
"loss": 4.182196140289307,
|
| 2615 |
-
"step": 371
|
| 2616 |
-
},
|
| 2617 |
-
{
|
| 2618 |
-
"epoch": 0.372,
|
| 2619 |
-
"grad_norm": 0.21232007443904877,
|
| 2620 |
-
"learning_rate": 0.024,
|
| 2621 |
-
"loss": 4.1692094802856445,
|
| 2622 |
-
"step": 372
|
| 2623 |
-
},
|
| 2624 |
-
{
|
| 2625 |
-
"epoch": 0.373,
|
| 2626 |
-
"grad_norm": 0.19661198556423187,
|
| 2627 |
-
"learning_rate": 0.024,
|
| 2628 |
-
"loss": 4.222322940826416,
|
| 2629 |
-
"step": 373
|
| 2630 |
-
},
|
| 2631 |
-
{
|
| 2632 |
-
"epoch": 0.374,
|
| 2633 |
-
"grad_norm": 0.17027661204338074,
|
| 2634 |
-
"learning_rate": 0.024,
|
| 2635 |
-
"loss": 4.234375,
|
| 2636 |
-
"step": 374
|
| 2637 |
-
},
|
| 2638 |
-
{
|
| 2639 |
-
"epoch": 0.375,
|
| 2640 |
-
"grad_norm": 0.16195537149906158,
|
| 2641 |
-
"learning_rate": 0.024,
|
| 2642 |
-
"loss": 4.184053897857666,
|
| 2643 |
-
"step": 375
|
| 2644 |
-
},
|
| 2645 |
-
{
|
| 2646 |
-
"epoch": 0.376,
|
| 2647 |
-
"grad_norm": 0.16979259252548218,
|
| 2648 |
-
"learning_rate": 0.024,
|
| 2649 |
-
"loss": 4.187852382659912,
|
| 2650 |
-
"step": 376
|
| 2651 |
-
},
|
| 2652 |
-
{
|
| 2653 |
-
"epoch": 0.377,
|
| 2654 |
-
"grad_norm": 0.17063412070274353,
|
| 2655 |
-
"learning_rate": 0.024,
|
| 2656 |
-
"loss": 4.214792251586914,
|
| 2657 |
-
"step": 377
|
| 2658 |
-
},
|
| 2659 |
-
{
|
| 2660 |
-
"epoch": 0.378,
|
| 2661 |
-
"grad_norm": 0.17546872794628143,
|
| 2662 |
-
"learning_rate": 0.024,
|
| 2663 |
-
"loss": 4.199203014373779,
|
| 2664 |
-
"step": 378
|
| 2665 |
-
},
|
| 2666 |
-
{
|
| 2667 |
-
"epoch": 0.379,
|
| 2668 |
-
"grad_norm": 0.1778363436460495,
|
| 2669 |
-
"learning_rate": 0.024,
|
| 2670 |
-
"loss": 4.202012538909912,
|
| 2671 |
-
"step": 379
|
| 2672 |
-
},
|
| 2673 |
-
{
|
| 2674 |
-
"epoch": 0.38,
|
| 2675 |
-
"grad_norm": 0.19184568524360657,
|
| 2676 |
-
"learning_rate": 0.024,
|
| 2677 |
-
"loss": 4.17599630355835,
|
| 2678 |
-
"step": 380
|
| 2679 |
-
},
|
| 2680 |
-
{
|
| 2681 |
-
"epoch": 0.381,
|
| 2682 |
-
"grad_norm": 0.20098869502544403,
|
| 2683 |
-
"learning_rate": 0.024,
|
| 2684 |
-
"loss": 4.203759670257568,
|
| 2685 |
-
"step": 381
|
| 2686 |
-
},
|
| 2687 |
-
{
|
| 2688 |
-
"epoch": 0.382,
|
| 2689 |
-
"grad_norm": 0.20209340751171112,
|
| 2690 |
-
"learning_rate": 0.024,
|
| 2691 |
-
"loss": 4.141083717346191,
|
| 2692 |
-
"step": 382
|
| 2693 |
-
},
|
| 2694 |
-
{
|
| 2695 |
-
"epoch": 0.383,
|
| 2696 |
-
"grad_norm": 0.18570493161678314,
|
| 2697 |
-
"learning_rate": 0.024,
|
| 2698 |
-
"loss": 4.133065700531006,
|
| 2699 |
-
"step": 383
|
| 2700 |
-
},
|
| 2701 |
-
{
|
| 2702 |
-
"epoch": 0.384,
|
| 2703 |
-
"grad_norm": 0.17936527729034424,
|
| 2704 |
-
"learning_rate": 0.024,
|
| 2705 |
-
"loss": 4.1899333000183105,
|
| 2706 |
-
"step": 384
|
| 2707 |
-
},
|
| 2708 |
-
{
|
| 2709 |
-
"epoch": 0.385,
|
| 2710 |
-
"grad_norm": 0.1793138086795807,
|
| 2711 |
-
"learning_rate": 0.024,
|
| 2712 |
-
"loss": 4.166260242462158,
|
| 2713 |
-
"step": 385
|
| 2714 |
-
},
|
| 2715 |
-
{
|
| 2716 |
-
"epoch": 0.386,
|
| 2717 |
-
"grad_norm": 0.17355501651763916,
|
| 2718 |
-
"learning_rate": 0.024,
|
| 2719 |
-
"loss": 4.183168888092041,
|
| 2720 |
-
"step": 386
|
| 2721 |
-
},
|
| 2722 |
-
{
|
| 2723 |
-
"epoch": 0.387,
|
| 2724 |
-
"grad_norm": 0.18104924261569977,
|
| 2725 |
-
"learning_rate": 0.024,
|
| 2726 |
-
"loss": 4.17315149307251,
|
| 2727 |
-
"step": 387
|
| 2728 |
-
},
|
| 2729 |
-
{
|
| 2730 |
-
"epoch": 0.388,
|
| 2731 |
-
"grad_norm": 0.1864767223596573,
|
| 2732 |
-
"learning_rate": 0.024,
|
| 2733 |
-
"loss": 4.167158126831055,
|
| 2734 |
-
"step": 388
|
| 2735 |
-
},
|
| 2736 |
-
{
|
| 2737 |
-
"epoch": 0.389,
|
| 2738 |
-
"grad_norm": 0.18434834480285645,
|
| 2739 |
-
"learning_rate": 0.024,
|
| 2740 |
-
"loss": 4.171673774719238,
|
| 2741 |
-
"step": 389
|
| 2742 |
-
},
|
| 2743 |
-
{
|
| 2744 |
-
"epoch": 0.39,
|
| 2745 |
-
"grad_norm": 0.17429886758327484,
|
| 2746 |
-
"learning_rate": 0.024,
|
| 2747 |
-
"loss": 4.132385730743408,
|
| 2748 |
-
"step": 390
|
| 2749 |
-
},
|
| 2750 |
-
{
|
| 2751 |
-
"epoch": 0.391,
|
| 2752 |
-
"grad_norm": 0.17167805135250092,
|
| 2753 |
-
"learning_rate": 0.024,
|
| 2754 |
-
"loss": 4.155993461608887,
|
| 2755 |
-
"step": 391
|
| 2756 |
-
},
|
| 2757 |
-
{
|
| 2758 |
-
"epoch": 0.392,
|
| 2759 |
-
"grad_norm": 0.17526599764823914,
|
| 2760 |
-
"learning_rate": 0.024,
|
| 2761 |
-
"loss": 4.153561115264893,
|
| 2762 |
-
"step": 392
|
| 2763 |
-
},
|
| 2764 |
-
{
|
| 2765 |
-
"epoch": 0.393,
|
| 2766 |
-
"grad_norm": 0.18514658510684967,
|
| 2767 |
-
"learning_rate": 0.024,
|
| 2768 |
-
"loss": 4.116867542266846,
|
| 2769 |
-
"step": 393
|
| 2770 |
-
},
|
| 2771 |
-
{
|
| 2772 |
-
"epoch": 0.394,
|
| 2773 |
-
"grad_norm": 0.1985170543193817,
|
| 2774 |
-
"learning_rate": 0.024,
|
| 2775 |
-
"loss": 4.198697090148926,
|
| 2776 |
-
"step": 394
|
| 2777 |
-
},
|
| 2778 |
-
{
|
| 2779 |
-
"epoch": 0.395,
|
| 2780 |
-
"grad_norm": 0.19961412250995636,
|
| 2781 |
-
"learning_rate": 0.024,
|
| 2782 |
-
"loss": 4.195560455322266,
|
| 2783 |
-
"step": 395
|
| 2784 |
-
},
|
| 2785 |
-
{
|
| 2786 |
-
"epoch": 0.396,
|
| 2787 |
-
"grad_norm": 0.1817820817232132,
|
| 2788 |
-
"learning_rate": 0.024,
|
| 2789 |
-
"loss": 4.170080184936523,
|
| 2790 |
-
"step": 396
|
| 2791 |
-
},
|
| 2792 |
-
{
|
| 2793 |
-
"epoch": 0.397,
|
| 2794 |
-
"grad_norm": 0.1824864149093628,
|
| 2795 |
-
"learning_rate": 0.024,
|
| 2796 |
-
"loss": 4.193587303161621,
|
| 2797 |
-
"step": 397
|
| 2798 |
-
},
|
| 2799 |
-
{
|
| 2800 |
-
"epoch": 0.398,
|
| 2801 |
-
"grad_norm": 0.18462251126766205,
|
| 2802 |
-
"learning_rate": 0.024,
|
| 2803 |
-
"loss": 4.170504570007324,
|
| 2804 |
-
"step": 398
|
| 2805 |
-
},
|
| 2806 |
-
{
|
| 2807 |
-
"epoch": 0.399,
|
| 2808 |
-
"grad_norm": 0.18598416447639465,
|
| 2809 |
-
"learning_rate": 0.024,
|
| 2810 |
-
"loss": 4.163930416107178,
|
| 2811 |
-
"step": 399
|
| 2812 |
-
},
|
| 2813 |
-
{
|
| 2814 |
-
"epoch": 0.4,
|
| 2815 |
-
"grad_norm": 0.18852722644805908,
|
| 2816 |
-
"learning_rate": 0.024,
|
| 2817 |
-
"loss": 4.165492534637451,
|
| 2818 |
-
"step": 400
|
| 2819 |
-
},
|
| 2820 |
-
{
|
| 2821 |
-
"epoch": 0.4,
|
| 2822 |
-
"eval_loss": 4.120884418487549,
|
| 2823 |
-
"eval_runtime": 75.8459,
|
| 2824 |
-
"eval_samples_per_second": 13.501,
|
| 2825 |
-
"eval_steps_per_second": 3.375,
|
| 2826 |
-
"step": 400
|
| 2827 |
}
|
| 2828 |
],
|
| 2829 |
"logging_steps": 1,
|
|
@@ -2843,7 +727,7 @@
|
|
| 2843 |
"attributes": {}
|
| 2844 |
}
|
| 2845 |
},
|
| 2846 |
-
"total_flos":
|
| 2847 |
"train_batch_size": 4,
|
| 2848 |
"trial_name": null,
|
| 2849 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.1,
|
| 6 |
"eval_steps": 200,
|
| 7 |
+
"global_step": 100,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.001,
|
| 14 |
+
"grad_norm": 6.191351890563965,
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
+
"loss": 11.086275100708008,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.002,
|
| 21 |
+
"grad_norm": 6.183396339416504,
|
| 22 |
+
"learning_rate": 0.00011999999999999999,
|
| 23 |
+
"loss": 11.084210395812988,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.003,
|
| 28 |
+
"grad_norm": 40.25852584838867,
|
| 29 |
+
"learning_rate": 0.00023999999999999998,
|
| 30 |
+
"loss": 11.06926441192627,
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.004,
|
| 35 |
+
"grad_norm": 3.6167638301849365,
|
| 36 |
+
"learning_rate": 0.00035999999999999997,
|
| 37 |
+
"loss": 12.345776557922363,
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.005,
|
| 42 |
+
"grad_norm": 37.04316711425781,
|
| 43 |
+
"learning_rate": 0.00047999999999999996,
|
| 44 |
+
"loss": 12.08117961883545,
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.006,
|
| 49 |
+
"grad_norm": 2.5832877159118652,
|
| 50 |
+
"learning_rate": 0.0006,
|
| 51 |
+
"loss": 10.664973258972168,
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.007,
|
| 56 |
+
"grad_norm": 2.420443296432495,
|
| 57 |
+
"learning_rate": 0.0007199999999999999,
|
| 58 |
+
"loss": 9.976428031921387,
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.008,
|
| 63 |
+
"grad_norm": 2.208857774734497,
|
| 64 |
+
"learning_rate": 0.0008399999999999999,
|
| 65 |
+
"loss": 9.32384204864502,
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.009,
|
| 70 |
+
"grad_norm": 1.4080778360366821,
|
| 71 |
+
"learning_rate": 0.0009599999999999999,
|
| 72 |
+
"loss": 8.844566345214844,
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.01,
|
| 77 |
+
"grad_norm": 0.932022750377655,
|
| 78 |
+
"learning_rate": 0.00108,
|
| 79 |
+
"loss": 8.574441909790039,
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.011,
|
| 84 |
+
"grad_norm": 0.905206024646759,
|
| 85 |
+
"learning_rate": 0.0012,
|
| 86 |
+
"loss": 8.570406913757324,
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.012,
|
| 91 |
+
"grad_norm": 1.9950780868530273,
|
| 92 |
+
"learning_rate": 0.0012,
|
| 93 |
+
"loss": 8.654431343078613,
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.013,
|
| 98 |
+
"grad_norm": 1.4423072338104248,
|
| 99 |
+
"learning_rate": 0.0012,
|
| 100 |
+
"loss": 8.775788307189941,
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.014,
|
| 105 |
+
"grad_norm": 1.3934134244918823,
|
| 106 |
+
"learning_rate": 0.0012,
|
| 107 |
+
"loss": 8.70737361907959,
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.015,
|
| 112 |
+
"grad_norm": 0.6730704307556152,
|
| 113 |
+
"learning_rate": 0.0012,
|
| 114 |
+
"loss": 8.58652114868164,
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.016,
|
| 119 |
+
"grad_norm": 1.3850765228271484,
|
| 120 |
+
"learning_rate": 0.0012,
|
| 121 |
+
"loss": 8.599067687988281,
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.017,
|
| 126 |
+
"grad_norm": 1.1575509309768677,
|
| 127 |
+
"learning_rate": 0.0012,
|
| 128 |
+
"loss": 8.577943801879883,
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.018,
|
| 133 |
+
"grad_norm": 1.094803810119629,
|
| 134 |
+
"learning_rate": 0.0012,
|
| 135 |
+
"loss": 8.583452224731445,
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.019,
|
| 140 |
+
"grad_norm": 0.9581138491630554,
|
| 141 |
+
"learning_rate": 0.0012,
|
| 142 |
+
"loss": 8.573091506958008,
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.02,
|
| 147 |
+
"grad_norm": 0.8642209768295288,
|
| 148 |
+
"learning_rate": 0.0012,
|
| 149 |
+
"loss": 8.558369636535645,
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.021,
|
| 154 |
+
"grad_norm": 0.5513532757759094,
|
| 155 |
+
"learning_rate": 0.0012,
|
| 156 |
+
"loss": 8.547396659851074,
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.022,
|
| 161 |
+
"grad_norm": 0.5867564678192139,
|
| 162 |
+
"learning_rate": 0.0012,
|
| 163 |
+
"loss": 8.557206153869629,
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.023,
|
| 168 |
+
"grad_norm": 0.5513554215431213,
|
| 169 |
+
"learning_rate": 0.0012,
|
| 170 |
+
"loss": 8.519376754760742,
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.024,
|
| 175 |
+
"grad_norm": 0.5908657908439636,
|
| 176 |
+
"learning_rate": 0.0012,
|
| 177 |
+
"loss": 8.571252822875977,
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.025,
|
| 182 |
+
"grad_norm": 0.5163703560829163,
|
| 183 |
+
"learning_rate": 0.0012,
|
| 184 |
+
"loss": 8.550788879394531,
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.026,
|
| 189 |
+
"grad_norm": 0.6450480818748474,
|
| 190 |
+
"learning_rate": 0.0012,
|
| 191 |
+
"loss": 8.553627014160156,
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.027,
|
| 196 |
+
"grad_norm": 0.4144176244735718,
|
| 197 |
+
"learning_rate": 0.0012,
|
| 198 |
+
"loss": 8.526437759399414,
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.028,
|
| 203 |
+
"grad_norm": 0.5550076961517334,
|
| 204 |
+
"learning_rate": 0.0012,
|
| 205 |
+
"loss": 8.547487258911133,
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.029,
|
| 210 |
+
"grad_norm": 0.3792600631713867,
|
| 211 |
+
"learning_rate": 0.0012,
|
| 212 |
+
"loss": 8.534337043762207,
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.03,
|
| 217 |
+
"grad_norm": 0.38315337896347046,
|
| 218 |
+
"learning_rate": 0.0012,
|
| 219 |
+
"loss": 8.537225723266602,
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.031,
|
| 224 |
+
"grad_norm": 0.44758984446525574,
|
| 225 |
+
"learning_rate": 0.0012,
|
| 226 |
+
"loss": 8.539337158203125,
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.032,
|
| 231 |
+
"grad_norm": 0.42012736201286316,
|
| 232 |
+
"learning_rate": 0.0012,
|
| 233 |
+
"loss": 8.50920581817627,
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.033,
|
| 238 |
+
"grad_norm": 0.4103575348854065,
|
| 239 |
+
"learning_rate": 0.0012,
|
| 240 |
+
"loss": 8.508569717407227,
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.034,
|
| 245 |
+
"grad_norm": 0.2795290946960449,
|
| 246 |
+
"learning_rate": 0.0012,
|
| 247 |
+
"loss": 8.499092102050781,
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.035,
|
| 252 |
+
"grad_norm": 0.47300985455513,
|
| 253 |
+
"learning_rate": 0.0012,
|
| 254 |
+
"loss": 8.508503913879395,
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.036,
|
| 259 |
+
"grad_norm": 0.27382028102874756,
|
| 260 |
+
"learning_rate": 0.0012,
|
| 261 |
+
"loss": 8.498414039611816,
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.037,
|
| 266 |
+
"grad_norm": 0.37145861983299255,
|
| 267 |
+
"learning_rate": 0.0012,
|
| 268 |
+
"loss": 8.516667366027832,
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.038,
|
| 273 |
+
"grad_norm": 0.26397329568862915,
|
| 274 |
+
"learning_rate": 0.0012,
|
| 275 |
+
"loss": 8.515829086303711,
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.039,
|
| 280 |
+
"grad_norm": 0.2639530599117279,
|
| 281 |
+
"learning_rate": 0.0012,
|
| 282 |
+
"loss": 8.509860038757324,
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.04,
|
| 287 |
+
"grad_norm": 0.22002728283405304,
|
| 288 |
+
"learning_rate": 0.0012,
|
| 289 |
+
"loss": 8.498579978942871,
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.041,
|
| 294 |
+
"grad_norm": 0.2718033194541931,
|
| 295 |
+
"learning_rate": 0.0012,
|
| 296 |
+
"loss": 8.4937105178833,
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.042,
|
| 301 |
+
"grad_norm": 0.2288224995136261,
|
| 302 |
+
"learning_rate": 0.0012,
|
| 303 |
+
"loss": 8.510730743408203,
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.043,
|
| 308 |
+
"grad_norm": 0.2424091100692749,
|
| 309 |
+
"learning_rate": 0.0012,
|
| 310 |
+
"loss": 8.484707832336426,
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.044,
|
| 315 |
+
"grad_norm": 0.2990792393684387,
|
| 316 |
+
"learning_rate": 0.0012,
|
| 317 |
+
"loss": 8.510951042175293,
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.045,
|
| 322 |
+
"grad_norm": 0.2716998755931854,
|
| 323 |
+
"learning_rate": 0.0012,
|
| 324 |
+
"loss": 8.487112998962402,
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.046,
|
| 329 |
+
"grad_norm": 0.31668758392333984,
|
| 330 |
+
"learning_rate": 0.0012,
|
| 331 |
+
"loss": 8.497540473937988,
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.047,
|
| 336 |
+
"grad_norm": 0.2717650830745697,
|
| 337 |
+
"learning_rate": 0.0012,
|
| 338 |
+
"loss": 8.497724533081055,
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.048,
|
| 343 |
+
"grad_norm": 0.2814731299877167,
|
| 344 |
+
"learning_rate": 0.0012,
|
| 345 |
+
"loss": 8.501847267150879,
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.049,
|
| 350 |
+
"grad_norm": 0.31272783875465393,
|
| 351 |
+
"learning_rate": 0.0012,
|
| 352 |
+
"loss": 8.50948715209961,
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.05,
|
| 357 |
+
"grad_norm": 0.24348217248916626,
|
| 358 |
+
"learning_rate": 0.0012,
|
| 359 |
+
"loss": 8.495022773742676,
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.051,
|
| 364 |
+
"grad_norm": 0.3011173605918884,
|
| 365 |
+
"learning_rate": 0.0012,
|
| 366 |
+
"loss": 8.496807098388672,
|
| 367 |
"step": 51
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.052,
|
| 371 |
+
"grad_norm": 0.29326388239860535,
|
| 372 |
+
"learning_rate": 0.0012,
|
| 373 |
+
"loss": 8.492173194885254,
|
| 374 |
"step": 52
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.053,
|
| 378 |
+
"grad_norm": 0.29722103476524353,
|
| 379 |
+
"learning_rate": 0.0012,
|
| 380 |
+
"loss": 8.480097770690918,
|
| 381 |
"step": 53
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.054,
|
| 385 |
+
"grad_norm": 0.21530401706695557,
|
| 386 |
+
"learning_rate": 0.0012,
|
| 387 |
+
"loss": 8.49035930633545,
|
| 388 |
"step": 54
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.055,
|
| 392 |
+
"grad_norm": 0.2602374851703644,
|
| 393 |
+
"learning_rate": 0.0012,
|
| 394 |
+
"loss": 8.483044624328613,
|
| 395 |
"step": 55
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.056,
|
| 399 |
+
"grad_norm": 0.26086798310279846,
|
| 400 |
+
"learning_rate": 0.0012,
|
| 401 |
+
"loss": 8.479232788085938,
|
| 402 |
"step": 56
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.057,
|
| 406 |
+
"grad_norm": 0.3874048590660095,
|
| 407 |
+
"learning_rate": 0.0012,
|
| 408 |
+
"loss": 8.478414535522461,
|
| 409 |
"step": 57
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.058,
|
| 413 |
+
"grad_norm": 0.7225645780563354,
|
| 414 |
+
"learning_rate": 0.0012,
|
| 415 |
+
"loss": 8.495964050292969,
|
| 416 |
"step": 58
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.059,
|
| 420 |
+
"grad_norm": 1.4137455224990845,
|
| 421 |
+
"learning_rate": 0.0012,
|
| 422 |
+
"loss": 8.577896118164062,
|
| 423 |
"step": 59
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.06,
|
| 427 |
+
"grad_norm": 0.5455415844917297,
|
| 428 |
+
"learning_rate": 0.0012,
|
| 429 |
+
"loss": 8.526337623596191,
|
| 430 |
"step": 60
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.061,
|
| 434 |
+
"grad_norm": 2.8609142303466797,
|
| 435 |
+
"learning_rate": 0.0012,
|
| 436 |
+
"loss": 8.514455795288086,
|
| 437 |
"step": 61
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.062,
|
| 441 |
+
"grad_norm": 1.440839171409607,
|
| 442 |
+
"learning_rate": 0.0012,
|
| 443 |
+
"loss": 8.634124755859375,
|
| 444 |
"step": 62
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.063,
|
| 448 |
+
"grad_norm": 1.2359699010849,
|
| 449 |
+
"learning_rate": 0.0012,
|
| 450 |
+
"loss": 8.568485260009766,
|
| 451 |
"step": 63
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.064,
|
| 455 |
+
"grad_norm": 1.180612564086914,
|
| 456 |
+
"learning_rate": 0.0012,
|
| 457 |
+
"loss": 8.51024341583252,
|
| 458 |
"step": 64
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.065,
|
| 462 |
+
"grad_norm": 1.0028116703033447,
|
| 463 |
+
"learning_rate": 0.0012,
|
| 464 |
+
"loss": 8.499656677246094,
|
| 465 |
"step": 65
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.066,
|
| 469 |
+
"grad_norm": 2.423152446746826,
|
| 470 |
+
"learning_rate": 0.0012,
|
| 471 |
+
"loss": 8.472902297973633,
|
| 472 |
"step": 66
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.067,
|
| 476 |
+
"grad_norm": 1.8138713836669922,
|
| 477 |
+
"learning_rate": 0.0012,
|
| 478 |
+
"loss": 8.483245849609375,
|
| 479 |
"step": 67
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.068,
|
| 483 |
+
"grad_norm": 0.959982693195343,
|
| 484 |
+
"learning_rate": 0.0012,
|
| 485 |
+
"loss": 8.483409881591797,
|
| 486 |
"step": 68
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.069,
|
| 490 |
+
"grad_norm": 0.8686690330505371,
|
| 491 |
+
"learning_rate": 0.0012,
|
| 492 |
+
"loss": 8.440892219543457,
|
| 493 |
"step": 69
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.07,
|
| 497 |
+
"grad_norm": 0.7423587441444397,
|
| 498 |
+
"learning_rate": 0.0012,
|
| 499 |
+
"loss": 8.409979820251465,
|
| 500 |
"step": 70
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.071,
|
| 504 |
+
"grad_norm": 0.88447505235672,
|
| 505 |
+
"learning_rate": 0.0012,
|
| 506 |
+
"loss": 8.390122413635254,
|
| 507 |
"step": 71
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.072,
|
| 511 |
+
"grad_norm": 0.982702910900116,
|
| 512 |
+
"learning_rate": 0.0012,
|
| 513 |
+
"loss": 8.37291431427002,
|
| 514 |
"step": 72
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.073,
|
| 518 |
+
"grad_norm": 0.6741560101509094,
|
| 519 |
+
"learning_rate": 0.0012,
|
| 520 |
+
"loss": 8.374544143676758,
|
| 521 |
"step": 73
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.074,
|
| 525 |
+
"grad_norm": 0.9037299156188965,
|
| 526 |
+
"learning_rate": 0.0012,
|
| 527 |
+
"loss": 8.357539176940918,
|
| 528 |
"step": 74
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.075,
|
| 532 |
+
"grad_norm": 0.5122299790382385,
|
| 533 |
+
"learning_rate": 0.0012,
|
| 534 |
+
"loss": 8.336565017700195,
|
| 535 |
"step": 75
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.076,
|
| 539 |
+
"grad_norm": 1.5479395389556885,
|
| 540 |
+
"learning_rate": 0.0012,
|
| 541 |
+
"loss": 8.325927734375,
|
| 542 |
"step": 76
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.077,
|
| 546 |
+
"grad_norm": 1.3528615236282349,
|
| 547 |
+
"learning_rate": 0.0012,
|
| 548 |
+
"loss": 8.327680587768555,
|
| 549 |
"step": 77
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.078,
|
| 553 |
+
"grad_norm": 0.7083619832992554,
|
| 554 |
+
"learning_rate": 0.0012,
|
| 555 |
+
"loss": 8.325244903564453,
|
| 556 |
"step": 78
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.079,
|
| 560 |
+
"grad_norm": 0.9852667450904846,
|
| 561 |
+
"learning_rate": 0.0012,
|
| 562 |
+
"loss": 8.305098533630371,
|
| 563 |
"step": 79
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.08,
|
| 567 |
+
"grad_norm": 0.5205386281013489,
|
| 568 |
+
"learning_rate": 0.0012,
|
| 569 |
+
"loss": 8.274360656738281,
|
| 570 |
"step": 80
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.081,
|
| 574 |
+
"grad_norm": 0.946526825428009,
|
| 575 |
+
"learning_rate": 0.0012,
|
| 576 |
+
"loss": 8.282258033752441,
|
| 577 |
"step": 81
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.082,
|
| 581 |
+
"grad_norm": 0.6880609393119812,
|
| 582 |
+
"learning_rate": 0.0012,
|
| 583 |
+
"loss": 8.254171371459961,
|
| 584 |
"step": 82
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.083,
|
| 588 |
+
"grad_norm": 0.5522202253341675,
|
| 589 |
+
"learning_rate": 0.0012,
|
| 590 |
+
"loss": 8.278489112854004,
|
| 591 |
"step": 83
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.084,
|
| 595 |
+
"grad_norm": 0.6842210292816162,
|
| 596 |
+
"learning_rate": 0.0012,
|
| 597 |
+
"loss": 8.220597267150879,
|
| 598 |
"step": 84
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.085,
|
| 602 |
+
"grad_norm": 0.29896456003189087,
|
| 603 |
+
"learning_rate": 0.0012,
|
| 604 |
+
"loss": 8.204936981201172,
|
| 605 |
"step": 85
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.086,
|
| 609 |
+
"grad_norm": 0.748399555683136,
|
| 610 |
+
"learning_rate": 0.0012,
|
| 611 |
+
"loss": 8.220271110534668,
|
| 612 |
"step": 86
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.087,
|
| 616 |
+
"grad_norm": 0.28956496715545654,
|
| 617 |
+
"learning_rate": 0.0012,
|
| 618 |
+
"loss": 8.216758728027344,
|
| 619 |
"step": 87
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.088,
|
| 623 |
+
"grad_norm": 0.5921151638031006,
|
| 624 |
+
"learning_rate": 0.0012,
|
| 625 |
+
"loss": 8.215239524841309,
|
| 626 |
"step": 88
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.089,
|
| 630 |
+
"grad_norm": 0.451276570558548,
|
| 631 |
+
"learning_rate": 0.0012,
|
| 632 |
+
"loss": 8.17009162902832,
|
| 633 |
"step": 89
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.09,
|
| 637 |
+
"grad_norm": 0.5010297894477844,
|
| 638 |
+
"learning_rate": 0.0012,
|
| 639 |
+
"loss": 8.161727905273438,
|
| 640 |
"step": 90
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.091,
|
| 644 |
+
"grad_norm": 0.30494871735572815,
|
| 645 |
+
"learning_rate": 0.0012,
|
| 646 |
+
"loss": 8.166382789611816,
|
| 647 |
"step": 91
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.092,
|
| 651 |
+
"grad_norm": 0.43188294768333435,
|
| 652 |
+
"learning_rate": 0.0012,
|
| 653 |
+
"loss": 8.167814254760742,
|
| 654 |
"step": 92
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.093,
|
| 658 |
+
"grad_norm": 0.28245413303375244,
|
| 659 |
+
"learning_rate": 0.0012,
|
| 660 |
+
"loss": 8.160371780395508,
|
| 661 |
"step": 93
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.094,
|
| 665 |
+
"grad_norm": 0.4027661383152008,
|
| 666 |
+
"learning_rate": 0.0012,
|
| 667 |
+
"loss": 8.109189987182617,
|
| 668 |
"step": 94
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.095,
|
| 672 |
+
"grad_norm": 0.2549852132797241,
|
| 673 |
+
"learning_rate": 0.0012,
|
| 674 |
+
"loss": 8.114521026611328,
|
| 675 |
"step": 95
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.096,
|
| 679 |
+
"grad_norm": 0.3145453929901123,
|
| 680 |
+
"learning_rate": 0.0012,
|
| 681 |
+
"loss": 8.10285472869873,
|
| 682 |
"step": 96
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.097,
|
| 686 |
+
"grad_norm": 0.37485775351524353,
|
| 687 |
+
"learning_rate": 0.0012,
|
| 688 |
+
"loss": 8.09223747253418,
|
| 689 |
"step": 97
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.098,
|
| 693 |
+
"grad_norm": 0.36243894696235657,
|
| 694 |
+
"learning_rate": 0.0012,
|
| 695 |
+
"loss": 8.064014434814453,
|
| 696 |
"step": 98
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.099,
|
| 700 |
+
"grad_norm": 0.2771179676055908,
|
| 701 |
+
"learning_rate": 0.0012,
|
| 702 |
+
"loss": 8.062220573425293,
|
| 703 |
"step": 99
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.1,
|
| 707 |
+
"grad_norm": 0.2642522156238556,
|
| 708 |
+
"learning_rate": 0.0012,
|
| 709 |
+
"loss": 8.07100772857666,
|
| 710 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 711 |
}
|
| 712 |
],
|
| 713 |
"logging_steps": 1,
|
|
|
|
| 727 |
"attributes": {}
|
| 728 |
}
|
| 729 |
},
|
| 730 |
+
"total_flos": 2.469577768971264e+17,
|
| 731 |
"train_batch_size": 4,
|
| 732 |
"trial_name": null,
|
| 733 |
"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 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d58e58b86ebc8ee74fda679712e3a52e109c5c26268157eabc8b5c0cd94155a
|
| 3 |
size 5777
|