Upload folder using huggingface_hub
Browse files
checkpoint-31500/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3702168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1eb2f06288fd7ff7e62da2d80f413f63d457bcb727ccfe7f8fbb8318ec3d1a72
|
| 3 |
size 3702168
|
checkpoint-31500/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7473658
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba614a2d4298c4933d28ad2db0f85d17efbe574236930bd61b1e74c1e0e32443
|
| 3 |
size 7473658
|
checkpoint-31500/trainer_state.json
CHANGED
|
@@ -10,2263 +10,2263 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.025,
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
"learning_rate": 0.000499987952239832,
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.05,
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
"learning_rate": 0.0004999518101205162,
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.075,
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
"learning_rate": 0.0004998915771255053,
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.1,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
"learning_rate": 0.0004998072590601808,
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.125,
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
"learning_rate": 0.0004996988640512931,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.15,
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
"learning_rate": 0.000499566402546179,
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.175,
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
"learning_rate": 0.0004994098873117539,
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.2,
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
"learning_rate": 0.000499229333433282,
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.225,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
"learning_rate": 0.0004990247583129218,
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.25,
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
"learning_rate": 0.0004987961816680492,
|
| 78 |
"loss": 1.0698,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.275,
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
"learning_rate": 0.0004985436255293571,
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.3,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
"learning_rate": 0.0004982671142387316,
|
| 92 |
"loss": 1.0906,
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.325,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
"learning_rate": 0.0004979666744469065,
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.35,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
"learning_rate": 0.0004976423351108943,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.375,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
"learning_rate": 0.0004972941274911952,
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.4,
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
"learning_rate": 0.0004969220851487844,
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.425,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
"learning_rate": 0.0004965262439418772,
|
| 127 |
"loss": 1.082,
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.45,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
"learning_rate": 0.0004961066420224729,
|
| 134 |
"loss": 1.0937,
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.475,
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
"learning_rate": 0.000495663319832678,
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.5,
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
"learning_rate": 0.0004951963201008077,
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.525,
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
"learning_rate": 0.0004947056878372681,
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.55,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
"learning_rate": 0.0004941914703302181,
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.575,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
"learning_rate": 0.0004936537171410112,
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.6,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
"learning_rate": 0.0004930924800994192,
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.625,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
"learning_rate": 0.000492507813298636,
|
| 183 |
"loss": 1.0858,
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.65,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
"learning_rate": 0.0004918997730900649,
|
| 190 |
-
"loss": 1.
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.675,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
"learning_rate": 0.0004912684180778869,
|
| 197 |
"loss": 1.0647,
|
| 198 |
"step": 2700
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.7,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
"learning_rate": 0.0004906138091134118,
|
| 204 |
"loss": 1.0484,
|
| 205 |
"step": 2800
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.725,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
"learning_rate": 0.0004899360092892143,
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 2900
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.75,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
"learning_rate": 0.0004892350839330522,
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.775,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
"learning_rate": 0.0004885111006015701,
|
| 225 |
-
"loss": 1.
|
| 226 |
"step": 3100
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.8,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
"learning_rate": 0.0004877641290737884,
|
| 232 |
-
"loss": 1.
|
| 233 |
"step": 3200
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.825,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
"learning_rate": 0.0004869942413443776,
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 3300
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.85,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
"learning_rate": 0.00048620151161671955,
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 3400
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.875,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
"learning_rate": 0.0004853860162957552,
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.9,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 0.0004845478339806211,
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 3600
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.925,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
"learning_rate": 0.0004836870454570731,
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 3700
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.95,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
"learning_rate": 0.00048280373368970086,
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 3800
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.975,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
"learning_rate": 0.000481897983813931,
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 3900
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 1.0,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
"learning_rate": 0.0004809698831278217,
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
-
"eval_loss": 1.
|
| 294 |
-
"eval_runtime":
|
| 295 |
-
"eval_samples_per_second":
|
| 296 |
-
"eval_steps_per_second":
|
| 297 |
"step": 4000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.025,
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
"learning_rate": 0.00048001952108364876,
|
| 303 |
-
"loss": 1.
|
| 304 |
"step": 4100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 1.05,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
"learning_rate": 0.00047904698927928404,
|
| 310 |
-
"loss": 1.
|
| 311 |
"step": 4200
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 1.075,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
"learning_rate": 0.0004780523814493669,
|
| 317 |
-
"loss": 1.
|
| 318 |
"step": 4300
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 1.1,
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
"learning_rate": 0.00047703579345627036,
|
| 324 |
-
"loss": 1.
|
| 325 |
"step": 4400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.125,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
"learning_rate": 0.0004759973232808609,
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 4500
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.15,
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
"learning_rate": 0.0004749370710130554,
|
| 338 |
-
"loss": 1.
|
| 339 |
"step": 4600
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.175,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
"learning_rate": 0.0004738551388421742,
|
| 345 |
"loss": 1.0291,
|
| 346 |
"step": 4700
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.2,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
"learning_rate": 0.00047275163104709196,
|
| 352 |
-
"loss": 1.
|
| 353 |
"step": 4800
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.225,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
"learning_rate": 0.00047162665398618666,
|
| 359 |
-
"loss": 1.
|
| 360 |
"step": 4900
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.25,
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
"learning_rate": 0.00047048031608708875,
|
| 366 |
-
"loss": 1.
|
| 367 |
"step": 5000
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.275,
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
"learning_rate": 0.00046931272783623106,
|
| 373 |
-
"loss": 1.
|
| 374 |
"step": 5100
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.3,
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
"learning_rate": 0.0004681240017681993,
|
| 380 |
-
"loss": 1.
|
| 381 |
"step": 5200
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.325,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
"learning_rate": 0.00046691425245488607,
|
| 387 |
-
"loss": 1.
|
| 388 |
"step": 5300
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 1.35,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
"learning_rate": 0.00046568359649444796,
|
| 394 |
-
"loss": 1.
|
| 395 |
"step": 5400
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.375,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
"learning_rate": 0.00046443215250006805,
|
| 401 |
-
"loss": 1.
|
| 402 |
"step": 5500
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.4,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
"learning_rate": 0.00046316004108852305,
|
| 408 |
-
"loss": 1.
|
| 409 |
"step": 5600
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 1.425,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
"learning_rate": 0.0004618673848685586,
|
| 415 |
-
"loss": 1.
|
| 416 |
"step": 5700
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 1.45,
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
"learning_rate": 0.0004605543084290716,
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 5800
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 1.475,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
"learning_rate": 0.0004592209383271023,
|
| 429 |
-
"loss": 1.
|
| 430 |
"step": 5900
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 1.5,
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
"learning_rate": 0.00045786740307563633,
|
| 436 |
-
"loss": 1.
|
| 437 |
"step": 6000
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 1.525,
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
"learning_rate": 0.0004564938331312183,
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 6100
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 1.55,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
"learning_rate": 0.0004551003608813784,
|
| 450 |
-
"loss": 1.
|
| 451 |
"step": 6200
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 1.575,
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
"learning_rate": 0.00045368712063187237,
|
| 457 |
-
"loss": 1.
|
| 458 |
"step": 6300
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 1.6,
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
"learning_rate": 0.0004522542485937369,
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 6400
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 1.625,
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
"learning_rate": 0.0004508018828701612,
|
| 471 |
-
"loss": 1.
|
| 472 |
"step": 6500
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 1.65,
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
"learning_rate": 0.0004493301634431768,
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 6600
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 1.675,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
"learning_rate": 0.00044783923216016507,
|
| 485 |
-
"loss": 1.
|
| 486 |
"step": 6700
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 1.7,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
"learning_rate": 0.0004463292327201862,
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 6800
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 1.725,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
"learning_rate": 0.00044480031066012916,
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 6900
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 1.75,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
"learning_rate": 0.0004432526133406842,
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 7000
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 1.775,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
"learning_rate": 0.00044168628993214036,
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 7100
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 1.8,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
"learning_rate": 0.0004401014914000078,
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 7200
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 1.825,
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
"learning_rate": 0.00043849837049046735,
|
| 527 |
"loss": 1.007,
|
| 528 |
"step": 7300
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 1.85,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
"learning_rate": 0.00043687708171564923,
|
| 534 |
-
"loss": 1.
|
| 535 |
"step": 7400
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 1.875,
|
| 539 |
-
"grad_norm": 0.
|
| 540 |
"learning_rate": 0.0004352377813387398,
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 7500
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 1.9,
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
"learning_rate": 0.0004335806273589214,
|
| 548 |
-
"loss": 1.
|
| 549 |
"step": 7600
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 1.925,
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
"learning_rate": 0.00043190577949614375,
|
| 555 |
-
"loss": 1.
|
| 556 |
"step": 7700
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.95,
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
"learning_rate": 0.0004302133991757297,
|
| 562 |
-
"loss": 0.
|
| 563 |
"step": 7800
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 1.975,
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
"learning_rate": 0.00042850364951281707,
|
| 569 |
-
"loss": 1.
|
| 570 |
"step": 7900
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 2.0,
|
| 574 |
-
"grad_norm": 0.
|
| 575 |
"learning_rate": 0.00042677669529663686,
|
| 576 |
-
"loss": 1.
|
| 577 |
"step": 8000
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 2.0,
|
| 581 |
-
"eval_loss": 1.
|
| 582 |
-
"eval_runtime":
|
| 583 |
-
"eval_samples_per_second":
|
| 584 |
-
"eval_steps_per_second":
|
| 585 |
"step": 8000
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 2.025,
|
| 589 |
-
"grad_norm": 0.
|
| 590 |
"learning_rate": 0.0004250327029746309,
|
| 591 |
-
"loss": 1.
|
| 592 |
"step": 8100
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 2.05,
|
| 596 |
-
"grad_norm": 0.
|
| 597 |
"learning_rate": 0.000423271840636409,
|
| 598 |
-
"loss": 0.
|
| 599 |
"step": 8200
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 2.075,
|
| 603 |
-
"grad_norm": 0.
|
| 604 |
"learning_rate": 0.00042149427799754817,
|
| 605 |
-
"loss": 0.
|
| 606 |
"step": 8300
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 2.1,
|
| 610 |
-
"grad_norm": 0.
|
| 611 |
"learning_rate": 0.00041970018638323546,
|
| 612 |
-
"loss": 0.
|
| 613 |
"step": 8400
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 2.125,
|
| 617 |
-
"grad_norm": 0.
|
| 618 |
"learning_rate": 0.00041788973871175465,
|
| 619 |
-
"loss": 0.
|
| 620 |
"step": 8500
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 2.15,
|
| 624 |
-
"grad_norm": 0.
|
| 625 |
"learning_rate": 0.00041606310947782046,
|
| 626 |
-
"loss": 0.
|
| 627 |
"step": 8600
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 2.175,
|
| 631 |
-
"grad_norm": 0.
|
| 632 |
"learning_rate": 0.00041422047473576033,
|
| 633 |
-
"loss": 1.
|
| 634 |
"step": 8700
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 2.2,
|
| 638 |
-
"grad_norm": 0.
|
| 639 |
"learning_rate": 0.0004123620120825459,
|
| 640 |
-
"loss": 1.
|
| 641 |
"step": 8800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 2.225,
|
| 645 |
-
"grad_norm": 0.
|
| 646 |
"learning_rate": 0.00041048790064067577,
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 8900
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 2.25,
|
| 652 |
-
"grad_norm": 0.
|
| 653 |
"learning_rate": 0.0004085983210409114,
|
| 654 |
-
"loss": 0.
|
| 655 |
"step": 9000
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 2.275,
|
| 659 |
-
"grad_norm": 0.
|
| 660 |
"learning_rate": 0.0004066934554048674,
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 9100
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 2.3,
|
| 666 |
-
"grad_norm": 0.
|
| 667 |
"learning_rate": 0.00040477348732745853,
|
| 668 |
"loss": 0.9858,
|
| 669 |
"step": 9200
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 2.325,
|
| 673 |
-
"grad_norm": 0.
|
| 674 |
"learning_rate": 0.0004028386018592041,
|
| 675 |
-
"loss": 0.
|
| 676 |
"step": 9300
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 2.35,
|
| 680 |
-
"grad_norm": 0.
|
| 681 |
"learning_rate": 0.0004008889854883929,
|
| 682 |
-
"loss": 0.
|
| 683 |
"step": 9400
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 2.375,
|
| 687 |
-
"grad_norm": 0.
|
| 688 |
"learning_rate": 0.0003989248261231084,
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 9500
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 2.4,
|
| 694 |
-
"grad_norm": 0.
|
| 695 |
"learning_rate": 0.0003969463130731183,
|
| 696 |
-
"loss": 0.
|
| 697 |
"step": 9600
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 2.425,
|
| 701 |
-
"grad_norm": 0.
|
| 702 |
"learning_rate": 0.00039495363703162843,
|
| 703 |
"loss": 0.978,
|
| 704 |
"step": 9700
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 2.45,
|
| 708 |
-
"grad_norm": 0.
|
| 709 |
"learning_rate": 0.000392946990056903,
|
| 710 |
-
"loss": 1.
|
| 711 |
"step": 9800
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 2.475,
|
| 715 |
-
"grad_norm": 0.
|
| 716 |
"learning_rate": 0.00039092656555375416,
|
| 717 |
-
"loss": 1.
|
| 718 |
"step": 9900
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 2.5,
|
| 722 |
-
"grad_norm": 0.
|
| 723 |
"learning_rate": 0.00038889255825490053,
|
| 724 |
-
"loss": 0.
|
| 725 |
"step": 10000
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 2.525,
|
| 729 |
-
"grad_norm": 0.
|
| 730 |
"learning_rate": 0.0003868451642021992,
|
| 731 |
"loss": 0.9745,
|
| 732 |
"step": 10100
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 2.55,
|
| 736 |
-
"grad_norm": 0.
|
| 737 |
"learning_rate": 0.0003847845807277501,
|
| 738 |
-
"loss": 0.
|
| 739 |
"step": 10200
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 2.575,
|
| 743 |
-
"grad_norm": 0.
|
| 744 |
"learning_rate": 0.0003827110064348773,
|
| 745 |
-
"loss": 0.
|
| 746 |
"step": 10300
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 2.6,
|
| 750 |
-
"grad_norm": 0.
|
| 751 |
"learning_rate": 0.0003806246411789872,
|
| 752 |
-
"loss": 0.
|
| 753 |
"step": 10400
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 2.625,
|
| 757 |
-
"grad_norm": 0.
|
| 758 |
"learning_rate": 0.0003785256860483054,
|
| 759 |
-
"loss": 0.
|
| 760 |
"step": 10500
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 2.65,
|
| 764 |
-
"grad_norm": 0.
|
| 765 |
"learning_rate": 0.0003764143433444962,
|
| 766 |
-
"loss": 0.
|
| 767 |
"step": 10600
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 2.675,
|
| 771 |
-
"grad_norm": 0.
|
| 772 |
"learning_rate": 0.0003742908165631636,
|
| 773 |
-
"loss": 0.
|
| 774 |
"step": 10700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 2.7,
|
| 778 |
-
"grad_norm": 0.
|
| 779 |
"learning_rate": 0.0003721553103742388,
|
| 780 |
"loss": 1.0125,
|
| 781 |
"step": 10800
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 2.725,
|
| 785 |
-
"grad_norm": 0.
|
| 786 |
"learning_rate": 0.0003700080306022528,
|
| 787 |
-
"loss": 0.
|
| 788 |
"step": 10900
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 2.75,
|
| 792 |
-
"grad_norm": 0.
|
| 793 |
"learning_rate": 0.0003678491842064995,
|
| 794 |
-
"loss": 0.
|
| 795 |
"step": 11000
|
| 796 |
},
|
| 797 |
{
|
| 798 |
"epoch": 2.775,
|
| 799 |
-
"grad_norm": 0.
|
| 800 |
"learning_rate": 0.00036567897926108756,
|
| 801 |
"loss": 0.9473,
|
| 802 |
"step": 11100
|
| 803 |
},
|
| 804 |
{
|
| 805 |
"epoch": 2.8,
|
| 806 |
-
"grad_norm": 0.
|
| 807 |
"learning_rate": 0.00036349762493488667,
|
| 808 |
-
"loss": 0.
|
| 809 |
"step": 11200
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 2.825,
|
| 813 |
-
"grad_norm": 0.
|
| 814 |
"learning_rate": 0.0003613053314713671,
|
| 815 |
-
"loss": 0.
|
| 816 |
"step": 11300
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 2.85,
|
| 820 |
-
"grad_norm": 0.
|
| 821 |
"learning_rate": 0.0003591023101683355,
|
| 822 |
-
"loss": 0.
|
| 823 |
"step": 11400
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 2.875,
|
| 827 |
-
"grad_norm": 0.
|
| 828 |
"learning_rate": 0.0003568887733575705,
|
| 829 |
-
"loss": 1.
|
| 830 |
"step": 11500
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 2.9,
|
| 834 |
-
"grad_norm": 0.
|
| 835 |
"learning_rate": 0.00035466493438435703,
|
| 836 |
"loss": 0.9675,
|
| 837 |
"step": 11600
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 2.925,
|
| 841 |
-
"grad_norm": 0.
|
| 842 |
"learning_rate": 0.0003524310075869239,
|
| 843 |
-
"loss": 0.
|
| 844 |
"step": 11700
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 2.95,
|
| 848 |
-
"grad_norm": 0.
|
| 849 |
"learning_rate": 0.0003501872082757852,
|
| 850 |
-
"loss": 0.
|
| 851 |
"step": 11800
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 2.975,
|
| 855 |
-
"grad_norm": 0.
|
| 856 |
"learning_rate": 0.000347933752712989,
|
| 857 |
-
"loss": 0.
|
| 858 |
"step": 11900
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 3.0,
|
| 862 |
-
"grad_norm": 0.
|
| 863 |
"learning_rate": 0.0003456708580912725,
|
| 864 |
"loss": 0.9995,
|
| 865 |
"step": 12000
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 3.0,
|
| 869 |
-
"eval_loss": 0.
|
| 870 |
-
"eval_runtime":
|
| 871 |
-
"eval_samples_per_second":
|
| 872 |
-
"eval_steps_per_second":
|
| 873 |
"step": 12000
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 3.025,
|
| 877 |
-
"grad_norm": 0.
|
| 878 |
"learning_rate": 0.0003433987425131291,
|
| 879 |
-
"loss": 0.
|
| 880 |
"step": 12100
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 3.05,
|
| 884 |
-
"grad_norm": 0.
|
| 885 |
"learning_rate": 0.0003411176249697875,
|
| 886 |
-
"loss": 0.
|
| 887 |
"step": 12200
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 3.075,
|
| 891 |
-
"grad_norm": 0.
|
| 892 |
"learning_rate": 0.00033882772532010404,
|
| 893 |
-
"loss": 0.
|
| 894 |
"step": 12300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 3.1,
|
| 898 |
-
"grad_norm": 0.
|
| 899 |
"learning_rate": 0.0003365292642693733,
|
| 900 |
-
"loss": 0.
|
| 901 |
"step": 12400
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 3.125,
|
| 905 |
-
"grad_norm": 0.
|
| 906 |
"learning_rate": 0.00033422246334805503,
|
| 907 |
-
"loss": 0.
|
| 908 |
"step": 12500
|
| 909 |
},
|
| 910 |
{
|
| 911 |
"epoch": 3.15,
|
| 912 |
-
"grad_norm": 0.
|
| 913 |
"learning_rate": 0.0003319075448904234,
|
| 914 |
-
"loss": 0.
|
| 915 |
"step": 12600
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 3.175,
|
| 919 |
-
"grad_norm": 0.
|
| 920 |
"learning_rate": 0.00032958473201313745,
|
| 921 |
-
"loss": 0.
|
| 922 |
"step": 12700
|
| 923 |
},
|
| 924 |
{
|
| 925 |
"epoch": 3.2,
|
| 926 |
-
"grad_norm": 0.
|
| 927 |
"learning_rate": 0.00032725424859373687,
|
| 928 |
-
"loss": 0.
|
| 929 |
"step": 12800
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 3.225,
|
| 933 |
-
"grad_norm": 0.
|
| 934 |
"learning_rate": 0.00032491631924906416,
|
| 935 |
-
"loss": 0.
|
| 936 |
"step": 12900
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 3.25,
|
| 940 |
-
"grad_norm": 0.
|
| 941 |
"learning_rate": 0.00032257116931361555,
|
| 942 |
-
"loss": 0.
|
| 943 |
"step": 13000
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 3.275,
|
| 947 |
-
"grad_norm": 0.
|
| 948 |
"learning_rate": 0.00032021902481782304,
|
| 949 |
"loss": 0.9536,
|
| 950 |
"step": 13100
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 3.3,
|
| 954 |
-
"grad_norm": 0.
|
| 955 |
"learning_rate": 0.00031786011246626855,
|
| 956 |
-
"loss": 0.
|
| 957 |
"step": 13200
|
| 958 |
},
|
| 959 |
{
|
| 960 |
"epoch": 3.325,
|
| 961 |
-
"grad_norm": 0.
|
| 962 |
"learning_rate": 0.0003154946596158343,
|
| 963 |
-
"loss": 0.
|
| 964 |
"step": 13300
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 3.35,
|
| 968 |
-
"grad_norm": 0.
|
| 969 |
"learning_rate": 0.0003131228942537895,
|
| 970 |
"loss": 0.9787,
|
| 971 |
"step": 13400
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 3.375,
|
| 975 |
-
"grad_norm": 0.
|
| 976 |
"learning_rate": 0.000310745044975816,
|
| 977 |
-
"loss": 0.
|
| 978 |
"step": 13500
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 3.4,
|
| 982 |
-
"grad_norm": 0.
|
| 983 |
"learning_rate": 0.0003083613409639764,
|
| 984 |
-
"loss": 0.
|
| 985 |
"step": 13600
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 3.425,
|
| 989 |
-
"grad_norm": 0.
|
| 990 |
"learning_rate": 0.00030597201196462466,
|
| 991 |
-
"loss": 0.
|
| 992 |
"step": 13700
|
| 993 |
},
|
| 994 |
{
|
| 995 |
"epoch": 3.45,
|
| 996 |
-
"grad_norm": 0.
|
| 997 |
"learning_rate": 0.00030357728826626266,
|
| 998 |
-
"loss": 0.
|
| 999 |
"step": 13800
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 3.475,
|
| 1003 |
-
"grad_norm": 0.
|
| 1004 |
"learning_rate": 0.00030117740067734495,
|
| 1005 |
"loss": 0.9835,
|
| 1006 |
"step": 13900
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
"epoch": 3.5,
|
| 1010 |
-
"grad_norm": 0.
|
| 1011 |
"learning_rate": 0.0002987725805040321,
|
| 1012 |
-
"loss": 0.
|
| 1013 |
"step": 14000
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
"epoch": 3.525,
|
| 1017 |
-
"grad_norm": 0.
|
| 1018 |
"learning_rate": 0.0002963630595278977,
|
| 1019 |
-
"loss": 0.
|
| 1020 |
"step": 14100
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
"epoch": 3.55,
|
| 1024 |
-
"grad_norm": 0.
|
| 1025 |
"learning_rate": 0.0002939490699835887,
|
| 1026 |
-
"loss": 0.
|
| 1027 |
"step": 14200
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
"epoch": 3.575,
|
| 1031 |
-
"grad_norm": 0.
|
| 1032 |
"learning_rate": 0.00029153084453644135,
|
| 1033 |
-
"loss": 0.
|
| 1034 |
"step": 14300
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 3.6,
|
| 1038 |
-
"grad_norm": 0.
|
| 1039 |
"learning_rate": 0.00028910861626005774,
|
| 1040 |
-
"loss": 0.
|
| 1041 |
"step": 14400
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 3.625,
|
| 1045 |
-
"grad_norm": 0.
|
| 1046 |
"learning_rate": 0.00028668261861384045,
|
| 1047 |
-
"loss": 0.
|
| 1048 |
"step": 14500
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
"epoch": 3.65,
|
| 1052 |
-
"grad_norm": 0.
|
| 1053 |
"learning_rate": 0.00028425308542049207,
|
| 1054 |
-
"loss": 0.
|
| 1055 |
"step": 14600
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
"epoch": 3.675,
|
| 1059 |
-
"grad_norm": 0.
|
| 1060 |
"learning_rate": 0.0002818202508434783,
|
| 1061 |
-
"loss": 0.
|
| 1062 |
"step": 14700
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 3.7,
|
| 1066 |
-
"grad_norm": 0.
|
| 1067 |
"learning_rate": 0.00027938434936445943,
|
| 1068 |
-
"loss": 0.
|
| 1069 |
"step": 14800
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
"epoch": 3.725,
|
| 1073 |
-
"grad_norm": 0.
|
| 1074 |
"learning_rate": 0.00027694561576068985,
|
| 1075 |
-
"loss": 0.
|
| 1076 |
"step": 14900
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
"epoch": 3.75,
|
| 1080 |
-
"grad_norm": 0.
|
| 1081 |
"learning_rate": 0.0002745042850823902,
|
| 1082 |
-
"loss": 0.
|
| 1083 |
"step": 15000
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
"epoch": 3.775,
|
| 1087 |
-
"grad_norm": 0.
|
| 1088 |
"learning_rate": 0.00027206059263009243,
|
| 1089 |
-
"loss": 0.
|
| 1090 |
"step": 15100
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 3.8,
|
| 1094 |
-
"grad_norm": 0.
|
| 1095 |
"learning_rate": 0.00026961477393196127,
|
| 1096 |
-
"loss": 0.
|
| 1097 |
"step": 15200
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 3.825,
|
| 1101 |
-
"grad_norm": 0.
|
| 1102 |
"learning_rate": 0.0002671670647210934,
|
| 1103 |
-
"loss": 0.
|
| 1104 |
"step": 15300
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 3.85,
|
| 1108 |
-
"grad_norm": 0.
|
| 1109 |
"learning_rate": 0.00026471770091279724,
|
| 1110 |
-
"loss": 0.
|
| 1111 |
"step": 15400
|
| 1112 |
},
|
| 1113 |
{
|
| 1114 |
"epoch": 3.875,
|
| 1115 |
-
"grad_norm": 0.
|
| 1116 |
"learning_rate": 0.00026226691858185456,
|
| 1117 |
-
"loss": 0.
|
| 1118 |
"step": 15500
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
"epoch": 3.9,
|
| 1122 |
-
"grad_norm": 0.
|
| 1123 |
"learning_rate": 0.00025981495393976716,
|
| 1124 |
-
"loss": 0.
|
| 1125 |
"step": 15600
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
"epoch": 3.925,
|
| 1129 |
-
"grad_norm": 0.
|
| 1130 |
"learning_rate": 0.00025736204331199084,
|
| 1131 |
-
"loss": 0.
|
| 1132 |
"step": 15700
|
| 1133 |
},
|
| 1134 |
{
|
| 1135 |
"epoch": 3.95,
|
| 1136 |
-
"grad_norm": 0.
|
| 1137 |
"learning_rate": 0.00025490842311515704,
|
| 1138 |
-
"loss": 0.
|
| 1139 |
"step": 15800
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
"epoch": 3.975,
|
| 1143 |
-
"grad_norm": 0.
|
| 1144 |
"learning_rate": 0.0002524543298342875,
|
| 1145 |
-
"loss": 0.
|
| 1146 |
"step": 15900
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
"epoch": 4.0,
|
| 1150 |
-
"grad_norm": 0.
|
| 1151 |
"learning_rate": 0.00025,
|
| 1152 |
-
"loss": 0.
|
| 1153 |
"step": 16000
|
| 1154 |
},
|
| 1155 |
{
|
| 1156 |
"epoch": 4.0,
|
| 1157 |
-
"eval_loss": 0.
|
| 1158 |
-
"eval_runtime":
|
| 1159 |
-
"eval_samples_per_second":
|
| 1160 |
-
"eval_steps_per_second":
|
| 1161 |
"step": 16000
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 4.025,
|
| 1165 |
-
"grad_norm": 0.
|
| 1166 |
"learning_rate": 0.0002475456701657126,
|
| 1167 |
-
"loss": 0.
|
| 1168 |
"step": 16100
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 4.05,
|
| 1172 |
-
"grad_norm": 0.
|
| 1173 |
"learning_rate": 0.00024509157688484297,
|
| 1174 |
-
"loss": 0.
|
| 1175 |
"step": 16200
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 4.075,
|
| 1179 |
-
"grad_norm": 0.
|
| 1180 |
"learning_rate": 0.0002426379566880092,
|
| 1181 |
-
"loss": 0.
|
| 1182 |
"step": 16300
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 4.1,
|
| 1186 |
-
"grad_norm": 0.
|
| 1187 |
"learning_rate": 0.00024018504606023293,
|
| 1188 |
-
"loss": 0.
|
| 1189 |
"step": 16400
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 4.125,
|
| 1193 |
-
"grad_norm": 0.
|
| 1194 |
"learning_rate": 0.0002377330814181455,
|
| 1195 |
-
"loss": 0.
|
| 1196 |
"step": 16500
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 4.15,
|
| 1200 |
-
"grad_norm": 0.
|
| 1201 |
"learning_rate": 0.00023528229908720272,
|
| 1202 |
-
"loss": 0.
|
| 1203 |
"step": 16600
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 4.175,
|
| 1207 |
-
"grad_norm": 0.
|
| 1208 |
"learning_rate": 0.00023283293527890658,
|
| 1209 |
-
"loss": 0.
|
| 1210 |
"step": 16700
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 4.2,
|
| 1214 |
-
"grad_norm": 0.
|
| 1215 |
"learning_rate": 0.0002303852260680388,
|
| 1216 |
-
"loss": 0.
|
| 1217 |
"step": 16800
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 4.225,
|
| 1221 |
-
"grad_norm": 0.
|
| 1222 |
"learning_rate": 0.00022793940736990766,
|
| 1223 |
-
"loss": 0.
|
| 1224 |
"step": 16900
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 4.25,
|
| 1228 |
-
"grad_norm": 0.
|
| 1229 |
"learning_rate": 0.00022549571491760985,
|
| 1230 |
-
"loss": 0.
|
| 1231 |
"step": 17000
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 4.275,
|
| 1235 |
-
"grad_norm": 0.
|
| 1236 |
"learning_rate": 0.00022305438423931017,
|
| 1237 |
-
"loss": 0.
|
| 1238 |
"step": 17100
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 4.3,
|
| 1242 |
-
"grad_norm":
|
| 1243 |
"learning_rate": 0.00022061565063554063,
|
| 1244 |
-
"loss": 0.
|
| 1245 |
"step": 17200
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 4.325,
|
| 1249 |
-
"grad_norm": 0.
|
| 1250 |
"learning_rate": 0.00021817974915652172,
|
| 1251 |
-
"loss": 0.
|
| 1252 |
"step": 17300
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 4.35,
|
| 1256 |
-
"grad_norm": 0.
|
| 1257 |
"learning_rate": 0.00021574691457950805,
|
| 1258 |
-
"loss": 0.
|
| 1259 |
"step": 17400
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 4.375,
|
| 1263 |
-
"grad_norm": 0.
|
| 1264 |
"learning_rate": 0.00021331738138615958,
|
| 1265 |
-
"loss": 0.
|
| 1266 |
"step": 17500
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 4.4,
|
| 1270 |
-
"grad_norm": 0.
|
| 1271 |
"learning_rate": 0.00021089138373994224,
|
| 1272 |
-
"loss": 0.
|
| 1273 |
"step": 17600
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 4.425,
|
| 1277 |
-
"grad_norm": 0.
|
| 1278 |
"learning_rate": 0.0002084691554635587,
|
| 1279 |
-
"loss": 0.
|
| 1280 |
"step": 17700
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 4.45,
|
| 1284 |
-
"grad_norm": 0.
|
| 1285 |
"learning_rate": 0.00020605093001641137,
|
| 1286 |
-
"loss": 0.
|
| 1287 |
"step": 17800
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 4.475,
|
| 1291 |
-
"grad_norm": 0.
|
| 1292 |
"learning_rate": 0.00020363694047210228,
|
| 1293 |
-
"loss": 0.
|
| 1294 |
"step": 17900
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 4.5,
|
| 1298 |
-
"grad_norm": 0.
|
| 1299 |
"learning_rate": 0.00020122741949596797,
|
| 1300 |
-
"loss": 0.
|
| 1301 |
"step": 18000
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 4.525,
|
| 1305 |
-
"grad_norm": 0.
|
| 1306 |
"learning_rate": 0.00019882259932265512,
|
| 1307 |
-
"loss": 0.
|
| 1308 |
"step": 18100
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 4.55,
|
| 1312 |
-
"grad_norm": 0.
|
| 1313 |
"learning_rate": 0.00019642271173373735,
|
| 1314 |
-
"loss": 0.
|
| 1315 |
"step": 18200
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 4.575,
|
| 1319 |
-
"grad_norm": 0.
|
| 1320 |
"learning_rate": 0.00019402798803537538,
|
| 1321 |
-
"loss": 0.
|
| 1322 |
"step": 18300
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 4.6,
|
| 1326 |
-
"grad_norm": 0.
|
| 1327 |
"learning_rate": 0.00019163865903602372,
|
| 1328 |
"loss": 0.941,
|
| 1329 |
"step": 18400
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 4.625,
|
| 1333 |
-
"grad_norm": 0.
|
| 1334 |
"learning_rate": 0.00018925495502418406,
|
| 1335 |
-
"loss": 0.
|
| 1336 |
"step": 18500
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 4.65,
|
| 1340 |
-
"grad_norm": 0.
|
| 1341 |
"learning_rate": 0.00018687710574621051,
|
| 1342 |
-
"loss": 0.
|
| 1343 |
"step": 18600
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 4.675,
|
| 1347 |
-
"grad_norm": 0.
|
| 1348 |
"learning_rate": 0.00018450534038416566,
|
| 1349 |
-
"loss": 0.
|
| 1350 |
"step": 18700
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 4.7,
|
| 1354 |
-
"grad_norm": 0.
|
| 1355 |
"learning_rate": 0.00018213988753373146,
|
| 1356 |
-
"loss": 0.
|
| 1357 |
"step": 18800
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 4.725,
|
| 1361 |
-
"grad_norm": 0.
|
| 1362 |
"learning_rate": 0.00017978097518217702,
|
| 1363 |
-
"loss": 0.
|
| 1364 |
"step": 18900
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 4.75,
|
| 1368 |
-
"grad_norm": 0.
|
| 1369 |
"learning_rate": 0.00017742883068638446,
|
| 1370 |
-
"loss": 0.
|
| 1371 |
"step": 19000
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 4.775,
|
| 1375 |
-
"grad_norm": 0.
|
| 1376 |
"learning_rate": 0.00017508368075093582,
|
| 1377 |
-
"loss": 0.
|
| 1378 |
"step": 19100
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 4.8,
|
| 1382 |
-
"grad_norm": 0.
|
| 1383 |
"learning_rate": 0.00017274575140626317,
|
| 1384 |
-
"loss": 0.
|
| 1385 |
"step": 19200
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 4.825,
|
| 1389 |
-
"grad_norm": 0.
|
| 1390 |
"learning_rate": 0.0001704152679868626,
|
| 1391 |
-
"loss": 0.
|
| 1392 |
"step": 19300
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 4.85,
|
| 1396 |
-
"grad_norm": 0.
|
| 1397 |
"learning_rate": 0.00016809245510957666,
|
| 1398 |
-
"loss": 0.
|
| 1399 |
"step": 19400
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 4.875,
|
| 1403 |
-
"grad_norm": 0.
|
| 1404 |
"learning_rate": 0.000165777536651945,
|
| 1405 |
-
"loss": 0.
|
| 1406 |
"step": 19500
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 4.9,
|
| 1410 |
-
"grad_norm": 0.
|
| 1411 |
"learning_rate": 0.0001634707357306267,
|
| 1412 |
-
"loss": 0.
|
| 1413 |
"step": 19600
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 4.925,
|
| 1417 |
-
"grad_norm": 0.
|
| 1418 |
"learning_rate": 0.00016117227467989602,
|
| 1419 |
-
"loss": 0.
|
| 1420 |
"step": 19700
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 4.95,
|
| 1424 |
-
"grad_norm": 0.
|
| 1425 |
"learning_rate": 0.0001588823750302126,
|
| 1426 |
-
"loss": 0.
|
| 1427 |
"step": 19800
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 4.975,
|
| 1431 |
-
"grad_norm": 0.
|
| 1432 |
"learning_rate": 0.00015660125748687094,
|
| 1433 |
-
"loss": 0.
|
| 1434 |
"step": 19900
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 5.0,
|
| 1438 |
-
"grad_norm": 0.
|
| 1439 |
"learning_rate": 0.00015432914190872756,
|
| 1440 |
-
"loss": 0.
|
| 1441 |
"step": 20000
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 5.0,
|
| 1445 |
-
"eval_loss": 0.
|
| 1446 |
-
"eval_runtime":
|
| 1447 |
-
"eval_samples_per_second":
|
| 1448 |
-
"eval_steps_per_second":
|
| 1449 |
"step": 20000
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 5.025,
|
| 1453 |
-
"grad_norm": 0.
|
| 1454 |
"learning_rate": 0.000152066247287011,
|
| 1455 |
-
"loss": 0.
|
| 1456 |
"step": 20100
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 5.05,
|
| 1460 |
-
"grad_norm": 0.
|
| 1461 |
"learning_rate": 0.00014981279172421482,
|
| 1462 |
"loss": 0.9414,
|
| 1463 |
"step": 20200
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 5.075,
|
| 1467 |
-
"grad_norm": 0.
|
| 1468 |
"learning_rate": 0.00014756899241307614,
|
| 1469 |
-
"loss": 0.
|
| 1470 |
"step": 20300
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 5.1,
|
| 1474 |
-
"grad_norm": 0.
|
| 1475 |
"learning_rate": 0.00014533506561564306,
|
| 1476 |
-
"loss": 0.
|
| 1477 |
"step": 20400
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 5.125,
|
| 1481 |
-
"grad_norm": 0.
|
| 1482 |
"learning_rate": 0.00014311122664242953,
|
| 1483 |
-
"loss": 0.
|
| 1484 |
"step": 20500
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 5.15,
|
| 1488 |
-
"grad_norm": 0.
|
| 1489 |
"learning_rate": 0.00014089768983166444,
|
| 1490 |
-
"loss": 0.
|
| 1491 |
"step": 20600
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 5.175,
|
| 1495 |
-
"grad_norm": 0.
|
| 1496 |
"learning_rate": 0.000138694668528633,
|
| 1497 |
-
"loss": 0.
|
| 1498 |
"step": 20700
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 5.2,
|
| 1502 |
-
"grad_norm": 0.
|
| 1503 |
"learning_rate": 0.00013650237506511331,
|
| 1504 |
-
"loss": 0.
|
| 1505 |
"step": 20800
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 5.225,
|
| 1509 |
-
"grad_norm": 0.
|
| 1510 |
"learning_rate": 0.0001343210207389125,
|
| 1511 |
-
"loss": 0.
|
| 1512 |
"step": 20900
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 5.25,
|
| 1516 |
-
"grad_norm": 0.
|
| 1517 |
"learning_rate": 0.00013215081579350058,
|
| 1518 |
-
"loss": 0.
|
| 1519 |
"step": 21000
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 5.275,
|
| 1523 |
-
"grad_norm": 0.
|
| 1524 |
"learning_rate": 0.00012999196939774722,
|
| 1525 |
-
"loss": 0.
|
| 1526 |
"step": 21100
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 5.3,
|
| 1530 |
-
"grad_norm": 0.
|
| 1531 |
"learning_rate": 0.00012784468962576134,
|
| 1532 |
-
"loss": 0.
|
| 1533 |
"step": 21200
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 5.325,
|
| 1537 |
-
"grad_norm": 0.
|
| 1538 |
"learning_rate": 0.00012570918343683636,
|
| 1539 |
-
"loss": 0.
|
| 1540 |
"step": 21300
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 5.35,
|
| 1544 |
-
"grad_norm": 0.
|
| 1545 |
"learning_rate": 0.0001235856566555039,
|
| 1546 |
-
"loss": 0.
|
| 1547 |
"step": 21400
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 5.375,
|
| 1551 |
-
"grad_norm": 0.
|
| 1552 |
"learning_rate": 0.0001214743139516946,
|
| 1553 |
-
"loss": 0.
|
| 1554 |
"step": 21500
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 5.4,
|
| 1558 |
-
"grad_norm": 0.
|
| 1559 |
"learning_rate": 0.00011937535882101281,
|
| 1560 |
-
"loss": 0.
|
| 1561 |
"step": 21600
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 5.425,
|
| 1565 |
-
"grad_norm": 0.
|
| 1566 |
"learning_rate": 0.00011728899356512265,
|
| 1567 |
-
"loss": 0.
|
| 1568 |
"step": 21700
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 5.45,
|
| 1572 |
-
"grad_norm": 0.
|
| 1573 |
"learning_rate": 0.00011521541927224994,
|
| 1574 |
-
"loss": 0.
|
| 1575 |
"step": 21800
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 5.475,
|
| 1579 |
-
"grad_norm": 0.
|
| 1580 |
"learning_rate": 0.00011315483579780094,
|
| 1581 |
-
"loss": 0.
|
| 1582 |
"step": 21900
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 5.5,
|
| 1586 |
-
"grad_norm": 0.
|
| 1587 |
"learning_rate": 0.00011110744174509952,
|
| 1588 |
-
"loss": 0.
|
| 1589 |
"step": 22000
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 5.525,
|
| 1593 |
-
"grad_norm": 0.
|
| 1594 |
"learning_rate": 0.00010907343444624579,
|
| 1595 |
-
"loss": 0.
|
| 1596 |
"step": 22100
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 5.55,
|
| 1600 |
-
"grad_norm": 0.
|
| 1601 |
"learning_rate": 0.00010705300994309697,
|
| 1602 |
"loss": 0.9286,
|
| 1603 |
"step": 22200
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 5.575,
|
| 1607 |
-
"grad_norm": 0.
|
| 1608 |
"learning_rate": 0.00010504636296837161,
|
| 1609 |
-
"loss": 0.
|
| 1610 |
"step": 22300
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 5.6,
|
| 1614 |
-
"grad_norm": 0.
|
| 1615 |
"learning_rate": 0.00010305368692688174,
|
| 1616 |
-
"loss": 0.
|
| 1617 |
"step": 22400
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 5.625,
|
| 1621 |
-
"grad_norm": 0.
|
| 1622 |
"learning_rate": 0.00010107517387689166,
|
| 1623 |
-
"loss": 0.
|
| 1624 |
"step": 22500
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 5.65,
|
| 1628 |
-
"grad_norm": 0.
|
| 1629 |
"learning_rate": 9.911101451160715e-05,
|
| 1630 |
-
"loss": 0.
|
| 1631 |
"step": 22600
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 5.675,
|
| 1635 |
-
"grad_norm": 0.
|
| 1636 |
"learning_rate": 9.716139814079594e-05,
|
| 1637 |
-
"loss": 0.
|
| 1638 |
"step": 22700
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 5.7,
|
| 1642 |
-
"grad_norm": 0.
|
| 1643 |
"learning_rate": 9.522651267254148e-05,
|
| 1644 |
-
"loss": 0.
|
| 1645 |
"step": 22800
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
"epoch": 5.725,
|
| 1649 |
-
"grad_norm": 0.
|
| 1650 |
"learning_rate": 9.330654459513265e-05,
|
| 1651 |
-
"loss": 0.
|
| 1652 |
"step": 22900
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
"epoch": 5.75,
|
| 1656 |
-
"grad_norm": 0.
|
| 1657 |
"learning_rate": 9.140167895908866e-05,
|
| 1658 |
-
"loss": 0.
|
| 1659 |
"step": 23000
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
"epoch": 5.775,
|
| 1663 |
-
"grad_norm": 0.
|
| 1664 |
"learning_rate": 8.951209935932425e-05,
|
| 1665 |
"loss": 0.9113,
|
| 1666 |
"step": 23100
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
"epoch": 5.8,
|
| 1670 |
-
"grad_norm": 0.
|
| 1671 |
"learning_rate": 8.763798791745412e-05,
|
| 1672 |
-
"loss": 0.
|
| 1673 |
"step": 23200
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
"epoch": 5.825,
|
| 1677 |
-
"grad_norm": 0.
|
| 1678 |
"learning_rate": 8.577952526423969e-05,
|
| 1679 |
-
"loss": 0.
|
| 1680 |
"step": 23300
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
"epoch": 5.85,
|
| 1684 |
-
"grad_norm": 0.
|
| 1685 |
"learning_rate": 8.393689052217964e-05,
|
| 1686 |
-
"loss": 0.
|
| 1687 |
"step": 23400
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
"epoch": 5.875,
|
| 1691 |
-
"grad_norm": 0.
|
| 1692 |
"learning_rate": 8.211026128824539e-05,
|
| 1693 |
-
"loss": 0.
|
| 1694 |
"step": 23500
|
| 1695 |
},
|
| 1696 |
{
|
| 1697 |
"epoch": 5.9,
|
| 1698 |
-
"grad_norm": 0.
|
| 1699 |
"learning_rate": 8.029981361676455e-05,
|
| 1700 |
-
"loss": 0.
|
| 1701 |
"step": 23600
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 5.925,
|
| 1705 |
-
"grad_norm": 0.
|
| 1706 |
"learning_rate": 7.850572200245185e-05,
|
| 1707 |
-
"loss": 0.
|
| 1708 |
"step": 23700
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
"epoch": 5.95,
|
| 1712 |
-
"grad_norm": 0.
|
| 1713 |
"learning_rate": 7.672815936359106e-05,
|
| 1714 |
-
"loss": 0.
|
| 1715 |
"step": 23800
|
| 1716 |
},
|
| 1717 |
{
|
| 1718 |
"epoch": 5.975,
|
| 1719 |
-
"grad_norm": 0.
|
| 1720 |
"learning_rate": 7.496729702536912e-05,
|
| 1721 |
-
"loss": 0.
|
| 1722 |
"step": 23900
|
| 1723 |
},
|
| 1724 |
{
|
| 1725 |
"epoch": 6.0,
|
| 1726 |
-
"grad_norm": 0.
|
| 1727 |
"learning_rate": 7.322330470336314e-05,
|
| 1728 |
-
"loss": 0.
|
| 1729 |
"step": 24000
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
"epoch": 6.0,
|
| 1733 |
-
"eval_loss": 0.
|
| 1734 |
-
"eval_runtime":
|
| 1735 |
-
"eval_samples_per_second":
|
| 1736 |
-
"eval_steps_per_second":
|
| 1737 |
"step": 24000
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 6.025,
|
| 1741 |
-
"grad_norm": 0.
|
| 1742 |
"learning_rate": 7.149635048718294e-05,
|
| 1743 |
-
"loss": 0.
|
| 1744 |
"step": 24100
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 6.05,
|
| 1748 |
-
"grad_norm": 0.
|
| 1749 |
"learning_rate": 6.97866008242703e-05,
|
| 1750 |
"loss": 0.9105,
|
| 1751 |
"step": 24200
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 6.075,
|
| 1755 |
-
"grad_norm": 0.
|
| 1756 |
"learning_rate": 6.809422050385628e-05,
|
| 1757 |
-
"loss": 0.
|
| 1758 |
"step": 24300
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 6.1,
|
| 1762 |
-
"grad_norm": 0.
|
| 1763 |
"learning_rate": 6.641937264107867e-05,
|
| 1764 |
-
"loss": 0.
|
| 1765 |
"step": 24400
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 6.125,
|
| 1769 |
-
"grad_norm": 0.
|
| 1770 |
"learning_rate": 6.476221866126028e-05,
|
| 1771 |
-
"loss": 0.
|
| 1772 |
"step": 24500
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 6.15,
|
| 1776 |
-
"grad_norm": 0.
|
| 1777 |
"learning_rate": 6.312291828435076e-05,
|
| 1778 |
-
"loss": 0.
|
| 1779 |
"step": 24600
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 6.175,
|
| 1783 |
-
"grad_norm": 0.
|
| 1784 |
"learning_rate": 6.150162950953264e-05,
|
| 1785 |
-
"loss": 0.
|
| 1786 |
"step": 24700
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 6.2,
|
| 1790 |
-
"grad_norm": 0.
|
| 1791 |
"learning_rate": 5.989850859999227e-05,
|
| 1792 |
-
"loss": 0.
|
| 1793 |
"step": 24800
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 6.225,
|
| 1797 |
-
"grad_norm": 0.
|
| 1798 |
"learning_rate": 5.831371006785963e-05,
|
| 1799 |
-
"loss": 0.
|
| 1800 |
"step": 24900
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 6.25,
|
| 1804 |
-
"grad_norm": 0.
|
| 1805 |
"learning_rate": 5.6747386659315755e-05,
|
| 1806 |
-
"loss": 0.
|
| 1807 |
"step": 25000
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 6.275,
|
| 1811 |
-
"grad_norm": 0.
|
| 1812 |
"learning_rate": 5.519968933987082e-05,
|
| 1813 |
-
"loss": 0.
|
| 1814 |
"step": 25100
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 6.3,
|
| 1818 |
-
"grad_norm": 0.
|
| 1819 |
"learning_rate": 5.367076727981382e-05,
|
| 1820 |
-
"loss": 0.
|
| 1821 |
"step": 25200
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 6.325,
|
| 1825 |
-
"grad_norm": 0.
|
| 1826 |
"learning_rate": 5.216076783983492e-05,
|
| 1827 |
-
"loss": 0.
|
| 1828 |
"step": 25300
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 6.35,
|
| 1832 |
-
"grad_norm": 0.
|
| 1833 |
"learning_rate": 5.066983655682325e-05,
|
| 1834 |
-
"loss": 0.
|
| 1835 |
"step": 25400
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 6.375,
|
| 1839 |
-
"grad_norm": 0.
|
| 1840 |
"learning_rate": 4.919811712983879e-05,
|
| 1841 |
-
"loss": 0.
|
| 1842 |
"step": 25500
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 6.4,
|
| 1846 |
-
"grad_norm": 0.
|
| 1847 |
"learning_rate": 4.7745751406263163e-05,
|
| 1848 |
-
"loss": 0.
|
| 1849 |
"step": 25600
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 6.425,
|
| 1853 |
-
"grad_norm": 0.
|
| 1854 |
"learning_rate": 4.6312879368127645e-05,
|
| 1855 |
-
"loss": 0.
|
| 1856 |
"step": 25700
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 6.45,
|
| 1860 |
-
"grad_norm": 0.
|
| 1861 |
"learning_rate": 4.4899639118621604e-05,
|
| 1862 |
-
"loss": 0.
|
| 1863 |
"step": 25800
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 6.475,
|
| 1867 |
-
"grad_norm": 0.
|
| 1868 |
"learning_rate": 4.350616686878175e-05,
|
| 1869 |
-
"loss": 0.
|
| 1870 |
"step": 25900
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 6.5,
|
| 1874 |
-
"grad_norm": 0.
|
| 1875 |
"learning_rate": 4.213259692436367e-05,
|
| 1876 |
-
"loss": 0.
|
| 1877 |
"step": 26000
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 6.525,
|
| 1881 |
-
"grad_norm": 0.
|
| 1882 |
"learning_rate": 4.077906167289766e-05,
|
| 1883 |
-
"loss": 0.
|
| 1884 |
"step": 26100
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 6.55,
|
| 1888 |
-
"grad_norm": 0.
|
| 1889 |
"learning_rate": 3.944569157092839e-05,
|
| 1890 |
-
"loss": 0.
|
| 1891 |
"step": 26200
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 6.575,
|
| 1895 |
-
"grad_norm": 0.
|
| 1896 |
"learning_rate": 3.8132615131441396e-05,
|
| 1897 |
-
"loss": 0.
|
| 1898 |
"step": 26300
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 6.6,
|
| 1902 |
-
"grad_norm": 0.
|
| 1903 |
"learning_rate": 3.6839958911476953e-05,
|
| 1904 |
-
"loss": 0.
|
| 1905 |
"step": 26400
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 6.625,
|
| 1909 |
-
"grad_norm": 0.
|
| 1910 |
"learning_rate": 3.5567847499932e-05,
|
| 1911 |
"loss": 0.9453,
|
| 1912 |
"step": 26500
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 6.65,
|
| 1916 |
-
"grad_norm": 0.
|
| 1917 |
"learning_rate": 3.431640350555204e-05,
|
| 1918 |
-
"loss": 0.
|
| 1919 |
"step": 26600
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 6.675,
|
| 1923 |
-
"grad_norm": 0.
|
| 1924 |
"learning_rate": 3.308574754511404e-05,
|
| 1925 |
-
"loss": 0.
|
| 1926 |
"step": 26700
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 6.7,
|
| 1930 |
-
"grad_norm": 0.
|
| 1931 |
"learning_rate": 3.187599823180071e-05,
|
| 1932 |
-
"loss": 0.
|
| 1933 |
"step": 26800
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 6.725,
|
| 1937 |
-
"grad_norm": 0.
|
| 1938 |
"learning_rate": 3.0687272163768986e-05,
|
| 1939 |
-
"loss": 0.
|
| 1940 |
"step": 26900
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 6.75,
|
| 1944 |
-
"grad_norm": 0.
|
| 1945 |
"learning_rate": 2.9519683912911265e-05,
|
| 1946 |
-
"loss": 0.
|
| 1947 |
"step": 27000
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 6.775,
|
| 1951 |
-
"grad_norm": 0.
|
| 1952 |
"learning_rate": 2.8373346013813417e-05,
|
| 1953 |
-
"loss": 0.
|
| 1954 |
"step": 27100
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 6.8,
|
| 1958 |
-
"grad_norm": 0.
|
| 1959 |
"learning_rate": 2.7248368952908055e-05,
|
| 1960 |
-
"loss": 0.
|
| 1961 |
"step": 27200
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 6.825,
|
| 1965 |
-
"grad_norm": 0.
|
| 1966 |
"learning_rate": 2.6144861157825773e-05,
|
| 1967 |
-
"loss": 0.
|
| 1968 |
"step": 27300
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 6.85,
|
| 1972 |
-
"grad_norm": 0.
|
| 1973 |
"learning_rate": 2.5062928986944677e-05,
|
| 1974 |
-
"loss": 0.
|
| 1975 |
"step": 27400
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 6.875,
|
| 1979 |
-
"grad_norm": 0.
|
| 1980 |
"learning_rate": 2.4002676719139166e-05,
|
| 1981 |
"loss": 0.9394,
|
| 1982 |
"step": 27500
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 6.9,
|
| 1986 |
-
"grad_norm": 0.
|
| 1987 |
"learning_rate": 2.296420654372966e-05,
|
| 1988 |
-
"loss": 0.
|
| 1989 |
"step": 27600
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 6.925,
|
| 1993 |
-
"grad_norm": 0.
|
| 1994 |
"learning_rate": 2.1947618550633096e-05,
|
| 1995 |
-
"loss": 0.
|
| 1996 |
"step": 27700
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 6.95,
|
| 2000 |
-
"grad_norm": 0.
|
| 2001 |
"learning_rate": 2.0953010720716037e-05,
|
| 2002 |
-
"loss": 0.
|
| 2003 |
"step": 27800
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 6.975,
|
| 2007 |
-
"grad_norm": 0.
|
| 2008 |
"learning_rate": 1.9980478916351297e-05,
|
| 2009 |
-
"loss": 0.
|
| 2010 |
"step": 27900
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 7.0,
|
| 2014 |
-
"grad_norm": 0.
|
| 2015 |
"learning_rate": 1.9030116872178316e-05,
|
| 2016 |
-
"loss": 0.
|
| 2017 |
"step": 28000
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 7.0,
|
| 2021 |
-
"eval_loss": 0.
|
| 2022 |
-
"eval_runtime":
|
| 2023 |
-
"eval_samples_per_second":
|
| 2024 |
-
"eval_steps_per_second":
|
| 2025 |
"step": 28000
|
| 2026 |
},
|
| 2027 |
{
|
| 2028 |
"epoch": 7.025,
|
| 2029 |
-
"grad_norm": 0.
|
| 2030 |
"learning_rate": 1.8102016186068992e-05,
|
| 2031 |
-
"loss": 0.
|
| 2032 |
"step": 28100
|
| 2033 |
},
|
| 2034 |
{
|
| 2035 |
"epoch": 7.05,
|
| 2036 |
-
"grad_norm": 0.
|
| 2037 |
"learning_rate": 1.719626631029911e-05,
|
| 2038 |
-
"loss": 0.
|
| 2039 |
"step": 28200
|
| 2040 |
},
|
| 2041 |
{
|
| 2042 |
"epoch": 7.075,
|
| 2043 |
-
"grad_norm": 0.
|
| 2044 |
"learning_rate": 1.6312954542926888e-05,
|
| 2045 |
-
"loss": 0.
|
| 2046 |
"step": 28300
|
| 2047 |
},
|
| 2048 |
{
|
| 2049 |
"epoch": 7.1,
|
| 2050 |
-
"grad_norm": 0.
|
| 2051 |
"learning_rate": 1.5452166019378987e-05,
|
| 2052 |
-
"loss": 0.
|
| 2053 |
"step": 28400
|
| 2054 |
},
|
| 2055 |
{
|
| 2056 |
"epoch": 7.125,
|
| 2057 |
-
"grad_norm": 0.
|
| 2058 |
"learning_rate": 1.4613983704244827e-05,
|
| 2059 |
-
"loss": 0.
|
| 2060 |
"step": 28500
|
| 2061 |
},
|
| 2062 |
{
|
| 2063 |
"epoch": 7.15,
|
| 2064 |
-
"grad_norm": 0.
|
| 2065 |
"learning_rate": 1.3798488383280488e-05,
|
| 2066 |
-
"loss": 0.
|
| 2067 |
"step": 28600
|
| 2068 |
},
|
| 2069 |
{
|
| 2070 |
"epoch": 7.175,
|
| 2071 |
-
"grad_norm": 0.
|
| 2072 |
"learning_rate": 1.3005758655622424e-05,
|
| 2073 |
-
"loss": 0.
|
| 2074 |
"step": 28700
|
| 2075 |
},
|
| 2076 |
{
|
| 2077 |
"epoch": 7.2,
|
| 2078 |
-
"grad_norm": 0.
|
| 2079 |
"learning_rate": 1.2235870926211617e-05,
|
| 2080 |
-
"loss": 0.
|
| 2081 |
"step": 28800
|
| 2082 |
},
|
| 2083 |
{
|
| 2084 |
"epoch": 7.225,
|
| 2085 |
-
"grad_norm": 0.
|
| 2086 |
"learning_rate": 1.1488899398429897e-05,
|
| 2087 |
-
"loss": 0.
|
| 2088 |
"step": 28900
|
| 2089 |
},
|
| 2090 |
{
|
| 2091 |
"epoch": 7.25,
|
| 2092 |
-
"grad_norm": 0.
|
| 2093 |
"learning_rate": 1.0764916066947795e-05,
|
| 2094 |
-
"loss": 0.
|
| 2095 |
"step": 29000
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
"epoch": 7.275,
|
| 2099 |
-
"grad_norm": 0.
|
| 2100 |
"learning_rate": 1.0063990710785648e-05,
|
| 2101 |
-
"loss": 0.
|
| 2102 |
"step": 29100
|
| 2103 |
},
|
| 2104 |
{
|
| 2105 |
"epoch": 7.3,
|
| 2106 |
-
"grad_norm": 0.
|
| 2107 |
"learning_rate": 9.386190886588208e-06,
|
| 2108 |
-
"loss": 0.
|
| 2109 |
"step": 29200
|
| 2110 |
},
|
| 2111 |
{
|
| 2112 |
"epoch": 7.325,
|
| 2113 |
-
"grad_norm": 0.
|
| 2114 |
"learning_rate": 8.731581922113152e-06,
|
| 2115 |
-
"loss": 0.
|
| 2116 |
"step": 29300
|
| 2117 |
},
|
| 2118 |
{
|
| 2119 |
"epoch": 7.35,
|
| 2120 |
-
"grad_norm": 0.
|
| 2121 |
"learning_rate": 8.10022690993506e-06,
|
| 2122 |
-
"loss": 0.
|
| 2123 |
"step": 29400
|
| 2124 |
},
|
| 2125 |
{
|
| 2126 |
"epoch": 7.375,
|
| 2127 |
-
"grad_norm": 0.
|
| 2128 |
"learning_rate": 7.4921867013640064e-06,
|
| 2129 |
-
"loss": 0.
|
| 2130 |
"step": 29500
|
| 2131 |
},
|
| 2132 |
{
|
| 2133 |
"epoch": 7.4,
|
| 2134 |
-
"grad_norm": 0.
|
| 2135 |
"learning_rate": 6.907519900580861e-06,
|
| 2136 |
-
"loss": 0.
|
| 2137 |
"step": 29600
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
"epoch": 7.425,
|
| 2141 |
-
"grad_norm": 0.
|
| 2142 |
"learning_rate": 6.34628285898875e-06,
|
| 2143 |
-
"loss": 0.
|
| 2144 |
"step": 29700
|
| 2145 |
},
|
| 2146 |
{
|
| 2147 |
"epoch": 7.45,
|
| 2148 |
-
"grad_norm": 0.
|
| 2149 |
"learning_rate": 5.808529669781903e-06,
|
| 2150 |
-
"loss": 0.
|
| 2151 |
"step": 29800
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
"epoch": 7.475,
|
| 2155 |
-
"grad_norm": 0.
|
| 2156 |
"learning_rate": 5.294312162731935e-06,
|
| 2157 |
-
"loss": 0.
|
| 2158 |
"step": 29900
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
"epoch": 7.5,
|
| 2162 |
-
"grad_norm": 0.
|
| 2163 |
"learning_rate": 4.803679899192393e-06,
|
| 2164 |
-
"loss": 0.
|
| 2165 |
"step": 30000
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
"epoch": 7.525,
|
| 2169 |
-
"grad_norm": 0.
|
| 2170 |
"learning_rate": 4.336680167322055e-06,
|
| 2171 |
-
"loss": 0.
|
| 2172 |
"step": 30100
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
"epoch": 7.55,
|
| 2176 |
-
"grad_norm": 0.
|
| 2177 |
"learning_rate": 3.893357977527101e-06,
|
| 2178 |
-
"loss": 0.
|
| 2179 |
"step": 30200
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
"epoch": 7.575,
|
| 2183 |
-
"grad_norm": 0.
|
| 2184 |
"learning_rate": 3.4737560581228343e-06,
|
| 2185 |
-
"loss": 0.
|
| 2186 |
"step": 30300
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
"epoch": 7.6,
|
| 2190 |
-
"grad_norm": 0.
|
| 2191 |
"learning_rate": 3.077914851215585e-06,
|
| 2192 |
-
"loss": 0.
|
| 2193 |
"step": 30400
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
"epoch": 7.625,
|
| 2197 |
-
"grad_norm": 0.
|
| 2198 |
"learning_rate": 2.7058725088047465e-06,
|
| 2199 |
-
"loss": 0.
|
| 2200 |
"step": 30500
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
"epoch": 7.65,
|
| 2204 |
-
"grad_norm": 0.
|
| 2205 |
"learning_rate": 2.357664889105687e-06,
|
| 2206 |
-
"loss": 0.
|
| 2207 |
"step": 30600
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
"epoch": 7.675,
|
| 2211 |
-
"grad_norm": 0.
|
| 2212 |
"learning_rate": 2.0333255530934903e-06,
|
| 2213 |
-
"loss": 0.
|
| 2214 |
"step": 30700
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 7.7,
|
| 2218 |
-
"grad_norm": 0.
|
| 2219 |
"learning_rate": 1.7328857612684267e-06,
|
| 2220 |
-
"loss": 0.
|
| 2221 |
"step": 30800
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
"epoch": 7.725,
|
| 2225 |
-
"grad_norm": 0.
|
| 2226 |
"learning_rate": 1.4563744706429517e-06,
|
| 2227 |
-
"loss": 0.
|
| 2228 |
"step": 30900
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
"epoch": 7.75,
|
| 2232 |
-
"grad_norm": 0.
|
| 2233 |
"learning_rate": 1.2038183319507957e-06,
|
| 2234 |
-
"loss": 0.
|
| 2235 |
"step": 31000
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
"epoch": 7.775,
|
| 2239 |
-
"grad_norm": 0.
|
| 2240 |
"learning_rate": 9.752416870782156e-07,
|
| 2241 |
-
"loss": 0.
|
| 2242 |
"step": 31100
|
| 2243 |
},
|
| 2244 |
{
|
| 2245 |
"epoch": 7.8,
|
| 2246 |
-
"grad_norm": 0.
|
| 2247 |
"learning_rate": 7.70666566718009e-07,
|
| 2248 |
-
"loss": 0.
|
| 2249 |
"step": 31200
|
| 2250 |
},
|
| 2251 |
{
|
| 2252 |
"epoch": 7.825,
|
| 2253 |
-
"grad_norm": 0.
|
| 2254 |
"learning_rate": 5.90112688246075e-07,
|
| 2255 |
-
"loss": 0.
|
| 2256 |
"step": 31300
|
| 2257 |
},
|
| 2258 |
{
|
| 2259 |
"epoch": 7.85,
|
| 2260 |
-
"grad_norm": 0.
|
| 2261 |
"learning_rate": 4.335974538210441e-07,
|
| 2262 |
-
"loss": 0.
|
| 2263 |
"step": 31400
|
| 2264 |
},
|
| 2265 |
{
|
| 2266 |
"epoch": 7.875,
|
| 2267 |
-
"grad_norm": 0.
|
| 2268 |
"learning_rate": 3.0113594870689873e-07,
|
| 2269 |
-
"loss": 0.
|
| 2270 |
"step": 31500
|
| 2271 |
}
|
| 2272 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.025,
|
| 13 |
+
"grad_norm": 0.20501533150672913,
|
| 14 |
"learning_rate": 0.000499987952239832,
|
| 15 |
+
"loss": 1.3493,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.05,
|
| 20 |
+
"grad_norm": 0.17424988746643066,
|
| 21 |
"learning_rate": 0.0004999518101205162,
|
| 22 |
+
"loss": 1.1647,
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.075,
|
| 27 |
+
"grad_norm": 0.24931852519512177,
|
| 28 |
"learning_rate": 0.0004998915771255053,
|
| 29 |
+
"loss": 1.1393,
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.1,
|
| 34 |
+
"grad_norm": 0.2063228040933609,
|
| 35 |
"learning_rate": 0.0004998072590601808,
|
| 36 |
+
"loss": 1.1057,
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.125,
|
| 41 |
+
"grad_norm": 0.23096567392349243,
|
| 42 |
"learning_rate": 0.0004996988640512931,
|
| 43 |
+
"loss": 1.1329,
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.15,
|
| 48 |
+
"grad_norm": 0.19355718791484833,
|
| 49 |
"learning_rate": 0.000499566402546179,
|
| 50 |
+
"loss": 1.1168,
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.175,
|
| 55 |
+
"grad_norm": 0.2221299558877945,
|
| 56 |
"learning_rate": 0.0004994098873117539,
|
| 57 |
+
"loss": 1.1316,
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.2,
|
| 62 |
+
"grad_norm": 0.2073143720626831,
|
| 63 |
"learning_rate": 0.000499229333433282,
|
| 64 |
+
"loss": 1.1141,
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.225,
|
| 69 |
+
"grad_norm": 0.21791169047355652,
|
| 70 |
"learning_rate": 0.0004990247583129218,
|
| 71 |
+
"loss": 1.1274,
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.25,
|
| 76 |
+
"grad_norm": 0.21467778086662292,
|
| 77 |
"learning_rate": 0.0004987961816680492,
|
| 78 |
"loss": 1.0698,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.275,
|
| 83 |
+
"grad_norm": 0.2048187404870987,
|
| 84 |
"learning_rate": 0.0004985436255293571,
|
| 85 |
+
"loss": 1.0928,
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.3,
|
| 90 |
+
"grad_norm": 0.21932852268218994,
|
| 91 |
"learning_rate": 0.0004982671142387316,
|
| 92 |
"loss": 1.0906,
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.325,
|
| 97 |
+
"grad_norm": 0.2301362156867981,
|
| 98 |
"learning_rate": 0.0004979666744469065,
|
| 99 |
+
"loss": 1.0949,
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.35,
|
| 104 |
+
"grad_norm": 0.2186669409275055,
|
| 105 |
"learning_rate": 0.0004976423351108943,
|
| 106 |
+
"loss": 1.0857,
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.375,
|
| 111 |
+
"grad_norm": 0.19886991381645203,
|
| 112 |
"learning_rate": 0.0004972941274911952,
|
| 113 |
+
"loss": 1.0642,
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.4,
|
| 118 |
+
"grad_norm": 0.23275479674339294,
|
| 119 |
"learning_rate": 0.0004969220851487844,
|
| 120 |
+
"loss": 1.0718,
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.425,
|
| 125 |
+
"grad_norm": 0.21335655450820923,
|
| 126 |
"learning_rate": 0.0004965262439418772,
|
| 127 |
"loss": 1.082,
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.45,
|
| 132 |
+
"grad_norm": 0.2357306331396103,
|
| 133 |
"learning_rate": 0.0004961066420224729,
|
| 134 |
"loss": 1.0937,
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.475,
|
| 139 |
+
"grad_norm": 0.23492847383022308,
|
| 140 |
"learning_rate": 0.000495663319832678,
|
| 141 |
+
"loss": 1.0846,
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.5,
|
| 146 |
+
"grad_norm": 0.2015356421470642,
|
| 147 |
"learning_rate": 0.0004951963201008077,
|
| 148 |
+
"loss": 1.0829,
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.525,
|
| 153 |
+
"grad_norm": 0.24353289604187012,
|
| 154 |
"learning_rate": 0.0004947056878372681,
|
| 155 |
+
"loss": 1.0688,
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.55,
|
| 160 |
+
"grad_norm": 0.27955085039138794,
|
| 161 |
"learning_rate": 0.0004941914703302181,
|
| 162 |
+
"loss": 1.0644,
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.575,
|
| 167 |
+
"grad_norm": 0.24444086849689484,
|
| 168 |
"learning_rate": 0.0004936537171410112,
|
| 169 |
+
"loss": 1.0946,
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.6,
|
| 174 |
+
"grad_norm": 0.22312547266483307,
|
| 175 |
"learning_rate": 0.0004930924800994192,
|
| 176 |
+
"loss": 1.0636,
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.625,
|
| 181 |
+
"grad_norm": 0.22395829856395721,
|
| 182 |
"learning_rate": 0.000492507813298636,
|
| 183 |
"loss": 1.0858,
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.65,
|
| 188 |
+
"grad_norm": 0.2249772697687149,
|
| 189 |
"learning_rate": 0.0004918997730900649,
|
| 190 |
+
"loss": 1.0885,
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.675,
|
| 195 |
+
"grad_norm": 0.22566111385822296,
|
| 196 |
"learning_rate": 0.0004912684180778869,
|
| 197 |
"loss": 1.0647,
|
| 198 |
"step": 2700
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.7,
|
| 202 |
+
"grad_norm": 0.256917804479599,
|
| 203 |
"learning_rate": 0.0004906138091134118,
|
| 204 |
"loss": 1.0484,
|
| 205 |
"step": 2800
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.725,
|
| 209 |
+
"grad_norm": 0.2598745822906494,
|
| 210 |
"learning_rate": 0.0004899360092892143,
|
| 211 |
+
"loss": 1.063,
|
| 212 |
"step": 2900
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.75,
|
| 216 |
+
"grad_norm": 0.2576007843017578,
|
| 217 |
"learning_rate": 0.0004892350839330522,
|
| 218 |
+
"loss": 1.0585,
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.775,
|
| 223 |
+
"grad_norm": 0.2577688992023468,
|
| 224 |
"learning_rate": 0.0004885111006015701,
|
| 225 |
+
"loss": 1.0452,
|
| 226 |
"step": 3100
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.8,
|
| 230 |
+
"grad_norm": 0.2334214746952057,
|
| 231 |
"learning_rate": 0.0004877641290737884,
|
| 232 |
+
"loss": 1.0473,
|
| 233 |
"step": 3200
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.825,
|
| 237 |
+
"grad_norm": 0.2869343161582947,
|
| 238 |
"learning_rate": 0.0004869942413443776,
|
| 239 |
+
"loss": 1.0412,
|
| 240 |
"step": 3300
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.85,
|
| 244 |
+
"grad_norm": 0.26476991176605225,
|
| 245 |
"learning_rate": 0.00048620151161671955,
|
| 246 |
+
"loss": 1.064,
|
| 247 |
"step": 3400
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.875,
|
| 251 |
+
"grad_norm": 0.2549433708190918,
|
| 252 |
"learning_rate": 0.0004853860162957552,
|
| 253 |
+
"loss": 1.0557,
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.9,
|
| 258 |
+
"grad_norm": 0.26020538806915283,
|
| 259 |
"learning_rate": 0.0004845478339806211,
|
| 260 |
+
"loss": 1.0424,
|
| 261 |
"step": 3600
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.925,
|
| 265 |
+
"grad_norm": 0.2770324945449829,
|
| 266 |
"learning_rate": 0.0004836870454570731,
|
| 267 |
+
"loss": 1.0552,
|
| 268 |
"step": 3700
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.95,
|
| 272 |
+
"grad_norm": 0.2561758756637573,
|
| 273 |
"learning_rate": 0.00048280373368970086,
|
| 274 |
+
"loss": 1.0336,
|
| 275 |
"step": 3800
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.975,
|
| 279 |
+
"grad_norm": 0.25469744205474854,
|
| 280 |
"learning_rate": 0.000481897983813931,
|
| 281 |
+
"loss": 1.06,
|
| 282 |
"step": 3900
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 1.0,
|
| 286 |
+
"grad_norm": 0.2685888707637787,
|
| 287 |
"learning_rate": 0.0004809698831278217,
|
| 288 |
+
"loss": 1.0151,
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
+
"eval_loss": 1.0406872034072876,
|
| 294 |
+
"eval_runtime": 152.4194,
|
| 295 |
+
"eval_samples_per_second": 26.243,
|
| 296 |
+
"eval_steps_per_second": 6.561,
|
| 297 |
"step": 4000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.025,
|
| 301 |
+
"grad_norm": 0.28361451625823975,
|
| 302 |
"learning_rate": 0.00048001952108364876,
|
| 303 |
+
"loss": 1.0277,
|
| 304 |
"step": 4100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 1.05,
|
| 308 |
+
"grad_norm": 0.26193392276763916,
|
| 309 |
"learning_rate": 0.00047904698927928404,
|
| 310 |
+
"loss": 1.0187,
|
| 311 |
"step": 4200
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 1.075,
|
| 315 |
+
"grad_norm": 0.2700245678424835,
|
| 316 |
"learning_rate": 0.0004780523814493669,
|
| 317 |
+
"loss": 1.0325,
|
| 318 |
"step": 4300
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 1.1,
|
| 322 |
+
"grad_norm": 0.2610750198364258,
|
| 323 |
"learning_rate": 0.00047703579345627036,
|
| 324 |
+
"loss": 1.0182,
|
| 325 |
"step": 4400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.125,
|
| 329 |
+
"grad_norm": 0.2965630888938904,
|
| 330 |
"learning_rate": 0.0004759973232808609,
|
| 331 |
+
"loss": 1.0105,
|
| 332 |
"step": 4500
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.15,
|
| 336 |
+
"grad_norm": 0.25167468190193176,
|
| 337 |
"learning_rate": 0.0004749370710130554,
|
| 338 |
+
"loss": 1.0404,
|
| 339 |
"step": 4600
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.175,
|
| 343 |
+
"grad_norm": 0.26605212688446045,
|
| 344 |
"learning_rate": 0.0004738551388421742,
|
| 345 |
"loss": 1.0291,
|
| 346 |
"step": 4700
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.2,
|
| 350 |
+
"grad_norm": 0.25520554184913635,
|
| 351 |
"learning_rate": 0.00047275163104709196,
|
| 352 |
+
"loss": 1.019,
|
| 353 |
"step": 4800
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.225,
|
| 357 |
+
"grad_norm": 0.26057741045951843,
|
| 358 |
"learning_rate": 0.00047162665398618666,
|
| 359 |
+
"loss": 1.0309,
|
| 360 |
"step": 4900
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.25,
|
| 364 |
+
"grad_norm": 0.2987058758735657,
|
| 365 |
"learning_rate": 0.00047048031608708875,
|
| 366 |
+
"loss": 1.0127,
|
| 367 |
"step": 5000
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.275,
|
| 371 |
+
"grad_norm": 0.2619779706001282,
|
| 372 |
"learning_rate": 0.00046931272783623106,
|
| 373 |
+
"loss": 1.0191,
|
| 374 |
"step": 5100
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.3,
|
| 378 |
+
"grad_norm": 0.27090466022491455,
|
| 379 |
"learning_rate": 0.0004681240017681993,
|
| 380 |
+
"loss": 1.0416,
|
| 381 |
"step": 5200
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.325,
|
| 385 |
+
"grad_norm": 0.24344772100448608,
|
| 386 |
"learning_rate": 0.00046691425245488607,
|
| 387 |
+
"loss": 1.0075,
|
| 388 |
"step": 5300
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 1.35,
|
| 392 |
+
"grad_norm": 0.2597128450870514,
|
| 393 |
"learning_rate": 0.00046568359649444796,
|
| 394 |
+
"loss": 1.0147,
|
| 395 |
"step": 5400
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.375,
|
| 399 |
+
"grad_norm": 0.3321911692619324,
|
| 400 |
"learning_rate": 0.00046443215250006805,
|
| 401 |
+
"loss": 1.0277,
|
| 402 |
"step": 5500
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.4,
|
| 406 |
+
"grad_norm": 0.2736588716506958,
|
| 407 |
"learning_rate": 0.00046316004108852305,
|
| 408 |
+
"loss": 1.041,
|
| 409 |
"step": 5600
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 1.425,
|
| 413 |
+
"grad_norm": 0.30011600255966187,
|
| 414 |
"learning_rate": 0.0004618673848685586,
|
| 415 |
+
"loss": 1.006,
|
| 416 |
"step": 5700
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 1.45,
|
| 420 |
+
"grad_norm": 0.27296823263168335,
|
| 421 |
"learning_rate": 0.0004605543084290716,
|
| 422 |
+
"loss": 1.0207,
|
| 423 |
"step": 5800
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 1.475,
|
| 427 |
+
"grad_norm": 0.2965295910835266,
|
| 428 |
"learning_rate": 0.0004592209383271023,
|
| 429 |
+
"loss": 1.0306,
|
| 430 |
"step": 5900
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 1.5,
|
| 434 |
+
"grad_norm": 0.27299028635025024,
|
| 435 |
"learning_rate": 0.00045786740307563633,
|
| 436 |
+
"loss": 1.0296,
|
| 437 |
"step": 6000
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 1.525,
|
| 441 |
+
"grad_norm": 0.2956341803073883,
|
| 442 |
"learning_rate": 0.0004564938331312183,
|
| 443 |
+
"loss": 0.9965,
|
| 444 |
"step": 6100
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 1.55,
|
| 448 |
+
"grad_norm": 0.2554098665714264,
|
| 449 |
"learning_rate": 0.0004551003608813784,
|
| 450 |
+
"loss": 1.0232,
|
| 451 |
"step": 6200
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 1.575,
|
| 455 |
+
"grad_norm": 0.2712690234184265,
|
| 456 |
"learning_rate": 0.00045368712063187237,
|
| 457 |
+
"loss": 1.02,
|
| 458 |
"step": 6300
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 1.6,
|
| 462 |
+
"grad_norm": 0.31331920623779297,
|
| 463 |
"learning_rate": 0.0004522542485937369,
|
| 464 |
+
"loss": 0.971,
|
| 465 |
"step": 6400
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 1.625,
|
| 469 |
+
"grad_norm": 0.3378168046474457,
|
| 470 |
"learning_rate": 0.0004508018828701612,
|
| 471 |
+
"loss": 1.0017,
|
| 472 |
"step": 6500
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 1.65,
|
| 476 |
+
"grad_norm": 0.3126436471939087,
|
| 477 |
"learning_rate": 0.0004493301634431768,
|
| 478 |
+
"loss": 0.9969,
|
| 479 |
"step": 6600
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 1.675,
|
| 483 |
+
"grad_norm": 0.2890426218509674,
|
| 484 |
"learning_rate": 0.00044783923216016507,
|
| 485 |
+
"loss": 1.0082,
|
| 486 |
"step": 6700
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 1.7,
|
| 490 |
+
"grad_norm": 0.3303642272949219,
|
| 491 |
"learning_rate": 0.0004463292327201862,
|
| 492 |
+
"loss": 0.9989,
|
| 493 |
"step": 6800
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 1.725,
|
| 497 |
+
"grad_norm": 0.3025239408016205,
|
| 498 |
"learning_rate": 0.00044480031066012916,
|
| 499 |
+
"loss": 1.0305,
|
| 500 |
"step": 6900
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 1.75,
|
| 504 |
+
"grad_norm": 0.2934839725494385,
|
| 505 |
"learning_rate": 0.0004432526133406842,
|
| 506 |
+
"loss": 1.0003,
|
| 507 |
"step": 7000
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 1.775,
|
| 511 |
+
"grad_norm": 0.31030675768852234,
|
| 512 |
"learning_rate": 0.00044168628993214036,
|
| 513 |
+
"loss": 0.982,
|
| 514 |
"step": 7100
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 1.8,
|
| 518 |
+
"grad_norm": 0.30027347803115845,
|
| 519 |
"learning_rate": 0.0004401014914000078,
|
| 520 |
+
"loss": 0.9825,
|
| 521 |
"step": 7200
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 1.825,
|
| 525 |
+
"grad_norm": 0.2832087278366089,
|
| 526 |
"learning_rate": 0.00043849837049046735,
|
| 527 |
"loss": 1.007,
|
| 528 |
"step": 7300
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 1.85,
|
| 532 |
+
"grad_norm": 0.30503541231155396,
|
| 533 |
"learning_rate": 0.00043687708171564923,
|
| 534 |
+
"loss": 1.0023,
|
| 535 |
"step": 7400
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 1.875,
|
| 539 |
+
"grad_norm": 0.30458569526672363,
|
| 540 |
"learning_rate": 0.0004352377813387398,
|
| 541 |
+
"loss": 1.0006,
|
| 542 |
"step": 7500
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 1.9,
|
| 546 |
+
"grad_norm": 0.3471212089061737,
|
| 547 |
"learning_rate": 0.0004335806273589214,
|
| 548 |
+
"loss": 1.0239,
|
| 549 |
"step": 7600
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 1.925,
|
| 553 |
+
"grad_norm": 0.29310837388038635,
|
| 554 |
"learning_rate": 0.00043190577949614375,
|
| 555 |
+
"loss": 1.0129,
|
| 556 |
"step": 7700
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.95,
|
| 560 |
+
"grad_norm": 0.294300377368927,
|
| 561 |
"learning_rate": 0.0004302133991757297,
|
| 562 |
+
"loss": 0.9956,
|
| 563 |
"step": 7800
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 1.975,
|
| 567 |
+
"grad_norm": 0.3151143193244934,
|
| 568 |
"learning_rate": 0.00042850364951281707,
|
| 569 |
+
"loss": 1.0224,
|
| 570 |
"step": 7900
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 2.0,
|
| 574 |
+
"grad_norm": 0.3241264820098877,
|
| 575 |
"learning_rate": 0.00042677669529663686,
|
| 576 |
+
"loss": 1.0234,
|
| 577 |
"step": 8000
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 2.0,
|
| 581 |
+
"eval_loss": 1.0086950063705444,
|
| 582 |
+
"eval_runtime": 152.329,
|
| 583 |
+
"eval_samples_per_second": 26.259,
|
| 584 |
+
"eval_steps_per_second": 6.565,
|
| 585 |
"step": 8000
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 2.025,
|
| 589 |
+
"grad_norm": 0.3218720853328705,
|
| 590 |
"learning_rate": 0.0004250327029746309,
|
| 591 |
+
"loss": 1.0029,
|
| 592 |
"step": 8100
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 2.05,
|
| 596 |
+
"grad_norm": 0.31330084800720215,
|
| 597 |
"learning_rate": 0.000423271840636409,
|
| 598 |
+
"loss": 0.9844,
|
| 599 |
"step": 8200
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 2.075,
|
| 603 |
+
"grad_norm": 0.30206719040870667,
|
| 604 |
"learning_rate": 0.00042149427799754817,
|
| 605 |
+
"loss": 0.989,
|
| 606 |
"step": 8300
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 2.1,
|
| 610 |
+
"grad_norm": 0.303648978471756,
|
| 611 |
"learning_rate": 0.00041970018638323546,
|
| 612 |
+
"loss": 0.9543,
|
| 613 |
"step": 8400
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 2.125,
|
| 617 |
+
"grad_norm": 0.317037433385849,
|
| 618 |
"learning_rate": 0.00041788973871175465,
|
| 619 |
+
"loss": 0.9945,
|
| 620 |
"step": 8500
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 2.15,
|
| 624 |
+
"grad_norm": 0.3210434913635254,
|
| 625 |
"learning_rate": 0.00041606310947782046,
|
| 626 |
+
"loss": 0.988,
|
| 627 |
"step": 8600
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 2.175,
|
| 631 |
+
"grad_norm": 0.33763307332992554,
|
| 632 |
"learning_rate": 0.00041422047473576033,
|
| 633 |
+
"loss": 1.0066,
|
| 634 |
"step": 8700
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 2.2,
|
| 638 |
+
"grad_norm": 0.30534419417381287,
|
| 639 |
"learning_rate": 0.0004123620120825459,
|
| 640 |
+
"loss": 1.0033,
|
| 641 |
"step": 8800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 2.225,
|
| 645 |
+
"grad_norm": 0.3363495171070099,
|
| 646 |
"learning_rate": 0.00041048790064067577,
|
| 647 |
+
"loss": 0.993,
|
| 648 |
"step": 8900
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 2.25,
|
| 652 |
+
"grad_norm": 0.2920531928539276,
|
| 653 |
"learning_rate": 0.0004085983210409114,
|
| 654 |
+
"loss": 0.9844,
|
| 655 |
"step": 9000
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 2.275,
|
| 659 |
+
"grad_norm": 0.33446577191352844,
|
| 660 |
"learning_rate": 0.0004066934554048674,
|
| 661 |
+
"loss": 0.9846,
|
| 662 |
"step": 9100
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 2.3,
|
| 666 |
+
"grad_norm": 0.33012858033180237,
|
| 667 |
"learning_rate": 0.00040477348732745853,
|
| 668 |
"loss": 0.9858,
|
| 669 |
"step": 9200
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 2.325,
|
| 673 |
+
"grad_norm": 0.3301050364971161,
|
| 674 |
"learning_rate": 0.0004028386018592041,
|
| 675 |
+
"loss": 0.9787,
|
| 676 |
"step": 9300
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 2.35,
|
| 680 |
+
"grad_norm": 0.3368571400642395,
|
| 681 |
"learning_rate": 0.0004008889854883929,
|
| 682 |
+
"loss": 0.9912,
|
| 683 |
"step": 9400
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 2.375,
|
| 687 |
+
"grad_norm": 0.285281777381897,
|
| 688 |
"learning_rate": 0.0003989248261231084,
|
| 689 |
+
"loss": 0.9655,
|
| 690 |
"step": 9500
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 2.4,
|
| 694 |
+
"grad_norm": 0.36711010336875916,
|
| 695 |
"learning_rate": 0.0003969463130731183,
|
| 696 |
+
"loss": 0.9753,
|
| 697 |
"step": 9600
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 2.425,
|
| 701 |
+
"grad_norm": 0.28881680965423584,
|
| 702 |
"learning_rate": 0.00039495363703162843,
|
| 703 |
"loss": 0.978,
|
| 704 |
"step": 9700
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 2.45,
|
| 708 |
+
"grad_norm": 0.3270690441131592,
|
| 709 |
"learning_rate": 0.000392946990056903,
|
| 710 |
+
"loss": 1.0242,
|
| 711 |
"step": 9800
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 2.475,
|
| 715 |
+
"grad_norm": 0.3338991105556488,
|
| 716 |
"learning_rate": 0.00039092656555375416,
|
| 717 |
+
"loss": 1.0066,
|
| 718 |
"step": 9900
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 2.5,
|
| 722 |
+
"grad_norm": 0.34082475304603577,
|
| 723 |
"learning_rate": 0.00038889255825490053,
|
| 724 |
+
"loss": 0.9781,
|
| 725 |
"step": 10000
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 2.525,
|
| 729 |
+
"grad_norm": 0.34000107645988464,
|
| 730 |
"learning_rate": 0.0003868451642021992,
|
| 731 |
"loss": 0.9745,
|
| 732 |
"step": 10100
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 2.55,
|
| 736 |
+
"grad_norm": 0.3570154309272766,
|
| 737 |
"learning_rate": 0.0003847845807277501,
|
| 738 |
+
"loss": 0.9759,
|
| 739 |
"step": 10200
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 2.575,
|
| 743 |
+
"grad_norm": 0.33699771761894226,
|
| 744 |
"learning_rate": 0.0003827110064348773,
|
| 745 |
+
"loss": 0.9882,
|
| 746 |
"step": 10300
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 2.6,
|
| 750 |
+
"grad_norm": 0.29268336296081543,
|
| 751 |
"learning_rate": 0.0003806246411789872,
|
| 752 |
+
"loss": 0.9745,
|
| 753 |
"step": 10400
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 2.625,
|
| 757 |
+
"grad_norm": 0.3088798224925995,
|
| 758 |
"learning_rate": 0.0003785256860483054,
|
| 759 |
+
"loss": 0.9501,
|
| 760 |
"step": 10500
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 2.65,
|
| 764 |
+
"grad_norm": 0.34749528765678406,
|
| 765 |
"learning_rate": 0.0003764143433444962,
|
| 766 |
+
"loss": 0.9702,
|
| 767 |
"step": 10600
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 2.675,
|
| 771 |
+
"grad_norm": 0.33005061745643616,
|
| 772 |
"learning_rate": 0.0003742908165631636,
|
| 773 |
+
"loss": 0.987,
|
| 774 |
"step": 10700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 2.7,
|
| 778 |
+
"grad_norm": 0.3165997266769409,
|
| 779 |
"learning_rate": 0.0003721553103742388,
|
| 780 |
"loss": 1.0125,
|
| 781 |
"step": 10800
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 2.725,
|
| 785 |
+
"grad_norm": 0.3213465213775635,
|
| 786 |
"learning_rate": 0.0003700080306022528,
|
| 787 |
+
"loss": 0.9613,
|
| 788 |
"step": 10900
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 2.75,
|
| 792 |
+
"grad_norm": 0.33442971110343933,
|
| 793 |
"learning_rate": 0.0003678491842064995,
|
| 794 |
+
"loss": 0.9825,
|
| 795 |
"step": 11000
|
| 796 |
},
|
| 797 |
{
|
| 798 |
"epoch": 2.775,
|
| 799 |
+
"grad_norm": 0.2766638994216919,
|
| 800 |
"learning_rate": 0.00036567897926108756,
|
| 801 |
"loss": 0.9473,
|
| 802 |
"step": 11100
|
| 803 |
},
|
| 804 |
{
|
| 805 |
"epoch": 2.8,
|
| 806 |
+
"grad_norm": 0.3419225215911865,
|
| 807 |
"learning_rate": 0.00036349762493488667,
|
| 808 |
+
"loss": 0.9462,
|
| 809 |
"step": 11200
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 2.825,
|
| 813 |
+
"grad_norm": 0.3595256209373474,
|
| 814 |
"learning_rate": 0.0003613053314713671,
|
| 815 |
+
"loss": 0.9879,
|
| 816 |
"step": 11300
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 2.85,
|
| 820 |
+
"grad_norm": 0.3483733534812927,
|
| 821 |
"learning_rate": 0.0003591023101683355,
|
| 822 |
+
"loss": 0.9997,
|
| 823 |
"step": 11400
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 2.875,
|
| 827 |
+
"grad_norm": 0.32273510098457336,
|
| 828 |
"learning_rate": 0.0003568887733575705,
|
| 829 |
+
"loss": 1.0057,
|
| 830 |
"step": 11500
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 2.9,
|
| 834 |
+
"grad_norm": 0.31241312623023987,
|
| 835 |
"learning_rate": 0.00035466493438435703,
|
| 836 |
"loss": 0.9675,
|
| 837 |
"step": 11600
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 2.925,
|
| 841 |
+
"grad_norm": 0.37985849380493164,
|
| 842 |
"learning_rate": 0.0003524310075869239,
|
| 843 |
+
"loss": 0.9927,
|
| 844 |
"step": 11700
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 2.95,
|
| 848 |
+
"grad_norm": 0.33704909682273865,
|
| 849 |
"learning_rate": 0.0003501872082757852,
|
| 850 |
+
"loss": 0.9779,
|
| 851 |
"step": 11800
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 2.975,
|
| 855 |
+
"grad_norm": 0.4041268229484558,
|
| 856 |
"learning_rate": 0.000347933752712989,
|
| 857 |
+
"loss": 0.9739,
|
| 858 |
"step": 11900
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 3.0,
|
| 862 |
+
"grad_norm": 0.3415277302265167,
|
| 863 |
"learning_rate": 0.0003456708580912725,
|
| 864 |
"loss": 0.9995,
|
| 865 |
"step": 12000
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 3.0,
|
| 869 |
+
"eval_loss": 0.9921265244483948,
|
| 870 |
+
"eval_runtime": 153.1095,
|
| 871 |
+
"eval_samples_per_second": 26.125,
|
| 872 |
+
"eval_steps_per_second": 6.531,
|
| 873 |
"step": 12000
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 3.025,
|
| 877 |
+
"grad_norm": 0.3260078430175781,
|
| 878 |
"learning_rate": 0.0003433987425131291,
|
| 879 |
+
"loss": 0.969,
|
| 880 |
"step": 12100
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 3.05,
|
| 884 |
+
"grad_norm": 0.3035444915294647,
|
| 885 |
"learning_rate": 0.0003411176249697875,
|
| 886 |
+
"loss": 0.9636,
|
| 887 |
"step": 12200
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 3.075,
|
| 891 |
+
"grad_norm": 0.3195238411426544,
|
| 892 |
"learning_rate": 0.00033882772532010404,
|
| 893 |
+
"loss": 0.9522,
|
| 894 |
"step": 12300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 3.1,
|
| 898 |
+
"grad_norm": 0.33937081694602966,
|
| 899 |
"learning_rate": 0.0003365292642693733,
|
| 900 |
+
"loss": 0.9425,
|
| 901 |
"step": 12400
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 3.125,
|
| 905 |
+
"grad_norm": 0.37307849526405334,
|
| 906 |
"learning_rate": 0.00033422246334805503,
|
| 907 |
+
"loss": 0.9683,
|
| 908 |
"step": 12500
|
| 909 |
},
|
| 910 |
{
|
| 911 |
"epoch": 3.15,
|
| 912 |
+
"grad_norm": 0.3180018365383148,
|
| 913 |
"learning_rate": 0.0003319075448904234,
|
| 914 |
+
"loss": 0.9761,
|
| 915 |
"step": 12600
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 3.175,
|
| 919 |
+
"grad_norm": 0.3785949647426605,
|
| 920 |
"learning_rate": 0.00032958473201313745,
|
| 921 |
+
"loss": 0.9748,
|
| 922 |
"step": 12700
|
| 923 |
},
|
| 924 |
{
|
| 925 |
"epoch": 3.2,
|
| 926 |
+
"grad_norm": 0.3146439790725708,
|
| 927 |
"learning_rate": 0.00032725424859373687,
|
| 928 |
+
"loss": 0.9657,
|
| 929 |
"step": 12800
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 3.225,
|
| 933 |
+
"grad_norm": 0.3667052686214447,
|
| 934 |
"learning_rate": 0.00032491631924906416,
|
| 935 |
+
"loss": 0.9685,
|
| 936 |
"step": 12900
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 3.25,
|
| 940 |
+
"grad_norm": 0.34998300671577454,
|
| 941 |
"learning_rate": 0.00032257116931361555,
|
| 942 |
+
"loss": 0.9821,
|
| 943 |
"step": 13000
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 3.275,
|
| 947 |
+
"grad_norm": 0.34160691499710083,
|
| 948 |
"learning_rate": 0.00032021902481782304,
|
| 949 |
"loss": 0.9536,
|
| 950 |
"step": 13100
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 3.3,
|
| 954 |
+
"grad_norm": 0.35554495453834534,
|
| 955 |
"learning_rate": 0.00031786011246626855,
|
| 956 |
+
"loss": 0.9626,
|
| 957 |
"step": 13200
|
| 958 |
},
|
| 959 |
{
|
| 960 |
"epoch": 3.325,
|
| 961 |
+
"grad_norm": 0.373349666595459,
|
| 962 |
"learning_rate": 0.0003154946596158343,
|
| 963 |
+
"loss": 0.9926,
|
| 964 |
"step": 13300
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 3.35,
|
| 968 |
+
"grad_norm": 0.38320091366767883,
|
| 969 |
"learning_rate": 0.0003131228942537895,
|
| 970 |
"loss": 0.9787,
|
| 971 |
"step": 13400
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 3.375,
|
| 975 |
+
"grad_norm": 0.3836865723133087,
|
| 976 |
"learning_rate": 0.000310745044975816,
|
| 977 |
+
"loss": 0.9583,
|
| 978 |
"step": 13500
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 3.4,
|
| 982 |
+
"grad_norm": 0.3358447551727295,
|
| 983 |
"learning_rate": 0.0003083613409639764,
|
| 984 |
+
"loss": 0.953,
|
| 985 |
"step": 13600
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 3.425,
|
| 989 |
+
"grad_norm": 0.3501192629337311,
|
| 990 |
"learning_rate": 0.00030597201196462466,
|
| 991 |
+
"loss": 0.9346,
|
| 992 |
"step": 13700
|
| 993 |
},
|
| 994 |
{
|
| 995 |
"epoch": 3.45,
|
| 996 |
+
"grad_norm": 0.3595263957977295,
|
| 997 |
"learning_rate": 0.00030357728826626266,
|
| 998 |
+
"loss": 0.9656,
|
| 999 |
"step": 13800
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 3.475,
|
| 1003 |
+
"grad_norm": 0.3395259380340576,
|
| 1004 |
"learning_rate": 0.00030117740067734495,
|
| 1005 |
"loss": 0.9835,
|
| 1006 |
"step": 13900
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
"epoch": 3.5,
|
| 1010 |
+
"grad_norm": 0.36751946806907654,
|
| 1011 |
"learning_rate": 0.0002987725805040321,
|
| 1012 |
+
"loss": 0.9465,
|
| 1013 |
"step": 14000
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
"epoch": 3.525,
|
| 1017 |
+
"grad_norm": 0.348893404006958,
|
| 1018 |
"learning_rate": 0.0002963630595278977,
|
| 1019 |
+
"loss": 0.9564,
|
| 1020 |
"step": 14100
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
"epoch": 3.55,
|
| 1024 |
+
"grad_norm": 0.39346957206726074,
|
| 1025 |
"learning_rate": 0.0002939490699835887,
|
| 1026 |
+
"loss": 0.9508,
|
| 1027 |
"step": 14200
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
"epoch": 3.575,
|
| 1031 |
+
"grad_norm": 0.33249151706695557,
|
| 1032 |
"learning_rate": 0.00029153084453644135,
|
| 1033 |
+
"loss": 0.9757,
|
| 1034 |
"step": 14300
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 3.6,
|
| 1038 |
+
"grad_norm": 0.3582550585269928,
|
| 1039 |
"learning_rate": 0.00028910861626005774,
|
| 1040 |
+
"loss": 0.9552,
|
| 1041 |
"step": 14400
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 3.625,
|
| 1045 |
+
"grad_norm": 0.3562643229961395,
|
| 1046 |
"learning_rate": 0.00028668261861384045,
|
| 1047 |
+
"loss": 0.95,
|
| 1048 |
"step": 14500
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
"epoch": 3.65,
|
| 1052 |
+
"grad_norm": 0.3955428898334503,
|
| 1053 |
"learning_rate": 0.00028425308542049207,
|
| 1054 |
+
"loss": 0.9699,
|
| 1055 |
"step": 14600
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
"epoch": 3.675,
|
| 1059 |
+
"grad_norm": 0.38744592666625977,
|
| 1060 |
"learning_rate": 0.0002818202508434783,
|
| 1061 |
+
"loss": 0.9838,
|
| 1062 |
"step": 14700
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 3.7,
|
| 1066 |
+
"grad_norm": 0.404632031917572,
|
| 1067 |
"learning_rate": 0.00027938434936445943,
|
| 1068 |
+
"loss": 0.9589,
|
| 1069 |
"step": 14800
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
"epoch": 3.725,
|
| 1073 |
+
"grad_norm": 0.36033356189727783,
|
| 1074 |
"learning_rate": 0.00027694561576068985,
|
| 1075 |
+
"loss": 0.973,
|
| 1076 |
"step": 14900
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
"epoch": 3.75,
|
| 1080 |
+
"grad_norm": 0.3457838296890259,
|
| 1081 |
"learning_rate": 0.0002745042850823902,
|
| 1082 |
+
"loss": 0.9801,
|
| 1083 |
"step": 15000
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
"epoch": 3.775,
|
| 1087 |
+
"grad_norm": 0.35734304785728455,
|
| 1088 |
"learning_rate": 0.00027206059263009243,
|
| 1089 |
+
"loss": 0.9469,
|
| 1090 |
"step": 15100
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 3.8,
|
| 1094 |
+
"grad_norm": 0.3542312681674957,
|
| 1095 |
"learning_rate": 0.00026961477393196127,
|
| 1096 |
+
"loss": 0.9336,
|
| 1097 |
"step": 15200
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 3.825,
|
| 1101 |
+
"grad_norm": 0.3671669065952301,
|
| 1102 |
"learning_rate": 0.0002671670647210934,
|
| 1103 |
+
"loss": 0.9793,
|
| 1104 |
"step": 15300
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 3.85,
|
| 1108 |
+
"grad_norm": 0.357881098985672,
|
| 1109 |
"learning_rate": 0.00026471770091279724,
|
| 1110 |
+
"loss": 0.9544,
|
| 1111 |
"step": 15400
|
| 1112 |
},
|
| 1113 |
{
|
| 1114 |
"epoch": 3.875,
|
| 1115 |
+
"grad_norm": 0.31071895360946655,
|
| 1116 |
"learning_rate": 0.00026226691858185456,
|
| 1117 |
+
"loss": 0.9656,
|
| 1118 |
"step": 15500
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
"epoch": 3.9,
|
| 1122 |
+
"grad_norm": 0.34388652443885803,
|
| 1123 |
"learning_rate": 0.00025981495393976716,
|
| 1124 |
+
"loss": 0.96,
|
| 1125 |
"step": 15600
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
"epoch": 3.925,
|
| 1129 |
+
"grad_norm": 0.3791946768760681,
|
| 1130 |
"learning_rate": 0.00025736204331199084,
|
| 1131 |
+
"loss": 0.9797,
|
| 1132 |
"step": 15700
|
| 1133 |
},
|
| 1134 |
{
|
| 1135 |
"epoch": 3.95,
|
| 1136 |
+
"grad_norm": 0.3630838096141815,
|
| 1137 |
"learning_rate": 0.00025490842311515704,
|
| 1138 |
+
"loss": 0.9592,
|
| 1139 |
"step": 15800
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
"epoch": 3.975,
|
| 1143 |
+
"grad_norm": 0.3712921440601349,
|
| 1144 |
"learning_rate": 0.0002524543298342875,
|
| 1145 |
+
"loss": 0.9731,
|
| 1146 |
"step": 15900
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
"epoch": 4.0,
|
| 1150 |
+
"grad_norm": 0.4159054756164551,
|
| 1151 |
"learning_rate": 0.00025,
|
| 1152 |
+
"loss": 0.9528,
|
| 1153 |
"step": 16000
|
| 1154 |
},
|
| 1155 |
{
|
| 1156 |
"epoch": 4.0,
|
| 1157 |
+
"eval_loss": 0.9823833107948303,
|
| 1158 |
+
"eval_runtime": 152.8782,
|
| 1159 |
+
"eval_samples_per_second": 26.165,
|
| 1160 |
+
"eval_steps_per_second": 6.541,
|
| 1161 |
"step": 16000
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 4.025,
|
| 1165 |
+
"grad_norm": 0.39106351137161255,
|
| 1166 |
"learning_rate": 0.0002475456701657126,
|
| 1167 |
+
"loss": 0.9814,
|
| 1168 |
"step": 16100
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 4.05,
|
| 1172 |
+
"grad_norm": 0.36192214488983154,
|
| 1173 |
"learning_rate": 0.00024509157688484297,
|
| 1174 |
+
"loss": 0.9369,
|
| 1175 |
"step": 16200
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 4.075,
|
| 1179 |
+
"grad_norm": 0.3404277563095093,
|
| 1180 |
"learning_rate": 0.0002426379566880092,
|
| 1181 |
+
"loss": 0.9501,
|
| 1182 |
"step": 16300
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 4.1,
|
| 1186 |
+
"grad_norm": 0.3791425824165344,
|
| 1187 |
"learning_rate": 0.00024018504606023293,
|
| 1188 |
+
"loss": 0.9336,
|
| 1189 |
"step": 16400
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 4.125,
|
| 1193 |
+
"grad_norm": 0.4009718894958496,
|
| 1194 |
"learning_rate": 0.0002377330814181455,
|
| 1195 |
+
"loss": 0.9487,
|
| 1196 |
"step": 16500
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 4.15,
|
| 1200 |
+
"grad_norm": 0.37880653142929077,
|
| 1201 |
"learning_rate": 0.00023528229908720272,
|
| 1202 |
+
"loss": 0.9542,
|
| 1203 |
"step": 16600
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 4.175,
|
| 1207 |
+
"grad_norm": 0.3936653733253479,
|
| 1208 |
"learning_rate": 0.00023283293527890658,
|
| 1209 |
+
"loss": 0.9388,
|
| 1210 |
"step": 16700
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 4.2,
|
| 1214 |
+
"grad_norm": 0.33316799998283386,
|
| 1215 |
"learning_rate": 0.0002303852260680388,
|
| 1216 |
+
"loss": 0.943,
|
| 1217 |
"step": 16800
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 4.225,
|
| 1221 |
+
"grad_norm": 0.3236179053783417,
|
| 1222 |
"learning_rate": 0.00022793940736990766,
|
| 1223 |
+
"loss": 0.9044,
|
| 1224 |
"step": 16900
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 4.25,
|
| 1228 |
+
"grad_norm": 0.3429449200630188,
|
| 1229 |
"learning_rate": 0.00022549571491760985,
|
| 1230 |
+
"loss": 0.9624,
|
| 1231 |
"step": 17000
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 4.275,
|
| 1235 |
+
"grad_norm": 0.35557985305786133,
|
| 1236 |
"learning_rate": 0.00022305438423931017,
|
| 1237 |
+
"loss": 0.9226,
|
| 1238 |
"step": 17100
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 4.3,
|
| 1242 |
+
"grad_norm": 1.0042773485183716,
|
| 1243 |
"learning_rate": 0.00022061565063554063,
|
| 1244 |
+
"loss": 0.9472,
|
| 1245 |
"step": 17200
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 4.325,
|
| 1249 |
+
"grad_norm": 0.3724925220012665,
|
| 1250 |
"learning_rate": 0.00021817974915652172,
|
| 1251 |
+
"loss": 0.947,
|
| 1252 |
"step": 17300
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 4.35,
|
| 1256 |
+
"grad_norm": 0.3926277756690979,
|
| 1257 |
"learning_rate": 0.00021574691457950805,
|
| 1258 |
+
"loss": 0.9478,
|
| 1259 |
"step": 17400
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 4.375,
|
| 1263 |
+
"grad_norm": 0.38053977489471436,
|
| 1264 |
"learning_rate": 0.00021331738138615958,
|
| 1265 |
+
"loss": 0.9801,
|
| 1266 |
"step": 17500
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 4.4,
|
| 1270 |
+
"grad_norm": 0.36115145683288574,
|
| 1271 |
"learning_rate": 0.00021089138373994224,
|
| 1272 |
+
"loss": 0.9619,
|
| 1273 |
"step": 17600
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 4.425,
|
| 1277 |
+
"grad_norm": 0.35729482769966125,
|
| 1278 |
"learning_rate": 0.0002084691554635587,
|
| 1279 |
+
"loss": 0.9549,
|
| 1280 |
"step": 17700
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 4.45,
|
| 1284 |
+
"grad_norm": 0.3762321174144745,
|
| 1285 |
"learning_rate": 0.00020605093001641137,
|
| 1286 |
+
"loss": 0.9359,
|
| 1287 |
"step": 17800
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 4.475,
|
| 1291 |
+
"grad_norm": 0.3317940831184387,
|
| 1292 |
"learning_rate": 0.00020363694047210228,
|
| 1293 |
+
"loss": 0.9571,
|
| 1294 |
"step": 17900
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 4.5,
|
| 1298 |
+
"grad_norm": 0.3733656406402588,
|
| 1299 |
"learning_rate": 0.00020122741949596797,
|
| 1300 |
+
"loss": 0.9409,
|
| 1301 |
"step": 18000
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 4.525,
|
| 1305 |
+
"grad_norm": 0.3790251612663269,
|
| 1306 |
"learning_rate": 0.00019882259932265512,
|
| 1307 |
+
"loss": 0.9768,
|
| 1308 |
"step": 18100
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 4.55,
|
| 1312 |
+
"grad_norm": 0.37900087237358093,
|
| 1313 |
"learning_rate": 0.00019642271173373735,
|
| 1314 |
+
"loss": 0.9208,
|
| 1315 |
"step": 18200
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 4.575,
|
| 1319 |
+
"grad_norm": 0.34171342849731445,
|
| 1320 |
"learning_rate": 0.00019402798803537538,
|
| 1321 |
+
"loss": 0.9831,
|
| 1322 |
"step": 18300
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 4.6,
|
| 1326 |
+
"grad_norm": 0.39172932505607605,
|
| 1327 |
"learning_rate": 0.00019163865903602372,
|
| 1328 |
"loss": 0.941,
|
| 1329 |
"step": 18400
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 4.625,
|
| 1333 |
+
"grad_norm": 0.36279597878456116,
|
| 1334 |
"learning_rate": 0.00018925495502418406,
|
| 1335 |
+
"loss": 0.9511,
|
| 1336 |
"step": 18500
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 4.65,
|
| 1340 |
+
"grad_norm": 0.35040661692619324,
|
| 1341 |
"learning_rate": 0.00018687710574621051,
|
| 1342 |
+
"loss": 0.9713,
|
| 1343 |
"step": 18600
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 4.675,
|
| 1347 |
+
"grad_norm": 0.35176095366477966,
|
| 1348 |
"learning_rate": 0.00018450534038416566,
|
| 1349 |
+
"loss": 0.9227,
|
| 1350 |
"step": 18700
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 4.7,
|
| 1354 |
+
"grad_norm": 0.41925740242004395,
|
| 1355 |
"learning_rate": 0.00018213988753373146,
|
| 1356 |
+
"loss": 0.9454,
|
| 1357 |
"step": 18800
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 4.725,
|
| 1361 |
+
"grad_norm": 0.4147469699382782,
|
| 1362 |
"learning_rate": 0.00017978097518217702,
|
| 1363 |
+
"loss": 0.9783,
|
| 1364 |
"step": 18900
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 4.75,
|
| 1368 |
+
"grad_norm": 0.364629864692688,
|
| 1369 |
"learning_rate": 0.00017742883068638446,
|
| 1370 |
+
"loss": 0.963,
|
| 1371 |
"step": 19000
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 4.775,
|
| 1375 |
+
"grad_norm": 0.39363783597946167,
|
| 1376 |
"learning_rate": 0.00017508368075093582,
|
| 1377 |
+
"loss": 0.9175,
|
| 1378 |
"step": 19100
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 4.8,
|
| 1382 |
+
"grad_norm": 0.39291703701019287,
|
| 1383 |
"learning_rate": 0.00017274575140626317,
|
| 1384 |
+
"loss": 0.9616,
|
| 1385 |
"step": 19200
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 4.825,
|
| 1389 |
+
"grad_norm": 0.3810083866119385,
|
| 1390 |
"learning_rate": 0.0001704152679868626,
|
| 1391 |
+
"loss": 0.9754,
|
| 1392 |
"step": 19300
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 4.85,
|
| 1396 |
+
"grad_norm": 0.36352667212486267,
|
| 1397 |
"learning_rate": 0.00016809245510957666,
|
| 1398 |
+
"loss": 0.953,
|
| 1399 |
"step": 19400
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 4.875,
|
| 1403 |
+
"grad_norm": 0.31296002864837646,
|
| 1404 |
"learning_rate": 0.000165777536651945,
|
| 1405 |
+
"loss": 0.9478,
|
| 1406 |
"step": 19500
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 4.9,
|
| 1410 |
+
"grad_norm": 0.4004424214363098,
|
| 1411 |
"learning_rate": 0.0001634707357306267,
|
| 1412 |
+
"loss": 0.9508,
|
| 1413 |
"step": 19600
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 4.925,
|
| 1417 |
+
"grad_norm": 0.3690031170845032,
|
| 1418 |
"learning_rate": 0.00016117227467989602,
|
| 1419 |
+
"loss": 0.9456,
|
| 1420 |
"step": 19700
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 4.95,
|
| 1424 |
+
"grad_norm": 0.3946553170681,
|
| 1425 |
"learning_rate": 0.0001588823750302126,
|
| 1426 |
+
"loss": 0.95,
|
| 1427 |
"step": 19800
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 4.975,
|
| 1431 |
+
"grad_norm": 0.39817288517951965,
|
| 1432 |
"learning_rate": 0.00015660125748687094,
|
| 1433 |
+
"loss": 0.9431,
|
| 1434 |
"step": 19900
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 5.0,
|
| 1438 |
+
"grad_norm": 0.37246620655059814,
|
| 1439 |
"learning_rate": 0.00015432914190872756,
|
| 1440 |
+
"loss": 0.9353,
|
| 1441 |
"step": 20000
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 5.0,
|
| 1445 |
+
"eval_loss": 0.9754775762557983,
|
| 1446 |
+
"eval_runtime": 152.4185,
|
| 1447 |
+
"eval_samples_per_second": 26.244,
|
| 1448 |
+
"eval_steps_per_second": 6.561,
|
| 1449 |
"step": 20000
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 5.025,
|
| 1453 |
+
"grad_norm": 0.36520034074783325,
|
| 1454 |
"learning_rate": 0.000152066247287011,
|
| 1455 |
+
"loss": 0.9377,
|
| 1456 |
"step": 20100
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 5.05,
|
| 1460 |
+
"grad_norm": 0.3535405397415161,
|
| 1461 |
"learning_rate": 0.00014981279172421482,
|
| 1462 |
"loss": 0.9414,
|
| 1463 |
"step": 20200
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 5.075,
|
| 1467 |
+
"grad_norm": 0.35102036595344543,
|
| 1468 |
"learning_rate": 0.00014756899241307614,
|
| 1469 |
+
"loss": 0.9164,
|
| 1470 |
"step": 20300
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 5.1,
|
| 1474 |
+
"grad_norm": 0.3539367616176605,
|
| 1475 |
"learning_rate": 0.00014533506561564306,
|
| 1476 |
+
"loss": 0.9187,
|
| 1477 |
"step": 20400
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 5.125,
|
| 1481 |
+
"grad_norm": 0.347476601600647,
|
| 1482 |
"learning_rate": 0.00014311122664242953,
|
| 1483 |
+
"loss": 0.9196,
|
| 1484 |
"step": 20500
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 5.15,
|
| 1488 |
+
"grad_norm": 0.3902195990085602,
|
| 1489 |
"learning_rate": 0.00014089768983166444,
|
| 1490 |
+
"loss": 0.9301,
|
| 1491 |
"step": 20600
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 5.175,
|
| 1495 |
+
"grad_norm": 0.3888334631919861,
|
| 1496 |
"learning_rate": 0.000138694668528633,
|
| 1497 |
+
"loss": 0.9241,
|
| 1498 |
"step": 20700
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 5.2,
|
| 1502 |
+
"grad_norm": 0.41042113304138184,
|
| 1503 |
"learning_rate": 0.00013650237506511331,
|
| 1504 |
+
"loss": 0.9504,
|
| 1505 |
"step": 20800
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 5.225,
|
| 1509 |
+
"grad_norm": 0.3549586236476898,
|
| 1510 |
"learning_rate": 0.0001343210207389125,
|
| 1511 |
+
"loss": 0.955,
|
| 1512 |
"step": 20900
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 5.25,
|
| 1516 |
+
"grad_norm": 0.34020912647247314,
|
| 1517 |
"learning_rate": 0.00013215081579350058,
|
| 1518 |
+
"loss": 0.9471,
|
| 1519 |
"step": 21000
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 5.275,
|
| 1523 |
+
"grad_norm": 0.33093366026878357,
|
| 1524 |
"learning_rate": 0.00012999196939774722,
|
| 1525 |
+
"loss": 0.9351,
|
| 1526 |
"step": 21100
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 5.3,
|
| 1530 |
+
"grad_norm": 0.4015860855579376,
|
| 1531 |
"learning_rate": 0.00012784468962576134,
|
| 1532 |
+
"loss": 0.9491,
|
| 1533 |
"step": 21200
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 5.325,
|
| 1537 |
+
"grad_norm": 0.37603822350502014,
|
| 1538 |
"learning_rate": 0.00012570918343683636,
|
| 1539 |
+
"loss": 0.958,
|
| 1540 |
"step": 21300
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 5.35,
|
| 1544 |
+
"grad_norm": 0.39434927701950073,
|
| 1545 |
"learning_rate": 0.0001235856566555039,
|
| 1546 |
+
"loss": 0.94,
|
| 1547 |
"step": 21400
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 5.375,
|
| 1551 |
+
"grad_norm": 0.43162479996681213,
|
| 1552 |
"learning_rate": 0.0001214743139516946,
|
| 1553 |
+
"loss": 0.9291,
|
| 1554 |
"step": 21500
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 5.4,
|
| 1558 |
+
"grad_norm": 0.38297414779663086,
|
| 1559 |
"learning_rate": 0.00011937535882101281,
|
| 1560 |
+
"loss": 0.9494,
|
| 1561 |
"step": 21600
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 5.425,
|
| 1565 |
+
"grad_norm": 0.40226617455482483,
|
| 1566 |
"learning_rate": 0.00011728899356512265,
|
| 1567 |
+
"loss": 0.9067,
|
| 1568 |
"step": 21700
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 5.45,
|
| 1572 |
+
"grad_norm": 0.38879096508026123,
|
| 1573 |
"learning_rate": 0.00011521541927224994,
|
| 1574 |
+
"loss": 0.949,
|
| 1575 |
"step": 21800
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 5.475,
|
| 1579 |
+
"grad_norm": 0.33627960085868835,
|
| 1580 |
"learning_rate": 0.00011315483579780094,
|
| 1581 |
+
"loss": 0.9358,
|
| 1582 |
"step": 21900
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 5.5,
|
| 1586 |
+
"grad_norm": 0.4547683596611023,
|
| 1587 |
"learning_rate": 0.00011110744174509952,
|
| 1588 |
+
"loss": 0.9179,
|
| 1589 |
"step": 22000
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 5.525,
|
| 1593 |
+
"grad_norm": 0.3883083462715149,
|
| 1594 |
"learning_rate": 0.00010907343444624579,
|
| 1595 |
+
"loss": 0.9662,
|
| 1596 |
"step": 22100
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 5.55,
|
| 1600 |
+
"grad_norm": 0.4168870449066162,
|
| 1601 |
"learning_rate": 0.00010705300994309697,
|
| 1602 |
"loss": 0.9286,
|
| 1603 |
"step": 22200
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 5.575,
|
| 1607 |
+
"grad_norm": 0.3810732066631317,
|
| 1608 |
"learning_rate": 0.00010504636296837161,
|
| 1609 |
+
"loss": 0.9443,
|
| 1610 |
"step": 22300
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 5.6,
|
| 1614 |
+
"grad_norm": 0.40432223677635193,
|
| 1615 |
"learning_rate": 0.00010305368692688174,
|
| 1616 |
+
"loss": 0.9138,
|
| 1617 |
"step": 22400
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 5.625,
|
| 1621 |
+
"grad_norm": 0.3702322542667389,
|
| 1622 |
"learning_rate": 0.00010107517387689166,
|
| 1623 |
+
"loss": 0.9474,
|
| 1624 |
"step": 22500
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 5.65,
|
| 1628 |
+
"grad_norm": 0.40569931268692017,
|
| 1629 |
"learning_rate": 9.911101451160715e-05,
|
| 1630 |
+
"loss": 0.9645,
|
| 1631 |
"step": 22600
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 5.675,
|
| 1635 |
+
"grad_norm": 0.39476191997528076,
|
| 1636 |
"learning_rate": 9.716139814079594e-05,
|
| 1637 |
+
"loss": 0.9478,
|
| 1638 |
"step": 22700
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 5.7,
|
| 1642 |
+
"grad_norm": 0.347662091255188,
|
| 1643 |
"learning_rate": 9.522651267254148e-05,
|
| 1644 |
+
"loss": 0.9627,
|
| 1645 |
"step": 22800
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
"epoch": 5.725,
|
| 1649 |
+
"grad_norm": 0.36813852190971375,
|
| 1650 |
"learning_rate": 9.330654459513265e-05,
|
| 1651 |
+
"loss": 0.9328,
|
| 1652 |
"step": 22900
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
"epoch": 5.75,
|
| 1656 |
+
"grad_norm": 0.32601600885391235,
|
| 1657 |
"learning_rate": 9.140167895908866e-05,
|
| 1658 |
+
"loss": 0.9547,
|
| 1659 |
"step": 23000
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
"epoch": 5.775,
|
| 1663 |
+
"grad_norm": 0.40840694308280945,
|
| 1664 |
"learning_rate": 8.951209935932425e-05,
|
| 1665 |
"loss": 0.9113,
|
| 1666 |
"step": 23100
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
"epoch": 5.8,
|
| 1670 |
+
"grad_norm": 0.38472625613212585,
|
| 1671 |
"learning_rate": 8.763798791745412e-05,
|
| 1672 |
+
"loss": 0.9667,
|
| 1673 |
"step": 23200
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
"epoch": 5.825,
|
| 1677 |
+
"grad_norm": 0.37892642617225647,
|
| 1678 |
"learning_rate": 8.577952526423969e-05,
|
| 1679 |
+
"loss": 0.9241,
|
| 1680 |
"step": 23300
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
"epoch": 5.85,
|
| 1684 |
+
"grad_norm": 0.3874419927597046,
|
| 1685 |
"learning_rate": 8.393689052217964e-05,
|
| 1686 |
+
"loss": 0.9539,
|
| 1687 |
"step": 23400
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
"epoch": 5.875,
|
| 1691 |
+
"grad_norm": 0.37583285570144653,
|
| 1692 |
"learning_rate": 8.211026128824539e-05,
|
| 1693 |
+
"loss": 0.9411,
|
| 1694 |
"step": 23500
|
| 1695 |
},
|
| 1696 |
{
|
| 1697 |
"epoch": 5.9,
|
| 1698 |
+
"grad_norm": 0.37817347049713135,
|
| 1699 |
"learning_rate": 8.029981361676455e-05,
|
| 1700 |
+
"loss": 0.9426,
|
| 1701 |
"step": 23600
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 5.925,
|
| 1705 |
+
"grad_norm": 0.32163432240486145,
|
| 1706 |
"learning_rate": 7.850572200245185e-05,
|
| 1707 |
+
"loss": 0.9512,
|
| 1708 |
"step": 23700
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
"epoch": 5.95,
|
| 1712 |
+
"grad_norm": 0.37549078464508057,
|
| 1713 |
"learning_rate": 7.672815936359106e-05,
|
| 1714 |
+
"loss": 0.9694,
|
| 1715 |
"step": 23800
|
| 1716 |
},
|
| 1717 |
{
|
| 1718 |
"epoch": 5.975,
|
| 1719 |
+
"grad_norm": 0.43039214611053467,
|
| 1720 |
"learning_rate": 7.496729702536912e-05,
|
| 1721 |
+
"loss": 0.9191,
|
| 1722 |
"step": 23900
|
| 1723 |
},
|
| 1724 |
{
|
| 1725 |
"epoch": 6.0,
|
| 1726 |
+
"grad_norm": 0.3925749361515045,
|
| 1727 |
"learning_rate": 7.322330470336314e-05,
|
| 1728 |
+
"loss": 0.9121,
|
| 1729 |
"step": 24000
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
"epoch": 6.0,
|
| 1733 |
+
"eval_loss": 0.9720398187637329,
|
| 1734 |
+
"eval_runtime": 152.9557,
|
| 1735 |
+
"eval_samples_per_second": 26.151,
|
| 1736 |
+
"eval_steps_per_second": 6.538,
|
| 1737 |
"step": 24000
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 6.025,
|
| 1741 |
+
"grad_norm": 0.41935107111930847,
|
| 1742 |
"learning_rate": 7.149635048718294e-05,
|
| 1743 |
+
"loss": 0.9072,
|
| 1744 |
"step": 24100
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 6.05,
|
| 1748 |
+
"grad_norm": 0.4102177321910858,
|
| 1749 |
"learning_rate": 6.97866008242703e-05,
|
| 1750 |
"loss": 0.9105,
|
| 1751 |
"step": 24200
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 6.075,
|
| 1755 |
+
"grad_norm": 0.3717847168445587,
|
| 1756 |
"learning_rate": 6.809422050385628e-05,
|
| 1757 |
+
"loss": 0.9399,
|
| 1758 |
"step": 24300
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 6.1,
|
| 1762 |
+
"grad_norm": 0.3842828869819641,
|
| 1763 |
"learning_rate": 6.641937264107867e-05,
|
| 1764 |
+
"loss": 0.8992,
|
| 1765 |
"step": 24400
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 6.125,
|
| 1769 |
+
"grad_norm": 0.4051499366760254,
|
| 1770 |
"learning_rate": 6.476221866126028e-05,
|
| 1771 |
+
"loss": 0.9274,
|
| 1772 |
"step": 24500
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 6.15,
|
| 1776 |
+
"grad_norm": 0.4197865128517151,
|
| 1777 |
"learning_rate": 6.312291828435076e-05,
|
| 1778 |
+
"loss": 0.9646,
|
| 1779 |
"step": 24600
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 6.175,
|
| 1783 |
+
"grad_norm": 0.41968774795532227,
|
| 1784 |
"learning_rate": 6.150162950953264e-05,
|
| 1785 |
+
"loss": 0.9412,
|
| 1786 |
"step": 24700
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 6.2,
|
| 1790 |
+
"grad_norm": 0.40155336260795593,
|
| 1791 |
"learning_rate": 5.989850859999227e-05,
|
| 1792 |
+
"loss": 0.9627,
|
| 1793 |
"step": 24800
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 6.225,
|
| 1797 |
+
"grad_norm": 0.39962807297706604,
|
| 1798 |
"learning_rate": 5.831371006785963e-05,
|
| 1799 |
+
"loss": 0.9337,
|
| 1800 |
"step": 24900
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 6.25,
|
| 1804 |
+
"grad_norm": 0.4056980609893799,
|
| 1805 |
"learning_rate": 5.6747386659315755e-05,
|
| 1806 |
+
"loss": 0.9221,
|
| 1807 |
"step": 25000
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 6.275,
|
| 1811 |
+
"grad_norm": 0.4260208308696747,
|
| 1812 |
"learning_rate": 5.519968933987082e-05,
|
| 1813 |
+
"loss": 0.9359,
|
| 1814 |
"step": 25100
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 6.3,
|
| 1818 |
+
"grad_norm": 0.39483314752578735,
|
| 1819 |
"learning_rate": 5.367076727981382e-05,
|
| 1820 |
+
"loss": 0.9187,
|
| 1821 |
"step": 25200
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 6.325,
|
| 1825 |
+
"grad_norm": 0.43257468938827515,
|
| 1826 |
"learning_rate": 5.216076783983492e-05,
|
| 1827 |
+
"loss": 0.9701,
|
| 1828 |
"step": 25300
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 6.35,
|
| 1832 |
+
"grad_norm": 0.39114078879356384,
|
| 1833 |
"learning_rate": 5.066983655682325e-05,
|
| 1834 |
+
"loss": 0.9664,
|
| 1835 |
"step": 25400
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 6.375,
|
| 1839 |
+
"grad_norm": 0.39986172318458557,
|
| 1840 |
"learning_rate": 4.919811712983879e-05,
|
| 1841 |
+
"loss": 0.9345,
|
| 1842 |
"step": 25500
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 6.4,
|
| 1846 |
+
"grad_norm": 0.41802579164505005,
|
| 1847 |
"learning_rate": 4.7745751406263163e-05,
|
| 1848 |
+
"loss": 0.9192,
|
| 1849 |
"step": 25600
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 6.425,
|
| 1853 |
+
"grad_norm": 0.34528008103370667,
|
| 1854 |
"learning_rate": 4.6312879368127645e-05,
|
| 1855 |
+
"loss": 0.953,
|
| 1856 |
"step": 25700
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 6.45,
|
| 1860 |
+
"grad_norm": 0.418544739484787,
|
| 1861 |
"learning_rate": 4.4899639118621604e-05,
|
| 1862 |
+
"loss": 0.9121,
|
| 1863 |
"step": 25800
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 6.475,
|
| 1867 |
+
"grad_norm": 0.40490132570266724,
|
| 1868 |
"learning_rate": 4.350616686878175e-05,
|
| 1869 |
+
"loss": 0.9199,
|
| 1870 |
"step": 25900
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 6.5,
|
| 1874 |
+
"grad_norm": 0.4175280034542084,
|
| 1875 |
"learning_rate": 4.213259692436367e-05,
|
| 1876 |
+
"loss": 0.9304,
|
| 1877 |
"step": 26000
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 6.525,
|
| 1881 |
+
"grad_norm": 0.3687520921230316,
|
| 1882 |
"learning_rate": 4.077906167289766e-05,
|
| 1883 |
+
"loss": 0.953,
|
| 1884 |
"step": 26100
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 6.55,
|
| 1888 |
+
"grad_norm": 0.389283686876297,
|
| 1889 |
"learning_rate": 3.944569157092839e-05,
|
| 1890 |
+
"loss": 0.9278,
|
| 1891 |
"step": 26200
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 6.575,
|
| 1895 |
+
"grad_norm": 0.353508323431015,
|
| 1896 |
"learning_rate": 3.8132615131441396e-05,
|
| 1897 |
+
"loss": 0.9488,
|
| 1898 |
"step": 26300
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 6.6,
|
| 1902 |
+
"grad_norm": 0.3994731605052948,
|
| 1903 |
"learning_rate": 3.6839958911476953e-05,
|
| 1904 |
+
"loss": 0.8955,
|
| 1905 |
"step": 26400
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 6.625,
|
| 1909 |
+
"grad_norm": 0.37398815155029297,
|
| 1910 |
"learning_rate": 3.5567847499932e-05,
|
| 1911 |
"loss": 0.9453,
|
| 1912 |
"step": 26500
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 6.65,
|
| 1916 |
+
"grad_norm": 0.37087592482566833,
|
| 1917 |
"learning_rate": 3.431640350555204e-05,
|
| 1918 |
+
"loss": 0.9449,
|
| 1919 |
"step": 26600
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 6.675,
|
| 1923 |
+
"grad_norm": 0.3406459093093872,
|
| 1924 |
"learning_rate": 3.308574754511404e-05,
|
| 1925 |
+
"loss": 0.9143,
|
| 1926 |
"step": 26700
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 6.7,
|
| 1930 |
+
"grad_norm": 0.36757832765579224,
|
| 1931 |
"learning_rate": 3.187599823180071e-05,
|
| 1932 |
+
"loss": 0.935,
|
| 1933 |
"step": 26800
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 6.725,
|
| 1937 |
+
"grad_norm": 0.40823671221733093,
|
| 1938 |
"learning_rate": 3.0687272163768986e-05,
|
| 1939 |
+
"loss": 0.9181,
|
| 1940 |
"step": 26900
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 6.75,
|
| 1944 |
+
"grad_norm": 0.3942839801311493,
|
| 1945 |
"learning_rate": 2.9519683912911265e-05,
|
| 1946 |
+
"loss": 0.9219,
|
| 1947 |
"step": 27000
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 6.775,
|
| 1951 |
+
"grad_norm": 0.3952127695083618,
|
| 1952 |
"learning_rate": 2.8373346013813417e-05,
|
| 1953 |
+
"loss": 0.9001,
|
| 1954 |
"step": 27100
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 6.8,
|
| 1958 |
+
"grad_norm": 0.4122380018234253,
|
| 1959 |
"learning_rate": 2.7248368952908055e-05,
|
| 1960 |
+
"loss": 0.9266,
|
| 1961 |
"step": 27200
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 6.825,
|
| 1965 |
+
"grad_norm": 0.4118337631225586,
|
| 1966 |
"learning_rate": 2.6144861157825773e-05,
|
| 1967 |
+
"loss": 0.9364,
|
| 1968 |
"step": 27300
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 6.85,
|
| 1972 |
+
"grad_norm": 0.39552587270736694,
|
| 1973 |
"learning_rate": 2.5062928986944677e-05,
|
| 1974 |
+
"loss": 0.9226,
|
| 1975 |
"step": 27400
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 6.875,
|
| 1979 |
+
"grad_norm": 0.4226832687854767,
|
| 1980 |
"learning_rate": 2.4002676719139166e-05,
|
| 1981 |
"loss": 0.9394,
|
| 1982 |
"step": 27500
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 6.9,
|
| 1986 |
+
"grad_norm": 0.34474310278892517,
|
| 1987 |
"learning_rate": 2.296420654372966e-05,
|
| 1988 |
+
"loss": 0.9255,
|
| 1989 |
"step": 27600
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 6.925,
|
| 1993 |
+
"grad_norm": 0.4371466636657715,
|
| 1994 |
"learning_rate": 2.1947618550633096e-05,
|
| 1995 |
+
"loss": 0.9298,
|
| 1996 |
"step": 27700
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 6.95,
|
| 2000 |
+
"grad_norm": 0.4318234622478485,
|
| 2001 |
"learning_rate": 2.0953010720716037e-05,
|
| 2002 |
+
"loss": 0.9533,
|
| 2003 |
"step": 27800
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 6.975,
|
| 2007 |
+
"grad_norm": 0.38867416977882385,
|
| 2008 |
"learning_rate": 1.9980478916351297e-05,
|
| 2009 |
+
"loss": 0.9612,
|
| 2010 |
"step": 27900
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 7.0,
|
| 2014 |
+
"grad_norm": 0.38006094098091125,
|
| 2015 |
"learning_rate": 1.9030116872178316e-05,
|
| 2016 |
+
"loss": 0.9175,
|
| 2017 |
"step": 28000
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 7.0,
|
| 2021 |
+
"eval_loss": 0.9707372784614563,
|
| 2022 |
+
"eval_runtime": 152.8648,
|
| 2023 |
+
"eval_samples_per_second": 26.167,
|
| 2024 |
+
"eval_steps_per_second": 6.542,
|
| 2025 |
"step": 28000
|
| 2026 |
},
|
| 2027 |
{
|
| 2028 |
"epoch": 7.025,
|
| 2029 |
+
"grad_norm": 0.4155292510986328,
|
| 2030 |
"learning_rate": 1.8102016186068992e-05,
|
| 2031 |
+
"loss": 0.936,
|
| 2032 |
"step": 28100
|
| 2033 |
},
|
| 2034 |
{
|
| 2035 |
"epoch": 7.05,
|
| 2036 |
+
"grad_norm": 0.3884825110435486,
|
| 2037 |
"learning_rate": 1.719626631029911e-05,
|
| 2038 |
+
"loss": 0.9398,
|
| 2039 |
"step": 28200
|
| 2040 |
},
|
| 2041 |
{
|
| 2042 |
"epoch": 7.075,
|
| 2043 |
+
"grad_norm": 0.39031410217285156,
|
| 2044 |
"learning_rate": 1.6312954542926888e-05,
|
| 2045 |
+
"loss": 0.9056,
|
| 2046 |
"step": 28300
|
| 2047 |
},
|
| 2048 |
{
|
| 2049 |
"epoch": 7.1,
|
| 2050 |
+
"grad_norm": 0.4356350004673004,
|
| 2051 |
"learning_rate": 1.5452166019378987e-05,
|
| 2052 |
+
"loss": 0.9178,
|
| 2053 |
"step": 28400
|
| 2054 |
},
|
| 2055 |
{
|
| 2056 |
"epoch": 7.125,
|
| 2057 |
+
"grad_norm": 0.3904324769973755,
|
| 2058 |
"learning_rate": 1.4613983704244827e-05,
|
| 2059 |
+
"loss": 0.9067,
|
| 2060 |
"step": 28500
|
| 2061 |
},
|
| 2062 |
{
|
| 2063 |
"epoch": 7.15,
|
| 2064 |
+
"grad_norm": 0.3213234543800354,
|
| 2065 |
"learning_rate": 1.3798488383280488e-05,
|
| 2066 |
+
"loss": 0.9436,
|
| 2067 |
"step": 28600
|
| 2068 |
},
|
| 2069 |
{
|
| 2070 |
"epoch": 7.175,
|
| 2071 |
+
"grad_norm": 0.38800540566444397,
|
| 2072 |
"learning_rate": 1.3005758655622424e-05,
|
| 2073 |
+
"loss": 0.9131,
|
| 2074 |
"step": 28700
|
| 2075 |
},
|
| 2076 |
{
|
| 2077 |
"epoch": 7.2,
|
| 2078 |
+
"grad_norm": 0.3734656870365143,
|
| 2079 |
"learning_rate": 1.2235870926211617e-05,
|
| 2080 |
+
"loss": 0.9466,
|
| 2081 |
"step": 28800
|
| 2082 |
},
|
| 2083 |
{
|
| 2084 |
"epoch": 7.225,
|
| 2085 |
+
"grad_norm": 0.4088476002216339,
|
| 2086 |
"learning_rate": 1.1488899398429897e-05,
|
| 2087 |
+
"loss": 0.9444,
|
| 2088 |
"step": 28900
|
| 2089 |
},
|
| 2090 |
{
|
| 2091 |
"epoch": 7.25,
|
| 2092 |
+
"grad_norm": 0.3788207471370697,
|
| 2093 |
"learning_rate": 1.0764916066947795e-05,
|
| 2094 |
+
"loss": 0.9015,
|
| 2095 |
"step": 29000
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
"epoch": 7.275,
|
| 2099 |
+
"grad_norm": 0.3710539638996124,
|
| 2100 |
"learning_rate": 1.0063990710785648e-05,
|
| 2101 |
+
"loss": 0.9356,
|
| 2102 |
"step": 29100
|
| 2103 |
},
|
| 2104 |
{
|
| 2105 |
"epoch": 7.3,
|
| 2106 |
+
"grad_norm": 0.39175015687942505,
|
| 2107 |
"learning_rate": 9.386190886588208e-06,
|
| 2108 |
+
"loss": 0.9277,
|
| 2109 |
"step": 29200
|
| 2110 |
},
|
| 2111 |
{
|
| 2112 |
"epoch": 7.325,
|
| 2113 |
+
"grad_norm": 0.39857038855552673,
|
| 2114 |
"learning_rate": 8.731581922113152e-06,
|
| 2115 |
+
"loss": 0.9597,
|
| 2116 |
"step": 29300
|
| 2117 |
},
|
| 2118 |
{
|
| 2119 |
"epoch": 7.35,
|
| 2120 |
+
"grad_norm": 0.42670151591300964,
|
| 2121 |
"learning_rate": 8.10022690993506e-06,
|
| 2122 |
+
"loss": 0.9124,
|
| 2123 |
"step": 29400
|
| 2124 |
},
|
| 2125 |
{
|
| 2126 |
"epoch": 7.375,
|
| 2127 |
+
"grad_norm": 0.37594255805015564,
|
| 2128 |
"learning_rate": 7.4921867013640064e-06,
|
| 2129 |
+
"loss": 0.9202,
|
| 2130 |
"step": 29500
|
| 2131 |
},
|
| 2132 |
{
|
| 2133 |
"epoch": 7.4,
|
| 2134 |
+
"grad_norm": 0.34726592898368835,
|
| 2135 |
"learning_rate": 6.907519900580861e-06,
|
| 2136 |
+
"loss": 0.9462,
|
| 2137 |
"step": 29600
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
"epoch": 7.425,
|
| 2141 |
+
"grad_norm": 0.3583536744117737,
|
| 2142 |
"learning_rate": 6.34628285898875e-06,
|
| 2143 |
+
"loss": 0.9356,
|
| 2144 |
"step": 29700
|
| 2145 |
},
|
| 2146 |
{
|
| 2147 |
"epoch": 7.45,
|
| 2148 |
+
"grad_norm": 0.40716230869293213,
|
| 2149 |
"learning_rate": 5.808529669781903e-06,
|
| 2150 |
+
"loss": 0.9204,
|
| 2151 |
"step": 29800
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
"epoch": 7.475,
|
| 2155 |
+
"grad_norm": 0.4574654698371887,
|
| 2156 |
"learning_rate": 5.294312162731935e-06,
|
| 2157 |
+
"loss": 0.9272,
|
| 2158 |
"step": 29900
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
"epoch": 7.5,
|
| 2162 |
+
"grad_norm": 0.41072556376457214,
|
| 2163 |
"learning_rate": 4.803679899192393e-06,
|
| 2164 |
+
"loss": 0.9181,
|
| 2165 |
"step": 30000
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
"epoch": 7.525,
|
| 2169 |
+
"grad_norm": 0.4989704191684723,
|
| 2170 |
"learning_rate": 4.336680167322055e-06,
|
| 2171 |
+
"loss": 0.9346,
|
| 2172 |
"step": 30100
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
"epoch": 7.55,
|
| 2176 |
+
"grad_norm": 0.4089345932006836,
|
| 2177 |
"learning_rate": 3.893357977527101e-06,
|
| 2178 |
+
"loss": 0.9392,
|
| 2179 |
"step": 30200
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
"epoch": 7.575,
|
| 2183 |
+
"grad_norm": 0.4189968407154083,
|
| 2184 |
"learning_rate": 3.4737560581228343e-06,
|
| 2185 |
+
"loss": 0.9338,
|
| 2186 |
"step": 30300
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
"epoch": 7.6,
|
| 2190 |
+
"grad_norm": 0.40372830629348755,
|
| 2191 |
"learning_rate": 3.077914851215585e-06,
|
| 2192 |
+
"loss": 0.9394,
|
| 2193 |
"step": 30400
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
"epoch": 7.625,
|
| 2197 |
+
"grad_norm": 0.3864797055721283,
|
| 2198 |
"learning_rate": 2.7058725088047465e-06,
|
| 2199 |
+
"loss": 0.9283,
|
| 2200 |
"step": 30500
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
"epoch": 7.65,
|
| 2204 |
+
"grad_norm": 0.4125591814517975,
|
| 2205 |
"learning_rate": 2.357664889105687e-06,
|
| 2206 |
+
"loss": 0.9731,
|
| 2207 |
"step": 30600
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
"epoch": 7.675,
|
| 2211 |
+
"grad_norm": 0.4122030436992645,
|
| 2212 |
"learning_rate": 2.0333255530934903e-06,
|
| 2213 |
+
"loss": 0.9257,
|
| 2214 |
"step": 30700
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 7.7,
|
| 2218 |
+
"grad_norm": 0.40973401069641113,
|
| 2219 |
"learning_rate": 1.7328857612684267e-06,
|
| 2220 |
+
"loss": 0.8964,
|
| 2221 |
"step": 30800
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
"epoch": 7.725,
|
| 2225 |
+
"grad_norm": 0.36363255977630615,
|
| 2226 |
"learning_rate": 1.4563744706429517e-06,
|
| 2227 |
+
"loss": 0.9012,
|
| 2228 |
"step": 30900
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
"epoch": 7.75,
|
| 2232 |
+
"grad_norm": 0.39668798446655273,
|
| 2233 |
"learning_rate": 1.2038183319507957e-06,
|
| 2234 |
+
"loss": 0.9399,
|
| 2235 |
"step": 31000
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
"epoch": 7.775,
|
| 2239 |
+
"grad_norm": 0.40102848410606384,
|
| 2240 |
"learning_rate": 9.752416870782156e-07,
|
| 2241 |
+
"loss": 0.9471,
|
| 2242 |
"step": 31100
|
| 2243 |
},
|
| 2244 |
{
|
| 2245 |
"epoch": 7.8,
|
| 2246 |
+
"grad_norm": 0.43242618441581726,
|
| 2247 |
"learning_rate": 7.70666566718009e-07,
|
| 2248 |
+
"loss": 0.9289,
|
| 2249 |
"step": 31200
|
| 2250 |
},
|
| 2251 |
{
|
| 2252 |
"epoch": 7.825,
|
| 2253 |
+
"grad_norm": 0.41685134172439575,
|
| 2254 |
"learning_rate": 5.90112688246075e-07,
|
| 2255 |
+
"loss": 0.9365,
|
| 2256 |
"step": 31300
|
| 2257 |
},
|
| 2258 |
{
|
| 2259 |
"epoch": 7.85,
|
| 2260 |
+
"grad_norm": 0.36857473850250244,
|
| 2261 |
"learning_rate": 4.335974538210441e-07,
|
| 2262 |
+
"loss": 0.9324,
|
| 2263 |
"step": 31400
|
| 2264 |
},
|
| 2265 |
{
|
| 2266 |
"epoch": 7.875,
|
| 2267 |
+
"grad_norm": 0.42975786328315735,
|
| 2268 |
"learning_rate": 3.0113594870689873e-07,
|
| 2269 |
+
"loss": 0.9379,
|
| 2270 |
"step": 31500
|
| 2271 |
}
|
| 2272 |
],
|
checkpoint-32000/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3702168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41a63c7fc27bf93b10bab03bee6d69c7097075cc4f0bf9af5d2b6839ce7b0137
|
| 3 |
size 3702168
|
checkpoint-32000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7473658
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44b2992101a08833f08bb0c94ab1bef70d5086e1fe2ef8c69f7f13130f92f460
|
| 3 |
size 7473658
|
checkpoint-32000/trainer_state.json
CHANGED
|
@@ -10,2298 +10,2298 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.025,
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
"learning_rate": 0.000499987952239832,
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.05,
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
"learning_rate": 0.0004999518101205162,
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.075,
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
"learning_rate": 0.0004998915771255053,
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.1,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
"learning_rate": 0.0004998072590601808,
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.125,
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
"learning_rate": 0.0004996988640512931,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.15,
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
"learning_rate": 0.000499566402546179,
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.175,
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
"learning_rate": 0.0004994098873117539,
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.2,
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
"learning_rate": 0.000499229333433282,
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.225,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
"learning_rate": 0.0004990247583129218,
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.25,
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
"learning_rate": 0.0004987961816680492,
|
| 78 |
"loss": 1.0698,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.275,
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
"learning_rate": 0.0004985436255293571,
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.3,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
"learning_rate": 0.0004982671142387316,
|
| 92 |
"loss": 1.0906,
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.325,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
"learning_rate": 0.0004979666744469065,
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.35,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
"learning_rate": 0.0004976423351108943,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.375,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
"learning_rate": 0.0004972941274911952,
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.4,
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
"learning_rate": 0.0004969220851487844,
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.425,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
"learning_rate": 0.0004965262439418772,
|
| 127 |
"loss": 1.082,
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.45,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
"learning_rate": 0.0004961066420224729,
|
| 134 |
"loss": 1.0937,
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.475,
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
"learning_rate": 0.000495663319832678,
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.5,
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
"learning_rate": 0.0004951963201008077,
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.525,
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
"learning_rate": 0.0004947056878372681,
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.55,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
"learning_rate": 0.0004941914703302181,
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.575,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
"learning_rate": 0.0004936537171410112,
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.6,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
"learning_rate": 0.0004930924800994192,
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.625,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
"learning_rate": 0.000492507813298636,
|
| 183 |
"loss": 1.0858,
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.65,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
"learning_rate": 0.0004918997730900649,
|
| 190 |
-
"loss": 1.
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.675,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
"learning_rate": 0.0004912684180778869,
|
| 197 |
"loss": 1.0647,
|
| 198 |
"step": 2700
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.7,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
"learning_rate": 0.0004906138091134118,
|
| 204 |
"loss": 1.0484,
|
| 205 |
"step": 2800
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.725,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
"learning_rate": 0.0004899360092892143,
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 2900
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.75,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
"learning_rate": 0.0004892350839330522,
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.775,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
"learning_rate": 0.0004885111006015701,
|
| 225 |
-
"loss": 1.
|
| 226 |
"step": 3100
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.8,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
"learning_rate": 0.0004877641290737884,
|
| 232 |
-
"loss": 1.
|
| 233 |
"step": 3200
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.825,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
"learning_rate": 0.0004869942413443776,
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 3300
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.85,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
"learning_rate": 0.00048620151161671955,
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 3400
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.875,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
"learning_rate": 0.0004853860162957552,
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.9,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 0.0004845478339806211,
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 3600
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.925,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
"learning_rate": 0.0004836870454570731,
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 3700
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.95,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
"learning_rate": 0.00048280373368970086,
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 3800
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.975,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
"learning_rate": 0.000481897983813931,
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 3900
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 1.0,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
"learning_rate": 0.0004809698831278217,
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
-
"eval_loss": 1.
|
| 294 |
-
"eval_runtime":
|
| 295 |
-
"eval_samples_per_second":
|
| 296 |
-
"eval_steps_per_second":
|
| 297 |
"step": 4000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.025,
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
"learning_rate": 0.00048001952108364876,
|
| 303 |
-
"loss": 1.
|
| 304 |
"step": 4100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 1.05,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
"learning_rate": 0.00047904698927928404,
|
| 310 |
-
"loss": 1.
|
| 311 |
"step": 4200
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 1.075,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
"learning_rate": 0.0004780523814493669,
|
| 317 |
-
"loss": 1.
|
| 318 |
"step": 4300
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 1.1,
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
"learning_rate": 0.00047703579345627036,
|
| 324 |
-
"loss": 1.
|
| 325 |
"step": 4400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.125,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
"learning_rate": 0.0004759973232808609,
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 4500
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.15,
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
"learning_rate": 0.0004749370710130554,
|
| 338 |
-
"loss": 1.
|
| 339 |
"step": 4600
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.175,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
"learning_rate": 0.0004738551388421742,
|
| 345 |
"loss": 1.0291,
|
| 346 |
"step": 4700
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.2,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
"learning_rate": 0.00047275163104709196,
|
| 352 |
-
"loss": 1.
|
| 353 |
"step": 4800
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.225,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
"learning_rate": 0.00047162665398618666,
|
| 359 |
-
"loss": 1.
|
| 360 |
"step": 4900
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.25,
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
"learning_rate": 0.00047048031608708875,
|
| 366 |
-
"loss": 1.
|
| 367 |
"step": 5000
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.275,
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
"learning_rate": 0.00046931272783623106,
|
| 373 |
-
"loss": 1.
|
| 374 |
"step": 5100
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.3,
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
"learning_rate": 0.0004681240017681993,
|
| 380 |
-
"loss": 1.
|
| 381 |
"step": 5200
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.325,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
"learning_rate": 0.00046691425245488607,
|
| 387 |
-
"loss": 1.
|
| 388 |
"step": 5300
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 1.35,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
"learning_rate": 0.00046568359649444796,
|
| 394 |
-
"loss": 1.
|
| 395 |
"step": 5400
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.375,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
"learning_rate": 0.00046443215250006805,
|
| 401 |
-
"loss": 1.
|
| 402 |
"step": 5500
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.4,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
"learning_rate": 0.00046316004108852305,
|
| 408 |
-
"loss": 1.
|
| 409 |
"step": 5600
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 1.425,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
"learning_rate": 0.0004618673848685586,
|
| 415 |
-
"loss": 1.
|
| 416 |
"step": 5700
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 1.45,
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
"learning_rate": 0.0004605543084290716,
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 5800
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 1.475,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
"learning_rate": 0.0004592209383271023,
|
| 429 |
-
"loss": 1.
|
| 430 |
"step": 5900
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 1.5,
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
"learning_rate": 0.00045786740307563633,
|
| 436 |
-
"loss": 1.
|
| 437 |
"step": 6000
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 1.525,
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
"learning_rate": 0.0004564938331312183,
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 6100
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 1.55,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
"learning_rate": 0.0004551003608813784,
|
| 450 |
-
"loss": 1.
|
| 451 |
"step": 6200
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 1.575,
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
"learning_rate": 0.00045368712063187237,
|
| 457 |
-
"loss": 1.
|
| 458 |
"step": 6300
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 1.6,
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
"learning_rate": 0.0004522542485937369,
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 6400
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 1.625,
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
"learning_rate": 0.0004508018828701612,
|
| 471 |
-
"loss": 1.
|
| 472 |
"step": 6500
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 1.65,
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
"learning_rate": 0.0004493301634431768,
|
| 478 |
-
"loss": 0.
|
| 479 |
"step": 6600
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 1.675,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
"learning_rate": 0.00044783923216016507,
|
| 485 |
-
"loss": 1.
|
| 486 |
"step": 6700
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 1.7,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
"learning_rate": 0.0004463292327201862,
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 6800
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 1.725,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
"learning_rate": 0.00044480031066012916,
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 6900
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 1.75,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
"learning_rate": 0.0004432526133406842,
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 7000
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 1.775,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
"learning_rate": 0.00044168628993214036,
|
| 513 |
-
"loss": 0.
|
| 514 |
"step": 7100
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 1.8,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
"learning_rate": 0.0004401014914000078,
|
| 520 |
-
"loss": 0.
|
| 521 |
"step": 7200
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 1.825,
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
"learning_rate": 0.00043849837049046735,
|
| 527 |
"loss": 1.007,
|
| 528 |
"step": 7300
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 1.85,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
"learning_rate": 0.00043687708171564923,
|
| 534 |
-
"loss": 1.
|
| 535 |
"step": 7400
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 1.875,
|
| 539 |
-
"grad_norm": 0.
|
| 540 |
"learning_rate": 0.0004352377813387398,
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 7500
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 1.9,
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
"learning_rate": 0.0004335806273589214,
|
| 548 |
-
"loss": 1.
|
| 549 |
"step": 7600
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 1.925,
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
"learning_rate": 0.00043190577949614375,
|
| 555 |
-
"loss": 1.
|
| 556 |
"step": 7700
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.95,
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
"learning_rate": 0.0004302133991757297,
|
| 562 |
-
"loss": 0.
|
| 563 |
"step": 7800
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 1.975,
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
"learning_rate": 0.00042850364951281707,
|
| 569 |
-
"loss": 1.
|
| 570 |
"step": 7900
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 2.0,
|
| 574 |
-
"grad_norm": 0.
|
| 575 |
"learning_rate": 0.00042677669529663686,
|
| 576 |
-
"loss": 1.
|
| 577 |
"step": 8000
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 2.0,
|
| 581 |
-
"eval_loss": 1.
|
| 582 |
-
"eval_runtime":
|
| 583 |
-
"eval_samples_per_second":
|
| 584 |
-
"eval_steps_per_second":
|
| 585 |
"step": 8000
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 2.025,
|
| 589 |
-
"grad_norm": 0.
|
| 590 |
"learning_rate": 0.0004250327029746309,
|
| 591 |
-
"loss": 1.
|
| 592 |
"step": 8100
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 2.05,
|
| 596 |
-
"grad_norm": 0.
|
| 597 |
"learning_rate": 0.000423271840636409,
|
| 598 |
-
"loss": 0.
|
| 599 |
"step": 8200
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 2.075,
|
| 603 |
-
"grad_norm": 0.
|
| 604 |
"learning_rate": 0.00042149427799754817,
|
| 605 |
-
"loss": 0.
|
| 606 |
"step": 8300
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 2.1,
|
| 610 |
-
"grad_norm": 0.
|
| 611 |
"learning_rate": 0.00041970018638323546,
|
| 612 |
-
"loss": 0.
|
| 613 |
"step": 8400
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 2.125,
|
| 617 |
-
"grad_norm": 0.
|
| 618 |
"learning_rate": 0.00041788973871175465,
|
| 619 |
-
"loss": 0.
|
| 620 |
"step": 8500
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 2.15,
|
| 624 |
-
"grad_norm": 0.
|
| 625 |
"learning_rate": 0.00041606310947782046,
|
| 626 |
-
"loss": 0.
|
| 627 |
"step": 8600
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 2.175,
|
| 631 |
-
"grad_norm": 0.
|
| 632 |
"learning_rate": 0.00041422047473576033,
|
| 633 |
-
"loss": 1.
|
| 634 |
"step": 8700
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 2.2,
|
| 638 |
-
"grad_norm": 0.
|
| 639 |
"learning_rate": 0.0004123620120825459,
|
| 640 |
-
"loss": 1.
|
| 641 |
"step": 8800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 2.225,
|
| 645 |
-
"grad_norm": 0.
|
| 646 |
"learning_rate": 0.00041048790064067577,
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 8900
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 2.25,
|
| 652 |
-
"grad_norm": 0.
|
| 653 |
"learning_rate": 0.0004085983210409114,
|
| 654 |
-
"loss": 0.
|
| 655 |
"step": 9000
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 2.275,
|
| 659 |
-
"grad_norm": 0.
|
| 660 |
"learning_rate": 0.0004066934554048674,
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 9100
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 2.3,
|
| 666 |
-
"grad_norm": 0.
|
| 667 |
"learning_rate": 0.00040477348732745853,
|
| 668 |
"loss": 0.9858,
|
| 669 |
"step": 9200
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 2.325,
|
| 673 |
-
"grad_norm": 0.
|
| 674 |
"learning_rate": 0.0004028386018592041,
|
| 675 |
-
"loss": 0.
|
| 676 |
"step": 9300
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 2.35,
|
| 680 |
-
"grad_norm": 0.
|
| 681 |
"learning_rate": 0.0004008889854883929,
|
| 682 |
-
"loss": 0.
|
| 683 |
"step": 9400
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 2.375,
|
| 687 |
-
"grad_norm": 0.
|
| 688 |
"learning_rate": 0.0003989248261231084,
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 9500
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 2.4,
|
| 694 |
-
"grad_norm": 0.
|
| 695 |
"learning_rate": 0.0003969463130731183,
|
| 696 |
-
"loss": 0.
|
| 697 |
"step": 9600
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 2.425,
|
| 701 |
-
"grad_norm": 0.
|
| 702 |
"learning_rate": 0.00039495363703162843,
|
| 703 |
"loss": 0.978,
|
| 704 |
"step": 9700
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 2.45,
|
| 708 |
-
"grad_norm": 0.
|
| 709 |
"learning_rate": 0.000392946990056903,
|
| 710 |
-
"loss": 1.
|
| 711 |
"step": 9800
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 2.475,
|
| 715 |
-
"grad_norm": 0.
|
| 716 |
"learning_rate": 0.00039092656555375416,
|
| 717 |
-
"loss": 1.
|
| 718 |
"step": 9900
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 2.5,
|
| 722 |
-
"grad_norm": 0.
|
| 723 |
"learning_rate": 0.00038889255825490053,
|
| 724 |
-
"loss": 0.
|
| 725 |
"step": 10000
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 2.525,
|
| 729 |
-
"grad_norm": 0.
|
| 730 |
"learning_rate": 0.0003868451642021992,
|
| 731 |
"loss": 0.9745,
|
| 732 |
"step": 10100
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 2.55,
|
| 736 |
-
"grad_norm": 0.
|
| 737 |
"learning_rate": 0.0003847845807277501,
|
| 738 |
-
"loss": 0.
|
| 739 |
"step": 10200
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 2.575,
|
| 743 |
-
"grad_norm": 0.
|
| 744 |
"learning_rate": 0.0003827110064348773,
|
| 745 |
-
"loss": 0.
|
| 746 |
"step": 10300
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 2.6,
|
| 750 |
-
"grad_norm": 0.
|
| 751 |
"learning_rate": 0.0003806246411789872,
|
| 752 |
-
"loss": 0.
|
| 753 |
"step": 10400
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 2.625,
|
| 757 |
-
"grad_norm": 0.
|
| 758 |
"learning_rate": 0.0003785256860483054,
|
| 759 |
-
"loss": 0.
|
| 760 |
"step": 10500
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 2.65,
|
| 764 |
-
"grad_norm": 0.
|
| 765 |
"learning_rate": 0.0003764143433444962,
|
| 766 |
-
"loss": 0.
|
| 767 |
"step": 10600
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 2.675,
|
| 771 |
-
"grad_norm": 0.
|
| 772 |
"learning_rate": 0.0003742908165631636,
|
| 773 |
-
"loss": 0.
|
| 774 |
"step": 10700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 2.7,
|
| 778 |
-
"grad_norm": 0.
|
| 779 |
"learning_rate": 0.0003721553103742388,
|
| 780 |
"loss": 1.0125,
|
| 781 |
"step": 10800
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 2.725,
|
| 785 |
-
"grad_norm": 0.
|
| 786 |
"learning_rate": 0.0003700080306022528,
|
| 787 |
-
"loss": 0.
|
| 788 |
"step": 10900
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 2.75,
|
| 792 |
-
"grad_norm": 0.
|
| 793 |
"learning_rate": 0.0003678491842064995,
|
| 794 |
-
"loss": 0.
|
| 795 |
"step": 11000
|
| 796 |
},
|
| 797 |
{
|
| 798 |
"epoch": 2.775,
|
| 799 |
-
"grad_norm": 0.
|
| 800 |
"learning_rate": 0.00036567897926108756,
|
| 801 |
"loss": 0.9473,
|
| 802 |
"step": 11100
|
| 803 |
},
|
| 804 |
{
|
| 805 |
"epoch": 2.8,
|
| 806 |
-
"grad_norm": 0.
|
| 807 |
"learning_rate": 0.00036349762493488667,
|
| 808 |
-
"loss": 0.
|
| 809 |
"step": 11200
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 2.825,
|
| 813 |
-
"grad_norm": 0.
|
| 814 |
"learning_rate": 0.0003613053314713671,
|
| 815 |
-
"loss": 0.
|
| 816 |
"step": 11300
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 2.85,
|
| 820 |
-
"grad_norm": 0.
|
| 821 |
"learning_rate": 0.0003591023101683355,
|
| 822 |
-
"loss": 0.
|
| 823 |
"step": 11400
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 2.875,
|
| 827 |
-
"grad_norm": 0.
|
| 828 |
"learning_rate": 0.0003568887733575705,
|
| 829 |
-
"loss": 1.
|
| 830 |
"step": 11500
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 2.9,
|
| 834 |
-
"grad_norm": 0.
|
| 835 |
"learning_rate": 0.00035466493438435703,
|
| 836 |
"loss": 0.9675,
|
| 837 |
"step": 11600
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 2.925,
|
| 841 |
-
"grad_norm": 0.
|
| 842 |
"learning_rate": 0.0003524310075869239,
|
| 843 |
-
"loss": 0.
|
| 844 |
"step": 11700
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 2.95,
|
| 848 |
-
"grad_norm": 0.
|
| 849 |
"learning_rate": 0.0003501872082757852,
|
| 850 |
-
"loss": 0.
|
| 851 |
"step": 11800
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 2.975,
|
| 855 |
-
"grad_norm": 0.
|
| 856 |
"learning_rate": 0.000347933752712989,
|
| 857 |
-
"loss": 0.
|
| 858 |
"step": 11900
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 3.0,
|
| 862 |
-
"grad_norm": 0.
|
| 863 |
"learning_rate": 0.0003456708580912725,
|
| 864 |
"loss": 0.9995,
|
| 865 |
"step": 12000
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 3.0,
|
| 869 |
-
"eval_loss": 0.
|
| 870 |
-
"eval_runtime":
|
| 871 |
-
"eval_samples_per_second":
|
| 872 |
-
"eval_steps_per_second":
|
| 873 |
"step": 12000
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 3.025,
|
| 877 |
-
"grad_norm": 0.
|
| 878 |
"learning_rate": 0.0003433987425131291,
|
| 879 |
-
"loss": 0.
|
| 880 |
"step": 12100
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 3.05,
|
| 884 |
-
"grad_norm": 0.
|
| 885 |
"learning_rate": 0.0003411176249697875,
|
| 886 |
-
"loss": 0.
|
| 887 |
"step": 12200
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 3.075,
|
| 891 |
-
"grad_norm": 0.
|
| 892 |
"learning_rate": 0.00033882772532010404,
|
| 893 |
-
"loss": 0.
|
| 894 |
"step": 12300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 3.1,
|
| 898 |
-
"grad_norm": 0.
|
| 899 |
"learning_rate": 0.0003365292642693733,
|
| 900 |
-
"loss": 0.
|
| 901 |
"step": 12400
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 3.125,
|
| 905 |
-
"grad_norm": 0.
|
| 906 |
"learning_rate": 0.00033422246334805503,
|
| 907 |
-
"loss": 0.
|
| 908 |
"step": 12500
|
| 909 |
},
|
| 910 |
{
|
| 911 |
"epoch": 3.15,
|
| 912 |
-
"grad_norm": 0.
|
| 913 |
"learning_rate": 0.0003319075448904234,
|
| 914 |
-
"loss": 0.
|
| 915 |
"step": 12600
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 3.175,
|
| 919 |
-
"grad_norm": 0.
|
| 920 |
"learning_rate": 0.00032958473201313745,
|
| 921 |
-
"loss": 0.
|
| 922 |
"step": 12700
|
| 923 |
},
|
| 924 |
{
|
| 925 |
"epoch": 3.2,
|
| 926 |
-
"grad_norm": 0.
|
| 927 |
"learning_rate": 0.00032725424859373687,
|
| 928 |
-
"loss": 0.
|
| 929 |
"step": 12800
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 3.225,
|
| 933 |
-
"grad_norm": 0.
|
| 934 |
"learning_rate": 0.00032491631924906416,
|
| 935 |
-
"loss": 0.
|
| 936 |
"step": 12900
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 3.25,
|
| 940 |
-
"grad_norm": 0.
|
| 941 |
"learning_rate": 0.00032257116931361555,
|
| 942 |
-
"loss": 0.
|
| 943 |
"step": 13000
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 3.275,
|
| 947 |
-
"grad_norm": 0.
|
| 948 |
"learning_rate": 0.00032021902481782304,
|
| 949 |
"loss": 0.9536,
|
| 950 |
"step": 13100
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 3.3,
|
| 954 |
-
"grad_norm": 0.
|
| 955 |
"learning_rate": 0.00031786011246626855,
|
| 956 |
-
"loss": 0.
|
| 957 |
"step": 13200
|
| 958 |
},
|
| 959 |
{
|
| 960 |
"epoch": 3.325,
|
| 961 |
-
"grad_norm": 0.
|
| 962 |
"learning_rate": 0.0003154946596158343,
|
| 963 |
-
"loss": 0.
|
| 964 |
"step": 13300
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 3.35,
|
| 968 |
-
"grad_norm": 0.
|
| 969 |
"learning_rate": 0.0003131228942537895,
|
| 970 |
"loss": 0.9787,
|
| 971 |
"step": 13400
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 3.375,
|
| 975 |
-
"grad_norm": 0.
|
| 976 |
"learning_rate": 0.000310745044975816,
|
| 977 |
-
"loss": 0.
|
| 978 |
"step": 13500
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 3.4,
|
| 982 |
-
"grad_norm": 0.
|
| 983 |
"learning_rate": 0.0003083613409639764,
|
| 984 |
-
"loss": 0.
|
| 985 |
"step": 13600
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 3.425,
|
| 989 |
-
"grad_norm": 0.
|
| 990 |
"learning_rate": 0.00030597201196462466,
|
| 991 |
-
"loss": 0.
|
| 992 |
"step": 13700
|
| 993 |
},
|
| 994 |
{
|
| 995 |
"epoch": 3.45,
|
| 996 |
-
"grad_norm": 0.
|
| 997 |
"learning_rate": 0.00030357728826626266,
|
| 998 |
-
"loss": 0.
|
| 999 |
"step": 13800
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 3.475,
|
| 1003 |
-
"grad_norm": 0.
|
| 1004 |
"learning_rate": 0.00030117740067734495,
|
| 1005 |
"loss": 0.9835,
|
| 1006 |
"step": 13900
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
"epoch": 3.5,
|
| 1010 |
-
"grad_norm": 0.
|
| 1011 |
"learning_rate": 0.0002987725805040321,
|
| 1012 |
-
"loss": 0.
|
| 1013 |
"step": 14000
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
"epoch": 3.525,
|
| 1017 |
-
"grad_norm": 0.
|
| 1018 |
"learning_rate": 0.0002963630595278977,
|
| 1019 |
-
"loss": 0.
|
| 1020 |
"step": 14100
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
"epoch": 3.55,
|
| 1024 |
-
"grad_norm": 0.
|
| 1025 |
"learning_rate": 0.0002939490699835887,
|
| 1026 |
-
"loss": 0.
|
| 1027 |
"step": 14200
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
"epoch": 3.575,
|
| 1031 |
-
"grad_norm": 0.
|
| 1032 |
"learning_rate": 0.00029153084453644135,
|
| 1033 |
-
"loss": 0.
|
| 1034 |
"step": 14300
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 3.6,
|
| 1038 |
-
"grad_norm": 0.
|
| 1039 |
"learning_rate": 0.00028910861626005774,
|
| 1040 |
-
"loss": 0.
|
| 1041 |
"step": 14400
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 3.625,
|
| 1045 |
-
"grad_norm": 0.
|
| 1046 |
"learning_rate": 0.00028668261861384045,
|
| 1047 |
-
"loss": 0.
|
| 1048 |
"step": 14500
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
"epoch": 3.65,
|
| 1052 |
-
"grad_norm": 0.
|
| 1053 |
"learning_rate": 0.00028425308542049207,
|
| 1054 |
-
"loss": 0.
|
| 1055 |
"step": 14600
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
"epoch": 3.675,
|
| 1059 |
-
"grad_norm": 0.
|
| 1060 |
"learning_rate": 0.0002818202508434783,
|
| 1061 |
-
"loss": 0.
|
| 1062 |
"step": 14700
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 3.7,
|
| 1066 |
-
"grad_norm": 0.
|
| 1067 |
"learning_rate": 0.00027938434936445943,
|
| 1068 |
-
"loss": 0.
|
| 1069 |
"step": 14800
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
"epoch": 3.725,
|
| 1073 |
-
"grad_norm": 0.
|
| 1074 |
"learning_rate": 0.00027694561576068985,
|
| 1075 |
-
"loss": 0.
|
| 1076 |
"step": 14900
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
"epoch": 3.75,
|
| 1080 |
-
"grad_norm": 0.
|
| 1081 |
"learning_rate": 0.0002745042850823902,
|
| 1082 |
-
"loss": 0.
|
| 1083 |
"step": 15000
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
"epoch": 3.775,
|
| 1087 |
-
"grad_norm": 0.
|
| 1088 |
"learning_rate": 0.00027206059263009243,
|
| 1089 |
-
"loss": 0.
|
| 1090 |
"step": 15100
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 3.8,
|
| 1094 |
-
"grad_norm": 0.
|
| 1095 |
"learning_rate": 0.00026961477393196127,
|
| 1096 |
-
"loss": 0.
|
| 1097 |
"step": 15200
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 3.825,
|
| 1101 |
-
"grad_norm": 0.
|
| 1102 |
"learning_rate": 0.0002671670647210934,
|
| 1103 |
-
"loss": 0.
|
| 1104 |
"step": 15300
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 3.85,
|
| 1108 |
-
"grad_norm": 0.
|
| 1109 |
"learning_rate": 0.00026471770091279724,
|
| 1110 |
-
"loss": 0.
|
| 1111 |
"step": 15400
|
| 1112 |
},
|
| 1113 |
{
|
| 1114 |
"epoch": 3.875,
|
| 1115 |
-
"grad_norm": 0.
|
| 1116 |
"learning_rate": 0.00026226691858185456,
|
| 1117 |
-
"loss": 0.
|
| 1118 |
"step": 15500
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
"epoch": 3.9,
|
| 1122 |
-
"grad_norm": 0.
|
| 1123 |
"learning_rate": 0.00025981495393976716,
|
| 1124 |
-
"loss": 0.
|
| 1125 |
"step": 15600
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
"epoch": 3.925,
|
| 1129 |
-
"grad_norm": 0.
|
| 1130 |
"learning_rate": 0.00025736204331199084,
|
| 1131 |
-
"loss": 0.
|
| 1132 |
"step": 15700
|
| 1133 |
},
|
| 1134 |
{
|
| 1135 |
"epoch": 3.95,
|
| 1136 |
-
"grad_norm": 0.
|
| 1137 |
"learning_rate": 0.00025490842311515704,
|
| 1138 |
-
"loss": 0.
|
| 1139 |
"step": 15800
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
"epoch": 3.975,
|
| 1143 |
-
"grad_norm": 0.
|
| 1144 |
"learning_rate": 0.0002524543298342875,
|
| 1145 |
-
"loss": 0.
|
| 1146 |
"step": 15900
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
"epoch": 4.0,
|
| 1150 |
-
"grad_norm": 0.
|
| 1151 |
"learning_rate": 0.00025,
|
| 1152 |
-
"loss": 0.
|
| 1153 |
"step": 16000
|
| 1154 |
},
|
| 1155 |
{
|
| 1156 |
"epoch": 4.0,
|
| 1157 |
-
"eval_loss": 0.
|
| 1158 |
-
"eval_runtime":
|
| 1159 |
-
"eval_samples_per_second":
|
| 1160 |
-
"eval_steps_per_second":
|
| 1161 |
"step": 16000
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 4.025,
|
| 1165 |
-
"grad_norm": 0.
|
| 1166 |
"learning_rate": 0.0002475456701657126,
|
| 1167 |
-
"loss": 0.
|
| 1168 |
"step": 16100
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 4.05,
|
| 1172 |
-
"grad_norm": 0.
|
| 1173 |
"learning_rate": 0.00024509157688484297,
|
| 1174 |
-
"loss": 0.
|
| 1175 |
"step": 16200
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 4.075,
|
| 1179 |
-
"grad_norm": 0.
|
| 1180 |
"learning_rate": 0.0002426379566880092,
|
| 1181 |
-
"loss": 0.
|
| 1182 |
"step": 16300
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 4.1,
|
| 1186 |
-
"grad_norm": 0.
|
| 1187 |
"learning_rate": 0.00024018504606023293,
|
| 1188 |
-
"loss": 0.
|
| 1189 |
"step": 16400
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 4.125,
|
| 1193 |
-
"grad_norm": 0.
|
| 1194 |
"learning_rate": 0.0002377330814181455,
|
| 1195 |
-
"loss": 0.
|
| 1196 |
"step": 16500
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 4.15,
|
| 1200 |
-
"grad_norm": 0.
|
| 1201 |
"learning_rate": 0.00023528229908720272,
|
| 1202 |
-
"loss": 0.
|
| 1203 |
"step": 16600
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 4.175,
|
| 1207 |
-
"grad_norm": 0.
|
| 1208 |
"learning_rate": 0.00023283293527890658,
|
| 1209 |
-
"loss": 0.
|
| 1210 |
"step": 16700
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 4.2,
|
| 1214 |
-
"grad_norm": 0.
|
| 1215 |
"learning_rate": 0.0002303852260680388,
|
| 1216 |
-
"loss": 0.
|
| 1217 |
"step": 16800
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 4.225,
|
| 1221 |
-
"grad_norm": 0.
|
| 1222 |
"learning_rate": 0.00022793940736990766,
|
| 1223 |
-
"loss": 0.
|
| 1224 |
"step": 16900
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 4.25,
|
| 1228 |
-
"grad_norm": 0.
|
| 1229 |
"learning_rate": 0.00022549571491760985,
|
| 1230 |
-
"loss": 0.
|
| 1231 |
"step": 17000
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 4.275,
|
| 1235 |
-
"grad_norm": 0.
|
| 1236 |
"learning_rate": 0.00022305438423931017,
|
| 1237 |
-
"loss": 0.
|
| 1238 |
"step": 17100
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 4.3,
|
| 1242 |
-
"grad_norm":
|
| 1243 |
"learning_rate": 0.00022061565063554063,
|
| 1244 |
-
"loss": 0.
|
| 1245 |
"step": 17200
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 4.325,
|
| 1249 |
-
"grad_norm": 0.
|
| 1250 |
"learning_rate": 0.00021817974915652172,
|
| 1251 |
-
"loss": 0.
|
| 1252 |
"step": 17300
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 4.35,
|
| 1256 |
-
"grad_norm": 0.
|
| 1257 |
"learning_rate": 0.00021574691457950805,
|
| 1258 |
-
"loss": 0.
|
| 1259 |
"step": 17400
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 4.375,
|
| 1263 |
-
"grad_norm": 0.
|
| 1264 |
"learning_rate": 0.00021331738138615958,
|
| 1265 |
-
"loss": 0.
|
| 1266 |
"step": 17500
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 4.4,
|
| 1270 |
-
"grad_norm": 0.
|
| 1271 |
"learning_rate": 0.00021089138373994224,
|
| 1272 |
-
"loss": 0.
|
| 1273 |
"step": 17600
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 4.425,
|
| 1277 |
-
"grad_norm": 0.
|
| 1278 |
"learning_rate": 0.0002084691554635587,
|
| 1279 |
-
"loss": 0.
|
| 1280 |
"step": 17700
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 4.45,
|
| 1284 |
-
"grad_norm": 0.
|
| 1285 |
"learning_rate": 0.00020605093001641137,
|
| 1286 |
-
"loss": 0.
|
| 1287 |
"step": 17800
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 4.475,
|
| 1291 |
-
"grad_norm": 0.
|
| 1292 |
"learning_rate": 0.00020363694047210228,
|
| 1293 |
-
"loss": 0.
|
| 1294 |
"step": 17900
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 4.5,
|
| 1298 |
-
"grad_norm": 0.
|
| 1299 |
"learning_rate": 0.00020122741949596797,
|
| 1300 |
-
"loss": 0.
|
| 1301 |
"step": 18000
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 4.525,
|
| 1305 |
-
"grad_norm": 0.
|
| 1306 |
"learning_rate": 0.00019882259932265512,
|
| 1307 |
-
"loss": 0.
|
| 1308 |
"step": 18100
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 4.55,
|
| 1312 |
-
"grad_norm": 0.
|
| 1313 |
"learning_rate": 0.00019642271173373735,
|
| 1314 |
-
"loss": 0.
|
| 1315 |
"step": 18200
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 4.575,
|
| 1319 |
-
"grad_norm": 0.
|
| 1320 |
"learning_rate": 0.00019402798803537538,
|
| 1321 |
-
"loss": 0.
|
| 1322 |
"step": 18300
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 4.6,
|
| 1326 |
-
"grad_norm": 0.
|
| 1327 |
"learning_rate": 0.00019163865903602372,
|
| 1328 |
"loss": 0.941,
|
| 1329 |
"step": 18400
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 4.625,
|
| 1333 |
-
"grad_norm": 0.
|
| 1334 |
"learning_rate": 0.00018925495502418406,
|
| 1335 |
-
"loss": 0.
|
| 1336 |
"step": 18500
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 4.65,
|
| 1340 |
-
"grad_norm": 0.
|
| 1341 |
"learning_rate": 0.00018687710574621051,
|
| 1342 |
-
"loss": 0.
|
| 1343 |
"step": 18600
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 4.675,
|
| 1347 |
-
"grad_norm": 0.
|
| 1348 |
"learning_rate": 0.00018450534038416566,
|
| 1349 |
-
"loss": 0.
|
| 1350 |
"step": 18700
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 4.7,
|
| 1354 |
-
"grad_norm": 0.
|
| 1355 |
"learning_rate": 0.00018213988753373146,
|
| 1356 |
-
"loss": 0.
|
| 1357 |
"step": 18800
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 4.725,
|
| 1361 |
-
"grad_norm": 0.
|
| 1362 |
"learning_rate": 0.00017978097518217702,
|
| 1363 |
-
"loss": 0.
|
| 1364 |
"step": 18900
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 4.75,
|
| 1368 |
-
"grad_norm": 0.
|
| 1369 |
"learning_rate": 0.00017742883068638446,
|
| 1370 |
-
"loss": 0.
|
| 1371 |
"step": 19000
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 4.775,
|
| 1375 |
-
"grad_norm": 0.
|
| 1376 |
"learning_rate": 0.00017508368075093582,
|
| 1377 |
-
"loss": 0.
|
| 1378 |
"step": 19100
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 4.8,
|
| 1382 |
-
"grad_norm": 0.
|
| 1383 |
"learning_rate": 0.00017274575140626317,
|
| 1384 |
-
"loss": 0.
|
| 1385 |
"step": 19200
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 4.825,
|
| 1389 |
-
"grad_norm": 0.
|
| 1390 |
"learning_rate": 0.0001704152679868626,
|
| 1391 |
-
"loss": 0.
|
| 1392 |
"step": 19300
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 4.85,
|
| 1396 |
-
"grad_norm": 0.
|
| 1397 |
"learning_rate": 0.00016809245510957666,
|
| 1398 |
-
"loss": 0.
|
| 1399 |
"step": 19400
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 4.875,
|
| 1403 |
-
"grad_norm": 0.
|
| 1404 |
"learning_rate": 0.000165777536651945,
|
| 1405 |
-
"loss": 0.
|
| 1406 |
"step": 19500
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 4.9,
|
| 1410 |
-
"grad_norm": 0.
|
| 1411 |
"learning_rate": 0.0001634707357306267,
|
| 1412 |
-
"loss": 0.
|
| 1413 |
"step": 19600
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 4.925,
|
| 1417 |
-
"grad_norm": 0.
|
| 1418 |
"learning_rate": 0.00016117227467989602,
|
| 1419 |
-
"loss": 0.
|
| 1420 |
"step": 19700
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 4.95,
|
| 1424 |
-
"grad_norm": 0.
|
| 1425 |
"learning_rate": 0.0001588823750302126,
|
| 1426 |
-
"loss": 0.
|
| 1427 |
"step": 19800
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 4.975,
|
| 1431 |
-
"grad_norm": 0.
|
| 1432 |
"learning_rate": 0.00015660125748687094,
|
| 1433 |
-
"loss": 0.
|
| 1434 |
"step": 19900
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 5.0,
|
| 1438 |
-
"grad_norm": 0.
|
| 1439 |
"learning_rate": 0.00015432914190872756,
|
| 1440 |
-
"loss": 0.
|
| 1441 |
"step": 20000
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 5.0,
|
| 1445 |
-
"eval_loss": 0.
|
| 1446 |
-
"eval_runtime":
|
| 1447 |
-
"eval_samples_per_second":
|
| 1448 |
-
"eval_steps_per_second":
|
| 1449 |
"step": 20000
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 5.025,
|
| 1453 |
-
"grad_norm": 0.
|
| 1454 |
"learning_rate": 0.000152066247287011,
|
| 1455 |
-
"loss": 0.
|
| 1456 |
"step": 20100
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 5.05,
|
| 1460 |
-
"grad_norm": 0.
|
| 1461 |
"learning_rate": 0.00014981279172421482,
|
| 1462 |
"loss": 0.9414,
|
| 1463 |
"step": 20200
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 5.075,
|
| 1467 |
-
"grad_norm": 0.
|
| 1468 |
"learning_rate": 0.00014756899241307614,
|
| 1469 |
-
"loss": 0.
|
| 1470 |
"step": 20300
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 5.1,
|
| 1474 |
-
"grad_norm": 0.
|
| 1475 |
"learning_rate": 0.00014533506561564306,
|
| 1476 |
-
"loss": 0.
|
| 1477 |
"step": 20400
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 5.125,
|
| 1481 |
-
"grad_norm": 0.
|
| 1482 |
"learning_rate": 0.00014311122664242953,
|
| 1483 |
-
"loss": 0.
|
| 1484 |
"step": 20500
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 5.15,
|
| 1488 |
-
"grad_norm": 0.
|
| 1489 |
"learning_rate": 0.00014089768983166444,
|
| 1490 |
-
"loss": 0.
|
| 1491 |
"step": 20600
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 5.175,
|
| 1495 |
-
"grad_norm": 0.
|
| 1496 |
"learning_rate": 0.000138694668528633,
|
| 1497 |
-
"loss": 0.
|
| 1498 |
"step": 20700
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 5.2,
|
| 1502 |
-
"grad_norm": 0.
|
| 1503 |
"learning_rate": 0.00013650237506511331,
|
| 1504 |
-
"loss": 0.
|
| 1505 |
"step": 20800
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 5.225,
|
| 1509 |
-
"grad_norm": 0.
|
| 1510 |
"learning_rate": 0.0001343210207389125,
|
| 1511 |
-
"loss": 0.
|
| 1512 |
"step": 20900
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 5.25,
|
| 1516 |
-
"grad_norm": 0.
|
| 1517 |
"learning_rate": 0.00013215081579350058,
|
| 1518 |
-
"loss": 0.
|
| 1519 |
"step": 21000
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 5.275,
|
| 1523 |
-
"grad_norm": 0.
|
| 1524 |
"learning_rate": 0.00012999196939774722,
|
| 1525 |
-
"loss": 0.
|
| 1526 |
"step": 21100
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 5.3,
|
| 1530 |
-
"grad_norm": 0.
|
| 1531 |
"learning_rate": 0.00012784468962576134,
|
| 1532 |
-
"loss": 0.
|
| 1533 |
"step": 21200
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 5.325,
|
| 1537 |
-
"grad_norm": 0.
|
| 1538 |
"learning_rate": 0.00012570918343683636,
|
| 1539 |
-
"loss": 0.
|
| 1540 |
"step": 21300
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 5.35,
|
| 1544 |
-
"grad_norm": 0.
|
| 1545 |
"learning_rate": 0.0001235856566555039,
|
| 1546 |
-
"loss": 0.
|
| 1547 |
"step": 21400
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 5.375,
|
| 1551 |
-
"grad_norm": 0.
|
| 1552 |
"learning_rate": 0.0001214743139516946,
|
| 1553 |
-
"loss": 0.
|
| 1554 |
"step": 21500
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 5.4,
|
| 1558 |
-
"grad_norm": 0.
|
| 1559 |
"learning_rate": 0.00011937535882101281,
|
| 1560 |
-
"loss": 0.
|
| 1561 |
"step": 21600
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 5.425,
|
| 1565 |
-
"grad_norm": 0.
|
| 1566 |
"learning_rate": 0.00011728899356512265,
|
| 1567 |
-
"loss": 0.
|
| 1568 |
"step": 21700
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 5.45,
|
| 1572 |
-
"grad_norm": 0.
|
| 1573 |
"learning_rate": 0.00011521541927224994,
|
| 1574 |
-
"loss": 0.
|
| 1575 |
"step": 21800
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 5.475,
|
| 1579 |
-
"grad_norm": 0.
|
| 1580 |
"learning_rate": 0.00011315483579780094,
|
| 1581 |
-
"loss": 0.
|
| 1582 |
"step": 21900
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 5.5,
|
| 1586 |
-
"grad_norm": 0.
|
| 1587 |
"learning_rate": 0.00011110744174509952,
|
| 1588 |
-
"loss": 0.
|
| 1589 |
"step": 22000
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 5.525,
|
| 1593 |
-
"grad_norm": 0.
|
| 1594 |
"learning_rate": 0.00010907343444624579,
|
| 1595 |
-
"loss": 0.
|
| 1596 |
"step": 22100
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 5.55,
|
| 1600 |
-
"grad_norm": 0.
|
| 1601 |
"learning_rate": 0.00010705300994309697,
|
| 1602 |
"loss": 0.9286,
|
| 1603 |
"step": 22200
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 5.575,
|
| 1607 |
-
"grad_norm": 0.
|
| 1608 |
"learning_rate": 0.00010504636296837161,
|
| 1609 |
-
"loss": 0.
|
| 1610 |
"step": 22300
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 5.6,
|
| 1614 |
-
"grad_norm": 0.
|
| 1615 |
"learning_rate": 0.00010305368692688174,
|
| 1616 |
-
"loss": 0.
|
| 1617 |
"step": 22400
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 5.625,
|
| 1621 |
-
"grad_norm": 0.
|
| 1622 |
"learning_rate": 0.00010107517387689166,
|
| 1623 |
-
"loss": 0.
|
| 1624 |
"step": 22500
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 5.65,
|
| 1628 |
-
"grad_norm": 0.
|
| 1629 |
"learning_rate": 9.911101451160715e-05,
|
| 1630 |
-
"loss": 0.
|
| 1631 |
"step": 22600
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 5.675,
|
| 1635 |
-
"grad_norm": 0.
|
| 1636 |
"learning_rate": 9.716139814079594e-05,
|
| 1637 |
-
"loss": 0.
|
| 1638 |
"step": 22700
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 5.7,
|
| 1642 |
-
"grad_norm": 0.
|
| 1643 |
"learning_rate": 9.522651267254148e-05,
|
| 1644 |
-
"loss": 0.
|
| 1645 |
"step": 22800
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
"epoch": 5.725,
|
| 1649 |
-
"grad_norm": 0.
|
| 1650 |
"learning_rate": 9.330654459513265e-05,
|
| 1651 |
-
"loss": 0.
|
| 1652 |
"step": 22900
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
"epoch": 5.75,
|
| 1656 |
-
"grad_norm": 0.
|
| 1657 |
"learning_rate": 9.140167895908866e-05,
|
| 1658 |
-
"loss": 0.
|
| 1659 |
"step": 23000
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
"epoch": 5.775,
|
| 1663 |
-
"grad_norm": 0.
|
| 1664 |
"learning_rate": 8.951209935932425e-05,
|
| 1665 |
"loss": 0.9113,
|
| 1666 |
"step": 23100
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
"epoch": 5.8,
|
| 1670 |
-
"grad_norm": 0.
|
| 1671 |
"learning_rate": 8.763798791745412e-05,
|
| 1672 |
-
"loss": 0.
|
| 1673 |
"step": 23200
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
"epoch": 5.825,
|
| 1677 |
-
"grad_norm": 0.
|
| 1678 |
"learning_rate": 8.577952526423969e-05,
|
| 1679 |
-
"loss": 0.
|
| 1680 |
"step": 23300
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
"epoch": 5.85,
|
| 1684 |
-
"grad_norm": 0.
|
| 1685 |
"learning_rate": 8.393689052217964e-05,
|
| 1686 |
-
"loss": 0.
|
| 1687 |
"step": 23400
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
"epoch": 5.875,
|
| 1691 |
-
"grad_norm": 0.
|
| 1692 |
"learning_rate": 8.211026128824539e-05,
|
| 1693 |
-
"loss": 0.
|
| 1694 |
"step": 23500
|
| 1695 |
},
|
| 1696 |
{
|
| 1697 |
"epoch": 5.9,
|
| 1698 |
-
"grad_norm": 0.
|
| 1699 |
"learning_rate": 8.029981361676455e-05,
|
| 1700 |
-
"loss": 0.
|
| 1701 |
"step": 23600
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 5.925,
|
| 1705 |
-
"grad_norm": 0.
|
| 1706 |
"learning_rate": 7.850572200245185e-05,
|
| 1707 |
-
"loss": 0.
|
| 1708 |
"step": 23700
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
"epoch": 5.95,
|
| 1712 |
-
"grad_norm": 0.
|
| 1713 |
"learning_rate": 7.672815936359106e-05,
|
| 1714 |
-
"loss": 0.
|
| 1715 |
"step": 23800
|
| 1716 |
},
|
| 1717 |
{
|
| 1718 |
"epoch": 5.975,
|
| 1719 |
-
"grad_norm": 0.
|
| 1720 |
"learning_rate": 7.496729702536912e-05,
|
| 1721 |
-
"loss": 0.
|
| 1722 |
"step": 23900
|
| 1723 |
},
|
| 1724 |
{
|
| 1725 |
"epoch": 6.0,
|
| 1726 |
-
"grad_norm": 0.
|
| 1727 |
"learning_rate": 7.322330470336314e-05,
|
| 1728 |
-
"loss": 0.
|
| 1729 |
"step": 24000
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
"epoch": 6.0,
|
| 1733 |
-
"eval_loss": 0.
|
| 1734 |
-
"eval_runtime":
|
| 1735 |
-
"eval_samples_per_second":
|
| 1736 |
-
"eval_steps_per_second":
|
| 1737 |
"step": 24000
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 6.025,
|
| 1741 |
-
"grad_norm": 0.
|
| 1742 |
"learning_rate": 7.149635048718294e-05,
|
| 1743 |
-
"loss": 0.
|
| 1744 |
"step": 24100
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 6.05,
|
| 1748 |
-
"grad_norm": 0.
|
| 1749 |
"learning_rate": 6.97866008242703e-05,
|
| 1750 |
"loss": 0.9105,
|
| 1751 |
"step": 24200
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 6.075,
|
| 1755 |
-
"grad_norm": 0.
|
| 1756 |
"learning_rate": 6.809422050385628e-05,
|
| 1757 |
-
"loss": 0.
|
| 1758 |
"step": 24300
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 6.1,
|
| 1762 |
-
"grad_norm": 0.
|
| 1763 |
"learning_rate": 6.641937264107867e-05,
|
| 1764 |
-
"loss": 0.
|
| 1765 |
"step": 24400
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 6.125,
|
| 1769 |
-
"grad_norm": 0.
|
| 1770 |
"learning_rate": 6.476221866126028e-05,
|
| 1771 |
-
"loss": 0.
|
| 1772 |
"step": 24500
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 6.15,
|
| 1776 |
-
"grad_norm": 0.
|
| 1777 |
"learning_rate": 6.312291828435076e-05,
|
| 1778 |
-
"loss": 0.
|
| 1779 |
"step": 24600
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 6.175,
|
| 1783 |
-
"grad_norm": 0.
|
| 1784 |
"learning_rate": 6.150162950953264e-05,
|
| 1785 |
-
"loss": 0.
|
| 1786 |
"step": 24700
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 6.2,
|
| 1790 |
-
"grad_norm": 0.
|
| 1791 |
"learning_rate": 5.989850859999227e-05,
|
| 1792 |
-
"loss": 0.
|
| 1793 |
"step": 24800
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 6.225,
|
| 1797 |
-
"grad_norm": 0.
|
| 1798 |
"learning_rate": 5.831371006785963e-05,
|
| 1799 |
-
"loss": 0.
|
| 1800 |
"step": 24900
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 6.25,
|
| 1804 |
-
"grad_norm": 0.
|
| 1805 |
"learning_rate": 5.6747386659315755e-05,
|
| 1806 |
-
"loss": 0.
|
| 1807 |
"step": 25000
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 6.275,
|
| 1811 |
-
"grad_norm": 0.
|
| 1812 |
"learning_rate": 5.519968933987082e-05,
|
| 1813 |
-
"loss": 0.
|
| 1814 |
"step": 25100
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 6.3,
|
| 1818 |
-
"grad_norm": 0.
|
| 1819 |
"learning_rate": 5.367076727981382e-05,
|
| 1820 |
-
"loss": 0.
|
| 1821 |
"step": 25200
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 6.325,
|
| 1825 |
-
"grad_norm": 0.
|
| 1826 |
"learning_rate": 5.216076783983492e-05,
|
| 1827 |
-
"loss": 0.
|
| 1828 |
"step": 25300
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 6.35,
|
| 1832 |
-
"grad_norm": 0.
|
| 1833 |
"learning_rate": 5.066983655682325e-05,
|
| 1834 |
-
"loss": 0.
|
| 1835 |
"step": 25400
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 6.375,
|
| 1839 |
-
"grad_norm": 0.
|
| 1840 |
"learning_rate": 4.919811712983879e-05,
|
| 1841 |
-
"loss": 0.
|
| 1842 |
"step": 25500
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 6.4,
|
| 1846 |
-
"grad_norm": 0.
|
| 1847 |
"learning_rate": 4.7745751406263163e-05,
|
| 1848 |
-
"loss": 0.
|
| 1849 |
"step": 25600
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 6.425,
|
| 1853 |
-
"grad_norm": 0.
|
| 1854 |
"learning_rate": 4.6312879368127645e-05,
|
| 1855 |
-
"loss": 0.
|
| 1856 |
"step": 25700
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 6.45,
|
| 1860 |
-
"grad_norm": 0.
|
| 1861 |
"learning_rate": 4.4899639118621604e-05,
|
| 1862 |
-
"loss": 0.
|
| 1863 |
"step": 25800
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 6.475,
|
| 1867 |
-
"grad_norm": 0.
|
| 1868 |
"learning_rate": 4.350616686878175e-05,
|
| 1869 |
-
"loss": 0.
|
| 1870 |
"step": 25900
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 6.5,
|
| 1874 |
-
"grad_norm": 0.
|
| 1875 |
"learning_rate": 4.213259692436367e-05,
|
| 1876 |
-
"loss": 0.
|
| 1877 |
"step": 26000
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 6.525,
|
| 1881 |
-
"grad_norm": 0.
|
| 1882 |
"learning_rate": 4.077906167289766e-05,
|
| 1883 |
-
"loss": 0.
|
| 1884 |
"step": 26100
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 6.55,
|
| 1888 |
-
"grad_norm": 0.
|
| 1889 |
"learning_rate": 3.944569157092839e-05,
|
| 1890 |
-
"loss": 0.
|
| 1891 |
"step": 26200
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 6.575,
|
| 1895 |
-
"grad_norm": 0.
|
| 1896 |
"learning_rate": 3.8132615131441396e-05,
|
| 1897 |
-
"loss": 0.
|
| 1898 |
"step": 26300
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 6.6,
|
| 1902 |
-
"grad_norm": 0.
|
| 1903 |
"learning_rate": 3.6839958911476953e-05,
|
| 1904 |
-
"loss": 0.
|
| 1905 |
"step": 26400
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 6.625,
|
| 1909 |
-
"grad_norm": 0.
|
| 1910 |
"learning_rate": 3.5567847499932e-05,
|
| 1911 |
"loss": 0.9453,
|
| 1912 |
"step": 26500
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 6.65,
|
| 1916 |
-
"grad_norm": 0.
|
| 1917 |
"learning_rate": 3.431640350555204e-05,
|
| 1918 |
-
"loss": 0.
|
| 1919 |
"step": 26600
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 6.675,
|
| 1923 |
-
"grad_norm": 0.
|
| 1924 |
"learning_rate": 3.308574754511404e-05,
|
| 1925 |
-
"loss": 0.
|
| 1926 |
"step": 26700
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 6.7,
|
| 1930 |
-
"grad_norm": 0.
|
| 1931 |
"learning_rate": 3.187599823180071e-05,
|
| 1932 |
-
"loss": 0.
|
| 1933 |
"step": 26800
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 6.725,
|
| 1937 |
-
"grad_norm": 0.
|
| 1938 |
"learning_rate": 3.0687272163768986e-05,
|
| 1939 |
-
"loss": 0.
|
| 1940 |
"step": 26900
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 6.75,
|
| 1944 |
-
"grad_norm": 0.
|
| 1945 |
"learning_rate": 2.9519683912911265e-05,
|
| 1946 |
-
"loss": 0.
|
| 1947 |
"step": 27000
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 6.775,
|
| 1951 |
-
"grad_norm": 0.
|
| 1952 |
"learning_rate": 2.8373346013813417e-05,
|
| 1953 |
-
"loss": 0.
|
| 1954 |
"step": 27100
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 6.8,
|
| 1958 |
-
"grad_norm": 0.
|
| 1959 |
"learning_rate": 2.7248368952908055e-05,
|
| 1960 |
-
"loss": 0.
|
| 1961 |
"step": 27200
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 6.825,
|
| 1965 |
-
"grad_norm": 0.
|
| 1966 |
"learning_rate": 2.6144861157825773e-05,
|
| 1967 |
-
"loss": 0.
|
| 1968 |
"step": 27300
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 6.85,
|
| 1972 |
-
"grad_norm": 0.
|
| 1973 |
"learning_rate": 2.5062928986944677e-05,
|
| 1974 |
-
"loss": 0.
|
| 1975 |
"step": 27400
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 6.875,
|
| 1979 |
-
"grad_norm": 0.
|
| 1980 |
"learning_rate": 2.4002676719139166e-05,
|
| 1981 |
"loss": 0.9394,
|
| 1982 |
"step": 27500
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 6.9,
|
| 1986 |
-
"grad_norm": 0.
|
| 1987 |
"learning_rate": 2.296420654372966e-05,
|
| 1988 |
-
"loss": 0.
|
| 1989 |
"step": 27600
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 6.925,
|
| 1993 |
-
"grad_norm": 0.
|
| 1994 |
"learning_rate": 2.1947618550633096e-05,
|
| 1995 |
-
"loss": 0.
|
| 1996 |
"step": 27700
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 6.95,
|
| 2000 |
-
"grad_norm": 0.
|
| 2001 |
"learning_rate": 2.0953010720716037e-05,
|
| 2002 |
-
"loss": 0.
|
| 2003 |
"step": 27800
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 6.975,
|
| 2007 |
-
"grad_norm": 0.
|
| 2008 |
"learning_rate": 1.9980478916351297e-05,
|
| 2009 |
-
"loss": 0.
|
| 2010 |
"step": 27900
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 7.0,
|
| 2014 |
-
"grad_norm": 0.
|
| 2015 |
"learning_rate": 1.9030116872178316e-05,
|
| 2016 |
-
"loss": 0.
|
| 2017 |
"step": 28000
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 7.0,
|
| 2021 |
-
"eval_loss": 0.
|
| 2022 |
-
"eval_runtime":
|
| 2023 |
-
"eval_samples_per_second":
|
| 2024 |
-
"eval_steps_per_second":
|
| 2025 |
"step": 28000
|
| 2026 |
},
|
| 2027 |
{
|
| 2028 |
"epoch": 7.025,
|
| 2029 |
-
"grad_norm": 0.
|
| 2030 |
"learning_rate": 1.8102016186068992e-05,
|
| 2031 |
-
"loss": 0.
|
| 2032 |
"step": 28100
|
| 2033 |
},
|
| 2034 |
{
|
| 2035 |
"epoch": 7.05,
|
| 2036 |
-
"grad_norm": 0.
|
| 2037 |
"learning_rate": 1.719626631029911e-05,
|
| 2038 |
-
"loss": 0.
|
| 2039 |
"step": 28200
|
| 2040 |
},
|
| 2041 |
{
|
| 2042 |
"epoch": 7.075,
|
| 2043 |
-
"grad_norm": 0.
|
| 2044 |
"learning_rate": 1.6312954542926888e-05,
|
| 2045 |
-
"loss": 0.
|
| 2046 |
"step": 28300
|
| 2047 |
},
|
| 2048 |
{
|
| 2049 |
"epoch": 7.1,
|
| 2050 |
-
"grad_norm": 0.
|
| 2051 |
"learning_rate": 1.5452166019378987e-05,
|
| 2052 |
-
"loss": 0.
|
| 2053 |
"step": 28400
|
| 2054 |
},
|
| 2055 |
{
|
| 2056 |
"epoch": 7.125,
|
| 2057 |
-
"grad_norm": 0.
|
| 2058 |
"learning_rate": 1.4613983704244827e-05,
|
| 2059 |
-
"loss": 0.
|
| 2060 |
"step": 28500
|
| 2061 |
},
|
| 2062 |
{
|
| 2063 |
"epoch": 7.15,
|
| 2064 |
-
"grad_norm": 0.
|
| 2065 |
"learning_rate": 1.3798488383280488e-05,
|
| 2066 |
-
"loss": 0.
|
| 2067 |
"step": 28600
|
| 2068 |
},
|
| 2069 |
{
|
| 2070 |
"epoch": 7.175,
|
| 2071 |
-
"grad_norm": 0.
|
| 2072 |
"learning_rate": 1.3005758655622424e-05,
|
| 2073 |
-
"loss": 0.
|
| 2074 |
"step": 28700
|
| 2075 |
},
|
| 2076 |
{
|
| 2077 |
"epoch": 7.2,
|
| 2078 |
-
"grad_norm": 0.
|
| 2079 |
"learning_rate": 1.2235870926211617e-05,
|
| 2080 |
-
"loss": 0.
|
| 2081 |
"step": 28800
|
| 2082 |
},
|
| 2083 |
{
|
| 2084 |
"epoch": 7.225,
|
| 2085 |
-
"grad_norm": 0.
|
| 2086 |
"learning_rate": 1.1488899398429897e-05,
|
| 2087 |
-
"loss": 0.
|
| 2088 |
"step": 28900
|
| 2089 |
},
|
| 2090 |
{
|
| 2091 |
"epoch": 7.25,
|
| 2092 |
-
"grad_norm": 0.
|
| 2093 |
"learning_rate": 1.0764916066947795e-05,
|
| 2094 |
-
"loss": 0.
|
| 2095 |
"step": 29000
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
"epoch": 7.275,
|
| 2099 |
-
"grad_norm": 0.
|
| 2100 |
"learning_rate": 1.0063990710785648e-05,
|
| 2101 |
-
"loss": 0.
|
| 2102 |
"step": 29100
|
| 2103 |
},
|
| 2104 |
{
|
| 2105 |
"epoch": 7.3,
|
| 2106 |
-
"grad_norm": 0.
|
| 2107 |
"learning_rate": 9.386190886588208e-06,
|
| 2108 |
-
"loss": 0.
|
| 2109 |
"step": 29200
|
| 2110 |
},
|
| 2111 |
{
|
| 2112 |
"epoch": 7.325,
|
| 2113 |
-
"grad_norm": 0.
|
| 2114 |
"learning_rate": 8.731581922113152e-06,
|
| 2115 |
-
"loss": 0.
|
| 2116 |
"step": 29300
|
| 2117 |
},
|
| 2118 |
{
|
| 2119 |
"epoch": 7.35,
|
| 2120 |
-
"grad_norm": 0.
|
| 2121 |
"learning_rate": 8.10022690993506e-06,
|
| 2122 |
-
"loss": 0.
|
| 2123 |
"step": 29400
|
| 2124 |
},
|
| 2125 |
{
|
| 2126 |
"epoch": 7.375,
|
| 2127 |
-
"grad_norm": 0.
|
| 2128 |
"learning_rate": 7.4921867013640064e-06,
|
| 2129 |
-
"loss": 0.
|
| 2130 |
"step": 29500
|
| 2131 |
},
|
| 2132 |
{
|
| 2133 |
"epoch": 7.4,
|
| 2134 |
-
"grad_norm": 0.
|
| 2135 |
"learning_rate": 6.907519900580861e-06,
|
| 2136 |
-
"loss": 0.
|
| 2137 |
"step": 29600
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
"epoch": 7.425,
|
| 2141 |
-
"grad_norm": 0.
|
| 2142 |
"learning_rate": 6.34628285898875e-06,
|
| 2143 |
-
"loss": 0.
|
| 2144 |
"step": 29700
|
| 2145 |
},
|
| 2146 |
{
|
| 2147 |
"epoch": 7.45,
|
| 2148 |
-
"grad_norm": 0.
|
| 2149 |
"learning_rate": 5.808529669781903e-06,
|
| 2150 |
-
"loss": 0.
|
| 2151 |
"step": 29800
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
"epoch": 7.475,
|
| 2155 |
-
"grad_norm": 0.
|
| 2156 |
"learning_rate": 5.294312162731935e-06,
|
| 2157 |
-
"loss": 0.
|
| 2158 |
"step": 29900
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
"epoch": 7.5,
|
| 2162 |
-
"grad_norm": 0.
|
| 2163 |
"learning_rate": 4.803679899192393e-06,
|
| 2164 |
-
"loss": 0.
|
| 2165 |
"step": 30000
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
"epoch": 7.525,
|
| 2169 |
-
"grad_norm": 0.
|
| 2170 |
"learning_rate": 4.336680167322055e-06,
|
| 2171 |
-
"loss": 0.
|
| 2172 |
"step": 30100
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
"epoch": 7.55,
|
| 2176 |
-
"grad_norm": 0.
|
| 2177 |
"learning_rate": 3.893357977527101e-06,
|
| 2178 |
-
"loss": 0.
|
| 2179 |
"step": 30200
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
"epoch": 7.575,
|
| 2183 |
-
"grad_norm": 0.
|
| 2184 |
"learning_rate": 3.4737560581228343e-06,
|
| 2185 |
-
"loss": 0.
|
| 2186 |
"step": 30300
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
"epoch": 7.6,
|
| 2190 |
-
"grad_norm": 0.
|
| 2191 |
"learning_rate": 3.077914851215585e-06,
|
| 2192 |
-
"loss": 0.
|
| 2193 |
"step": 30400
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
"epoch": 7.625,
|
| 2197 |
-
"grad_norm": 0.
|
| 2198 |
"learning_rate": 2.7058725088047465e-06,
|
| 2199 |
-
"loss": 0.
|
| 2200 |
"step": 30500
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
"epoch": 7.65,
|
| 2204 |
-
"grad_norm": 0.
|
| 2205 |
"learning_rate": 2.357664889105687e-06,
|
| 2206 |
-
"loss": 0.
|
| 2207 |
"step": 30600
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
"epoch": 7.675,
|
| 2211 |
-
"grad_norm": 0.
|
| 2212 |
"learning_rate": 2.0333255530934903e-06,
|
| 2213 |
-
"loss": 0.
|
| 2214 |
"step": 30700
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 7.7,
|
| 2218 |
-
"grad_norm": 0.
|
| 2219 |
"learning_rate": 1.7328857612684267e-06,
|
| 2220 |
-
"loss": 0.
|
| 2221 |
"step": 30800
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
"epoch": 7.725,
|
| 2225 |
-
"grad_norm": 0.
|
| 2226 |
"learning_rate": 1.4563744706429517e-06,
|
| 2227 |
-
"loss": 0.
|
| 2228 |
"step": 30900
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
"epoch": 7.75,
|
| 2232 |
-
"grad_norm": 0.
|
| 2233 |
"learning_rate": 1.2038183319507957e-06,
|
| 2234 |
-
"loss": 0.
|
| 2235 |
"step": 31000
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
"epoch": 7.775,
|
| 2239 |
-
"grad_norm": 0.
|
| 2240 |
"learning_rate": 9.752416870782156e-07,
|
| 2241 |
-
"loss": 0.
|
| 2242 |
"step": 31100
|
| 2243 |
},
|
| 2244 |
{
|
| 2245 |
"epoch": 7.8,
|
| 2246 |
-
"grad_norm": 0.
|
| 2247 |
"learning_rate": 7.70666566718009e-07,
|
| 2248 |
-
"loss": 0.
|
| 2249 |
"step": 31200
|
| 2250 |
},
|
| 2251 |
{
|
| 2252 |
"epoch": 7.825,
|
| 2253 |
-
"grad_norm": 0.
|
| 2254 |
"learning_rate": 5.90112688246075e-07,
|
| 2255 |
-
"loss": 0.
|
| 2256 |
"step": 31300
|
| 2257 |
},
|
| 2258 |
{
|
| 2259 |
"epoch": 7.85,
|
| 2260 |
-
"grad_norm": 0.
|
| 2261 |
"learning_rate": 4.335974538210441e-07,
|
| 2262 |
-
"loss": 0.
|
| 2263 |
"step": 31400
|
| 2264 |
},
|
| 2265 |
{
|
| 2266 |
"epoch": 7.875,
|
| 2267 |
-
"grad_norm": 0.
|
| 2268 |
"learning_rate": 3.0113594870689873e-07,
|
| 2269 |
-
"loss": 0.
|
| 2270 |
"step": 31500
|
| 2271 |
},
|
| 2272 |
{
|
| 2273 |
"epoch": 7.9,
|
| 2274 |
-
"grad_norm": 0.
|
| 2275 |
"learning_rate": 1.9274093981927476e-07,
|
| 2276 |
-
"loss": 0.
|
| 2277 |
"step": 31600
|
| 2278 |
},
|
| 2279 |
{
|
| 2280 |
"epoch": 7.925,
|
| 2281 |
-
"grad_norm": 0.
|
| 2282 |
"learning_rate": 1.0842287449469579e-07,
|
| 2283 |
-
"loss": 0.
|
| 2284 |
"step": 31700
|
| 2285 |
},
|
| 2286 |
{
|
| 2287 |
"epoch": 7.95,
|
| 2288 |
-
"grad_norm": 0.
|
| 2289 |
"learning_rate": 4.818987948379538e-08,
|
| 2290 |
-
"loss": 0.
|
| 2291 |
"step": 31800
|
| 2292 |
},
|
| 2293 |
{
|
| 2294 |
"epoch": 7.975,
|
| 2295 |
-
"grad_norm": 0.
|
| 2296 |
"learning_rate": 1.2047760167999133e-08,
|
| 2297 |
-
"loss": 0.
|
| 2298 |
"step": 31900
|
| 2299 |
},
|
| 2300 |
{
|
| 2301 |
"epoch": 8.0,
|
| 2302 |
-
"grad_norm": 0.
|
| 2303 |
"learning_rate": 0.0,
|
| 2304 |
-
"loss": 0.
|
| 2305 |
"step": 32000
|
| 2306 |
}
|
| 2307 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.025,
|
| 13 |
+
"grad_norm": 0.20501533150672913,
|
| 14 |
"learning_rate": 0.000499987952239832,
|
| 15 |
+
"loss": 1.3493,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.05,
|
| 20 |
+
"grad_norm": 0.17424988746643066,
|
| 21 |
"learning_rate": 0.0004999518101205162,
|
| 22 |
+
"loss": 1.1647,
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.075,
|
| 27 |
+
"grad_norm": 0.24931852519512177,
|
| 28 |
"learning_rate": 0.0004998915771255053,
|
| 29 |
+
"loss": 1.1393,
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.1,
|
| 34 |
+
"grad_norm": 0.2063228040933609,
|
| 35 |
"learning_rate": 0.0004998072590601808,
|
| 36 |
+
"loss": 1.1057,
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.125,
|
| 41 |
+
"grad_norm": 0.23096567392349243,
|
| 42 |
"learning_rate": 0.0004996988640512931,
|
| 43 |
+
"loss": 1.1329,
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.15,
|
| 48 |
+
"grad_norm": 0.19355718791484833,
|
| 49 |
"learning_rate": 0.000499566402546179,
|
| 50 |
+
"loss": 1.1168,
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.175,
|
| 55 |
+
"grad_norm": 0.2221299558877945,
|
| 56 |
"learning_rate": 0.0004994098873117539,
|
| 57 |
+
"loss": 1.1316,
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.2,
|
| 62 |
+
"grad_norm": 0.2073143720626831,
|
| 63 |
"learning_rate": 0.000499229333433282,
|
| 64 |
+
"loss": 1.1141,
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.225,
|
| 69 |
+
"grad_norm": 0.21791169047355652,
|
| 70 |
"learning_rate": 0.0004990247583129218,
|
| 71 |
+
"loss": 1.1274,
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.25,
|
| 76 |
+
"grad_norm": 0.21467778086662292,
|
| 77 |
"learning_rate": 0.0004987961816680492,
|
| 78 |
"loss": 1.0698,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.275,
|
| 83 |
+
"grad_norm": 0.2048187404870987,
|
| 84 |
"learning_rate": 0.0004985436255293571,
|
| 85 |
+
"loss": 1.0928,
|
| 86 |
"step": 1100
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.3,
|
| 90 |
+
"grad_norm": 0.21932852268218994,
|
| 91 |
"learning_rate": 0.0004982671142387316,
|
| 92 |
"loss": 1.0906,
|
| 93 |
"step": 1200
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.325,
|
| 97 |
+
"grad_norm": 0.2301362156867981,
|
| 98 |
"learning_rate": 0.0004979666744469065,
|
| 99 |
+
"loss": 1.0949,
|
| 100 |
"step": 1300
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.35,
|
| 104 |
+
"grad_norm": 0.2186669409275055,
|
| 105 |
"learning_rate": 0.0004976423351108943,
|
| 106 |
+
"loss": 1.0857,
|
| 107 |
"step": 1400
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.375,
|
| 111 |
+
"grad_norm": 0.19886991381645203,
|
| 112 |
"learning_rate": 0.0004972941274911952,
|
| 113 |
+
"loss": 1.0642,
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.4,
|
| 118 |
+
"grad_norm": 0.23275479674339294,
|
| 119 |
"learning_rate": 0.0004969220851487844,
|
| 120 |
+
"loss": 1.0718,
|
| 121 |
"step": 1600
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.425,
|
| 125 |
+
"grad_norm": 0.21335655450820923,
|
| 126 |
"learning_rate": 0.0004965262439418772,
|
| 127 |
"loss": 1.082,
|
| 128 |
"step": 1700
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.45,
|
| 132 |
+
"grad_norm": 0.2357306331396103,
|
| 133 |
"learning_rate": 0.0004961066420224729,
|
| 134 |
"loss": 1.0937,
|
| 135 |
"step": 1800
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.475,
|
| 139 |
+
"grad_norm": 0.23492847383022308,
|
| 140 |
"learning_rate": 0.000495663319832678,
|
| 141 |
+
"loss": 1.0846,
|
| 142 |
"step": 1900
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.5,
|
| 146 |
+
"grad_norm": 0.2015356421470642,
|
| 147 |
"learning_rate": 0.0004951963201008077,
|
| 148 |
+
"loss": 1.0829,
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.525,
|
| 153 |
+
"grad_norm": 0.24353289604187012,
|
| 154 |
"learning_rate": 0.0004947056878372681,
|
| 155 |
+
"loss": 1.0688,
|
| 156 |
"step": 2100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.55,
|
| 160 |
+
"grad_norm": 0.27955085039138794,
|
| 161 |
"learning_rate": 0.0004941914703302181,
|
| 162 |
+
"loss": 1.0644,
|
| 163 |
"step": 2200
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.575,
|
| 167 |
+
"grad_norm": 0.24444086849689484,
|
| 168 |
"learning_rate": 0.0004936537171410112,
|
| 169 |
+
"loss": 1.0946,
|
| 170 |
"step": 2300
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.6,
|
| 174 |
+
"grad_norm": 0.22312547266483307,
|
| 175 |
"learning_rate": 0.0004930924800994192,
|
| 176 |
+
"loss": 1.0636,
|
| 177 |
"step": 2400
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.625,
|
| 181 |
+
"grad_norm": 0.22395829856395721,
|
| 182 |
"learning_rate": 0.000492507813298636,
|
| 183 |
"loss": 1.0858,
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.65,
|
| 188 |
+
"grad_norm": 0.2249772697687149,
|
| 189 |
"learning_rate": 0.0004918997730900649,
|
| 190 |
+
"loss": 1.0885,
|
| 191 |
"step": 2600
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.675,
|
| 195 |
+
"grad_norm": 0.22566111385822296,
|
| 196 |
"learning_rate": 0.0004912684180778869,
|
| 197 |
"loss": 1.0647,
|
| 198 |
"step": 2700
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.7,
|
| 202 |
+
"grad_norm": 0.256917804479599,
|
| 203 |
"learning_rate": 0.0004906138091134118,
|
| 204 |
"loss": 1.0484,
|
| 205 |
"step": 2800
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.725,
|
| 209 |
+
"grad_norm": 0.2598745822906494,
|
| 210 |
"learning_rate": 0.0004899360092892143,
|
| 211 |
+
"loss": 1.063,
|
| 212 |
"step": 2900
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.75,
|
| 216 |
+
"grad_norm": 0.2576007843017578,
|
| 217 |
"learning_rate": 0.0004892350839330522,
|
| 218 |
+
"loss": 1.0585,
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.775,
|
| 223 |
+
"grad_norm": 0.2577688992023468,
|
| 224 |
"learning_rate": 0.0004885111006015701,
|
| 225 |
+
"loss": 1.0452,
|
| 226 |
"step": 3100
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.8,
|
| 230 |
+
"grad_norm": 0.2334214746952057,
|
| 231 |
"learning_rate": 0.0004877641290737884,
|
| 232 |
+
"loss": 1.0473,
|
| 233 |
"step": 3200
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.825,
|
| 237 |
+
"grad_norm": 0.2869343161582947,
|
| 238 |
"learning_rate": 0.0004869942413443776,
|
| 239 |
+
"loss": 1.0412,
|
| 240 |
"step": 3300
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.85,
|
| 244 |
+
"grad_norm": 0.26476991176605225,
|
| 245 |
"learning_rate": 0.00048620151161671955,
|
| 246 |
+
"loss": 1.064,
|
| 247 |
"step": 3400
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.875,
|
| 251 |
+
"grad_norm": 0.2549433708190918,
|
| 252 |
"learning_rate": 0.0004853860162957552,
|
| 253 |
+
"loss": 1.0557,
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.9,
|
| 258 |
+
"grad_norm": 0.26020538806915283,
|
| 259 |
"learning_rate": 0.0004845478339806211,
|
| 260 |
+
"loss": 1.0424,
|
| 261 |
"step": 3600
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.925,
|
| 265 |
+
"grad_norm": 0.2770324945449829,
|
| 266 |
"learning_rate": 0.0004836870454570731,
|
| 267 |
+
"loss": 1.0552,
|
| 268 |
"step": 3700
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.95,
|
| 272 |
+
"grad_norm": 0.2561758756637573,
|
| 273 |
"learning_rate": 0.00048280373368970086,
|
| 274 |
+
"loss": 1.0336,
|
| 275 |
"step": 3800
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.975,
|
| 279 |
+
"grad_norm": 0.25469744205474854,
|
| 280 |
"learning_rate": 0.000481897983813931,
|
| 281 |
+
"loss": 1.06,
|
| 282 |
"step": 3900
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 1.0,
|
| 286 |
+
"grad_norm": 0.2685888707637787,
|
| 287 |
"learning_rate": 0.0004809698831278217,
|
| 288 |
+
"loss": 1.0151,
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
+
"eval_loss": 1.0406872034072876,
|
| 294 |
+
"eval_runtime": 152.4194,
|
| 295 |
+
"eval_samples_per_second": 26.243,
|
| 296 |
+
"eval_steps_per_second": 6.561,
|
| 297 |
"step": 4000
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 1.025,
|
| 301 |
+
"grad_norm": 0.28361451625823975,
|
| 302 |
"learning_rate": 0.00048001952108364876,
|
| 303 |
+
"loss": 1.0277,
|
| 304 |
"step": 4100
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 1.05,
|
| 308 |
+
"grad_norm": 0.26193392276763916,
|
| 309 |
"learning_rate": 0.00047904698927928404,
|
| 310 |
+
"loss": 1.0187,
|
| 311 |
"step": 4200
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 1.075,
|
| 315 |
+
"grad_norm": 0.2700245678424835,
|
| 316 |
"learning_rate": 0.0004780523814493669,
|
| 317 |
+
"loss": 1.0325,
|
| 318 |
"step": 4300
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 1.1,
|
| 322 |
+
"grad_norm": 0.2610750198364258,
|
| 323 |
"learning_rate": 0.00047703579345627036,
|
| 324 |
+
"loss": 1.0182,
|
| 325 |
"step": 4400
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 1.125,
|
| 329 |
+
"grad_norm": 0.2965630888938904,
|
| 330 |
"learning_rate": 0.0004759973232808609,
|
| 331 |
+
"loss": 1.0105,
|
| 332 |
"step": 4500
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 1.15,
|
| 336 |
+
"grad_norm": 0.25167468190193176,
|
| 337 |
"learning_rate": 0.0004749370710130554,
|
| 338 |
+
"loss": 1.0404,
|
| 339 |
"step": 4600
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 1.175,
|
| 343 |
+
"grad_norm": 0.26605212688446045,
|
| 344 |
"learning_rate": 0.0004738551388421742,
|
| 345 |
"loss": 1.0291,
|
| 346 |
"step": 4700
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 1.2,
|
| 350 |
+
"grad_norm": 0.25520554184913635,
|
| 351 |
"learning_rate": 0.00047275163104709196,
|
| 352 |
+
"loss": 1.019,
|
| 353 |
"step": 4800
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.225,
|
| 357 |
+
"grad_norm": 0.26057741045951843,
|
| 358 |
"learning_rate": 0.00047162665398618666,
|
| 359 |
+
"loss": 1.0309,
|
| 360 |
"step": 4900
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.25,
|
| 364 |
+
"grad_norm": 0.2987058758735657,
|
| 365 |
"learning_rate": 0.00047048031608708875,
|
| 366 |
+
"loss": 1.0127,
|
| 367 |
"step": 5000
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.275,
|
| 371 |
+
"grad_norm": 0.2619779706001282,
|
| 372 |
"learning_rate": 0.00046931272783623106,
|
| 373 |
+
"loss": 1.0191,
|
| 374 |
"step": 5100
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.3,
|
| 378 |
+
"grad_norm": 0.27090466022491455,
|
| 379 |
"learning_rate": 0.0004681240017681993,
|
| 380 |
+
"loss": 1.0416,
|
| 381 |
"step": 5200
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.325,
|
| 385 |
+
"grad_norm": 0.24344772100448608,
|
| 386 |
"learning_rate": 0.00046691425245488607,
|
| 387 |
+
"loss": 1.0075,
|
| 388 |
"step": 5300
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 1.35,
|
| 392 |
+
"grad_norm": 0.2597128450870514,
|
| 393 |
"learning_rate": 0.00046568359649444796,
|
| 394 |
+
"loss": 1.0147,
|
| 395 |
"step": 5400
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.375,
|
| 399 |
+
"grad_norm": 0.3321911692619324,
|
| 400 |
"learning_rate": 0.00046443215250006805,
|
| 401 |
+
"loss": 1.0277,
|
| 402 |
"step": 5500
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.4,
|
| 406 |
+
"grad_norm": 0.2736588716506958,
|
| 407 |
"learning_rate": 0.00046316004108852305,
|
| 408 |
+
"loss": 1.041,
|
| 409 |
"step": 5600
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 1.425,
|
| 413 |
+
"grad_norm": 0.30011600255966187,
|
| 414 |
"learning_rate": 0.0004618673848685586,
|
| 415 |
+
"loss": 1.006,
|
| 416 |
"step": 5700
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 1.45,
|
| 420 |
+
"grad_norm": 0.27296823263168335,
|
| 421 |
"learning_rate": 0.0004605543084290716,
|
| 422 |
+
"loss": 1.0207,
|
| 423 |
"step": 5800
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 1.475,
|
| 427 |
+
"grad_norm": 0.2965295910835266,
|
| 428 |
"learning_rate": 0.0004592209383271023,
|
| 429 |
+
"loss": 1.0306,
|
| 430 |
"step": 5900
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 1.5,
|
| 434 |
+
"grad_norm": 0.27299028635025024,
|
| 435 |
"learning_rate": 0.00045786740307563633,
|
| 436 |
+
"loss": 1.0296,
|
| 437 |
"step": 6000
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 1.525,
|
| 441 |
+
"grad_norm": 0.2956341803073883,
|
| 442 |
"learning_rate": 0.0004564938331312183,
|
| 443 |
+
"loss": 0.9965,
|
| 444 |
"step": 6100
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 1.55,
|
| 448 |
+
"grad_norm": 0.2554098665714264,
|
| 449 |
"learning_rate": 0.0004551003608813784,
|
| 450 |
+
"loss": 1.0232,
|
| 451 |
"step": 6200
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 1.575,
|
| 455 |
+
"grad_norm": 0.2712690234184265,
|
| 456 |
"learning_rate": 0.00045368712063187237,
|
| 457 |
+
"loss": 1.02,
|
| 458 |
"step": 6300
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 1.6,
|
| 462 |
+
"grad_norm": 0.31331920623779297,
|
| 463 |
"learning_rate": 0.0004522542485937369,
|
| 464 |
+
"loss": 0.971,
|
| 465 |
"step": 6400
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 1.625,
|
| 469 |
+
"grad_norm": 0.3378168046474457,
|
| 470 |
"learning_rate": 0.0004508018828701612,
|
| 471 |
+
"loss": 1.0017,
|
| 472 |
"step": 6500
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 1.65,
|
| 476 |
+
"grad_norm": 0.3126436471939087,
|
| 477 |
"learning_rate": 0.0004493301634431768,
|
| 478 |
+
"loss": 0.9969,
|
| 479 |
"step": 6600
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 1.675,
|
| 483 |
+
"grad_norm": 0.2890426218509674,
|
| 484 |
"learning_rate": 0.00044783923216016507,
|
| 485 |
+
"loss": 1.0082,
|
| 486 |
"step": 6700
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 1.7,
|
| 490 |
+
"grad_norm": 0.3303642272949219,
|
| 491 |
"learning_rate": 0.0004463292327201862,
|
| 492 |
+
"loss": 0.9989,
|
| 493 |
"step": 6800
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 1.725,
|
| 497 |
+
"grad_norm": 0.3025239408016205,
|
| 498 |
"learning_rate": 0.00044480031066012916,
|
| 499 |
+
"loss": 1.0305,
|
| 500 |
"step": 6900
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 1.75,
|
| 504 |
+
"grad_norm": 0.2934839725494385,
|
| 505 |
"learning_rate": 0.0004432526133406842,
|
| 506 |
+
"loss": 1.0003,
|
| 507 |
"step": 7000
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 1.775,
|
| 511 |
+
"grad_norm": 0.31030675768852234,
|
| 512 |
"learning_rate": 0.00044168628993214036,
|
| 513 |
+
"loss": 0.982,
|
| 514 |
"step": 7100
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 1.8,
|
| 518 |
+
"grad_norm": 0.30027347803115845,
|
| 519 |
"learning_rate": 0.0004401014914000078,
|
| 520 |
+
"loss": 0.9825,
|
| 521 |
"step": 7200
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 1.825,
|
| 525 |
+
"grad_norm": 0.2832087278366089,
|
| 526 |
"learning_rate": 0.00043849837049046735,
|
| 527 |
"loss": 1.007,
|
| 528 |
"step": 7300
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 1.85,
|
| 532 |
+
"grad_norm": 0.30503541231155396,
|
| 533 |
"learning_rate": 0.00043687708171564923,
|
| 534 |
+
"loss": 1.0023,
|
| 535 |
"step": 7400
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 1.875,
|
| 539 |
+
"grad_norm": 0.30458569526672363,
|
| 540 |
"learning_rate": 0.0004352377813387398,
|
| 541 |
+
"loss": 1.0006,
|
| 542 |
"step": 7500
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 1.9,
|
| 546 |
+
"grad_norm": 0.3471212089061737,
|
| 547 |
"learning_rate": 0.0004335806273589214,
|
| 548 |
+
"loss": 1.0239,
|
| 549 |
"step": 7600
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 1.925,
|
| 553 |
+
"grad_norm": 0.29310837388038635,
|
| 554 |
"learning_rate": 0.00043190577949614375,
|
| 555 |
+
"loss": 1.0129,
|
| 556 |
"step": 7700
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.95,
|
| 560 |
+
"grad_norm": 0.294300377368927,
|
| 561 |
"learning_rate": 0.0004302133991757297,
|
| 562 |
+
"loss": 0.9956,
|
| 563 |
"step": 7800
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 1.975,
|
| 567 |
+
"grad_norm": 0.3151143193244934,
|
| 568 |
"learning_rate": 0.00042850364951281707,
|
| 569 |
+
"loss": 1.0224,
|
| 570 |
"step": 7900
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 2.0,
|
| 574 |
+
"grad_norm": 0.3241264820098877,
|
| 575 |
"learning_rate": 0.00042677669529663686,
|
| 576 |
+
"loss": 1.0234,
|
| 577 |
"step": 8000
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 2.0,
|
| 581 |
+
"eval_loss": 1.0086950063705444,
|
| 582 |
+
"eval_runtime": 152.329,
|
| 583 |
+
"eval_samples_per_second": 26.259,
|
| 584 |
+
"eval_steps_per_second": 6.565,
|
| 585 |
"step": 8000
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 2.025,
|
| 589 |
+
"grad_norm": 0.3218720853328705,
|
| 590 |
"learning_rate": 0.0004250327029746309,
|
| 591 |
+
"loss": 1.0029,
|
| 592 |
"step": 8100
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 2.05,
|
| 596 |
+
"grad_norm": 0.31330084800720215,
|
| 597 |
"learning_rate": 0.000423271840636409,
|
| 598 |
+
"loss": 0.9844,
|
| 599 |
"step": 8200
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 2.075,
|
| 603 |
+
"grad_norm": 0.30206719040870667,
|
| 604 |
"learning_rate": 0.00042149427799754817,
|
| 605 |
+
"loss": 0.989,
|
| 606 |
"step": 8300
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 2.1,
|
| 610 |
+
"grad_norm": 0.303648978471756,
|
| 611 |
"learning_rate": 0.00041970018638323546,
|
| 612 |
+
"loss": 0.9543,
|
| 613 |
"step": 8400
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 2.125,
|
| 617 |
+
"grad_norm": 0.317037433385849,
|
| 618 |
"learning_rate": 0.00041788973871175465,
|
| 619 |
+
"loss": 0.9945,
|
| 620 |
"step": 8500
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 2.15,
|
| 624 |
+
"grad_norm": 0.3210434913635254,
|
| 625 |
"learning_rate": 0.00041606310947782046,
|
| 626 |
+
"loss": 0.988,
|
| 627 |
"step": 8600
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 2.175,
|
| 631 |
+
"grad_norm": 0.33763307332992554,
|
| 632 |
"learning_rate": 0.00041422047473576033,
|
| 633 |
+
"loss": 1.0066,
|
| 634 |
"step": 8700
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 2.2,
|
| 638 |
+
"grad_norm": 0.30534419417381287,
|
| 639 |
"learning_rate": 0.0004123620120825459,
|
| 640 |
+
"loss": 1.0033,
|
| 641 |
"step": 8800
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 2.225,
|
| 645 |
+
"grad_norm": 0.3363495171070099,
|
| 646 |
"learning_rate": 0.00041048790064067577,
|
| 647 |
+
"loss": 0.993,
|
| 648 |
"step": 8900
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 2.25,
|
| 652 |
+
"grad_norm": 0.2920531928539276,
|
| 653 |
"learning_rate": 0.0004085983210409114,
|
| 654 |
+
"loss": 0.9844,
|
| 655 |
"step": 9000
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 2.275,
|
| 659 |
+
"grad_norm": 0.33446577191352844,
|
| 660 |
"learning_rate": 0.0004066934554048674,
|
| 661 |
+
"loss": 0.9846,
|
| 662 |
"step": 9100
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 2.3,
|
| 666 |
+
"grad_norm": 0.33012858033180237,
|
| 667 |
"learning_rate": 0.00040477348732745853,
|
| 668 |
"loss": 0.9858,
|
| 669 |
"step": 9200
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 2.325,
|
| 673 |
+
"grad_norm": 0.3301050364971161,
|
| 674 |
"learning_rate": 0.0004028386018592041,
|
| 675 |
+
"loss": 0.9787,
|
| 676 |
"step": 9300
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 2.35,
|
| 680 |
+
"grad_norm": 0.3368571400642395,
|
| 681 |
"learning_rate": 0.0004008889854883929,
|
| 682 |
+
"loss": 0.9912,
|
| 683 |
"step": 9400
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 2.375,
|
| 687 |
+
"grad_norm": 0.285281777381897,
|
| 688 |
"learning_rate": 0.0003989248261231084,
|
| 689 |
+
"loss": 0.9655,
|
| 690 |
"step": 9500
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 2.4,
|
| 694 |
+
"grad_norm": 0.36711010336875916,
|
| 695 |
"learning_rate": 0.0003969463130731183,
|
| 696 |
+
"loss": 0.9753,
|
| 697 |
"step": 9600
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 2.425,
|
| 701 |
+
"grad_norm": 0.28881680965423584,
|
| 702 |
"learning_rate": 0.00039495363703162843,
|
| 703 |
"loss": 0.978,
|
| 704 |
"step": 9700
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 2.45,
|
| 708 |
+
"grad_norm": 0.3270690441131592,
|
| 709 |
"learning_rate": 0.000392946990056903,
|
| 710 |
+
"loss": 1.0242,
|
| 711 |
"step": 9800
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 2.475,
|
| 715 |
+
"grad_norm": 0.3338991105556488,
|
| 716 |
"learning_rate": 0.00039092656555375416,
|
| 717 |
+
"loss": 1.0066,
|
| 718 |
"step": 9900
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 2.5,
|
| 722 |
+
"grad_norm": 0.34082475304603577,
|
| 723 |
"learning_rate": 0.00038889255825490053,
|
| 724 |
+
"loss": 0.9781,
|
| 725 |
"step": 10000
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 2.525,
|
| 729 |
+
"grad_norm": 0.34000107645988464,
|
| 730 |
"learning_rate": 0.0003868451642021992,
|
| 731 |
"loss": 0.9745,
|
| 732 |
"step": 10100
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 2.55,
|
| 736 |
+
"grad_norm": 0.3570154309272766,
|
| 737 |
"learning_rate": 0.0003847845807277501,
|
| 738 |
+
"loss": 0.9759,
|
| 739 |
"step": 10200
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 2.575,
|
| 743 |
+
"grad_norm": 0.33699771761894226,
|
| 744 |
"learning_rate": 0.0003827110064348773,
|
| 745 |
+
"loss": 0.9882,
|
| 746 |
"step": 10300
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 2.6,
|
| 750 |
+
"grad_norm": 0.29268336296081543,
|
| 751 |
"learning_rate": 0.0003806246411789872,
|
| 752 |
+
"loss": 0.9745,
|
| 753 |
"step": 10400
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 2.625,
|
| 757 |
+
"grad_norm": 0.3088798224925995,
|
| 758 |
"learning_rate": 0.0003785256860483054,
|
| 759 |
+
"loss": 0.9501,
|
| 760 |
"step": 10500
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 2.65,
|
| 764 |
+
"grad_norm": 0.34749528765678406,
|
| 765 |
"learning_rate": 0.0003764143433444962,
|
| 766 |
+
"loss": 0.9702,
|
| 767 |
"step": 10600
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 2.675,
|
| 771 |
+
"grad_norm": 0.33005061745643616,
|
| 772 |
"learning_rate": 0.0003742908165631636,
|
| 773 |
+
"loss": 0.987,
|
| 774 |
"step": 10700
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 2.7,
|
| 778 |
+
"grad_norm": 0.3165997266769409,
|
| 779 |
"learning_rate": 0.0003721553103742388,
|
| 780 |
"loss": 1.0125,
|
| 781 |
"step": 10800
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 2.725,
|
| 785 |
+
"grad_norm": 0.3213465213775635,
|
| 786 |
"learning_rate": 0.0003700080306022528,
|
| 787 |
+
"loss": 0.9613,
|
| 788 |
"step": 10900
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 2.75,
|
| 792 |
+
"grad_norm": 0.33442971110343933,
|
| 793 |
"learning_rate": 0.0003678491842064995,
|
| 794 |
+
"loss": 0.9825,
|
| 795 |
"step": 11000
|
| 796 |
},
|
| 797 |
{
|
| 798 |
"epoch": 2.775,
|
| 799 |
+
"grad_norm": 0.2766638994216919,
|
| 800 |
"learning_rate": 0.00036567897926108756,
|
| 801 |
"loss": 0.9473,
|
| 802 |
"step": 11100
|
| 803 |
},
|
| 804 |
{
|
| 805 |
"epoch": 2.8,
|
| 806 |
+
"grad_norm": 0.3419225215911865,
|
| 807 |
"learning_rate": 0.00036349762493488667,
|
| 808 |
+
"loss": 0.9462,
|
| 809 |
"step": 11200
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 2.825,
|
| 813 |
+
"grad_norm": 0.3595256209373474,
|
| 814 |
"learning_rate": 0.0003613053314713671,
|
| 815 |
+
"loss": 0.9879,
|
| 816 |
"step": 11300
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 2.85,
|
| 820 |
+
"grad_norm": 0.3483733534812927,
|
| 821 |
"learning_rate": 0.0003591023101683355,
|
| 822 |
+
"loss": 0.9997,
|
| 823 |
"step": 11400
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 2.875,
|
| 827 |
+
"grad_norm": 0.32273510098457336,
|
| 828 |
"learning_rate": 0.0003568887733575705,
|
| 829 |
+
"loss": 1.0057,
|
| 830 |
"step": 11500
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 2.9,
|
| 834 |
+
"grad_norm": 0.31241312623023987,
|
| 835 |
"learning_rate": 0.00035466493438435703,
|
| 836 |
"loss": 0.9675,
|
| 837 |
"step": 11600
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 2.925,
|
| 841 |
+
"grad_norm": 0.37985849380493164,
|
| 842 |
"learning_rate": 0.0003524310075869239,
|
| 843 |
+
"loss": 0.9927,
|
| 844 |
"step": 11700
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 2.95,
|
| 848 |
+
"grad_norm": 0.33704909682273865,
|
| 849 |
"learning_rate": 0.0003501872082757852,
|
| 850 |
+
"loss": 0.9779,
|
| 851 |
"step": 11800
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 2.975,
|
| 855 |
+
"grad_norm": 0.4041268229484558,
|
| 856 |
"learning_rate": 0.000347933752712989,
|
| 857 |
+
"loss": 0.9739,
|
| 858 |
"step": 11900
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 3.0,
|
| 862 |
+
"grad_norm": 0.3415277302265167,
|
| 863 |
"learning_rate": 0.0003456708580912725,
|
| 864 |
"loss": 0.9995,
|
| 865 |
"step": 12000
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 3.0,
|
| 869 |
+
"eval_loss": 0.9921265244483948,
|
| 870 |
+
"eval_runtime": 153.1095,
|
| 871 |
+
"eval_samples_per_second": 26.125,
|
| 872 |
+
"eval_steps_per_second": 6.531,
|
| 873 |
"step": 12000
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 3.025,
|
| 877 |
+
"grad_norm": 0.3260078430175781,
|
| 878 |
"learning_rate": 0.0003433987425131291,
|
| 879 |
+
"loss": 0.969,
|
| 880 |
"step": 12100
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 3.05,
|
| 884 |
+
"grad_norm": 0.3035444915294647,
|
| 885 |
"learning_rate": 0.0003411176249697875,
|
| 886 |
+
"loss": 0.9636,
|
| 887 |
"step": 12200
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 3.075,
|
| 891 |
+
"grad_norm": 0.3195238411426544,
|
| 892 |
"learning_rate": 0.00033882772532010404,
|
| 893 |
+
"loss": 0.9522,
|
| 894 |
"step": 12300
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 3.1,
|
| 898 |
+
"grad_norm": 0.33937081694602966,
|
| 899 |
"learning_rate": 0.0003365292642693733,
|
| 900 |
+
"loss": 0.9425,
|
| 901 |
"step": 12400
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 3.125,
|
| 905 |
+
"grad_norm": 0.37307849526405334,
|
| 906 |
"learning_rate": 0.00033422246334805503,
|
| 907 |
+
"loss": 0.9683,
|
| 908 |
"step": 12500
|
| 909 |
},
|
| 910 |
{
|
| 911 |
"epoch": 3.15,
|
| 912 |
+
"grad_norm": 0.3180018365383148,
|
| 913 |
"learning_rate": 0.0003319075448904234,
|
| 914 |
+
"loss": 0.9761,
|
| 915 |
"step": 12600
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 3.175,
|
| 919 |
+
"grad_norm": 0.3785949647426605,
|
| 920 |
"learning_rate": 0.00032958473201313745,
|
| 921 |
+
"loss": 0.9748,
|
| 922 |
"step": 12700
|
| 923 |
},
|
| 924 |
{
|
| 925 |
"epoch": 3.2,
|
| 926 |
+
"grad_norm": 0.3146439790725708,
|
| 927 |
"learning_rate": 0.00032725424859373687,
|
| 928 |
+
"loss": 0.9657,
|
| 929 |
"step": 12800
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 3.225,
|
| 933 |
+
"grad_norm": 0.3667052686214447,
|
| 934 |
"learning_rate": 0.00032491631924906416,
|
| 935 |
+
"loss": 0.9685,
|
| 936 |
"step": 12900
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 3.25,
|
| 940 |
+
"grad_norm": 0.34998300671577454,
|
| 941 |
"learning_rate": 0.00032257116931361555,
|
| 942 |
+
"loss": 0.9821,
|
| 943 |
"step": 13000
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 3.275,
|
| 947 |
+
"grad_norm": 0.34160691499710083,
|
| 948 |
"learning_rate": 0.00032021902481782304,
|
| 949 |
"loss": 0.9536,
|
| 950 |
"step": 13100
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 3.3,
|
| 954 |
+
"grad_norm": 0.35554495453834534,
|
| 955 |
"learning_rate": 0.00031786011246626855,
|
| 956 |
+
"loss": 0.9626,
|
| 957 |
"step": 13200
|
| 958 |
},
|
| 959 |
{
|
| 960 |
"epoch": 3.325,
|
| 961 |
+
"grad_norm": 0.373349666595459,
|
| 962 |
"learning_rate": 0.0003154946596158343,
|
| 963 |
+
"loss": 0.9926,
|
| 964 |
"step": 13300
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 3.35,
|
| 968 |
+
"grad_norm": 0.38320091366767883,
|
| 969 |
"learning_rate": 0.0003131228942537895,
|
| 970 |
"loss": 0.9787,
|
| 971 |
"step": 13400
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 3.375,
|
| 975 |
+
"grad_norm": 0.3836865723133087,
|
| 976 |
"learning_rate": 0.000310745044975816,
|
| 977 |
+
"loss": 0.9583,
|
| 978 |
"step": 13500
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 3.4,
|
| 982 |
+
"grad_norm": 0.3358447551727295,
|
| 983 |
"learning_rate": 0.0003083613409639764,
|
| 984 |
+
"loss": 0.953,
|
| 985 |
"step": 13600
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 3.425,
|
| 989 |
+
"grad_norm": 0.3501192629337311,
|
| 990 |
"learning_rate": 0.00030597201196462466,
|
| 991 |
+
"loss": 0.9346,
|
| 992 |
"step": 13700
|
| 993 |
},
|
| 994 |
{
|
| 995 |
"epoch": 3.45,
|
| 996 |
+
"grad_norm": 0.3595263957977295,
|
| 997 |
"learning_rate": 0.00030357728826626266,
|
| 998 |
+
"loss": 0.9656,
|
| 999 |
"step": 13800
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 3.475,
|
| 1003 |
+
"grad_norm": 0.3395259380340576,
|
| 1004 |
"learning_rate": 0.00030117740067734495,
|
| 1005 |
"loss": 0.9835,
|
| 1006 |
"step": 13900
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
"epoch": 3.5,
|
| 1010 |
+
"grad_norm": 0.36751946806907654,
|
| 1011 |
"learning_rate": 0.0002987725805040321,
|
| 1012 |
+
"loss": 0.9465,
|
| 1013 |
"step": 14000
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
"epoch": 3.525,
|
| 1017 |
+
"grad_norm": 0.348893404006958,
|
| 1018 |
"learning_rate": 0.0002963630595278977,
|
| 1019 |
+
"loss": 0.9564,
|
| 1020 |
"step": 14100
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
"epoch": 3.55,
|
| 1024 |
+
"grad_norm": 0.39346957206726074,
|
| 1025 |
"learning_rate": 0.0002939490699835887,
|
| 1026 |
+
"loss": 0.9508,
|
| 1027 |
"step": 14200
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
"epoch": 3.575,
|
| 1031 |
+
"grad_norm": 0.33249151706695557,
|
| 1032 |
"learning_rate": 0.00029153084453644135,
|
| 1033 |
+
"loss": 0.9757,
|
| 1034 |
"step": 14300
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 3.6,
|
| 1038 |
+
"grad_norm": 0.3582550585269928,
|
| 1039 |
"learning_rate": 0.00028910861626005774,
|
| 1040 |
+
"loss": 0.9552,
|
| 1041 |
"step": 14400
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 3.625,
|
| 1045 |
+
"grad_norm": 0.3562643229961395,
|
| 1046 |
"learning_rate": 0.00028668261861384045,
|
| 1047 |
+
"loss": 0.95,
|
| 1048 |
"step": 14500
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
"epoch": 3.65,
|
| 1052 |
+
"grad_norm": 0.3955428898334503,
|
| 1053 |
"learning_rate": 0.00028425308542049207,
|
| 1054 |
+
"loss": 0.9699,
|
| 1055 |
"step": 14600
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
"epoch": 3.675,
|
| 1059 |
+
"grad_norm": 0.38744592666625977,
|
| 1060 |
"learning_rate": 0.0002818202508434783,
|
| 1061 |
+
"loss": 0.9838,
|
| 1062 |
"step": 14700
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 3.7,
|
| 1066 |
+
"grad_norm": 0.404632031917572,
|
| 1067 |
"learning_rate": 0.00027938434936445943,
|
| 1068 |
+
"loss": 0.9589,
|
| 1069 |
"step": 14800
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
"epoch": 3.725,
|
| 1073 |
+
"grad_norm": 0.36033356189727783,
|
| 1074 |
"learning_rate": 0.00027694561576068985,
|
| 1075 |
+
"loss": 0.973,
|
| 1076 |
"step": 14900
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
"epoch": 3.75,
|
| 1080 |
+
"grad_norm": 0.3457838296890259,
|
| 1081 |
"learning_rate": 0.0002745042850823902,
|
| 1082 |
+
"loss": 0.9801,
|
| 1083 |
"step": 15000
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
"epoch": 3.775,
|
| 1087 |
+
"grad_norm": 0.35734304785728455,
|
| 1088 |
"learning_rate": 0.00027206059263009243,
|
| 1089 |
+
"loss": 0.9469,
|
| 1090 |
"step": 15100
|
| 1091 |
},
|
| 1092 |
{
|
| 1093 |
"epoch": 3.8,
|
| 1094 |
+
"grad_norm": 0.3542312681674957,
|
| 1095 |
"learning_rate": 0.00026961477393196127,
|
| 1096 |
+
"loss": 0.9336,
|
| 1097 |
"step": 15200
|
| 1098 |
},
|
| 1099 |
{
|
| 1100 |
"epoch": 3.825,
|
| 1101 |
+
"grad_norm": 0.3671669065952301,
|
| 1102 |
"learning_rate": 0.0002671670647210934,
|
| 1103 |
+
"loss": 0.9793,
|
| 1104 |
"step": 15300
|
| 1105 |
},
|
| 1106 |
{
|
| 1107 |
"epoch": 3.85,
|
| 1108 |
+
"grad_norm": 0.357881098985672,
|
| 1109 |
"learning_rate": 0.00026471770091279724,
|
| 1110 |
+
"loss": 0.9544,
|
| 1111 |
"step": 15400
|
| 1112 |
},
|
| 1113 |
{
|
| 1114 |
"epoch": 3.875,
|
| 1115 |
+
"grad_norm": 0.31071895360946655,
|
| 1116 |
"learning_rate": 0.00026226691858185456,
|
| 1117 |
+
"loss": 0.9656,
|
| 1118 |
"step": 15500
|
| 1119 |
},
|
| 1120 |
{
|
| 1121 |
"epoch": 3.9,
|
| 1122 |
+
"grad_norm": 0.34388652443885803,
|
| 1123 |
"learning_rate": 0.00025981495393976716,
|
| 1124 |
+
"loss": 0.96,
|
| 1125 |
"step": 15600
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
"epoch": 3.925,
|
| 1129 |
+
"grad_norm": 0.3791946768760681,
|
| 1130 |
"learning_rate": 0.00025736204331199084,
|
| 1131 |
+
"loss": 0.9797,
|
| 1132 |
"step": 15700
|
| 1133 |
},
|
| 1134 |
{
|
| 1135 |
"epoch": 3.95,
|
| 1136 |
+
"grad_norm": 0.3630838096141815,
|
| 1137 |
"learning_rate": 0.00025490842311515704,
|
| 1138 |
+
"loss": 0.9592,
|
| 1139 |
"step": 15800
|
| 1140 |
},
|
| 1141 |
{
|
| 1142 |
"epoch": 3.975,
|
| 1143 |
+
"grad_norm": 0.3712921440601349,
|
| 1144 |
"learning_rate": 0.0002524543298342875,
|
| 1145 |
+
"loss": 0.9731,
|
| 1146 |
"step": 15900
|
| 1147 |
},
|
| 1148 |
{
|
| 1149 |
"epoch": 4.0,
|
| 1150 |
+
"grad_norm": 0.4159054756164551,
|
| 1151 |
"learning_rate": 0.00025,
|
| 1152 |
+
"loss": 0.9528,
|
| 1153 |
"step": 16000
|
| 1154 |
},
|
| 1155 |
{
|
| 1156 |
"epoch": 4.0,
|
| 1157 |
+
"eval_loss": 0.9823833107948303,
|
| 1158 |
+
"eval_runtime": 152.8782,
|
| 1159 |
+
"eval_samples_per_second": 26.165,
|
| 1160 |
+
"eval_steps_per_second": 6.541,
|
| 1161 |
"step": 16000
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 4.025,
|
| 1165 |
+
"grad_norm": 0.39106351137161255,
|
| 1166 |
"learning_rate": 0.0002475456701657126,
|
| 1167 |
+
"loss": 0.9814,
|
| 1168 |
"step": 16100
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 4.05,
|
| 1172 |
+
"grad_norm": 0.36192214488983154,
|
| 1173 |
"learning_rate": 0.00024509157688484297,
|
| 1174 |
+
"loss": 0.9369,
|
| 1175 |
"step": 16200
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 4.075,
|
| 1179 |
+
"grad_norm": 0.3404277563095093,
|
| 1180 |
"learning_rate": 0.0002426379566880092,
|
| 1181 |
+
"loss": 0.9501,
|
| 1182 |
"step": 16300
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 4.1,
|
| 1186 |
+
"grad_norm": 0.3791425824165344,
|
| 1187 |
"learning_rate": 0.00024018504606023293,
|
| 1188 |
+
"loss": 0.9336,
|
| 1189 |
"step": 16400
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 4.125,
|
| 1193 |
+
"grad_norm": 0.4009718894958496,
|
| 1194 |
"learning_rate": 0.0002377330814181455,
|
| 1195 |
+
"loss": 0.9487,
|
| 1196 |
"step": 16500
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 4.15,
|
| 1200 |
+
"grad_norm": 0.37880653142929077,
|
| 1201 |
"learning_rate": 0.00023528229908720272,
|
| 1202 |
+
"loss": 0.9542,
|
| 1203 |
"step": 16600
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 4.175,
|
| 1207 |
+
"grad_norm": 0.3936653733253479,
|
| 1208 |
"learning_rate": 0.00023283293527890658,
|
| 1209 |
+
"loss": 0.9388,
|
| 1210 |
"step": 16700
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 4.2,
|
| 1214 |
+
"grad_norm": 0.33316799998283386,
|
| 1215 |
"learning_rate": 0.0002303852260680388,
|
| 1216 |
+
"loss": 0.943,
|
| 1217 |
"step": 16800
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 4.225,
|
| 1221 |
+
"grad_norm": 0.3236179053783417,
|
| 1222 |
"learning_rate": 0.00022793940736990766,
|
| 1223 |
+
"loss": 0.9044,
|
| 1224 |
"step": 16900
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 4.25,
|
| 1228 |
+
"grad_norm": 0.3429449200630188,
|
| 1229 |
"learning_rate": 0.00022549571491760985,
|
| 1230 |
+
"loss": 0.9624,
|
| 1231 |
"step": 17000
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 4.275,
|
| 1235 |
+
"grad_norm": 0.35557985305786133,
|
| 1236 |
"learning_rate": 0.00022305438423931017,
|
| 1237 |
+
"loss": 0.9226,
|
| 1238 |
"step": 17100
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 4.3,
|
| 1242 |
+
"grad_norm": 1.0042773485183716,
|
| 1243 |
"learning_rate": 0.00022061565063554063,
|
| 1244 |
+
"loss": 0.9472,
|
| 1245 |
"step": 17200
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 4.325,
|
| 1249 |
+
"grad_norm": 0.3724925220012665,
|
| 1250 |
"learning_rate": 0.00021817974915652172,
|
| 1251 |
+
"loss": 0.947,
|
| 1252 |
"step": 17300
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 4.35,
|
| 1256 |
+
"grad_norm": 0.3926277756690979,
|
| 1257 |
"learning_rate": 0.00021574691457950805,
|
| 1258 |
+
"loss": 0.9478,
|
| 1259 |
"step": 17400
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 4.375,
|
| 1263 |
+
"grad_norm": 0.38053977489471436,
|
| 1264 |
"learning_rate": 0.00021331738138615958,
|
| 1265 |
+
"loss": 0.9801,
|
| 1266 |
"step": 17500
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 4.4,
|
| 1270 |
+
"grad_norm": 0.36115145683288574,
|
| 1271 |
"learning_rate": 0.00021089138373994224,
|
| 1272 |
+
"loss": 0.9619,
|
| 1273 |
"step": 17600
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 4.425,
|
| 1277 |
+
"grad_norm": 0.35729482769966125,
|
| 1278 |
"learning_rate": 0.0002084691554635587,
|
| 1279 |
+
"loss": 0.9549,
|
| 1280 |
"step": 17700
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 4.45,
|
| 1284 |
+
"grad_norm": 0.3762321174144745,
|
| 1285 |
"learning_rate": 0.00020605093001641137,
|
| 1286 |
+
"loss": 0.9359,
|
| 1287 |
"step": 17800
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 4.475,
|
| 1291 |
+
"grad_norm": 0.3317940831184387,
|
| 1292 |
"learning_rate": 0.00020363694047210228,
|
| 1293 |
+
"loss": 0.9571,
|
| 1294 |
"step": 17900
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 4.5,
|
| 1298 |
+
"grad_norm": 0.3733656406402588,
|
| 1299 |
"learning_rate": 0.00020122741949596797,
|
| 1300 |
+
"loss": 0.9409,
|
| 1301 |
"step": 18000
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 4.525,
|
| 1305 |
+
"grad_norm": 0.3790251612663269,
|
| 1306 |
"learning_rate": 0.00019882259932265512,
|
| 1307 |
+
"loss": 0.9768,
|
| 1308 |
"step": 18100
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 4.55,
|
| 1312 |
+
"grad_norm": 0.37900087237358093,
|
| 1313 |
"learning_rate": 0.00019642271173373735,
|
| 1314 |
+
"loss": 0.9208,
|
| 1315 |
"step": 18200
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 4.575,
|
| 1319 |
+
"grad_norm": 0.34171342849731445,
|
| 1320 |
"learning_rate": 0.00019402798803537538,
|
| 1321 |
+
"loss": 0.9831,
|
| 1322 |
"step": 18300
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 4.6,
|
| 1326 |
+
"grad_norm": 0.39172932505607605,
|
| 1327 |
"learning_rate": 0.00019163865903602372,
|
| 1328 |
"loss": 0.941,
|
| 1329 |
"step": 18400
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 4.625,
|
| 1333 |
+
"grad_norm": 0.36279597878456116,
|
| 1334 |
"learning_rate": 0.00018925495502418406,
|
| 1335 |
+
"loss": 0.9511,
|
| 1336 |
"step": 18500
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 4.65,
|
| 1340 |
+
"grad_norm": 0.35040661692619324,
|
| 1341 |
"learning_rate": 0.00018687710574621051,
|
| 1342 |
+
"loss": 0.9713,
|
| 1343 |
"step": 18600
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 4.675,
|
| 1347 |
+
"grad_norm": 0.35176095366477966,
|
| 1348 |
"learning_rate": 0.00018450534038416566,
|
| 1349 |
+
"loss": 0.9227,
|
| 1350 |
"step": 18700
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 4.7,
|
| 1354 |
+
"grad_norm": 0.41925740242004395,
|
| 1355 |
"learning_rate": 0.00018213988753373146,
|
| 1356 |
+
"loss": 0.9454,
|
| 1357 |
"step": 18800
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 4.725,
|
| 1361 |
+
"grad_norm": 0.4147469699382782,
|
| 1362 |
"learning_rate": 0.00017978097518217702,
|
| 1363 |
+
"loss": 0.9783,
|
| 1364 |
"step": 18900
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 4.75,
|
| 1368 |
+
"grad_norm": 0.364629864692688,
|
| 1369 |
"learning_rate": 0.00017742883068638446,
|
| 1370 |
+
"loss": 0.963,
|
| 1371 |
"step": 19000
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 4.775,
|
| 1375 |
+
"grad_norm": 0.39363783597946167,
|
| 1376 |
"learning_rate": 0.00017508368075093582,
|
| 1377 |
+
"loss": 0.9175,
|
| 1378 |
"step": 19100
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 4.8,
|
| 1382 |
+
"grad_norm": 0.39291703701019287,
|
| 1383 |
"learning_rate": 0.00017274575140626317,
|
| 1384 |
+
"loss": 0.9616,
|
| 1385 |
"step": 19200
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 4.825,
|
| 1389 |
+
"grad_norm": 0.3810083866119385,
|
| 1390 |
"learning_rate": 0.0001704152679868626,
|
| 1391 |
+
"loss": 0.9754,
|
| 1392 |
"step": 19300
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 4.85,
|
| 1396 |
+
"grad_norm": 0.36352667212486267,
|
| 1397 |
"learning_rate": 0.00016809245510957666,
|
| 1398 |
+
"loss": 0.953,
|
| 1399 |
"step": 19400
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 4.875,
|
| 1403 |
+
"grad_norm": 0.31296002864837646,
|
| 1404 |
"learning_rate": 0.000165777536651945,
|
| 1405 |
+
"loss": 0.9478,
|
| 1406 |
"step": 19500
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 4.9,
|
| 1410 |
+
"grad_norm": 0.4004424214363098,
|
| 1411 |
"learning_rate": 0.0001634707357306267,
|
| 1412 |
+
"loss": 0.9508,
|
| 1413 |
"step": 19600
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 4.925,
|
| 1417 |
+
"grad_norm": 0.3690031170845032,
|
| 1418 |
"learning_rate": 0.00016117227467989602,
|
| 1419 |
+
"loss": 0.9456,
|
| 1420 |
"step": 19700
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 4.95,
|
| 1424 |
+
"grad_norm": 0.3946553170681,
|
| 1425 |
"learning_rate": 0.0001588823750302126,
|
| 1426 |
+
"loss": 0.95,
|
| 1427 |
"step": 19800
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 4.975,
|
| 1431 |
+
"grad_norm": 0.39817288517951965,
|
| 1432 |
"learning_rate": 0.00015660125748687094,
|
| 1433 |
+
"loss": 0.9431,
|
| 1434 |
"step": 19900
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 5.0,
|
| 1438 |
+
"grad_norm": 0.37246620655059814,
|
| 1439 |
"learning_rate": 0.00015432914190872756,
|
| 1440 |
+
"loss": 0.9353,
|
| 1441 |
"step": 20000
|
| 1442 |
},
|
| 1443 |
{
|
| 1444 |
"epoch": 5.0,
|
| 1445 |
+
"eval_loss": 0.9754775762557983,
|
| 1446 |
+
"eval_runtime": 152.4185,
|
| 1447 |
+
"eval_samples_per_second": 26.244,
|
| 1448 |
+
"eval_steps_per_second": 6.561,
|
| 1449 |
"step": 20000
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 5.025,
|
| 1453 |
+
"grad_norm": 0.36520034074783325,
|
| 1454 |
"learning_rate": 0.000152066247287011,
|
| 1455 |
+
"loss": 0.9377,
|
| 1456 |
"step": 20100
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 5.05,
|
| 1460 |
+
"grad_norm": 0.3535405397415161,
|
| 1461 |
"learning_rate": 0.00014981279172421482,
|
| 1462 |
"loss": 0.9414,
|
| 1463 |
"step": 20200
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 5.075,
|
| 1467 |
+
"grad_norm": 0.35102036595344543,
|
| 1468 |
"learning_rate": 0.00014756899241307614,
|
| 1469 |
+
"loss": 0.9164,
|
| 1470 |
"step": 20300
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 5.1,
|
| 1474 |
+
"grad_norm": 0.3539367616176605,
|
| 1475 |
"learning_rate": 0.00014533506561564306,
|
| 1476 |
+
"loss": 0.9187,
|
| 1477 |
"step": 20400
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 5.125,
|
| 1481 |
+
"grad_norm": 0.347476601600647,
|
| 1482 |
"learning_rate": 0.00014311122664242953,
|
| 1483 |
+
"loss": 0.9196,
|
| 1484 |
"step": 20500
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 5.15,
|
| 1488 |
+
"grad_norm": 0.3902195990085602,
|
| 1489 |
"learning_rate": 0.00014089768983166444,
|
| 1490 |
+
"loss": 0.9301,
|
| 1491 |
"step": 20600
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 5.175,
|
| 1495 |
+
"grad_norm": 0.3888334631919861,
|
| 1496 |
"learning_rate": 0.000138694668528633,
|
| 1497 |
+
"loss": 0.9241,
|
| 1498 |
"step": 20700
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 5.2,
|
| 1502 |
+
"grad_norm": 0.41042113304138184,
|
| 1503 |
"learning_rate": 0.00013650237506511331,
|
| 1504 |
+
"loss": 0.9504,
|
| 1505 |
"step": 20800
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 5.225,
|
| 1509 |
+
"grad_norm": 0.3549586236476898,
|
| 1510 |
"learning_rate": 0.0001343210207389125,
|
| 1511 |
+
"loss": 0.955,
|
| 1512 |
"step": 20900
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 5.25,
|
| 1516 |
+
"grad_norm": 0.34020912647247314,
|
| 1517 |
"learning_rate": 0.00013215081579350058,
|
| 1518 |
+
"loss": 0.9471,
|
| 1519 |
"step": 21000
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 5.275,
|
| 1523 |
+
"grad_norm": 0.33093366026878357,
|
| 1524 |
"learning_rate": 0.00012999196939774722,
|
| 1525 |
+
"loss": 0.9351,
|
| 1526 |
"step": 21100
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 5.3,
|
| 1530 |
+
"grad_norm": 0.4015860855579376,
|
| 1531 |
"learning_rate": 0.00012784468962576134,
|
| 1532 |
+
"loss": 0.9491,
|
| 1533 |
"step": 21200
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 5.325,
|
| 1537 |
+
"grad_norm": 0.37603822350502014,
|
| 1538 |
"learning_rate": 0.00012570918343683636,
|
| 1539 |
+
"loss": 0.958,
|
| 1540 |
"step": 21300
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 5.35,
|
| 1544 |
+
"grad_norm": 0.39434927701950073,
|
| 1545 |
"learning_rate": 0.0001235856566555039,
|
| 1546 |
+
"loss": 0.94,
|
| 1547 |
"step": 21400
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 5.375,
|
| 1551 |
+
"grad_norm": 0.43162479996681213,
|
| 1552 |
"learning_rate": 0.0001214743139516946,
|
| 1553 |
+
"loss": 0.9291,
|
| 1554 |
"step": 21500
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 5.4,
|
| 1558 |
+
"grad_norm": 0.38297414779663086,
|
| 1559 |
"learning_rate": 0.00011937535882101281,
|
| 1560 |
+
"loss": 0.9494,
|
| 1561 |
"step": 21600
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 5.425,
|
| 1565 |
+
"grad_norm": 0.40226617455482483,
|
| 1566 |
"learning_rate": 0.00011728899356512265,
|
| 1567 |
+
"loss": 0.9067,
|
| 1568 |
"step": 21700
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 5.45,
|
| 1572 |
+
"grad_norm": 0.38879096508026123,
|
| 1573 |
"learning_rate": 0.00011521541927224994,
|
| 1574 |
+
"loss": 0.949,
|
| 1575 |
"step": 21800
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 5.475,
|
| 1579 |
+
"grad_norm": 0.33627960085868835,
|
| 1580 |
"learning_rate": 0.00011315483579780094,
|
| 1581 |
+
"loss": 0.9358,
|
| 1582 |
"step": 21900
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 5.5,
|
| 1586 |
+
"grad_norm": 0.4547683596611023,
|
| 1587 |
"learning_rate": 0.00011110744174509952,
|
| 1588 |
+
"loss": 0.9179,
|
| 1589 |
"step": 22000
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 5.525,
|
| 1593 |
+
"grad_norm": 0.3883083462715149,
|
| 1594 |
"learning_rate": 0.00010907343444624579,
|
| 1595 |
+
"loss": 0.9662,
|
| 1596 |
"step": 22100
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 5.55,
|
| 1600 |
+
"grad_norm": 0.4168870449066162,
|
| 1601 |
"learning_rate": 0.00010705300994309697,
|
| 1602 |
"loss": 0.9286,
|
| 1603 |
"step": 22200
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 5.575,
|
| 1607 |
+
"grad_norm": 0.3810732066631317,
|
| 1608 |
"learning_rate": 0.00010504636296837161,
|
| 1609 |
+
"loss": 0.9443,
|
| 1610 |
"step": 22300
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 5.6,
|
| 1614 |
+
"grad_norm": 0.40432223677635193,
|
| 1615 |
"learning_rate": 0.00010305368692688174,
|
| 1616 |
+
"loss": 0.9138,
|
| 1617 |
"step": 22400
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 5.625,
|
| 1621 |
+
"grad_norm": 0.3702322542667389,
|
| 1622 |
"learning_rate": 0.00010107517387689166,
|
| 1623 |
+
"loss": 0.9474,
|
| 1624 |
"step": 22500
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 5.65,
|
| 1628 |
+
"grad_norm": 0.40569931268692017,
|
| 1629 |
"learning_rate": 9.911101451160715e-05,
|
| 1630 |
+
"loss": 0.9645,
|
| 1631 |
"step": 22600
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 5.675,
|
| 1635 |
+
"grad_norm": 0.39476191997528076,
|
| 1636 |
"learning_rate": 9.716139814079594e-05,
|
| 1637 |
+
"loss": 0.9478,
|
| 1638 |
"step": 22700
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 5.7,
|
| 1642 |
+
"grad_norm": 0.347662091255188,
|
| 1643 |
"learning_rate": 9.522651267254148e-05,
|
| 1644 |
+
"loss": 0.9627,
|
| 1645 |
"step": 22800
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
"epoch": 5.725,
|
| 1649 |
+
"grad_norm": 0.36813852190971375,
|
| 1650 |
"learning_rate": 9.330654459513265e-05,
|
| 1651 |
+
"loss": 0.9328,
|
| 1652 |
"step": 22900
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
"epoch": 5.75,
|
| 1656 |
+
"grad_norm": 0.32601600885391235,
|
| 1657 |
"learning_rate": 9.140167895908866e-05,
|
| 1658 |
+
"loss": 0.9547,
|
| 1659 |
"step": 23000
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
"epoch": 5.775,
|
| 1663 |
+
"grad_norm": 0.40840694308280945,
|
| 1664 |
"learning_rate": 8.951209935932425e-05,
|
| 1665 |
"loss": 0.9113,
|
| 1666 |
"step": 23100
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
"epoch": 5.8,
|
| 1670 |
+
"grad_norm": 0.38472625613212585,
|
| 1671 |
"learning_rate": 8.763798791745412e-05,
|
| 1672 |
+
"loss": 0.9667,
|
| 1673 |
"step": 23200
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
"epoch": 5.825,
|
| 1677 |
+
"grad_norm": 0.37892642617225647,
|
| 1678 |
"learning_rate": 8.577952526423969e-05,
|
| 1679 |
+
"loss": 0.9241,
|
| 1680 |
"step": 23300
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
"epoch": 5.85,
|
| 1684 |
+
"grad_norm": 0.3874419927597046,
|
| 1685 |
"learning_rate": 8.393689052217964e-05,
|
| 1686 |
+
"loss": 0.9539,
|
| 1687 |
"step": 23400
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
"epoch": 5.875,
|
| 1691 |
+
"grad_norm": 0.37583285570144653,
|
| 1692 |
"learning_rate": 8.211026128824539e-05,
|
| 1693 |
+
"loss": 0.9411,
|
| 1694 |
"step": 23500
|
| 1695 |
},
|
| 1696 |
{
|
| 1697 |
"epoch": 5.9,
|
| 1698 |
+
"grad_norm": 0.37817347049713135,
|
| 1699 |
"learning_rate": 8.029981361676455e-05,
|
| 1700 |
+
"loss": 0.9426,
|
| 1701 |
"step": 23600
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 5.925,
|
| 1705 |
+
"grad_norm": 0.32163432240486145,
|
| 1706 |
"learning_rate": 7.850572200245185e-05,
|
| 1707 |
+
"loss": 0.9512,
|
| 1708 |
"step": 23700
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
"epoch": 5.95,
|
| 1712 |
+
"grad_norm": 0.37549078464508057,
|
| 1713 |
"learning_rate": 7.672815936359106e-05,
|
| 1714 |
+
"loss": 0.9694,
|
| 1715 |
"step": 23800
|
| 1716 |
},
|
| 1717 |
{
|
| 1718 |
"epoch": 5.975,
|
| 1719 |
+
"grad_norm": 0.43039214611053467,
|
| 1720 |
"learning_rate": 7.496729702536912e-05,
|
| 1721 |
+
"loss": 0.9191,
|
| 1722 |
"step": 23900
|
| 1723 |
},
|
| 1724 |
{
|
| 1725 |
"epoch": 6.0,
|
| 1726 |
+
"grad_norm": 0.3925749361515045,
|
| 1727 |
"learning_rate": 7.322330470336314e-05,
|
| 1728 |
+
"loss": 0.9121,
|
| 1729 |
"step": 24000
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
"epoch": 6.0,
|
| 1733 |
+
"eval_loss": 0.9720398187637329,
|
| 1734 |
+
"eval_runtime": 152.9557,
|
| 1735 |
+
"eval_samples_per_second": 26.151,
|
| 1736 |
+
"eval_steps_per_second": 6.538,
|
| 1737 |
"step": 24000
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 6.025,
|
| 1741 |
+
"grad_norm": 0.41935107111930847,
|
| 1742 |
"learning_rate": 7.149635048718294e-05,
|
| 1743 |
+
"loss": 0.9072,
|
| 1744 |
"step": 24100
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 6.05,
|
| 1748 |
+
"grad_norm": 0.4102177321910858,
|
| 1749 |
"learning_rate": 6.97866008242703e-05,
|
| 1750 |
"loss": 0.9105,
|
| 1751 |
"step": 24200
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 6.075,
|
| 1755 |
+
"grad_norm": 0.3717847168445587,
|
| 1756 |
"learning_rate": 6.809422050385628e-05,
|
| 1757 |
+
"loss": 0.9399,
|
| 1758 |
"step": 24300
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 6.1,
|
| 1762 |
+
"grad_norm": 0.3842828869819641,
|
| 1763 |
"learning_rate": 6.641937264107867e-05,
|
| 1764 |
+
"loss": 0.8992,
|
| 1765 |
"step": 24400
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 6.125,
|
| 1769 |
+
"grad_norm": 0.4051499366760254,
|
| 1770 |
"learning_rate": 6.476221866126028e-05,
|
| 1771 |
+
"loss": 0.9274,
|
| 1772 |
"step": 24500
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 6.15,
|
| 1776 |
+
"grad_norm": 0.4197865128517151,
|
| 1777 |
"learning_rate": 6.312291828435076e-05,
|
| 1778 |
+
"loss": 0.9646,
|
| 1779 |
"step": 24600
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 6.175,
|
| 1783 |
+
"grad_norm": 0.41968774795532227,
|
| 1784 |
"learning_rate": 6.150162950953264e-05,
|
| 1785 |
+
"loss": 0.9412,
|
| 1786 |
"step": 24700
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 6.2,
|
| 1790 |
+
"grad_norm": 0.40155336260795593,
|
| 1791 |
"learning_rate": 5.989850859999227e-05,
|
| 1792 |
+
"loss": 0.9627,
|
| 1793 |
"step": 24800
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 6.225,
|
| 1797 |
+
"grad_norm": 0.39962807297706604,
|
| 1798 |
"learning_rate": 5.831371006785963e-05,
|
| 1799 |
+
"loss": 0.9337,
|
| 1800 |
"step": 24900
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 6.25,
|
| 1804 |
+
"grad_norm": 0.4056980609893799,
|
| 1805 |
"learning_rate": 5.6747386659315755e-05,
|
| 1806 |
+
"loss": 0.9221,
|
| 1807 |
"step": 25000
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 6.275,
|
| 1811 |
+
"grad_norm": 0.4260208308696747,
|
| 1812 |
"learning_rate": 5.519968933987082e-05,
|
| 1813 |
+
"loss": 0.9359,
|
| 1814 |
"step": 25100
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 6.3,
|
| 1818 |
+
"grad_norm": 0.39483314752578735,
|
| 1819 |
"learning_rate": 5.367076727981382e-05,
|
| 1820 |
+
"loss": 0.9187,
|
| 1821 |
"step": 25200
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 6.325,
|
| 1825 |
+
"grad_norm": 0.43257468938827515,
|
| 1826 |
"learning_rate": 5.216076783983492e-05,
|
| 1827 |
+
"loss": 0.9701,
|
| 1828 |
"step": 25300
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 6.35,
|
| 1832 |
+
"grad_norm": 0.39114078879356384,
|
| 1833 |
"learning_rate": 5.066983655682325e-05,
|
| 1834 |
+
"loss": 0.9664,
|
| 1835 |
"step": 25400
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 6.375,
|
| 1839 |
+
"grad_norm": 0.39986172318458557,
|
| 1840 |
"learning_rate": 4.919811712983879e-05,
|
| 1841 |
+
"loss": 0.9345,
|
| 1842 |
"step": 25500
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 6.4,
|
| 1846 |
+
"grad_norm": 0.41802579164505005,
|
| 1847 |
"learning_rate": 4.7745751406263163e-05,
|
| 1848 |
+
"loss": 0.9192,
|
| 1849 |
"step": 25600
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 6.425,
|
| 1853 |
+
"grad_norm": 0.34528008103370667,
|
| 1854 |
"learning_rate": 4.6312879368127645e-05,
|
| 1855 |
+
"loss": 0.953,
|
| 1856 |
"step": 25700
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 6.45,
|
| 1860 |
+
"grad_norm": 0.418544739484787,
|
| 1861 |
"learning_rate": 4.4899639118621604e-05,
|
| 1862 |
+
"loss": 0.9121,
|
| 1863 |
"step": 25800
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 6.475,
|
| 1867 |
+
"grad_norm": 0.40490132570266724,
|
| 1868 |
"learning_rate": 4.350616686878175e-05,
|
| 1869 |
+
"loss": 0.9199,
|
| 1870 |
"step": 25900
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 6.5,
|
| 1874 |
+
"grad_norm": 0.4175280034542084,
|
| 1875 |
"learning_rate": 4.213259692436367e-05,
|
| 1876 |
+
"loss": 0.9304,
|
| 1877 |
"step": 26000
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 6.525,
|
| 1881 |
+
"grad_norm": 0.3687520921230316,
|
| 1882 |
"learning_rate": 4.077906167289766e-05,
|
| 1883 |
+
"loss": 0.953,
|
| 1884 |
"step": 26100
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 6.55,
|
| 1888 |
+
"grad_norm": 0.389283686876297,
|
| 1889 |
"learning_rate": 3.944569157092839e-05,
|
| 1890 |
+
"loss": 0.9278,
|
| 1891 |
"step": 26200
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 6.575,
|
| 1895 |
+
"grad_norm": 0.353508323431015,
|
| 1896 |
"learning_rate": 3.8132615131441396e-05,
|
| 1897 |
+
"loss": 0.9488,
|
| 1898 |
"step": 26300
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 6.6,
|
| 1902 |
+
"grad_norm": 0.3994731605052948,
|
| 1903 |
"learning_rate": 3.6839958911476953e-05,
|
| 1904 |
+
"loss": 0.8955,
|
| 1905 |
"step": 26400
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 6.625,
|
| 1909 |
+
"grad_norm": 0.37398815155029297,
|
| 1910 |
"learning_rate": 3.5567847499932e-05,
|
| 1911 |
"loss": 0.9453,
|
| 1912 |
"step": 26500
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 6.65,
|
| 1916 |
+
"grad_norm": 0.37087592482566833,
|
| 1917 |
"learning_rate": 3.431640350555204e-05,
|
| 1918 |
+
"loss": 0.9449,
|
| 1919 |
"step": 26600
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 6.675,
|
| 1923 |
+
"grad_norm": 0.3406459093093872,
|
| 1924 |
"learning_rate": 3.308574754511404e-05,
|
| 1925 |
+
"loss": 0.9143,
|
| 1926 |
"step": 26700
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 6.7,
|
| 1930 |
+
"grad_norm": 0.36757832765579224,
|
| 1931 |
"learning_rate": 3.187599823180071e-05,
|
| 1932 |
+
"loss": 0.935,
|
| 1933 |
"step": 26800
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 6.725,
|
| 1937 |
+
"grad_norm": 0.40823671221733093,
|
| 1938 |
"learning_rate": 3.0687272163768986e-05,
|
| 1939 |
+
"loss": 0.9181,
|
| 1940 |
"step": 26900
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 6.75,
|
| 1944 |
+
"grad_norm": 0.3942839801311493,
|
| 1945 |
"learning_rate": 2.9519683912911265e-05,
|
| 1946 |
+
"loss": 0.9219,
|
| 1947 |
"step": 27000
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 6.775,
|
| 1951 |
+
"grad_norm": 0.3952127695083618,
|
| 1952 |
"learning_rate": 2.8373346013813417e-05,
|
| 1953 |
+
"loss": 0.9001,
|
| 1954 |
"step": 27100
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 6.8,
|
| 1958 |
+
"grad_norm": 0.4122380018234253,
|
| 1959 |
"learning_rate": 2.7248368952908055e-05,
|
| 1960 |
+
"loss": 0.9266,
|
| 1961 |
"step": 27200
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 6.825,
|
| 1965 |
+
"grad_norm": 0.4118337631225586,
|
| 1966 |
"learning_rate": 2.6144861157825773e-05,
|
| 1967 |
+
"loss": 0.9364,
|
| 1968 |
"step": 27300
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 6.85,
|
| 1972 |
+
"grad_norm": 0.39552587270736694,
|
| 1973 |
"learning_rate": 2.5062928986944677e-05,
|
| 1974 |
+
"loss": 0.9226,
|
| 1975 |
"step": 27400
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 6.875,
|
| 1979 |
+
"grad_norm": 0.4226832687854767,
|
| 1980 |
"learning_rate": 2.4002676719139166e-05,
|
| 1981 |
"loss": 0.9394,
|
| 1982 |
"step": 27500
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 6.9,
|
| 1986 |
+
"grad_norm": 0.34474310278892517,
|
| 1987 |
"learning_rate": 2.296420654372966e-05,
|
| 1988 |
+
"loss": 0.9255,
|
| 1989 |
"step": 27600
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 6.925,
|
| 1993 |
+
"grad_norm": 0.4371466636657715,
|
| 1994 |
"learning_rate": 2.1947618550633096e-05,
|
| 1995 |
+
"loss": 0.9298,
|
| 1996 |
"step": 27700
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 6.95,
|
| 2000 |
+
"grad_norm": 0.4318234622478485,
|
| 2001 |
"learning_rate": 2.0953010720716037e-05,
|
| 2002 |
+
"loss": 0.9533,
|
| 2003 |
"step": 27800
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 6.975,
|
| 2007 |
+
"grad_norm": 0.38867416977882385,
|
| 2008 |
"learning_rate": 1.9980478916351297e-05,
|
| 2009 |
+
"loss": 0.9612,
|
| 2010 |
"step": 27900
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 7.0,
|
| 2014 |
+
"grad_norm": 0.38006094098091125,
|
| 2015 |
"learning_rate": 1.9030116872178316e-05,
|
| 2016 |
+
"loss": 0.9175,
|
| 2017 |
"step": 28000
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 7.0,
|
| 2021 |
+
"eval_loss": 0.9707372784614563,
|
| 2022 |
+
"eval_runtime": 152.8648,
|
| 2023 |
+
"eval_samples_per_second": 26.167,
|
| 2024 |
+
"eval_steps_per_second": 6.542,
|
| 2025 |
"step": 28000
|
| 2026 |
},
|
| 2027 |
{
|
| 2028 |
"epoch": 7.025,
|
| 2029 |
+
"grad_norm": 0.4155292510986328,
|
| 2030 |
"learning_rate": 1.8102016186068992e-05,
|
| 2031 |
+
"loss": 0.936,
|
| 2032 |
"step": 28100
|
| 2033 |
},
|
| 2034 |
{
|
| 2035 |
"epoch": 7.05,
|
| 2036 |
+
"grad_norm": 0.3884825110435486,
|
| 2037 |
"learning_rate": 1.719626631029911e-05,
|
| 2038 |
+
"loss": 0.9398,
|
| 2039 |
"step": 28200
|
| 2040 |
},
|
| 2041 |
{
|
| 2042 |
"epoch": 7.075,
|
| 2043 |
+
"grad_norm": 0.39031410217285156,
|
| 2044 |
"learning_rate": 1.6312954542926888e-05,
|
| 2045 |
+
"loss": 0.9056,
|
| 2046 |
"step": 28300
|
| 2047 |
},
|
| 2048 |
{
|
| 2049 |
"epoch": 7.1,
|
| 2050 |
+
"grad_norm": 0.4356350004673004,
|
| 2051 |
"learning_rate": 1.5452166019378987e-05,
|
| 2052 |
+
"loss": 0.9178,
|
| 2053 |
"step": 28400
|
| 2054 |
},
|
| 2055 |
{
|
| 2056 |
"epoch": 7.125,
|
| 2057 |
+
"grad_norm": 0.3904324769973755,
|
| 2058 |
"learning_rate": 1.4613983704244827e-05,
|
| 2059 |
+
"loss": 0.9067,
|
| 2060 |
"step": 28500
|
| 2061 |
},
|
| 2062 |
{
|
| 2063 |
"epoch": 7.15,
|
| 2064 |
+
"grad_norm": 0.3213234543800354,
|
| 2065 |
"learning_rate": 1.3798488383280488e-05,
|
| 2066 |
+
"loss": 0.9436,
|
| 2067 |
"step": 28600
|
| 2068 |
},
|
| 2069 |
{
|
| 2070 |
"epoch": 7.175,
|
| 2071 |
+
"grad_norm": 0.38800540566444397,
|
| 2072 |
"learning_rate": 1.3005758655622424e-05,
|
| 2073 |
+
"loss": 0.9131,
|
| 2074 |
"step": 28700
|
| 2075 |
},
|
| 2076 |
{
|
| 2077 |
"epoch": 7.2,
|
| 2078 |
+
"grad_norm": 0.3734656870365143,
|
| 2079 |
"learning_rate": 1.2235870926211617e-05,
|
| 2080 |
+
"loss": 0.9466,
|
| 2081 |
"step": 28800
|
| 2082 |
},
|
| 2083 |
{
|
| 2084 |
"epoch": 7.225,
|
| 2085 |
+
"grad_norm": 0.4088476002216339,
|
| 2086 |
"learning_rate": 1.1488899398429897e-05,
|
| 2087 |
+
"loss": 0.9444,
|
| 2088 |
"step": 28900
|
| 2089 |
},
|
| 2090 |
{
|
| 2091 |
"epoch": 7.25,
|
| 2092 |
+
"grad_norm": 0.3788207471370697,
|
| 2093 |
"learning_rate": 1.0764916066947795e-05,
|
| 2094 |
+
"loss": 0.9015,
|
| 2095 |
"step": 29000
|
| 2096 |
},
|
| 2097 |
{
|
| 2098 |
"epoch": 7.275,
|
| 2099 |
+
"grad_norm": 0.3710539638996124,
|
| 2100 |
"learning_rate": 1.0063990710785648e-05,
|
| 2101 |
+
"loss": 0.9356,
|
| 2102 |
"step": 29100
|
| 2103 |
},
|
| 2104 |
{
|
| 2105 |
"epoch": 7.3,
|
| 2106 |
+
"grad_norm": 0.39175015687942505,
|
| 2107 |
"learning_rate": 9.386190886588208e-06,
|
| 2108 |
+
"loss": 0.9277,
|
| 2109 |
"step": 29200
|
| 2110 |
},
|
| 2111 |
{
|
| 2112 |
"epoch": 7.325,
|
| 2113 |
+
"grad_norm": 0.39857038855552673,
|
| 2114 |
"learning_rate": 8.731581922113152e-06,
|
| 2115 |
+
"loss": 0.9597,
|
| 2116 |
"step": 29300
|
| 2117 |
},
|
| 2118 |
{
|
| 2119 |
"epoch": 7.35,
|
| 2120 |
+
"grad_norm": 0.42670151591300964,
|
| 2121 |
"learning_rate": 8.10022690993506e-06,
|
| 2122 |
+
"loss": 0.9124,
|
| 2123 |
"step": 29400
|
| 2124 |
},
|
| 2125 |
{
|
| 2126 |
"epoch": 7.375,
|
| 2127 |
+
"grad_norm": 0.37594255805015564,
|
| 2128 |
"learning_rate": 7.4921867013640064e-06,
|
| 2129 |
+
"loss": 0.9202,
|
| 2130 |
"step": 29500
|
| 2131 |
},
|
| 2132 |
{
|
| 2133 |
"epoch": 7.4,
|
| 2134 |
+
"grad_norm": 0.34726592898368835,
|
| 2135 |
"learning_rate": 6.907519900580861e-06,
|
| 2136 |
+
"loss": 0.9462,
|
| 2137 |
"step": 29600
|
| 2138 |
},
|
| 2139 |
{
|
| 2140 |
"epoch": 7.425,
|
| 2141 |
+
"grad_norm": 0.3583536744117737,
|
| 2142 |
"learning_rate": 6.34628285898875e-06,
|
| 2143 |
+
"loss": 0.9356,
|
| 2144 |
"step": 29700
|
| 2145 |
},
|
| 2146 |
{
|
| 2147 |
"epoch": 7.45,
|
| 2148 |
+
"grad_norm": 0.40716230869293213,
|
| 2149 |
"learning_rate": 5.808529669781903e-06,
|
| 2150 |
+
"loss": 0.9204,
|
| 2151 |
"step": 29800
|
| 2152 |
},
|
| 2153 |
{
|
| 2154 |
"epoch": 7.475,
|
| 2155 |
+
"grad_norm": 0.4574654698371887,
|
| 2156 |
"learning_rate": 5.294312162731935e-06,
|
| 2157 |
+
"loss": 0.9272,
|
| 2158 |
"step": 29900
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
"epoch": 7.5,
|
| 2162 |
+
"grad_norm": 0.41072556376457214,
|
| 2163 |
"learning_rate": 4.803679899192393e-06,
|
| 2164 |
+
"loss": 0.9181,
|
| 2165 |
"step": 30000
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
"epoch": 7.525,
|
| 2169 |
+
"grad_norm": 0.4989704191684723,
|
| 2170 |
"learning_rate": 4.336680167322055e-06,
|
| 2171 |
+
"loss": 0.9346,
|
| 2172 |
"step": 30100
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
"epoch": 7.55,
|
| 2176 |
+
"grad_norm": 0.4089345932006836,
|
| 2177 |
"learning_rate": 3.893357977527101e-06,
|
| 2178 |
+
"loss": 0.9392,
|
| 2179 |
"step": 30200
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
"epoch": 7.575,
|
| 2183 |
+
"grad_norm": 0.4189968407154083,
|
| 2184 |
"learning_rate": 3.4737560581228343e-06,
|
| 2185 |
+
"loss": 0.9338,
|
| 2186 |
"step": 30300
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
"epoch": 7.6,
|
| 2190 |
+
"grad_norm": 0.40372830629348755,
|
| 2191 |
"learning_rate": 3.077914851215585e-06,
|
| 2192 |
+
"loss": 0.9394,
|
| 2193 |
"step": 30400
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
"epoch": 7.625,
|
| 2197 |
+
"grad_norm": 0.3864797055721283,
|
| 2198 |
"learning_rate": 2.7058725088047465e-06,
|
| 2199 |
+
"loss": 0.9283,
|
| 2200 |
"step": 30500
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
"epoch": 7.65,
|
| 2204 |
+
"grad_norm": 0.4125591814517975,
|
| 2205 |
"learning_rate": 2.357664889105687e-06,
|
| 2206 |
+
"loss": 0.9731,
|
| 2207 |
"step": 30600
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
"epoch": 7.675,
|
| 2211 |
+
"grad_norm": 0.4122030436992645,
|
| 2212 |
"learning_rate": 2.0333255530934903e-06,
|
| 2213 |
+
"loss": 0.9257,
|
| 2214 |
"step": 30700
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 7.7,
|
| 2218 |
+
"grad_norm": 0.40973401069641113,
|
| 2219 |
"learning_rate": 1.7328857612684267e-06,
|
| 2220 |
+
"loss": 0.8964,
|
| 2221 |
"step": 30800
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
"epoch": 7.725,
|
| 2225 |
+
"grad_norm": 0.36363255977630615,
|
| 2226 |
"learning_rate": 1.4563744706429517e-06,
|
| 2227 |
+
"loss": 0.9012,
|
| 2228 |
"step": 30900
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
"epoch": 7.75,
|
| 2232 |
+
"grad_norm": 0.39668798446655273,
|
| 2233 |
"learning_rate": 1.2038183319507957e-06,
|
| 2234 |
+
"loss": 0.9399,
|
| 2235 |
"step": 31000
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
"epoch": 7.775,
|
| 2239 |
+
"grad_norm": 0.40102848410606384,
|
| 2240 |
"learning_rate": 9.752416870782156e-07,
|
| 2241 |
+
"loss": 0.9471,
|
| 2242 |
"step": 31100
|
| 2243 |
},
|
| 2244 |
{
|
| 2245 |
"epoch": 7.8,
|
| 2246 |
+
"grad_norm": 0.43242618441581726,
|
| 2247 |
"learning_rate": 7.70666566718009e-07,
|
| 2248 |
+
"loss": 0.9289,
|
| 2249 |
"step": 31200
|
| 2250 |
},
|
| 2251 |
{
|
| 2252 |
"epoch": 7.825,
|
| 2253 |
+
"grad_norm": 0.41685134172439575,
|
| 2254 |
"learning_rate": 5.90112688246075e-07,
|
| 2255 |
+
"loss": 0.9365,
|
| 2256 |
"step": 31300
|
| 2257 |
},
|
| 2258 |
{
|
| 2259 |
"epoch": 7.85,
|
| 2260 |
+
"grad_norm": 0.36857473850250244,
|
| 2261 |
"learning_rate": 4.335974538210441e-07,
|
| 2262 |
+
"loss": 0.9324,
|
| 2263 |
"step": 31400
|
| 2264 |
},
|
| 2265 |
{
|
| 2266 |
"epoch": 7.875,
|
| 2267 |
+
"grad_norm": 0.42975786328315735,
|
| 2268 |
"learning_rate": 3.0113594870689873e-07,
|
| 2269 |
+
"loss": 0.9379,
|
| 2270 |
"step": 31500
|
| 2271 |
},
|
| 2272 |
{
|
| 2273 |
"epoch": 7.9,
|
| 2274 |
+
"grad_norm": 0.42332813143730164,
|
| 2275 |
"learning_rate": 1.9274093981927476e-07,
|
| 2276 |
+
"loss": 0.9453,
|
| 2277 |
"step": 31600
|
| 2278 |
},
|
| 2279 |
{
|
| 2280 |
"epoch": 7.925,
|
| 2281 |
+
"grad_norm": 0.41806164383888245,
|
| 2282 |
"learning_rate": 1.0842287449469579e-07,
|
| 2283 |
+
"loss": 0.9055,
|
| 2284 |
"step": 31700
|
| 2285 |
},
|
| 2286 |
{
|
| 2287 |
"epoch": 7.95,
|
| 2288 |
+
"grad_norm": 0.35621610283851624,
|
| 2289 |
"learning_rate": 4.818987948379538e-08,
|
| 2290 |
+
"loss": 0.8726,
|
| 2291 |
"step": 31800
|
| 2292 |
},
|
| 2293 |
{
|
| 2294 |
"epoch": 7.975,
|
| 2295 |
+
"grad_norm": 0.43373915553092957,
|
| 2296 |
"learning_rate": 1.2047760167999133e-08,
|
| 2297 |
+
"loss": 0.946,
|
| 2298 |
"step": 31900
|
| 2299 |
},
|
| 2300 |
{
|
| 2301 |
"epoch": 8.0,
|
| 2302 |
+
"grad_norm": 0.4070851504802704,
|
| 2303 |
"learning_rate": 0.0,
|
| 2304 |
+
"loss": 0.9197,
|
| 2305 |
"step": 32000
|
| 2306 |
}
|
| 2307 |
],
|