Training in progress, step 400, checkpoint
Browse files
last-checkpoint/adapter_config.json
CHANGED
|
@@ -20,10 +20,10 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"dense_4h_to_h",
|
| 24 |
-
"query_key_value",
|
| 25 |
"dense",
|
| 26 |
-
"
|
|
|
|
|
|
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
| 23 |
"dense",
|
| 24 |
+
"dense_4h_to_h",
|
| 25 |
+
"dense_h_to_4h",
|
| 26 |
+
"query_key_value"
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1579384
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6122248efad15f08bb35ef89dbfb4db02d69860b1a9b820e62900a2fb0d6353
|
| 3 |
size 1579384
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 857338
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6ba50451f2f4098994e44c0276099046264f1481fc025e6ae04dd5638e8e654
|
| 3 |
size 857338
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:27038ab56031f405c10aacd6fb0edec15d29b713b4ea2a4062288739d9c449ed
|
| 3 |
size 14244
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 1.
|
| 5 |
"eval_steps": 200,
|
| 6 |
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -10,2826 +10,2826 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.002641135688345989,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 1.0000000000000002e-06,
|
| 15 |
-
"loss":
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.002641135688345989,
|
| 20 |
-
"eval_loss": 10.
|
| 21 |
-
"eval_runtime": 2.
|
| 22 |
-
"eval_samples_per_second":
|
| 23 |
-
"eval_steps_per_second":
|
| 24 |
"step": 1
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.005282271376691978,
|
| 28 |
-
"grad_norm":
|
| 29 |
"learning_rate": 2.0000000000000003e-06,
|
| 30 |
-
"loss":
|
| 31 |
"step": 2
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.007923407065037967,
|
| 35 |
-
"grad_norm":
|
| 36 |
"learning_rate": 3e-06,
|
| 37 |
-
"loss":
|
| 38 |
"step": 3
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.010564542753383956,
|
| 42 |
-
"grad_norm":
|
| 43 |
"learning_rate": 4.000000000000001e-06,
|
| 44 |
-
"loss":
|
| 45 |
"step": 4
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.013205678441729944,
|
| 49 |
-
"grad_norm":
|
| 50 |
"learning_rate": 5e-06,
|
| 51 |
-
"loss":
|
| 52 |
"step": 5
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.015846814130075933,
|
| 56 |
-
"grad_norm":
|
| 57 |
"learning_rate": 6e-06,
|
| 58 |
-
"loss":
|
| 59 |
"step": 6
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.01848794981842192,
|
| 63 |
-
"grad_norm":
|
| 64 |
"learning_rate": 7.000000000000001e-06,
|
| 65 |
-
"loss":
|
| 66 |
"step": 7
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.02112908550676791,
|
| 70 |
-
"grad_norm":
|
| 71 |
"learning_rate": 8.000000000000001e-06,
|
| 72 |
-
"loss":
|
| 73 |
"step": 8
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.0237702211951139,
|
| 77 |
-
"grad_norm":
|
| 78 |
"learning_rate": 9e-06,
|
| 79 |
-
"loss":
|
| 80 |
"step": 9
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.02641135688345989,
|
| 84 |
-
"grad_norm":
|
| 85 |
"learning_rate": 1e-05,
|
| 86 |
-
"loss":
|
| 87 |
"step": 10
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.029052492571805876,
|
| 91 |
-
"grad_norm":
|
| 92 |
"learning_rate": 1.1000000000000001e-05,
|
| 93 |
-
"loss":
|
| 94 |
"step": 11
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.03169362826015187,
|
| 98 |
-
"grad_norm":
|
| 99 |
"learning_rate": 1.2e-05,
|
| 100 |
-
"loss":
|
| 101 |
"step": 12
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.034334763948497854,
|
| 105 |
-
"grad_norm":
|
| 106 |
"learning_rate": 1.3000000000000001e-05,
|
| 107 |
-
"loss":
|
| 108 |
"step": 13
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.03697589963684384,
|
| 112 |
-
"grad_norm":
|
| 113 |
"learning_rate": 1.4000000000000001e-05,
|
| 114 |
-
"loss":
|
| 115 |
"step": 14
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.03961703532518983,
|
| 119 |
-
"grad_norm":
|
| 120 |
"learning_rate": 1.5e-05,
|
| 121 |
-
"loss":
|
| 122 |
"step": 15
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.04225817101353582,
|
| 126 |
-
"grad_norm":
|
| 127 |
"learning_rate": 1.6000000000000003e-05,
|
| 128 |
-
"loss":
|
| 129 |
"step": 16
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.04489930670188181,
|
| 133 |
-
"grad_norm":
|
| 134 |
"learning_rate": 1.7000000000000003e-05,
|
| 135 |
-
"loss":
|
| 136 |
"step": 17
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.0475404423902278,
|
| 140 |
-
"grad_norm":
|
| 141 |
"learning_rate": 1.8e-05,
|
| 142 |
-
"loss":
|
| 143 |
"step": 18
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.050181578078573784,
|
| 147 |
-
"grad_norm":
|
| 148 |
"learning_rate": 1.9e-05,
|
| 149 |
-
"loss":
|
| 150 |
"step": 19
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.05282271376691978,
|
| 154 |
-
"grad_norm":
|
| 155 |
"learning_rate": 2e-05,
|
| 156 |
-
"loss":
|
| 157 |
"step": 20
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.055463849455265765,
|
| 161 |
-
"grad_norm":
|
| 162 |
"learning_rate": 2.1e-05,
|
| 163 |
-
"loss":
|
| 164 |
"step": 21
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.05810498514361175,
|
| 168 |
-
"grad_norm":
|
| 169 |
"learning_rate": 2.2000000000000003e-05,
|
| 170 |
-
"loss":
|
| 171 |
"step": 22
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.06074612083195774,
|
| 175 |
-
"grad_norm":
|
| 176 |
"learning_rate": 2.3000000000000003e-05,
|
| 177 |
-
"loss":
|
| 178 |
"step": 23
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.06338725652030373,
|
| 182 |
-
"grad_norm":
|
| 183 |
"learning_rate": 2.4e-05,
|
| 184 |
-
"loss":
|
| 185 |
"step": 24
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.06602839220864971,
|
| 189 |
-
"grad_norm":
|
| 190 |
"learning_rate": 2.5e-05,
|
| 191 |
-
"loss":
|
| 192 |
"step": 25
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.06866952789699571,
|
| 196 |
-
"grad_norm":
|
| 197 |
"learning_rate": 2.6000000000000002e-05,
|
| 198 |
-
"loss":
|
| 199 |
"step": 26
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.0713106635853417,
|
| 203 |
-
"grad_norm":
|
| 204 |
"learning_rate": 2.7000000000000002e-05,
|
| 205 |
-
"loss":
|
| 206 |
"step": 27
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.07395179927368768,
|
| 210 |
-
"grad_norm":
|
| 211 |
"learning_rate": 2.8000000000000003e-05,
|
| 212 |
-
"loss":
|
| 213 |
"step": 28
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.07659293496203368,
|
| 217 |
-
"grad_norm":
|
| 218 |
"learning_rate": 2.9e-05,
|
| 219 |
-
"loss":
|
| 220 |
"step": 29
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.07923407065037966,
|
| 224 |
-
"grad_norm":
|
| 225 |
"learning_rate": 3e-05,
|
| 226 |
-
"loss":
|
| 227 |
"step": 30
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.08187520633872565,
|
| 231 |
-
"grad_norm":
|
| 232 |
"learning_rate": 3.1e-05,
|
| 233 |
-
"loss":
|
| 234 |
"step": 31
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.08451634202707164,
|
| 238 |
-
"grad_norm":
|
| 239 |
"learning_rate": 3.2000000000000005e-05,
|
| 240 |
-
"loss":
|
| 241 |
"step": 32
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.08715747771541763,
|
| 245 |
-
"grad_norm":
|
| 246 |
"learning_rate": 3.3e-05,
|
| 247 |
-
"loss":
|
| 248 |
"step": 33
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.08979861340376362,
|
| 252 |
-
"grad_norm":
|
| 253 |
"learning_rate": 3.4000000000000007e-05,
|
| 254 |
-
"loss":
|
| 255 |
"step": 34
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.09243974909210961,
|
| 259 |
-
"grad_norm":
|
| 260 |
"learning_rate": 3.5e-05,
|
| 261 |
-
"loss":
|
| 262 |
"step": 35
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.0950808847804556,
|
| 266 |
-
"grad_norm":
|
| 267 |
"learning_rate": 3.6e-05,
|
| 268 |
-
"loss":
|
| 269 |
"step": 36
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.09772202046880159,
|
| 273 |
-
"grad_norm":
|
| 274 |
"learning_rate": 3.7e-05,
|
| 275 |
-
"loss":
|
| 276 |
"step": 37
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.10036315615714757,
|
| 280 |
-
"grad_norm":
|
| 281 |
"learning_rate": 3.8e-05,
|
| 282 |
-
"loss":
|
| 283 |
"step": 38
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.10300429184549356,
|
| 287 |
-
"grad_norm":
|
| 288 |
"learning_rate": 3.9000000000000006e-05,
|
| 289 |
-
"loss":
|
| 290 |
"step": 39
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.10564542753383956,
|
| 294 |
-
"grad_norm":
|
| 295 |
"learning_rate": 4e-05,
|
| 296 |
-
"loss":
|
| 297 |
"step": 40
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.10828656322218554,
|
| 301 |
-
"grad_norm":
|
| 302 |
"learning_rate": 4.1e-05,
|
| 303 |
-
"loss":
|
| 304 |
"step": 41
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.11092769891053153,
|
| 308 |
-
"grad_norm":
|
| 309 |
"learning_rate": 4.2e-05,
|
| 310 |
-
"loss":
|
| 311 |
"step": 42
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.11356883459887751,
|
| 315 |
-
"grad_norm":
|
| 316 |
"learning_rate": 4.3e-05,
|
| 317 |
-
"loss":
|
| 318 |
"step": 43
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.1162099702872235,
|
| 322 |
-
"grad_norm":
|
| 323 |
"learning_rate": 4.4000000000000006e-05,
|
| 324 |
-
"loss":
|
| 325 |
"step": 44
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.1188511059755695,
|
| 329 |
-
"grad_norm":
|
| 330 |
"learning_rate": 4.5e-05,
|
| 331 |
-
"loss":
|
| 332 |
"step": 45
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.12149224166391548,
|
| 336 |
-
"grad_norm":
|
| 337 |
"learning_rate": 4.600000000000001e-05,
|
| 338 |
-
"loss":
|
| 339 |
"step": 46
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.12413337735226147,
|
| 343 |
-
"grad_norm":
|
| 344 |
"learning_rate": 4.7e-05,
|
| 345 |
-
"loss":
|
| 346 |
"step": 47
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.12677451304060747,
|
| 350 |
-
"grad_norm":
|
| 351 |
"learning_rate": 4.8e-05,
|
| 352 |
-
"loss":
|
| 353 |
"step": 48
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.12941564872895345,
|
| 357 |
-
"grad_norm":
|
| 358 |
"learning_rate": 4.9e-05,
|
| 359 |
-
"loss":
|
| 360 |
"step": 49
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.13205678441729943,
|
| 364 |
-
"grad_norm":
|
| 365 |
"learning_rate": 5e-05,
|
| 366 |
-
"loss":
|
| 367 |
"step": 50
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.13469792010564544,
|
| 371 |
-
"grad_norm":
|
| 372 |
"learning_rate": 5.1000000000000006e-05,
|
| 373 |
-
"loss":
|
| 374 |
"step": 51
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.13733905579399142,
|
| 378 |
-
"grad_norm":
|
| 379 |
"learning_rate": 5.2000000000000004e-05,
|
| 380 |
-
"loss":
|
| 381 |
"step": 52
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.1399801914823374,
|
| 385 |
-
"grad_norm":
|
| 386 |
"learning_rate": 5.300000000000001e-05,
|
| 387 |
-
"loss":
|
| 388 |
"step": 53
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.1426213271706834,
|
| 392 |
-
"grad_norm":
|
| 393 |
"learning_rate": 5.4000000000000005e-05,
|
| 394 |
-
"loss":
|
| 395 |
"step": 54
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.14526246285902938,
|
| 399 |
-
"grad_norm":
|
| 400 |
"learning_rate": 5.500000000000001e-05,
|
| 401 |
-
"loss":
|
| 402 |
"step": 55
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.14790359854737536,
|
| 406 |
-
"grad_norm":
|
| 407 |
"learning_rate": 5.6000000000000006e-05,
|
| 408 |
-
"loss":
|
| 409 |
"step": 56
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.15054473423572137,
|
| 413 |
-
"grad_norm":
|
| 414 |
"learning_rate": 5.6999999999999996e-05,
|
| 415 |
-
"loss":
|
| 416 |
"step": 57
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.15318586992406735,
|
| 420 |
-
"grad_norm":
|
| 421 |
"learning_rate": 5.8e-05,
|
| 422 |
-
"loss":
|
| 423 |
"step": 58
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.15582700561241333,
|
| 427 |
-
"grad_norm":
|
| 428 |
"learning_rate": 5.9e-05,
|
| 429 |
-
"loss":
|
| 430 |
"step": 59
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.1584681413007593,
|
| 434 |
-
"grad_norm":
|
| 435 |
"learning_rate": 6e-05,
|
| 436 |
-
"loss":
|
| 437 |
"step": 60
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.16110927698910532,
|
| 441 |
-
"grad_norm":
|
| 442 |
"learning_rate": 6.1e-05,
|
| 443 |
-
"loss":
|
| 444 |
"step": 61
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.1637504126774513,
|
| 448 |
-
"grad_norm":
|
| 449 |
"learning_rate": 6.2e-05,
|
| 450 |
-
"loss":
|
| 451 |
"step": 62
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.16639154836579728,
|
| 455 |
-
"grad_norm":
|
| 456 |
"learning_rate": 6.3e-05,
|
| 457 |
-
"loss":
|
| 458 |
"step": 63
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.1690326840541433,
|
| 462 |
-
"grad_norm":
|
| 463 |
"learning_rate": 6.400000000000001e-05,
|
| 464 |
-
"loss":
|
| 465 |
"step": 64
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.17167381974248927,
|
| 469 |
-
"grad_norm":
|
| 470 |
"learning_rate": 6.500000000000001e-05,
|
| 471 |
-
"loss":
|
| 472 |
"step": 65
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.17431495543083525,
|
| 476 |
-
"grad_norm":
|
| 477 |
"learning_rate": 6.6e-05,
|
| 478 |
-
"loss":
|
| 479 |
"step": 66
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.17695609111918126,
|
| 483 |
-
"grad_norm":
|
| 484 |
"learning_rate": 6.7e-05,
|
| 485 |
-
"loss":
|
| 486 |
"step": 67
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.17959722680752724,
|
| 490 |
-
"grad_norm":
|
| 491 |
"learning_rate": 6.800000000000001e-05,
|
| 492 |
-
"loss":
|
| 493 |
"step": 68
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.18223836249587322,
|
| 497 |
-
"grad_norm":
|
| 498 |
"learning_rate": 6.9e-05,
|
| 499 |
-
"loss":
|
| 500 |
"step": 69
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.18487949818421923,
|
| 504 |
-
"grad_norm":
|
| 505 |
"learning_rate": 7e-05,
|
| 506 |
-
"loss":
|
| 507 |
"step": 70
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.1875206338725652,
|
| 511 |
-
"grad_norm":
|
| 512 |
"learning_rate": 7.1e-05,
|
| 513 |
-
"loss":
|
| 514 |
"step": 71
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.1901617695609112,
|
| 518 |
-
"grad_norm":
|
| 519 |
"learning_rate": 7.2e-05,
|
| 520 |
-
"loss":
|
| 521 |
"step": 72
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.19280290524925717,
|
| 525 |
-
"grad_norm":
|
| 526 |
"learning_rate": 7.3e-05,
|
| 527 |
-
"loss":
|
| 528 |
"step": 73
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.19544404093760318,
|
| 532 |
-
"grad_norm":
|
| 533 |
"learning_rate": 7.4e-05,
|
| 534 |
-
"loss":
|
| 535 |
"step": 74
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.19808517662594916,
|
| 539 |
-
"grad_norm":
|
| 540 |
"learning_rate": 7.500000000000001e-05,
|
| 541 |
-
"loss":
|
| 542 |
"step": 75
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.20072631231429514,
|
| 546 |
-
"grad_norm":
|
| 547 |
"learning_rate": 7.6e-05,
|
| 548 |
-
"loss":
|
| 549 |
"step": 76
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.20336744800264114,
|
| 553 |
-
"grad_norm":
|
| 554 |
"learning_rate": 7.7e-05,
|
| 555 |
-
"loss":
|
| 556 |
"step": 77
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.20600858369098712,
|
| 560 |
-
"grad_norm":
|
| 561 |
"learning_rate": 7.800000000000001e-05,
|
| 562 |
-
"loss":
|
| 563 |
"step": 78
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.2086497193793331,
|
| 567 |
-
"grad_norm":
|
| 568 |
"learning_rate": 7.900000000000001e-05,
|
| 569 |
-
"loss":
|
| 570 |
"step": 79
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.2112908550676791,
|
| 574 |
-
"grad_norm":
|
| 575 |
"learning_rate": 8e-05,
|
| 576 |
-
"loss":
|
| 577 |
"step": 80
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.2139319907560251,
|
| 581 |
-
"grad_norm":
|
| 582 |
"learning_rate": 8.1e-05,
|
| 583 |
-
"loss":
|
| 584 |
"step": 81
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.21657312644437107,
|
| 588 |
-
"grad_norm":
|
| 589 |
"learning_rate": 8.2e-05,
|
| 590 |
-
"loss":
|
| 591 |
"step": 82
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.21921426213271708,
|
| 595 |
-
"grad_norm":
|
| 596 |
"learning_rate": 8.3e-05,
|
| 597 |
-
"loss":
|
| 598 |
"step": 83
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.22185539782106306,
|
| 602 |
-
"grad_norm":
|
| 603 |
"learning_rate": 8.4e-05,
|
| 604 |
-
"loss":
|
| 605 |
"step": 84
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.22449653350940904,
|
| 609 |
-
"grad_norm":
|
| 610 |
"learning_rate": 8.5e-05,
|
| 611 |
-
"loss":
|
| 612 |
"step": 85
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.22713766919775502,
|
| 616 |
-
"grad_norm":
|
| 617 |
"learning_rate": 8.6e-05,
|
| 618 |
-
"loss":
|
| 619 |
"step": 86
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.22977880488610103,
|
| 623 |
-
"grad_norm":
|
| 624 |
"learning_rate": 8.7e-05,
|
| 625 |
-
"loss":
|
| 626 |
"step": 87
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.232419940574447,
|
| 630 |
-
"grad_norm":
|
| 631 |
"learning_rate": 8.800000000000001e-05,
|
| 632 |
-
"loss":
|
| 633 |
"step": 88
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.235061076262793,
|
| 637 |
-
"grad_norm":
|
| 638 |
"learning_rate": 8.900000000000001e-05,
|
| 639 |
-
"loss":
|
| 640 |
"step": 89
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.237702211951139,
|
| 644 |
-
"grad_norm":
|
| 645 |
"learning_rate": 9e-05,
|
| 646 |
-
"loss":
|
| 647 |
"step": 90
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.24034334763948498,
|
| 651 |
-
"grad_norm":
|
| 652 |
"learning_rate": 9.1e-05,
|
| 653 |
-
"loss":
|
| 654 |
"step": 91
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.24298448332783096,
|
| 658 |
-
"grad_norm":
|
| 659 |
"learning_rate": 9.200000000000001e-05,
|
| 660 |
-
"loss":
|
| 661 |
"step": 92
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.24562561901617697,
|
| 665 |
-
"grad_norm":
|
| 666 |
"learning_rate": 9.300000000000001e-05,
|
| 667 |
-
"loss":
|
| 668 |
"step": 93
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.24826675470452295,
|
| 672 |
-
"grad_norm":
|
| 673 |
"learning_rate": 9.4e-05,
|
| 674 |
-
"loss":
|
| 675 |
"step": 94
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.2509078903928689,
|
| 679 |
-
"grad_norm":
|
| 680 |
"learning_rate": 9.5e-05,
|
| 681 |
-
"loss":
|
| 682 |
"step": 95
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.25354902608121493,
|
| 686 |
-
"grad_norm":
|
| 687 |
"learning_rate": 9.6e-05,
|
| 688 |
-
"loss":
|
| 689 |
"step": 96
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.2561901617695609,
|
| 693 |
-
"grad_norm":
|
| 694 |
"learning_rate": 9.7e-05,
|
| 695 |
-
"loss":
|
| 696 |
"step": 97
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.2588312974579069,
|
| 700 |
-
"grad_norm":
|
| 701 |
"learning_rate": 9.8e-05,
|
| 702 |
-
"loss":
|
| 703 |
"step": 98
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.2614724331462529,
|
| 707 |
-
"grad_norm":
|
| 708 |
"learning_rate": 9.900000000000001e-05,
|
| 709 |
-
"loss":
|
| 710 |
"step": 99
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.26411356883459886,
|
| 714 |
-
"grad_norm":
|
| 715 |
"learning_rate": 0.0001,
|
| 716 |
-
"loss":
|
| 717 |
"step": 100
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.26675470452294486,
|
| 721 |
-
"grad_norm":
|
| 722 |
"learning_rate": 9.99999993018897e-05,
|
| 723 |
-
"loss":
|
| 724 |
"step": 101
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.26939584021129087,
|
| 728 |
-
"grad_norm":
|
| 729 |
"learning_rate": 9.999999720755877e-05,
|
| 730 |
-
"loss":
|
| 731 |
"step": 102
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.2720369758996368,
|
| 735 |
-
"grad_norm":
|
| 736 |
"learning_rate": 9.99999937170073e-05,
|
| 737 |
-
"loss":
|
| 738 |
"step": 103
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.27467811158798283,
|
| 742 |
-
"grad_norm":
|
| 743 |
"learning_rate": 9.999998883023537e-05,
|
| 744 |
-
"loss":
|
| 745 |
"step": 104
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.27731924727632884,
|
| 749 |
-
"grad_norm":
|
| 750 |
"learning_rate": 9.999998254724313e-05,
|
| 751 |
-
"loss":
|
| 752 |
"step": 105
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.2799603829646748,
|
| 756 |
-
"grad_norm":
|
| 757 |
"learning_rate": 9.999997486803075e-05,
|
| 758 |
-
"loss":
|
| 759 |
"step": 106
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.2826015186530208,
|
| 763 |
-
"grad_norm":
|
| 764 |
"learning_rate": 9.999996579259843e-05,
|
| 765 |
-
"loss":
|
| 766 |
"step": 107
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.2852426543413668,
|
| 770 |
-
"grad_norm":
|
| 771 |
"learning_rate": 9.999995532094644e-05,
|
| 772 |
-
"loss":
|
| 773 |
"step": 108
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.28788379002971276,
|
| 777 |
-
"grad_norm":
|
| 778 |
"learning_rate": 9.999994345307508e-05,
|
| 779 |
-
"loss":
|
| 780 |
"step": 109
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.29052492571805877,
|
| 784 |
-
"grad_norm":
|
| 785 |
"learning_rate": 9.999993018898466e-05,
|
| 786 |
-
"loss":
|
| 787 |
"step": 110
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.2931660614064048,
|
| 791 |
-
"grad_norm":
|
| 792 |
"learning_rate": 9.999991552867558e-05,
|
| 793 |
-
"loss":
|
| 794 |
"step": 111
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.29580719709475073,
|
| 798 |
-
"grad_norm":
|
| 799 |
"learning_rate": 9.99998994721482e-05,
|
| 800 |
-
"loss":
|
| 801 |
"step": 112
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.29844833278309674,
|
| 805 |
-
"grad_norm":
|
| 806 |
"learning_rate": 9.999988201940302e-05,
|
| 807 |
-
"loss":
|
| 808 |
"step": 113
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.30108946847144274,
|
| 812 |
-
"grad_norm":
|
| 813 |
"learning_rate": 9.999986317044051e-05,
|
| 814 |
-
"loss":
|
| 815 |
"step": 114
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.3037306041597887,
|
| 819 |
-
"grad_norm":
|
| 820 |
"learning_rate": 9.999984292526118e-05,
|
| 821 |
-
"loss":
|
| 822 |
"step": 115
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 0.3063717398481347,
|
| 826 |
-
"grad_norm":
|
| 827 |
"learning_rate": 9.999982128386562e-05,
|
| 828 |
-
"loss":
|
| 829 |
"step": 116
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 0.3090128755364807,
|
| 833 |
-
"grad_norm":
|
| 834 |
"learning_rate": 9.99997982462544e-05,
|
| 835 |
-
"loss":
|
| 836 |
"step": 117
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.31165401122482667,
|
| 840 |
-
"grad_norm":
|
| 841 |
"learning_rate": 9.999977381242821e-05,
|
| 842 |
-
"loss":
|
| 843 |
"step": 118
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 0.3142951469131727,
|
| 847 |
-
"grad_norm":
|
| 848 |
"learning_rate": 9.999974798238769e-05,
|
| 849 |
-
"loss":
|
| 850 |
"step": 119
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 0.3169362826015186,
|
| 854 |
-
"grad_norm":
|
| 855 |
"learning_rate": 9.99997207561336e-05,
|
| 856 |
-
"loss":
|
| 857 |
"step": 120
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.31957741828986463,
|
| 861 |
-
"grad_norm":
|
| 862 |
"learning_rate": 9.999969213366667e-05,
|
| 863 |
-
"loss":
|
| 864 |
"step": 121
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.32221855397821064,
|
| 868 |
-
"grad_norm":
|
| 869 |
"learning_rate": 9.99996621149877e-05,
|
| 870 |
-
"loss":
|
| 871 |
"step": 122
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 0.3248596896665566,
|
| 875 |
-
"grad_norm":
|
| 876 |
"learning_rate": 9.999963070009755e-05,
|
| 877 |
-
"loss":
|
| 878 |
"step": 123
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.3275008253549026,
|
| 882 |
-
"grad_norm":
|
| 883 |
"learning_rate": 9.999959788899706e-05,
|
| 884 |
-
"loss":
|
| 885 |
"step": 124
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 0.3301419610432486,
|
| 889 |
-
"grad_norm":
|
| 890 |
"learning_rate": 9.999956368168719e-05,
|
| 891 |
-
"loss":
|
| 892 |
"step": 125
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 0.33278309673159456,
|
| 896 |
-
"grad_norm":
|
| 897 |
"learning_rate": 9.999952807816888e-05,
|
| 898 |
-
"loss":
|
| 899 |
"step": 126
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 0.33542423241994057,
|
| 903 |
-
"grad_norm":
|
| 904 |
"learning_rate": 9.99994910784431e-05,
|
| 905 |
-
"loss":
|
| 906 |
"step": 127
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 0.3380653681082866,
|
| 910 |
-
"grad_norm":
|
| 911 |
"learning_rate": 9.999945268251092e-05,
|
| 912 |
-
"loss":
|
| 913 |
"step": 128
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 0.34070650379663253,
|
| 917 |
-
"grad_norm":
|
| 918 |
"learning_rate": 9.999941289037338e-05,
|
| 919 |
-
"loss":
|
| 920 |
"step": 129
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 0.34334763948497854,
|
| 924 |
-
"grad_norm":
|
| 925 |
"learning_rate": 9.999937170203162e-05,
|
| 926 |
-
"loss":
|
| 927 |
"step": 130
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.34598877517332455,
|
| 931 |
-
"grad_norm":
|
| 932 |
"learning_rate": 9.999932911748678e-05,
|
| 933 |
-
"loss":
|
| 934 |
"step": 131
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 0.3486299108616705,
|
| 938 |
-
"grad_norm":
|
| 939 |
"learning_rate": 9.999928513674004e-05,
|
| 940 |
-
"loss":
|
| 941 |
"step": 132
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.3512710465500165,
|
| 945 |
-
"grad_norm":
|
| 946 |
"learning_rate": 9.999923975979262e-05,
|
| 947 |
-
"loss":
|
| 948 |
"step": 133
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 0.3539121822383625,
|
| 952 |
-
"grad_norm":
|
| 953 |
"learning_rate": 9.999919298664582e-05,
|
| 954 |
-
"loss":
|
| 955 |
"step": 134
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 0.35655331792670847,
|
| 959 |
-
"grad_norm":
|
| 960 |
"learning_rate": 9.999914481730092e-05,
|
| 961 |
-
"loss":
|
| 962 |
"step": 135
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.3591944536150545,
|
| 966 |
-
"grad_norm":
|
| 967 |
"learning_rate": 9.999909525175927e-05,
|
| 968 |
-
"loss":
|
| 969 |
"step": 136
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.3618355893034005,
|
| 973 |
-
"grad_norm":
|
| 974 |
"learning_rate": 9.999904429002225e-05,
|
| 975 |
-
"loss":
|
| 976 |
"step": 137
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 0.36447672499174644,
|
| 980 |
-
"grad_norm":
|
| 981 |
"learning_rate": 9.99989919320913e-05,
|
| 982 |
-
"loss":
|
| 983 |
"step": 138
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 0.36711786068009244,
|
| 987 |
-
"grad_norm":
|
| 988 |
"learning_rate": 9.999893817796786e-05,
|
| 989 |
-
"loss":
|
| 990 |
"step": 139
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.36975899636843845,
|
| 994 |
-
"grad_norm":
|
| 995 |
"learning_rate": 9.999888302765345e-05,
|
| 996 |
-
"loss":
|
| 997 |
"step": 140
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 0.3724001320567844,
|
| 1001 |
-
"grad_norm":
|
| 1002 |
"learning_rate": 9.99988264811496e-05,
|
| 1003 |
-
"loss":
|
| 1004 |
"step": 141
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 0.3750412677451304,
|
| 1008 |
-
"grad_norm":
|
| 1009 |
"learning_rate": 9.99987685384579e-05,
|
| 1010 |
-
"loss":
|
| 1011 |
"step": 142
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 0.3776824034334764,
|
| 1015 |
-
"grad_norm":
|
| 1016 |
"learning_rate": 9.999870919957996e-05,
|
| 1017 |
-
"loss":
|
| 1018 |
"step": 143
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 0.3803235391218224,
|
| 1022 |
-
"grad_norm":
|
| 1023 |
"learning_rate": 9.999864846451744e-05,
|
| 1024 |
-
"loss":
|
| 1025 |
"step": 144
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 0.3829646748101684,
|
| 1029 |
-
"grad_norm":
|
| 1030 |
"learning_rate": 9.999858633327201e-05,
|
| 1031 |
-
"loss":
|
| 1032 |
"step": 145
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 0.38560581049851433,
|
| 1036 |
-
"grad_norm":
|
| 1037 |
"learning_rate": 9.999852280584544e-05,
|
| 1038 |
-
"loss":
|
| 1039 |
"step": 146
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
"epoch": 0.38824694618686034,
|
| 1043 |
-
"grad_norm":
|
| 1044 |
"learning_rate": 9.999845788223949e-05,
|
| 1045 |
-
"loss":
|
| 1046 |
"step": 147
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 0.39088808187520635,
|
| 1050 |
-
"grad_norm":
|
| 1051 |
"learning_rate": 9.999839156245598e-05,
|
| 1052 |
-
"loss":
|
| 1053 |
"step": 148
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
"epoch": 0.3935292175635523,
|
| 1057 |
-
"grad_norm":
|
| 1058 |
"learning_rate": 9.999832384649674e-05,
|
| 1059 |
-
"loss":
|
| 1060 |
"step": 149
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 0.3961703532518983,
|
| 1064 |
-
"grad_norm":
|
| 1065 |
"learning_rate": 9.999825473436369e-05,
|
| 1066 |
-
"loss":
|
| 1067 |
"step": 150
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 0.3988114889402443,
|
| 1071 |
-
"grad_norm":
|
| 1072 |
"learning_rate": 9.999818422605875e-05,
|
| 1073 |
-
"loss":
|
| 1074 |
"step": 151
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 0.40145262462859027,
|
| 1078 |
-
"grad_norm":
|
| 1079 |
"learning_rate": 9.999811232158389e-05,
|
| 1080 |
-
"loss":
|
| 1081 |
"step": 152
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.4040937603169363,
|
| 1085 |
-
"grad_norm":
|
| 1086 |
"learning_rate": 9.999803902094109e-05,
|
| 1087 |
-
"loss":
|
| 1088 |
"step": 153
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 0.4067348960052823,
|
| 1092 |
-
"grad_norm":
|
| 1093 |
"learning_rate": 9.999796432413244e-05,
|
| 1094 |
-
"loss":
|
| 1095 |
"step": 154
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 0.40937603169362824,
|
| 1099 |
-
"grad_norm":
|
| 1100 |
"learning_rate": 9.999788823116001e-05,
|
| 1101 |
-
"loss":
|
| 1102 |
"step": 155
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 0.41201716738197425,
|
| 1106 |
-
"grad_norm":
|
| 1107 |
"learning_rate": 9.999781074202592e-05,
|
| 1108 |
-
"loss":
|
| 1109 |
"step": 156
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 0.41465830307032026,
|
| 1113 |
-
"grad_norm":
|
| 1114 |
"learning_rate": 9.999773185673232e-05,
|
| 1115 |
-
"loss":
|
| 1116 |
"step": 157
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 0.4172994387586662,
|
| 1120 |
-
"grad_norm":
|
| 1121 |
"learning_rate": 9.999765157528145e-05,
|
| 1122 |
-
"loss":
|
| 1123 |
"step": 158
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 0.4199405744470122,
|
| 1127 |
-
"grad_norm":
|
| 1128 |
"learning_rate": 9.99975698976755e-05,
|
| 1129 |
-
"loss":
|
| 1130 |
"step": 159
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 0.4225817101353582,
|
| 1134 |
-
"grad_norm":
|
| 1135 |
"learning_rate": 9.99974868239168e-05,
|
| 1136 |
-
"loss":
|
| 1137 |
"step": 160
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 0.4252228458237042,
|
| 1141 |
-
"grad_norm":
|
| 1142 |
"learning_rate": 9.999740235400765e-05,
|
| 1143 |
-
"loss":
|
| 1144 |
"step": 161
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 0.4278639815120502,
|
| 1148 |
-
"grad_norm":
|
| 1149 |
"learning_rate": 9.999731648795041e-05,
|
| 1150 |
-
"loss":
|
| 1151 |
"step": 162
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 0.4305051172003962,
|
| 1155 |
-
"grad_norm":
|
| 1156 |
"learning_rate": 9.999722922574749e-05,
|
| 1157 |
-
"loss":
|
| 1158 |
"step": 163
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 0.43314625288874214,
|
| 1162 |
-
"grad_norm":
|
| 1163 |
"learning_rate": 9.999714056740129e-05,
|
| 1164 |
-
"loss":
|
| 1165 |
"step": 164
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 0.43578738857708815,
|
| 1169 |
-
"grad_norm":
|
| 1170 |
"learning_rate": 9.999705051291432e-05,
|
| 1171 |
-
"loss":
|
| 1172 |
"step": 165
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 0.43842852426543416,
|
| 1176 |
-
"grad_norm":
|
| 1177 |
"learning_rate": 9.999695906228908e-05,
|
| 1178 |
-
"loss":
|
| 1179 |
"step": 166
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.4410696599537801,
|
| 1183 |
-
"grad_norm":
|
| 1184 |
"learning_rate": 9.999686621552813e-05,
|
| 1185 |
-
"loss":
|
| 1186 |
"step": 167
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 0.4437107956421261,
|
| 1190 |
-
"grad_norm":
|
| 1191 |
"learning_rate": 9.999677197263406e-05,
|
| 1192 |
-
"loss":
|
| 1193 |
"step": 168
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.44635193133047213,
|
| 1197 |
-
"grad_norm":
|
| 1198 |
"learning_rate": 9.999667633360952e-05,
|
| 1199 |
-
"loss":
|
| 1200 |
"step": 169
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.4489930670188181,
|
| 1204 |
-
"grad_norm":
|
| 1205 |
"learning_rate": 9.999657929845714e-05,
|
| 1206 |
-
"loss":
|
| 1207 |
"step": 170
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.4516342027071641,
|
| 1211 |
-
"grad_norm":
|
| 1212 |
"learning_rate": 9.999648086717966e-05,
|
| 1213 |
-
"loss":
|
| 1214 |
"step": 171
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.45427533839551004,
|
| 1218 |
-
"grad_norm":
|
| 1219 |
"learning_rate": 9.999638103977982e-05,
|
| 1220 |
-
"loss":
|
| 1221 |
"step": 172
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.45691647408385605,
|
| 1225 |
-
"grad_norm":
|
| 1226 |
"learning_rate": 9.999627981626041e-05,
|
| 1227 |
-
"loss":
|
| 1228 |
"step": 173
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.45955760977220206,
|
| 1232 |
-
"grad_norm":
|
| 1233 |
"learning_rate": 9.999617719662426e-05,
|
| 1234 |
-
"loss":
|
| 1235 |
"step": 174
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.462198745460548,
|
| 1239 |
-
"grad_norm":
|
| 1240 |
"learning_rate": 9.999607318087423e-05,
|
| 1241 |
-
"loss":
|
| 1242 |
"step": 175
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.464839881148894,
|
| 1246 |
-
"grad_norm":
|
| 1247 |
"learning_rate": 9.999596776901322e-05,
|
| 1248 |
-
"loss":
|
| 1249 |
"step": 176
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.46748101683724,
|
| 1253 |
-
"grad_norm":
|
| 1254 |
"learning_rate": 9.999586096104419e-05,
|
| 1255 |
-
"loss":
|
| 1256 |
"step": 177
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.470122152525586,
|
| 1260 |
-
"grad_norm":
|
| 1261 |
"learning_rate": 9.99957527569701e-05,
|
| 1262 |
-
"loss":
|
| 1263 |
"step": 178
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.472763288213932,
|
| 1267 |
-
"grad_norm":
|
| 1268 |
"learning_rate": 9.999564315679398e-05,
|
| 1269 |
-
"loss":
|
| 1270 |
"step": 179
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.475404423902278,
|
| 1274 |
-
"grad_norm":
|
| 1275 |
"learning_rate": 9.99955321605189e-05,
|
| 1276 |
-
"loss":
|
| 1277 |
"step": 180
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.47804555959062395,
|
| 1281 |
-
"grad_norm":
|
| 1282 |
"learning_rate": 9.999541976814796e-05,
|
| 1283 |
-
"loss":
|
| 1284 |
"step": 181
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.48068669527896996,
|
| 1288 |
-
"grad_norm":
|
| 1289 |
"learning_rate": 9.999530597968428e-05,
|
| 1290 |
-
"loss":
|
| 1291 |
"step": 182
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.48332783096731596,
|
| 1295 |
-
"grad_norm":
|
| 1296 |
"learning_rate": 9.999519079513107e-05,
|
| 1297 |
-
"loss":
|
| 1298 |
"step": 183
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.4859689666556619,
|
| 1302 |
-
"grad_norm":
|
| 1303 |
"learning_rate": 9.999507421449151e-05,
|
| 1304 |
-
"loss":
|
| 1305 |
"step": 184
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.4886101023440079,
|
| 1309 |
-
"grad_norm":
|
| 1310 |
"learning_rate": 9.999495623776886e-05,
|
| 1311 |
-
"loss":
|
| 1312 |
"step": 185
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.49125123803235393,
|
| 1316 |
-
"grad_norm":
|
| 1317 |
"learning_rate": 9.999483686496645e-05,
|
| 1318 |
-
"loss":
|
| 1319 |
"step": 186
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.4938923737206999,
|
| 1323 |
-
"grad_norm":
|
| 1324 |
"learning_rate": 9.999471609608757e-05,
|
| 1325 |
-
"loss":
|
| 1326 |
"step": 187
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.4965335094090459,
|
| 1330 |
-
"grad_norm":
|
| 1331 |
"learning_rate": 9.999459393113561e-05,
|
| 1332 |
-
"loss":
|
| 1333 |
"step": 188
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.4991746450973919,
|
| 1337 |
-
"grad_norm":
|
| 1338 |
"learning_rate": 9.9994470370114e-05,
|
| 1339 |
-
"loss":
|
| 1340 |
"step": 189
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.5018157807857379,
|
| 1344 |
-
"grad_norm":
|
| 1345 |
"learning_rate": 9.999434541302616e-05,
|
| 1346 |
-
"loss":
|
| 1347 |
"step": 190
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 0.5044569164740839,
|
| 1351 |
-
"grad_norm":
|
| 1352 |
"learning_rate": 9.99942190598756e-05,
|
| 1353 |
-
"loss":
|
| 1354 |
"step": 191
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 0.5070980521624299,
|
| 1358 |
-
"grad_norm":
|
| 1359 |
"learning_rate": 9.999409131066583e-05,
|
| 1360 |
-
"loss":
|
| 1361 |
"step": 192
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 0.5097391878507759,
|
| 1365 |
-
"grad_norm":
|
| 1366 |
"learning_rate": 9.999396216540044e-05,
|
| 1367 |
-
"loss":
|
| 1368 |
"step": 193
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 0.5123803235391218,
|
| 1372 |
-
"grad_norm":
|
| 1373 |
"learning_rate": 9.999383162408304e-05,
|
| 1374 |
-
"loss":
|
| 1375 |
"step": 194
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 0.5150214592274678,
|
| 1379 |
-
"grad_norm":
|
| 1380 |
"learning_rate": 9.999369968671723e-05,
|
| 1381 |
-
"loss":
|
| 1382 |
"step": 195
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 0.5176625949158138,
|
| 1386 |
-
"grad_norm":
|
| 1387 |
"learning_rate": 9.999356635330674e-05,
|
| 1388 |
-
"loss":
|
| 1389 |
"step": 196
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 0.5203037306041598,
|
| 1393 |
-
"grad_norm":
|
| 1394 |
"learning_rate": 9.999343162385529e-05,
|
| 1395 |
-
"loss":
|
| 1396 |
"step": 197
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 0.5229448662925058,
|
| 1400 |
-
"grad_norm":
|
| 1401 |
"learning_rate": 9.99932954983666e-05,
|
| 1402 |
-
"loss":
|
| 1403 |
"step": 198
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 0.5255860019808518,
|
| 1407 |
-
"grad_norm":
|
| 1408 |
"learning_rate": 9.999315797684451e-05,
|
| 1409 |
-
"loss":
|
| 1410 |
"step": 199
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 0.5282271376691977,
|
| 1414 |
-
"grad_norm":
|
| 1415 |
"learning_rate": 9.999301905929286e-05,
|
| 1416 |
-
"loss":
|
| 1417 |
"step": 200
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 0.5282271376691977,
|
| 1421 |
-
"eval_loss":
|
| 1422 |
-
"eval_runtime": 2.
|
| 1423 |
-
"eval_samples_per_second":
|
| 1424 |
-
"eval_steps_per_second":
|
| 1425 |
"step": 200
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 0.5308682733575437,
|
| 1429 |
-
"grad_norm":
|
| 1430 |
"learning_rate": 9.999287874571552e-05,
|
| 1431 |
-
"loss":
|
| 1432 |
"step": 201
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 0.5335094090458897,
|
| 1436 |
-
"grad_norm":
|
| 1437 |
"learning_rate": 9.99927370361164e-05,
|
| 1438 |
-
"loss":
|
| 1439 |
"step": 202
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 0.5361505447342357,
|
| 1443 |
-
"grad_norm":
|
| 1444 |
"learning_rate": 9.999259393049947e-05,
|
| 1445 |
-
"loss":
|
| 1446 |
"step": 203
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 0.5387916804225817,
|
| 1450 |
-
"grad_norm":
|
| 1451 |
"learning_rate": 9.999244942886871e-05,
|
| 1452 |
-
"loss":
|
| 1453 |
"step": 204
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 0.5414328161109278,
|
| 1457 |
-
"grad_norm":
|
| 1458 |
"learning_rate": 9.999230353122819e-05,
|
| 1459 |
-
"loss":
|
| 1460 |
"step": 205
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 0.5440739517992736,
|
| 1464 |
-
"grad_norm":
|
| 1465 |
"learning_rate": 9.999215623758194e-05,
|
| 1466 |
-
"loss":
|
| 1467 |
"step": 206
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.5467150874876197,
|
| 1471 |
-
"grad_norm":
|
| 1472 |
"learning_rate": 9.99920075479341e-05,
|
| 1473 |
-
"loss":
|
| 1474 |
"step": 207
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 0.5493562231759657,
|
| 1478 |
-
"grad_norm":
|
| 1479 |
"learning_rate": 9.999185746228882e-05,
|
| 1480 |
-
"loss":
|
| 1481 |
"step": 208
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.5519973588643117,
|
| 1485 |
-
"grad_norm":
|
| 1486 |
"learning_rate": 9.999170598065028e-05,
|
| 1487 |
-
"loss":
|
| 1488 |
"step": 209
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 0.5546384945526577,
|
| 1492 |
-
"grad_norm":
|
| 1493 |
"learning_rate": 9.999155310302273e-05,
|
| 1494 |
-
"loss":
|
| 1495 |
"step": 210
|
| 1496 |
},
|
| 1497 |
{
|
| 1498 |
"epoch": 0.5572796302410036,
|
| 1499 |
-
"grad_norm":
|
| 1500 |
"learning_rate": 9.999139882941043e-05,
|
| 1501 |
-
"loss":
|
| 1502 |
"step": 211
|
| 1503 |
},
|
| 1504 |
{
|
| 1505 |
"epoch": 0.5599207659293496,
|
| 1506 |
-
"grad_norm":
|
| 1507 |
"learning_rate": 9.999124315981766e-05,
|
| 1508 |
-
"loss":
|
| 1509 |
"step": 212
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
"epoch": 0.5625619016176956,
|
| 1513 |
-
"grad_norm":
|
| 1514 |
"learning_rate": 9.999108609424881e-05,
|
| 1515 |
-
"loss":
|
| 1516 |
"step": 213
|
| 1517 |
},
|
| 1518 |
{
|
| 1519 |
"epoch": 0.5652030373060416,
|
| 1520 |
-
"grad_norm":
|
| 1521 |
"learning_rate": 9.999092763270823e-05,
|
| 1522 |
-
"loss":
|
| 1523 |
"step": 214
|
| 1524 |
},
|
| 1525 |
{
|
| 1526 |
"epoch": 0.5678441729943876,
|
| 1527 |
-
"grad_norm":
|
| 1528 |
"learning_rate": 9.999076777520037e-05,
|
| 1529 |
-
"loss":
|
| 1530 |
"step": 215
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 0.5704853086827336,
|
| 1534 |
-
"grad_norm":
|
| 1535 |
"learning_rate": 9.99906065217297e-05,
|
| 1536 |
-
"loss":
|
| 1537 |
"step": 216
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
"epoch": 0.5731264443710795,
|
| 1541 |
-
"grad_norm":
|
| 1542 |
"learning_rate": 9.99904438723007e-05,
|
| 1543 |
-
"loss":
|
| 1544 |
"step": 217
|
| 1545 |
},
|
| 1546 |
{
|
| 1547 |
"epoch": 0.5757675800594255,
|
| 1548 |
-
"grad_norm":
|
| 1549 |
"learning_rate": 9.999027982691793e-05,
|
| 1550 |
-
"loss":
|
| 1551 |
"step": 218
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
"epoch": 0.5784087157477715,
|
| 1555 |
-
"grad_norm":
|
| 1556 |
"learning_rate": 9.999011438558595e-05,
|
| 1557 |
-
"loss":
|
| 1558 |
"step": 219
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
"epoch": 0.5810498514361175,
|
| 1562 |
-
"grad_norm":
|
| 1563 |
"learning_rate": 9.99899475483094e-05,
|
| 1564 |
-
"loss":
|
| 1565 |
"step": 220
|
| 1566 |
},
|
| 1567 |
{
|
| 1568 |
"epoch": 0.5836909871244635,
|
| 1569 |
-
"grad_norm":
|
| 1570 |
"learning_rate": 9.998977931509291e-05,
|
| 1571 |
-
"loss":
|
| 1572 |
"step": 221
|
| 1573 |
},
|
| 1574 |
{
|
| 1575 |
"epoch": 0.5863321228128096,
|
| 1576 |
-
"grad_norm":
|
| 1577 |
"learning_rate": 9.998960968594121e-05,
|
| 1578 |
-
"loss":
|
| 1579 |
"step": 222
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
"epoch": 0.5889732585011554,
|
| 1583 |
-
"grad_norm":
|
| 1584 |
"learning_rate": 9.998943866085903e-05,
|
| 1585 |
-
"loss":
|
| 1586 |
"step": 223
|
| 1587 |
},
|
| 1588 |
{
|
| 1589 |
"epoch": 0.5916143941895015,
|
| 1590 |
-
"grad_norm":
|
| 1591 |
"learning_rate": 9.998926623985114e-05,
|
| 1592 |
-
"loss":
|
| 1593 |
"step": 224
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 0.5942555298778475,
|
| 1597 |
-
"grad_norm":
|
| 1598 |
"learning_rate": 9.998909242292235e-05,
|
| 1599 |
-
"loss":
|
| 1600 |
"step": 225
|
| 1601 |
},
|
| 1602 |
{
|
| 1603 |
"epoch": 0.5968966655661935,
|
| 1604 |
-
"grad_norm":
|
| 1605 |
"learning_rate": 9.998891721007752e-05,
|
| 1606 |
-
"loss":
|
| 1607 |
"step": 226
|
| 1608 |
},
|
| 1609 |
{
|
| 1610 |
"epoch": 0.5995378012545395,
|
| 1611 |
-
"grad_norm":
|
| 1612 |
"learning_rate": 9.998874060132155e-05,
|
| 1613 |
-
"loss":
|
| 1614 |
"step": 227
|
| 1615 |
},
|
| 1616 |
{
|
| 1617 |
"epoch": 0.6021789369428855,
|
| 1618 |
-
"grad_norm":
|
| 1619 |
"learning_rate": 9.998856259665936e-05,
|
| 1620 |
-
"loss":
|
| 1621 |
"step": 228
|
| 1622 |
},
|
| 1623 |
{
|
| 1624 |
"epoch": 0.6048200726312314,
|
| 1625 |
-
"grad_norm":
|
| 1626 |
"learning_rate": 9.998838319609591e-05,
|
| 1627 |
-
"loss":
|
| 1628 |
"step": 229
|
| 1629 |
},
|
| 1630 |
{
|
| 1631 |
"epoch": 0.6074612083195774,
|
| 1632 |
-
"grad_norm":
|
| 1633 |
"learning_rate": 9.998820239963624e-05,
|
| 1634 |
-
"loss":
|
| 1635 |
"step": 230
|
| 1636 |
},
|
| 1637 |
{
|
| 1638 |
"epoch": 0.6101023440079234,
|
| 1639 |
-
"grad_norm":
|
| 1640 |
"learning_rate": 9.998802020728537e-05,
|
| 1641 |
-
"loss":
|
| 1642 |
"step": 231
|
| 1643 |
},
|
| 1644 |
{
|
| 1645 |
"epoch": 0.6127434796962694,
|
| 1646 |
-
"grad_norm":
|
| 1647 |
"learning_rate": 9.998783661904843e-05,
|
| 1648 |
-
"loss":
|
| 1649 |
"step": 232
|
| 1650 |
},
|
| 1651 |
{
|
| 1652 |
"epoch": 0.6153846153846154,
|
| 1653 |
-
"grad_norm":
|
| 1654 |
"learning_rate": 9.99876516349305e-05,
|
| 1655 |
-
"loss":
|
| 1656 |
"step": 233
|
| 1657 |
},
|
| 1658 |
{
|
| 1659 |
"epoch": 0.6180257510729614,
|
| 1660 |
-
"grad_norm":
|
| 1661 |
"learning_rate": 9.998746525493674e-05,
|
| 1662 |
-
"loss":
|
| 1663 |
"step": 234
|
| 1664 |
},
|
| 1665 |
{
|
| 1666 |
"epoch": 0.6206668867613073,
|
| 1667 |
-
"grad_norm":
|
| 1668 |
"learning_rate": 9.99872774790724e-05,
|
| 1669 |
-
"loss":
|
| 1670 |
"step": 235
|
| 1671 |
},
|
| 1672 |
{
|
| 1673 |
"epoch": 0.6233080224496533,
|
| 1674 |
-
"grad_norm":
|
| 1675 |
"learning_rate": 9.99870883073427e-05,
|
| 1676 |
-
"loss":
|
| 1677 |
"step": 236
|
| 1678 |
},
|
| 1679 |
{
|
| 1680 |
"epoch": 0.6259491581379993,
|
| 1681 |
-
"grad_norm":
|
| 1682 |
"learning_rate": 9.998689773975291e-05,
|
| 1683 |
-
"loss":
|
| 1684 |
"step": 237
|
| 1685 |
},
|
| 1686 |
{
|
| 1687 |
"epoch": 0.6285902938263453,
|
| 1688 |
-
"grad_norm":
|
| 1689 |
"learning_rate": 9.998670577630838e-05,
|
| 1690 |
-
"loss":
|
| 1691 |
"step": 238
|
| 1692 |
},
|
| 1693 |
{
|
| 1694 |
"epoch": 0.6312314295146914,
|
| 1695 |
-
"grad_norm":
|
| 1696 |
"learning_rate": 9.998651241701445e-05,
|
| 1697 |
-
"loss":
|
| 1698 |
"step": 239
|
| 1699 |
},
|
| 1700 |
{
|
| 1701 |
"epoch": 0.6338725652030373,
|
| 1702 |
-
"grad_norm":
|
| 1703 |
"learning_rate": 9.998631766187651e-05,
|
| 1704 |
-
"loss":
|
| 1705 |
"step": 240
|
| 1706 |
},
|
| 1707 |
{
|
| 1708 |
"epoch": 0.6365137008913833,
|
| 1709 |
-
"grad_norm":
|
| 1710 |
"learning_rate": 9.998612151090003e-05,
|
| 1711 |
-
"loss":
|
| 1712 |
"step": 241
|
| 1713 |
},
|
| 1714 |
{
|
| 1715 |
"epoch": 0.6391548365797293,
|
| 1716 |
-
"grad_norm":
|
| 1717 |
"learning_rate": 9.998592396409047e-05,
|
| 1718 |
-
"loss":
|
| 1719 |
"step": 242
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
"epoch": 0.6417959722680753,
|
| 1723 |
-
"grad_norm":
|
| 1724 |
"learning_rate": 9.998572502145334e-05,
|
| 1725 |
-
"loss":
|
| 1726 |
"step": 243
|
| 1727 |
},
|
| 1728 |
{
|
| 1729 |
"epoch": 0.6444371079564213,
|
| 1730 |
-
"grad_norm":
|
| 1731 |
"learning_rate": 9.998552468299421e-05,
|
| 1732 |
-
"loss":
|
| 1733 |
"step": 244
|
| 1734 |
},
|
| 1735 |
{
|
| 1736 |
"epoch": 0.6470782436447673,
|
| 1737 |
-
"grad_norm":
|
| 1738 |
"learning_rate": 9.998532294871866e-05,
|
| 1739 |
-
"loss":
|
| 1740 |
"step": 245
|
| 1741 |
},
|
| 1742 |
{
|
| 1743 |
"epoch": 0.6497193793331132,
|
| 1744 |
-
"grad_norm":
|
| 1745 |
"learning_rate": 9.998511981863232e-05,
|
| 1746 |
-
"loss":
|
| 1747 |
"step": 246
|
| 1748 |
},
|
| 1749 |
{
|
| 1750 |
"epoch": 0.6523605150214592,
|
| 1751 |
-
"grad_norm":
|
| 1752 |
"learning_rate": 9.998491529274089e-05,
|
| 1753 |
-
"loss":
|
| 1754 |
"step": 247
|
| 1755 |
},
|
| 1756 |
{
|
| 1757 |
"epoch": 0.6550016507098052,
|
| 1758 |
-
"grad_norm":
|
| 1759 |
"learning_rate": 9.998470937105006e-05,
|
| 1760 |
-
"loss":
|
| 1761 |
"step": 248
|
| 1762 |
},
|
| 1763 |
{
|
| 1764 |
"epoch": 0.6576427863981512,
|
| 1765 |
-
"grad_norm":
|
| 1766 |
"learning_rate": 9.998450205356557e-05,
|
| 1767 |
-
"loss":
|
| 1768 |
"step": 249
|
| 1769 |
},
|
| 1770 |
{
|
| 1771 |
"epoch": 0.6602839220864972,
|
| 1772 |
-
"grad_norm":
|
| 1773 |
"learning_rate": 9.998429334029323e-05,
|
| 1774 |
-
"loss":
|
| 1775 |
"step": 250
|
| 1776 |
},
|
| 1777 |
{
|
| 1778 |
"epoch": 0.6629250577748432,
|
| 1779 |
-
"grad_norm":
|
| 1780 |
"learning_rate": 9.998408323123887e-05,
|
| 1781 |
-
"loss":
|
| 1782 |
"step": 251
|
| 1783 |
},
|
| 1784 |
{
|
| 1785 |
"epoch": 0.6655661934631891,
|
| 1786 |
-
"grad_norm":
|
| 1787 |
"learning_rate": 9.998387172640834e-05,
|
| 1788 |
-
"loss":
|
| 1789 |
"step": 252
|
| 1790 |
},
|
| 1791 |
{
|
| 1792 |
"epoch": 0.6682073291515351,
|
| 1793 |
-
"grad_norm":
|
| 1794 |
"learning_rate": 9.998365882580756e-05,
|
| 1795 |
-
"loss":
|
| 1796 |
"step": 253
|
| 1797 |
},
|
| 1798 |
{
|
| 1799 |
"epoch": 0.6708484648398811,
|
| 1800 |
-
"grad_norm":
|
| 1801 |
"learning_rate": 9.998344452944247e-05,
|
| 1802 |
-
"loss":
|
| 1803 |
"step": 254
|
| 1804 |
},
|
| 1805 |
{
|
| 1806 |
"epoch": 0.6734896005282272,
|
| 1807 |
-
"grad_norm":
|
| 1808 |
"learning_rate": 9.998322883731903e-05,
|
| 1809 |
-
"loss":
|
| 1810 |
"step": 255
|
| 1811 |
},
|
| 1812 |
{
|
| 1813 |
"epoch": 0.6761307362165732,
|
| 1814 |
-
"grad_norm":
|
| 1815 |
"learning_rate": 9.998301174944332e-05,
|
| 1816 |
-
"loss":
|
| 1817 |
"step": 256
|
| 1818 |
},
|
| 1819 |
{
|
| 1820 |
"epoch": 0.6787718719049192,
|
| 1821 |
-
"grad_norm":
|
| 1822 |
"learning_rate": 9.998279326582134e-05,
|
| 1823 |
-
"loss":
|
| 1824 |
"step": 257
|
| 1825 |
},
|
| 1826 |
{
|
| 1827 |
"epoch": 0.6814130075932651,
|
| 1828 |
-
"grad_norm":
|
| 1829 |
"learning_rate": 9.998257338645924e-05,
|
| 1830 |
-
"loss":
|
| 1831 |
"step": 258
|
| 1832 |
},
|
| 1833 |
{
|
| 1834 |
"epoch": 0.6840541432816111,
|
| 1835 |
-
"grad_norm":
|
| 1836 |
"learning_rate": 9.998235211136312e-05,
|
| 1837 |
-
"loss":
|
| 1838 |
"step": 259
|
| 1839 |
},
|
| 1840 |
{
|
| 1841 |
"epoch": 0.6866952789699571,
|
| 1842 |
-
"grad_norm":
|
| 1843 |
"learning_rate": 9.99821294405392e-05,
|
| 1844 |
-
"loss":
|
| 1845 |
"step": 260
|
| 1846 |
},
|
| 1847 |
{
|
| 1848 |
"epoch": 0.6893364146583031,
|
| 1849 |
-
"grad_norm":
|
| 1850 |
"learning_rate": 9.998190537399366e-05,
|
| 1851 |
-
"loss":
|
| 1852 |
"step": 261
|
| 1853 |
},
|
| 1854 |
{
|
| 1855 |
"epoch": 0.6919775503466491,
|
| 1856 |
-
"grad_norm":
|
| 1857 |
"learning_rate": 9.998167991173277e-05,
|
| 1858 |
-
"loss":
|
| 1859 |
"step": 262
|
| 1860 |
},
|
| 1861 |
{
|
| 1862 |
"epoch": 0.6946186860349951,
|
| 1863 |
-
"grad_norm":
|
| 1864 |
"learning_rate": 9.998145305376286e-05,
|
| 1865 |
-
"loss":
|
| 1866 |
"step": 263
|
| 1867 |
},
|
| 1868 |
{
|
| 1869 |
"epoch": 0.697259821723341,
|
| 1870 |
-
"grad_norm":
|
| 1871 |
"learning_rate": 9.99812248000902e-05,
|
| 1872 |
-
"loss":
|
| 1873 |
"step": 264
|
| 1874 |
},
|
| 1875 |
{
|
| 1876 |
"epoch": 0.699900957411687,
|
| 1877 |
-
"grad_norm":
|
| 1878 |
"learning_rate": 9.998099515072122e-05,
|
| 1879 |
-
"loss":
|
| 1880 |
"step": 265
|
| 1881 |
},
|
| 1882 |
{
|
| 1883 |
"epoch": 0.702542093100033,
|
| 1884 |
-
"grad_norm":
|
| 1885 |
"learning_rate": 9.998076410566229e-05,
|
| 1886 |
-
"loss":
|
| 1887 |
"step": 266
|
| 1888 |
},
|
| 1889 |
{
|
| 1890 |
"epoch": 0.705183228788379,
|
| 1891 |
-
"grad_norm":
|
| 1892 |
"learning_rate": 9.99805316649199e-05,
|
| 1893 |
-
"loss":
|
| 1894 |
"step": 267
|
| 1895 |
},
|
| 1896 |
{
|
| 1897 |
"epoch": 0.707824364476725,
|
| 1898 |
-
"grad_norm":
|
| 1899 |
"learning_rate": 9.998029782850051e-05,
|
| 1900 |
-
"loss":
|
| 1901 |
"step": 268
|
| 1902 |
},
|
| 1903 |
{
|
| 1904 |
"epoch": 0.7104655001650709,
|
| 1905 |
-
"grad_norm":
|
| 1906 |
"learning_rate": 9.998006259641068e-05,
|
| 1907 |
-
"loss":
|
| 1908 |
"step": 269
|
| 1909 |
},
|
| 1910 |
{
|
| 1911 |
"epoch": 0.7131066358534169,
|
| 1912 |
-
"grad_norm":
|
| 1913 |
"learning_rate": 9.997982596865695e-05,
|
| 1914 |
-
"loss":
|
| 1915 |
"step": 270
|
| 1916 |
},
|
| 1917 |
{
|
| 1918 |
"epoch": 0.715747771541763,
|
| 1919 |
-
"grad_norm":
|
| 1920 |
"learning_rate": 9.997958794524594e-05,
|
| 1921 |
-
"loss":
|
| 1922 |
"step": 271
|
| 1923 |
},
|
| 1924 |
{
|
| 1925 |
"epoch": 0.718388907230109,
|
| 1926 |
-
"grad_norm":
|
| 1927 |
"learning_rate": 9.99793485261843e-05,
|
| 1928 |
-
"loss":
|
| 1929 |
"step": 272
|
| 1930 |
},
|
| 1931 |
{
|
| 1932 |
"epoch": 0.721030042918455,
|
| 1933 |
-
"grad_norm":
|
| 1934 |
"learning_rate": 9.997910771147872e-05,
|
| 1935 |
-
"loss":
|
| 1936 |
"step": 273
|
| 1937 |
},
|
| 1938 |
{
|
| 1939 |
"epoch": 0.723671178606801,
|
| 1940 |
-
"grad_norm":
|
| 1941 |
"learning_rate": 9.99788655011359e-05,
|
| 1942 |
-
"loss":
|
| 1943 |
"step": 274
|
| 1944 |
},
|
| 1945 |
{
|
| 1946 |
"epoch": 0.7263123142951469,
|
| 1947 |
-
"grad_norm":
|
| 1948 |
"learning_rate": 9.997862189516263e-05,
|
| 1949 |
-
"loss":
|
| 1950 |
"step": 275
|
| 1951 |
},
|
| 1952 |
{
|
| 1953 |
"epoch": 0.7289534499834929,
|
| 1954 |
-
"grad_norm":
|
| 1955 |
"learning_rate": 9.99783768935657e-05,
|
| 1956 |
-
"loss":
|
| 1957 |
"step": 276
|
| 1958 |
},
|
| 1959 |
{
|
| 1960 |
"epoch": 0.7315945856718389,
|
| 1961 |
-
"grad_norm":
|
| 1962 |
"learning_rate": 9.997813049635195e-05,
|
| 1963 |
-
"loss":
|
| 1964 |
"step": 277
|
| 1965 |
},
|
| 1966 |
{
|
| 1967 |
"epoch": 0.7342357213601849,
|
| 1968 |
-
"grad_norm":
|
| 1969 |
"learning_rate": 9.997788270352827e-05,
|
| 1970 |
-
"loss":
|
| 1971 |
"step": 278
|
| 1972 |
},
|
| 1973 |
{
|
| 1974 |
"epoch": 0.7368768570485309,
|
| 1975 |
-
"grad_norm":
|
| 1976 |
"learning_rate": 9.997763351510157e-05,
|
| 1977 |
-
"loss":
|
| 1978 |
"step": 279
|
| 1979 |
},
|
| 1980 |
{
|
| 1981 |
"epoch": 0.7395179927368769,
|
| 1982 |
-
"grad_norm":
|
| 1983 |
"learning_rate": 9.997738293107881e-05,
|
| 1984 |
-
"loss":
|
| 1985 |
"step": 280
|
| 1986 |
},
|
| 1987 |
{
|
| 1988 |
"epoch": 0.7421591284252228,
|
| 1989 |
-
"grad_norm":
|
| 1990 |
"learning_rate": 9.9977130951467e-05,
|
| 1991 |
-
"loss":
|
| 1992 |
"step": 281
|
| 1993 |
},
|
| 1994 |
{
|
| 1995 |
"epoch": 0.7448002641135688,
|
| 1996 |
-
"grad_norm":
|
| 1997 |
"learning_rate": 9.997687757627316e-05,
|
| 1998 |
-
"loss":
|
| 1999 |
"step": 282
|
| 2000 |
},
|
| 2001 |
{
|
| 2002 |
"epoch": 0.7474413998019148,
|
| 2003 |
-
"grad_norm":
|
| 2004 |
"learning_rate": 9.997662280550437e-05,
|
| 2005 |
-
"loss":
|
| 2006 |
"step": 283
|
| 2007 |
},
|
| 2008 |
{
|
| 2009 |
"epoch": 0.7500825354902608,
|
| 2010 |
-
"grad_norm":
|
| 2011 |
"learning_rate": 9.997636663916776e-05,
|
| 2012 |
-
"loss":
|
| 2013 |
"step": 284
|
| 2014 |
},
|
| 2015 |
{
|
| 2016 |
"epoch": 0.7527236711786068,
|
| 2017 |
-
"grad_norm":
|
| 2018 |
"learning_rate": 9.997610907727046e-05,
|
| 2019 |
-
"loss":
|
| 2020 |
"step": 285
|
| 2021 |
},
|
| 2022 |
{
|
| 2023 |
"epoch": 0.7553648068669528,
|
| 2024 |
-
"grad_norm":
|
| 2025 |
"learning_rate": 9.997585011981966e-05,
|
| 2026 |
-
"loss":
|
| 2027 |
"step": 286
|
| 2028 |
},
|
| 2029 |
{
|
| 2030 |
"epoch": 0.7580059425552987,
|
| 2031 |
-
"grad_norm":
|
| 2032 |
"learning_rate": 9.997558976682262e-05,
|
| 2033 |
-
"loss":
|
| 2034 |
"step": 287
|
| 2035 |
},
|
| 2036 |
{
|
| 2037 |
"epoch": 0.7606470782436447,
|
| 2038 |
-
"grad_norm":
|
| 2039 |
"learning_rate": 9.997532801828658e-05,
|
| 2040 |
-
"loss":
|
| 2041 |
"step": 288
|
| 2042 |
},
|
| 2043 |
{
|
| 2044 |
"epoch": 0.7632882139319908,
|
| 2045 |
-
"grad_norm":
|
| 2046 |
"learning_rate": 9.997506487421888e-05,
|
| 2047 |
-
"loss":
|
| 2048 |
"step": 289
|
| 2049 |
},
|
| 2050 |
{
|
| 2051 |
"epoch": 0.7659293496203368,
|
| 2052 |
-
"grad_norm":
|
| 2053 |
"learning_rate": 9.997480033462683e-05,
|
| 2054 |
-
"loss":
|
| 2055 |
"step": 290
|
| 2056 |
},
|
| 2057 |
{
|
| 2058 |
"epoch": 0.7685704853086828,
|
| 2059 |
-
"grad_norm":
|
| 2060 |
"learning_rate": 9.997453439951784e-05,
|
| 2061 |
-
"loss":
|
| 2062 |
"step": 291
|
| 2063 |
},
|
| 2064 |
{
|
| 2065 |
"epoch": 0.7712116209970287,
|
| 2066 |
-
"grad_norm":
|
| 2067 |
"learning_rate": 9.997426706889935e-05,
|
| 2068 |
-
"loss":
|
| 2069 |
"step": 292
|
| 2070 |
},
|
| 2071 |
{
|
| 2072 |
"epoch": 0.7738527566853747,
|
| 2073 |
-
"grad_norm":
|
| 2074 |
"learning_rate": 9.997399834277878e-05,
|
| 2075 |
-
"loss":
|
| 2076 |
"step": 293
|
| 2077 |
},
|
| 2078 |
{
|
| 2079 |
"epoch": 0.7764938923737207,
|
| 2080 |
-
"grad_norm":
|
| 2081 |
"learning_rate": 9.997372822116368e-05,
|
| 2082 |
-
"loss":
|
| 2083 |
"step": 294
|
| 2084 |
},
|
| 2085 |
{
|
| 2086 |
"epoch": 0.7791350280620667,
|
| 2087 |
-
"grad_norm":
|
| 2088 |
"learning_rate": 9.99734567040616e-05,
|
| 2089 |
-
"loss":
|
| 2090 |
"step": 295
|
| 2091 |
},
|
| 2092 |
{
|
| 2093 |
"epoch": 0.7817761637504127,
|
| 2094 |
-
"grad_norm":
|
| 2095 |
"learning_rate": 9.997318379148007e-05,
|
| 2096 |
-
"loss":
|
| 2097 |
"step": 296
|
| 2098 |
},
|
| 2099 |
{
|
| 2100 |
"epoch": 0.7844172994387587,
|
| 2101 |
-
"grad_norm":
|
| 2102 |
"learning_rate": 9.997290948342673e-05,
|
| 2103 |
-
"loss":
|
| 2104 |
"step": 297
|
| 2105 |
},
|
| 2106 |
{
|
| 2107 |
"epoch": 0.7870584351271046,
|
| 2108 |
-
"grad_norm":
|
| 2109 |
"learning_rate": 9.997263377990926e-05,
|
| 2110 |
-
"loss":
|
| 2111 |
"step": 298
|
| 2112 |
},
|
| 2113 |
{
|
| 2114 |
"epoch": 0.7896995708154506,
|
| 2115 |
-
"grad_norm":
|
| 2116 |
"learning_rate": 9.997235668093535e-05,
|
| 2117 |
-
"loss":
|
| 2118 |
"step": 299
|
| 2119 |
},
|
| 2120 |
{
|
| 2121 |
"epoch": 0.7923407065037966,
|
| 2122 |
-
"grad_norm":
|
| 2123 |
"learning_rate": 9.997207818651274e-05,
|
| 2124 |
-
"loss":
|
| 2125 |
"step": 300
|
| 2126 |
},
|
| 2127 |
{
|
| 2128 |
"epoch": 0.7949818421921426,
|
| 2129 |
-
"grad_norm":
|
| 2130 |
"learning_rate": 9.997179829664918e-05,
|
| 2131 |
-
"loss":
|
| 2132 |
"step": 301
|
| 2133 |
},
|
| 2134 |
{
|
| 2135 |
"epoch": 0.7976229778804886,
|
| 2136 |
-
"grad_norm":
|
| 2137 |
"learning_rate": 9.997151701135253e-05,
|
| 2138 |
-
"loss":
|
| 2139 |
"step": 302
|
| 2140 |
},
|
| 2141 |
{
|
| 2142 |
"epoch": 0.8002641135688346,
|
| 2143 |
-
"grad_norm":
|
| 2144 |
"learning_rate": 9.997123433063062e-05,
|
| 2145 |
-
"loss":
|
| 2146 |
"step": 303
|
| 2147 |
},
|
| 2148 |
{
|
| 2149 |
"epoch": 0.8029052492571805,
|
| 2150 |
-
"grad_norm":
|
| 2151 |
"learning_rate": 9.997095025449134e-05,
|
| 2152 |
-
"loss":
|
| 2153 |
"step": 304
|
| 2154 |
},
|
| 2155 |
{
|
| 2156 |
"epoch": 0.8055463849455266,
|
| 2157 |
-
"grad_norm":
|
| 2158 |
"learning_rate": 9.997066478294262e-05,
|
| 2159 |
-
"loss":
|
| 2160 |
"step": 305
|
| 2161 |
},
|
| 2162 |
{
|
| 2163 |
"epoch": 0.8081875206338726,
|
| 2164 |
-
"grad_norm":
|
| 2165 |
"learning_rate": 9.997037791599245e-05,
|
| 2166 |
-
"loss":
|
| 2167 |
"step": 306
|
| 2168 |
},
|
| 2169 |
{
|
| 2170 |
"epoch": 0.8108286563222186,
|
| 2171 |
-
"grad_norm":
|
| 2172 |
"learning_rate": 9.997008965364884e-05,
|
| 2173 |
-
"loss":
|
| 2174 |
"step": 307
|
| 2175 |
},
|
| 2176 |
{
|
| 2177 |
"epoch": 0.8134697920105646,
|
| 2178 |
-
"grad_norm":
|
| 2179 |
"learning_rate": 9.996979999591983e-05,
|
| 2180 |
-
"loss":
|
| 2181 |
"step": 308
|
| 2182 |
},
|
| 2183 |
{
|
| 2184 |
"epoch": 0.8161109276989106,
|
| 2185 |
-
"grad_norm":
|
| 2186 |
"learning_rate": 9.996950894281349e-05,
|
| 2187 |
-
"loss":
|
| 2188 |
"step": 309
|
| 2189 |
},
|
| 2190 |
{
|
| 2191 |
"epoch": 0.8187520633872565,
|
| 2192 |
-
"grad_norm":
|
| 2193 |
"learning_rate": 9.996921649433796e-05,
|
| 2194 |
-
"loss":
|
| 2195 |
"step": 310
|
| 2196 |
},
|
| 2197 |
{
|
| 2198 |
"epoch": 0.8213931990756025,
|
| 2199 |
-
"grad_norm":
|
| 2200 |
"learning_rate": 9.996892265050144e-05,
|
| 2201 |
-
"loss":
|
| 2202 |
"step": 311
|
| 2203 |
},
|
| 2204 |
{
|
| 2205 |
"epoch": 0.8240343347639485,
|
| 2206 |
-
"grad_norm":
|
| 2207 |
"learning_rate": 9.99686274113121e-05,
|
| 2208 |
-
"loss":
|
| 2209 |
"step": 312
|
| 2210 |
},
|
| 2211 |
{
|
| 2212 |
"epoch": 0.8266754704522945,
|
| 2213 |
-
"grad_norm":
|
| 2214 |
"learning_rate": 9.996833077677819e-05,
|
| 2215 |
-
"loss":
|
| 2216 |
"step": 313
|
| 2217 |
},
|
| 2218 |
{
|
| 2219 |
"epoch": 0.8293166061406405,
|
| 2220 |
-
"grad_norm":
|
| 2221 |
"learning_rate": 9.9968032746908e-05,
|
| 2222 |
-
"loss":
|
| 2223 |
"step": 314
|
| 2224 |
},
|
| 2225 |
{
|
| 2226 |
"epoch": 0.8319577418289865,
|
| 2227 |
-
"grad_norm":
|
| 2228 |
"learning_rate": 9.996773332170983e-05,
|
| 2229 |
-
"loss":
|
| 2230 |
"step": 315
|
| 2231 |
},
|
| 2232 |
{
|
| 2233 |
"epoch": 0.8345988775173324,
|
| 2234 |
-
"grad_norm":
|
| 2235 |
"learning_rate": 9.996743250119209e-05,
|
| 2236 |
-
"loss":
|
| 2237 |
"step": 316
|
| 2238 |
},
|
| 2239 |
{
|
| 2240 |
"epoch": 0.8372400132056784,
|
| 2241 |
-
"grad_norm":
|
| 2242 |
"learning_rate": 9.996713028536313e-05,
|
| 2243 |
-
"loss":
|
| 2244 |
"step": 317
|
| 2245 |
},
|
| 2246 |
{
|
| 2247 |
"epoch": 0.8398811488940244,
|
| 2248 |
-
"grad_norm":
|
| 2249 |
"learning_rate": 9.99668266742314e-05,
|
| 2250 |
-
"loss":
|
| 2251 |
"step": 318
|
| 2252 |
},
|
| 2253 |
{
|
| 2254 |
"epoch": 0.8425222845823704,
|
| 2255 |
-
"grad_norm":
|
| 2256 |
"learning_rate": 9.99665216678054e-05,
|
| 2257 |
-
"loss":
|
| 2258 |
"step": 319
|
| 2259 |
},
|
| 2260 |
{
|
| 2261 |
"epoch": 0.8451634202707164,
|
| 2262 |
-
"grad_norm":
|
| 2263 |
"learning_rate": 9.996621526609364e-05,
|
| 2264 |
-
"loss":
|
| 2265 |
"step": 320
|
| 2266 |
},
|
| 2267 |
{
|
| 2268 |
"epoch": 0.8478045559590623,
|
| 2269 |
-
"grad_norm":
|
| 2270 |
"learning_rate": 9.996590746910467e-05,
|
| 2271 |
-
"loss":
|
| 2272 |
"step": 321
|
| 2273 |
},
|
| 2274 |
{
|
| 2275 |
"epoch": 0.8504456916474084,
|
| 2276 |
-
"grad_norm":
|
| 2277 |
"learning_rate": 9.996559827684709e-05,
|
| 2278 |
-
"loss":
|
| 2279 |
"step": 322
|
| 2280 |
},
|
| 2281 |
{
|
| 2282 |
"epoch": 0.8530868273357544,
|
| 2283 |
-
"grad_norm":
|
| 2284 |
"learning_rate": 9.996528768932951e-05,
|
| 2285 |
-
"loss":
|
| 2286 |
"step": 323
|
| 2287 |
},
|
| 2288 |
{
|
| 2289 |
"epoch": 0.8557279630241004,
|
| 2290 |
-
"grad_norm":
|
| 2291 |
"learning_rate": 9.996497570656062e-05,
|
| 2292 |
-
"loss":
|
| 2293 |
"step": 324
|
| 2294 |
},
|
| 2295 |
{
|
| 2296 |
"epoch": 0.8583690987124464,
|
| 2297 |
-
"grad_norm":
|
| 2298 |
"learning_rate": 9.996466232854915e-05,
|
| 2299 |
-
"loss":
|
| 2300 |
"step": 325
|
| 2301 |
},
|
| 2302 |
{
|
| 2303 |
"epoch": 0.8610102344007924,
|
| 2304 |
-
"grad_norm":
|
| 2305 |
"learning_rate": 9.996434755530384e-05,
|
| 2306 |
-
"loss":
|
| 2307 |
"step": 326
|
| 2308 |
},
|
| 2309 |
{
|
| 2310 |
"epoch": 0.8636513700891383,
|
| 2311 |
-
"grad_norm":
|
| 2312 |
"learning_rate": 9.996403138683347e-05,
|
| 2313 |
-
"loss":
|
| 2314 |
"step": 327
|
| 2315 |
},
|
| 2316 |
{
|
| 2317 |
"epoch": 0.8662925057774843,
|
| 2318 |
-
"grad_norm":
|
| 2319 |
"learning_rate": 9.996371382314686e-05,
|
| 2320 |
-
"loss":
|
| 2321 |
"step": 328
|
| 2322 |
},
|
| 2323 |
{
|
| 2324 |
"epoch": 0.8689336414658303,
|
| 2325 |
-
"grad_norm":
|
| 2326 |
"learning_rate": 9.996339486425291e-05,
|
| 2327 |
-
"loss":
|
| 2328 |
"step": 329
|
| 2329 |
},
|
| 2330 |
{
|
| 2331 |
"epoch": 0.8715747771541763,
|
| 2332 |
-
"grad_norm":
|
| 2333 |
"learning_rate": 9.99630745101605e-05,
|
| 2334 |
-
"loss":
|
| 2335 |
"step": 330
|
| 2336 |
},
|
| 2337 |
{
|
| 2338 |
"epoch": 0.8742159128425223,
|
| 2339 |
-
"grad_norm":
|
| 2340 |
"learning_rate": 9.996275276087859e-05,
|
| 2341 |
-
"loss":
|
| 2342 |
"step": 331
|
| 2343 |
},
|
| 2344 |
{
|
| 2345 |
"epoch": 0.8768570485308683,
|
| 2346 |
-
"grad_norm":
|
| 2347 |
"learning_rate": 9.996242961641615e-05,
|
| 2348 |
-
"loss":
|
| 2349 |
"step": 332
|
| 2350 |
},
|
| 2351 |
{
|
| 2352 |
"epoch": 0.8794981842192142,
|
| 2353 |
-
"grad_norm":
|
| 2354 |
"learning_rate": 9.996210507678223e-05,
|
| 2355 |
-
"loss":
|
| 2356 |
"step": 333
|
| 2357 |
},
|
| 2358 |
{
|
| 2359 |
"epoch": 0.8821393199075602,
|
| 2360 |
-
"grad_norm":
|
| 2361 |
"learning_rate": 9.996177914198586e-05,
|
| 2362 |
-
"loss":
|
| 2363 |
"step": 334
|
| 2364 |
},
|
| 2365 |
{
|
| 2366 |
"epoch": 0.8847804555959062,
|
| 2367 |
-
"grad_norm":
|
| 2368 |
"learning_rate": 9.996145181203615e-05,
|
| 2369 |
-
"loss":
|
| 2370 |
"step": 335
|
| 2371 |
},
|
| 2372 |
{
|
| 2373 |
"epoch": 0.8874215912842522,
|
| 2374 |
-
"grad_norm":
|
| 2375 |
"learning_rate": 9.996112308694225e-05,
|
| 2376 |
-
"loss":
|
| 2377 |
"step": 336
|
| 2378 |
},
|
| 2379 |
{
|
| 2380 |
"epoch": 0.8900627269725983,
|
| 2381 |
-
"grad_norm":
|
| 2382 |
"learning_rate": 9.996079296671334e-05,
|
| 2383 |
-
"loss":
|
| 2384 |
"step": 337
|
| 2385 |
},
|
| 2386 |
{
|
| 2387 |
"epoch": 0.8927038626609443,
|
| 2388 |
-
"grad_norm":
|
| 2389 |
"learning_rate": 9.996046145135865e-05,
|
| 2390 |
-
"loss":
|
| 2391 |
"step": 338
|
| 2392 |
},
|
| 2393 |
{
|
| 2394 |
"epoch": 0.8953449983492902,
|
| 2395 |
-
"grad_norm":
|
| 2396 |
"learning_rate": 9.99601285408874e-05,
|
| 2397 |
-
"loss":
|
| 2398 |
"step": 339
|
| 2399 |
},
|
| 2400 |
{
|
| 2401 |
"epoch": 0.8979861340376362,
|
| 2402 |
-
"grad_norm":
|
| 2403 |
"learning_rate": 9.995979423530892e-05,
|
| 2404 |
-
"loss":
|
| 2405 |
"step": 340
|
| 2406 |
},
|
| 2407 |
{
|
| 2408 |
"epoch": 0.9006272697259822,
|
| 2409 |
-
"grad_norm":
|
| 2410 |
"learning_rate": 9.995945853463253e-05,
|
| 2411 |
-
"loss":
|
| 2412 |
"step": 341
|
| 2413 |
},
|
| 2414 |
{
|
| 2415 |
"epoch": 0.9032684054143282,
|
| 2416 |
-
"grad_norm":
|
| 2417 |
"learning_rate": 9.995912143886763e-05,
|
| 2418 |
-
"loss":
|
| 2419 |
"step": 342
|
| 2420 |
},
|
| 2421 |
{
|
| 2422 |
"epoch": 0.9059095411026742,
|
| 2423 |
-
"grad_norm":
|
| 2424 |
"learning_rate": 9.995878294802357e-05,
|
| 2425 |
-
"loss":
|
| 2426 |
"step": 343
|
| 2427 |
},
|
| 2428 |
{
|
| 2429 |
"epoch": 0.9085506767910201,
|
| 2430 |
-
"grad_norm":
|
| 2431 |
"learning_rate": 9.995844306210988e-05,
|
| 2432 |
-
"loss":
|
| 2433 |
"step": 344
|
| 2434 |
},
|
| 2435 |
{
|
| 2436 |
"epoch": 0.9111918124793661,
|
| 2437 |
-
"grad_norm":
|
| 2438 |
"learning_rate": 9.995810178113599e-05,
|
| 2439 |
-
"loss":
|
| 2440 |
"step": 345
|
| 2441 |
},
|
| 2442 |
{
|
| 2443 |
"epoch": 0.9138329481677121,
|
| 2444 |
-
"grad_norm":
|
| 2445 |
"learning_rate": 9.995775910511147e-05,
|
| 2446 |
-
"loss":
|
| 2447 |
"step": 346
|
| 2448 |
},
|
| 2449 |
{
|
| 2450 |
"epoch": 0.9164740838560581,
|
| 2451 |
-
"grad_norm":
|
| 2452 |
"learning_rate": 9.995741503404587e-05,
|
| 2453 |
-
"loss":
|
| 2454 |
"step": 347
|
| 2455 |
},
|
| 2456 |
{
|
| 2457 |
"epoch": 0.9191152195444041,
|
| 2458 |
-
"grad_norm":
|
| 2459 |
"learning_rate": 9.995706956794879e-05,
|
| 2460 |
-
"loss":
|
| 2461 |
"step": 348
|
| 2462 |
},
|
| 2463 |
{
|
| 2464 |
"epoch": 0.9217563552327501,
|
| 2465 |
-
"grad_norm":
|
| 2466 |
"learning_rate": 9.99567227068299e-05,
|
| 2467 |
-
"loss":
|
| 2468 |
"step": 349
|
| 2469 |
},
|
| 2470 |
{
|
| 2471 |
"epoch": 0.924397490921096,
|
| 2472 |
-
"grad_norm":
|
| 2473 |
"learning_rate": 9.995637445069887e-05,
|
| 2474 |
-
"loss":
|
| 2475 |
"step": 350
|
| 2476 |
},
|
| 2477 |
{
|
| 2478 |
"epoch": 0.927038626609442,
|
| 2479 |
-
"grad_norm":
|
| 2480 |
"learning_rate": 9.995602479956545e-05,
|
| 2481 |
-
"loss":
|
| 2482 |
"step": 351
|
| 2483 |
},
|
| 2484 |
{
|
| 2485 |
"epoch": 0.929679762297788,
|
| 2486 |
-
"grad_norm":
|
| 2487 |
"learning_rate": 9.995567375343937e-05,
|
| 2488 |
-
"loss":
|
| 2489 |
"step": 352
|
| 2490 |
},
|
| 2491 |
{
|
| 2492 |
"epoch": 0.932320897986134,
|
| 2493 |
-
"grad_norm":
|
| 2494 |
"learning_rate": 9.995532131233044e-05,
|
| 2495 |
-
"loss":
|
| 2496 |
"step": 353
|
| 2497 |
},
|
| 2498 |
{
|
| 2499 |
"epoch": 0.93496203367448,
|
| 2500 |
-
"grad_norm":
|
| 2501 |
"learning_rate": 9.99549674762485e-05,
|
| 2502 |
-
"loss":
|
| 2503 |
"step": 354
|
| 2504 |
},
|
| 2505 |
{
|
| 2506 |
"epoch": 0.9376031693628261,
|
| 2507 |
-
"grad_norm":
|
| 2508 |
"learning_rate": 9.995461224520345e-05,
|
| 2509 |
-
"loss":
|
| 2510 |
"step": 355
|
| 2511 |
},
|
| 2512 |
{
|
| 2513 |
"epoch": 0.940244305051172,
|
| 2514 |
-
"grad_norm":
|
| 2515 |
"learning_rate": 9.995425561920519e-05,
|
| 2516 |
-
"loss":
|
| 2517 |
"step": 356
|
| 2518 |
},
|
| 2519 |
{
|
| 2520 |
"epoch": 0.942885440739518,
|
| 2521 |
-
"grad_norm":
|
| 2522 |
"learning_rate": 9.99538975982637e-05,
|
| 2523 |
-
"loss":
|
| 2524 |
"step": 357
|
| 2525 |
},
|
| 2526 |
{
|
| 2527 |
"epoch": 0.945526576427864,
|
| 2528 |
-
"grad_norm":
|
| 2529 |
"learning_rate": 9.995353818238895e-05,
|
| 2530 |
-
"loss":
|
| 2531 |
"step": 358
|
| 2532 |
},
|
| 2533 |
{
|
| 2534 |
"epoch": 0.94816771211621,
|
| 2535 |
-
"grad_norm":
|
| 2536 |
"learning_rate": 9.9953177371591e-05,
|
| 2537 |
-
"loss":
|
| 2538 |
"step": 359
|
| 2539 |
},
|
| 2540 |
{
|
| 2541 |
"epoch": 0.950808847804556,
|
| 2542 |
-
"grad_norm":
|
| 2543 |
"learning_rate": 9.995281516587991e-05,
|
| 2544 |
-
"loss":
|
| 2545 |
"step": 360
|
| 2546 |
},
|
| 2547 |
{
|
| 2548 |
"epoch": 0.953449983492902,
|
| 2549 |
-
"grad_norm":
|
| 2550 |
"learning_rate": 9.99524515652658e-05,
|
| 2551 |
-
"loss":
|
| 2552 |
"step": 361
|
| 2553 |
},
|
| 2554 |
{
|
| 2555 |
"epoch": 0.9560911191812479,
|
| 2556 |
-
"grad_norm":
|
| 2557 |
"learning_rate": 9.995208656975884e-05,
|
| 2558 |
-
"loss":
|
| 2559 |
"step": 362
|
| 2560 |
},
|
| 2561 |
{
|
| 2562 |
"epoch": 0.9587322548695939,
|
| 2563 |
-
"grad_norm":
|
| 2564 |
"learning_rate": 9.995172017936919e-05,
|
| 2565 |
-
"loss":
|
| 2566 |
"step": 363
|
| 2567 |
},
|
| 2568 |
{
|
| 2569 |
"epoch": 0.9613733905579399,
|
| 2570 |
-
"grad_norm":
|
| 2571 |
"learning_rate": 9.99513523941071e-05,
|
| 2572 |
-
"loss":
|
| 2573 |
"step": 364
|
| 2574 |
},
|
| 2575 |
{
|
| 2576 |
"epoch": 0.9640145262462859,
|
| 2577 |
-
"grad_norm":
|
| 2578 |
"learning_rate": 9.995098321398284e-05,
|
| 2579 |
-
"loss":
|
| 2580 |
"step": 365
|
| 2581 |
},
|
| 2582 |
{
|
| 2583 |
"epoch": 0.9666556619346319,
|
| 2584 |
-
"grad_norm":
|
| 2585 |
"learning_rate": 9.995061263900671e-05,
|
| 2586 |
-
"loss":
|
| 2587 |
"step": 366
|
| 2588 |
},
|
| 2589 |
{
|
| 2590 |
"epoch": 0.9692967976229779,
|
| 2591 |
-
"grad_norm":
|
| 2592 |
"learning_rate": 9.995024066918908e-05,
|
| 2593 |
-
"loss":
|
| 2594 |
"step": 367
|
| 2595 |
},
|
| 2596 |
{
|
| 2597 |
"epoch": 0.9719379333113238,
|
| 2598 |
-
"grad_norm":
|
| 2599 |
"learning_rate": 9.994986730454031e-05,
|
| 2600 |
-
"loss":
|
| 2601 |
"step": 368
|
| 2602 |
},
|
| 2603 |
{
|
| 2604 |
"epoch": 0.9745790689996698,
|
| 2605 |
-
"grad_norm":
|
| 2606 |
"learning_rate": 9.994949254507084e-05,
|
| 2607 |
-
"loss":
|
| 2608 |
"step": 369
|
| 2609 |
},
|
| 2610 |
{
|
| 2611 |
"epoch": 0.9772202046880158,
|
| 2612 |
-
"grad_norm":
|
| 2613 |
"learning_rate": 9.994911639079112e-05,
|
| 2614 |
-
"loss":
|
| 2615 |
"step": 370
|
| 2616 |
},
|
| 2617 |
{
|
| 2618 |
"epoch": 0.9798613403763619,
|
| 2619 |
-
"grad_norm":
|
| 2620 |
"learning_rate": 9.994873884171167e-05,
|
| 2621 |
-
"loss":
|
| 2622 |
"step": 371
|
| 2623 |
},
|
| 2624 |
{
|
| 2625 |
"epoch": 0.9825024760647079,
|
| 2626 |
-
"grad_norm":
|
| 2627 |
"learning_rate": 9.994835989784305e-05,
|
| 2628 |
-
"loss":
|
| 2629 |
"step": 372
|
| 2630 |
},
|
| 2631 |
{
|
| 2632 |
"epoch": 0.9851436117530538,
|
| 2633 |
-
"grad_norm":
|
| 2634 |
"learning_rate": 9.994797955919581e-05,
|
| 2635 |
-
"loss":
|
| 2636 |
"step": 373
|
| 2637 |
},
|
| 2638 |
{
|
| 2639 |
"epoch": 0.9877847474413998,
|
| 2640 |
-
"grad_norm":
|
| 2641 |
"learning_rate": 9.994759782578058e-05,
|
| 2642 |
-
"loss":
|
| 2643 |
"step": 374
|
| 2644 |
},
|
| 2645 |
{
|
| 2646 |
"epoch": 0.9904258831297458,
|
| 2647 |
-
"grad_norm":
|
| 2648 |
"learning_rate": 9.994721469760801e-05,
|
| 2649 |
-
"loss":
|
| 2650 |
"step": 375
|
| 2651 |
},
|
| 2652 |
{
|
| 2653 |
"epoch": 0.9930670188180918,
|
| 2654 |
-
"grad_norm":
|
| 2655 |
"learning_rate": 9.994683017468883e-05,
|
| 2656 |
-
"loss":
|
| 2657 |
"step": 376
|
| 2658 |
},
|
| 2659 |
{
|
| 2660 |
"epoch": 0.9957081545064378,
|
| 2661 |
-
"grad_norm":
|
| 2662 |
"learning_rate": 9.994644425703374e-05,
|
| 2663 |
-
"loss":
|
| 2664 |
"step": 377
|
| 2665 |
},
|
| 2666 |
{
|
| 2667 |
"epoch": 0.9983492901947838,
|
| 2668 |
-
"grad_norm":
|
| 2669 |
"learning_rate": 9.994605694465355e-05,
|
| 2670 |
-
"loss":
|
| 2671 |
"step": 378
|
| 2672 |
},
|
| 2673 |
{
|
| 2674 |
-
"epoch": 1.
|
| 2675 |
-
"grad_norm":
|
| 2676 |
"learning_rate": 9.994566823755907e-05,
|
| 2677 |
-
"loss":
|
| 2678 |
"step": 379
|
| 2679 |
},
|
| 2680 |
{
|
| 2681 |
-
"epoch": 1.
|
| 2682 |
-
"grad_norm":
|
| 2683 |
"learning_rate": 9.99452781357611e-05,
|
| 2684 |
-
"loss":
|
| 2685 |
"step": 380
|
| 2686 |
},
|
| 2687 |
{
|
| 2688 |
-
"epoch": 1.
|
| 2689 |
-
"grad_norm":
|
| 2690 |
"learning_rate": 9.994488663927062e-05,
|
| 2691 |
-
"loss":
|
| 2692 |
"step": 381
|
| 2693 |
},
|
| 2694 |
{
|
| 2695 |
-
"epoch": 1.
|
| 2696 |
-
"grad_norm":
|
| 2697 |
"learning_rate": 9.994449374809851e-05,
|
| 2698 |
-
"loss":
|
| 2699 |
"step": 382
|
| 2700 |
},
|
| 2701 |
{
|
| 2702 |
-
"epoch": 1.
|
| 2703 |
-
"grad_norm":
|
| 2704 |
"learning_rate": 9.994409946225574e-05,
|
| 2705 |
-
"loss":
|
| 2706 |
"step": 383
|
| 2707 |
},
|
| 2708 |
{
|
| 2709 |
-
"epoch": 1.
|
| 2710 |
-
"grad_norm":
|
| 2711 |
"learning_rate": 9.994370378175332e-05,
|
| 2712 |
-
"loss":
|
| 2713 |
"step": 384
|
| 2714 |
},
|
| 2715 |
{
|
| 2716 |
-
"epoch": 1.
|
| 2717 |
-
"grad_norm":
|
| 2718 |
"learning_rate": 9.994330670660235e-05,
|
| 2719 |
-
"loss":
|
| 2720 |
"step": 385
|
| 2721 |
},
|
| 2722 |
{
|
| 2723 |
-
"epoch": 1.
|
| 2724 |
-
"grad_norm":
|
| 2725 |
"learning_rate": 9.994290823681385e-05,
|
| 2726 |
-
"loss":
|
| 2727 |
"step": 386
|
| 2728 |
},
|
| 2729 |
{
|
| 2730 |
-
"epoch": 1.
|
| 2731 |
-
"grad_norm":
|
| 2732 |
"learning_rate": 9.994250837239897e-05,
|
| 2733 |
-
"loss":
|
| 2734 |
"step": 387
|
| 2735 |
},
|
| 2736 |
{
|
| 2737 |
-
"epoch": 1.
|
| 2738 |
-
"grad_norm":
|
| 2739 |
"learning_rate": 9.994210711336891e-05,
|
| 2740 |
-
"loss":
|
| 2741 |
"step": 388
|
| 2742 |
},
|
| 2743 |
{
|
| 2744 |
-
"epoch": 1.
|
| 2745 |
-
"grad_norm":
|
| 2746 |
"learning_rate": 9.994170445973483e-05,
|
| 2747 |
-
"loss":
|
| 2748 |
"step": 389
|
| 2749 |
},
|
| 2750 |
{
|
| 2751 |
-
"epoch": 1.
|
| 2752 |
-
"grad_norm":
|
| 2753 |
"learning_rate": 9.994130041150798e-05,
|
| 2754 |
-
"loss":
|
| 2755 |
"step": 390
|
| 2756 |
},
|
| 2757 |
{
|
| 2758 |
-
"epoch": 1.
|
| 2759 |
-
"grad_norm":
|
| 2760 |
"learning_rate": 9.994089496869968e-05,
|
| 2761 |
-
"loss":
|
| 2762 |
"step": 391
|
| 2763 |
},
|
| 2764 |
{
|
| 2765 |
-
"epoch": 1.
|
| 2766 |
-
"grad_norm":
|
| 2767 |
"learning_rate": 9.994048813132119e-05,
|
| 2768 |
-
"loss":
|
| 2769 |
"step": 392
|
| 2770 |
},
|
| 2771 |
{
|
| 2772 |
-
"epoch": 1.
|
| 2773 |
-
"grad_norm":
|
| 2774 |
"learning_rate": 9.994007989938392e-05,
|
| 2775 |
-
"loss":
|
| 2776 |
"step": 393
|
| 2777 |
},
|
| 2778 |
{
|
| 2779 |
-
"epoch": 1.
|
| 2780 |
-
"grad_norm":
|
| 2781 |
"learning_rate": 9.993967027289927e-05,
|
| 2782 |
-
"loss":
|
| 2783 |
"step": 394
|
| 2784 |
},
|
| 2785 |
{
|
| 2786 |
-
"epoch": 1.
|
| 2787 |
-
"grad_norm":
|
| 2788 |
"learning_rate": 9.993925925187865e-05,
|
| 2789 |
-
"loss":
|
| 2790 |
"step": 395
|
| 2791 |
},
|
| 2792 |
{
|
| 2793 |
-
"epoch": 1.
|
| 2794 |
-
"grad_norm":
|
| 2795 |
"learning_rate": 9.993884683633354e-05,
|
| 2796 |
-
"loss":
|
| 2797 |
"step": 396
|
| 2798 |
},
|
| 2799 |
{
|
| 2800 |
-
"epoch": 1.
|
| 2801 |
-
"grad_norm":
|
| 2802 |
"learning_rate": 9.993843302627549e-05,
|
| 2803 |
-
"loss":
|
| 2804 |
"step": 397
|
| 2805 |
},
|
| 2806 |
{
|
| 2807 |
-
"epoch": 1.
|
| 2808 |
-
"grad_norm":
|
| 2809 |
"learning_rate": 9.993801782171603e-05,
|
| 2810 |
-
"loss":
|
| 2811 |
"step": 398
|
| 2812 |
},
|
| 2813 |
{
|
| 2814 |
-
"epoch": 1.
|
| 2815 |
-
"grad_norm":
|
| 2816 |
"learning_rate": 9.993760122266676e-05,
|
| 2817 |
-
"loss":
|
| 2818 |
"step": 399
|
| 2819 |
},
|
| 2820 |
{
|
| 2821 |
-
"epoch": 1.
|
| 2822 |
-
"grad_norm":
|
| 2823 |
"learning_rate": 9.99371832291393e-05,
|
| 2824 |
-
"loss":
|
| 2825 |
"step": 400
|
| 2826 |
},
|
| 2827 |
{
|
| 2828 |
-
"epoch": 1.
|
| 2829 |
-
"eval_loss": 8.
|
| 2830 |
-
"eval_runtime": 2.
|
| 2831 |
-
"eval_samples_per_second":
|
| 2832 |
-
"eval_steps_per_second":
|
| 2833 |
"step": 400
|
| 2834 |
}
|
| 2835 |
],
|
|
@@ -2850,8 +2850,8 @@
|
|
| 2850 |
"attributes": {}
|
| 2851 |
}
|
| 2852 |
},
|
| 2853 |
-
"total_flos":
|
| 2854 |
-
"train_batch_size":
|
| 2855 |
"trial_name": null,
|
| 2856 |
"trial_params": null
|
| 2857 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0564542753383954,
|
| 5 |
"eval_steps": 200,
|
| 6 |
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.002641135688345989,
|
| 13 |
+
"grad_norm": 38990.80078125,
|
| 14 |
"learning_rate": 1.0000000000000002e-06,
|
| 15 |
+
"loss": 39.1249,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.002641135688345989,
|
| 20 |
+
"eval_loss": 10.096823692321777,
|
| 21 |
+
"eval_runtime": 2.1873,
|
| 22 |
+
"eval_samples_per_second": 226.308,
|
| 23 |
+
"eval_steps_per_second": 28.346,
|
| 24 |
"step": 1
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.005282271376691978,
|
| 28 |
+
"grad_norm": 22047.08203125,
|
| 29 |
"learning_rate": 2.0000000000000003e-06,
|
| 30 |
+
"loss": 36.867,
|
| 31 |
"step": 2
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.007923407065037967,
|
| 35 |
+
"grad_norm": 81682.6796875,
|
| 36 |
"learning_rate": 3e-06,
|
| 37 |
+
"loss": 39.9853,
|
| 38 |
"step": 3
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.010564542753383956,
|
| 42 |
+
"grad_norm": 26123.87109375,
|
| 43 |
"learning_rate": 4.000000000000001e-06,
|
| 44 |
+
"loss": 38.4879,
|
| 45 |
"step": 4
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.013205678441729944,
|
| 49 |
+
"grad_norm": 29102.25390625,
|
| 50 |
"learning_rate": 5e-06,
|
| 51 |
+
"loss": 37.9034,
|
| 52 |
"step": 5
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.015846814130075933,
|
| 56 |
+
"grad_norm": 24646.03125,
|
| 57 |
"learning_rate": 6e-06,
|
| 58 |
+
"loss": 37.7639,
|
| 59 |
"step": 6
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.01848794981842192,
|
| 63 |
+
"grad_norm": 56427.59375,
|
| 64 |
"learning_rate": 7.000000000000001e-06,
|
| 65 |
+
"loss": 37.4074,
|
| 66 |
"step": 7
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.02112908550676791,
|
| 70 |
+
"grad_norm": 28639.47265625,
|
| 71 |
"learning_rate": 8.000000000000001e-06,
|
| 72 |
+
"loss": 37.1164,
|
| 73 |
"step": 8
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.0237702211951139,
|
| 77 |
+
"grad_norm": 18426.78515625,
|
| 78 |
"learning_rate": 9e-06,
|
| 79 |
+
"loss": 37.2912,
|
| 80 |
"step": 9
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.02641135688345989,
|
| 84 |
+
"grad_norm": 29681.62109375,
|
| 85 |
"learning_rate": 1e-05,
|
| 86 |
+
"loss": 37.1851,
|
| 87 |
"step": 10
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.029052492571805876,
|
| 91 |
+
"grad_norm": 61535.3671875,
|
| 92 |
"learning_rate": 1.1000000000000001e-05,
|
| 93 |
+
"loss": 38.4262,
|
| 94 |
"step": 11
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.03169362826015187,
|
| 98 |
+
"grad_norm": 28930.455078125,
|
| 99 |
"learning_rate": 1.2e-05,
|
| 100 |
+
"loss": 37.3699,
|
| 101 |
"step": 12
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.034334763948497854,
|
| 105 |
+
"grad_norm": 31548.685546875,
|
| 106 |
"learning_rate": 1.3000000000000001e-05,
|
| 107 |
+
"loss": 36.4952,
|
| 108 |
"step": 13
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.03697589963684384,
|
| 112 |
+
"grad_norm": 18435.33203125,
|
| 113 |
"learning_rate": 1.4000000000000001e-05,
|
| 114 |
+
"loss": 36.2176,
|
| 115 |
"step": 14
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.03961703532518983,
|
| 119 |
+
"grad_norm": 28708.28515625,
|
| 120 |
"learning_rate": 1.5e-05,
|
| 121 |
+
"loss": 36.6194,
|
| 122 |
"step": 15
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.04225817101353582,
|
| 126 |
+
"grad_norm": 20564.423828125,
|
| 127 |
"learning_rate": 1.6000000000000003e-05,
|
| 128 |
+
"loss": 36.3482,
|
| 129 |
"step": 16
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.04489930670188181,
|
| 133 |
+
"grad_norm": 17695.943359375,
|
| 134 |
"learning_rate": 1.7000000000000003e-05,
|
| 135 |
+
"loss": 35.715,
|
| 136 |
"step": 17
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.0475404423902278,
|
| 140 |
+
"grad_norm": 15335.0712890625,
|
| 141 |
"learning_rate": 1.8e-05,
|
| 142 |
+
"loss": 34.6268,
|
| 143 |
"step": 18
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.050181578078573784,
|
| 147 |
+
"grad_norm": 13583.33203125,
|
| 148 |
"learning_rate": 1.9e-05,
|
| 149 |
+
"loss": 35.5606,
|
| 150 |
"step": 19
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.05282271376691978,
|
| 154 |
+
"grad_norm": 26019.890625,
|
| 155 |
"learning_rate": 2e-05,
|
| 156 |
+
"loss": 35.5296,
|
| 157 |
"step": 20
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.055463849455265765,
|
| 161 |
+
"grad_norm": 14379.9052734375,
|
| 162 |
"learning_rate": 2.1e-05,
|
| 163 |
+
"loss": 34.4806,
|
| 164 |
"step": 21
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.05810498514361175,
|
| 168 |
+
"grad_norm": 27736.314453125,
|
| 169 |
"learning_rate": 2.2000000000000003e-05,
|
| 170 |
+
"loss": 36.2501,
|
| 171 |
"step": 22
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.06074612083195774,
|
| 175 |
+
"grad_norm": 260855.0,
|
| 176 |
"learning_rate": 2.3000000000000003e-05,
|
| 177 |
+
"loss": 150.3672,
|
| 178 |
"step": 23
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.06338725652030373,
|
| 182 |
+
"grad_norm": 422515.90625,
|
| 183 |
"learning_rate": 2.4e-05,
|
| 184 |
+
"loss": 292.9531,
|
| 185 |
"step": 24
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.06602839220864971,
|
| 189 |
+
"grad_norm": 1018015.1875,
|
| 190 |
"learning_rate": 2.5e-05,
|
| 191 |
+
"loss": 355.2773,
|
| 192 |
"step": 25
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.06866952789699571,
|
| 196 |
+
"grad_norm": 817252.75,
|
| 197 |
"learning_rate": 2.6000000000000002e-05,
|
| 198 |
+
"loss": 268.3635,
|
| 199 |
"step": 26
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.0713106635853417,
|
| 203 |
+
"grad_norm": 650477.3125,
|
| 204 |
"learning_rate": 2.7000000000000002e-05,
|
| 205 |
+
"loss": 304.9252,
|
| 206 |
"step": 27
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.07395179927368768,
|
| 210 |
+
"grad_norm": 405537.28125,
|
| 211 |
"learning_rate": 2.8000000000000003e-05,
|
| 212 |
+
"loss": 298.1852,
|
| 213 |
"step": 28
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.07659293496203368,
|
| 217 |
+
"grad_norm": 1309428.0,
|
| 218 |
"learning_rate": 2.9e-05,
|
| 219 |
+
"loss": 405.8535,
|
| 220 |
"step": 29
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.07923407065037966,
|
| 224 |
+
"grad_norm": 574544.0625,
|
| 225 |
"learning_rate": 3e-05,
|
| 226 |
+
"loss": 341.7129,
|
| 227 |
"step": 30
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.08187520633872565,
|
| 231 |
+
"grad_norm": 425538.625,
|
| 232 |
"learning_rate": 3.1e-05,
|
| 233 |
+
"loss": 260.0619,
|
| 234 |
"step": 31
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.08451634202707164,
|
| 238 |
+
"grad_norm": 566855.6875,
|
| 239 |
"learning_rate": 3.2000000000000005e-05,
|
| 240 |
+
"loss": 272.2906,
|
| 241 |
"step": 32
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.08715747771541763,
|
| 245 |
+
"grad_norm": 300651.28125,
|
| 246 |
"learning_rate": 3.3e-05,
|
| 247 |
+
"loss": 79.3429,
|
| 248 |
"step": 33
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.08979861340376362,
|
| 252 |
+
"grad_norm": 14083.6494140625,
|
| 253 |
"learning_rate": 3.4000000000000007e-05,
|
| 254 |
+
"loss": 35.4547,
|
| 255 |
"step": 34
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.09243974909210961,
|
| 259 |
+
"grad_norm": 22081.201171875,
|
| 260 |
"learning_rate": 3.5e-05,
|
| 261 |
+
"loss": 34.5699,
|
| 262 |
"step": 35
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.0950808847804556,
|
| 266 |
+
"grad_norm": 11022.6884765625,
|
| 267 |
"learning_rate": 3.6e-05,
|
| 268 |
+
"loss": 34.3495,
|
| 269 |
"step": 36
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.09772202046880159,
|
| 273 |
+
"grad_norm": 11900.990234375,
|
| 274 |
"learning_rate": 3.7e-05,
|
| 275 |
+
"loss": 35.6864,
|
| 276 |
"step": 37
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.10036315615714757,
|
| 280 |
+
"grad_norm": 13156.771484375,
|
| 281 |
"learning_rate": 3.8e-05,
|
| 282 |
+
"loss": 34.4444,
|
| 283 |
"step": 38
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.10300429184549356,
|
| 287 |
+
"grad_norm": 11813.6083984375,
|
| 288 |
"learning_rate": 3.9000000000000006e-05,
|
| 289 |
+
"loss": 34.9737,
|
| 290 |
"step": 39
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.10564542753383956,
|
| 294 |
+
"grad_norm": 15030.2021484375,
|
| 295 |
"learning_rate": 4e-05,
|
| 296 |
+
"loss": 34.0348,
|
| 297 |
"step": 40
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.10828656322218554,
|
| 301 |
+
"grad_norm": 11196.2529296875,
|
| 302 |
"learning_rate": 4.1e-05,
|
| 303 |
+
"loss": 34.3456,
|
| 304 |
"step": 41
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.11092769891053153,
|
| 308 |
+
"grad_norm": 11016.130859375,
|
| 309 |
"learning_rate": 4.2e-05,
|
| 310 |
+
"loss": 34.3275,
|
| 311 |
"step": 42
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.11356883459887751,
|
| 315 |
+
"grad_norm": 14342.5283203125,
|
| 316 |
"learning_rate": 4.3e-05,
|
| 317 |
+
"loss": 33.6461,
|
| 318 |
"step": 43
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.1162099702872235,
|
| 322 |
+
"grad_norm": 13592.828125,
|
| 323 |
"learning_rate": 4.4000000000000006e-05,
|
| 324 |
+
"loss": 35.0608,
|
| 325 |
"step": 44
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.1188511059755695,
|
| 329 |
+
"grad_norm": 14278.0205078125,
|
| 330 |
"learning_rate": 4.5e-05,
|
| 331 |
+
"loss": 33.6034,
|
| 332 |
"step": 45
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.12149224166391548,
|
| 336 |
+
"grad_norm": 15676.8076171875,
|
| 337 |
"learning_rate": 4.600000000000001e-05,
|
| 338 |
+
"loss": 34.7689,
|
| 339 |
"step": 46
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.12413337735226147,
|
| 343 |
+
"grad_norm": 16533.037109375,
|
| 344 |
"learning_rate": 4.7e-05,
|
| 345 |
+
"loss": 34.1397,
|
| 346 |
"step": 47
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.12677451304060747,
|
| 350 |
+
"grad_norm": 17516.21484375,
|
| 351 |
"learning_rate": 4.8e-05,
|
| 352 |
+
"loss": 35.4853,
|
| 353 |
"step": 48
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.12941564872895345,
|
| 357 |
+
"grad_norm": 22093.806640625,
|
| 358 |
"learning_rate": 4.9e-05,
|
| 359 |
+
"loss": 36.8646,
|
| 360 |
"step": 49
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.13205678441729943,
|
| 364 |
+
"grad_norm": 34389.921875,
|
| 365 |
"learning_rate": 5e-05,
|
| 366 |
+
"loss": 39.5266,
|
| 367 |
"step": 50
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.13469792010564544,
|
| 371 |
+
"grad_norm": 7149.9775390625,
|
| 372 |
"learning_rate": 5.1000000000000006e-05,
|
| 373 |
+
"loss": 35.8088,
|
| 374 |
"step": 51
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.13733905579399142,
|
| 378 |
+
"grad_norm": 6511.89306640625,
|
| 379 |
"learning_rate": 5.2000000000000004e-05,
|
| 380 |
+
"loss": 35.0745,
|
| 381 |
"step": 52
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.1399801914823374,
|
| 385 |
+
"grad_norm": 11293.515625,
|
| 386 |
"learning_rate": 5.300000000000001e-05,
|
| 387 |
+
"loss": 33.7064,
|
| 388 |
"step": 53
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.1426213271706834,
|
| 392 |
+
"grad_norm": 7394.4853515625,
|
| 393 |
"learning_rate": 5.4000000000000005e-05,
|
| 394 |
+
"loss": 34.612,
|
| 395 |
"step": 54
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.14526246285902938,
|
| 399 |
+
"grad_norm": 7513.56982421875,
|
| 400 |
"learning_rate": 5.500000000000001e-05,
|
| 401 |
+
"loss": 34.2795,
|
| 402 |
"step": 55
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.14790359854737536,
|
| 406 |
+
"grad_norm": 12561.0849609375,
|
| 407 |
"learning_rate": 5.6000000000000006e-05,
|
| 408 |
+
"loss": 34.1079,
|
| 409 |
"step": 56
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.15054473423572137,
|
| 413 |
+
"grad_norm": 7255.42724609375,
|
| 414 |
"learning_rate": 5.6999999999999996e-05,
|
| 415 |
+
"loss": 33.7773,
|
| 416 |
"step": 57
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.15318586992406735,
|
| 420 |
+
"grad_norm": 8305.197265625,
|
| 421 |
"learning_rate": 5.8e-05,
|
| 422 |
+
"loss": 33.5425,
|
| 423 |
"step": 58
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.15582700561241333,
|
| 427 |
+
"grad_norm": 7724.32666015625,
|
| 428 |
"learning_rate": 5.9e-05,
|
| 429 |
+
"loss": 34.3069,
|
| 430 |
"step": 59
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.1584681413007593,
|
| 434 |
+
"grad_norm": 6973.86669921875,
|
| 435 |
"learning_rate": 6e-05,
|
| 436 |
+
"loss": 31.8323,
|
| 437 |
"step": 60
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.16110927698910532,
|
| 441 |
+
"grad_norm": 8178.408203125,
|
| 442 |
"learning_rate": 6.1e-05,
|
| 443 |
+
"loss": 33.4728,
|
| 444 |
"step": 61
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.1637504126774513,
|
| 448 |
+
"grad_norm": 7446.3310546875,
|
| 449 |
"learning_rate": 6.2e-05,
|
| 450 |
+
"loss": 32.2049,
|
| 451 |
"step": 62
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.16639154836579728,
|
| 455 |
+
"grad_norm": 7538.81494140625,
|
| 456 |
"learning_rate": 6.3e-05,
|
| 457 |
+
"loss": 31.9451,
|
| 458 |
"step": 63
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.1690326840541433,
|
| 462 |
+
"grad_norm": 7067.33154296875,
|
| 463 |
"learning_rate": 6.400000000000001e-05,
|
| 464 |
+
"loss": 32.0696,
|
| 465 |
"step": 64
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.17167381974248927,
|
| 469 |
+
"grad_norm": 7199.02294921875,
|
| 470 |
"learning_rate": 6.500000000000001e-05,
|
| 471 |
+
"loss": 31.7234,
|
| 472 |
"step": 65
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.17431495543083525,
|
| 476 |
+
"grad_norm": 6351.2900390625,
|
| 477 |
"learning_rate": 6.6e-05,
|
| 478 |
+
"loss": 31.4103,
|
| 479 |
"step": 66
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.17695609111918126,
|
| 483 |
+
"grad_norm": 9954.1572265625,
|
| 484 |
"learning_rate": 6.7e-05,
|
| 485 |
+
"loss": 31.2581,
|
| 486 |
"step": 67
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.17959722680752724,
|
| 490 |
+
"grad_norm": 6812.11083984375,
|
| 491 |
"learning_rate": 6.800000000000001e-05,
|
| 492 |
+
"loss": 31.0586,
|
| 493 |
"step": 68
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.18223836249587322,
|
| 497 |
+
"grad_norm": 6788.81787109375,
|
| 498 |
"learning_rate": 6.9e-05,
|
| 499 |
+
"loss": 31.2012,
|
| 500 |
"step": 69
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.18487949818421923,
|
| 504 |
+
"grad_norm": 6330.77880859375,
|
| 505 |
"learning_rate": 7e-05,
|
| 506 |
+
"loss": 31.758,
|
| 507 |
"step": 70
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.1875206338725652,
|
| 511 |
+
"grad_norm": 6925.2958984375,
|
| 512 |
"learning_rate": 7.1e-05,
|
| 513 |
+
"loss": 31.7811,
|
| 514 |
"step": 71
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.1901617695609112,
|
| 518 |
+
"grad_norm": 15530.548828125,
|
| 519 |
"learning_rate": 7.2e-05,
|
| 520 |
+
"loss": 38.1656,
|
| 521 |
"step": 72
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.19280290524925717,
|
| 525 |
+
"grad_norm": 248175.0,
|
| 526 |
"learning_rate": 7.3e-05,
|
| 527 |
+
"loss": 314.6807,
|
| 528 |
"step": 73
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.19544404093760318,
|
| 532 |
+
"grad_norm": 348192.3125,
|
| 533 |
"learning_rate": 7.4e-05,
|
| 534 |
+
"loss": 319.9785,
|
| 535 |
"step": 74
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.19808517662594916,
|
| 539 |
+
"grad_norm": 399153.90625,
|
| 540 |
"learning_rate": 7.500000000000001e-05,
|
| 541 |
+
"loss": 245.481,
|
| 542 |
"step": 75
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.20072631231429514,
|
| 546 |
+
"grad_norm": 293389.5,
|
| 547 |
"learning_rate": 7.6e-05,
|
| 548 |
+
"loss": 269.3301,
|
| 549 |
"step": 76
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.20336744800264114,
|
| 553 |
+
"grad_norm": 536027.375,
|
| 554 |
"learning_rate": 7.7e-05,
|
| 555 |
+
"loss": 240.3848,
|
| 556 |
"step": 77
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.20600858369098712,
|
| 560 |
+
"grad_norm": 340636.96875,
|
| 561 |
"learning_rate": 7.800000000000001e-05,
|
| 562 |
+
"loss": 257.0401,
|
| 563 |
"step": 78
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.2086497193793331,
|
| 567 |
+
"grad_norm": 343861.0625,
|
| 568 |
"learning_rate": 7.900000000000001e-05,
|
| 569 |
+
"loss": 246.5806,
|
| 570 |
"step": 79
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.2112908550676791,
|
| 574 |
+
"grad_norm": 377362.75,
|
| 575 |
"learning_rate": 8e-05,
|
| 576 |
+
"loss": 204.2622,
|
| 577 |
"step": 80
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.2139319907560251,
|
| 581 |
+
"grad_norm": 495172.15625,
|
| 582 |
"learning_rate": 8.1e-05,
|
| 583 |
+
"loss": 152.9565,
|
| 584 |
"step": 81
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.21657312644437107,
|
| 588 |
+
"grad_norm": 410514.21875,
|
| 589 |
"learning_rate": 8.2e-05,
|
| 590 |
+
"loss": 120.1336,
|
| 591 |
"step": 82
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.21921426213271708,
|
| 595 |
+
"grad_norm": 37318.89453125,
|
| 596 |
"learning_rate": 8.3e-05,
|
| 597 |
+
"loss": 43.507,
|
| 598 |
"step": 83
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.22185539782106306,
|
| 602 |
+
"grad_norm": 46563.8515625,
|
| 603 |
"learning_rate": 8.4e-05,
|
| 604 |
+
"loss": 42.6661,
|
| 605 |
"step": 84
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.22449653350940904,
|
| 609 |
+
"grad_norm": 25882.45703125,
|
| 610 |
"learning_rate": 8.5e-05,
|
| 611 |
+
"loss": 41.1904,
|
| 612 |
"step": 85
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.22713766919775502,
|
| 616 |
+
"grad_norm": 21462.017578125,
|
| 617 |
"learning_rate": 8.6e-05,
|
| 618 |
+
"loss": 35.6957,
|
| 619 |
"step": 86
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.22977880488610103,
|
| 623 |
+
"grad_norm": 11826.3798828125,
|
| 624 |
"learning_rate": 8.7e-05,
|
| 625 |
+
"loss": 33.1654,
|
| 626 |
"step": 87
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.232419940574447,
|
| 630 |
+
"grad_norm": 10408.4365234375,
|
| 631 |
"learning_rate": 8.800000000000001e-05,
|
| 632 |
+
"loss": 32.1405,
|
| 633 |
"step": 88
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.235061076262793,
|
| 637 |
+
"grad_norm": 9028.2587890625,
|
| 638 |
"learning_rate": 8.900000000000001e-05,
|
| 639 |
+
"loss": 30.806,
|
| 640 |
"step": 89
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.237702211951139,
|
| 644 |
+
"grad_norm": 14064.7021484375,
|
| 645 |
"learning_rate": 9e-05,
|
| 646 |
+
"loss": 32.0673,
|
| 647 |
"step": 90
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.24034334763948498,
|
| 651 |
+
"grad_norm": 10274.6611328125,
|
| 652 |
"learning_rate": 9.1e-05,
|
| 653 |
+
"loss": 30.8923,
|
| 654 |
"step": 91
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.24298448332783096,
|
| 658 |
+
"grad_norm": 13376.0947265625,
|
| 659 |
"learning_rate": 9.200000000000001e-05,
|
| 660 |
+
"loss": 32.0376,
|
| 661 |
"step": 92
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.24562561901617697,
|
| 665 |
+
"grad_norm": 13412.4970703125,
|
| 666 |
"learning_rate": 9.300000000000001e-05,
|
| 667 |
+
"loss": 32.5937,
|
| 668 |
"step": 93
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.24826675470452295,
|
| 672 |
+
"grad_norm": 17289.099609375,
|
| 673 |
"learning_rate": 9.4e-05,
|
| 674 |
+
"loss": 32.0219,
|
| 675 |
"step": 94
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.2509078903928689,
|
| 679 |
+
"grad_norm": 10165.4990234375,
|
| 680 |
"learning_rate": 9.5e-05,
|
| 681 |
+
"loss": 32.7753,
|
| 682 |
"step": 95
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.25354902608121493,
|
| 686 |
+
"grad_norm": 16371.439453125,
|
| 687 |
"learning_rate": 9.6e-05,
|
| 688 |
+
"loss": 31.7399,
|
| 689 |
"step": 96
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.2561901617695609,
|
| 693 |
+
"grad_norm": 28360.642578125,
|
| 694 |
"learning_rate": 9.7e-05,
|
| 695 |
+
"loss": 32.4525,
|
| 696 |
"step": 97
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.2588312974579069,
|
| 700 |
+
"grad_norm": 19952.9296875,
|
| 701 |
"learning_rate": 9.8e-05,
|
| 702 |
+
"loss": 33.4285,
|
| 703 |
"step": 98
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.2614724331462529,
|
| 707 |
+
"grad_norm": 20724.11328125,
|
| 708 |
"learning_rate": 9.900000000000001e-05,
|
| 709 |
+
"loss": 34.1331,
|
| 710 |
"step": 99
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.26411356883459886,
|
| 714 |
+
"grad_norm": 38431.6328125,
|
| 715 |
"learning_rate": 0.0001,
|
| 716 |
+
"loss": 40.0183,
|
| 717 |
"step": 100
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.26675470452294486,
|
| 721 |
+
"grad_norm": 9295.7626953125,
|
| 722 |
"learning_rate": 9.99999993018897e-05,
|
| 723 |
+
"loss": 32.908,
|
| 724 |
"step": 101
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.26939584021129087,
|
| 728 |
+
"grad_norm": 9068.3134765625,
|
| 729 |
"learning_rate": 9.999999720755877e-05,
|
| 730 |
+
"loss": 32.7796,
|
| 731 |
"step": 102
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.2720369758996368,
|
| 735 |
+
"grad_norm": 9507.033203125,
|
| 736 |
"learning_rate": 9.99999937170073e-05,
|
| 737 |
+
"loss": 33.0735,
|
| 738 |
"step": 103
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.27467811158798283,
|
| 742 |
+
"grad_norm": 9898.73046875,
|
| 743 |
"learning_rate": 9.999998883023537e-05,
|
| 744 |
+
"loss": 34.4524,
|
| 745 |
"step": 104
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.27731924727632884,
|
| 749 |
+
"grad_norm": 8197.7294921875,
|
| 750 |
"learning_rate": 9.999998254724313e-05,
|
| 751 |
+
"loss": 34.3429,
|
| 752 |
"step": 105
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.2799603829646748,
|
| 756 |
+
"grad_norm": 7723.392578125,
|
| 757 |
"learning_rate": 9.999997486803075e-05,
|
| 758 |
+
"loss": 33.3132,
|
| 759 |
"step": 106
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.2826015186530208,
|
| 763 |
+
"grad_norm": 9174.4091796875,
|
| 764 |
"learning_rate": 9.999996579259843e-05,
|
| 765 |
+
"loss": 32.9465,
|
| 766 |
"step": 107
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.2852426543413668,
|
| 770 |
+
"grad_norm": 10098.0283203125,
|
| 771 |
"learning_rate": 9.999995532094644e-05,
|
| 772 |
+
"loss": 34.1124,
|
| 773 |
"step": 108
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.28788379002971276,
|
| 777 |
+
"grad_norm": 7904.126953125,
|
| 778 |
"learning_rate": 9.999994345307508e-05,
|
| 779 |
+
"loss": 32.68,
|
| 780 |
"step": 109
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.29052492571805877,
|
| 784 |
+
"grad_norm": 7395.32177734375,
|
| 785 |
"learning_rate": 9.999993018898466e-05,
|
| 786 |
+
"loss": 32.1147,
|
| 787 |
"step": 110
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.2931660614064048,
|
| 791 |
+
"grad_norm": 15490.7314453125,
|
| 792 |
"learning_rate": 9.999991552867558e-05,
|
| 793 |
+
"loss": 32.7157,
|
| 794 |
"step": 111
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.29580719709475073,
|
| 798 |
+
"grad_norm": 6962.9326171875,
|
| 799 |
"learning_rate": 9.99998994721482e-05,
|
| 800 |
+
"loss": 32.5103,
|
| 801 |
"step": 112
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.29844833278309674,
|
| 805 |
+
"grad_norm": 6746.60546875,
|
| 806 |
"learning_rate": 9.999988201940302e-05,
|
| 807 |
+
"loss": 31.9245,
|
| 808 |
"step": 113
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.30108946847144274,
|
| 812 |
+
"grad_norm": 6943.94140625,
|
| 813 |
"learning_rate": 9.999986317044051e-05,
|
| 814 |
+
"loss": 30.399,
|
| 815 |
"step": 114
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 0.3037306041597887,
|
| 819 |
+
"grad_norm": 6095.4384765625,
|
| 820 |
"learning_rate": 9.999984292526118e-05,
|
| 821 |
+
"loss": 29.941,
|
| 822 |
"step": 115
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 0.3063717398481347,
|
| 826 |
+
"grad_norm": 6518.970703125,
|
| 827 |
"learning_rate": 9.999982128386562e-05,
|
| 828 |
+
"loss": 30.5093,
|
| 829 |
"step": 116
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 0.3090128755364807,
|
| 833 |
+
"grad_norm": 5806.0927734375,
|
| 834 |
"learning_rate": 9.99997982462544e-05,
|
| 835 |
+
"loss": 29.6937,
|
| 836 |
"step": 117
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.31165401122482667,
|
| 840 |
+
"grad_norm": 6216.46435546875,
|
| 841 |
"learning_rate": 9.999977381242821e-05,
|
| 842 |
+
"loss": 29.7115,
|
| 843 |
"step": 118
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 0.3142951469131727,
|
| 847 |
+
"grad_norm": 5445.48828125,
|
| 848 |
"learning_rate": 9.999974798238769e-05,
|
| 849 |
+
"loss": 28.9644,
|
| 850 |
"step": 119
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 0.3169362826015186,
|
| 854 |
+
"grad_norm": 4930.64453125,
|
| 855 |
"learning_rate": 9.99997207561336e-05,
|
| 856 |
+
"loss": 29.803,
|
| 857 |
"step": 120
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 0.31957741828986463,
|
| 861 |
+
"grad_norm": 5866.5478515625,
|
| 862 |
"learning_rate": 9.999969213366667e-05,
|
| 863 |
+
"loss": 29.2732,
|
| 864 |
"step": 121
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 0.32221855397821064,
|
| 868 |
+
"grad_norm": 13160.4111328125,
|
| 869 |
"learning_rate": 9.99996621149877e-05,
|
| 870 |
+
"loss": 31.3179,
|
| 871 |
"step": 122
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 0.3248596896665566,
|
| 875 |
+
"grad_norm": 577362.4375,
|
| 876 |
"learning_rate": 9.999963070009755e-05,
|
| 877 |
+
"loss": 192.1116,
|
| 878 |
"step": 123
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.3275008253549026,
|
| 882 |
+
"grad_norm": 447577.625,
|
| 883 |
"learning_rate": 9.999959788899706e-05,
|
| 884 |
+
"loss": 353.353,
|
| 885 |
"step": 124
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 0.3301419610432486,
|
| 889 |
+
"grad_norm": 422884.03125,
|
| 890 |
"learning_rate": 9.999956368168719e-05,
|
| 891 |
+
"loss": 328.2871,
|
| 892 |
"step": 125
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 0.33278309673159456,
|
| 896 |
+
"grad_norm": 217372.875,
|
| 897 |
"learning_rate": 9.999952807816888e-05,
|
| 898 |
+
"loss": 311.041,
|
| 899 |
"step": 126
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 0.33542423241994057,
|
| 903 |
+
"grad_norm": 197269.15625,
|
| 904 |
"learning_rate": 9.99994910784431e-05,
|
| 905 |
+
"loss": 331.5454,
|
| 906 |
"step": 127
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 0.3380653681082866,
|
| 910 |
+
"grad_norm": 447190.15625,
|
| 911 |
"learning_rate": 9.999945268251092e-05,
|
| 912 |
+
"loss": 292.5098,
|
| 913 |
"step": 128
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 0.34070650379663253,
|
| 917 |
+
"grad_norm": 156708.53125,
|
| 918 |
"learning_rate": 9.999941289037338e-05,
|
| 919 |
+
"loss": 329.5899,
|
| 920 |
"step": 129
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 0.34334763948497854,
|
| 924 |
+
"grad_norm": 214527.265625,
|
| 925 |
"learning_rate": 9.999937170203162e-05,
|
| 926 |
+
"loss": 295.6437,
|
| 927 |
"step": 130
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 0.34598877517332455,
|
| 931 |
+
"grad_norm": 144792.09375,
|
| 932 |
"learning_rate": 9.999932911748678e-05,
|
| 933 |
+
"loss": 321.7724,
|
| 934 |
"step": 131
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 0.3486299108616705,
|
| 938 |
+
"grad_norm": 183092.328125,
|
| 939 |
"learning_rate": 9.999928513674004e-05,
|
| 940 |
+
"loss": 138.0811,
|
| 941 |
"step": 132
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 0.3512710465500165,
|
| 945 |
+
"grad_norm": 6041.10107421875,
|
| 946 |
"learning_rate": 9.999923975979262e-05,
|
| 947 |
+
"loss": 30.1601,
|
| 948 |
"step": 133
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 0.3539121822383625,
|
| 952 |
+
"grad_norm": 5054.18798828125,
|
| 953 |
"learning_rate": 9.999919298664582e-05,
|
| 954 |
+
"loss": 29.4563,
|
| 955 |
"step": 134
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 0.35655331792670847,
|
| 959 |
+
"grad_norm": 9742.12890625,
|
| 960 |
"learning_rate": 9.999914481730092e-05,
|
| 961 |
+
"loss": 29.5483,
|
| 962 |
"step": 135
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 0.3591944536150545,
|
| 966 |
+
"grad_norm": 13321.4970703125,
|
| 967 |
"learning_rate": 9.999909525175927e-05,
|
| 968 |
+
"loss": 29.7589,
|
| 969 |
"step": 136
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 0.3618355893034005,
|
| 973 |
+
"grad_norm": 9211.091796875,
|
| 974 |
"learning_rate": 9.999904429002225e-05,
|
| 975 |
+
"loss": 30.0795,
|
| 976 |
"step": 137
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 0.36447672499174644,
|
| 980 |
+
"grad_norm": 10673.2529296875,
|
| 981 |
"learning_rate": 9.99989919320913e-05,
|
| 982 |
+
"loss": 29.8073,
|
| 983 |
"step": 138
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 0.36711786068009244,
|
| 987 |
+
"grad_norm": 9673.37109375,
|
| 988 |
"learning_rate": 9.999893817796786e-05,
|
| 989 |
+
"loss": 30.8933,
|
| 990 |
"step": 139
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 0.36975899636843845,
|
| 994 |
+
"grad_norm": 10085.38671875,
|
| 995 |
"learning_rate": 9.999888302765345e-05,
|
| 996 |
+
"loss": 29.8822,
|
| 997 |
"step": 140
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 0.3724001320567844,
|
| 1001 |
+
"grad_norm": 10791.521484375,
|
| 1002 |
"learning_rate": 9.99988264811496e-05,
|
| 1003 |
+
"loss": 30.1218,
|
| 1004 |
"step": 141
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 0.3750412677451304,
|
| 1008 |
+
"grad_norm": 11358.93359375,
|
| 1009 |
"learning_rate": 9.99987685384579e-05,
|
| 1010 |
+
"loss": 30.735,
|
| 1011 |
"step": 142
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 0.3776824034334764,
|
| 1015 |
+
"grad_norm": 7013.380859375,
|
| 1016 |
"learning_rate": 9.999870919957996e-05,
|
| 1017 |
+
"loss": 29.9077,
|
| 1018 |
"step": 143
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 0.3803235391218224,
|
| 1022 |
+
"grad_norm": 7458.63525390625,
|
| 1023 |
"learning_rate": 9.999864846451744e-05,
|
| 1024 |
+
"loss": 30.7425,
|
| 1025 |
"step": 144
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 0.3829646748101684,
|
| 1029 |
+
"grad_norm": 8038.50732421875,
|
| 1030 |
"learning_rate": 9.999858633327201e-05,
|
| 1031 |
+
"loss": 31.053,
|
| 1032 |
"step": 145
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 0.38560581049851433,
|
| 1036 |
+
"grad_norm": 7841.15283203125,
|
| 1037 |
"learning_rate": 9.999852280584544e-05,
|
| 1038 |
+
"loss": 30.7345,
|
| 1039 |
"step": 146
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
"epoch": 0.38824694618686034,
|
| 1043 |
+
"grad_norm": 7719.5048828125,
|
| 1044 |
"learning_rate": 9.999845788223949e-05,
|
| 1045 |
+
"loss": 30.9241,
|
| 1046 |
"step": 147
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 0.39088808187520635,
|
| 1050 |
+
"grad_norm": 13179.359375,
|
| 1051 |
"learning_rate": 9.999839156245598e-05,
|
| 1052 |
+
"loss": 31.945,
|
| 1053 |
"step": 148
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
"epoch": 0.3935292175635523,
|
| 1057 |
+
"grad_norm": 11153.3046875,
|
| 1058 |
"learning_rate": 9.999832384649674e-05,
|
| 1059 |
+
"loss": 34.644,
|
| 1060 |
"step": 149
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 0.3961703532518983,
|
| 1064 |
+
"grad_norm": 47252.56640625,
|
| 1065 |
"learning_rate": 9.999825473436369e-05,
|
| 1066 |
+
"loss": 39.1459,
|
| 1067 |
"step": 150
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 0.3988114889402443,
|
| 1071 |
+
"grad_norm": 5307.408203125,
|
| 1072 |
"learning_rate": 9.999818422605875e-05,
|
| 1073 |
+
"loss": 32.3124,
|
| 1074 |
"step": 151
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 0.40145262462859027,
|
| 1078 |
+
"grad_norm": 8414.1484375,
|
| 1079 |
"learning_rate": 9.999811232158389e-05,
|
| 1080 |
+
"loss": 31.5456,
|
| 1081 |
"step": 152
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 0.4040937603169363,
|
| 1085 |
+
"grad_norm": 5779.16943359375,
|
| 1086 |
"learning_rate": 9.999803902094109e-05,
|
| 1087 |
+
"loss": 32.0291,
|
| 1088 |
"step": 153
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 0.4067348960052823,
|
| 1092 |
+
"grad_norm": 6989.2958984375,
|
| 1093 |
"learning_rate": 9.999796432413244e-05,
|
| 1094 |
+
"loss": 32.4468,
|
| 1095 |
"step": 154
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 0.40937603169362824,
|
| 1099 |
+
"grad_norm": 10169.005859375,
|
| 1100 |
"learning_rate": 9.999788823116001e-05,
|
| 1101 |
+
"loss": 33.1476,
|
| 1102 |
"step": 155
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 0.41201716738197425,
|
| 1106 |
+
"grad_norm": 6967.77197265625,
|
| 1107 |
"learning_rate": 9.999781074202592e-05,
|
| 1108 |
+
"loss": 32.4884,
|
| 1109 |
"step": 156
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 0.41465830307032026,
|
| 1113 |
+
"grad_norm": 8052.6611328125,
|
| 1114 |
"learning_rate": 9.999773185673232e-05,
|
| 1115 |
+
"loss": 33.8162,
|
| 1116 |
"step": 157
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 0.4172994387586662,
|
| 1120 |
+
"grad_norm": 13675.26953125,
|
| 1121 |
"learning_rate": 9.999765157528145e-05,
|
| 1122 |
+
"loss": 33.4981,
|
| 1123 |
"step": 158
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 0.4199405744470122,
|
| 1127 |
+
"grad_norm": 23900.8515625,
|
| 1128 |
"learning_rate": 9.99975698976755e-05,
|
| 1129 |
+
"loss": 33.6758,
|
| 1130 |
"step": 159
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 0.4225817101353582,
|
| 1134 |
+
"grad_norm": 8697.0146484375,
|
| 1135 |
"learning_rate": 9.99974868239168e-05,
|
| 1136 |
+
"loss": 33.4007,
|
| 1137 |
"step": 160
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 0.4252228458237042,
|
| 1141 |
+
"grad_norm": 7423.0234375,
|
| 1142 |
"learning_rate": 9.999740235400765e-05,
|
| 1143 |
+
"loss": 32.8796,
|
| 1144 |
"step": 161
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 0.4278639815120502,
|
| 1148 |
+
"grad_norm": 8968.0107421875,
|
| 1149 |
"learning_rate": 9.999731648795041e-05,
|
| 1150 |
+
"loss": 35.1091,
|
| 1151 |
"step": 162
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 0.4305051172003962,
|
| 1155 |
+
"grad_norm": 8960.2470703125,
|
| 1156 |
"learning_rate": 9.999722922574749e-05,
|
| 1157 |
+
"loss": 34.6028,
|
| 1158 |
"step": 163
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 0.43314625288874214,
|
| 1162 |
+
"grad_norm": 9324.4716796875,
|
| 1163 |
"learning_rate": 9.999714056740129e-05,
|
| 1164 |
+
"loss": 35.0468,
|
| 1165 |
"step": 164
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 0.43578738857708815,
|
| 1169 |
+
"grad_norm": 15031.443359375,
|
| 1170 |
"learning_rate": 9.999705051291432e-05,
|
| 1171 |
+
"loss": 33.8078,
|
| 1172 |
"step": 165
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 0.43842852426543416,
|
| 1176 |
+
"grad_norm": 10380.2470703125,
|
| 1177 |
"learning_rate": 9.999695906228908e-05,
|
| 1178 |
+
"loss": 34.8672,
|
| 1179 |
"step": 166
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 0.4410696599537801,
|
| 1183 |
+
"grad_norm": 18920.16796875,
|
| 1184 |
"learning_rate": 9.999686621552813e-05,
|
| 1185 |
+
"loss": 34.697,
|
| 1186 |
"step": 167
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 0.4437107956421261,
|
| 1190 |
+
"grad_norm": 17273.609375,
|
| 1191 |
"learning_rate": 9.999677197263406e-05,
|
| 1192 |
+
"loss": 35.5471,
|
| 1193 |
"step": 168
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 0.44635193133047213,
|
| 1197 |
+
"grad_norm": 10327.810546875,
|
| 1198 |
"learning_rate": 9.999667633360952e-05,
|
| 1199 |
+
"loss": 33.3773,
|
| 1200 |
"step": 169
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.4489930670188181,
|
| 1204 |
+
"grad_norm": 15529.2939453125,
|
| 1205 |
"learning_rate": 9.999657929845714e-05,
|
| 1206 |
+
"loss": 35.3255,
|
| 1207 |
"step": 170
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.4516342027071641,
|
| 1211 |
+
"grad_norm": 15885.65625,
|
| 1212 |
"learning_rate": 9.999648086717966e-05,
|
| 1213 |
+
"loss": 35.0333,
|
| 1214 |
"step": 171
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.45427533839551004,
|
| 1218 |
+
"grad_norm": 16440.353515625,
|
| 1219 |
"learning_rate": 9.999638103977982e-05,
|
| 1220 |
+
"loss": 36.1782,
|
| 1221 |
"step": 172
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.45691647408385605,
|
| 1225 |
+
"grad_norm": 623966.3125,
|
| 1226 |
"learning_rate": 9.999627981626041e-05,
|
| 1227 |
+
"loss": 117.4766,
|
| 1228 |
"step": 173
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.45955760977220206,
|
| 1232 |
+
"grad_norm": 443642.1875,
|
| 1233 |
"learning_rate": 9.999617719662426e-05,
|
| 1234 |
+
"loss": 256.9298,
|
| 1235 |
"step": 174
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.462198745460548,
|
| 1239 |
+
"grad_norm": 426303.78125,
|
| 1240 |
"learning_rate": 9.999607318087423e-05,
|
| 1241 |
+
"loss": 213.1021,
|
| 1242 |
"step": 175
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.464839881148894,
|
| 1246 |
+
"grad_norm": 753837.9375,
|
| 1247 |
"learning_rate": 9.999596776901322e-05,
|
| 1248 |
+
"loss": 234.3458,
|
| 1249 |
"step": 176
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.46748101683724,
|
| 1253 |
+
"grad_norm": 500841.875,
|
| 1254 |
"learning_rate": 9.999586096104419e-05,
|
| 1255 |
+
"loss": 242.1502,
|
| 1256 |
"step": 177
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.470122152525586,
|
| 1260 |
+
"grad_norm": 488348.28125,
|
| 1261 |
"learning_rate": 9.99957527569701e-05,
|
| 1262 |
+
"loss": 259.3533,
|
| 1263 |
"step": 178
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.472763288213932,
|
| 1267 |
+
"grad_norm": 599034.6875,
|
| 1268 |
"learning_rate": 9.999564315679398e-05,
|
| 1269 |
+
"loss": 254.9457,
|
| 1270 |
"step": 179
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.475404423902278,
|
| 1274 |
+
"grad_norm": 740236.3125,
|
| 1275 |
"learning_rate": 9.99955321605189e-05,
|
| 1276 |
+
"loss": 200.2197,
|
| 1277 |
"step": 180
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.47804555959062395,
|
| 1281 |
+
"grad_norm": 279145.40625,
|
| 1282 |
"learning_rate": 9.999541976814796e-05,
|
| 1283 |
+
"loss": 211.6974,
|
| 1284 |
"step": 181
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.48068669527896996,
|
| 1288 |
+
"grad_norm": 565175.5625,
|
| 1289 |
"learning_rate": 9.999530597968428e-05,
|
| 1290 |
+
"loss": 152.1028,
|
| 1291 |
"step": 182
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.48332783096731596,
|
| 1295 |
+
"grad_norm": 29268.025390625,
|
| 1296 |
"learning_rate": 9.999519079513107e-05,
|
| 1297 |
+
"loss": 37.5746,
|
| 1298 |
"step": 183
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.4859689666556619,
|
| 1302 |
+
"grad_norm": 19322.490234375,
|
| 1303 |
"learning_rate": 9.999507421449151e-05,
|
| 1304 |
+
"loss": 38.4138,
|
| 1305 |
"step": 184
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.4886101023440079,
|
| 1309 |
+
"grad_norm": 27010.8203125,
|
| 1310 |
"learning_rate": 9.999495623776886e-05,
|
| 1311 |
+
"loss": 35.2608,
|
| 1312 |
"step": 185
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.49125123803235393,
|
| 1316 |
+
"grad_norm": 25924.7890625,
|
| 1317 |
"learning_rate": 9.999483686496645e-05,
|
| 1318 |
+
"loss": 38.389,
|
| 1319 |
"step": 186
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.4938923737206999,
|
| 1323 |
+
"grad_norm": 33607.66015625,
|
| 1324 |
"learning_rate": 9.999471609608757e-05,
|
| 1325 |
+
"loss": 35.7422,
|
| 1326 |
"step": 187
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.4965335094090459,
|
| 1330 |
+
"grad_norm": 19824.349609375,
|
| 1331 |
"learning_rate": 9.999459393113561e-05,
|
| 1332 |
+
"loss": 37.8325,
|
| 1333 |
"step": 188
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.4991746450973919,
|
| 1337 |
+
"grad_norm": 16384.638671875,
|
| 1338 |
"learning_rate": 9.9994470370114e-05,
|
| 1339 |
+
"loss": 37.3911,
|
| 1340 |
"step": 189
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.5018157807857379,
|
| 1344 |
+
"grad_norm": 15732.8330078125,
|
| 1345 |
"learning_rate": 9.999434541302616e-05,
|
| 1346 |
+
"loss": 35.9949,
|
| 1347 |
"step": 190
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 0.5044569164740839,
|
| 1351 |
+
"grad_norm": 23623.61328125,
|
| 1352 |
"learning_rate": 9.99942190598756e-05,
|
| 1353 |
+
"loss": 36.3237,
|
| 1354 |
"step": 191
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 0.5070980521624299,
|
| 1358 |
+
"grad_norm": 32387.189453125,
|
| 1359 |
"learning_rate": 9.999409131066583e-05,
|
| 1360 |
+
"loss": 36.7266,
|
| 1361 |
"step": 192
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 0.5097391878507759,
|
| 1365 |
+
"grad_norm": 19656.185546875,
|
| 1366 |
"learning_rate": 9.999396216540044e-05,
|
| 1367 |
+
"loss": 36.9575,
|
| 1368 |
"step": 193
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 0.5123803235391218,
|
| 1372 |
+
"grad_norm": 20705.455078125,
|
| 1373 |
"learning_rate": 9.999383162408304e-05,
|
| 1374 |
+
"loss": 37.2048,
|
| 1375 |
"step": 194
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 0.5150214592274678,
|
| 1379 |
+
"grad_norm": 21470.52734375,
|
| 1380 |
"learning_rate": 9.999369968671723e-05,
|
| 1381 |
+
"loss": 36.1668,
|
| 1382 |
"step": 195
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 0.5176625949158138,
|
| 1386 |
+
"grad_norm": 19358.25,
|
| 1387 |
"learning_rate": 9.999356635330674e-05,
|
| 1388 |
+
"loss": 33.7397,
|
| 1389 |
"step": 196
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 0.5203037306041598,
|
| 1393 |
+
"grad_norm": 19253.916015625,
|
| 1394 |
"learning_rate": 9.999343162385529e-05,
|
| 1395 |
+
"loss": 36.7927,
|
| 1396 |
"step": 197
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 0.5229448662925058,
|
| 1400 |
+
"grad_norm": 41119.46875,
|
| 1401 |
"learning_rate": 9.99932954983666e-05,
|
| 1402 |
+
"loss": 36.7557,
|
| 1403 |
"step": 198
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 0.5255860019808518,
|
| 1407 |
+
"grad_norm": 23741.87109375,
|
| 1408 |
"learning_rate": 9.999315797684451e-05,
|
| 1409 |
+
"loss": 38.4819,
|
| 1410 |
"step": 199
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 0.5282271376691977,
|
| 1414 |
+
"grad_norm": 33874.09765625,
|
| 1415 |
"learning_rate": 9.999301905929286e-05,
|
| 1416 |
+
"loss": 42.3858,
|
| 1417 |
"step": 200
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 0.5282271376691977,
|
| 1421 |
+
"eval_loss": 7.334134578704834,
|
| 1422 |
+
"eval_runtime": 2.2174,
|
| 1423 |
+
"eval_samples_per_second": 223.237,
|
| 1424 |
+
"eval_steps_per_second": 27.961,
|
| 1425 |
"step": 200
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 0.5308682733575437,
|
| 1429 |
+
"grad_norm": 15801.6083984375,
|
| 1430 |
"learning_rate": 9.999287874571552e-05,
|
| 1431 |
+
"loss": 38.8128,
|
| 1432 |
"step": 201
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 0.5335094090458897,
|
| 1436 |
+
"grad_norm": 12974.27734375,
|
| 1437 |
"learning_rate": 9.99927370361164e-05,
|
| 1438 |
+
"loss": 38.6081,
|
| 1439 |
"step": 202
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 0.5361505447342357,
|
| 1443 |
+
"grad_norm": 12007.9013671875,
|
| 1444 |
"learning_rate": 9.999259393049947e-05,
|
| 1445 |
+
"loss": 37.1496,
|
| 1446 |
"step": 203
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 0.5387916804225817,
|
| 1450 |
+
"grad_norm": 13070.220703125,
|
| 1451 |
"learning_rate": 9.999244942886871e-05,
|
| 1452 |
+
"loss": 38.7187,
|
| 1453 |
"step": 204
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 0.5414328161109278,
|
| 1457 |
+
"grad_norm": 16807.220703125,
|
| 1458 |
"learning_rate": 9.999230353122819e-05,
|
| 1459 |
+
"loss": 41.07,
|
| 1460 |
"step": 205
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 0.5440739517992736,
|
| 1464 |
+
"grad_norm": 14268.9052734375,
|
| 1465 |
"learning_rate": 9.999215623758194e-05,
|
| 1466 |
+
"loss": 40.1817,
|
| 1467 |
"step": 206
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.5467150874876197,
|
| 1471 |
+
"grad_norm": 13336.4287109375,
|
| 1472 |
"learning_rate": 9.99920075479341e-05,
|
| 1473 |
+
"loss": 37.3859,
|
| 1474 |
"step": 207
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 0.5493562231759657,
|
| 1478 |
+
"grad_norm": 15000.0390625,
|
| 1479 |
"learning_rate": 9.999185746228882e-05,
|
| 1480 |
+
"loss": 37.9181,
|
| 1481 |
"step": 208
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.5519973588643117,
|
| 1485 |
+
"grad_norm": 11059.775390625,
|
| 1486 |
"learning_rate": 9.999170598065028e-05,
|
| 1487 |
+
"loss": 37.7867,
|
| 1488 |
"step": 209
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 0.5546384945526577,
|
| 1492 |
+
"grad_norm": 12954.494140625,
|
| 1493 |
"learning_rate": 9.999155310302273e-05,
|
| 1494 |
+
"loss": 38.3371,
|
| 1495 |
"step": 210
|
| 1496 |
},
|
| 1497 |
{
|
| 1498 |
"epoch": 0.5572796302410036,
|
| 1499 |
+
"grad_norm": 10920.3037109375,
|
| 1500 |
"learning_rate": 9.999139882941043e-05,
|
| 1501 |
+
"loss": 35.1785,
|
| 1502 |
"step": 211
|
| 1503 |
},
|
| 1504 |
{
|
| 1505 |
"epoch": 0.5599207659293496,
|
| 1506 |
+
"grad_norm": 15022.30078125,
|
| 1507 |
"learning_rate": 9.999124315981766e-05,
|
| 1508 |
+
"loss": 35.528,
|
| 1509 |
"step": 212
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
"epoch": 0.5625619016176956,
|
| 1513 |
+
"grad_norm": 10339.8525390625,
|
| 1514 |
"learning_rate": 9.999108609424881e-05,
|
| 1515 |
+
"loss": 34.5773,
|
| 1516 |
"step": 213
|
| 1517 |
},
|
| 1518 |
{
|
| 1519 |
"epoch": 0.5652030373060416,
|
| 1520 |
+
"grad_norm": 9615.1484375,
|
| 1521 |
"learning_rate": 9.999092763270823e-05,
|
| 1522 |
+
"loss": 34.6027,
|
| 1523 |
"step": 214
|
| 1524 |
},
|
| 1525 |
{
|
| 1526 |
"epoch": 0.5678441729943876,
|
| 1527 |
+
"grad_norm": 13707.630859375,
|
| 1528 |
"learning_rate": 9.999076777520037e-05,
|
| 1529 |
+
"loss": 34.8469,
|
| 1530 |
"step": 215
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 0.5704853086827336,
|
| 1534 |
+
"grad_norm": 13718.404296875,
|
| 1535 |
"learning_rate": 9.99906065217297e-05,
|
| 1536 |
+
"loss": 34.0409,
|
| 1537 |
"step": 216
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
"epoch": 0.5731264443710795,
|
| 1541 |
+
"grad_norm": 12160.12109375,
|
| 1542 |
"learning_rate": 9.99904438723007e-05,
|
| 1543 |
+
"loss": 32.9267,
|
| 1544 |
"step": 217
|
| 1545 |
},
|
| 1546 |
{
|
| 1547 |
"epoch": 0.5757675800594255,
|
| 1548 |
+
"grad_norm": 9693.056640625,
|
| 1549 |
"learning_rate": 9.999027982691793e-05,
|
| 1550 |
+
"loss": 33.0474,
|
| 1551 |
"step": 218
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
"epoch": 0.5784087157477715,
|
| 1555 |
+
"grad_norm": 14817.9755859375,
|
| 1556 |
"learning_rate": 9.999011438558595e-05,
|
| 1557 |
+
"loss": 33.6275,
|
| 1558 |
"step": 219
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
"epoch": 0.5810498514361175,
|
| 1562 |
+
"grad_norm": 12656.400390625,
|
| 1563 |
"learning_rate": 9.99899475483094e-05,
|
| 1564 |
+
"loss": 33.9675,
|
| 1565 |
"step": 220
|
| 1566 |
},
|
| 1567 |
{
|
| 1568 |
"epoch": 0.5836909871244635,
|
| 1569 |
+
"grad_norm": 17197.283203125,
|
| 1570 |
"learning_rate": 9.998977931509291e-05,
|
| 1571 |
+
"loss": 35.6857,
|
| 1572 |
"step": 221
|
| 1573 |
},
|
| 1574 |
{
|
| 1575 |
"epoch": 0.5863321228128096,
|
| 1576 |
+
"grad_norm": 215147.109375,
|
| 1577 |
"learning_rate": 9.998960968594121e-05,
|
| 1578 |
+
"loss": 88.1464,
|
| 1579 |
"step": 222
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
"epoch": 0.5889732585011554,
|
| 1583 |
+
"grad_norm": 625456.3125,
|
| 1584 |
"learning_rate": 9.998943866085903e-05,
|
| 1585 |
+
"loss": 186.8345,
|
| 1586 |
"step": 223
|
| 1587 |
},
|
| 1588 |
{
|
| 1589 |
"epoch": 0.5916143941895015,
|
| 1590 |
+
"grad_norm": 491068.96875,
|
| 1591 |
"learning_rate": 9.998926623985114e-05,
|
| 1592 |
+
"loss": 158.0338,
|
| 1593 |
"step": 224
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 0.5942555298778475,
|
| 1597 |
+
"grad_norm": 626101.125,
|
| 1598 |
"learning_rate": 9.998909242292235e-05,
|
| 1599 |
+
"loss": 218.7658,
|
| 1600 |
"step": 225
|
| 1601 |
},
|
| 1602 |
{
|
| 1603 |
"epoch": 0.5968966655661935,
|
| 1604 |
+
"grad_norm": 303837.34375,
|
| 1605 |
"learning_rate": 9.998891721007752e-05,
|
| 1606 |
+
"loss": 186.0703,
|
| 1607 |
"step": 226
|
| 1608 |
},
|
| 1609 |
{
|
| 1610 |
"epoch": 0.5995378012545395,
|
| 1611 |
+
"grad_norm": 354231.84375,
|
| 1612 |
"learning_rate": 9.998874060132155e-05,
|
| 1613 |
+
"loss": 162.2602,
|
| 1614 |
"step": 227
|
| 1615 |
},
|
| 1616 |
{
|
| 1617 |
"epoch": 0.6021789369428855,
|
| 1618 |
+
"grad_norm": 570096.0625,
|
| 1619 |
"learning_rate": 9.998856259665936e-05,
|
| 1620 |
+
"loss": 165.2661,
|
| 1621 |
"step": 228
|
| 1622 |
},
|
| 1623 |
{
|
| 1624 |
"epoch": 0.6048200726312314,
|
| 1625 |
+
"grad_norm": 405688.65625,
|
| 1626 |
"learning_rate": 9.998838319609591e-05,
|
| 1627 |
+
"loss": 159.5345,
|
| 1628 |
"step": 229
|
| 1629 |
},
|
| 1630 |
{
|
| 1631 |
"epoch": 0.6074612083195774,
|
| 1632 |
+
"grad_norm": 592211.125,
|
| 1633 |
"learning_rate": 9.998820239963624e-05,
|
| 1634 |
+
"loss": 141.6046,
|
| 1635 |
"step": 230
|
| 1636 |
},
|
| 1637 |
{
|
| 1638 |
"epoch": 0.6101023440079234,
|
| 1639 |
+
"grad_norm": 678225.0625,
|
| 1640 |
"learning_rate": 9.998802020728537e-05,
|
| 1641 |
+
"loss": 84.9725,
|
| 1642 |
"step": 231
|
| 1643 |
},
|
| 1644 |
{
|
| 1645 |
"epoch": 0.6127434796962694,
|
| 1646 |
+
"grad_norm": 22088.375,
|
| 1647 |
"learning_rate": 9.998783661904843e-05,
|
| 1648 |
+
"loss": 38.1227,
|
| 1649 |
"step": 232
|
| 1650 |
},
|
| 1651 |
{
|
| 1652 |
"epoch": 0.6153846153846154,
|
| 1653 |
+
"grad_norm": 19927.962890625,
|
| 1654 |
"learning_rate": 9.99876516349305e-05,
|
| 1655 |
+
"loss": 37.8816,
|
| 1656 |
"step": 233
|
| 1657 |
},
|
| 1658 |
{
|
| 1659 |
"epoch": 0.6180257510729614,
|
| 1660 |
+
"grad_norm": 33203.27734375,
|
| 1661 |
"learning_rate": 9.998746525493674e-05,
|
| 1662 |
+
"loss": 34.0087,
|
| 1663 |
"step": 234
|
| 1664 |
},
|
| 1665 |
{
|
| 1666 |
"epoch": 0.6206668867613073,
|
| 1667 |
+
"grad_norm": 10135.03515625,
|
| 1668 |
"learning_rate": 9.99872774790724e-05,
|
| 1669 |
+
"loss": 34.0175,
|
| 1670 |
"step": 235
|
| 1671 |
},
|
| 1672 |
{
|
| 1673 |
"epoch": 0.6233080224496533,
|
| 1674 |
+
"grad_norm": 11513.166015625,
|
| 1675 |
"learning_rate": 9.99870883073427e-05,
|
| 1676 |
+
"loss": 32.6651,
|
| 1677 |
"step": 236
|
| 1678 |
},
|
| 1679 |
{
|
| 1680 |
"epoch": 0.6259491581379993,
|
| 1681 |
+
"grad_norm": 7397.00732421875,
|
| 1682 |
"learning_rate": 9.998689773975291e-05,
|
| 1683 |
+
"loss": 32.2064,
|
| 1684 |
"step": 237
|
| 1685 |
},
|
| 1686 |
{
|
| 1687 |
"epoch": 0.6285902938263453,
|
| 1688 |
+
"grad_norm": 10573.4638671875,
|
| 1689 |
"learning_rate": 9.998670577630838e-05,
|
| 1690 |
+
"loss": 32.1057,
|
| 1691 |
"step": 238
|
| 1692 |
},
|
| 1693 |
{
|
| 1694 |
"epoch": 0.6312314295146914,
|
| 1695 |
+
"grad_norm": 10578.8310546875,
|
| 1696 |
"learning_rate": 9.998651241701445e-05,
|
| 1697 |
+
"loss": 32.1381,
|
| 1698 |
"step": 239
|
| 1699 |
},
|
| 1700 |
{
|
| 1701 |
"epoch": 0.6338725652030373,
|
| 1702 |
+
"grad_norm": 9302.189453125,
|
| 1703 |
"learning_rate": 9.998631766187651e-05,
|
| 1704 |
+
"loss": 32.8179,
|
| 1705 |
"step": 240
|
| 1706 |
},
|
| 1707 |
{
|
| 1708 |
"epoch": 0.6365137008913833,
|
| 1709 |
+
"grad_norm": 8694.892578125,
|
| 1710 |
"learning_rate": 9.998612151090003e-05,
|
| 1711 |
+
"loss": 32.7711,
|
| 1712 |
"step": 241
|
| 1713 |
},
|
| 1714 |
{
|
| 1715 |
"epoch": 0.6391548365797293,
|
| 1716 |
+
"grad_norm": 10467.7099609375,
|
| 1717 |
"learning_rate": 9.998592396409047e-05,
|
| 1718 |
+
"loss": 33.1121,
|
| 1719 |
"step": 242
|
| 1720 |
},
|
| 1721 |
{
|
| 1722 |
"epoch": 0.6417959722680753,
|
| 1723 |
+
"grad_norm": 11832.251953125,
|
| 1724 |
"learning_rate": 9.998572502145334e-05,
|
| 1725 |
+
"loss": 32.8568,
|
| 1726 |
"step": 243
|
| 1727 |
},
|
| 1728 |
{
|
| 1729 |
"epoch": 0.6444371079564213,
|
| 1730 |
+
"grad_norm": 14376.9228515625,
|
| 1731 |
"learning_rate": 9.998552468299421e-05,
|
| 1732 |
+
"loss": 32.5907,
|
| 1733 |
"step": 244
|
| 1734 |
},
|
| 1735 |
{
|
| 1736 |
"epoch": 0.6470782436447673,
|
| 1737 |
+
"grad_norm": 13190.787109375,
|
| 1738 |
"learning_rate": 9.998532294871866e-05,
|
| 1739 |
+
"loss": 32.6583,
|
| 1740 |
"step": 245
|
| 1741 |
},
|
| 1742 |
{
|
| 1743 |
"epoch": 0.6497193793331132,
|
| 1744 |
+
"grad_norm": 10301.1328125,
|
| 1745 |
"learning_rate": 9.998511981863232e-05,
|
| 1746 |
+
"loss": 31.7794,
|
| 1747 |
"step": 246
|
| 1748 |
},
|
| 1749 |
{
|
| 1750 |
"epoch": 0.6523605150214592,
|
| 1751 |
+
"grad_norm": 18970.587890625,
|
| 1752 |
"learning_rate": 9.998491529274089e-05,
|
| 1753 |
+
"loss": 32.5321,
|
| 1754 |
"step": 247
|
| 1755 |
},
|
| 1756 |
{
|
| 1757 |
"epoch": 0.6550016507098052,
|
| 1758 |
+
"grad_norm": 10323.8408203125,
|
| 1759 |
"learning_rate": 9.998470937105006e-05,
|
| 1760 |
+
"loss": 32.6962,
|
| 1761 |
"step": 248
|
| 1762 |
},
|
| 1763 |
{
|
| 1764 |
"epoch": 0.6576427863981512,
|
| 1765 |
+
"grad_norm": 13553.1123046875,
|
| 1766 |
"learning_rate": 9.998450205356557e-05,
|
| 1767 |
+
"loss": 34.1782,
|
| 1768 |
"step": 249
|
| 1769 |
},
|
| 1770 |
{
|
| 1771 |
"epoch": 0.6602839220864972,
|
| 1772 |
+
"grad_norm": 34080.28125,
|
| 1773 |
"learning_rate": 9.998429334029323e-05,
|
| 1774 |
+
"loss": 37.3095,
|
| 1775 |
"step": 250
|
| 1776 |
},
|
| 1777 |
{
|
| 1778 |
"epoch": 0.6629250577748432,
|
| 1779 |
+
"grad_norm": 12205.15234375,
|
| 1780 |
"learning_rate": 9.998408323123887e-05,
|
| 1781 |
+
"loss": 33.7182,
|
| 1782 |
"step": 251
|
| 1783 |
},
|
| 1784 |
{
|
| 1785 |
"epoch": 0.6655661934631891,
|
| 1786 |
+
"grad_norm": 11019.15234375,
|
| 1787 |
"learning_rate": 9.998387172640834e-05,
|
| 1788 |
+
"loss": 34.2941,
|
| 1789 |
"step": 252
|
| 1790 |
},
|
| 1791 |
{
|
| 1792 |
"epoch": 0.6682073291515351,
|
| 1793 |
+
"grad_norm": 10185.3310546875,
|
| 1794 |
"learning_rate": 9.998365882580756e-05,
|
| 1795 |
+
"loss": 34.5573,
|
| 1796 |
"step": 253
|
| 1797 |
},
|
| 1798 |
{
|
| 1799 |
"epoch": 0.6708484648398811,
|
| 1800 |
+
"grad_norm": 8710.2685546875,
|
| 1801 |
"learning_rate": 9.998344452944247e-05,
|
| 1802 |
+
"loss": 33.6592,
|
| 1803 |
"step": 254
|
| 1804 |
},
|
| 1805 |
{
|
| 1806 |
"epoch": 0.6734896005282272,
|
| 1807 |
+
"grad_norm": 8050.28759765625,
|
| 1808 |
"learning_rate": 9.998322883731903e-05,
|
| 1809 |
+
"loss": 33.1733,
|
| 1810 |
"step": 255
|
| 1811 |
},
|
| 1812 |
{
|
| 1813 |
"epoch": 0.6761307362165732,
|
| 1814 |
+
"grad_norm": 6891.90673828125,
|
| 1815 |
"learning_rate": 9.998301174944332e-05,
|
| 1816 |
+
"loss": 32.2699,
|
| 1817 |
"step": 256
|
| 1818 |
},
|
| 1819 |
{
|
| 1820 |
"epoch": 0.6787718719049192,
|
| 1821 |
+
"grad_norm": 6904.37060546875,
|
| 1822 |
"learning_rate": 9.998279326582134e-05,
|
| 1823 |
+
"loss": 33.2969,
|
| 1824 |
"step": 257
|
| 1825 |
},
|
| 1826 |
{
|
| 1827 |
"epoch": 0.6814130075932651,
|
| 1828 |
+
"grad_norm": 6681.41162109375,
|
| 1829 |
"learning_rate": 9.998257338645924e-05,
|
| 1830 |
+
"loss": 32.5617,
|
| 1831 |
"step": 258
|
| 1832 |
},
|
| 1833 |
{
|
| 1834 |
"epoch": 0.6840541432816111,
|
| 1835 |
+
"grad_norm": 7499.51025390625,
|
| 1836 |
"learning_rate": 9.998235211136312e-05,
|
| 1837 |
+
"loss": 31.2502,
|
| 1838 |
"step": 259
|
| 1839 |
},
|
| 1840 |
{
|
| 1841 |
"epoch": 0.6866952789699571,
|
| 1842 |
+
"grad_norm": 5850.79931640625,
|
| 1843 |
"learning_rate": 9.99821294405392e-05,
|
| 1844 |
+
"loss": 31.384,
|
| 1845 |
"step": 260
|
| 1846 |
},
|
| 1847 |
{
|
| 1848 |
"epoch": 0.6893364146583031,
|
| 1849 |
+
"grad_norm": 5846.03271484375,
|
| 1850 |
"learning_rate": 9.998190537399366e-05,
|
| 1851 |
+
"loss": 31.2545,
|
| 1852 |
"step": 261
|
| 1853 |
},
|
| 1854 |
{
|
| 1855 |
"epoch": 0.6919775503466491,
|
| 1856 |
+
"grad_norm": 7224.54833984375,
|
| 1857 |
"learning_rate": 9.998167991173277e-05,
|
| 1858 |
+
"loss": 31.2568,
|
| 1859 |
"step": 262
|
| 1860 |
},
|
| 1861 |
{
|
| 1862 |
"epoch": 0.6946186860349951,
|
| 1863 |
+
"grad_norm": 6079.56982421875,
|
| 1864 |
"learning_rate": 9.998145305376286e-05,
|
| 1865 |
+
"loss": 31.7204,
|
| 1866 |
"step": 263
|
| 1867 |
},
|
| 1868 |
{
|
| 1869 |
"epoch": 0.697259821723341,
|
| 1870 |
+
"grad_norm": 7802.859375,
|
| 1871 |
"learning_rate": 9.99812248000902e-05,
|
| 1872 |
+
"loss": 30.3375,
|
| 1873 |
"step": 264
|
| 1874 |
},
|
| 1875 |
{
|
| 1876 |
"epoch": 0.699900957411687,
|
| 1877 |
+
"grad_norm": 7014.5146484375,
|
| 1878 |
"learning_rate": 9.998099515072122e-05,
|
| 1879 |
+
"loss": 30.6416,
|
| 1880 |
"step": 265
|
| 1881 |
},
|
| 1882 |
{
|
| 1883 |
"epoch": 0.702542093100033,
|
| 1884 |
+
"grad_norm": 6766.64208984375,
|
| 1885 |
"learning_rate": 9.998076410566229e-05,
|
| 1886 |
+
"loss": 30.4145,
|
| 1887 |
"step": 266
|
| 1888 |
},
|
| 1889 |
{
|
| 1890 |
"epoch": 0.705183228788379,
|
| 1891 |
+
"grad_norm": 6723.0986328125,
|
| 1892 |
"learning_rate": 9.99805316649199e-05,
|
| 1893 |
+
"loss": 29.3229,
|
| 1894 |
"step": 267
|
| 1895 |
},
|
| 1896 |
{
|
| 1897 |
"epoch": 0.707824364476725,
|
| 1898 |
+
"grad_norm": 8847.9677734375,
|
| 1899 |
"learning_rate": 9.998029782850051e-05,
|
| 1900 |
+
"loss": 29.2886,
|
| 1901 |
"step": 268
|
| 1902 |
},
|
| 1903 |
{
|
| 1904 |
"epoch": 0.7104655001650709,
|
| 1905 |
+
"grad_norm": 5896.45458984375,
|
| 1906 |
"learning_rate": 9.998006259641068e-05,
|
| 1907 |
+
"loss": 29.5852,
|
| 1908 |
"step": 269
|
| 1909 |
},
|
| 1910 |
{
|
| 1911 |
"epoch": 0.7131066358534169,
|
| 1912 |
+
"grad_norm": 7112.9150390625,
|
| 1913 |
"learning_rate": 9.997982596865695e-05,
|
| 1914 |
+
"loss": 29.5084,
|
| 1915 |
"step": 270
|
| 1916 |
},
|
| 1917 |
{
|
| 1918 |
"epoch": 0.715747771541763,
|
| 1919 |
+
"grad_norm": 8039.98876953125,
|
| 1920 |
"learning_rate": 9.997958794524594e-05,
|
| 1921 |
+
"loss": 31.9893,
|
| 1922 |
"step": 271
|
| 1923 |
},
|
| 1924 |
{
|
| 1925 |
"epoch": 0.718388907230109,
|
| 1926 |
+
"grad_norm": 179267.265625,
|
| 1927 |
"learning_rate": 9.99793485261843e-05,
|
| 1928 |
+
"loss": 140.9562,
|
| 1929 |
"step": 272
|
| 1930 |
},
|
| 1931 |
{
|
| 1932 |
"epoch": 0.721030042918455,
|
| 1933 |
+
"grad_norm": 578681.125,
|
| 1934 |
"learning_rate": 9.997910771147872e-05,
|
| 1935 |
+
"loss": 262.198,
|
| 1936 |
"step": 273
|
| 1937 |
},
|
| 1938 |
{
|
| 1939 |
"epoch": 0.723671178606801,
|
| 1940 |
+
"grad_norm": 322541.34375,
|
| 1941 |
"learning_rate": 9.99788655011359e-05,
|
| 1942 |
+
"loss": 237.3132,
|
| 1943 |
"step": 274
|
| 1944 |
},
|
| 1945 |
{
|
| 1946 |
"epoch": 0.7263123142951469,
|
| 1947 |
+
"grad_norm": 235946.640625,
|
| 1948 |
"learning_rate": 9.997862189516263e-05,
|
| 1949 |
+
"loss": 300.6354,
|
| 1950 |
"step": 275
|
| 1951 |
},
|
| 1952 |
{
|
| 1953 |
"epoch": 0.7289534499834929,
|
| 1954 |
+
"grad_norm": 262057.515625,
|
| 1955 |
"learning_rate": 9.99783768935657e-05,
|
| 1956 |
+
"loss": 209.6862,
|
| 1957 |
"step": 276
|
| 1958 |
},
|
| 1959 |
{
|
| 1960 |
"epoch": 0.7315945856718389,
|
| 1961 |
+
"grad_norm": 221274.765625,
|
| 1962 |
"learning_rate": 9.997813049635195e-05,
|
| 1963 |
+
"loss": 208.7495,
|
| 1964 |
"step": 277
|
| 1965 |
},
|
| 1966 |
{
|
| 1967 |
"epoch": 0.7342357213601849,
|
| 1968 |
+
"grad_norm": 363778.46875,
|
| 1969 |
"learning_rate": 9.997788270352827e-05,
|
| 1970 |
+
"loss": 234.0036,
|
| 1971 |
"step": 278
|
| 1972 |
},
|
| 1973 |
{
|
| 1974 |
"epoch": 0.7368768570485309,
|
| 1975 |
+
"grad_norm": 198016.546875,
|
| 1976 |
"learning_rate": 9.997763351510157e-05,
|
| 1977 |
+
"loss": 221.2396,
|
| 1978 |
"step": 279
|
| 1979 |
},
|
| 1980 |
{
|
| 1981 |
"epoch": 0.7395179927368769,
|
| 1982 |
+
"grad_norm": 383717.4375,
|
| 1983 |
"learning_rate": 9.997738293107881e-05,
|
| 1984 |
+
"loss": 166.7505,
|
| 1985 |
"step": 280
|
| 1986 |
},
|
| 1987 |
{
|
| 1988 |
"epoch": 0.7421591284252228,
|
| 1989 |
+
"grad_norm": 471310.09375,
|
| 1990 |
"learning_rate": 9.9977130951467e-05,
|
| 1991 |
+
"loss": 155.5116,
|
| 1992 |
"step": 281
|
| 1993 |
},
|
| 1994 |
{
|
| 1995 |
"epoch": 0.7448002641135688,
|
| 1996 |
+
"grad_norm": 135402.15625,
|
| 1997 |
"learning_rate": 9.997687757627316e-05,
|
| 1998 |
+
"loss": 71.9904,
|
| 1999 |
"step": 282
|
| 2000 |
},
|
| 2001 |
{
|
| 2002 |
"epoch": 0.7474413998019148,
|
| 2003 |
+
"grad_norm": 6735.1005859375,
|
| 2004 |
"learning_rate": 9.997662280550437e-05,
|
| 2005 |
+
"loss": 30.8698,
|
| 2006 |
"step": 283
|
| 2007 |
},
|
| 2008 |
{
|
| 2009 |
"epoch": 0.7500825354902608,
|
| 2010 |
+
"grad_norm": 11189.4736328125,
|
| 2011 |
"learning_rate": 9.997636663916776e-05,
|
| 2012 |
+
"loss": 30.6788,
|
| 2013 |
"step": 284
|
| 2014 |
},
|
| 2015 |
{
|
| 2016 |
"epoch": 0.7527236711786068,
|
| 2017 |
+
"grad_norm": 9472.00390625,
|
| 2018 |
"learning_rate": 9.997610907727046e-05,
|
| 2019 |
+
"loss": 32.5548,
|
| 2020 |
"step": 285
|
| 2021 |
},
|
| 2022 |
{
|
| 2023 |
"epoch": 0.7553648068669528,
|
| 2024 |
+
"grad_norm": 10074.7333984375,
|
| 2025 |
"learning_rate": 9.997585011981966e-05,
|
| 2026 |
+
"loss": 30.9945,
|
| 2027 |
"step": 286
|
| 2028 |
},
|
| 2029 |
{
|
| 2030 |
"epoch": 0.7580059425552987,
|
| 2031 |
+
"grad_norm": 11928.4619140625,
|
| 2032 |
"learning_rate": 9.997558976682262e-05,
|
| 2033 |
+
"loss": 30.6684,
|
| 2034 |
"step": 287
|
| 2035 |
},
|
| 2036 |
{
|
| 2037 |
"epoch": 0.7606470782436447,
|
| 2038 |
+
"grad_norm": 13231.986328125,
|
| 2039 |
"learning_rate": 9.997532801828658e-05,
|
| 2040 |
+
"loss": 30.9457,
|
| 2041 |
"step": 288
|
| 2042 |
},
|
| 2043 |
{
|
| 2044 |
"epoch": 0.7632882139319908,
|
| 2045 |
+
"grad_norm": 8904.8466796875,
|
| 2046 |
"learning_rate": 9.997506487421888e-05,
|
| 2047 |
+
"loss": 31.3361,
|
| 2048 |
"step": 289
|
| 2049 |
},
|
| 2050 |
{
|
| 2051 |
"epoch": 0.7659293496203368,
|
| 2052 |
+
"grad_norm": 9125.240234375,
|
| 2053 |
"learning_rate": 9.997480033462683e-05,
|
| 2054 |
+
"loss": 30.7196,
|
| 2055 |
"step": 290
|
| 2056 |
},
|
| 2057 |
{
|
| 2058 |
"epoch": 0.7685704853086828,
|
| 2059 |
+
"grad_norm": 9812.6181640625,
|
| 2060 |
"learning_rate": 9.997453439951784e-05,
|
| 2061 |
+
"loss": 30.7277,
|
| 2062 |
"step": 291
|
| 2063 |
},
|
| 2064 |
{
|
| 2065 |
"epoch": 0.7712116209970287,
|
| 2066 |
+
"grad_norm": 7082.22607421875,
|
| 2067 |
"learning_rate": 9.997426706889935e-05,
|
| 2068 |
+
"loss": 31.2053,
|
| 2069 |
"step": 292
|
| 2070 |
},
|
| 2071 |
{
|
| 2072 |
"epoch": 0.7738527566853747,
|
| 2073 |
+
"grad_norm": 9316.9384765625,
|
| 2074 |
"learning_rate": 9.997399834277878e-05,
|
| 2075 |
+
"loss": 31.5169,
|
| 2076 |
"step": 293
|
| 2077 |
},
|
| 2078 |
{
|
| 2079 |
"epoch": 0.7764938923737207,
|
| 2080 |
+
"grad_norm": 19302.771484375,
|
| 2081 |
"learning_rate": 9.997372822116368e-05,
|
| 2082 |
+
"loss": 31.651,
|
| 2083 |
"step": 294
|
| 2084 |
},
|
| 2085 |
{
|
| 2086 |
"epoch": 0.7791350280620667,
|
| 2087 |
+
"grad_norm": 10954.8271484375,
|
| 2088 |
"learning_rate": 9.99734567040616e-05,
|
| 2089 |
+
"loss": 30.4,
|
| 2090 |
"step": 295
|
| 2091 |
},
|
| 2092 |
{
|
| 2093 |
"epoch": 0.7817761637504127,
|
| 2094 |
+
"grad_norm": 9081.9521484375,
|
| 2095 |
"learning_rate": 9.997318379148007e-05,
|
| 2096 |
+
"loss": 30.8718,
|
| 2097 |
"step": 296
|
| 2098 |
},
|
| 2099 |
{
|
| 2100 |
"epoch": 0.7844172994387587,
|
| 2101 |
+
"grad_norm": 6827.958984375,
|
| 2102 |
"learning_rate": 9.997290948342673e-05,
|
| 2103 |
+
"loss": 31.0843,
|
| 2104 |
"step": 297
|
| 2105 |
},
|
| 2106 |
{
|
| 2107 |
"epoch": 0.7870584351271046,
|
| 2108 |
+
"grad_norm": 10805.7939453125,
|
| 2109 |
"learning_rate": 9.997263377990926e-05,
|
| 2110 |
+
"loss": 31.6845,
|
| 2111 |
"step": 298
|
| 2112 |
},
|
| 2113 |
{
|
| 2114 |
"epoch": 0.7896995708154506,
|
| 2115 |
+
"grad_norm": 11347.0078125,
|
| 2116 |
"learning_rate": 9.997235668093535e-05,
|
| 2117 |
+
"loss": 33.4166,
|
| 2118 |
"step": 299
|
| 2119 |
},
|
| 2120 |
{
|
| 2121 |
"epoch": 0.7923407065037966,
|
| 2122 |
+
"grad_norm": 16983.841796875,
|
| 2123 |
"learning_rate": 9.997207818651274e-05,
|
| 2124 |
+
"loss": 35.7603,
|
| 2125 |
"step": 300
|
| 2126 |
},
|
| 2127 |
{
|
| 2128 |
"epoch": 0.7949818421921426,
|
| 2129 |
+
"grad_norm": 3815.614990234375,
|
| 2130 |
"learning_rate": 9.997179829664918e-05,
|
| 2131 |
+
"loss": 33.1237,
|
| 2132 |
"step": 301
|
| 2133 |
},
|
| 2134 |
{
|
| 2135 |
"epoch": 0.7976229778804886,
|
| 2136 |
+
"grad_norm": 4439.759765625,
|
| 2137 |
"learning_rate": 9.997151701135253e-05,
|
| 2138 |
+
"loss": 32.6201,
|
| 2139 |
"step": 302
|
| 2140 |
},
|
| 2141 |
{
|
| 2142 |
"epoch": 0.8002641135688346,
|
| 2143 |
+
"grad_norm": 6584.0,
|
| 2144 |
"learning_rate": 9.997123433063062e-05,
|
| 2145 |
+
"loss": 31.9738,
|
| 2146 |
"step": 303
|
| 2147 |
},
|
| 2148 |
{
|
| 2149 |
"epoch": 0.8029052492571805,
|
| 2150 |
+
"grad_norm": 8394.333984375,
|
| 2151 |
"learning_rate": 9.997095025449134e-05,
|
| 2152 |
+
"loss": 34.1952,
|
| 2153 |
"step": 304
|
| 2154 |
},
|
| 2155 |
{
|
| 2156 |
"epoch": 0.8055463849455266,
|
| 2157 |
+
"grad_norm": 8264.888671875,
|
| 2158 |
"learning_rate": 9.997066478294262e-05,
|
| 2159 |
+
"loss": 34.1646,
|
| 2160 |
"step": 305
|
| 2161 |
},
|
| 2162 |
{
|
| 2163 |
"epoch": 0.8081875206338726,
|
| 2164 |
+
"grad_norm": 6815.27587890625,
|
| 2165 |
"learning_rate": 9.997037791599245e-05,
|
| 2166 |
+
"loss": 32.8399,
|
| 2167 |
"step": 306
|
| 2168 |
},
|
| 2169 |
{
|
| 2170 |
"epoch": 0.8108286563222186,
|
| 2171 |
+
"grad_norm": 6638.54296875,
|
| 2172 |
"learning_rate": 9.997008965364884e-05,
|
| 2173 |
+
"loss": 32.737,
|
| 2174 |
"step": 307
|
| 2175 |
},
|
| 2176 |
{
|
| 2177 |
"epoch": 0.8134697920105646,
|
| 2178 |
+
"grad_norm": 6356.19287109375,
|
| 2179 |
"learning_rate": 9.996979999591983e-05,
|
| 2180 |
+
"loss": 33.2864,
|
| 2181 |
"step": 308
|
| 2182 |
},
|
| 2183 |
{
|
| 2184 |
"epoch": 0.8161109276989106,
|
| 2185 |
+
"grad_norm": 10876.560546875,
|
| 2186 |
"learning_rate": 9.996950894281349e-05,
|
| 2187 |
+
"loss": 32.8353,
|
| 2188 |
"step": 309
|
| 2189 |
},
|
| 2190 |
{
|
| 2191 |
"epoch": 0.8187520633872565,
|
| 2192 |
+
"grad_norm": 18334.380859375,
|
| 2193 |
"learning_rate": 9.996921649433796e-05,
|
| 2194 |
+
"loss": 33.1125,
|
| 2195 |
"step": 310
|
| 2196 |
},
|
| 2197 |
{
|
| 2198 |
"epoch": 0.8213931990756025,
|
| 2199 |
+
"grad_norm": 5925.57080078125,
|
| 2200 |
"learning_rate": 9.996892265050144e-05,
|
| 2201 |
+
"loss": 33.4775,
|
| 2202 |
"step": 311
|
| 2203 |
},
|
| 2204 |
{
|
| 2205 |
"epoch": 0.8240343347639485,
|
| 2206 |
+
"grad_norm": 5512.29541015625,
|
| 2207 |
"learning_rate": 9.99686274113121e-05,
|
| 2208 |
+
"loss": 32.4073,
|
| 2209 |
"step": 312
|
| 2210 |
},
|
| 2211 |
{
|
| 2212 |
"epoch": 0.8266754704522945,
|
| 2213 |
+
"grad_norm": 6770.63232421875,
|
| 2214 |
"learning_rate": 9.996833077677819e-05,
|
| 2215 |
+
"loss": 33.0255,
|
| 2216 |
"step": 313
|
| 2217 |
},
|
| 2218 |
{
|
| 2219 |
"epoch": 0.8293166061406405,
|
| 2220 |
+
"grad_norm": 9025.830078125,
|
| 2221 |
"learning_rate": 9.9968032746908e-05,
|
| 2222 |
+
"loss": 31.7732,
|
| 2223 |
"step": 314
|
| 2224 |
},
|
| 2225 |
{
|
| 2226 |
"epoch": 0.8319577418289865,
|
| 2227 |
+
"grad_norm": 5815.4296875,
|
| 2228 |
"learning_rate": 9.996773332170983e-05,
|
| 2229 |
+
"loss": 31.5946,
|
| 2230 |
"step": 315
|
| 2231 |
},
|
| 2232 |
{
|
| 2233 |
"epoch": 0.8345988775173324,
|
| 2234 |
+
"grad_norm": 7221.68603515625,
|
| 2235 |
"learning_rate": 9.996743250119209e-05,
|
| 2236 |
+
"loss": 31.5973,
|
| 2237 |
"step": 316
|
| 2238 |
},
|
| 2239 |
{
|
| 2240 |
"epoch": 0.8372400132056784,
|
| 2241 |
+
"grad_norm": 7172.86962890625,
|
| 2242 |
"learning_rate": 9.996713028536313e-05,
|
| 2243 |
+
"loss": 31.4948,
|
| 2244 |
"step": 317
|
| 2245 |
},
|
| 2246 |
{
|
| 2247 |
"epoch": 0.8398811488940244,
|
| 2248 |
+
"grad_norm": 11000.0458984375,
|
| 2249 |
"learning_rate": 9.99668266742314e-05,
|
| 2250 |
+
"loss": 31.3127,
|
| 2251 |
"step": 318
|
| 2252 |
},
|
| 2253 |
{
|
| 2254 |
"epoch": 0.8425222845823704,
|
| 2255 |
+
"grad_norm": 8431.4716796875,
|
| 2256 |
"learning_rate": 9.99665216678054e-05,
|
| 2257 |
+
"loss": 30.8608,
|
| 2258 |
"step": 319
|
| 2259 |
},
|
| 2260 |
{
|
| 2261 |
"epoch": 0.8451634202707164,
|
| 2262 |
+
"grad_norm": 7308.78466796875,
|
| 2263 |
"learning_rate": 9.996621526609364e-05,
|
| 2264 |
+
"loss": 30.8716,
|
| 2265 |
"step": 320
|
| 2266 |
},
|
| 2267 |
{
|
| 2268 |
"epoch": 0.8478045559590623,
|
| 2269 |
+
"grad_norm": 8358.787109375,
|
| 2270 |
"learning_rate": 9.996590746910467e-05,
|
| 2271 |
+
"loss": 31.0737,
|
| 2272 |
"step": 321
|
| 2273 |
},
|
| 2274 |
{
|
| 2275 |
"epoch": 0.8504456916474084,
|
| 2276 |
+
"grad_norm": 29319.46484375,
|
| 2277 |
"learning_rate": 9.996559827684709e-05,
|
| 2278 |
+
"loss": 46.2402,
|
| 2279 |
"step": 322
|
| 2280 |
},
|
| 2281 |
{
|
| 2282 |
"epoch": 0.8530868273357544,
|
| 2283 |
+
"grad_norm": 903961.25,
|
| 2284 |
"learning_rate": 9.996528768932951e-05,
|
| 2285 |
+
"loss": 161.367,
|
| 2286 |
"step": 323
|
| 2287 |
},
|
| 2288 |
{
|
| 2289 |
"epoch": 0.8557279630241004,
|
| 2290 |
+
"grad_norm": 436229.9375,
|
| 2291 |
"learning_rate": 9.996497570656062e-05,
|
| 2292 |
+
"loss": 215.2534,
|
| 2293 |
"step": 324
|
| 2294 |
},
|
| 2295 |
{
|
| 2296 |
"epoch": 0.8583690987124464,
|
| 2297 |
+
"grad_norm": 310716.5,
|
| 2298 |
"learning_rate": 9.996466232854915e-05,
|
| 2299 |
+
"loss": 218.9532,
|
| 2300 |
"step": 325
|
| 2301 |
},
|
| 2302 |
{
|
| 2303 |
"epoch": 0.8610102344007924,
|
| 2304 |
+
"grad_norm": 935038.75,
|
| 2305 |
"learning_rate": 9.996434755530384e-05,
|
| 2306 |
+
"loss": 204.1668,
|
| 2307 |
"step": 326
|
| 2308 |
},
|
| 2309 |
{
|
| 2310 |
"epoch": 0.8636513700891383,
|
| 2311 |
+
"grad_norm": 577125.0625,
|
| 2312 |
"learning_rate": 9.996403138683347e-05,
|
| 2313 |
+
"loss": 225.0228,
|
| 2314 |
"step": 327
|
| 2315 |
},
|
| 2316 |
{
|
| 2317 |
"epoch": 0.8662925057774843,
|
| 2318 |
+
"grad_norm": 429562.09375,
|
| 2319 |
"learning_rate": 9.996371382314686e-05,
|
| 2320 |
+
"loss": 221.4529,
|
| 2321 |
"step": 328
|
| 2322 |
},
|
| 2323 |
{
|
| 2324 |
"epoch": 0.8689336414658303,
|
| 2325 |
+
"grad_norm": 469087.0625,
|
| 2326 |
"learning_rate": 9.996339486425291e-05,
|
| 2327 |
+
"loss": 161.132,
|
| 2328 |
"step": 329
|
| 2329 |
},
|
| 2330 |
{
|
| 2331 |
"epoch": 0.8715747771541763,
|
| 2332 |
+
"grad_norm": 946113.1875,
|
| 2333 |
"learning_rate": 9.99630745101605e-05,
|
| 2334 |
+
"loss": 169.4336,
|
| 2335 |
"step": 330
|
| 2336 |
},
|
| 2337 |
{
|
| 2338 |
"epoch": 0.8742159128425223,
|
| 2339 |
+
"grad_norm": 537740.1875,
|
| 2340 |
"learning_rate": 9.996275276087859e-05,
|
| 2341 |
+
"loss": 166.9042,
|
| 2342 |
"step": 331
|
| 2343 |
},
|
| 2344 |
{
|
| 2345 |
"epoch": 0.8768570485308683,
|
| 2346 |
+
"grad_norm": 377986.5,
|
| 2347 |
"learning_rate": 9.996242961641615e-05,
|
| 2348 |
+
"loss": 139.4483,
|
| 2349 |
"step": 332
|
| 2350 |
},
|
| 2351 |
{
|
| 2352 |
"epoch": 0.8794981842192142,
|
| 2353 |
+
"grad_norm": 6949.21044921875,
|
| 2354 |
"learning_rate": 9.996210507678223e-05,
|
| 2355 |
+
"loss": 32.8323,
|
| 2356 |
"step": 333
|
| 2357 |
},
|
| 2358 |
{
|
| 2359 |
"epoch": 0.8821393199075602,
|
| 2360 |
+
"grad_norm": 6551.869140625,
|
| 2361 |
"learning_rate": 9.996177914198586e-05,
|
| 2362 |
+
"loss": 31.1956,
|
| 2363 |
"step": 334
|
| 2364 |
},
|
| 2365 |
{
|
| 2366 |
"epoch": 0.8847804555959062,
|
| 2367 |
+
"grad_norm": 8210.8974609375,
|
| 2368 |
"learning_rate": 9.996145181203615e-05,
|
| 2369 |
+
"loss": 30.2494,
|
| 2370 |
"step": 335
|
| 2371 |
},
|
| 2372 |
{
|
| 2373 |
"epoch": 0.8874215912842522,
|
| 2374 |
+
"grad_norm": 12632.7666015625,
|
| 2375 |
"learning_rate": 9.996112308694225e-05,
|
| 2376 |
+
"loss": 30.7789,
|
| 2377 |
"step": 336
|
| 2378 |
},
|
| 2379 |
{
|
| 2380 |
"epoch": 0.8900627269725983,
|
| 2381 |
+
"grad_norm": 11905.80078125,
|
| 2382 |
"learning_rate": 9.996079296671334e-05,
|
| 2383 |
+
"loss": 30.9992,
|
| 2384 |
"step": 337
|
| 2385 |
},
|
| 2386 |
{
|
| 2387 |
"epoch": 0.8927038626609443,
|
| 2388 |
+
"grad_norm": 11776.396484375,
|
| 2389 |
"learning_rate": 9.996046145135865e-05,
|
| 2390 |
+
"loss": 30.6118,
|
| 2391 |
"step": 338
|
| 2392 |
},
|
| 2393 |
{
|
| 2394 |
"epoch": 0.8953449983492902,
|
| 2395 |
+
"grad_norm": 10494.625,
|
| 2396 |
"learning_rate": 9.99601285408874e-05,
|
| 2397 |
+
"loss": 30.6983,
|
| 2398 |
"step": 339
|
| 2399 |
},
|
| 2400 |
{
|
| 2401 |
"epoch": 0.8979861340376362,
|
| 2402 |
+
"grad_norm": 8309.9296875,
|
| 2403 |
"learning_rate": 9.995979423530892e-05,
|
| 2404 |
+
"loss": 30.6617,
|
| 2405 |
"step": 340
|
| 2406 |
},
|
| 2407 |
{
|
| 2408 |
"epoch": 0.9006272697259822,
|
| 2409 |
+
"grad_norm": 11482.9853515625,
|
| 2410 |
"learning_rate": 9.995945853463253e-05,
|
| 2411 |
+
"loss": 30.5696,
|
| 2412 |
"step": 341
|
| 2413 |
},
|
| 2414 |
{
|
| 2415 |
"epoch": 0.9032684054143282,
|
| 2416 |
+
"grad_norm": 8950.994140625,
|
| 2417 |
"learning_rate": 9.995912143886763e-05,
|
| 2418 |
+
"loss": 29.6077,
|
| 2419 |
"step": 342
|
| 2420 |
},
|
| 2421 |
{
|
| 2422 |
"epoch": 0.9059095411026742,
|
| 2423 |
+
"grad_norm": 8950.931640625,
|
| 2424 |
"learning_rate": 9.995878294802357e-05,
|
| 2425 |
+
"loss": 30.4176,
|
| 2426 |
"step": 343
|
| 2427 |
},
|
| 2428 |
{
|
| 2429 |
"epoch": 0.9085506767910201,
|
| 2430 |
+
"grad_norm": 6688.57470703125,
|
| 2431 |
"learning_rate": 9.995844306210988e-05,
|
| 2432 |
+
"loss": 29.8723,
|
| 2433 |
"step": 344
|
| 2434 |
},
|
| 2435 |
{
|
| 2436 |
"epoch": 0.9111918124793661,
|
| 2437 |
+
"grad_norm": 7882.67431640625,
|
| 2438 |
"learning_rate": 9.995810178113599e-05,
|
| 2439 |
+
"loss": 30.049,
|
| 2440 |
"step": 345
|
| 2441 |
},
|
| 2442 |
{
|
| 2443 |
"epoch": 0.9138329481677121,
|
| 2444 |
+
"grad_norm": 9309.5625,
|
| 2445 |
"learning_rate": 9.995775910511147e-05,
|
| 2446 |
+
"loss": 30.2998,
|
| 2447 |
"step": 346
|
| 2448 |
},
|
| 2449 |
{
|
| 2450 |
"epoch": 0.9164740838560581,
|
| 2451 |
+
"grad_norm": 9403.8974609375,
|
| 2452 |
"learning_rate": 9.995741503404587e-05,
|
| 2453 |
+
"loss": 30.4171,
|
| 2454 |
"step": 347
|
| 2455 |
},
|
| 2456 |
{
|
| 2457 |
"epoch": 0.9191152195444041,
|
| 2458 |
+
"grad_norm": 10254.1376953125,
|
| 2459 |
"learning_rate": 9.995706956794879e-05,
|
| 2460 |
+
"loss": 32.398,
|
| 2461 |
"step": 348
|
| 2462 |
},
|
| 2463 |
{
|
| 2464 |
"epoch": 0.9217563552327501,
|
| 2465 |
+
"grad_norm": 11519.509765625,
|
| 2466 |
"learning_rate": 9.99567227068299e-05,
|
| 2467 |
+
"loss": 33.4377,
|
| 2468 |
"step": 349
|
| 2469 |
},
|
| 2470 |
{
|
| 2471 |
"epoch": 0.924397490921096,
|
| 2472 |
+
"grad_norm": 17227.236328125,
|
| 2473 |
"learning_rate": 9.995637445069887e-05,
|
| 2474 |
+
"loss": 36.9788,
|
| 2475 |
"step": 350
|
| 2476 |
},
|
| 2477 |
{
|
| 2478 |
"epoch": 0.927038626609442,
|
| 2479 |
+
"grad_norm": 8033.53369140625,
|
| 2480 |
"learning_rate": 9.995602479956545e-05,
|
| 2481 |
+
"loss": 32.1,
|
| 2482 |
"step": 351
|
| 2483 |
},
|
| 2484 |
{
|
| 2485 |
"epoch": 0.929679762297788,
|
| 2486 |
+
"grad_norm": 10333.927734375,
|
| 2487 |
"learning_rate": 9.995567375343937e-05,
|
| 2488 |
+
"loss": 32.4024,
|
| 2489 |
"step": 352
|
| 2490 |
},
|
| 2491 |
{
|
| 2492 |
"epoch": 0.932320897986134,
|
| 2493 |
+
"grad_norm": 5577.73486328125,
|
| 2494 |
"learning_rate": 9.995532131233044e-05,
|
| 2495 |
+
"loss": 33.2651,
|
| 2496 |
"step": 353
|
| 2497 |
},
|
| 2498 |
{
|
| 2499 |
"epoch": 0.93496203367448,
|
| 2500 |
+
"grad_norm": 5001.80615234375,
|
| 2501 |
"learning_rate": 9.99549674762485e-05,
|
| 2502 |
+
"loss": 33.2199,
|
| 2503 |
"step": 354
|
| 2504 |
},
|
| 2505 |
{
|
| 2506 |
"epoch": 0.9376031693628261,
|
| 2507 |
+
"grad_norm": 6995.62255859375,
|
| 2508 |
"learning_rate": 9.995461224520345e-05,
|
| 2509 |
+
"loss": 33.0332,
|
| 2510 |
"step": 355
|
| 2511 |
},
|
| 2512 |
{
|
| 2513 |
"epoch": 0.940244305051172,
|
| 2514 |
+
"grad_norm": 5345.10888671875,
|
| 2515 |
"learning_rate": 9.995425561920519e-05,
|
| 2516 |
+
"loss": 32.4465,
|
| 2517 |
"step": 356
|
| 2518 |
},
|
| 2519 |
{
|
| 2520 |
"epoch": 0.942885440739518,
|
| 2521 |
+
"grad_norm": 5311.36376953125,
|
| 2522 |
"learning_rate": 9.99538975982637e-05,
|
| 2523 |
+
"loss": 33.3183,
|
| 2524 |
"step": 357
|
| 2525 |
},
|
| 2526 |
{
|
| 2527 |
"epoch": 0.945526576427864,
|
| 2528 |
+
"grad_norm": 4239.72021484375,
|
| 2529 |
"learning_rate": 9.995353818238895e-05,
|
| 2530 |
+
"loss": 30.5123,
|
| 2531 |
"step": 358
|
| 2532 |
},
|
| 2533 |
{
|
| 2534 |
"epoch": 0.94816771211621,
|
| 2535 |
+
"grad_norm": 6135.8544921875,
|
| 2536 |
"learning_rate": 9.9953177371591e-05,
|
| 2537 |
+
"loss": 30.1126,
|
| 2538 |
"step": 359
|
| 2539 |
},
|
| 2540 |
{
|
| 2541 |
"epoch": 0.950808847804556,
|
| 2542 |
+
"grad_norm": 3885.701904296875,
|
| 2543 |
"learning_rate": 9.995281516587991e-05,
|
| 2544 |
+
"loss": 30.1448,
|
| 2545 |
"step": 360
|
| 2546 |
},
|
| 2547 |
{
|
| 2548 |
"epoch": 0.953449983492902,
|
| 2549 |
+
"grad_norm": 17259.177734375,
|
| 2550 |
"learning_rate": 9.99524515652658e-05,
|
| 2551 |
+
"loss": 30.9694,
|
| 2552 |
"step": 361
|
| 2553 |
},
|
| 2554 |
{
|
| 2555 |
"epoch": 0.9560911191812479,
|
| 2556 |
+
"grad_norm": 5949.1728515625,
|
| 2557 |
"learning_rate": 9.995208656975884e-05,
|
| 2558 |
+
"loss": 30.8493,
|
| 2559 |
"step": 362
|
| 2560 |
},
|
| 2561 |
{
|
| 2562 |
"epoch": 0.9587322548695939,
|
| 2563 |
+
"grad_norm": 231986.453125,
|
| 2564 |
"learning_rate": 9.995172017936919e-05,
|
| 2565 |
+
"loss": 141.9035,
|
| 2566 |
"step": 363
|
| 2567 |
},
|
| 2568 |
{
|
| 2569 |
"epoch": 0.9613733905579399,
|
| 2570 |
+
"grad_norm": 103330.5546875,
|
| 2571 |
"learning_rate": 9.99513523941071e-05,
|
| 2572 |
+
"loss": 188.4911,
|
| 2573 |
"step": 364
|
| 2574 |
},
|
| 2575 |
{
|
| 2576 |
"epoch": 0.9640145262462859,
|
| 2577 |
+
"grad_norm": 307991.03125,
|
| 2578 |
"learning_rate": 9.995098321398284e-05,
|
| 2579 |
+
"loss": 160.2285,
|
| 2580 |
"step": 365
|
| 2581 |
},
|
| 2582 |
{
|
| 2583 |
"epoch": 0.9666556619346319,
|
| 2584 |
+
"grad_norm": 190517.765625,
|
| 2585 |
"learning_rate": 9.995061263900671e-05,
|
| 2586 |
+
"loss": 152.4148,
|
| 2587 |
"step": 366
|
| 2588 |
},
|
| 2589 |
{
|
| 2590 |
"epoch": 0.9692967976229779,
|
| 2591 |
+
"grad_norm": 134986.078125,
|
| 2592 |
"learning_rate": 9.995024066918908e-05,
|
| 2593 |
+
"loss": 119.8174,
|
| 2594 |
"step": 367
|
| 2595 |
},
|
| 2596 |
{
|
| 2597 |
"epoch": 0.9719379333113238,
|
| 2598 |
+
"grad_norm": 215117.609375,
|
| 2599 |
"learning_rate": 9.994986730454031e-05,
|
| 2600 |
+
"loss": 125.8479,
|
| 2601 |
"step": 368
|
| 2602 |
},
|
| 2603 |
{
|
| 2604 |
"epoch": 0.9745790689996698,
|
| 2605 |
+
"grad_norm": 22283.35546875,
|
| 2606 |
"learning_rate": 9.994949254507084e-05,
|
| 2607 |
+
"loss": 34.5446,
|
| 2608 |
"step": 369
|
| 2609 |
},
|
| 2610 |
{
|
| 2611 |
"epoch": 0.9772202046880158,
|
| 2612 |
+
"grad_norm": 12405.2236328125,
|
| 2613 |
"learning_rate": 9.994911639079112e-05,
|
| 2614 |
+
"loss": 35.1761,
|
| 2615 |
"step": 370
|
| 2616 |
},
|
| 2617 |
{
|
| 2618 |
"epoch": 0.9798613403763619,
|
| 2619 |
+
"grad_norm": 13683.02734375,
|
| 2620 |
"learning_rate": 9.994873884171167e-05,
|
| 2621 |
+
"loss": 32.3272,
|
| 2622 |
"step": 371
|
| 2623 |
},
|
| 2624 |
{
|
| 2625 |
"epoch": 0.9825024760647079,
|
| 2626 |
+
"grad_norm": 8963.904296875,
|
| 2627 |
"learning_rate": 9.994835989784305e-05,
|
| 2628 |
+
"loss": 31.0019,
|
| 2629 |
"step": 372
|
| 2630 |
},
|
| 2631 |
{
|
| 2632 |
"epoch": 0.9851436117530538,
|
| 2633 |
+
"grad_norm": 19926.2734375,
|
| 2634 |
"learning_rate": 9.994797955919581e-05,
|
| 2635 |
+
"loss": 30.4514,
|
| 2636 |
"step": 373
|
| 2637 |
},
|
| 2638 |
{
|
| 2639 |
"epoch": 0.9877847474413998,
|
| 2640 |
+
"grad_norm": 37221.25,
|
| 2641 |
"learning_rate": 9.994759782578058e-05,
|
| 2642 |
+
"loss": 32.0492,
|
| 2643 |
"step": 374
|
| 2644 |
},
|
| 2645 |
{
|
| 2646 |
"epoch": 0.9904258831297458,
|
| 2647 |
+
"grad_norm": 10019.828125,
|
| 2648 |
"learning_rate": 9.994721469760801e-05,
|
| 2649 |
+
"loss": 31.7783,
|
| 2650 |
"step": 375
|
| 2651 |
},
|
| 2652 |
{
|
| 2653 |
"epoch": 0.9930670188180918,
|
| 2654 |
+
"grad_norm": 8898.4228515625,
|
| 2655 |
"learning_rate": 9.994683017468883e-05,
|
| 2656 |
+
"loss": 30.9381,
|
| 2657 |
"step": 376
|
| 2658 |
},
|
| 2659 |
{
|
| 2660 |
"epoch": 0.9957081545064378,
|
| 2661 |
+
"grad_norm": 13350.8203125,
|
| 2662 |
"learning_rate": 9.994644425703374e-05,
|
| 2663 |
+
"loss": 32.4939,
|
| 2664 |
"step": 377
|
| 2665 |
},
|
| 2666 |
{
|
| 2667 |
"epoch": 0.9983492901947838,
|
| 2668 |
+
"grad_norm": 29945.037109375,
|
| 2669 |
"learning_rate": 9.994605694465355e-05,
|
| 2670 |
+
"loss": 34.0366,
|
| 2671 |
"step": 378
|
| 2672 |
},
|
| 2673 |
{
|
| 2674 |
+
"epoch": 1.0009904258831297,
|
| 2675 |
+
"grad_norm": 19480.009765625,
|
| 2676 |
"learning_rate": 9.994566823755907e-05,
|
| 2677 |
+
"loss": 37.1069,
|
| 2678 |
"step": 379
|
| 2679 |
},
|
| 2680 |
{
|
| 2681 |
+
"epoch": 1.0036315615714757,
|
| 2682 |
+
"grad_norm": 4824.83544921875,
|
| 2683 |
"learning_rate": 9.99452781357611e-05,
|
| 2684 |
+
"loss": 35.9486,
|
| 2685 |
"step": 380
|
| 2686 |
},
|
| 2687 |
{
|
| 2688 |
+
"epoch": 1.0062726972598217,
|
| 2689 |
+
"grad_norm": 4898.34423828125,
|
| 2690 |
"learning_rate": 9.994488663927062e-05,
|
| 2691 |
+
"loss": 34.3521,
|
| 2692 |
"step": 381
|
| 2693 |
},
|
| 2694 |
{
|
| 2695 |
+
"epoch": 1.0089138329481677,
|
| 2696 |
+
"grad_norm": 7551.79736328125,
|
| 2697 |
"learning_rate": 9.994449374809851e-05,
|
| 2698 |
+
"loss": 36.7028,
|
| 2699 |
"step": 382
|
| 2700 |
},
|
| 2701 |
{
|
| 2702 |
+
"epoch": 1.0115549686365137,
|
| 2703 |
+
"grad_norm": 8357.705078125,
|
| 2704 |
"learning_rate": 9.994409946225574e-05,
|
| 2705 |
+
"loss": 36.5134,
|
| 2706 |
"step": 383
|
| 2707 |
},
|
| 2708 |
{
|
| 2709 |
+
"epoch": 1.0141961043248597,
|
| 2710 |
+
"grad_norm": 5780.6787109375,
|
| 2711 |
"learning_rate": 9.994370378175332e-05,
|
| 2712 |
+
"loss": 37.3621,
|
| 2713 |
"step": 384
|
| 2714 |
},
|
| 2715 |
{
|
| 2716 |
+
"epoch": 1.0168372400132057,
|
| 2717 |
+
"grad_norm": 5624.93896484375,
|
| 2718 |
"learning_rate": 9.994330670660235e-05,
|
| 2719 |
+
"loss": 37.6676,
|
| 2720 |
"step": 385
|
| 2721 |
},
|
| 2722 |
{
|
| 2723 |
+
"epoch": 1.0194783757015518,
|
| 2724 |
+
"grad_norm": 6545.541015625,
|
| 2725 |
"learning_rate": 9.994290823681385e-05,
|
| 2726 |
+
"loss": 37.2885,
|
| 2727 |
"step": 386
|
| 2728 |
},
|
| 2729 |
{
|
| 2730 |
+
"epoch": 1.0221195113898978,
|
| 2731 |
+
"grad_norm": 9896.431640625,
|
| 2732 |
"learning_rate": 9.994250837239897e-05,
|
| 2733 |
+
"loss": 37.8031,
|
| 2734 |
"step": 387
|
| 2735 |
},
|
| 2736 |
{
|
| 2737 |
+
"epoch": 1.0247606470782435,
|
| 2738 |
+
"grad_norm": 6628.89453125,
|
| 2739 |
"learning_rate": 9.994210711336891e-05,
|
| 2740 |
+
"loss": 39.4998,
|
| 2741 |
"step": 388
|
| 2742 |
},
|
| 2743 |
{
|
| 2744 |
+
"epoch": 1.0274017827665896,
|
| 2745 |
+
"grad_norm": 7230.349609375,
|
| 2746 |
"learning_rate": 9.994170445973483e-05,
|
| 2747 |
+
"loss": 37.6952,
|
| 2748 |
"step": 389
|
| 2749 |
},
|
| 2750 |
{
|
| 2751 |
+
"epoch": 1.0300429184549356,
|
| 2752 |
+
"grad_norm": 5001.923828125,
|
| 2753 |
"learning_rate": 9.994130041150798e-05,
|
| 2754 |
+
"loss": 37.2387,
|
| 2755 |
"step": 390
|
| 2756 |
},
|
| 2757 |
{
|
| 2758 |
+
"epoch": 1.0326840541432816,
|
| 2759 |
+
"grad_norm": 8473.236328125,
|
| 2760 |
"learning_rate": 9.994089496869968e-05,
|
| 2761 |
+
"loss": 37.7243,
|
| 2762 |
"step": 391
|
| 2763 |
},
|
| 2764 |
{
|
| 2765 |
+
"epoch": 1.0353251898316276,
|
| 2766 |
+
"grad_norm": 12679.2109375,
|
| 2767 |
"learning_rate": 9.994048813132119e-05,
|
| 2768 |
+
"loss": 35.9025,
|
| 2769 |
"step": 392
|
| 2770 |
},
|
| 2771 |
{
|
| 2772 |
+
"epoch": 1.0379663255199736,
|
| 2773 |
+
"grad_norm": 7488.9248046875,
|
| 2774 |
"learning_rate": 9.994007989938392e-05,
|
| 2775 |
+
"loss": 36.2572,
|
| 2776 |
"step": 393
|
| 2777 |
},
|
| 2778 |
{
|
| 2779 |
+
"epoch": 1.0406074612083196,
|
| 2780 |
+
"grad_norm": 8192.458984375,
|
| 2781 |
"learning_rate": 9.993967027289927e-05,
|
| 2782 |
+
"loss": 38.7854,
|
| 2783 |
"step": 394
|
| 2784 |
},
|
| 2785 |
{
|
| 2786 |
+
"epoch": 1.0432485968966656,
|
| 2787 |
+
"grad_norm": 6160.6787109375,
|
| 2788 |
"learning_rate": 9.993925925187865e-05,
|
| 2789 |
+
"loss": 35.9352,
|
| 2790 |
"step": 395
|
| 2791 |
},
|
| 2792 |
{
|
| 2793 |
+
"epoch": 1.0458897325850116,
|
| 2794 |
+
"grad_norm": 6419.31103515625,
|
| 2795 |
"learning_rate": 9.993884683633354e-05,
|
| 2796 |
+
"loss": 37.7825,
|
| 2797 |
"step": 396
|
| 2798 |
},
|
| 2799 |
{
|
| 2800 |
+
"epoch": 1.0485308682733576,
|
| 2801 |
+
"grad_norm": 8226.6005859375,
|
| 2802 |
"learning_rate": 9.993843302627549e-05,
|
| 2803 |
+
"loss": 35.5052,
|
| 2804 |
"step": 397
|
| 2805 |
},
|
| 2806 |
{
|
| 2807 |
+
"epoch": 1.0511720039617036,
|
| 2808 |
+
"grad_norm": 8380.81640625,
|
| 2809 |
"learning_rate": 9.993801782171603e-05,
|
| 2810 |
+
"loss": 36.5649,
|
| 2811 |
"step": 398
|
| 2812 |
},
|
| 2813 |
{
|
| 2814 |
+
"epoch": 1.0538131396500496,
|
| 2815 |
+
"grad_norm": 10895.78515625,
|
| 2816 |
"learning_rate": 9.993760122266676e-05,
|
| 2817 |
+
"loss": 37.1919,
|
| 2818 |
"step": 399
|
| 2819 |
},
|
| 2820 |
{
|
| 2821 |
+
"epoch": 1.0564542753383954,
|
| 2822 |
+
"grad_norm": 14454.5390625,
|
| 2823 |
"learning_rate": 9.99371832291393e-05,
|
| 2824 |
+
"loss": 38.3564,
|
| 2825 |
"step": 400
|
| 2826 |
},
|
| 2827 |
{
|
| 2828 |
+
"epoch": 1.0564542753383954,
|
| 2829 |
+
"eval_loss": 8.69857406616211,
|
| 2830 |
+
"eval_runtime": 2.1301,
|
| 2831 |
+
"eval_samples_per_second": 232.388,
|
| 2832 |
+
"eval_steps_per_second": 29.107,
|
| 2833 |
"step": 400
|
| 2834 |
}
|
| 2835 |
],
|
|
|
|
| 2850 |
"attributes": {}
|
| 2851 |
}
|
| 2852 |
},
|
| 2853 |
+
"total_flos": 1043682507620352.0,
|
| 2854 |
+
"train_batch_size": 8,
|
| 2855 |
"trial_name": null,
|
| 2856 |
"trial_params": null
|
| 2857 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c25f947011cdcfeb2e2cd6b6797fb101f0c43353904140320fa5ceb0b490467
|
| 3 |
+
size 6712
|