Update model
Browse files- trainer_state.json +993 -993
trainer_state.json
CHANGED
|
@@ -10,2234 +10,2234 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004032258064516129,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 4.999799414013322e-06,
|
| 15 |
-
"loss": 3.
|
| 16 |
-
"mean_token_accuracy": 0.
|
| 17 |
-
"num_tokens":
|
| 18 |
"step": 1
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 0.008064516129032258,
|
| 22 |
-
"grad_norm":
|
| 23 |
"learning_rate": 4.999197688241076e-06,
|
| 24 |
-
"loss": 2.
|
| 25 |
-
"mean_token_accuracy": 0.
|
| 26 |
-
"num_tokens":
|
| 27 |
"step": 2
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.012096774193548387,
|
| 31 |
-
"grad_norm":
|
| 32 |
"learning_rate": 4.998194919241471e-06,
|
| 33 |
-
"loss":
|
| 34 |
-
"mean_token_accuracy": 0.
|
| 35 |
-
"num_tokens":
|
| 36 |
"step": 3
|
| 37 |
},
|
| 38 |
{
|
| 39 |
"epoch": 0.016129032258064516,
|
| 40 |
-
"grad_norm":
|
| 41 |
"learning_rate": 4.996791267927632e-06,
|
| 42 |
-
"loss": 1.
|
| 43 |
-
"mean_token_accuracy": 0.
|
| 44 |
-
"num_tokens":
|
| 45 |
"step": 4
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.020161290322580645,
|
| 49 |
-
"grad_norm":
|
| 50 |
"learning_rate": 4.994986959541788e-06,
|
| 51 |
-
"loss":
|
| 52 |
-
"mean_token_accuracy": 0.
|
| 53 |
-
"num_tokens":
|
| 54 |
"step": 5
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.024193548387096774,
|
| 58 |
-
"grad_norm": 8.
|
| 59 |
"learning_rate": 4.9927822836191185e-06,
|
| 60 |
-
"loss": 1.
|
| 61 |
-
"mean_token_accuracy": 0.
|
| 62 |
-
"num_tokens":
|
| 63 |
"step": 6
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.028225806451612902,
|
| 67 |
-
"grad_norm": 8.
|
| 68 |
"learning_rate": 4.990177593941303e-06,
|
| 69 |
-
"loss": 1.
|
| 70 |
-
"mean_token_accuracy": 0.
|
| 71 |
-
"num_tokens":
|
| 72 |
"step": 7
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.03225806451612903,
|
| 76 |
-
"grad_norm": 8.
|
| 77 |
"learning_rate": 4.987173308479738e-06,
|
| 78 |
-
"loss": 1.
|
| 79 |
-
"mean_token_accuracy": 0.
|
| 80 |
-
"num_tokens":
|
| 81 |
"step": 8
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.036290322580645164,
|
| 85 |
-
"grad_norm": 7.
|
| 86 |
"learning_rate": 4.9837699093284765e-06,
|
| 87 |
-
"loss": 1.
|
| 88 |
-
"mean_token_accuracy": 0.
|
| 89 |
-
"num_tokens":
|
| 90 |
"step": 9
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"epoch": 0.04032258064516129,
|
| 94 |
-
"grad_norm":
|
| 95 |
"learning_rate": 4.9799679426268575e-06,
|
| 96 |
-
"loss": 1.
|
| 97 |
-
"mean_token_accuracy": 0.
|
| 98 |
-
"num_tokens":
|
| 99 |
"step": 10
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.04435483870967742,
|
| 103 |
-
"grad_norm": 7.
|
| 104 |
"learning_rate": 4.975768018471877e-06,
|
| 105 |
-
"loss": 1.
|
| 106 |
-
"mean_token_accuracy": 0.
|
| 107 |
-
"num_tokens":
|
| 108 |
"step": 11
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.04838709677419355,
|
| 112 |
-
"grad_norm":
|
| 113 |
"learning_rate": 4.971170810820279e-06,
|
| 114 |
-
"loss": 1.
|
| 115 |
-
"mean_token_accuracy": 0.
|
| 116 |
-
"num_tokens":
|
| 117 |
"step": 12
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 0.05241935483870968,
|
| 121 |
-
"grad_norm": 7.
|
| 122 |
"learning_rate": 4.966177057380409e-06,
|
| 123 |
-
"loss": 1.
|
| 124 |
-
"mean_token_accuracy": 0.
|
| 125 |
-
"num_tokens":
|
| 126 |
"step": 13
|
| 127 |
},
|
| 128 |
{
|
| 129 |
"epoch": 0.056451612903225805,
|
| 130 |
-
"grad_norm": 7.
|
| 131 |
"learning_rate": 4.960787559493836e-06,
|
| 132 |
-
"loss": 1.
|
| 133 |
-
"mean_token_accuracy": 0.
|
| 134 |
-
"num_tokens":
|
| 135 |
"step": 14
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.06048387096774194,
|
| 139 |
-
"grad_norm": 6.
|
| 140 |
"learning_rate": 4.955003182006761e-06,
|
| 141 |
-
"loss": 1.
|
| 142 |
-
"mean_token_accuracy": 0.
|
| 143 |
-
"num_tokens":
|
| 144 |
"step": 15
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.06451612903225806,
|
| 148 |
-
"grad_norm": 7.
|
| 149 |
"learning_rate": 4.948824853131237e-06,
|
| 150 |
-
"loss": 1.
|
| 151 |
-
"mean_token_accuracy": 0.
|
| 152 |
-
"num_tokens":
|
| 153 |
"step": 16
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.06854838709677419,
|
| 157 |
-
"grad_norm": 6.
|
| 158 |
"learning_rate": 4.942253564296217e-06,
|
| 159 |
-
"loss": 1.
|
| 160 |
-
"mean_token_accuracy": 0.
|
| 161 |
-
"num_tokens":
|
| 162 |
"step": 17
|
| 163 |
},
|
| 164 |
{
|
| 165 |
"epoch": 0.07258064516129033,
|
| 166 |
-
"grad_norm": 6.
|
| 167 |
"learning_rate": 4.935290369988468e-06,
|
| 168 |
-
"loss": 1.
|
| 169 |
-
"mean_token_accuracy": 0.
|
| 170 |
-
"num_tokens":
|
| 171 |
"step": 18
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.07661290322580645,
|
| 175 |
-
"grad_norm":
|
| 176 |
"learning_rate": 4.927936387583348e-06,
|
| 177 |
-
"loss": 1.
|
| 178 |
-
"mean_token_accuracy": 0.
|
| 179 |
-
"num_tokens":
|
| 180 |
"step": 19
|
| 181 |
},
|
| 182 |
{
|
| 183 |
"epoch": 0.08064516129032258,
|
| 184 |
-
"grad_norm": 6.
|
| 185 |
"learning_rate": 4.920192797165511e-06,
|
| 186 |
-
"loss": 1.
|
| 187 |
-
"mean_token_accuracy": 0.
|
| 188 |
-
"num_tokens":
|
| 189 |
"step": 20
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.0846774193548387,
|
| 193 |
-
"grad_norm": 6.
|
| 194 |
"learning_rate": 4.912060841339536e-06,
|
| 195 |
-
"loss": 1.
|
| 196 |
-
"mean_token_accuracy": 0.
|
| 197 |
-
"num_tokens":
|
| 198 |
"step": 21
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.08870967741935484,
|
| 202 |
-
"grad_norm": 7.
|
| 203 |
"learning_rate": 4.9035418250305314e-06,
|
| 204 |
-
"loss": 1.
|
| 205 |
-
"mean_token_accuracy": 0.
|
| 206 |
-
"num_tokens":
|
| 207 |
"step": 22
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.09274193548387097,
|
| 211 |
-
"grad_norm": 6.
|
| 212 |
"learning_rate": 4.894637115274728e-06,
|
| 213 |
-
"loss": 1.
|
| 214 |
-
"mean_token_accuracy": 0.
|
| 215 |
-
"num_tokens":
|
| 216 |
"step": 23
|
| 217 |
},
|
| 218 |
{
|
| 219 |
"epoch": 0.0967741935483871,
|
| 220 |
-
"grad_norm": 6.
|
| 221 |
"learning_rate": 4.8853481410001225e-06,
|
| 222 |
-
"loss": 1.
|
| 223 |
-
"mean_token_accuracy": 0.
|
| 224 |
-
"num_tokens":
|
| 225 |
"step": 24
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 0.10080645161290322,
|
| 229 |
-
"grad_norm": 6.
|
| 230 |
"learning_rate": 4.875676392797169e-06,
|
| 231 |
-
"loss": 1.
|
| 232 |
-
"mean_token_accuracy": 0.
|
| 233 |
-
"num_tokens":
|
| 234 |
"step": 25
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.10483870967741936,
|
| 238 |
-
"grad_norm": 6.
|
| 239 |
"learning_rate": 4.865623422679593e-06,
|
| 240 |
-
"loss": 1.
|
| 241 |
-
"mean_token_accuracy": 0.
|
| 242 |
-
"num_tokens":
|
| 243 |
"step": 26
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 0.10887096774193548,
|
| 247 |
-
"grad_norm":
|
| 248 |
"learning_rate": 4.855190843835338e-06,
|
| 249 |
-
"loss": 1.
|
| 250 |
-
"mean_token_accuracy": 0.
|
| 251 |
-
"num_tokens":
|
| 252 |
"step": 27
|
| 253 |
},
|
| 254 |
{
|
| 255 |
"epoch": 0.11290322580645161,
|
| 256 |
-
"grad_norm": 6.
|
| 257 |
"learning_rate": 4.844380330367701e-06,
|
| 258 |
-
"loss": 1.
|
| 259 |
-
"mean_token_accuracy": 0.
|
| 260 |
-
"num_tokens":
|
| 261 |
"step": 28
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.11693548387096774,
|
| 265 |
-
"grad_norm": 6.
|
| 266 |
"learning_rate": 4.833193617026692e-06,
|
| 267 |
-
"loss": 1.
|
| 268 |
-
"mean_token_accuracy": 0.
|
| 269 |
-
"num_tokens":
|
| 270 |
"step": 29
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 0.12096774193548387,
|
| 274 |
-
"grad_norm": 6.
|
| 275 |
"learning_rate": 4.821632498930656e-06,
|
| 276 |
-
"loss": 1.
|
| 277 |
-
"mean_token_accuracy": 0.
|
| 278 |
-
"num_tokens":
|
| 279 |
"step": 30
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 0.125,
|
| 283 |
-
"grad_norm": 6.
|
| 284 |
"learning_rate": 4.809698831278217e-06,
|
| 285 |
-
"loss": 1.
|
| 286 |
-
"mean_token_accuracy": 0.
|
| 287 |
-
"num_tokens":
|
| 288 |
"step": 31
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"epoch": 0.12903225806451613,
|
| 292 |
-
"grad_norm": 6.
|
| 293 |
"learning_rate": 4.797394529050577e-06,
|
| 294 |
-
"loss": 1.
|
| 295 |
-
"mean_token_accuracy": 0.
|
| 296 |
-
"num_tokens":
|
| 297 |
"step": 32
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.13306451612903225,
|
| 301 |
-
"grad_norm": 7.
|
| 302 |
"learning_rate": 4.784721566704217e-06,
|
| 303 |
-
"loss": 1.
|
| 304 |
-
"mean_token_accuracy": 0.
|
| 305 |
-
"num_tokens":
|
| 306 |
"step": 33
|
| 307 |
},
|
| 308 |
{
|
| 309 |
"epoch": 0.13709677419354838,
|
| 310 |
-
"grad_norm": 6.
|
| 311 |
"learning_rate": 4.771681977854062e-06,
|
| 312 |
-
"loss": 1.
|
| 313 |
-
"mean_token_accuracy": 0.
|
| 314 |
-
"num_tokens":
|
| 315 |
"step": 34
|
| 316 |
},
|
| 317 |
{
|
| 318 |
"epoch": 0.14112903225806453,
|
| 319 |
-
"grad_norm": 6.
|
| 320 |
"learning_rate": 4.75827785494715e-06,
|
| 321 |
-
"loss": 1.
|
| 322 |
-
"mean_token_accuracy": 0.
|
| 323 |
-
"num_tokens":
|
| 324 |
"step": 35
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.14516129032258066,
|
| 328 |
-
"grad_norm": 6.
|
| 329 |
"learning_rate": 4.744511348926855e-06,
|
| 330 |
-
"loss": 1.
|
| 331 |
-
"mean_token_accuracy": 0.
|
| 332 |
-
"num_tokens":
|
| 333 |
"step": 36
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 0.14919354838709678,
|
| 337 |
-
"grad_norm": 5.
|
| 338 |
"learning_rate": 4.730384668887731e-06,
|
| 339 |
-
"loss": 1.
|
| 340 |
-
"mean_token_accuracy": 0.
|
| 341 |
-
"num_tokens":
|
| 342 |
"step": 37
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 0.1532258064516129,
|
| 346 |
-
"grad_norm": 6.
|
| 347 |
"learning_rate": 4.715900081721021e-06,
|
| 348 |
-
"loss": 1.
|
| 349 |
-
"mean_token_accuracy": 0.
|
| 350 |
-
"num_tokens":
|
| 351 |
"step": 38
|
| 352 |
},
|
| 353 |
{
|
| 354 |
"epoch": 0.15725806451612903,
|
| 355 |
-
"grad_norm":
|
| 356 |
"learning_rate": 4.7010599117508936e-06,
|
| 357 |
-
"loss": 1.
|
| 358 |
-
"mean_token_accuracy": 0.
|
| 359 |
-
"num_tokens":
|
| 360 |
"step": 39
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.16129032258064516,
|
| 364 |
-
"grad_norm": 6.
|
| 365 |
"learning_rate": 4.685866540361456e-06,
|
| 366 |
-
"loss": 1.
|
| 367 |
-
"mean_token_accuracy": 0.
|
| 368 |
-
"num_tokens":
|
| 369 |
"step": 40
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 0.16532258064516128,
|
| 373 |
-
"grad_norm":
|
| 374 |
"learning_rate": 4.670322405614621e-06,
|
| 375 |
-
"loss": 1.
|
| 376 |
-
"mean_token_accuracy": 0.
|
| 377 |
-
"num_tokens":
|
| 378 |
"step": 41
|
| 379 |
},
|
| 380 |
{
|
| 381 |
"epoch": 0.1693548387096774,
|
| 382 |
-
"grad_norm":
|
| 383 |
"learning_rate": 4.654430001858874e-06,
|
| 384 |
-
"loss": 1.
|
| 385 |
-
"mean_token_accuracy": 0.
|
| 386 |
-
"num_tokens":
|
| 387 |
"step": 42
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.17338709677419356,
|
| 391 |
-
"grad_norm": 6.
|
| 392 |
"learning_rate": 4.638191879329005e-06,
|
| 393 |
-
"loss": 1.
|
| 394 |
-
"mean_token_accuracy": 0.
|
| 395 |
-
"num_tokens":
|
| 396 |
"step": 43
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.1774193548387097,
|
| 400 |
-
"grad_norm": 6.
|
| 401 |
"learning_rate": 4.621610643736878e-06,
|
| 402 |
-
"loss": 1.
|
| 403 |
-
"mean_token_accuracy": 0.
|
| 404 |
-
"num_tokens":
|
| 405 |
"step": 44
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 0.1814516129032258,
|
| 409 |
-
"grad_norm": 11.
|
| 410 |
"learning_rate": 4.6046889558532925e-06,
|
| 411 |
-
"loss": 1.
|
| 412 |
-
"mean_token_accuracy": 0.
|
| 413 |
-
"num_tokens":
|
| 414 |
"step": 45
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 0.18548387096774194,
|
| 418 |
-
"grad_norm": 5.
|
| 419 |
"learning_rate": 4.587429531081019e-06,
|
| 420 |
-
"loss": 1.
|
| 421 |
-
"mean_token_accuracy": 0.
|
| 422 |
-
"num_tokens":
|
| 423 |
"step": 46
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.18951612903225806,
|
| 427 |
-
"grad_norm": 6.
|
| 428 |
"learning_rate": 4.569835139019054e-06,
|
| 429 |
-
"loss": 1.
|
| 430 |
-
"mean_token_accuracy": 0.
|
| 431 |
-
"num_tokens":
|
| 432 |
"step": 47
|
| 433 |
},
|
| 434 |
{
|
| 435 |
"epoch": 0.1935483870967742,
|
| 436 |
-
"grad_norm":
|
| 437 |
"learning_rate": 4.551908603018191e-06,
|
| 438 |
-
"loss": 1.
|
| 439 |
-
"mean_token_accuracy": 0.
|
| 440 |
-
"num_tokens":
|
| 441 |
"step": 48
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 0.1975806451612903,
|
| 445 |
-
"grad_norm": 5.
|
| 446 |
"learning_rate": 4.53365279972796e-06,
|
| 447 |
-
"loss": 1.
|
| 448 |
-
"mean_token_accuracy": 0.
|
| 449 |
-
"num_tokens":
|
| 450 |
"step": 49
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.20161290322580644,
|
| 454 |
-
"grad_norm": 5.
|
| 455 |
"learning_rate": 4.515070658635013e-06,
|
| 456 |
-
"loss": 1.
|
| 457 |
-
"mean_token_accuracy": 0.
|
| 458 |
-
"num_tokens":
|
| 459 |
"step": 50
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.2056451612903226,
|
| 463 |
-
"grad_norm": 6.
|
| 464 |
"learning_rate": 4.4961651615930344e-06,
|
| 465 |
-
"loss": 1.
|
| 466 |
-
"mean_token_accuracy": 0.
|
| 467 |
-
"num_tokens":
|
| 468 |
"step": 51
|
| 469 |
},
|
| 470 |
{
|
| 471 |
"epoch": 0.20967741935483872,
|
| 472 |
-
"grad_norm": 5.
|
| 473 |
"learning_rate": 4.476939342344246e-06,
|
| 474 |
-
"loss": 1.
|
| 475 |
-
"mean_token_accuracy": 0.
|
| 476 |
-
"num_tokens":
|
| 477 |
"step": 52
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 0.21370967741935484,
|
| 481 |
-
"grad_norm":
|
| 482 |
"learning_rate": 4.457396286032589e-06,
|
| 483 |
-
"loss": 1.
|
| 484 |
-
"mean_token_accuracy": 0.
|
| 485 |
-
"num_tokens":
|
| 486 |
"step": 53
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.21774193548387097,
|
| 490 |
-
"grad_norm": 6.
|
| 491 |
"learning_rate": 4.437539128708647e-06,
|
| 492 |
-
"loss": 1.
|
| 493 |
-
"mean_token_accuracy": 0.
|
| 494 |
-
"num_tokens":
|
| 495 |
"step": 54
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 0.2217741935483871,
|
| 499 |
-
"grad_norm": 6.
|
| 500 |
"learning_rate": 4.417371056826417e-06,
|
| 501 |
-
"loss": 1.
|
| 502 |
-
"mean_token_accuracy": 0.
|
| 503 |
-
"num_tokens":
|
| 504 |
"step": 55
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 0.22580645161290322,
|
| 508 |
-
"grad_norm":
|
| 509 |
"learning_rate": 4.396895306731978e-06,
|
| 510 |
-
"loss": 1.
|
| 511 |
-
"mean_token_accuracy": 0.
|
| 512 |
-
"num_tokens":
|
| 513 |
"step": 56
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.22983870967741934,
|
| 517 |
-
"grad_norm": 6.
|
| 518 |
"learning_rate": 4.376115164144157e-06,
|
| 519 |
-
"loss": 1.
|
| 520 |
-
"mean_token_accuracy": 0.
|
| 521 |
-
"num_tokens":
|
| 522 |
"step": 57
|
| 523 |
},
|
| 524 |
{
|
| 525 |
"epoch": 0.23387096774193547,
|
| 526 |
-
"grad_norm": 6.
|
| 527 |
"learning_rate": 4.355033963627277e-06,
|
| 528 |
-
"loss": 1.
|
| 529 |
-
"mean_token_accuracy": 0.
|
| 530 |
-
"num_tokens":
|
| 531 |
"step": 58
|
| 532 |
},
|
| 533 |
{
|
| 534 |
"epoch": 0.23790322580645162,
|
| 535 |
-
"grad_norm": 5.
|
| 536 |
"learning_rate": 4.333655088056065e-06,
|
| 537 |
-
"loss": 1.
|
| 538 |
-
"mean_token_accuracy": 0.
|
| 539 |
-
"num_tokens":
|
| 540 |
"step": 59
|
| 541 |
},
|
| 542 |
{
|
| 543 |
"epoch": 0.24193548387096775,
|
| 544 |
-
"grad_norm": 6.
|
| 545 |
"learning_rate": 4.3119819680728e-06,
|
| 546 |
-
"loss": 1.
|
| 547 |
-
"mean_token_accuracy": 0.
|
| 548 |
-
"num_tokens":
|
| 549 |
"step": 60
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.24596774193548387,
|
| 553 |
-
"grad_norm": 5.
|
| 554 |
"learning_rate": 4.290018081536807e-06,
|
| 555 |
-
"loss": 1.
|
| 556 |
-
"mean_token_accuracy": 0.
|
| 557 |
-
"num_tokens":
|
| 558 |
"step": 61
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 0.25,
|
| 562 |
-
"grad_norm": 6.
|
| 563 |
"learning_rate": 4.267766952966369e-06,
|
| 564 |
-
"loss": 1.
|
| 565 |
-
"mean_token_accuracy": 0.
|
| 566 |
-
"num_tokens":
|
| 567 |
"step": 62
|
| 568 |
},
|
| 569 |
{
|
| 570 |
"epoch": 0.2540322580645161,
|
| 571 |
-
"grad_norm": 6.
|
| 572 |
"learning_rate": 4.245232152973148e-06,
|
| 573 |
-
"loss": 1.
|
| 574 |
-
"mean_token_accuracy": 0.
|
| 575 |
-
"num_tokens":
|
| 576 |
"step": 63
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.25806451612903225,
|
| 580 |
-
"grad_norm": 6.
|
| 581 |
"learning_rate": 4.222417297689217e-06,
|
| 582 |
-
"loss": 1.
|
| 583 |
-
"mean_token_accuracy": 0.
|
| 584 |
-
"num_tokens":
|
| 585 |
"step": 64
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.2620967741935484,
|
| 589 |
-
"grad_norm": 5.
|
| 590 |
"learning_rate": 4.199326048186783e-06,
|
| 591 |
-
"loss": 1.
|
| 592 |
-
"mean_token_accuracy": 0.
|
| 593 |
-
"num_tokens":
|
| 594 |
"step": 65
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 0.2661290322580645,
|
| 598 |
-
"grad_norm": 6.
|
| 599 |
"learning_rate": 4.175962109890697e-06,
|
| 600 |
-
"loss": 1.
|
| 601 |
-
"mean_token_accuracy": 0.
|
| 602 |
-
"num_tokens":
|
| 603 |
"step": 66
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 0.2701612903225806,
|
| 607 |
-
"grad_norm": 7.
|
| 608 |
"learning_rate": 4.152329231983852e-06,
|
| 609 |
-
"loss": 1.
|
| 610 |
-
"mean_token_accuracy": 0.
|
| 611 |
-
"num_tokens":
|
| 612 |
"step": 67
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.27419354838709675,
|
| 616 |
-
"grad_norm": 6.
|
| 617 |
"learning_rate": 4.128431206805556e-06,
|
| 618 |
-
"loss": 1.
|
| 619 |
-
"mean_token_accuracy": 0.
|
| 620 |
-
"num_tokens":
|
| 621 |
"step": 68
|
| 622 |
},
|
| 623 |
{
|
| 624 |
"epoch": 0.2782258064516129,
|
| 625 |
-
"grad_norm": 5.
|
| 626 |
"learning_rate": 4.104271869242975e-06,
|
| 627 |
-
"loss": 1.
|
| 628 |
-
"mean_token_accuracy": 0.
|
| 629 |
-
"num_tokens":
|
| 630 |
"step": 69
|
| 631 |
},
|
| 632 |
{
|
| 633 |
"epoch": 0.28225806451612906,
|
| 634 |
-
"grad_norm": 6.
|
| 635 |
"learning_rate": 4.07985509611576e-06,
|
| 636 |
-
"loss": 1.
|
| 637 |
-
"mean_token_accuracy": 0.
|
| 638 |
-
"num_tokens":
|
| 639 |
"step": 70
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.2862903225806452,
|
| 643 |
-
"grad_norm": 5.
|
| 644 |
"learning_rate": 4.0551848055539345e-06,
|
| 645 |
-
"loss": 1.
|
| 646 |
-
"mean_token_accuracy": 0.
|
| 647 |
-
"num_tokens":
|
| 648 |
"step": 71
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 0.2903225806451613,
|
| 652 |
-
"grad_norm":
|
| 653 |
"learning_rate": 4.030264956369158e-06,
|
| 654 |
-
"loss": 1.
|
| 655 |
-
"mean_token_accuracy": 0.
|
| 656 |
-
"num_tokens":
|
| 657 |
"step": 72
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 0.29435483870967744,
|
| 661 |
-
"grad_norm":
|
| 662 |
"learning_rate": 4.005099547419458e-06,
|
| 663 |
-
"loss": 1.
|
| 664 |
-
"mean_token_accuracy": 0.
|
| 665 |
-
"num_tokens":
|
| 666 |
"step": 73
|
| 667 |
},
|
| 668 |
{
|
| 669 |
"epoch": 0.29838709677419356,
|
| 670 |
-
"grad_norm": 6.
|
| 671 |
"learning_rate": 3.979692616967543e-06,
|
| 672 |
-
"loss": 1.
|
| 673 |
-
"mean_token_accuracy": 0.
|
| 674 |
-
"num_tokens":
|
| 675 |
"step": 74
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.3024193548387097,
|
| 679 |
-
"grad_norm": 6.
|
| 680 |
"learning_rate": 3.9540482420327845e-06,
|
| 681 |
-
"loss": 1.
|
| 682 |
-
"mean_token_accuracy": 0.
|
| 683 |
-
"num_tokens":
|
| 684 |
"step": 75
|
| 685 |
},
|
| 686 |
{
|
| 687 |
"epoch": 0.3064516129032258,
|
| 688 |
-
"grad_norm":
|
| 689 |
"learning_rate": 3.9281705377369814e-06,
|
| 690 |
-
"loss": 1.
|
| 691 |
-
"mean_token_accuracy": 0.
|
| 692 |
-
"num_tokens":
|
| 693 |
"step": 76
|
| 694 |
},
|
| 695 |
{
|
| 696 |
"epoch": 0.31048387096774194,
|
| 697 |
-
"grad_norm": 6.
|
| 698 |
"learning_rate": 3.902063656644012e-06,
|
| 699 |
-
"loss": 1.
|
| 700 |
-
"mean_token_accuracy": 0.
|
| 701 |
-
"num_tokens":
|
| 702 |
"step": 77
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.31451612903225806,
|
| 706 |
-
"grad_norm": 6.
|
| 707 |
"learning_rate": 3.875731788093478e-06,
|
| 708 |
-
"loss": 1.
|
| 709 |
-
"mean_token_accuracy": 0.
|
| 710 |
-
"num_tokens":
|
| 711 |
"step": 78
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 0.3185483870967742,
|
| 715 |
-
"grad_norm": 6.
|
| 716 |
"learning_rate": 3.84917915752845e-06,
|
| 717 |
-
"loss": 1.
|
| 718 |
-
"mean_token_accuracy": 0.
|
| 719 |
-
"num_tokens":
|
| 720 |
"step": 79
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 0.3225806451612903,
|
| 724 |
-
"grad_norm": 5.
|
| 725 |
"learning_rate": 3.8224100258174066e-06,
|
| 726 |
-
"loss": 1.
|
| 727 |
-
"mean_token_accuracy": 0.
|
| 728 |
-
"num_tokens":
|
| 729 |
"step": 80
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.32661290322580644,
|
| 733 |
-
"grad_norm": 6.
|
| 734 |
"learning_rate": 3.795428688570505e-06,
|
| 735 |
-
"loss": 1.
|
| 736 |
-
"mean_token_accuracy": 0.
|
| 737 |
-
"num_tokens":
|
| 738 |
"step": 81
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.33064516129032256,
|
| 742 |
-
"grad_norm": 6.
|
| 743 |
"learning_rate": 3.7682394754502687e-06,
|
| 744 |
-
"loss": 1.
|
| 745 |
-
"mean_token_accuracy": 0.
|
| 746 |
-
"num_tokens":
|
| 747 |
"step": 82
|
| 748 |
},
|
| 749 |
{
|
| 750 |
"epoch": 0.3346774193548387,
|
| 751 |
-
"grad_norm": 6.
|
| 752 |
"learning_rate": 3.7408467494768104e-06,
|
| 753 |
-
"loss": 1.
|
| 754 |
-
"mean_token_accuracy": 0.
|
| 755 |
-
"num_tokens":
|
| 756 |
"step": 83
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"epoch": 0.3387096774193548,
|
| 760 |
-
"grad_norm":
|
| 761 |
"learning_rate": 3.7132549063277033e-06,
|
| 762 |
-
"loss": 1.
|
| 763 |
-
"mean_token_accuracy": 0.
|
| 764 |
-
"num_tokens":
|
| 765 |
"step": 84
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.34274193548387094,
|
| 769 |
-
"grad_norm": 6.
|
| 770 |
"learning_rate": 3.685468373632613e-06,
|
| 771 |
-
"loss": 1.
|
| 772 |
-
"mean_token_accuracy": 0.
|
| 773 |
-
"num_tokens":
|
| 774 |
"step": 85
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 0.3467741935483871,
|
| 778 |
-
"grad_norm": 5.
|
| 779 |
"learning_rate": 3.657491610262802e-06,
|
| 780 |
-
"loss": 1.
|
| 781 |
-
"mean_token_accuracy": 0.
|
| 782 |
-
"num_tokens":
|
| 783 |
"step": 86
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 0.35080645161290325,
|
| 787 |
-
"grad_norm": 5.
|
| 788 |
"learning_rate": 3.6293291056156178e-06,
|
| 789 |
-
"loss": 1.
|
| 790 |
-
"mean_token_accuracy": 0.
|
| 791 |
-
"num_tokens":
|
| 792 |
"step": 87
|
| 793 |
},
|
| 794 |
{
|
| 795 |
"epoch": 0.3548387096774194,
|
| 796 |
-
"grad_norm": 5.
|
| 797 |
"learning_rate": 3.600985378894086e-06,
|
| 798 |
-
"loss": 1.
|
| 799 |
-
"mean_token_accuracy": 0.
|
| 800 |
-
"num_tokens":
|
| 801 |
"step": 88
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.3588709677419355,
|
| 805 |
-
"grad_norm": 5.
|
| 806 |
"learning_rate": 3.572464978381719e-06,
|
| 807 |
-
"loss": 1.
|
| 808 |
-
"mean_token_accuracy": 0.
|
| 809 |
-
"num_tokens":
|
| 810 |
"step": 89
|
| 811 |
},
|
| 812 |
{
|
| 813 |
"epoch": 0.3629032258064516,
|
| 814 |
-
"grad_norm": 6.
|
| 815 |
"learning_rate": 3.5437724807126583e-06,
|
| 816 |
-
"loss": 1.
|
| 817 |
-
"mean_token_accuracy": 0.
|
| 818 |
-
"num_tokens":
|
| 819 |
"step": 90
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.36693548387096775,
|
| 823 |
-
"grad_norm": 5.
|
| 824 |
"learning_rate": 3.514912490137268e-06,
|
| 825 |
-
"loss": 1.
|
| 826 |
-
"mean_token_accuracy": 0.
|
| 827 |
-
"num_tokens":
|
| 828 |
"step": 91
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.3709677419354839,
|
| 832 |
-
"grad_norm": 6.
|
| 833 |
"learning_rate": 3.4858896377832966e-06,
|
| 834 |
-
"loss": 1.
|
| 835 |
-
"mean_token_accuracy": 0.
|
| 836 |
-
"num_tokens":
|
| 837 |
"step": 92
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 0.375,
|
| 841 |
-
"grad_norm": 5.
|
| 842 |
"learning_rate": 3.4567085809127247e-06,
|
| 843 |
-
"loss": 1.
|
| 844 |
-
"mean_token_accuracy": 0.
|
| 845 |
-
"num_tokens":
|
| 846 |
"step": 93
|
| 847 |
},
|
| 848 |
{
|
| 849 |
"epoch": 0.3790322580645161,
|
| 850 |
-
"grad_norm":
|
| 851 |
"learning_rate": 3.42737400217442e-06,
|
| 852 |
-
"loss": 1.
|
| 853 |
-
"mean_token_accuracy": 0.
|
| 854 |
-
"num_tokens":
|
| 855 |
"step": 94
|
| 856 |
},
|
| 857 |
{
|
| 858 |
"epoch": 0.38306451612903225,
|
| 859 |
-
"grad_norm": 6.
|
| 860 |
"learning_rate": 3.397890608852718e-06,
|
| 861 |
-
"loss": 1.
|
| 862 |
-
"mean_token_accuracy": 0.
|
| 863 |
-
"num_tokens":
|
| 864 |
"step": 95
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.3870967741935484,
|
| 868 |
-
"grad_norm": 6.
|
| 869 |
"learning_rate": 3.3682631321120507e-06,
|
| 870 |
-
"loss": 1.
|
| 871 |
-
"mean_token_accuracy": 0.
|
| 872 |
-
"num_tokens":
|
| 873 |
"step": 96
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.3911290322580645,
|
| 877 |
-
"grad_norm": 6.
|
| 878 |
"learning_rate": 3.3384963262377434e-06,
|
| 879 |
-
"loss": 1.
|
| 880 |
-
"mean_token_accuracy": 0.
|
| 881 |
-
"num_tokens":
|
| 882 |
"step": 97
|
| 883 |
},
|
| 884 |
{
|
| 885 |
"epoch": 0.3951612903225806,
|
| 886 |
-
"grad_norm": 5.
|
| 887 |
"learning_rate": 3.3085949678730953e-06,
|
| 888 |
-
"loss": 1.
|
| 889 |
-
"mean_token_accuracy": 0.
|
| 890 |
-
"num_tokens":
|
| 891 |
"step": 98
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.39919354838709675,
|
| 895 |
-
"grad_norm": 6.
|
| 896 |
"learning_rate": 3.278563855252885e-06,
|
| 897 |
-
"loss": 1.
|
| 898 |
-
"mean_token_accuracy": 0.
|
| 899 |
-
"num_tokens":
|
| 900 |
"step": 99
|
| 901 |
},
|
| 902 |
{
|
| 903 |
"epoch": 0.4032258064516129,
|
| 904 |
-
"grad_norm": 6.
|
| 905 |
"learning_rate": 3.248407807433396e-06,
|
| 906 |
-
"loss": 1.
|
| 907 |
-
"mean_token_accuracy": 0.
|
| 908 |
-
"num_tokens":
|
| 909 |
"step": 100
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 0.40725806451612906,
|
| 913 |
-
"grad_norm": 6.
|
| 914 |
"learning_rate": 3.2181316635191125e-06,
|
| 915 |
-
"loss": 1.
|
| 916 |
-
"mean_token_accuracy": 0.
|
| 917 |
-
"num_tokens":
|
| 918 |
"step": 101
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 0.4112903225806452,
|
| 922 |
-
"grad_norm": 5.
|
| 923 |
"learning_rate": 3.1877402818861954e-06,
|
| 924 |
-
"loss": 1.
|
| 925 |
-
"mean_token_accuracy": 0.
|
| 926 |
-
"num_tokens":
|
| 927 |
"step": 102
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.4153225806451613,
|
| 931 |
-
"grad_norm": 5.
|
| 932 |
"learning_rate": 3.157238539402862e-06,
|
| 933 |
-
"loss": 1.
|
| 934 |
-
"mean_token_accuracy": 0.
|
| 935 |
-
"num_tokens":
|
| 936 |
"step": 103
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 0.41935483870967744,
|
| 940 |
-
"grad_norm": 6.
|
| 941 |
"learning_rate": 3.1266313306468018e-06,
|
| 942 |
-
"loss": 1.
|
| 943 |
-
"mean_token_accuracy": 0.
|
| 944 |
-
"num_tokens":
|
| 945 |
"step": 104
|
| 946 |
},
|
| 947 |
{
|
| 948 |
"epoch": 0.42338709677419356,
|
| 949 |
-
"grad_norm":
|
| 950 |
"learning_rate": 3.095923567119748e-06,
|
| 951 |
-
"loss": 1.
|
| 952 |
-
"mean_token_accuracy": 0.
|
| 953 |
-
"num_tokens":
|
| 954 |
"step": 105
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.4274193548387097,
|
| 958 |
-
"grad_norm":
|
| 959 |
"learning_rate": 3.0651201764593375e-06,
|
| 960 |
-
"loss": 1.
|
| 961 |
-
"mean_token_accuracy": 0.
|
| 962 |
-
"num_tokens":
|
| 963 |
"step": 106
|
| 964 |
},
|
| 965 |
{
|
| 966 |
"epoch": 0.4314516129032258,
|
| 967 |
-
"grad_norm": 5.
|
| 968 |
"learning_rate": 3.034226101648377e-06,
|
| 969 |
-
"loss": 1.
|
| 970 |
-
"mean_token_accuracy": 0.
|
| 971 |
-
"num_tokens":
|
| 972 |
"step": 107
|
| 973 |
},
|
| 974 |
{
|
| 975 |
"epoch": 0.43548387096774194,
|
| 976 |
-
"grad_norm": 5.
|
| 977 |
"learning_rate": 3.0032463002216504e-06,
|
| 978 |
-
"loss": 1.
|
| 979 |
-
"mean_token_accuracy": 0.
|
| 980 |
-
"num_tokens":
|
| 981 |
"step": 108
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 0.43951612903225806,
|
| 985 |
-
"grad_norm": 5.
|
| 986 |
"learning_rate": 2.972185743470386e-06,
|
| 987 |
-
"loss": 1.
|
| 988 |
-
"mean_token_accuracy": 0.
|
| 989 |
-
"num_tokens":
|
| 990 |
"step": 109
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.4435483870967742,
|
| 994 |
-
"grad_norm": 6.
|
| 995 |
"learning_rate": 2.941049415644522e-06,
|
| 996 |
-
"loss": 1.
|
| 997 |
-
"mean_token_accuracy": 0.
|
| 998 |
-
"num_tokens":
|
| 999 |
"step": 110
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 0.4475806451612903,
|
| 1003 |
-
"grad_norm":
|
| 1004 |
"learning_rate": 2.909842313152888e-06,
|
| 1005 |
-
"loss": 1.
|
| 1006 |
-
"mean_token_accuracy": 0.
|
| 1007 |
-
"num_tokens":
|
| 1008 |
"step": 111
|
| 1009 |
},
|
| 1010 |
{
|
| 1011 |
"epoch": 0.45161290322580644,
|
| 1012 |
-
"grad_norm": 5.
|
| 1013 |
"learning_rate": 2.878569443761442e-06,
|
| 1014 |
-
"loss": 1.
|
| 1015 |
-
"mean_token_accuracy": 0.
|
| 1016 |
-
"num_tokens":
|
| 1017 |
"step": 112
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.45564516129032256,
|
| 1021 |
-
"grad_norm":
|
| 1022 |
"learning_rate": 2.847235825789673e-06,
|
| 1023 |
-
"loss": 1.
|
| 1024 |
-
"mean_token_accuracy": 0.
|
| 1025 |
-
"num_tokens":
|
| 1026 |
"step": 113
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
"epoch": 0.4596774193548387,
|
| 1030 |
-
"grad_norm": 6.
|
| 1031 |
"learning_rate": 2.8158464873053236e-06,
|
| 1032 |
-
"loss": 1.
|
| 1033 |
-
"mean_token_accuracy": 0.
|
| 1034 |
-
"num_tokens":
|
| 1035 |
"step": 114
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
"epoch": 0.4637096774193548,
|
| 1039 |
-
"grad_norm": 6.
|
| 1040 |
"learning_rate": 2.784406465317538e-06,
|
| 1041 |
-
"loss": 1.
|
| 1042 |
-
"mean_token_accuracy": 0.
|
| 1043 |
-
"num_tokens":
|
| 1044 |
"step": 115
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.46774193548387094,
|
| 1048 |
-
"grad_norm": 5.
|
| 1049 |
"learning_rate": 2.752920804968581e-06,
|
| 1050 |
-
"loss": 1.
|
| 1051 |
-
"mean_token_accuracy": 0.
|
| 1052 |
-
"num_tokens":
|
| 1053 |
"step": 116
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
"epoch": 0.4717741935483871,
|
| 1057 |
-
"grad_norm": 5.
|
| 1058 |
"learning_rate": 2.7213945587242507e-06,
|
| 1059 |
-
"loss": 1.
|
| 1060 |
-
"mean_token_accuracy": 0.
|
| 1061 |
-
"num_tokens":
|
| 1062 |
"step": 117
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 0.47580645161290325,
|
| 1066 |
-
"grad_norm": 5.
|
| 1067 |
"learning_rate": 2.689832785563116e-06,
|
| 1068 |
-
"loss": 1.
|
| 1069 |
-
"mean_token_accuracy": 0.
|
| 1070 |
-
"num_tokens":
|
| 1071 |
"step": 118
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
"epoch": 0.4798387096774194,
|
| 1075 |
-
"grad_norm": 5.
|
| 1076 |
"learning_rate": 2.658240550164704e-06,
|
| 1077 |
-
"loss": 1.
|
| 1078 |
-
"mean_token_accuracy": 0.
|
| 1079 |
-
"num_tokens":
|
| 1080 |
"step": 119
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.4838709677419355,
|
| 1084 |
-
"grad_norm":
|
| 1085 |
"learning_rate": 2.626622922096782e-06,
|
| 1086 |
-
"loss": 1.
|
| 1087 |
-
"mean_token_accuracy": 0.
|
| 1088 |
-
"num_tokens":
|
| 1089 |
"step": 120
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 0.4879032258064516,
|
| 1093 |
-
"grad_norm": 5.
|
| 1094 |
"learning_rate": 2.5949849750018486e-06,
|
| 1095 |
-
"loss": 1.
|
| 1096 |
-
"mean_token_accuracy": 0.
|
| 1097 |
-
"num_tokens":
|
| 1098 |
"step": 121
|
| 1099 |
},
|
| 1100 |
{
|
| 1101 |
"epoch": 0.49193548387096775,
|
| 1102 |
-
"grad_norm": 5.
|
| 1103 |
"learning_rate": 2.56333178578297e-06,
|
| 1104 |
-
"loss": 1.
|
| 1105 |
-
"mean_token_accuracy": 0.
|
| 1106 |
-
"num_tokens":
|
| 1107 |
"step": 122
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
"epoch": 0.4959677419354839,
|
| 1111 |
-
"grad_norm": 5.
|
| 1112 |
"learning_rate": 2.5316684337891005e-06,
|
| 1113 |
-
"loss": 1.
|
| 1114 |
-
"mean_token_accuracy": 0.
|
| 1115 |
-
"num_tokens":
|
| 1116 |
"step": 123
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 0.5,
|
| 1120 |
-
"grad_norm": 6.
|
| 1121 |
"learning_rate": 2.5e-06,
|
| 1122 |
-
"loss": 1.
|
| 1123 |
-
"mean_token_accuracy": 0.
|
| 1124 |
-
"num_tokens":
|
| 1125 |
"step": 124
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
"epoch": 0.5040322580645161,
|
| 1129 |
-
"grad_norm": 5.
|
| 1130 |
"learning_rate": 2.4683315662109003e-06,
|
| 1131 |
-
"loss": 1.
|
| 1132 |
-
"mean_token_accuracy": 0.
|
| 1133 |
-
"num_tokens":
|
| 1134 |
"step": 125
|
| 1135 |
},
|
| 1136 |
{
|
| 1137 |
"epoch": 0.5080645161290323,
|
| 1138 |
-
"grad_norm": 5.
|
| 1139 |
"learning_rate": 2.436668214217031e-06,
|
| 1140 |
-
"loss": 1.
|
| 1141 |
-
"mean_token_accuracy": 0.
|
| 1142 |
-
"num_tokens":
|
| 1143 |
"step": 126
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.5120967741935484,
|
| 1147 |
-
"grad_norm": 5.
|
| 1148 |
"learning_rate": 2.4050150249981522e-06,
|
| 1149 |
-
"loss": 1.
|
| 1150 |
-
"mean_token_accuracy": 0.
|
| 1151 |
-
"num_tokens":
|
| 1152 |
"step": 127
|
| 1153 |
},
|
| 1154 |
{
|
| 1155 |
"epoch": 0.5161290322580645,
|
| 1156 |
-
"grad_norm": 5.
|
| 1157 |
"learning_rate": 2.3733770779032185e-06,
|
| 1158 |
-
"loss": 1.
|
| 1159 |
-
"mean_token_accuracy": 0.
|
| 1160 |
-
"num_tokens":
|
| 1161 |
"step": 128
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.5201612903225806,
|
| 1165 |
-
"grad_norm": 5.
|
| 1166 |
"learning_rate": 2.341759449835297e-06,
|
| 1167 |
-
"loss": 1.
|
| 1168 |
-
"mean_token_accuracy": 0.
|
| 1169 |
-
"num_tokens":
|
| 1170 |
"step": 129
|
| 1171 |
},
|
| 1172 |
{
|
| 1173 |
"epoch": 0.5241935483870968,
|
| 1174 |
-
"grad_norm": 5.
|
| 1175 |
"learning_rate": 2.310167214436885e-06,
|
| 1176 |
-
"loss": 1.
|
| 1177 |
-
"mean_token_accuracy": 0.
|
| 1178 |
-
"num_tokens":
|
| 1179 |
"step": 130
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.5282258064516129,
|
| 1183 |
-
"grad_norm": 5.
|
| 1184 |
"learning_rate": 2.27860544127575e-06,
|
| 1185 |
-
"loss": 1.
|
| 1186 |
-
"mean_token_accuracy": 0.
|
| 1187 |
-
"num_tokens":
|
| 1188 |
"step": 131
|
| 1189 |
},
|
| 1190 |
{
|
| 1191 |
"epoch": 0.532258064516129,
|
| 1192 |
-
"grad_norm": 5.
|
| 1193 |
"learning_rate": 2.24707919503142e-06,
|
| 1194 |
-
"loss": 1.
|
| 1195 |
-
"mean_token_accuracy": 0.
|
| 1196 |
-
"num_tokens":
|
| 1197 |
"step": 132
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
"epoch": 0.5362903225806451,
|
| 1201 |
-
"grad_norm": 5.
|
| 1202 |
"learning_rate": 2.2155935346824634e-06,
|
| 1203 |
-
"loss": 1.
|
| 1204 |
-
"mean_token_accuracy": 0.
|
| 1205 |
-
"num_tokens":
|
| 1206 |
"step": 133
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.5403225806451613,
|
| 1210 |
-
"grad_norm": 6.
|
| 1211 |
"learning_rate": 2.1841535126946777e-06,
|
| 1212 |
-
"loss": 1.
|
| 1213 |
-
"mean_token_accuracy": 0.
|
| 1214 |
-
"num_tokens":
|
| 1215 |
"step": 134
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
"epoch": 0.5443548387096774,
|
| 1219 |
-
"grad_norm": 5.
|
| 1220 |
"learning_rate": 2.1527641742103282e-06,
|
| 1221 |
-
"loss": 1.
|
| 1222 |
-
"mean_token_accuracy": 0.
|
| 1223 |
-
"num_tokens":
|
| 1224 |
"step": 135
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 0.5483870967741935,
|
| 1228 |
-
"grad_norm": 6.
|
| 1229 |
"learning_rate": 2.1214305562385592e-06,
|
| 1230 |
-
"loss": 1.
|
| 1231 |
-
"mean_token_accuracy": 0.
|
| 1232 |
-
"num_tokens":
|
| 1233 |
"step": 136
|
| 1234 |
},
|
| 1235 |
{
|
| 1236 |
"epoch": 0.5524193548387096,
|
| 1237 |
-
"grad_norm": 5.
|
| 1238 |
"learning_rate": 2.0901576868471125e-06,
|
| 1239 |
-
"loss": 1.
|
| 1240 |
-
"mean_token_accuracy": 0.
|
| 1241 |
-
"num_tokens":
|
| 1242 |
"step": 137
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.5564516129032258,
|
| 1246 |
-
"grad_norm": 5.
|
| 1247 |
"learning_rate": 2.05895058435548e-06,
|
| 1248 |
-
"loss": 1.
|
| 1249 |
-
"mean_token_accuracy": 0.
|
| 1250 |
-
"num_tokens":
|
| 1251 |
"step": 138
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
"epoch": 0.5604838709677419,
|
| 1255 |
-
"grad_norm": 5.
|
| 1256 |
"learning_rate": 2.0278142565296153e-06,
|
| 1257 |
-
"loss": 1.
|
| 1258 |
-
"mean_token_accuracy": 0.
|
| 1259 |
-
"num_tokens":
|
| 1260 |
"step": 139
|
| 1261 |
},
|
| 1262 |
{
|
| 1263 |
"epoch": 0.5645161290322581,
|
| 1264 |
-
"grad_norm": 5.
|
| 1265 |
"learning_rate": 1.9967536997783495e-06,
|
| 1266 |
-
"loss": 1.
|
| 1267 |
-
"mean_token_accuracy": 0.
|
| 1268 |
-
"num_tokens":
|
| 1269 |
"step": 140
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.5685483870967742,
|
| 1273 |
-
"grad_norm":
|
| 1274 |
"learning_rate": 1.9657738983516227e-06,
|
| 1275 |
-
"loss": 1.
|
| 1276 |
-
"mean_token_accuracy": 0.
|
| 1277 |
-
"num_tokens":
|
| 1278 |
"step": 141
|
| 1279 |
},
|
| 1280 |
{
|
| 1281 |
"epoch": 0.5725806451612904,
|
| 1282 |
-
"grad_norm": 5.
|
| 1283 |
"learning_rate": 1.934879823540663e-06,
|
| 1284 |
-
"loss": 1.
|
| 1285 |
-
"mean_token_accuracy": 0.
|
| 1286 |
-
"num_tokens":
|
| 1287 |
"step": 142
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 0.5766129032258065,
|
| 1291 |
-
"grad_norm": 5.
|
| 1292 |
"learning_rate": 1.9040764328802523e-06,
|
| 1293 |
-
"loss": 1.
|
| 1294 |
-
"mean_token_accuracy": 0.
|
| 1295 |
-
"num_tokens":
|
| 1296 |
"step": 143
|
| 1297 |
},
|
| 1298 |
{
|
| 1299 |
"epoch": 0.5806451612903226,
|
| 1300 |
-
"grad_norm": 5.
|
| 1301 |
"learning_rate": 1.8733686693531986e-06,
|
| 1302 |
-
"loss": 1.
|
| 1303 |
-
"mean_token_accuracy": 0.
|
| 1304 |
-
"num_tokens":
|
| 1305 |
"step": 144
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.5846774193548387,
|
| 1309 |
-
"grad_norm":
|
| 1310 |
"learning_rate": 1.842761460597138e-06,
|
| 1311 |
-
"loss": 1.
|
| 1312 |
-
"mean_token_accuracy": 0.
|
| 1313 |
-
"num_tokens":
|
| 1314 |
"step": 145
|
| 1315 |
},
|
| 1316 |
{
|
| 1317 |
"epoch": 0.5887096774193549,
|
| 1318 |
-
"grad_norm": 7.
|
| 1319 |
"learning_rate": 1.812259718113805e-06,
|
| 1320 |
-
"loss": 1.
|
| 1321 |
-
"mean_token_accuracy": 0.
|
| 1322 |
-
"num_tokens":
|
| 1323 |
"step": 146
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
"epoch": 0.592741935483871,
|
| 1327 |
-
"grad_norm": 6.
|
| 1328 |
"learning_rate": 1.7818683364808883e-06,
|
| 1329 |
-
"loss": 1.
|
| 1330 |
-
"mean_token_accuracy": 0.
|
| 1331 |
-
"num_tokens":
|
| 1332 |
"step": 147
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
"epoch": 0.5967741935483871,
|
| 1336 |
-
"grad_norm": 6.
|
| 1337 |
"learning_rate": 1.7515921925666053e-06,
|
| 1338 |
-
"loss": 1.
|
| 1339 |
-
"mean_token_accuracy": 0.
|
| 1340 |
-
"num_tokens":
|
| 1341 |
"step": 148
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
"epoch": 0.6008064516129032,
|
| 1345 |
-
"grad_norm":
|
| 1346 |
"learning_rate": 1.7214361447471156e-06,
|
| 1347 |
-
"loss": 1.
|
| 1348 |
-
"mean_token_accuracy": 0.
|
| 1349 |
-
"num_tokens":
|
| 1350 |
"step": 149
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 0.6048387096774194,
|
| 1354 |
-
"grad_norm": 5.
|
| 1355 |
"learning_rate": 1.6914050321269049e-06,
|
| 1356 |
-
"loss": 1.
|
| 1357 |
-
"mean_token_accuracy": 0.
|
| 1358 |
-
"num_tokens":
|
| 1359 |
"step": 150
|
| 1360 |
},
|
| 1361 |
{
|
| 1362 |
"epoch": 0.6088709677419355,
|
| 1363 |
-
"grad_norm": 6.
|
| 1364 |
"learning_rate": 1.6615036737622574e-06,
|
| 1365 |
-
"loss": 1.
|
| 1366 |
-
"mean_token_accuracy": 0.
|
| 1367 |
-
"num_tokens":
|
| 1368 |
"step": 151
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 0.6129032258064516,
|
| 1372 |
-
"grad_norm":
|
| 1373 |
"learning_rate": 1.6317368678879497e-06,
|
| 1374 |
-
"loss": 1.
|
| 1375 |
-
"mean_token_accuracy": 0.
|
| 1376 |
-
"num_tokens":
|
| 1377 |
"step": 152
|
| 1378 |
},
|
| 1379 |
{
|
| 1380 |
"epoch": 0.6169354838709677,
|
| 1381 |
-
"grad_norm":
|
| 1382 |
"learning_rate": 1.6021093911472825e-06,
|
| 1383 |
-
"loss": 1.
|
| 1384 |
-
"mean_token_accuracy": 0.
|
| 1385 |
-
"num_tokens":
|
| 1386 |
"step": 153
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
"epoch": 0.6209677419354839,
|
| 1390 |
-
"grad_norm": 5.
|
| 1391 |
"learning_rate": 1.572625997825581e-06,
|
| 1392 |
-
"loss": 1.
|
| 1393 |
-
"mean_token_accuracy": 0.
|
| 1394 |
-
"num_tokens":
|
| 1395 |
"step": 154
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
"epoch": 0.625,
|
| 1399 |
-
"grad_norm": 5.
|
| 1400 |
"learning_rate": 1.5432914190872757e-06,
|
| 1401 |
-
"loss": 1.
|
| 1402 |
-
"mean_token_accuracy": 0.
|
| 1403 |
-
"num_tokens":
|
| 1404 |
"step": 155
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 0.6290322580645161,
|
| 1408 |
-
"grad_norm": 5.
|
| 1409 |
"learning_rate": 1.5141103622167042e-06,
|
| 1410 |
-
"loss": 1.
|
| 1411 |
-
"mean_token_accuracy": 0.
|
| 1412 |
-
"num_tokens":
|
| 1413 |
"step": 156
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 0.6330645161290323,
|
| 1417 |
-
"grad_norm": 6.
|
| 1418 |
"learning_rate": 1.4850875098627326e-06,
|
| 1419 |
-
"loss": 1.
|
| 1420 |
-
"mean_token_accuracy": 0.
|
| 1421 |
-
"num_tokens":
|
| 1422 |
"step": 157
|
| 1423 |
},
|
| 1424 |
{
|
| 1425 |
"epoch": 0.6370967741935484,
|
| 1426 |
-
"grad_norm": 5.
|
| 1427 |
"learning_rate": 1.456227519287343e-06,
|
| 1428 |
-
"loss": 1.
|
| 1429 |
-
"mean_token_accuracy": 0.
|
| 1430 |
-
"num_tokens":
|
| 1431 |
"step": 158
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 0.6411290322580645,
|
| 1435 |
-
"grad_norm":
|
| 1436 |
"learning_rate": 1.4275350216182824e-06,
|
| 1437 |
-
"loss": 1.
|
| 1438 |
-
"mean_token_accuracy": 0.
|
| 1439 |
-
"num_tokens":
|
| 1440 |
"step": 159
|
| 1441 |
},
|
| 1442 |
{
|
| 1443 |
"epoch": 0.6451612903225806,
|
| 1444 |
-
"grad_norm": 5.
|
| 1445 |
"learning_rate": 1.3990146211059141e-06,
|
| 1446 |
-
"loss": 1.
|
| 1447 |
-
"mean_token_accuracy": 0.
|
| 1448 |
-
"num_tokens":
|
| 1449 |
"step": 160
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.6491935483870968,
|
| 1453 |
-
"grad_norm": 5.
|
| 1454 |
"learning_rate": 1.3706708943843822e-06,
|
| 1455 |
-
"loss": 1.
|
| 1456 |
-
"mean_token_accuracy": 0.
|
| 1457 |
-
"num_tokens":
|
| 1458 |
"step": 161
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 0.6532258064516129,
|
| 1462 |
-
"grad_norm":
|
| 1463 |
"learning_rate": 1.3425083897371983e-06,
|
| 1464 |
-
"loss": 1.
|
| 1465 |
-
"mean_token_accuracy": 0.
|
| 1466 |
-
"num_tokens":
|
| 1467 |
"step": 162
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.657258064516129,
|
| 1471 |
-
"grad_norm": 5.
|
| 1472 |
"learning_rate": 1.3145316263673874e-06,
|
| 1473 |
-
"loss": 1.
|
| 1474 |
-
"mean_token_accuracy": 0.
|
| 1475 |
-
"num_tokens":
|
| 1476 |
"step": 163
|
| 1477 |
},
|
| 1478 |
{
|
| 1479 |
"epoch": 0.6612903225806451,
|
| 1480 |
-
"grad_norm":
|
| 1481 |
"learning_rate": 1.286745093672298e-06,
|
| 1482 |
-
"loss": 1.
|
| 1483 |
-
"mean_token_accuracy": 0.
|
| 1484 |
-
"num_tokens":
|
| 1485 |
"step": 164
|
| 1486 |
},
|
| 1487 |
{
|
| 1488 |
"epoch": 0.6653225806451613,
|
| 1489 |
-
"grad_norm": 5.
|
| 1490 |
"learning_rate": 1.2591532505231906e-06,
|
| 1491 |
-
"loss": 1.
|
| 1492 |
-
"mean_token_accuracy": 0.
|
| 1493 |
-
"num_tokens":
|
| 1494 |
"step": 165
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
"epoch": 0.6693548387096774,
|
| 1498 |
-
"grad_norm": 5.
|
| 1499 |
"learning_rate": 1.2317605245497324e-06,
|
| 1500 |
-
"loss": 1.
|
| 1501 |
-
"mean_token_accuracy": 0.
|
| 1502 |
-
"num_tokens":
|
| 1503 |
"step": 166
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 0.6733870967741935,
|
| 1507 |
-
"grad_norm": 5.
|
| 1508 |
"learning_rate": 1.204571311429496e-06,
|
| 1509 |
-
"loss": 1.
|
| 1510 |
-
"mean_token_accuracy": 0.
|
| 1511 |
-
"num_tokens":
|
| 1512 |
"step": 167
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 0.6774193548387096,
|
| 1516 |
-
"grad_norm": 6.
|
| 1517 |
"learning_rate": 1.1775899741825947e-06,
|
| 1518 |
-
"loss": 1.
|
| 1519 |
-
"mean_token_accuracy": 0.
|
| 1520 |
-
"num_tokens":
|
| 1521 |
"step": 168
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 0.6814516129032258,
|
| 1525 |
-
"grad_norm": 6.
|
| 1526 |
"learning_rate": 1.1508208424715511e-06,
|
| 1527 |
-
"loss": 1.
|
| 1528 |
-
"mean_token_accuracy": 0.
|
| 1529 |
-
"num_tokens":
|
| 1530 |
"step": 169
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 0.6854838709677419,
|
| 1534 |
-
"grad_norm": 5.
|
| 1535 |
"learning_rate": 1.1242682119065217e-06,
|
| 1536 |
-
"loss": 1.
|
| 1537 |
-
"mean_token_accuracy": 0.
|
| 1538 |
-
"num_tokens":
|
| 1539 |
"step": 170
|
| 1540 |
},
|
| 1541 |
{
|
| 1542 |
"epoch": 0.6895161290322581,
|
| 1543 |
-
"grad_norm":
|
| 1544 |
"learning_rate": 1.0979363433559892e-06,
|
| 1545 |
-
"loss": 1.
|
| 1546 |
-
"mean_token_accuracy": 0.
|
| 1547 |
-
"num_tokens":
|
| 1548 |
"step": 171
|
| 1549 |
},
|
| 1550 |
{
|
| 1551 |
"epoch": 0.6935483870967742,
|
| 1552 |
-
"grad_norm": 6.
|
| 1553 |
"learning_rate": 1.0718294622630188e-06,
|
| 1554 |
-
"loss": 1.
|
| 1555 |
-
"mean_token_accuracy": 0.
|
| 1556 |
-
"num_tokens":
|
| 1557 |
"step": 172
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
"epoch": 0.6975806451612904,
|
| 1561 |
-
"grad_norm": 5.
|
| 1562 |
"learning_rate": 1.045951757967215e-06,
|
| 1563 |
-
"loss": 1.
|
| 1564 |
-
"mean_token_accuracy": 0.
|
| 1565 |
-
"num_tokens":
|
| 1566 |
"step": 173
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 0.7016129032258065,
|
| 1570 |
-
"grad_norm": 6.
|
| 1571 |
"learning_rate": 1.0203073830324566e-06,
|
| 1572 |
-
"loss": 1.
|
| 1573 |
-
"mean_token_accuracy": 0.
|
| 1574 |
-
"num_tokens":
|
| 1575 |
"step": 174
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 0.7056451612903226,
|
| 1579 |
-
"grad_norm": 5.
|
| 1580 |
"learning_rate": 9.949004525805423e-07,
|
| 1581 |
-
"loss": 1.
|
| 1582 |
-
"mean_token_accuracy": 0.
|
| 1583 |
-
"num_tokens":
|
| 1584 |
"step": 175
|
| 1585 |
},
|
| 1586 |
{
|
| 1587 |
"epoch": 0.7096774193548387,
|
| 1588 |
-
"grad_norm": 6.
|
| 1589 |
"learning_rate": 9.697350436308428e-07,
|
| 1590 |
-
"loss": 1.
|
| 1591 |
-
"mean_token_accuracy": 0.
|
| 1592 |
-
"num_tokens":
|
| 1593 |
"step": 176
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 0.7137096774193549,
|
| 1597 |
-
"grad_norm":
|
| 1598 |
"learning_rate": 9.448151944460657e-07,
|
| 1599 |
-
"loss": 1.
|
| 1600 |
-
"mean_token_accuracy": 0.
|
| 1601 |
-
"num_tokens":
|
| 1602 |
"step": 177
|
| 1603 |
},
|
| 1604 |
{
|
| 1605 |
"epoch": 0.717741935483871,
|
| 1606 |
-
"grad_norm": 5.
|
| 1607 |
"learning_rate": 9.201449038842403e-07,
|
| 1608 |
-
"loss": 1.
|
| 1609 |
-
"mean_token_accuracy": 0.
|
| 1610 |
-
"num_tokens":
|
| 1611 |
"step": 178
|
| 1612 |
},
|
| 1613 |
{
|
| 1614 |
"epoch": 0.7217741935483871,
|
| 1615 |
-
"grad_norm": 6.
|
| 1616 |
"learning_rate": 8.957281307570254e-07,
|
| 1617 |
-
"loss": 1.
|
| 1618 |
-
"mean_token_accuracy": 0.
|
| 1619 |
-
"num_tokens":
|
| 1620 |
"step": 179
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
"epoch": 0.7258064516129032,
|
| 1624 |
-
"grad_norm": 5.
|
| 1625 |
"learning_rate": 8.71568793194445e-07,
|
| 1626 |
-
"loss": 1.
|
| 1627 |
-
"mean_token_accuracy": 0.
|
| 1628 |
-
"num_tokens":
|
| 1629 |
"step": 180
|
| 1630 |
},
|
| 1631 |
{
|
| 1632 |
"epoch": 0.7298387096774194,
|
| 1633 |
-
"grad_norm": 5.
|
| 1634 |
"learning_rate": 8.476707680161486e-07,
|
| 1635 |
-
"loss": 1.
|
| 1636 |
-
"mean_token_accuracy": 0.
|
| 1637 |
-
"num_tokens":
|
| 1638 |
"step": 181
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 0.7338709677419355,
|
| 1642 |
-
"grad_norm": 5.
|
| 1643 |
"learning_rate": 8.240378901093035e-07,
|
| 1644 |
-
"loss": 1.
|
| 1645 |
-
"mean_token_accuracy": 0.
|
| 1646 |
-
"num_tokens":
|
| 1647 |
"step": 182
|
| 1648 |
},
|
| 1649 |
{
|
| 1650 |
"epoch": 0.7379032258064516,
|
| 1651 |
-
"grad_norm": 6.
|
| 1652 |
"learning_rate": 8.006739518132179e-07,
|
| 1653 |
-
"loss": 1.
|
| 1654 |
-
"mean_token_accuracy": 0.
|
| 1655 |
-
"num_tokens":
|
| 1656 |
"step": 183
|
| 1657 |
},
|
| 1658 |
{
|
| 1659 |
"epoch": 0.7419354838709677,
|
| 1660 |
-
"grad_norm": 6.
|
| 1661 |
"learning_rate": 7.775827023107835e-07,
|
| 1662 |
-
"loss": 1.
|
| 1663 |
-
"mean_token_accuracy": 0.
|
| 1664 |
-
"num_tokens":
|
| 1665 |
"step": 184
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 0.7459677419354839,
|
| 1669 |
-
"grad_norm": 5.
|
| 1670 |
"learning_rate": 7.547678470268526e-07,
|
| 1671 |
-
"loss": 1.
|
| 1672 |
-
"mean_token_accuracy": 0.
|
| 1673 |
-
"num_tokens":
|
| 1674 |
"step": 185
|
| 1675 |
},
|
| 1676 |
{
|
| 1677 |
"epoch": 0.75,
|
| 1678 |
-
"grad_norm": 5.
|
| 1679 |
"learning_rate": 7.322330470336314e-07,
|
| 1680 |
-
"loss": 1.
|
| 1681 |
-
"mean_token_accuracy": 0.
|
| 1682 |
-
"num_tokens":
|
| 1683 |
"step": 186
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
"epoch": 0.7540322580645161,
|
| 1687 |
-
"grad_norm": 5.
|
| 1688 |
"learning_rate": 7.099819184631929e-07,
|
| 1689 |
-
"loss": 1.
|
| 1690 |
-
"mean_token_accuracy": 0.
|
| 1691 |
-
"num_tokens":
|
| 1692 |
"step": 187
|
| 1693 |
},
|
| 1694 |
{
|
| 1695 |
"epoch": 0.7580645161290323,
|
| 1696 |
-
"grad_norm": 5.
|
| 1697 |
"learning_rate": 6.880180319272006e-07,
|
| 1698 |
-
"loss": 1.
|
| 1699 |
-
"mean_token_accuracy": 0.
|
| 1700 |
-
"num_tokens":
|
| 1701 |
"step": 188
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 0.7620967741935484,
|
| 1705 |
-
"grad_norm": 5.
|
| 1706 |
"learning_rate": 6.663449119439358e-07,
|
| 1707 |
-
"loss": 1.
|
| 1708 |
-
"mean_token_accuracy": 0.
|
| 1709 |
-
"num_tokens":
|
| 1710 |
"step": 189
|
| 1711 |
},
|
| 1712 |
{
|
| 1713 |
"epoch": 0.7661290322580645,
|
| 1714 |
-
"grad_norm": 5.
|
| 1715 |
"learning_rate": 6.449660363727236e-07,
|
| 1716 |
-
"loss": 1.
|
| 1717 |
-
"mean_token_accuracy": 0.
|
| 1718 |
-
"num_tokens":
|
| 1719 |
"step": 190
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
"epoch": 0.7701612903225806,
|
| 1723 |
-
"grad_norm": 6.
|
| 1724 |
"learning_rate": 6.238848358558439e-07,
|
| 1725 |
-
"loss": 1.
|
| 1726 |
-
"mean_token_accuracy": 0.
|
| 1727 |
-
"num_tokens":
|
| 1728 |
"step": 191
|
| 1729 |
},
|
| 1730 |
{
|
| 1731 |
"epoch": 0.7741935483870968,
|
| 1732 |
-
"grad_norm": 6.
|
| 1733 |
"learning_rate": 6.031046932680229e-07,
|
| 1734 |
-
"loss": 1.
|
| 1735 |
-
"mean_token_accuracy": 0.
|
| 1736 |
-
"num_tokens":
|
| 1737 |
"step": 192
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 0.7782258064516129,
|
| 1741 |
-
"grad_norm": 5.
|
| 1742 |
"learning_rate": 5.826289431735832e-07,
|
| 1743 |
-
"loss": 1.
|
| 1744 |
-
"mean_token_accuracy": 0.
|
| 1745 |
-
"num_tokens":
|
| 1746 |
"step": 193
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
"epoch": 0.782258064516129,
|
| 1750 |
-
"grad_norm": 6.
|
| 1751 |
"learning_rate": 5.624608712913531e-07,
|
| 1752 |
-
"loss": 1.
|
| 1753 |
-
"mean_token_accuracy": 0.
|
| 1754 |
-
"num_tokens":
|
| 1755 |
"step": 194
|
| 1756 |
},
|
| 1757 |
{
|
| 1758 |
"epoch": 0.7862903225806451,
|
| 1759 |
-
"grad_norm": 5.
|
| 1760 |
"learning_rate": 5.426037139674117e-07,
|
| 1761 |
-
"loss": 1.
|
| 1762 |
-
"mean_token_accuracy": 0.
|
| 1763 |
-
"num_tokens":
|
| 1764 |
"step": 195
|
| 1765 |
},
|
| 1766 |
{
|
| 1767 |
"epoch": 0.7903225806451613,
|
| 1768 |
-
"grad_norm": 5.
|
| 1769 |
"learning_rate": 5.23060657655754e-07,
|
| 1770 |
-
"loss": 1.
|
| 1771 |
-
"mean_token_accuracy": 0.
|
| 1772 |
-
"num_tokens":
|
| 1773 |
"step": 196
|
| 1774 |
},
|
| 1775 |
{
|
| 1776 |
"epoch": 0.7943548387096774,
|
| 1777 |
-
"grad_norm": 5.
|
| 1778 |
"learning_rate": 5.038348384069663e-07,
|
| 1779 |
-
"loss": 1.
|
| 1780 |
-
"mean_token_accuracy": 0.
|
| 1781 |
-
"num_tokens":
|
| 1782 |
"step": 197
|
| 1783 |
},
|
| 1784 |
{
|
| 1785 |
"epoch": 0.7983870967741935,
|
| 1786 |
-
"grad_norm": 5.
|
| 1787 |
"learning_rate": 4.84929341364988e-07,
|
| 1788 |
-
"loss": 1.
|
| 1789 |
-
"mean_token_accuracy": 0.
|
| 1790 |
-
"num_tokens":
|
| 1791 |
"step": 198
|
| 1792 |
},
|
| 1793 |
{
|
| 1794 |
"epoch": 0.8024193548387096,
|
| 1795 |
-
"grad_norm": 5.
|
| 1796 |
"learning_rate": 4.6634720027204093e-07,
|
| 1797 |
-
"loss": 1.
|
| 1798 |
-
"mean_token_accuracy": 0.
|
| 1799 |
-
"num_tokens":
|
| 1800 |
"step": 199
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 0.8064516129032258,
|
| 1804 |
-
"grad_norm": 5.
|
| 1805 |
"learning_rate": 4.480913969818099e-07,
|
| 1806 |
-
"loss": 1.
|
| 1807 |
-
"mean_token_accuracy": 0.
|
| 1808 |
-
"num_tokens":
|
| 1809 |
"step": 200
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 0.8104838709677419,
|
| 1813 |
-
"grad_norm": 5.
|
| 1814 |
"learning_rate": 4.3016486098094667e-07,
|
| 1815 |
-
"loss": 1.
|
| 1816 |
-
"mean_token_accuracy": 0.
|
| 1817 |
-
"num_tokens":
|
| 1818 |
"step": 201
|
| 1819 |
},
|
| 1820 |
{
|
| 1821 |
"epoch": 0.8145161290322581,
|
| 1822 |
-
"grad_norm": 5.
|
| 1823 |
"learning_rate": 4.125704689189819e-07,
|
| 1824 |
-
"loss": 1.
|
| 1825 |
-
"mean_token_accuracy": 0.
|
| 1826 |
-
"num_tokens":
|
| 1827 |
"step": 202
|
| 1828 |
},
|
| 1829 |
{
|
| 1830 |
"epoch": 0.8185483870967742,
|
| 1831 |
-
"grad_norm": 5.
|
| 1832 |
"learning_rate": 3.953110441467073e-07,
|
| 1833 |
-
"loss": 1.
|
| 1834 |
-
"mean_token_accuracy": 0.
|
| 1835 |
-
"num_tokens":
|
| 1836 |
"step": 203
|
| 1837 |
},
|
| 1838 |
{
|
| 1839 |
"epoch": 0.8225806451612904,
|
| 1840 |
-
"grad_norm": 5.
|
| 1841 |
"learning_rate": 3.7838935626312246e-07,
|
| 1842 |
-
"loss": 1.
|
| 1843 |
-
"mean_token_accuracy": 0.
|
| 1844 |
-
"num_tokens":
|
| 1845 |
"step": 204
|
| 1846 |
},
|
| 1847 |
{
|
| 1848 |
"epoch": 0.8266129032258065,
|
| 1849 |
-
"grad_norm": 5.
|
| 1850 |
"learning_rate": 3.6180812067099477e-07,
|
| 1851 |
-
"loss": 1.
|
| 1852 |
-
"mean_token_accuracy": 0.
|
| 1853 |
-
"num_tokens":
|
| 1854 |
"step": 205
|
| 1855 |
},
|
| 1856 |
{
|
| 1857 |
"epoch": 0.8306451612903226,
|
| 1858 |
-
"grad_norm": 5.
|
| 1859 |
"learning_rate": 3.455699981411259e-07,
|
| 1860 |
-
"loss": 1.
|
| 1861 |
-
"mean_token_accuracy": 0.
|
| 1862 |
-
"num_tokens":
|
| 1863 |
"step": 206
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 0.8346774193548387,
|
| 1867 |
-
"grad_norm": 5.
|
| 1868 |
"learning_rate": 3.296775943853789e-07,
|
| 1869 |
-
"loss": 1.
|
| 1870 |
-
"mean_token_accuracy": 0.
|
| 1871 |
-
"num_tokens":
|
| 1872 |
"step": 207
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 0.8387096774193549,
|
| 1876 |
-
"grad_norm": 5.
|
| 1877 |
"learning_rate": 3.141334596385448e-07,
|
| 1878 |
-
"loss": 1.
|
| 1879 |
-
"mean_token_accuracy": 0.
|
| 1880 |
-
"num_tokens":
|
| 1881 |
"step": 208
|
| 1882 |
},
|
| 1883 |
{
|
| 1884 |
"epoch": 0.842741935483871,
|
| 1885 |
-
"grad_norm": 5.
|
| 1886 |
"learning_rate": 2.9894008824910726e-07,
|
| 1887 |
-
"loss": 1.
|
| 1888 |
-
"mean_token_accuracy": 0.
|
| 1889 |
-
"num_tokens":
|
| 1890 |
"step": 209
|
| 1891 |
},
|
| 1892 |
{
|
| 1893 |
"epoch": 0.8467741935483871,
|
| 1894 |
-
"grad_norm": 5.
|
| 1895 |
"learning_rate": 2.840999182789797e-07,
|
| 1896 |
-
"loss": 1.
|
| 1897 |
-
"mean_token_accuracy": 0.
|
| 1898 |
-
"num_tokens":
|
| 1899 |
"step": 210
|
| 1900 |
},
|
| 1901 |
{
|
| 1902 |
"epoch": 0.8508064516129032,
|
| 1903 |
-
"grad_norm": 5.
|
| 1904 |
"learning_rate": 2.696153311122704e-07,
|
| 1905 |
-
"loss": 1.
|
| 1906 |
-
"mean_token_accuracy": 0.
|
| 1907 |
-
"num_tokens":
|
| 1908 |
"step": 211
|
| 1909 |
},
|
| 1910 |
{
|
| 1911 |
"epoch": 0.8548387096774194,
|
| 1912 |
-
"grad_norm":
|
| 1913 |
"learning_rate": 2.5548865107314606e-07,
|
| 1914 |
-
"loss": 1.
|
| 1915 |
-
"mean_token_accuracy": 0.
|
| 1916 |
-
"num_tokens":
|
| 1917 |
"step": 212
|
| 1918 |
},
|
| 1919 |
{
|
| 1920 |
"epoch": 0.8588709677419355,
|
| 1921 |
-
"grad_norm": 5.
|
| 1922 |
"learning_rate": 2.4172214505285006e-07,
|
| 1923 |
-
"loss": 1.
|
| 1924 |
-
"mean_token_accuracy": 0.
|
| 1925 |
-
"num_tokens":
|
| 1926 |
"step": 213
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 0.8629032258064516,
|
| 1930 |
-
"grad_norm": 5.
|
| 1931 |
"learning_rate": 2.2831802214593774e-07,
|
| 1932 |
-
"loss": 1.
|
| 1933 |
-
"mean_token_accuracy": 0.
|
| 1934 |
-
"num_tokens":
|
| 1935 |
"step": 214
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
"epoch": 0.8669354838709677,
|
| 1939 |
-
"grad_norm": 5.
|
| 1940 |
"learning_rate": 2.1527843329578328e-07,
|
| 1941 |
-
"loss": 1.
|
| 1942 |
-
"mean_token_accuracy": 0.
|
| 1943 |
-
"num_tokens":
|
| 1944 |
"step": 215
|
| 1945 |
},
|
| 1946 |
{
|
| 1947 |
"epoch": 0.8709677419354839,
|
| 1948 |
-
"grad_norm": 5.
|
| 1949 |
"learning_rate": 2.026054709494235e-07,
|
| 1950 |
-
"loss": 1.
|
| 1951 |
-
"mean_token_accuracy": 0.
|
| 1952 |
-
"num_tokens":
|
| 1953 |
"step": 216
|
| 1954 |
},
|
| 1955 |
{
|
| 1956 |
"epoch": 0.875,
|
| 1957 |
-
"grad_norm":
|
| 1958 |
"learning_rate": 1.9030116872178317e-07,
|
| 1959 |
-
"loss": 1.
|
| 1960 |
-
"mean_token_accuracy": 0.
|
| 1961 |
-
"num_tokens":
|
| 1962 |
"step": 217
|
| 1963 |
},
|
| 1964 |
{
|
| 1965 |
"epoch": 0.8790322580645161,
|
| 1966 |
-
"grad_norm": 5.
|
| 1967 |
"learning_rate": 1.7836750106934475e-07,
|
| 1968 |
-
"loss": 1.
|
| 1969 |
-
"mean_token_accuracy": 0.
|
| 1970 |
-
"num_tokens":
|
| 1971 |
"step": 218
|
| 1972 |
},
|
| 1973 |
{
|
| 1974 |
"epoch": 0.8830645161290323,
|
| 1975 |
-
"grad_norm": 6.
|
| 1976 |
"learning_rate": 1.6680638297330854e-07,
|
| 1977 |
-
"loss": 1.
|
| 1978 |
-
"mean_token_accuracy": 0.
|
| 1979 |
-
"num_tokens":
|
| 1980 |
"step": 219
|
| 1981 |
},
|
| 1982 |
{
|
| 1983 |
"epoch": 0.8870967741935484,
|
| 1984 |
-
"grad_norm": 6.
|
| 1985 |
"learning_rate": 1.5561966963229925e-07,
|
| 1986 |
-
"loss": 1.
|
| 1987 |
-
"mean_token_accuracy": 0.
|
| 1988 |
-
"num_tokens":
|
| 1989 |
"step": 220
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 0.8911290322580645,
|
| 1993 |
-
"grad_norm": 5.
|
| 1994 |
"learning_rate": 1.448091561646628e-07,
|
| 1995 |
-
"loss": 1.
|
| 1996 |
-
"mean_token_accuracy": 0.
|
| 1997 |
-
"num_tokens":
|
| 1998 |
"step": 221
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
"epoch": 0.8951612903225806,
|
| 2002 |
-
"grad_norm": 6.
|
| 2003 |
"learning_rate": 1.3437657732040783e-07,
|
| 2004 |
-
"loss": 1.
|
| 2005 |
-
"mean_token_accuracy": 0.
|
| 2006 |
-
"num_tokens":
|
| 2007 |
"step": 222
|
| 2008 |
},
|
| 2009 |
{
|
| 2010 |
"epoch": 0.8991935483870968,
|
| 2011 |
-
"grad_norm": 5.
|
| 2012 |
"learning_rate": 1.243236072028317e-07,
|
| 2013 |
-
"loss": 1.
|
| 2014 |
-
"mean_token_accuracy": 0.
|
| 2015 |
-
"num_tokens":
|
| 2016 |
"step": 223
|
| 2017 |
},
|
| 2018 |
{
|
| 2019 |
"epoch": 0.9032258064516129,
|
| 2020 |
-
"grad_norm": 5.
|
| 2021 |
"learning_rate": 1.1465185899987797e-07,
|
| 2022 |
-
"loss": 1.
|
| 2023 |
-
"mean_token_accuracy": 0.
|
| 2024 |
-
"num_tokens":
|
| 2025 |
"step": 224
|
| 2026 |
},
|
| 2027 |
{
|
| 2028 |
"epoch": 0.907258064516129,
|
| 2029 |
-
"grad_norm": 5.
|
| 2030 |
"learning_rate": 1.0536288472527162e-07,
|
| 2031 |
-
"loss": 1.
|
| 2032 |
-
"mean_token_accuracy": 0.
|
| 2033 |
-
"num_tokens":
|
| 2034 |
"step": 225
|
| 2035 |
},
|
| 2036 |
{
|
| 2037 |
"epoch": 0.9112903225806451,
|
| 2038 |
-
"grad_norm": 5.
|
| 2039 |
"learning_rate": 9.645817496946902e-08,
|
| 2040 |
-
"loss": 1.
|
| 2041 |
-
"mean_token_accuracy": 0.
|
| 2042 |
-
"num_tokens":
|
| 2043 |
"step": 226
|
| 2044 |
},
|
| 2045 |
{
|
| 2046 |
"epoch": 0.9153225806451613,
|
| 2047 |
-
"grad_norm":
|
| 2048 |
"learning_rate": 8.79391586604636e-08,
|
| 2049 |
-
"loss": 1.
|
| 2050 |
-
"mean_token_accuracy": 0.
|
| 2051 |
-
"num_tokens":
|
| 2052 |
"step": 227
|
| 2053 |
},
|
| 2054 |
{
|
| 2055 |
"epoch": 0.9193548387096774,
|
| 2056 |
-
"grad_norm":
|
| 2057 |
"learning_rate": 7.980720283448957e-08,
|
| 2058 |
-
"loss": 1.
|
| 2059 |
-
"mean_token_accuracy": 0.
|
| 2060 |
-
"num_tokens":
|
| 2061 |
"step": 228
|
| 2062 |
},
|
| 2063 |
{
|
| 2064 |
"epoch": 0.9233870967741935,
|
| 2065 |
-
"grad_norm": 6.
|
| 2066 |
"learning_rate": 7.206361241665266e-08,
|
| 2067 |
-
"loss": 1.
|
| 2068 |
-
"mean_token_accuracy": 0.
|
| 2069 |
-
"num_tokens":
|
| 2070 |
"step": 229
|
| 2071 |
},
|
| 2072 |
{
|
| 2073 |
"epoch": 0.9274193548387096,
|
| 2074 |
-
"grad_norm": 5.
|
| 2075 |
"learning_rate": 6.470963001153268e-08,
|
| 2076 |
-
"loss": 1.
|
| 2077 |
-
"mean_token_accuracy": 0.
|
| 2078 |
-
"num_tokens":
|
| 2079 |
"step": 230
|
| 2080 |
},
|
| 2081 |
{
|
| 2082 |
"epoch": 0.9314516129032258,
|
| 2083 |
-
"grad_norm": 5.
|
| 2084 |
"learning_rate": 5.774643570378296e-08,
|
| 2085 |
-
"loss": 1.
|
| 2086 |
-
"mean_token_accuracy": 0.
|
| 2087 |
-
"num_tokens":
|
| 2088 |
"step": 231
|
| 2089 |
},
|
| 2090 |
{
|
| 2091 |
"epoch": 0.9354838709677419,
|
| 2092 |
-
"grad_norm": 5.
|
| 2093 |
"learning_rate": 5.117514686876379e-08,
|
| 2094 |
-
"loss": 1.
|
| 2095 |
-
"mean_token_accuracy": 0.
|
| 2096 |
-
"num_tokens":
|
| 2097 |
"step": 232
|
| 2098 |
},
|
| 2099 |
{
|
| 2100 |
"epoch": 0.9395161290322581,
|
| 2101 |
-
"grad_norm": 5.
|
| 2102 |
"learning_rate": 4.4996817993239464e-08,
|
| 2103 |
-
"loss": 1.
|
| 2104 |
-
"mean_token_accuracy": 0.
|
| 2105 |
-
"num_tokens":
|
| 2106 |
"step": 233
|
| 2107 |
},
|
| 2108 |
{
|
| 2109 |
"epoch": 0.9435483870967742,
|
| 2110 |
-
"grad_norm": 6.
|
| 2111 |
"learning_rate": 3.9212440506164465e-08,
|
| 2112 |
-
"loss": 1.
|
| 2113 |
-
"mean_token_accuracy": 0.
|
| 2114 |
-
"num_tokens":
|
| 2115 |
"step": 234
|
| 2116 |
},
|
| 2117 |
{
|
| 2118 |
"epoch": 0.9475806451612904,
|
| 2119 |
-
"grad_norm":
|
| 2120 |
"learning_rate": 3.382294261959157e-08,
|
| 2121 |
-
"loss": 1.
|
| 2122 |
-
"mean_token_accuracy": 0.
|
| 2123 |
-
"num_tokens":
|
| 2124 |
"step": 235
|
| 2125 |
},
|
| 2126 |
{
|
| 2127 |
"epoch": 0.9516129032258065,
|
| 2128 |
-
"grad_norm": 6.
|
| 2129 |
"learning_rate": 2.8829189179721552e-08,
|
| 2130 |
-
"loss": 1.
|
| 2131 |
-
"mean_token_accuracy": 0.
|
| 2132 |
-
"num_tokens":
|
| 2133 |
"step": 236
|
| 2134 |
},
|
| 2135 |
{
|
| 2136 |
"epoch": 0.9556451612903226,
|
| 2137 |
-
"grad_norm": 5.
|
| 2138 |
"learning_rate": 2.423198152812306e-08,
|
| 2139 |
-
"loss": 1.
|
| 2140 |
-
"mean_token_accuracy": 0.
|
| 2141 |
-
"num_tokens":
|
| 2142 |
"step": 237
|
| 2143 |
},
|
| 2144 |
{
|
| 2145 |
"epoch": 0.9596774193548387,
|
| 2146 |
-
"grad_norm": 5.
|
| 2147 |
"learning_rate": 2.0032057373142453e-08,
|
| 2148 |
-
"loss": 1.
|
| 2149 |
-
"mean_token_accuracy": 0.
|
| 2150 |
-
"num_tokens":
|
| 2151 |
"step": 238
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
"epoch": 0.9637096774193549,
|
| 2155 |
-
"grad_norm": 6.
|
| 2156 |
"learning_rate": 1.6230090671524312e-08,
|
| 2157 |
-
"loss": 1.
|
| 2158 |
-
"mean_token_accuracy": 0.
|
| 2159 |
-
"num_tokens":
|
| 2160 |
"step": 239
|
| 2161 |
},
|
| 2162 |
{
|
| 2163 |
"epoch": 0.967741935483871,
|
| 2164 |
-
"grad_norm": 6.
|
| 2165 |
"learning_rate": 1.2826691520262114e-08,
|
| 2166 |
-
"loss": 1.
|
| 2167 |
-
"mean_token_accuracy": 0.
|
| 2168 |
-
"num_tokens":
|
| 2169 |
"step": 240
|
| 2170 |
},
|
| 2171 |
{
|
| 2172 |
"epoch": 0.9717741935483871,
|
| 2173 |
-
"grad_norm": 5.
|
| 2174 |
"learning_rate": 9.822406058697665e-09,
|
| 2175 |
-
"loss": 1.
|
| 2176 |
-
"mean_token_accuracy": 0.
|
| 2177 |
-
"num_tokens":
|
| 2178 |
"step": 241
|
| 2179 |
},
|
| 2180 |
{
|
| 2181 |
"epoch": 0.9758064516129032,
|
| 2182 |
-
"grad_norm": 5.
|
| 2183 |
"learning_rate": 7.217716380881479e-09,
|
| 2184 |
-
"loss": 1.
|
| 2185 |
-
"mean_token_accuracy": 0.
|
| 2186 |
-
"num_tokens":
|
| 2187 |
"step": 242
|
| 2188 |
},
|
| 2189 |
{
|
| 2190 |
"epoch": 0.9798387096774194,
|
| 2191 |
-
"grad_norm": 6.
|
| 2192 |
"learning_rate": 5.0130404582127144e-09,
|
| 2193 |
-
"loss": 1.
|
| 2194 |
-
"mean_token_accuracy": 0.
|
| 2195 |
-
"num_tokens":
|
| 2196 |
"step": 243
|
| 2197 |
},
|
| 2198 |
{
|
| 2199 |
"epoch": 0.9838709677419355,
|
| 2200 |
-
"grad_norm": 5.
|
| 2201 |
"learning_rate": 3.208732072368104e-09,
|
| 2202 |
-
"loss": 1.
|
| 2203 |
-
"mean_token_accuracy": 0.
|
| 2204 |
-
"num_tokens":
|
| 2205 |
"step": 244
|
| 2206 |
},
|
| 2207 |
{
|
| 2208 |
"epoch": 0.9879032258064516,
|
| 2209 |
-
"grad_norm": 5.
|
| 2210 |
"learning_rate": 1.8050807585293095e-09,
|
| 2211 |
-
"loss": 1.
|
| 2212 |
-
"mean_token_accuracy": 0.
|
| 2213 |
-
"num_tokens":
|
| 2214 |
"step": 245
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 0.9919354838709677,
|
| 2218 |
-
"grad_norm": 5.
|
| 2219 |
"learning_rate": 8.023117589237017e-10,
|
| 2220 |
-
"loss": 1.
|
| 2221 |
-
"mean_token_accuracy": 0.
|
| 2222 |
-
"num_tokens":
|
| 2223 |
"step": 246
|
| 2224 |
},
|
| 2225 |
{
|
| 2226 |
"epoch": 0.9959677419354839,
|
| 2227 |
-
"grad_norm":
|
| 2228 |
"learning_rate": 2.0058598667854755e-10,
|
| 2229 |
-
"loss": 1.
|
| 2230 |
-
"mean_token_accuracy": 0.
|
| 2231 |
-
"num_tokens":
|
| 2232 |
"step": 247
|
| 2233 |
},
|
| 2234 |
{
|
| 2235 |
"epoch": 1.0,
|
| 2236 |
-
"grad_norm": 6.
|
| 2237 |
"learning_rate": 0.0,
|
| 2238 |
-
"loss": 1.
|
| 2239 |
-
"mean_token_accuracy": 0.
|
| 2240 |
-
"num_tokens":
|
| 2241 |
"step": 248
|
| 2242 |
}
|
| 2243 |
],
|
|
@@ -2258,7 +2258,7 @@
|
|
| 2258 |
"attributes": {}
|
| 2259 |
}
|
| 2260 |
},
|
| 2261 |
-
"total_flos":
|
| 2262 |
"train_batch_size": 1,
|
| 2263 |
"trial_name": null,
|
| 2264 |
"trial_params": null
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004032258064516129,
|
| 13 |
+
"grad_norm": 99.0960693359375,
|
| 14 |
"learning_rate": 4.999799414013322e-06,
|
| 15 |
+
"loss": 3.2792,
|
| 16 |
+
"mean_token_accuracy": 0.4705558717250824,
|
| 17 |
+
"num_tokens": 1819.0,
|
| 18 |
"step": 1
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 0.008064516129032258,
|
| 22 |
+
"grad_norm": 47.43061828613281,
|
| 23 |
"learning_rate": 4.999197688241076e-06,
|
| 24 |
+
"loss": 2.6949,
|
| 25 |
+
"mean_token_accuracy": 0.4918949007987976,
|
| 26 |
+
"num_tokens": 3610.0,
|
| 27 |
"step": 2
|
| 28 |
},
|
| 29 |
{
|
| 30 |
"epoch": 0.012096774193548387,
|
| 31 |
+
"grad_norm": 40.37574005126953,
|
| 32 |
"learning_rate": 4.998194919241471e-06,
|
| 33 |
+
"loss": 2.1322,
|
| 34 |
+
"mean_token_accuracy": 0.5253481864929199,
|
| 35 |
+
"num_tokens": 5407.0,
|
| 36 |
"step": 3
|
| 37 |
},
|
| 38 |
{
|
| 39 |
"epoch": 0.016129032258064516,
|
| 40 |
+
"grad_norm": 21.76544189453125,
|
| 41 |
"learning_rate": 4.996791267927632e-06,
|
| 42 |
+
"loss": 1.924,
|
| 43 |
+
"mean_token_accuracy": 0.5501400828361511,
|
| 44 |
+
"num_tokens": 7194.0,
|
| 45 |
"step": 4
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.020161290322580645,
|
| 49 |
+
"grad_norm": 30.451719284057617,
|
| 50 |
"learning_rate": 4.994986959541788e-06,
|
| 51 |
+
"loss": 2.0233,
|
| 52 |
+
"mean_token_accuracy": 0.5161290168762207,
|
| 53 |
+
"num_tokens": 9242.0,
|
| 54 |
"step": 5
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.024193548387096774,
|
| 58 |
+
"grad_norm": 8.43026065826416,
|
| 59 |
"learning_rate": 4.9927822836191185e-06,
|
| 60 |
+
"loss": 1.8731,
|
| 61 |
+
"mean_token_accuracy": 0.5502092242240906,
|
| 62 |
+
"num_tokens": 11156.0,
|
| 63 |
"step": 6
|
| 64 |
},
|
| 65 |
{
|
| 66 |
"epoch": 0.028225806451612902,
|
| 67 |
+
"grad_norm": 8.538537979125977,
|
| 68 |
"learning_rate": 4.990177593941303e-06,
|
| 69 |
+
"loss": 1.8274,
|
| 70 |
+
"mean_token_accuracy": 0.5549915432929993,
|
| 71 |
+
"num_tokens": 12931.0,
|
| 72 |
"step": 7
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.03225806451612903,
|
| 76 |
+
"grad_norm": 8.681036949157715,
|
| 77 |
"learning_rate": 4.987173308479738e-06,
|
| 78 |
+
"loss": 1.8148,
|
| 79 |
+
"mean_token_accuracy": 0.563049852848053,
|
| 80 |
+
"num_tokens": 14979.0,
|
| 81 |
"step": 8
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.036290322580645164,
|
| 85 |
+
"grad_norm": 7.441074848175049,
|
| 86 |
"learning_rate": 4.9837699093284765e-06,
|
| 87 |
+
"loss": 1.7176,
|
| 88 |
+
"mean_token_accuracy": 0.572820782661438,
|
| 89 |
+
"num_tokens": 16828.0,
|
| 90 |
"step": 9
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"epoch": 0.04032258064516129,
|
| 94 |
+
"grad_norm": 7.5016093254089355,
|
| 95 |
"learning_rate": 4.9799679426268575e-06,
|
| 96 |
+
"loss": 1.826,
|
| 97 |
+
"mean_token_accuracy": 0.5537056922912598,
|
| 98 |
+
"num_tokens": 18692.0,
|
| 99 |
"step": 10
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.04435483870967742,
|
| 103 |
+
"grad_norm": 7.0608062744140625,
|
| 104 |
"learning_rate": 4.975768018471877e-06,
|
| 105 |
+
"loss": 1.6788,
|
| 106 |
+
"mean_token_accuracy": 0.5741225481033325,
|
| 107 |
+
"num_tokens": 20603.0,
|
| 108 |
"step": 11
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.04838709677419355,
|
| 112 |
+
"grad_norm": 7.283655166625977,
|
| 113 |
"learning_rate": 4.971170810820279e-06,
|
| 114 |
+
"loss": 1.6694,
|
| 115 |
+
"mean_token_accuracy": 0.5708954930305481,
|
| 116 |
+
"num_tokens": 22213.0,
|
| 117 |
"step": 12
|
| 118 |
},
|
| 119 |
{
|
| 120 |
"epoch": 0.05241935483870968,
|
| 121 |
+
"grad_norm": 7.800687313079834,
|
| 122 |
"learning_rate": 4.966177057380409e-06,
|
| 123 |
+
"loss": 1.7202,
|
| 124 |
+
"mean_token_accuracy": 0.5798588991165161,
|
| 125 |
+
"num_tokens": 23774.0,
|
| 126 |
"step": 13
|
| 127 |
},
|
| 128 |
{
|
| 129 |
"epoch": 0.056451612903225805,
|
| 130 |
+
"grad_norm": 7.64439058303833,
|
| 131 |
"learning_rate": 4.960787559493836e-06,
|
| 132 |
+
"loss": 1.6408,
|
| 133 |
+
"mean_token_accuracy": 0.5935754179954529,
|
| 134 |
+
"num_tokens": 25208.0,
|
| 135 |
"step": 14
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.06048387096774194,
|
| 139 |
+
"grad_norm": 6.813995361328125,
|
| 140 |
"learning_rate": 4.955003182006761e-06,
|
| 141 |
+
"loss": 1.6397,
|
| 142 |
+
"mean_token_accuracy": 0.5809217691421509,
|
| 143 |
+
"num_tokens": 27076.0,
|
| 144 |
"step": 15
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 0.06451612903225806,
|
| 148 |
+
"grad_norm": 7.024792194366455,
|
| 149 |
"learning_rate": 4.948824853131237e-06,
|
| 150 |
+
"loss": 1.7807,
|
| 151 |
+
"mean_token_accuracy": 0.5546666383743286,
|
| 152 |
+
"num_tokens": 28953.0,
|
| 153 |
"step": 16
|
| 154 |
},
|
| 155 |
{
|
| 156 |
"epoch": 0.06854838709677419,
|
| 157 |
+
"grad_norm": 6.979730606079102,
|
| 158 |
"learning_rate": 4.942253564296217e-06,
|
| 159 |
+
"loss": 1.5861,
|
| 160 |
+
"mean_token_accuracy": 0.5993921160697937,
|
| 161 |
+
"num_tokens": 30600.0,
|
| 162 |
"step": 17
|
| 163 |
},
|
| 164 |
{
|
| 165 |
"epoch": 0.07258064516129033,
|
| 166 |
+
"grad_norm": 6.971430778503418,
|
| 167 |
"learning_rate": 4.935290369988468e-06,
|
| 168 |
+
"loss": 1.7106,
|
| 169 |
+
"mean_token_accuracy": 0.5747259259223938,
|
| 170 |
+
"num_tokens": 32335.0,
|
| 171 |
"step": 18
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.07661290322580645,
|
| 175 |
+
"grad_norm": 7.085752487182617,
|
| 176 |
"learning_rate": 4.927936387583348e-06,
|
| 177 |
+
"loss": 1.7075,
|
| 178 |
+
"mean_token_accuracy": 0.5898089408874512,
|
| 179 |
+
"num_tokens": 33907.0,
|
| 180 |
"step": 19
|
| 181 |
},
|
| 182 |
{
|
| 183 |
"epoch": 0.08064516129032258,
|
| 184 |
+
"grad_norm": 6.59229040145874,
|
| 185 |
"learning_rate": 4.920192797165511e-06,
|
| 186 |
+
"loss": 1.7039,
|
| 187 |
+
"mean_token_accuracy": 0.569328784942627,
|
| 188 |
+
"num_tokens": 35950.0,
|
| 189 |
"step": 20
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 0.0846774193548387,
|
| 193 |
+
"grad_norm": 6.934664249420166,
|
| 194 |
"learning_rate": 4.912060841339536e-06,
|
| 195 |
+
"loss": 1.582,
|
| 196 |
+
"mean_token_accuracy": 0.6080306172370911,
|
| 197 |
+
"num_tokens": 37521.0,
|
| 198 |
"step": 21
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.08870967741935484,
|
| 202 |
+
"grad_norm": 7.222076416015625,
|
| 203 |
"learning_rate": 4.9035418250305314e-06,
|
| 204 |
+
"loss": 1.6762,
|
| 205 |
+
"mean_token_accuracy": 0.5719424486160278,
|
| 206 |
+
"num_tokens": 39191.0,
|
| 207 |
"step": 22
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.09274193548387097,
|
| 211 |
+
"grad_norm": 6.332627773284912,
|
| 212 |
"learning_rate": 4.894637115274728e-06,
|
| 213 |
+
"loss": 1.6376,
|
| 214 |
+
"mean_token_accuracy": 0.5898959636688232,
|
| 215 |
+
"num_tokens": 41212.0,
|
| 216 |
"step": 23
|
| 217 |
},
|
| 218 |
{
|
| 219 |
"epoch": 0.0967741935483871,
|
| 220 |
+
"grad_norm": 6.263752460479736,
|
| 221 |
"learning_rate": 4.8853481410001225e-06,
|
| 222 |
+
"loss": 1.6816,
|
| 223 |
+
"mean_token_accuracy": 0.5732259750366211,
|
| 224 |
+
"num_tokens": 43201.0,
|
| 225 |
"step": 24
|
| 226 |
},
|
| 227 |
{
|
| 228 |
"epoch": 0.10080645161290322,
|
| 229 |
+
"grad_norm": 6.8030476570129395,
|
| 230 |
"learning_rate": 4.875676392797169e-06,
|
| 231 |
+
"loss": 1.5712,
|
| 232 |
+
"mean_token_accuracy": 0.6030839681625366,
|
| 233 |
+
"num_tokens": 44954.0,
|
| 234 |
"step": 25
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.10483870967741936,
|
| 238 |
+
"grad_norm": 6.477639198303223,
|
| 239 |
"learning_rate": 4.865623422679593e-06,
|
| 240 |
+
"loss": 1.6535,
|
| 241 |
+
"mean_token_accuracy": 0.5837808847427368,
|
| 242 |
+
"num_tokens": 46818.0,
|
| 243 |
"step": 26
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 0.10887096774193548,
|
| 247 |
+
"grad_norm": 8.109735488891602,
|
| 248 |
"learning_rate": 4.855190843835338e-06,
|
| 249 |
+
"loss": 1.5428,
|
| 250 |
+
"mean_token_accuracy": 0.60636305809021,
|
| 251 |
+
"num_tokens": 48423.0,
|
| 252 |
"step": 27
|
| 253 |
},
|
| 254 |
{
|
| 255 |
"epoch": 0.11290322580645161,
|
| 256 |
+
"grad_norm": 6.8654046058654785,
|
| 257 |
"learning_rate": 4.844380330367701e-06,
|
| 258 |
+
"loss": 1.5027,
|
| 259 |
+
"mean_token_accuracy": 0.6173325777053833,
|
| 260 |
+
"num_tokens": 50202.0,
|
| 261 |
"step": 28
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.11693548387096774,
|
| 265 |
+
"grad_norm": 6.878509044647217,
|
| 266 |
"learning_rate": 4.833193617026692e-06,
|
| 267 |
+
"loss": 1.6905,
|
| 268 |
+
"mean_token_accuracy": 0.5734870433807373,
|
| 269 |
+
"num_tokens": 51939.0,
|
| 270 |
"step": 29
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 0.12096774193548387,
|
| 274 |
+
"grad_norm": 6.781070709228516,
|
| 275 |
"learning_rate": 4.821632498930656e-06,
|
| 276 |
+
"loss": 1.584,
|
| 277 |
+
"mean_token_accuracy": 0.5995055437088013,
|
| 278 |
+
"num_tokens": 53559.0,
|
| 279 |
"step": 30
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 0.125,
|
| 283 |
+
"grad_norm": 6.2921271324157715,
|
| 284 |
"learning_rate": 4.809698831278217e-06,
|
| 285 |
+
"loss": 1.6781,
|
| 286 |
+
"mean_token_accuracy": 0.582112193107605,
|
| 287 |
+
"num_tokens": 55540.0,
|
| 288 |
"step": 31
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"epoch": 0.12903225806451613,
|
| 292 |
+
"grad_norm": 6.02636194229126,
|
| 293 |
"learning_rate": 4.797394529050577e-06,
|
| 294 |
+
"loss": 1.5655,
|
| 295 |
+
"mean_token_accuracy": 0.6006144285202026,
|
| 296 |
+
"num_tokens": 57495.0,
|
| 297 |
"step": 32
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.13306451612903225,
|
| 301 |
+
"grad_norm": 7.5229902267456055,
|
| 302 |
"learning_rate": 4.784721566704217e-06,
|
| 303 |
+
"loss": 1.7261,
|
| 304 |
+
"mean_token_accuracy": 0.5760197639465332,
|
| 305 |
+
"num_tokens": 59115.0,
|
| 306 |
"step": 33
|
| 307 |
},
|
| 308 |
{
|
| 309 |
"epoch": 0.13709677419354838,
|
| 310 |
+
"grad_norm": 6.765624046325684,
|
| 311 |
"learning_rate": 4.771681977854062e-06,
|
| 312 |
+
"loss": 1.625,
|
| 313 |
+
"mean_token_accuracy": 0.5822873115539551,
|
| 314 |
+
"num_tokens": 60551.0,
|
| 315 |
"step": 34
|
| 316 |
},
|
| 317 |
{
|
| 318 |
"epoch": 0.14112903225806453,
|
| 319 |
+
"grad_norm": 6.864407062530518,
|
| 320 |
"learning_rate": 4.75827785494715e-06,
|
| 321 |
+
"loss": 1.6879,
|
| 322 |
+
"mean_token_accuracy": 0.5782891511917114,
|
| 323 |
+
"num_tokens": 62552.0,
|
| 324 |
"step": 35
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.14516129032258066,
|
| 328 |
+
"grad_norm": 6.733999252319336,
|
| 329 |
"learning_rate": 4.744511348926855e-06,
|
| 330 |
+
"loss": 1.5214,
|
| 331 |
+
"mean_token_accuracy": 0.6006430983543396,
|
| 332 |
+
"num_tokens": 64109.0,
|
| 333 |
"step": 36
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 0.14919354838709678,
|
| 337 |
+
"grad_norm": 5.977267265319824,
|
| 338 |
"learning_rate": 4.730384668887731e-06,
|
| 339 |
+
"loss": 1.6555,
|
| 340 |
+
"mean_token_accuracy": 0.575398862361908,
|
| 341 |
+
"num_tokens": 66054.0,
|
| 342 |
"step": 37
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 0.1532258064516129,
|
| 346 |
+
"grad_norm": 6.319035053253174,
|
| 347 |
"learning_rate": 4.715900081721021e-06,
|
| 348 |
+
"loss": 1.657,
|
| 349 |
+
"mean_token_accuracy": 0.5774725079536438,
|
| 350 |
+
"num_tokens": 67876.0,
|
| 351 |
"step": 38
|
| 352 |
},
|
| 353 |
{
|
| 354 |
"epoch": 0.15725806451612903,
|
| 355 |
+
"grad_norm": 6.248798847198486,
|
| 356 |
"learning_rate": 4.7010599117508936e-06,
|
| 357 |
+
"loss": 1.5393,
|
| 358 |
+
"mean_token_accuracy": 0.5967366099357605,
|
| 359 |
+
"num_tokens": 69594.0,
|
| 360 |
"step": 39
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.16129032258064516,
|
| 364 |
+
"grad_norm": 6.949793815612793,
|
| 365 |
"learning_rate": 4.685866540361456e-06,
|
| 366 |
+
"loss": 1.6971,
|
| 367 |
+
"mean_token_accuracy": 0.5807833671569824,
|
| 368 |
+
"num_tokens": 71230.0,
|
| 369 |
"step": 40
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 0.16532258064516128,
|
| 373 |
+
"grad_norm": 7.044243812561035,
|
| 374 |
"learning_rate": 4.670322405614621e-06,
|
| 375 |
+
"loss": 1.5476,
|
| 376 |
+
"mean_token_accuracy": 0.6068921685218811,
|
| 377 |
+
"num_tokens": 72799.0,
|
| 378 |
"step": 41
|
| 379 |
},
|
| 380 |
{
|
| 381 |
"epoch": 0.1693548387096774,
|
| 382 |
+
"grad_norm": 6.1895670890808105,
|
| 383 |
"learning_rate": 4.654430001858874e-06,
|
| 384 |
+
"loss": 1.5685,
|
| 385 |
+
"mean_token_accuracy": 0.5987124443054199,
|
| 386 |
+
"num_tokens": 74665.0,
|
| 387 |
"step": 42
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.17338709677419356,
|
| 391 |
+
"grad_norm": 6.357958793640137,
|
| 392 |
"learning_rate": 4.638191879329005e-06,
|
| 393 |
+
"loss": 1.6369,
|
| 394 |
+
"mean_token_accuracy": 0.6065057516098022,
|
| 395 |
+
"num_tokens": 76573.0,
|
| 396 |
"step": 43
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.1774193548387097,
|
| 400 |
+
"grad_norm": 6.44213342666626,
|
| 401 |
"learning_rate": 4.621610643736878e-06,
|
| 402 |
+
"loss": 1.477,
|
| 403 |
+
"mean_token_accuracy": 0.6165943741798401,
|
| 404 |
+
"num_tokens": 78419.0,
|
| 405 |
"step": 44
|
| 406 |
},
|
| 407 |
{
|
| 408 |
"epoch": 0.1814516129032258,
|
| 409 |
+
"grad_norm": 11.772004127502441,
|
| 410 |
"learning_rate": 4.6046889558532925e-06,
|
| 411 |
+
"loss": 1.2458,
|
| 412 |
+
"mean_token_accuracy": 0.6105882525444031,
|
| 413 |
+
"num_tokens": 79271.0,
|
| 414 |
"step": 45
|
| 415 |
},
|
| 416 |
{
|
| 417 |
"epoch": 0.18548387096774194,
|
| 418 |
+
"grad_norm": 5.835907936096191,
|
| 419 |
"learning_rate": 4.587429531081019e-06,
|
| 420 |
+
"loss": 1.5937,
|
| 421 |
+
"mean_token_accuracy": 0.5787923336029053,
|
| 422 |
+
"num_tokens": 81310.0,
|
| 423 |
"step": 46
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.18951612903225806,
|
| 427 |
+
"grad_norm": 6.897541046142578,
|
| 428 |
"learning_rate": 4.569835139019054e-06,
|
| 429 |
+
"loss": 1.6269,
|
| 430 |
+
"mean_token_accuracy": 0.5900123119354248,
|
| 431 |
+
"num_tokens": 82934.0,
|
| 432 |
"step": 47
|
| 433 |
},
|
| 434 |
{
|
| 435 |
"epoch": 0.1935483870967742,
|
| 436 |
+
"grad_norm": 6.150733947753906,
|
| 437 |
"learning_rate": 4.551908603018191e-06,
|
| 438 |
+
"loss": 1.515,
|
| 439 |
+
"mean_token_accuracy": 0.5997865796089172,
|
| 440 |
+
"num_tokens": 84810.0,
|
| 441 |
"step": 48
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 0.1975806451612903,
|
| 445 |
+
"grad_norm": 5.933169841766357,
|
| 446 |
"learning_rate": 4.53365279972796e-06,
|
| 447 |
+
"loss": 1.6289,
|
| 448 |
+
"mean_token_accuracy": 0.5821572542190552,
|
| 449 |
+
"num_tokens": 86796.0,
|
| 450 |
"step": 49
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.20161290322580644,
|
| 454 |
+
"grad_norm": 5.93269681930542,
|
| 455 |
"learning_rate": 4.515070658635013e-06,
|
| 456 |
+
"loss": 1.6608,
|
| 457 |
+
"mean_token_accuracy": 0.5925726294517517,
|
| 458 |
+
"num_tokens": 88656.0,
|
| 459 |
"step": 50
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.2056451612903226,
|
| 463 |
+
"grad_norm": 6.56980037689209,
|
| 464 |
"learning_rate": 4.4961651615930344e-06,
|
| 465 |
+
"loss": 1.6368,
|
| 466 |
+
"mean_token_accuracy": 0.5721311569213867,
|
| 467 |
+
"num_tokens": 90488.0,
|
| 468 |
"step": 51
|
| 469 |
},
|
| 470 |
{
|
| 471 |
"epoch": 0.20967741935483872,
|
| 472 |
+
"grad_norm": 5.973373889923096,
|
| 473 |
"learning_rate": 4.476939342344246e-06,
|
| 474 |
+
"loss": 1.5851,
|
| 475 |
+
"mean_token_accuracy": 0.586832582950592,
|
| 476 |
+
"num_tokens": 92419.0,
|
| 477 |
"step": 52
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 0.21370967741935484,
|
| 481 |
+
"grad_norm": 6.0411553382873535,
|
| 482 |
"learning_rate": 4.457396286032589e-06,
|
| 483 |
+
"loss": 1.6096,
|
| 484 |
+
"mean_token_accuracy": 0.5805962681770325,
|
| 485 |
+
"num_tokens": 94400.0,
|
| 486 |
"step": 53
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.21774193548387097,
|
| 490 |
+
"grad_norm": 6.7846221923828125,
|
| 491 |
"learning_rate": 4.437539128708647e-06,
|
| 492 |
+
"loss": 1.7701,
|
| 493 |
+
"mean_token_accuracy": 0.5684491991996765,
|
| 494 |
+
"num_tokens": 96272.0,
|
| 495 |
"step": 54
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 0.2217741935483871,
|
| 499 |
+
"grad_norm": 6.365209579467773,
|
| 500 |
"learning_rate": 4.417371056826417e-06,
|
| 501 |
+
"loss": 1.5444,
|
| 502 |
+
"mean_token_accuracy": 0.6007281541824341,
|
| 503 |
+
"num_tokens": 97922.0,
|
| 504 |
"step": 55
|
| 505 |
},
|
| 506 |
{
|
| 507 |
"epoch": 0.22580645161290322,
|
| 508 |
+
"grad_norm": 6.159636974334717,
|
| 509 |
"learning_rate": 4.396895306731978e-06,
|
| 510 |
+
"loss": 1.6374,
|
| 511 |
+
"mean_token_accuracy": 0.5812357068061829,
|
| 512 |
+
"num_tokens": 99672.0,
|
| 513 |
"step": 56
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.22983870967741934,
|
| 517 |
+
"grad_norm": 6.037807464599609,
|
| 518 |
"learning_rate": 4.376115164144157e-06,
|
| 519 |
+
"loss": 1.5427,
|
| 520 |
+
"mean_token_accuracy": 0.6033584475517273,
|
| 521 |
+
"num_tokens": 101401.0,
|
| 522 |
"step": 57
|
| 523 |
},
|
| 524 |
{
|
| 525 |
"epoch": 0.23387096774193547,
|
| 526 |
+
"grad_norm": 6.224438190460205,
|
| 527 |
"learning_rate": 4.355033963627277e-06,
|
| 528 |
+
"loss": 1.5835,
|
| 529 |
+
"mean_token_accuracy": 0.5897436141967773,
|
| 530 |
+
"num_tokens": 103158.0,
|
| 531 |
"step": 58
|
| 532 |
},
|
| 533 |
{
|
| 534 |
"epoch": 0.23790322580645162,
|
| 535 |
+
"grad_norm": 5.788248062133789,
|
| 536 |
"learning_rate": 4.333655088056065e-06,
|
| 537 |
+
"loss": 1.5512,
|
| 538 |
+
"mean_token_accuracy": 0.5878172516822815,
|
| 539 |
+
"num_tokens": 105130.0,
|
| 540 |
"step": 59
|
| 541 |
},
|
| 542 |
{
|
| 543 |
"epoch": 0.24193548387096775,
|
| 544 |
+
"grad_norm": 6.5106024742126465,
|
| 545 |
"learning_rate": 4.3119819680728e-06,
|
| 546 |
+
"loss": 1.6323,
|
| 547 |
+
"mean_token_accuracy": 0.5971143245697021,
|
| 548 |
+
"num_tokens": 106934.0,
|
| 549 |
"step": 60
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.24596774193548387,
|
| 553 |
+
"grad_norm": 5.9823737144470215,
|
| 554 |
"learning_rate": 4.290018081536807e-06,
|
| 555 |
+
"loss": 1.5736,
|
| 556 |
+
"mean_token_accuracy": 0.5935134887695312,
|
| 557 |
+
"num_tokens": 108786.0,
|
| 558 |
"step": 61
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 0.25,
|
| 562 |
+
"grad_norm": 6.844268798828125,
|
| 563 |
"learning_rate": 4.267766952966369e-06,
|
| 564 |
+
"loss": 1.6744,
|
| 565 |
+
"mean_token_accuracy": 0.5678654313087463,
|
| 566 |
+
"num_tokens": 110512.0,
|
| 567 |
"step": 62
|
| 568 |
},
|
| 569 |
{
|
| 570 |
"epoch": 0.2540322580645161,
|
| 571 |
+
"grad_norm": 6.168908596038818,
|
| 572 |
"learning_rate": 4.245232152973148e-06,
|
| 573 |
+
"loss": 1.6208,
|
| 574 |
+
"mean_token_accuracy": 0.5953115224838257,
|
| 575 |
+
"num_tokens": 112135.0,
|
| 576 |
"step": 63
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.25806451612903225,
|
| 580 |
+
"grad_norm": 6.162073135375977,
|
| 581 |
"learning_rate": 4.222417297689217e-06,
|
| 582 |
+
"loss": 1.6455,
|
| 583 |
+
"mean_token_accuracy": 0.6012843251228333,
|
| 584 |
+
"num_tokens": 113850.0,
|
| 585 |
"step": 64
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.2620967741935484,
|
| 589 |
+
"grad_norm": 5.909740447998047,
|
| 590 |
"learning_rate": 4.199326048186783e-06,
|
| 591 |
+
"loss": 1.6408,
|
| 592 |
+
"mean_token_accuracy": 0.5889984369277954,
|
| 593 |
+
"num_tokens": 115779.0,
|
| 594 |
"step": 65
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 0.2661290322580645,
|
| 598 |
+
"grad_norm": 6.424287796020508,
|
| 599 |
"learning_rate": 4.175962109890697e-06,
|
| 600 |
+
"loss": 1.4827,
|
| 601 |
+
"mean_token_accuracy": 0.6072772741317749,
|
| 602 |
+
"num_tokens": 117375.0,
|
| 603 |
"step": 66
|
| 604 |
},
|
| 605 |
{
|
| 606 |
"epoch": 0.2701612903225806,
|
| 607 |
+
"grad_norm": 7.354248523712158,
|
| 608 |
"learning_rate": 4.152329231983852e-06,
|
| 609 |
+
"loss": 1.5259,
|
| 610 |
+
"mean_token_accuracy": 0.6074517369270325,
|
| 611 |
+
"num_tokens": 118880.0,
|
| 612 |
"step": 67
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.27419354838709675,
|
| 616 |
+
"grad_norm": 6.386468887329102,
|
| 617 |
"learning_rate": 4.128431206805556e-06,
|
| 618 |
+
"loss": 1.4579,
|
| 619 |
+
"mean_token_accuracy": 0.6266149878501892,
|
| 620 |
+
"num_tokens": 120430.0,
|
| 621 |
"step": 68
|
| 622 |
},
|
| 623 |
{
|
| 624 |
"epoch": 0.2782258064516129,
|
| 625 |
+
"grad_norm": 5.8017168045043945,
|
| 626 |
"learning_rate": 4.104271869242975e-06,
|
| 627 |
+
"loss": 1.5997,
|
| 628 |
+
"mean_token_accuracy": 0.5914376378059387,
|
| 629 |
+
"num_tokens": 122324.0,
|
| 630 |
"step": 69
|
| 631 |
},
|
| 632 |
{
|
| 633 |
"epoch": 0.28225806451612906,
|
| 634 |
+
"grad_norm": 6.298964500427246,
|
| 635 |
"learning_rate": 4.07985509611576e-06,
|
| 636 |
+
"loss": 1.5175,
|
| 637 |
+
"mean_token_accuracy": 0.6027742624282837,
|
| 638 |
+
"num_tokens": 123912.0,
|
| 639 |
"step": 70
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.2862903225806452,
|
| 643 |
+
"grad_norm": 5.610908031463623,
|
| 644 |
"learning_rate": 4.0551848055539345e-06,
|
| 645 |
+
"loss": 1.4782,
|
| 646 |
+
"mean_token_accuracy": 0.6146214008331299,
|
| 647 |
+
"num_tokens": 125829.0,
|
| 648 |
"step": 71
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 0.2903225806451613,
|
| 652 |
+
"grad_norm": 5.9410014152526855,
|
| 653 |
"learning_rate": 4.030264956369158e-06,
|
| 654 |
+
"loss": 1.5216,
|
| 655 |
+
"mean_token_accuracy": 0.6010044813156128,
|
| 656 |
+
"num_tokens": 127623.0,
|
| 657 |
"step": 72
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 0.29435483870967744,
|
| 661 |
+
"grad_norm": 6.027655601501465,
|
| 662 |
"learning_rate": 4.005099547419458e-06,
|
| 663 |
+
"loss": 1.5922,
|
| 664 |
+
"mean_token_accuracy": 0.604594349861145,
|
| 665 |
+
"num_tokens": 129671.0,
|
| 666 |
"step": 73
|
| 667 |
},
|
| 668 |
{
|
| 669 |
"epoch": 0.29838709677419356,
|
| 670 |
+
"grad_norm": 6.194504261016846,
|
| 671 |
"learning_rate": 3.979692616967543e-06,
|
| 672 |
+
"loss": 1.6518,
|
| 673 |
+
"mean_token_accuracy": 0.5671883225440979,
|
| 674 |
+
"num_tokens": 131526.0,
|
| 675 |
"step": 74
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.3024193548387097,
|
| 679 |
+
"grad_norm": 6.842575550079346,
|
| 680 |
"learning_rate": 3.9540482420327845e-06,
|
| 681 |
+
"loss": 1.7604,
|
| 682 |
+
"mean_token_accuracy": 0.5674124956130981,
|
| 683 |
+
"num_tokens": 133271.0,
|
| 684 |
"step": 75
|
| 685 |
},
|
| 686 |
{
|
| 687 |
"epoch": 0.3064516129032258,
|
| 688 |
+
"grad_norm": 7.071008682250977,
|
| 689 |
"learning_rate": 3.9281705377369814e-06,
|
| 690 |
+
"loss": 1.5444,
|
| 691 |
+
"mean_token_accuracy": 0.592570424079895,
|
| 692 |
+
"num_tokens": 134942.0,
|
| 693 |
"step": 76
|
| 694 |
},
|
| 695 |
{
|
| 696 |
"epoch": 0.31048387096774194,
|
| 697 |
+
"grad_norm": 6.561822414398193,
|
| 698 |
"learning_rate": 3.902063656644012e-06,
|
| 699 |
+
"loss": 1.6206,
|
| 700 |
+
"mean_token_accuracy": 0.5793358087539673,
|
| 701 |
+
"num_tokens": 136841.0,
|
| 702 |
"step": 77
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.31451612903225806,
|
| 706 |
+
"grad_norm": 6.077663421630859,
|
| 707 |
"learning_rate": 3.875731788093478e-06,
|
| 708 |
+
"loss": 1.5986,
|
| 709 |
+
"mean_token_accuracy": 0.591160237789154,
|
| 710 |
+
"num_tokens": 138653.0,
|
| 711 |
"step": 78
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 0.3185483870967742,
|
| 715 |
+
"grad_norm": 6.567859649658203,
|
| 716 |
"learning_rate": 3.84917915752845e-06,
|
| 717 |
+
"loss": 1.5392,
|
| 718 |
+
"mean_token_accuracy": 0.5913461446762085,
|
| 719 |
+
"num_tokens": 140111.0,
|
| 720 |
"step": 79
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 0.3225806451612903,
|
| 724 |
+
"grad_norm": 5.775151252746582,
|
| 725 |
"learning_rate": 3.8224100258174066e-06,
|
| 726 |
+
"loss": 1.5684,
|
| 727 |
+
"mean_token_accuracy": 0.5914602279663086,
|
| 728 |
+
"num_tokens": 142010.0,
|
| 729 |
"step": 80
|
| 730 |
},
|
| 731 |
{
|
| 732 |
"epoch": 0.32661290322580644,
|
| 733 |
+
"grad_norm": 6.075804233551025,
|
| 734 |
"learning_rate": 3.795428688570505e-06,
|
| 735 |
+
"loss": 1.5588,
|
| 736 |
+
"mean_token_accuracy": 0.5793555974960327,
|
| 737 |
+
"num_tokens": 143688.0,
|
| 738 |
"step": 81
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.33064516129032256,
|
| 742 |
+
"grad_norm": 6.182637691497803,
|
| 743 |
"learning_rate": 3.7682394754502687e-06,
|
| 744 |
+
"loss": 1.6354,
|
| 745 |
+
"mean_token_accuracy": 0.5826601982116699,
|
| 746 |
+
"num_tokens": 145547.0,
|
| 747 |
"step": 82
|
| 748 |
},
|
| 749 |
{
|
| 750 |
"epoch": 0.3346774193548387,
|
| 751 |
+
"grad_norm": 6.141689777374268,
|
| 752 |
"learning_rate": 3.7408467494768104e-06,
|
| 753 |
+
"loss": 1.5793,
|
| 754 |
+
"mean_token_accuracy": 0.5980230569839478,
|
| 755 |
+
"num_tokens": 147370.0,
|
| 756 |
"step": 83
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"epoch": 0.3387096774193548,
|
| 760 |
+
"grad_norm": 5.8948140144348145,
|
| 761 |
"learning_rate": 3.7132549063277033e-06,
|
| 762 |
+
"loss": 1.6276,
|
| 763 |
+
"mean_token_accuracy": 0.5829015374183655,
|
| 764 |
+
"num_tokens": 149302.0,
|
| 765 |
"step": 84
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.34274193548387094,
|
| 769 |
+
"grad_norm": 6.238958358764648,
|
| 770 |
"learning_rate": 3.685468373632613e-06,
|
| 771 |
+
"loss": 1.6904,
|
| 772 |
+
"mean_token_accuracy": 0.5823293328285217,
|
| 773 |
+
"num_tokens": 151047.0,
|
| 774 |
"step": 85
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 0.3467741935483871,
|
| 778 |
+
"grad_norm": 5.967896461486816,
|
| 779 |
"learning_rate": 3.657491610262802e-06,
|
| 780 |
+
"loss": 1.5618,
|
| 781 |
+
"mean_token_accuracy": 0.5967926979064941,
|
| 782 |
+
"num_tokens": 152795.0,
|
| 783 |
"step": 86
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 0.35080645161290325,
|
| 787 |
+
"grad_norm": 5.502566337585449,
|
| 788 |
"learning_rate": 3.6293291056156178e-06,
|
| 789 |
+
"loss": 1.5015,
|
| 790 |
+
"mean_token_accuracy": 0.5971802473068237,
|
| 791 |
+
"num_tokens": 154783.0,
|
| 792 |
"step": 87
|
| 793 |
},
|
| 794 |
{
|
| 795 |
"epoch": 0.3548387096774194,
|
| 796 |
+
"grad_norm": 5.598520755767822,
|
| 797 |
"learning_rate": 3.600985378894086e-06,
|
| 798 |
+
"loss": 1.6132,
|
| 799 |
+
"mean_token_accuracy": 0.5847665667533875,
|
| 800 |
+
"num_tokens": 156820.0,
|
| 801 |
"step": 88
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.3588709677419355,
|
| 805 |
+
"grad_norm": 5.761752605438232,
|
| 806 |
"learning_rate": 3.572464978381719e-06,
|
| 807 |
+
"loss": 1.5378,
|
| 808 |
+
"mean_token_accuracy": 0.5998993515968323,
|
| 809 |
+
"num_tokens": 158809.0,
|
| 810 |
"step": 89
|
| 811 |
},
|
| 812 |
{
|
| 813 |
"epoch": 0.3629032258064516,
|
| 814 |
+
"grad_norm": 6.394772529602051,
|
| 815 |
"learning_rate": 3.5437724807126583e-06,
|
| 816 |
+
"loss": 1.5489,
|
| 817 |
+
"mean_token_accuracy": 0.597762405872345,
|
| 818 |
+
"num_tokens": 160688.0,
|
| 819 |
"step": 90
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.36693548387096775,
|
| 823 |
+
"grad_norm": 5.892634868621826,
|
| 824 |
"learning_rate": 3.514912490137268e-06,
|
| 825 |
+
"loss": 1.3987,
|
| 826 |
+
"mean_token_accuracy": 0.634549617767334,
|
| 827 |
+
"num_tokens": 162444.0,
|
| 828 |
"step": 91
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.3709677419354839,
|
| 832 |
+
"grad_norm": 6.148153305053711,
|
| 833 |
"learning_rate": 3.4858896377832966e-06,
|
| 834 |
+
"loss": 1.5568,
|
| 835 |
+
"mean_token_accuracy": 0.5997700095176697,
|
| 836 |
+
"num_tokens": 164185.0,
|
| 837 |
"step": 92
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 0.375,
|
| 841 |
+
"grad_norm": 5.827703952789307,
|
| 842 |
"learning_rate": 3.4567085809127247e-06,
|
| 843 |
+
"loss": 1.6176,
|
| 844 |
+
"mean_token_accuracy": 0.6020779013633728,
|
| 845 |
+
"num_tokens": 166112.0,
|
| 846 |
"step": 93
|
| 847 |
},
|
| 848 |
{
|
| 849 |
"epoch": 0.3790322580645161,
|
| 850 |
+
"grad_norm": 5.979760646820068,
|
| 851 |
"learning_rate": 3.42737400217442e-06,
|
| 852 |
+
"loss": 1.6762,
|
| 853 |
+
"mean_token_accuracy": 0.574064314365387,
|
| 854 |
+
"num_tokens": 168011.0,
|
| 855 |
"step": 94
|
| 856 |
},
|
| 857 |
{
|
| 858 |
"epoch": 0.38306451612903225,
|
| 859 |
+
"grad_norm": 6.755500793457031,
|
| 860 |
"learning_rate": 3.397890608852718e-06,
|
| 861 |
+
"loss": 1.6051,
|
| 862 |
+
"mean_token_accuracy": 0.5879656076431274,
|
| 863 |
+
"num_tokens": 169758.0,
|
| 864 |
"step": 95
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.3870967741935484,
|
| 868 |
+
"grad_norm": 6.7287068367004395,
|
| 869 |
"learning_rate": 3.3682631321120507e-06,
|
| 870 |
+
"loss": 1.5288,
|
| 871 |
+
"mean_token_accuracy": 0.6041095852851868,
|
| 872 |
+
"num_tokens": 171220.0,
|
| 873 |
"step": 96
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.3911290322580645,
|
| 877 |
+
"grad_norm": 6.124728679656982,
|
| 878 |
"learning_rate": 3.3384963262377434e-06,
|
| 879 |
+
"loss": 1.5012,
|
| 880 |
+
"mean_token_accuracy": 0.5954991579055786,
|
| 881 |
+
"num_tokens": 172955.0,
|
| 882 |
"step": 97
|
| 883 |
},
|
| 884 |
{
|
| 885 |
"epoch": 0.3951612903225806,
|
| 886 |
+
"grad_norm": 5.964875221252441,
|
| 887 |
"learning_rate": 3.3085949678730953e-06,
|
| 888 |
+
"loss": 1.4252,
|
| 889 |
+
"mean_token_accuracy": 0.6224677562713623,
|
| 890 |
+
"num_tokens": 174586.0,
|
| 891 |
"step": 98
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.39919354838709675,
|
| 895 |
+
"grad_norm": 6.314966678619385,
|
| 896 |
"learning_rate": 3.278563855252885e-06,
|
| 897 |
+
"loss": 1.4519,
|
| 898 |
+
"mean_token_accuracy": 0.630832850933075,
|
| 899 |
+
"num_tokens": 176281.0,
|
| 900 |
"step": 99
|
| 901 |
},
|
| 902 |
{
|
| 903 |
"epoch": 0.4032258064516129,
|
| 904 |
+
"grad_norm": 6.121771812438965,
|
| 905 |
"learning_rate": 3.248407807433396e-06,
|
| 906 |
+
"loss": 1.541,
|
| 907 |
+
"mean_token_accuracy": 0.5919907093048096,
|
| 908 |
+
"num_tokens": 178006.0,
|
| 909 |
"step": 100
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 0.40725806451612906,
|
| 913 |
+
"grad_norm": 6.246324062347412,
|
| 914 |
"learning_rate": 3.2181316635191125e-06,
|
| 915 |
+
"loss": 1.4925,
|
| 916 |
+
"mean_token_accuracy": 0.6121912598609924,
|
| 917 |
+
"num_tokens": 179911.0,
|
| 918 |
"step": 101
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 0.4112903225806452,
|
| 922 |
+
"grad_norm": 5.7846221923828125,
|
| 923 |
"learning_rate": 3.1877402818861954e-06,
|
| 924 |
+
"loss": 1.5096,
|
| 925 |
+
"mean_token_accuracy": 0.6097561120986938,
|
| 926 |
+
"num_tokens": 181799.0,
|
| 927 |
"step": 102
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.4153225806451613,
|
| 931 |
+
"grad_norm": 5.7426252365112305,
|
| 932 |
"learning_rate": 3.157238539402862e-06,
|
| 933 |
+
"loss": 1.4982,
|
| 934 |
+
"mean_token_accuracy": 0.5930666923522949,
|
| 935 |
+
"num_tokens": 183676.0,
|
| 936 |
"step": 103
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 0.41935483870967744,
|
| 940 |
+
"grad_norm": 6.007601261138916,
|
| 941 |
"learning_rate": 3.1266313306468018e-06,
|
| 942 |
+
"loss": 1.515,
|
| 943 |
+
"mean_token_accuracy": 0.6020166277885437,
|
| 944 |
+
"num_tokens": 185364.0,
|
| 945 |
"step": 104
|
| 946 |
},
|
| 947 |
{
|
| 948 |
"epoch": 0.42338709677419356,
|
| 949 |
+
"grad_norm": 6.339498519897461,
|
| 950 |
"learning_rate": 3.095923567119748e-06,
|
| 951 |
+
"loss": 1.5713,
|
| 952 |
+
"mean_token_accuracy": 0.5824423432350159,
|
| 953 |
+
"num_tokens": 187143.0,
|
| 954 |
"step": 105
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.4274193548387097,
|
| 958 |
+
"grad_norm": 6.260862827301025,
|
| 959 |
"learning_rate": 3.0651201764593375e-06,
|
| 960 |
+
"loss": 1.6099,
|
| 961 |
+
"mean_token_accuracy": 0.5965318083763123,
|
| 962 |
+
"num_tokens": 188875.0,
|
| 963 |
"step": 106
|
| 964 |
},
|
| 965 |
{
|
| 966 |
"epoch": 0.4314516129032258,
|
| 967 |
+
"grad_norm": 5.941509246826172,
|
| 968 |
"learning_rate": 3.034226101648377e-06,
|
| 969 |
+
"loss": 1.6252,
|
| 970 |
+
"mean_token_accuracy": 0.586152195930481,
|
| 971 |
+
"num_tokens": 190769.0,
|
| 972 |
"step": 107
|
| 973 |
},
|
| 974 |
{
|
| 975 |
"epoch": 0.43548387096774194,
|
| 976 |
+
"grad_norm": 5.4027509689331055,
|
| 977 |
"learning_rate": 3.0032463002216504e-06,
|
| 978 |
+
"loss": 1.3669,
|
| 979 |
+
"mean_token_accuracy": 0.6161772012710571,
|
| 980 |
+
"num_tokens": 192712.0,
|
| 981 |
"step": 108
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 0.43951612903225806,
|
| 985 |
+
"grad_norm": 5.507197856903076,
|
| 986 |
"learning_rate": 2.972185743470386e-06,
|
| 987 |
+
"loss": 1.5409,
|
| 988 |
+
"mean_token_accuracy": 0.6028755307197571,
|
| 989 |
+
"num_tokens": 194731.0,
|
| 990 |
"step": 109
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.4435483870967742,
|
| 994 |
+
"grad_norm": 6.563897609710693,
|
| 995 |
"learning_rate": 2.941049415644522e-06,
|
| 996 |
+
"loss": 1.5292,
|
| 997 |
+
"mean_token_accuracy": 0.6081171035766602,
|
| 998 |
+
"num_tokens": 196236.0,
|
| 999 |
"step": 110
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 0.4475806451612903,
|
| 1003 |
+
"grad_norm": 6.108341693878174,
|
| 1004 |
"learning_rate": 2.909842313152888e-06,
|
| 1005 |
+
"loss": 1.5694,
|
| 1006 |
+
"mean_token_accuracy": 0.5977421402931213,
|
| 1007 |
+
"num_tokens": 197921.0,
|
| 1008 |
"step": 111
|
| 1009 |
},
|
| 1010 |
{
|
| 1011 |
"epoch": 0.45161290322580644,
|
| 1012 |
+
"grad_norm": 5.4269256591796875,
|
| 1013 |
"learning_rate": 2.878569443761442e-06,
|
| 1014 |
+
"loss": 1.4808,
|
| 1015 |
+
"mean_token_accuracy": 0.6034912467002869,
|
| 1016 |
+
"num_tokens": 199928.0,
|
| 1017 |
"step": 112
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.45564516129032256,
|
| 1021 |
+
"grad_norm": 6.073188304901123,
|
| 1022 |
"learning_rate": 2.847235825789673e-06,
|
| 1023 |
+
"loss": 1.5251,
|
| 1024 |
+
"mean_token_accuracy": 0.6071428656578064,
|
| 1025 |
+
"num_tokens": 201750.0,
|
| 1026 |
"step": 113
|
| 1027 |
},
|
| 1028 |
{
|
| 1029 |
"epoch": 0.4596774193548387,
|
| 1030 |
+
"grad_norm": 6.206169605255127,
|
| 1031 |
"learning_rate": 2.8158464873053236e-06,
|
| 1032 |
+
"loss": 1.6275,
|
| 1033 |
+
"mean_token_accuracy": 0.5718509554862976,
|
| 1034 |
+
"num_tokens": 203443.0,
|
| 1035 |
"step": 114
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
"epoch": 0.4637096774193548,
|
| 1039 |
+
"grad_norm": 6.5099616050720215,
|
| 1040 |
"learning_rate": 2.784406465317538e-06,
|
| 1041 |
+
"loss": 1.4364,
|
| 1042 |
+
"mean_token_accuracy": 0.6256089210510254,
|
| 1043 |
+
"num_tokens": 204882.0,
|
| 1044 |
"step": 115
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.46774193548387094,
|
| 1048 |
+
"grad_norm": 5.677121639251709,
|
| 1049 |
"learning_rate": 2.752920804968581e-06,
|
| 1050 |
+
"loss": 1.515,
|
| 1051 |
+
"mean_token_accuracy": 0.5985877513885498,
|
| 1052 |
+
"num_tokens": 206725.0,
|
| 1053 |
"step": 116
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
"epoch": 0.4717741935483871,
|
| 1057 |
+
"grad_norm": 5.99458646774292,
|
| 1058 |
"learning_rate": 2.7213945587242507e-06,
|
| 1059 |
+
"loss": 1.6479,
|
| 1060 |
+
"mean_token_accuracy": 0.5737704634666443,
|
| 1061 |
+
"num_tokens": 208740.0,
|
| 1062 |
"step": 117
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 0.47580645161290325,
|
| 1066 |
+
"grad_norm": 5.447024822235107,
|
| 1067 |
"learning_rate": 2.689832785563116e-06,
|
| 1068 |
+
"loss": 1.5805,
|
| 1069 |
+
"mean_token_accuracy": 0.5884652733802795,
|
| 1070 |
+
"num_tokens": 210788.0,
|
| 1071 |
"step": 118
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
"epoch": 0.4798387096774194,
|
| 1075 |
+
"grad_norm": 5.383421421051025,
|
| 1076 |
"learning_rate": 2.658240550164704e-06,
|
| 1077 |
+
"loss": 1.5772,
|
| 1078 |
+
"mean_token_accuracy": 0.5977517366409302,
|
| 1079 |
+
"num_tokens": 212836.0,
|
| 1080 |
"step": 119
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.4838709677419355,
|
| 1084 |
+
"grad_norm": 6.088338375091553,
|
| 1085 |
"learning_rate": 2.626622922096782e-06,
|
| 1086 |
+
"loss": 1.565,
|
| 1087 |
+
"mean_token_accuracy": 0.5961331725120544,
|
| 1088 |
+
"num_tokens": 214700.0,
|
| 1089 |
"step": 120
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
"epoch": 0.4879032258064516,
|
| 1093 |
+
"grad_norm": 5.788651466369629,
|
| 1094 |
"learning_rate": 2.5949849750018486e-06,
|
| 1095 |
+
"loss": 1.3632,
|
| 1096 |
+
"mean_token_accuracy": 0.6446384191513062,
|
| 1097 |
+
"num_tokens": 216306.0,
|
| 1098 |
"step": 121
|
| 1099 |
},
|
| 1100 |
{
|
| 1101 |
"epoch": 0.49193548387096775,
|
| 1102 |
+
"grad_norm": 5.829148769378662,
|
| 1103 |
"learning_rate": 2.56333178578297e-06,
|
| 1104 |
+
"loss": 1.6738,
|
| 1105 |
+
"mean_token_accuracy": 0.574331521987915,
|
| 1106 |
+
"num_tokens": 218178.0,
|
| 1107 |
"step": 122
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
"epoch": 0.4959677419354839,
|
| 1111 |
+
"grad_norm": 5.958081245422363,
|
| 1112 |
"learning_rate": 2.5316684337891005e-06,
|
| 1113 |
+
"loss": 1.3764,
|
| 1114 |
+
"mean_token_accuracy": 0.6442367434501648,
|
| 1115 |
+
"num_tokens": 219785.0,
|
| 1116 |
"step": 123
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 0.5,
|
| 1120 |
+
"grad_norm": 6.186583995819092,
|
| 1121 |
"learning_rate": 2.5e-06,
|
| 1122 |
+
"loss": 1.4931,
|
| 1123 |
+
"mean_token_accuracy": 0.6139564514160156,
|
| 1124 |
+
"num_tokens": 221349.0,
|
| 1125 |
"step": 124
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
"epoch": 0.5040322580645161,
|
| 1129 |
+
"grad_norm": 5.803173542022705,
|
| 1130 |
"learning_rate": 2.4683315662109003e-06,
|
| 1131 |
+
"loss": 1.5739,
|
| 1132 |
+
"mean_token_accuracy": 0.58859783411026,
|
| 1133 |
+
"num_tokens": 223298.0,
|
| 1134 |
"step": 125
|
| 1135 |
},
|
| 1136 |
{
|
| 1137 |
"epoch": 0.5080645161290323,
|
| 1138 |
+
"grad_norm": 5.776984214782715,
|
| 1139 |
"learning_rate": 2.436668214217031e-06,
|
| 1140 |
+
"loss": 1.5603,
|
| 1141 |
+
"mean_token_accuracy": 0.5867098569869995,
|
| 1142 |
+
"num_tokens": 225151.0,
|
| 1143 |
"step": 126
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.5120967741935484,
|
| 1147 |
+
"grad_norm": 5.5442094802856445,
|
| 1148 |
"learning_rate": 2.4050150249981522e-06,
|
| 1149 |
+
"loss": 1.4753,
|
| 1150 |
+
"mean_token_accuracy": 0.6018254160881042,
|
| 1151 |
+
"num_tokens": 226906.0,
|
| 1152 |
"step": 127
|
| 1153 |
},
|
| 1154 |
{
|
| 1155 |
"epoch": 0.5161290322580645,
|
| 1156 |
+
"grad_norm": 5.492681503295898,
|
| 1157 |
"learning_rate": 2.3733770779032185e-06,
|
| 1158 |
+
"loss": 1.6577,
|
| 1159 |
+
"mean_token_accuracy": 0.5733202695846558,
|
| 1160 |
+
"num_tokens": 228947.0,
|
| 1161 |
"step": 128
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.5201612903225806,
|
| 1165 |
+
"grad_norm": 5.890382766723633,
|
| 1166 |
"learning_rate": 2.341759449835297e-06,
|
| 1167 |
+
"loss": 1.5399,
|
| 1168 |
+
"mean_token_accuracy": 0.5982404947280884,
|
| 1169 |
+
"num_tokens": 230995.0,
|
| 1170 |
"step": 129
|
| 1171 |
},
|
| 1172 |
{
|
| 1173 |
"epoch": 0.5241935483870968,
|
| 1174 |
+
"grad_norm": 5.885677337646484,
|
| 1175 |
"learning_rate": 2.310167214436885e-06,
|
| 1176 |
+
"loss": 1.5541,
|
| 1177 |
+
"mean_token_accuracy": 0.6021798253059387,
|
| 1178 |
+
"num_tokens": 232832.0,
|
| 1179 |
"step": 130
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.5282258064516129,
|
| 1183 |
+
"grad_norm": 5.711495876312256,
|
| 1184 |
"learning_rate": 2.27860544127575e-06,
|
| 1185 |
+
"loss": 1.5311,
|
| 1186 |
+
"mean_token_accuracy": 0.5873016119003296,
|
| 1187 |
+
"num_tokens": 234598.0,
|
| 1188 |
"step": 131
|
| 1189 |
},
|
| 1190 |
{
|
| 1191 |
"epoch": 0.532258064516129,
|
| 1192 |
+
"grad_norm": 5.489191055297852,
|
| 1193 |
"learning_rate": 2.24707919503142e-06,
|
| 1194 |
+
"loss": 1.4668,
|
| 1195 |
+
"mean_token_accuracy": 0.6193058490753174,
|
| 1196 |
+
"num_tokens": 236444.0,
|
| 1197 |
"step": 132
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
"epoch": 0.5362903225806451,
|
| 1201 |
+
"grad_norm": 5.6918110847473145,
|
| 1202 |
"learning_rate": 2.2155935346824634e-06,
|
| 1203 |
+
"loss": 1.5273,
|
| 1204 |
+
"mean_token_accuracy": 0.6107238531112671,
|
| 1205 |
+
"num_tokens": 238311.0,
|
| 1206 |
"step": 133
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.5403225806451613,
|
| 1210 |
+
"grad_norm": 6.0905938148498535,
|
| 1211 |
"learning_rate": 2.1841535126946777e-06,
|
| 1212 |
+
"loss": 1.5814,
|
| 1213 |
+
"mean_token_accuracy": 0.6002208590507507,
|
| 1214 |
+
"num_tokens": 240124.0,
|
| 1215 |
"step": 134
|
| 1216 |
},
|
| 1217 |
{
|
| 1218 |
"epoch": 0.5443548387096774,
|
| 1219 |
+
"grad_norm": 5.952730178833008,
|
| 1220 |
"learning_rate": 2.1527641742103282e-06,
|
| 1221 |
+
"loss": 1.4447,
|
| 1222 |
+
"mean_token_accuracy": 0.6165267825126648,
|
| 1223 |
+
"num_tokens": 241675.0,
|
| 1224 |
"step": 135
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 0.5483870967741935,
|
| 1228 |
+
"grad_norm": 6.087653160095215,
|
| 1229 |
"learning_rate": 2.1214305562385592e-06,
|
| 1230 |
+
"loss": 1.6968,
|
| 1231 |
+
"mean_token_accuracy": 0.5795266628265381,
|
| 1232 |
+
"num_tokens": 243494.0,
|
| 1233 |
"step": 136
|
| 1234 |
},
|
| 1235 |
{
|
| 1236 |
"epoch": 0.5524193548387096,
|
| 1237 |
+
"grad_norm": 5.839648246765137,
|
| 1238 |
"learning_rate": 2.0901576868471125e-06,
|
| 1239 |
+
"loss": 1.7217,
|
| 1240 |
+
"mean_token_accuracy": 0.5630123019218445,
|
| 1241 |
+
"num_tokens": 245448.0,
|
| 1242 |
"step": 137
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.5564516129032258,
|
| 1246 |
+
"grad_norm": 5.999613285064697,
|
| 1247 |
"learning_rate": 2.05895058435548e-06,
|
| 1248 |
+
"loss": 1.3948,
|
| 1249 |
+
"mean_token_accuracy": 0.6161738038063049,
|
| 1250 |
+
"num_tokens": 247107.0,
|
| 1251 |
"step": 138
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
"epoch": 0.5604838709677419,
|
| 1255 |
+
"grad_norm": 5.091090202331543,
|
| 1256 |
"learning_rate": 2.0278142565296153e-06,
|
| 1257 |
+
"loss": 1.4922,
|
| 1258 |
+
"mean_token_accuracy": 0.607038140296936,
|
| 1259 |
+
"num_tokens": 249155.0,
|
| 1260 |
"step": 139
|
| 1261 |
},
|
| 1262 |
{
|
| 1263 |
"epoch": 0.5645161290322581,
|
| 1264 |
+
"grad_norm": 5.5148539543151855,
|
| 1265 |
"learning_rate": 1.9967536997783495e-06,
|
| 1266 |
+
"loss": 1.5327,
|
| 1267 |
+
"mean_token_accuracy": 0.6058357954025269,
|
| 1268 |
+
"num_tokens": 251179.0,
|
| 1269 |
"step": 140
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.5685483870967742,
|
| 1273 |
+
"grad_norm": 6.003233432769775,
|
| 1274 |
"learning_rate": 1.9657738983516227e-06,
|
| 1275 |
+
"loss": 1.5269,
|
| 1276 |
+
"mean_token_accuracy": 0.592024564743042,
|
| 1277 |
+
"num_tokens": 253137.0,
|
| 1278 |
"step": 141
|
| 1279 |
},
|
| 1280 |
{
|
| 1281 |
"epoch": 0.5725806451612904,
|
| 1282 |
+
"grad_norm": 5.6866068840026855,
|
| 1283 |
"learning_rate": 1.934879823540663e-06,
|
| 1284 |
+
"loss": 1.4626,
|
| 1285 |
+
"mean_token_accuracy": 0.6116965413093567,
|
| 1286 |
+
"num_tokens": 255037.0,
|
| 1287 |
"step": 142
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 0.5766129032258065,
|
| 1291 |
+
"grad_norm": 5.541101455688477,
|
| 1292 |
"learning_rate": 1.9040764328802523e-06,
|
| 1293 |
+
"loss": 1.3471,
|
| 1294 |
+
"mean_token_accuracy": 0.6364172101020813,
|
| 1295 |
+
"num_tokens": 256736.0,
|
| 1296 |
"step": 143
|
| 1297 |
},
|
| 1298 |
{
|
| 1299 |
"epoch": 0.5806451612903226,
|
| 1300 |
+
"grad_norm": 5.830052375793457,
|
| 1301 |
"learning_rate": 1.8733686693531986e-06,
|
| 1302 |
+
"loss": 1.5507,
|
| 1303 |
+
"mean_token_accuracy": 0.5932571887969971,
|
| 1304 |
+
"num_tokens": 258577.0,
|
| 1305 |
"step": 144
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.5846774193548387,
|
| 1309 |
+
"grad_norm": 6.131454944610596,
|
| 1310 |
"learning_rate": 1.842761460597138e-06,
|
| 1311 |
+
"loss": 1.4502,
|
| 1312 |
+
"mean_token_accuracy": 0.633840024471283,
|
| 1313 |
+
"num_tokens": 260092.0,
|
| 1314 |
"step": 145
|
| 1315 |
},
|
| 1316 |
{
|
| 1317 |
"epoch": 0.5887096774193549,
|
| 1318 |
+
"grad_norm": 7.0200324058532715,
|
| 1319 |
"learning_rate": 1.812259718113805e-06,
|
| 1320 |
+
"loss": 1.5706,
|
| 1321 |
+
"mean_token_accuracy": 0.5800653696060181,
|
| 1322 |
+
"num_tokens": 261930.0,
|
| 1323 |
"step": 146
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
"epoch": 0.592741935483871,
|
| 1327 |
+
"grad_norm": 6.529550552368164,
|
| 1328 |
"learning_rate": 1.7818683364808883e-06,
|
| 1329 |
+
"loss": 1.6779,
|
| 1330 |
+
"mean_token_accuracy": 0.5723379850387573,
|
| 1331 |
+
"num_tokens": 263660.0,
|
| 1332 |
"step": 147
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
"epoch": 0.5967741935483871,
|
| 1336 |
+
"grad_norm": 6.201499938964844,
|
| 1337 |
"learning_rate": 1.7515921925666053e-06,
|
| 1338 |
+
"loss": 1.5537,
|
| 1339 |
+
"mean_token_accuracy": 0.5979142785072327,
|
| 1340 |
+
"num_tokens": 265388.0,
|
| 1341 |
"step": 148
|
| 1342 |
},
|
| 1343 |
{
|
| 1344 |
"epoch": 0.6008064516129032,
|
| 1345 |
+
"grad_norm": 6.028550624847412,
|
| 1346 |
"learning_rate": 1.7214361447471156e-06,
|
| 1347 |
+
"loss": 1.5809,
|
| 1348 |
+
"mean_token_accuracy": 0.5865436792373657,
|
| 1349 |
+
"num_tokens": 267233.0,
|
| 1350 |
"step": 149
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 0.6048387096774194,
|
| 1354 |
+
"grad_norm": 5.58845853805542,
|
| 1355 |
"learning_rate": 1.6914050321269049e-06,
|
| 1356 |
+
"loss": 1.4061,
|
| 1357 |
+
"mean_token_accuracy": 0.623369574546814,
|
| 1358 |
+
"num_tokens": 269075.0,
|
| 1359 |
"step": 150
|
| 1360 |
},
|
| 1361 |
{
|
| 1362 |
"epoch": 0.6088709677419355,
|
| 1363 |
+
"grad_norm": 6.210185527801514,
|
| 1364 |
"learning_rate": 1.6615036737622574e-06,
|
| 1365 |
+
"loss": 1.6096,
|
| 1366 |
+
"mean_token_accuracy": 0.5861111283302307,
|
| 1367 |
+
"num_tokens": 270877.0,
|
| 1368 |
"step": 151
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 0.6129032258064516,
|
| 1372 |
+
"grad_norm": 6.132157325744629,
|
| 1373 |
"learning_rate": 1.6317368678879497e-06,
|
| 1374 |
+
"loss": 1.6248,
|
| 1375 |
+
"mean_token_accuracy": 0.5724177956581116,
|
| 1376 |
+
"num_tokens": 272612.0,
|
| 1377 |
"step": 152
|
| 1378 |
},
|
| 1379 |
{
|
| 1380 |
"epoch": 0.6169354838709677,
|
| 1381 |
+
"grad_norm": 6.025227069854736,
|
| 1382 |
"learning_rate": 1.6021093911472825e-06,
|
| 1383 |
+
"loss": 1.5701,
|
| 1384 |
+
"mean_token_accuracy": 0.5928652286529541,
|
| 1385 |
+
"num_tokens": 274380.0,
|
| 1386 |
"step": 153
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
"epoch": 0.6209677419354839,
|
| 1390 |
+
"grad_norm": 5.912363529205322,
|
| 1391 |
"learning_rate": 1.572625997825581e-06,
|
| 1392 |
+
"loss": 1.4989,
|
| 1393 |
+
"mean_token_accuracy": 0.5981630086898804,
|
| 1394 |
+
"num_tokens": 276124.0,
|
| 1395 |
"step": 154
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
"epoch": 0.625,
|
| 1399 |
+
"grad_norm": 5.779432773590088,
|
| 1400 |
"learning_rate": 1.5432914190872757e-06,
|
| 1401 |
+
"loss": 1.4394,
|
| 1402 |
+
"mean_token_accuracy": 0.6316964030265808,
|
| 1403 |
+
"num_tokens": 277918.0,
|
| 1404 |
"step": 155
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 0.6290322580645161,
|
| 1408 |
+
"grad_norm": 5.328639984130859,
|
| 1409 |
"learning_rate": 1.5141103622167042e-06,
|
| 1410 |
+
"loss": 1.4466,
|
| 1411 |
+
"mean_token_accuracy": 0.6170212626457214,
|
| 1412 |
+
"num_tokens": 279800.0,
|
| 1413 |
"step": 156
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 0.6330645161290323,
|
| 1417 |
+
"grad_norm": 6.32019567489624,
|
| 1418 |
"learning_rate": 1.4850875098627326e-06,
|
| 1419 |
+
"loss": 1.5899,
|
| 1420 |
+
"mean_token_accuracy": 0.6091743111610413,
|
| 1421 |
+
"num_tokens": 281437.0,
|
| 1422 |
"step": 157
|
| 1423 |
},
|
| 1424 |
{
|
| 1425 |
"epoch": 0.6370967741935484,
|
| 1426 |
+
"grad_norm": 5.477099418640137,
|
| 1427 |
"learning_rate": 1.456227519287343e-06,
|
| 1428 |
+
"loss": 1.5322,
|
| 1429 |
+
"mean_token_accuracy": 0.5986914038658142,
|
| 1430 |
+
"num_tokens": 283273.0,
|
| 1431 |
"step": 158
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 0.6411290322580645,
|
| 1435 |
+
"grad_norm": 6.026065349578857,
|
| 1436 |
"learning_rate": 1.4275350216182824e-06,
|
| 1437 |
+
"loss": 1.3874,
|
| 1438 |
+
"mean_token_accuracy": 0.6260971426963806,
|
| 1439 |
+
"num_tokens": 284984.0,
|
| 1440 |
"step": 159
|
| 1441 |
},
|
| 1442 |
{
|
| 1443 |
"epoch": 0.6451612903225806,
|
| 1444 |
+
"grad_norm": 5.370904922485352,
|
| 1445 |
"learning_rate": 1.3990146211059141e-06,
|
| 1446 |
+
"loss": 1.5299,
|
| 1447 |
+
"mean_token_accuracy": 0.6030451059341431,
|
| 1448 |
+
"num_tokens": 286825.0,
|
| 1449 |
"step": 160
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.6491935483870968,
|
| 1453 |
+
"grad_norm": 5.8816633224487305,
|
| 1454 |
"learning_rate": 1.3706708943843822e-06,
|
| 1455 |
+
"loss": 1.4889,
|
| 1456 |
+
"mean_token_accuracy": 0.6019198298454285,
|
| 1457 |
+
"num_tokens": 288598.0,
|
| 1458 |
"step": 161
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 0.6532258064516129,
|
| 1462 |
+
"grad_norm": 6.100244522094727,
|
| 1463 |
"learning_rate": 1.3425083897371983e-06,
|
| 1464 |
+
"loss": 1.4759,
|
| 1465 |
+
"mean_token_accuracy": 0.6230722069740295,
|
| 1466 |
+
"num_tokens": 290221.0,
|
| 1467 |
"step": 162
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.657258064516129,
|
| 1471 |
+
"grad_norm": 5.920684337615967,
|
| 1472 |
"learning_rate": 1.3145316263673874e-06,
|
| 1473 |
+
"loss": 1.5585,
|
| 1474 |
+
"mean_token_accuracy": 0.5871710777282715,
|
| 1475 |
+
"num_tokens": 292047.0,
|
| 1476 |
"step": 163
|
| 1477 |
},
|
| 1478 |
{
|
| 1479 |
"epoch": 0.6612903225806451,
|
| 1480 |
+
"grad_norm": 6.059237003326416,
|
| 1481 |
"learning_rate": 1.286745093672298e-06,
|
| 1482 |
+
"loss": 1.5276,
|
| 1483 |
+
"mean_token_accuracy": 0.6024376153945923,
|
| 1484 |
+
"num_tokens": 293772.0,
|
| 1485 |
"step": 164
|
| 1486 |
},
|
| 1487 |
{
|
| 1488 |
"epoch": 0.6653225806451613,
|
| 1489 |
+
"grad_norm": 5.7648701667785645,
|
| 1490 |
"learning_rate": 1.2591532505231906e-06,
|
| 1491 |
+
"loss": 1.5058,
|
| 1492 |
+
"mean_token_accuracy": 0.5996025800704956,
|
| 1493 |
+
"num_tokens": 295787.0,
|
| 1494 |
"step": 165
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
"epoch": 0.6693548387096774,
|
| 1498 |
+
"grad_norm": 5.578222274780273,
|
| 1499 |
"learning_rate": 1.2317605245497324e-06,
|
| 1500 |
+
"loss": 1.5277,
|
| 1501 |
+
"mean_token_accuracy": 0.5919329524040222,
|
| 1502 |
+
"num_tokens": 297698.0,
|
| 1503 |
"step": 166
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 0.6733870967741935,
|
| 1507 |
+
"grad_norm": 5.404395580291748,
|
| 1508 |
"learning_rate": 1.204571311429496e-06,
|
| 1509 |
+
"loss": 1.4855,
|
| 1510 |
+
"mean_token_accuracy": 0.5967162847518921,
|
| 1511 |
+
"num_tokens": 299649.0,
|
| 1512 |
"step": 167
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 0.6774193548387096,
|
| 1516 |
+
"grad_norm": 6.532902240753174,
|
| 1517 |
"learning_rate": 1.1775899741825947e-06,
|
| 1518 |
+
"loss": 1.4942,
|
| 1519 |
+
"mean_token_accuracy": 0.5987738370895386,
|
| 1520 |
+
"num_tokens": 301119.0,
|
| 1521 |
"step": 168
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 0.6814516129032258,
|
| 1525 |
+
"grad_norm": 6.508765697479248,
|
| 1526 |
"learning_rate": 1.1508208424715511e-06,
|
| 1527 |
+
"loss": 1.5848,
|
| 1528 |
+
"mean_token_accuracy": 0.5865746736526489,
|
| 1529 |
+
"num_tokens": 302715.0,
|
| 1530 |
"step": 169
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 0.6854838709677419,
|
| 1534 |
+
"grad_norm": 5.411791801452637,
|
| 1535 |
"learning_rate": 1.1242682119065217e-06,
|
| 1536 |
+
"loss": 1.4479,
|
| 1537 |
+
"mean_token_accuracy": 0.6116002202033997,
|
| 1538 |
+
"num_tokens": 304648.0,
|
| 1539 |
"step": 170
|
| 1540 |
},
|
| 1541 |
{
|
| 1542 |
"epoch": 0.6895161290322581,
|
| 1543 |
+
"grad_norm": 6.0362548828125,
|
| 1544 |
"learning_rate": 1.0979363433559892e-06,
|
| 1545 |
+
"loss": 1.5337,
|
| 1546 |
+
"mean_token_accuracy": 0.5946902632713318,
|
| 1547 |
+
"num_tokens": 306345.0,
|
| 1548 |
"step": 171
|
| 1549 |
},
|
| 1550 |
{
|
| 1551 |
"epoch": 0.6935483870967742,
|
| 1552 |
+
"grad_norm": 6.6355743408203125,
|
| 1553 |
"learning_rate": 1.0718294622630188e-06,
|
| 1554 |
+
"loss": 1.5346,
|
| 1555 |
+
"mean_token_accuracy": 0.6140100359916687,
|
| 1556 |
+
"num_tokens": 307746.0,
|
| 1557 |
"step": 172
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
"epoch": 0.6975806451612904,
|
| 1561 |
+
"grad_norm": 5.8390421867370605,
|
| 1562 |
"learning_rate": 1.045951757967215e-06,
|
| 1563 |
+
"loss": 1.5117,
|
| 1564 |
+
"mean_token_accuracy": 0.6113694906234741,
|
| 1565 |
+
"num_tokens": 309683.0,
|
| 1566 |
"step": 173
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 0.7016129032258065,
|
| 1570 |
+
"grad_norm": 6.209187030792236,
|
| 1571 |
"learning_rate": 1.0203073830324566e-06,
|
| 1572 |
+
"loss": 1.4862,
|
| 1573 |
+
"mean_token_accuracy": 0.6215522885322571,
|
| 1574 |
+
"num_tokens": 311244.0,
|
| 1575 |
"step": 174
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 0.7056451612903226,
|
| 1579 |
+
"grad_norm": 5.811971664428711,
|
| 1580 |
"learning_rate": 9.949004525805423e-07,
|
| 1581 |
+
"loss": 1.4113,
|
| 1582 |
+
"mean_token_accuracy": 0.625806450843811,
|
| 1583 |
+
"num_tokens": 312796.0,
|
| 1584 |
"step": 175
|
| 1585 |
},
|
| 1586 |
{
|
| 1587 |
"epoch": 0.7096774193548387,
|
| 1588 |
+
"grad_norm": 6.348616600036621,
|
| 1589 |
"learning_rate": 9.697350436308428e-07,
|
| 1590 |
+
"loss": 1.4407,
|
| 1591 |
+
"mean_token_accuracy": 0.614819347858429,
|
| 1592 |
+
"num_tokens": 314431.0,
|
| 1593 |
"step": 176
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 0.7137096774193549,
|
| 1597 |
+
"grad_norm": 5.936633110046387,
|
| 1598 |
"learning_rate": 9.448151944460657e-07,
|
| 1599 |
+
"loss": 1.6396,
|
| 1600 |
+
"mean_token_accuracy": 0.5756798386573792,
|
| 1601 |
+
"num_tokens": 316382.0,
|
| 1602 |
"step": 177
|
| 1603 |
},
|
| 1604 |
{
|
| 1605 |
"epoch": 0.717741935483871,
|
| 1606 |
+
"grad_norm": 5.856295585632324,
|
| 1607 |
"learning_rate": 9.201449038842403e-07,
|
| 1608 |
+
"loss": 1.5693,
|
| 1609 |
+
"mean_token_accuracy": 0.6030020713806152,
|
| 1610 |
+
"num_tokens": 318316.0,
|
| 1611 |
"step": 178
|
| 1612 |
},
|
| 1613 |
{
|
| 1614 |
"epoch": 0.7217741935483871,
|
| 1615 |
+
"grad_norm": 6.206198215484619,
|
| 1616 |
"learning_rate": 8.957281307570254e-07,
|
| 1617 |
+
"loss": 1.4272,
|
| 1618 |
+
"mean_token_accuracy": 0.6132633090019226,
|
| 1619 |
+
"num_tokens": 319841.0,
|
| 1620 |
"step": 179
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
"epoch": 0.7258064516129032,
|
| 1624 |
+
"grad_norm": 5.620049476623535,
|
| 1625 |
"learning_rate": 8.71568793194445e-07,
|
| 1626 |
+
"loss": 1.5291,
|
| 1627 |
+
"mean_token_accuracy": 0.6036797761917114,
|
| 1628 |
+
"num_tokens": 321854.0,
|
| 1629 |
"step": 180
|
| 1630 |
},
|
| 1631 |
{
|
| 1632 |
"epoch": 0.7298387096774194,
|
| 1633 |
+
"grad_norm": 5.8175177574157715,
|
| 1634 |
"learning_rate": 8.476707680161486e-07,
|
| 1635 |
+
"loss": 1.4948,
|
| 1636 |
+
"mean_token_accuracy": 0.5953565239906311,
|
| 1637 |
+
"num_tokens": 323665.0,
|
| 1638 |
"step": 181
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 0.7338709677419355,
|
| 1642 |
+
"grad_norm": 5.765270709991455,
|
| 1643 |
"learning_rate": 8.240378901093035e-07,
|
| 1644 |
+
"loss": 1.5781,
|
| 1645 |
+
"mean_token_accuracy": 0.6022663116455078,
|
| 1646 |
+
"num_tokens": 325432.0,
|
| 1647 |
"step": 182
|
| 1648 |
},
|
| 1649 |
{
|
| 1650 |
"epoch": 0.7379032258064516,
|
| 1651 |
+
"grad_norm": 6.488076686859131,
|
| 1652 |
"learning_rate": 8.006739518132179e-07,
|
| 1653 |
+
"loss": 1.5952,
|
| 1654 |
+
"mean_token_accuracy": 0.5865209698677063,
|
| 1655 |
+
"num_tokens": 327081.0,
|
| 1656 |
"step": 183
|
| 1657 |
},
|
| 1658 |
{
|
| 1659 |
"epoch": 0.7419354838709677,
|
| 1660 |
+
"grad_norm": 6.85787296295166,
|
| 1661 |
"learning_rate": 7.775827023107835e-07,
|
| 1662 |
+
"loss": 1.5112,
|
| 1663 |
+
"mean_token_accuracy": 0.6159420013427734,
|
| 1664 |
+
"num_tokens": 328463.0,
|
| 1665 |
"step": 184
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 0.7459677419354839,
|
| 1669 |
+
"grad_norm": 5.89924955368042,
|
| 1670 |
"learning_rate": 7.547678470268526e-07,
|
| 1671 |
+
"loss": 1.4708,
|
| 1672 |
+
"mean_token_accuracy": 0.615646243095398,
|
| 1673 |
+
"num_tokens": 330229.0,
|
| 1674 |
"step": 185
|
| 1675 |
},
|
| 1676 |
{
|
| 1677 |
"epoch": 0.75,
|
| 1678 |
+
"grad_norm": 5.8790812492370605,
|
| 1679 |
"learning_rate": 7.322330470336314e-07,
|
| 1680 |
+
"loss": 1.6306,
|
| 1681 |
+
"mean_token_accuracy": 0.5779411792755127,
|
| 1682 |
+
"num_tokens": 332271.0,
|
| 1683 |
"step": 186
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
"epoch": 0.7540322580645161,
|
| 1687 |
+
"grad_norm": 5.292142391204834,
|
| 1688 |
"learning_rate": 7.099819184631929e-07,
|
| 1689 |
+
"loss": 1.4507,
|
| 1690 |
+
"mean_token_accuracy": 0.6181613206863403,
|
| 1691 |
+
"num_tokens": 334046.0,
|
| 1692 |
"step": 187
|
| 1693 |
},
|
| 1694 |
{
|
| 1695 |
"epoch": 0.7580645161290323,
|
| 1696 |
+
"grad_norm": 5.520077228546143,
|
| 1697 |
"learning_rate": 6.880180319272006e-07,
|
| 1698 |
+
"loss": 1.4124,
|
| 1699 |
+
"mean_token_accuracy": 0.6184003949165344,
|
| 1700 |
+
"num_tokens": 335961.0,
|
| 1701 |
"step": 188
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 0.7620967741935484,
|
| 1705 |
+
"grad_norm": 5.447739124298096,
|
| 1706 |
"learning_rate": 6.663449119439358e-07,
|
| 1707 |
+
"loss": 1.5101,
|
| 1708 |
+
"mean_token_accuracy": 0.6079999804496765,
|
| 1709 |
+
"num_tokens": 337838.0,
|
| 1710 |
"step": 189
|
| 1711 |
},
|
| 1712 |
{
|
| 1713 |
"epoch": 0.7661290322580645,
|
| 1714 |
+
"grad_norm": 5.440929889678955,
|
| 1715 |
"learning_rate": 6.449660363727236e-07,
|
| 1716 |
+
"loss": 1.5131,
|
| 1717 |
+
"mean_token_accuracy": 0.5955055952072144,
|
| 1718 |
+
"num_tokens": 339709.0,
|
| 1719 |
"step": 190
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
"epoch": 0.7701612903225806,
|
| 1723 |
+
"grad_norm": 6.182394504547119,
|
| 1724 |
"learning_rate": 6.238848358558439e-07,
|
| 1725 |
+
"loss": 1.485,
|
| 1726 |
+
"mean_token_accuracy": 0.6035979986190796,
|
| 1727 |
+
"num_tokens": 341323.0,
|
| 1728 |
"step": 191
|
| 1729 |
},
|
| 1730 |
{
|
| 1731 |
"epoch": 0.7741935483870968,
|
| 1732 |
+
"grad_norm": 6.693000793457031,
|
| 1733 |
"learning_rate": 6.031046932680229e-07,
|
| 1734 |
+
"loss": 1.4743,
|
| 1735 |
+
"mean_token_accuracy": 0.6190476417541504,
|
| 1736 |
+
"num_tokens": 342627.0,
|
| 1737 |
"step": 192
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 0.7782258064516129,
|
| 1741 |
+
"grad_norm": 5.218486309051514,
|
| 1742 |
"learning_rate": 5.826289431735832e-07,
|
| 1743 |
+
"loss": 1.4456,
|
| 1744 |
+
"mean_token_accuracy": 0.6149131655693054,
|
| 1745 |
+
"num_tokens": 344587.0,
|
| 1746 |
"step": 193
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
"epoch": 0.782258064516129,
|
| 1750 |
+
"grad_norm": 6.235224723815918,
|
| 1751 |
"learning_rate": 5.624608712913531e-07,
|
| 1752 |
+
"loss": 1.3816,
|
| 1753 |
+
"mean_token_accuracy": 0.6143410801887512,
|
| 1754 |
+
"num_tokens": 346137.0,
|
| 1755 |
"step": 194
|
| 1756 |
},
|
| 1757 |
{
|
| 1758 |
"epoch": 0.7862903225806451,
|
| 1759 |
+
"grad_norm": 5.2699761390686035,
|
| 1760 |
"learning_rate": 5.426037139674117e-07,
|
| 1761 |
+
"loss": 1.4791,
|
| 1762 |
+
"mean_token_accuracy": 0.6150712966918945,
|
| 1763 |
+
"num_tokens": 348103.0,
|
| 1764 |
"step": 195
|
| 1765 |
},
|
| 1766 |
{
|
| 1767 |
"epoch": 0.7903225806451613,
|
| 1768 |
+
"grad_norm": 5.843770980834961,
|
| 1769 |
"learning_rate": 5.23060657655754e-07,
|
| 1770 |
+
"loss": 1.5063,
|
| 1771 |
+
"mean_token_accuracy": 0.6077738404273987,
|
| 1772 |
+
"num_tokens": 350086.0,
|
| 1773 |
"step": 196
|
| 1774 |
},
|
| 1775 |
{
|
| 1776 |
"epoch": 0.7943548387096774,
|
| 1777 |
+
"grad_norm": 5.2697248458862305,
|
| 1778 |
"learning_rate": 5.038348384069663e-07,
|
| 1779 |
+
"loss": 1.4611,
|
| 1780 |
+
"mean_token_accuracy": 0.6173697113990784,
|
| 1781 |
+
"num_tokens": 352103.0,
|
| 1782 |
"step": 197
|
| 1783 |
},
|
| 1784 |
{
|
| 1785 |
"epoch": 0.7983870967741935,
|
| 1786 |
+
"grad_norm": 5.984487056732178,
|
| 1787 |
"learning_rate": 4.84929341364988e-07,
|
| 1788 |
+
"loss": 1.5012,
|
| 1789 |
+
"mean_token_accuracy": 0.6050000190734863,
|
| 1790 |
+
"num_tokens": 353905.0,
|
| 1791 |
"step": 198
|
| 1792 |
},
|
| 1793 |
{
|
| 1794 |
"epoch": 0.8024193548387096,
|
| 1795 |
+
"grad_norm": 5.947335243225098,
|
| 1796 |
"learning_rate": 4.6634720027204093e-07,
|
| 1797 |
+
"loss": 1.4547,
|
| 1798 |
+
"mean_token_accuracy": 0.6180400848388672,
|
| 1799 |
+
"num_tokens": 355703.0,
|
| 1800 |
"step": 199
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 0.8064516129032258,
|
| 1804 |
+
"grad_norm": 5.749611854553223,
|
| 1805 |
"learning_rate": 4.480913969818099e-07,
|
| 1806 |
+
"loss": 1.5199,
|
| 1807 |
+
"mean_token_accuracy": 0.5952813029289246,
|
| 1808 |
+
"num_tokens": 357358.0,
|
| 1809 |
"step": 200
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 0.8104838709677419,
|
| 1813 |
+
"grad_norm": 5.199901580810547,
|
| 1814 |
"learning_rate": 4.3016486098094667e-07,
|
| 1815 |
+
"loss": 1.47,
|
| 1816 |
+
"mean_token_accuracy": 0.6163245439529419,
|
| 1817 |
+
"num_tokens": 359406.0,
|
| 1818 |
"step": 201
|
| 1819 |
},
|
| 1820 |
{
|
| 1821 |
"epoch": 0.8145161290322581,
|
| 1822 |
+
"grad_norm": 5.465948581695557,
|
| 1823 |
"learning_rate": 4.125704689189819e-07,
|
| 1824 |
+
"loss": 1.5307,
|
| 1825 |
+
"mean_token_accuracy": 0.6006842851638794,
|
| 1826 |
+
"num_tokens": 361454.0,
|
| 1827 |
"step": 202
|
| 1828 |
},
|
| 1829 |
{
|
| 1830 |
"epoch": 0.8185483870967742,
|
| 1831 |
+
"grad_norm": 5.9000749588012695,
|
| 1832 |
"learning_rate": 3.953110441467073e-07,
|
| 1833 |
+
"loss": 1.4202,
|
| 1834 |
+
"mean_token_accuracy": 0.6333949565887451,
|
| 1835 |
+
"num_tokens": 363079.0,
|
| 1836 |
"step": 203
|
| 1837 |
},
|
| 1838 |
{
|
| 1839 |
"epoch": 0.8225806451612904,
|
| 1840 |
+
"grad_norm": 5.953113555908203,
|
| 1841 |
"learning_rate": 3.7838935626312246e-07,
|
| 1842 |
+
"loss": 1.4819,
|
| 1843 |
+
"mean_token_accuracy": 0.6164458990097046,
|
| 1844 |
+
"num_tokens": 364893.0,
|
| 1845 |
"step": 204
|
| 1846 |
},
|
| 1847 |
{
|
| 1848 |
"epoch": 0.8266129032258065,
|
| 1849 |
+
"grad_norm": 5.6844892501831055,
|
| 1850 |
"learning_rate": 3.6180812067099477e-07,
|
| 1851 |
+
"loss": 1.4199,
|
| 1852 |
+
"mean_token_accuracy": 0.6286211013793945,
|
| 1853 |
+
"num_tokens": 366621.0,
|
| 1854 |
"step": 205
|
| 1855 |
},
|
| 1856 |
{
|
| 1857 |
"epoch": 0.8306451612903226,
|
| 1858 |
+
"grad_norm": 5.685678482055664,
|
| 1859 |
"learning_rate": 3.455699981411259e-07,
|
| 1860 |
+
"loss": 1.5947,
|
| 1861 |
+
"mean_token_accuracy": 0.5794117450714111,
|
| 1862 |
+
"num_tokens": 368663.0,
|
| 1863 |
"step": 206
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 0.8346774193548387,
|
| 1867 |
+
"grad_norm": 5.832515716552734,
|
| 1868 |
"learning_rate": 3.296775943853789e-07,
|
| 1869 |
+
"loss": 1.419,
|
| 1870 |
+
"mean_token_accuracy": 0.6132450103759766,
|
| 1871 |
+
"num_tokens": 370175.0,
|
| 1872 |
"step": 207
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 0.8387096774193549,
|
| 1876 |
+
"grad_norm": 5.6208720207214355,
|
| 1877 |
"learning_rate": 3.141334596385448e-07,
|
| 1878 |
+
"loss": 1.4981,
|
| 1879 |
+
"mean_token_accuracy": 0.6034953594207764,
|
| 1880 |
+
"num_tokens": 372008.0,
|
| 1881 |
"step": 208
|
| 1882 |
},
|
| 1883 |
{
|
| 1884 |
"epoch": 0.842741935483871,
|
| 1885 |
+
"grad_norm": 5.669355869293213,
|
| 1886 |
"learning_rate": 2.9894008824910726e-07,
|
| 1887 |
+
"loss": 1.5807,
|
| 1888 |
+
"mean_token_accuracy": 0.591269850730896,
|
| 1889 |
+
"num_tokens": 374026.0,
|
| 1890 |
"step": 209
|
| 1891 |
},
|
| 1892 |
{
|
| 1893 |
"epoch": 0.8467741935483871,
|
| 1894 |
+
"grad_norm": 5.443574905395508,
|
| 1895 |
"learning_rate": 2.840999182789797e-07,
|
| 1896 |
+
"loss": 1.4487,
|
| 1897 |
+
"mean_token_accuracy": 0.6099706888198853,
|
| 1898 |
+
"num_tokens": 376074.0,
|
| 1899 |
"step": 210
|
| 1900 |
},
|
| 1901 |
{
|
| 1902 |
"epoch": 0.8508064516129032,
|
| 1903 |
+
"grad_norm": 5.724096298217773,
|
| 1904 |
"learning_rate": 2.696153311122704e-07,
|
| 1905 |
+
"loss": 1.6509,
|
| 1906 |
+
"mean_token_accuracy": 0.5762162208557129,
|
| 1907 |
+
"num_tokens": 377926.0,
|
| 1908 |
"step": 211
|
| 1909 |
},
|
| 1910 |
{
|
| 1911 |
"epoch": 0.8548387096774194,
|
| 1912 |
+
"grad_norm": 6.077378749847412,
|
| 1913 |
"learning_rate": 2.5548865107314606e-07,
|
| 1914 |
+
"loss": 1.461,
|
| 1915 |
+
"mean_token_accuracy": 0.6061684489250183,
|
| 1916 |
+
"num_tokens": 379614.0,
|
| 1917 |
"step": 212
|
| 1918 |
},
|
| 1919 |
{
|
| 1920 |
"epoch": 0.8588709677419355,
|
| 1921 |
+
"grad_norm": 5.7304229736328125,
|
| 1922 |
"learning_rate": 2.4172214505285006e-07,
|
| 1923 |
+
"loss": 1.4963,
|
| 1924 |
+
"mean_token_accuracy": 0.6054667234420776,
|
| 1925 |
+
"num_tokens": 381555.0,
|
| 1926 |
"step": 213
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 0.8629032258064516,
|
| 1930 |
+
"grad_norm": 5.604650020599365,
|
| 1931 |
"learning_rate": 2.2831802214593774e-07,
|
| 1932 |
+
"loss": 1.6199,
|
| 1933 |
+
"mean_token_accuracy": 0.5726895332336426,
|
| 1934 |
+
"num_tokens": 383483.0,
|
| 1935 |
"step": 214
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
"epoch": 0.8669354838709677,
|
| 1939 |
+
"grad_norm": 5.319303512573242,
|
| 1940 |
"learning_rate": 2.1527843329578328e-07,
|
| 1941 |
+
"loss": 1.5539,
|
| 1942 |
+
"mean_token_accuracy": 0.597201406955719,
|
| 1943 |
+
"num_tokens": 385486.0,
|
| 1944 |
"step": 215
|
| 1945 |
},
|
| 1946 |
{
|
| 1947 |
"epoch": 0.8709677419354839,
|
| 1948 |
+
"grad_norm": 5.75916862487793,
|
| 1949 |
"learning_rate": 2.026054709494235e-07,
|
| 1950 |
+
"loss": 1.498,
|
| 1951 |
+
"mean_token_accuracy": 0.6148062348365784,
|
| 1952 |
+
"num_tokens": 387217.0,
|
| 1953 |
"step": 216
|
| 1954 |
},
|
| 1955 |
{
|
| 1956 |
"epoch": 0.875,
|
| 1957 |
+
"grad_norm": 6.274486541748047,
|
| 1958 |
"learning_rate": 1.9030116872178317e-07,
|
| 1959 |
+
"loss": 1.4691,
|
| 1960 |
+
"mean_token_accuracy": 0.6093660593032837,
|
| 1961 |
+
"num_tokens": 388970.0,
|
| 1962 |
"step": 217
|
| 1963 |
},
|
| 1964 |
{
|
| 1965 |
"epoch": 0.8790322580645161,
|
| 1966 |
+
"grad_norm": 5.682790756225586,
|
| 1967 |
"learning_rate": 1.7836750106934475e-07,
|
| 1968 |
+
"loss": 1.4487,
|
| 1969 |
+
"mean_token_accuracy": 0.6198156476020813,
|
| 1970 |
+
"num_tokens": 390708.0,
|
| 1971 |
"step": 218
|
| 1972 |
},
|
| 1973 |
{
|
| 1974 |
"epoch": 0.8830645161290323,
|
| 1975 |
+
"grad_norm": 6.133249282836914,
|
| 1976 |
"learning_rate": 1.6680638297330854e-07,
|
| 1977 |
+
"loss": 1.4905,
|
| 1978 |
+
"mean_token_accuracy": 0.5962441563606262,
|
| 1979 |
+
"num_tokens": 392414.0,
|
| 1980 |
"step": 219
|
| 1981 |
},
|
| 1982 |
{
|
| 1983 |
"epoch": 0.8870967741935484,
|
| 1984 |
+
"grad_norm": 6.2251386642456055,
|
| 1985 |
"learning_rate": 1.5561966963229925e-07,
|
| 1986 |
+
"loss": 1.4176,
|
| 1987 |
+
"mean_token_accuracy": 0.6197279095649719,
|
| 1988 |
+
"num_tokens": 393886.0,
|
| 1989 |
"step": 220
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 0.8911290322580645,
|
| 1993 |
+
"grad_norm": 5.665816307067871,
|
| 1994 |
"learning_rate": 1.448091561646628e-07,
|
| 1995 |
+
"loss": 1.5031,
|
| 1996 |
+
"mean_token_accuracy": 0.6201764345169067,
|
| 1997 |
+
"num_tokens": 395702.0,
|
| 1998 |
"step": 221
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
"epoch": 0.8951612903225806,
|
| 2002 |
+
"grad_norm": 6.105023384094238,
|
| 2003 |
"learning_rate": 1.3437657732040783e-07,
|
| 2004 |
+
"loss": 1.3976,
|
| 2005 |
+
"mean_token_accuracy": 0.618686854839325,
|
| 2006 |
+
"num_tokens": 397288.0,
|
| 2007 |
"step": 222
|
| 2008 |
},
|
| 2009 |
{
|
| 2010 |
"epoch": 0.8991935483870968,
|
| 2011 |
+
"grad_norm": 5.846229553222656,
|
| 2012 |
"learning_rate": 1.243236072028317e-07,
|
| 2013 |
+
"loss": 1.4722,
|
| 2014 |
+
"mean_token_accuracy": 0.6098484992980957,
|
| 2015 |
+
"num_tokens": 399138.0,
|
| 2016 |
"step": 223
|
| 2017 |
},
|
| 2018 |
{
|
| 2019 |
"epoch": 0.9032258064516129,
|
| 2020 |
+
"grad_norm": 5.863070964813232,
|
| 2021 |
"learning_rate": 1.1465185899987797e-07,
|
| 2022 |
+
"loss": 1.6153,
|
| 2023 |
+
"mean_token_accuracy": 0.5908849835395813,
|
| 2024 |
+
"num_tokens": 401027.0,
|
| 2025 |
"step": 224
|
| 2026 |
},
|
| 2027 |
{
|
| 2028 |
"epoch": 0.907258064516129,
|
| 2029 |
+
"grad_norm": 5.625185966491699,
|
| 2030 |
"learning_rate": 1.0536288472527162e-07,
|
| 2031 |
+
"loss": 1.6113,
|
| 2032 |
+
"mean_token_accuracy": 0.5997909903526306,
|
| 2033 |
+
"num_tokens": 402943.0,
|
| 2034 |
"step": 225
|
| 2035 |
},
|
| 2036 |
{
|
| 2037 |
"epoch": 0.9112903225806451,
|
| 2038 |
+
"grad_norm": 5.376096725463867,
|
| 2039 |
"learning_rate": 9.645817496946902e-08,
|
| 2040 |
+
"loss": 1.4866,
|
| 2041 |
+
"mean_token_accuracy": 0.6109374761581421,
|
| 2042 |
+
"num_tokens": 404865.0,
|
| 2043 |
"step": 226
|
| 2044 |
},
|
| 2045 |
{
|
| 2046 |
"epoch": 0.9153225806451613,
|
| 2047 |
+
"grad_norm": 6.09489631652832,
|
| 2048 |
"learning_rate": 8.79391586604636e-08,
|
| 2049 |
+
"loss": 1.6137,
|
| 2050 |
+
"mean_token_accuracy": 0.5798553228378296,
|
| 2051 |
+
"num_tokens": 406664.0,
|
| 2052 |
"step": 227
|
| 2053 |
},
|
| 2054 |
{
|
| 2055 |
"epoch": 0.9193548387096774,
|
| 2056 |
+
"grad_norm": 6.106083393096924,
|
| 2057 |
"learning_rate": 7.980720283448957e-08,
|
| 2058 |
+
"loss": 1.7006,
|
| 2059 |
+
"mean_token_accuracy": 0.5664711594581604,
|
| 2060 |
+
"num_tokens": 408712.0,
|
| 2061 |
"step": 228
|
| 2062 |
},
|
| 2063 |
{
|
| 2064 |
"epoch": 0.9233870967741935,
|
| 2065 |
+
"grad_norm": 6.476028919219971,
|
| 2066 |
"learning_rate": 7.206361241665266e-08,
|
| 2067 |
+
"loss": 1.5162,
|
| 2068 |
+
"mean_token_accuracy": 0.608540952205658,
|
| 2069 |
+
"num_tokens": 410119.0,
|
| 2070 |
"step": 229
|
| 2071 |
},
|
| 2072 |
{
|
| 2073 |
"epoch": 0.9274193548387096,
|
| 2074 |
+
"grad_norm": 5.52684211730957,
|
| 2075 |
"learning_rate": 6.470963001153268e-08,
|
| 2076 |
+
"loss": 1.5155,
|
| 2077 |
+
"mean_token_accuracy": 0.6171218752861023,
|
| 2078 |
+
"num_tokens": 412025.0,
|
| 2079 |
"step": 230
|
| 2080 |
},
|
| 2081 |
{
|
| 2082 |
"epoch": 0.9314516129032258,
|
| 2083 |
+
"grad_norm": 5.9519782066345215,
|
| 2084 |
"learning_rate": 5.774643570378296e-08,
|
| 2085 |
+
"loss": 1.3252,
|
| 2086 |
+
"mean_token_accuracy": 0.6445659399032593,
|
| 2087 |
+
"num_tokens": 413490.0,
|
| 2088 |
"step": 231
|
| 2089 |
},
|
| 2090 |
{
|
| 2091 |
"epoch": 0.9354838709677419,
|
| 2092 |
+
"grad_norm": 5.545828342437744,
|
| 2093 |
"learning_rate": 5.117514686876379e-08,
|
| 2094 |
+
"loss": 1.4661,
|
| 2095 |
+
"mean_token_accuracy": 0.6093666553497314,
|
| 2096 |
+
"num_tokens": 415371.0,
|
| 2097 |
"step": 232
|
| 2098 |
},
|
| 2099 |
{
|
| 2100 |
"epoch": 0.9395161290322581,
|
| 2101 |
+
"grad_norm": 5.3752617835998535,
|
| 2102 |
"learning_rate": 4.4996817993239464e-08,
|
| 2103 |
+
"loss": 1.5721,
|
| 2104 |
+
"mean_token_accuracy": 0.589442789554596,
|
| 2105 |
+
"num_tokens": 417419.0,
|
| 2106 |
"step": 233
|
| 2107 |
},
|
| 2108 |
{
|
| 2109 |
"epoch": 0.9435483870967742,
|
| 2110 |
+
"grad_norm": 6.181939601898193,
|
| 2111 |
"learning_rate": 3.9212440506164465e-08,
|
| 2112 |
+
"loss": 1.4851,
|
| 2113 |
+
"mean_token_accuracy": 0.6140567064285278,
|
| 2114 |
+
"num_tokens": 419043.0,
|
| 2115 |
"step": 234
|
| 2116 |
},
|
| 2117 |
{
|
| 2118 |
"epoch": 0.9475806451612904,
|
| 2119 |
+
"grad_norm": 6.1111836433410645,
|
| 2120 |
"learning_rate": 3.382294261959157e-08,
|
| 2121 |
+
"loss": 1.5793,
|
| 2122 |
+
"mean_token_accuracy": 0.6070796251296997,
|
| 2123 |
+
"num_tokens": 420740.0,
|
| 2124 |
"step": 235
|
| 2125 |
},
|
| 2126 |
{
|
| 2127 |
"epoch": 0.9516129032258065,
|
| 2128 |
+
"grad_norm": 6.486725330352783,
|
| 2129 |
"learning_rate": 2.8829189179721552e-08,
|
| 2130 |
+
"loss": 1.4531,
|
| 2131 |
+
"mean_token_accuracy": 0.6059664487838745,
|
| 2132 |
+
"num_tokens": 422351.0,
|
| 2133 |
"step": 236
|
| 2134 |
},
|
| 2135 |
{
|
| 2136 |
"epoch": 0.9556451612903226,
|
| 2137 |
+
"grad_norm": 5.367029190063477,
|
| 2138 |
"learning_rate": 2.423198152812306e-08,
|
| 2139 |
+
"loss": 1.412,
|
| 2140 |
+
"mean_token_accuracy": 0.6007625460624695,
|
| 2141 |
+
"num_tokens": 424189.0,
|
| 2142 |
"step": 237
|
| 2143 |
},
|
| 2144 |
{
|
| 2145 |
"epoch": 0.9596774193548387,
|
| 2146 |
+
"grad_norm": 5.885536193847656,
|
| 2147 |
"learning_rate": 2.0032057373142453e-08,
|
| 2148 |
+
"loss": 1.4751,
|
| 2149 |
+
"mean_token_accuracy": 0.6121867895126343,
|
| 2150 |
+
"num_tokens": 425947.0,
|
| 2151 |
"step": 238
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
"epoch": 0.9637096774193549,
|
| 2155 |
+
"grad_norm": 6.31535005569458,
|
| 2156 |
"learning_rate": 1.6230090671524312e-08,
|
| 2157 |
+
"loss": 1.5074,
|
| 2158 |
+
"mean_token_accuracy": 0.60240238904953,
|
| 2159 |
+
"num_tokens": 427614.0,
|
| 2160 |
"step": 239
|
| 2161 |
},
|
| 2162 |
{
|
| 2163 |
"epoch": 0.967741935483871,
|
| 2164 |
+
"grad_norm": 6.376657485961914,
|
| 2165 |
"learning_rate": 1.2826691520262114e-08,
|
| 2166 |
+
"loss": 1.6623,
|
| 2167 |
+
"mean_token_accuracy": 0.5947854518890381,
|
| 2168 |
+
"num_tokens": 429457.0,
|
| 2169 |
"step": 240
|
| 2170 |
},
|
| 2171 |
{
|
| 2172 |
"epoch": 0.9717741935483871,
|
| 2173 |
+
"grad_norm": 5.587056636810303,
|
| 2174 |
"learning_rate": 9.822406058697665e-09,
|
| 2175 |
+
"loss": 1.5953,
|
| 2176 |
+
"mean_token_accuracy": 0.5949429869651794,
|
| 2177 |
+
"num_tokens": 431476.0,
|
| 2178 |
"step": 241
|
| 2179 |
},
|
| 2180 |
{
|
| 2181 |
"epoch": 0.9758064516129032,
|
| 2182 |
+
"grad_norm": 5.475006580352783,
|
| 2183 |
"learning_rate": 7.217716380881479e-09,
|
| 2184 |
+
"loss": 1.476,
|
| 2185 |
+
"mean_token_accuracy": 0.6025062799453735,
|
| 2186 |
+
"num_tokens": 433473.0,
|
| 2187 |
"step": 242
|
| 2188 |
},
|
| 2189 |
{
|
| 2190 |
"epoch": 0.9798387096774194,
|
| 2191 |
+
"grad_norm": 6.063528537750244,
|
| 2192 |
"learning_rate": 5.0130404582127144e-09,
|
| 2193 |
+
"loss": 1.3995,
|
| 2194 |
+
"mean_token_accuracy": 0.6313099265098572,
|
| 2195 |
+
"num_tokens": 435040.0,
|
| 2196 |
"step": 243
|
| 2197 |
},
|
| 2198 |
{
|
| 2199 |
"epoch": 0.9838709677419355,
|
| 2200 |
+
"grad_norm": 5.9405035972595215,
|
| 2201 |
"learning_rate": 3.208732072368104e-09,
|
| 2202 |
+
"loss": 1.5641,
|
| 2203 |
+
"mean_token_accuracy": 0.5938547253608704,
|
| 2204 |
+
"num_tokens": 436832.0,
|
| 2205 |
"step": 244
|
| 2206 |
},
|
| 2207 |
{
|
| 2208 |
"epoch": 0.9879032258064516,
|
| 2209 |
+
"grad_norm": 5.634851932525635,
|
| 2210 |
"learning_rate": 1.8050807585293095e-09,
|
| 2211 |
+
"loss": 1.4805,
|
| 2212 |
+
"mean_token_accuracy": 0.5975820422172546,
|
| 2213 |
+
"num_tokens": 438571.0,
|
| 2214 |
"step": 245
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 0.9919354838709677,
|
| 2218 |
+
"grad_norm": 5.381149768829346,
|
| 2219 |
"learning_rate": 8.023117589237017e-10,
|
| 2220 |
+
"loss": 1.513,
|
| 2221 |
+
"mean_token_accuracy": 0.6109482049942017,
|
| 2222 |
+
"num_tokens": 440619.0,
|
| 2223 |
"step": 246
|
| 2224 |
},
|
| 2225 |
{
|
| 2226 |
"epoch": 0.9959677419354839,
|
| 2227 |
+
"grad_norm": 6.20208215713501,
|
| 2228 |
"learning_rate": 2.0058598667854755e-10,
|
| 2229 |
+
"loss": 1.4039,
|
| 2230 |
+
"mean_token_accuracy": 0.6231250166893005,
|
| 2231 |
+
"num_tokens": 442221.0,
|
| 2232 |
"step": 247
|
| 2233 |
},
|
| 2234 |
{
|
| 2235 |
"epoch": 1.0,
|
| 2236 |
+
"grad_norm": 6.332696914672852,
|
| 2237 |
"learning_rate": 0.0,
|
| 2238 |
+
"loss": 1.4206,
|
| 2239 |
+
"mean_token_accuracy": 0.5873016119003296,
|
| 2240 |
+
"num_tokens": 442852.0,
|
| 2241 |
"step": 248
|
| 2242 |
}
|
| 2243 |
],
|
|
|
|
| 2258 |
"attributes": {}
|
| 2259 |
}
|
| 2260 |
},
|
| 2261 |
+
"total_flos": 9984447243878400.0,
|
| 2262 |
"train_batch_size": 1,
|
| 2263 |
"trial_name": null,
|
| 2264 |
"trial_params": null
|