Training in progress, step 338, checkpoint
Browse files
checkpoint-338/adapter_config.json
CHANGED
|
@@ -33,13 +33,13 @@
|
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
| 36 |
-
"gate_proj",
|
| 37 |
-
"o_proj",
|
| 38 |
"q_proj",
|
| 39 |
-
"
|
|
|
|
| 40 |
"k_proj",
|
|
|
|
| 41 |
"up_proj",
|
| 42 |
-
"
|
| 43 |
],
|
| 44 |
"target_parameters": null,
|
| 45 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
|
|
|
|
|
|
| 36 |
"q_proj",
|
| 37 |
+
"down_proj",
|
| 38 |
+
"gate_proj",
|
| 39 |
"k_proj",
|
| 40 |
+
"o_proj",
|
| 41 |
"up_proj",
|
| 42 |
+
"v_proj"
|
| 43 |
],
|
| 44 |
"target_parameters": null,
|
| 45 |
"task_type": "CAUSAL_LM",
|
checkpoint-338/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 550593184
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:863cd485a07ec47beacb1d9c7a66d147bcef59ce4952b437e957c18fbd6aa8b9
|
| 3 |
size 550593184
|
checkpoint-338/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 280349285
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88bbc45fd5e1b603798197831427656851a891b7d9fc81e33d1cbabb829d972d
|
| 3 |
size 280349285
|
checkpoint-338/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6188fa82cfe12362753ceda20c0780cd93b0d548a6bce0d48ce1507f84a84360
|
| 3 |
size 1465
|
checkpoint-338/trainer_state.json
CHANGED
|
@@ -11,2416 +11,2416 @@
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.002962962962962963,
|
| 14 |
-
"grad_norm": 11.
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
"loss": 1.4603,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.005925925925925926,
|
| 21 |
-
"grad_norm": 12.
|
| 22 |
-
"learning_rate":
|
| 23 |
"loss": 1.662,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.008888888888888889,
|
| 28 |
-
"grad_norm": 10.
|
| 29 |
-
"learning_rate": 1.
|
| 30 |
-
"loss": 1.
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.011851851851851851,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 1.
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.014814814814814815,
|
| 42 |
-
"grad_norm": 5.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.017777777777777778,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 0.
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.02074074074074074,
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 0.
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.023703703703703703,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 0.
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.02666666666666667,
|
| 70 |
-
"grad_norm": 1.
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 0.
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.02962962962962963,
|
| 77 |
-
"grad_norm": 1.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 0.
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.03259259259259259,
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.035555555555555556,
|
| 91 |
-
"grad_norm":
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.03851851851851852,
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.04148148148148148,
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.044444444444444446,
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
-
"learning_rate":
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.047407407407407405,
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.05037037037037037,
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.05333333333333334,
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.056296296296296296,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.05925925925925926,
|
| 147 |
-
"grad_norm": 1.
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.06222222222222222,
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.06518518518518518,
|
| 161 |
-
"grad_norm": 1.
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.06814814814814815,
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.07111111111111111,
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.07407407407407407,
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss": 0.
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.07703703703703704,
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.08,
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.08296296296296296,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.08592592592592592,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.08888888888888889,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.09185185185185185,
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss": 0.
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.09481481481481481,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.09777777777777778,
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 0.
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.10074074074074074,
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
-
"learning_rate":
|
| 247 |
"loss": 0.1996,
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.1037037037037037,
|
| 252 |
-
"grad_norm": 0.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 0.
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.10666666666666667,
|
| 259 |
-
"grad_norm": 0.
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 0.
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.10962962962962963,
|
| 266 |
-
"grad_norm": 0.
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 0.
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.11259259259259259,
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.11555555555555555,
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 0.
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.11851851851851852,
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss": 0.
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.12148148148148148,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.12444444444444444,
|
| 301 |
-
"grad_norm":
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss": 0.
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.1274074074074074,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss": 0.
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.13037037037037036,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss": 0.
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.13333333333333333,
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss": 0.
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.1362962962962963,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 0.
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.13925925925925925,
|
| 336 |
-
"grad_norm": 1.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.14222222222222222,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 0.
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.1451851851851852,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 0.
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.14814814814814814,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.14814814814814814,
|
| 364 |
-
"eval_loss": 0.
|
| 365 |
-
"eval_runtime": 14.
|
| 366 |
-
"eval_samples_per_second": 41.
|
| 367 |
-
"eval_steps_per_second": 5.
|
| 368 |
"step": 50
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.1511111111111111,
|
| 372 |
-
"grad_norm": 0.
|
| 373 |
-
"learning_rate":
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 51
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 0.15407407407407409,
|
| 379 |
-
"grad_norm": 0.
|
| 380 |
-
"learning_rate":
|
| 381 |
-
"loss": 0.
|
| 382 |
"step": 52
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 0.15703703703703703,
|
| 386 |
-
"grad_norm": 0.
|
| 387 |
-
"learning_rate":
|
| 388 |
-
"loss": 0.
|
| 389 |
"step": 53
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 0.16,
|
| 393 |
-
"grad_norm": 0.
|
| 394 |
-
"learning_rate":
|
| 395 |
-
"loss": 0.
|
| 396 |
"step": 54
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.16296296296296298,
|
| 400 |
-
"grad_norm": 0.
|
| 401 |
-
"learning_rate":
|
| 402 |
-
"loss": 0.
|
| 403 |
"step": 55
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 0.16592592592592592,
|
| 407 |
-
"grad_norm": 0.
|
| 408 |
-
"learning_rate":
|
| 409 |
-
"loss": 0.
|
| 410 |
"step": 56
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.1688888888888889,
|
| 414 |
-
"grad_norm": 0.
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 57
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.17185185185185184,
|
| 421 |
-
"grad_norm": 0.
|
| 422 |
-
"learning_rate":
|
| 423 |
-
"loss": 0.
|
| 424 |
"step": 58
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 0.1748148148148148,
|
| 428 |
-
"grad_norm": 0.
|
| 429 |
-
"learning_rate":
|
| 430 |
-
"loss": 0.
|
| 431 |
"step": 59
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 0.17777777777777778,
|
| 435 |
-
"grad_norm": 0.
|
| 436 |
-
"learning_rate":
|
| 437 |
-
"loss": 0.
|
| 438 |
"step": 60
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 0.18074074074074073,
|
| 442 |
-
"grad_norm": 0.
|
| 443 |
-
"learning_rate":
|
| 444 |
-
"loss": 0.
|
| 445 |
"step": 61
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 0.1837037037037037,
|
| 449 |
-
"grad_norm": 0.
|
| 450 |
-
"learning_rate":
|
| 451 |
-
"loss": 0.
|
| 452 |
"step": 62
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.18666666666666668,
|
| 456 |
-
"grad_norm": 0.
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 63
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.18962962962962962,
|
| 463 |
-
"grad_norm": 0.
|
| 464 |
-
"learning_rate":
|
| 465 |
-
"loss": 0.
|
| 466 |
"step": 64
|
| 467 |
},
|
| 468 |
{
|
| 469 |
"epoch": 0.1925925925925926,
|
| 470 |
-
"grad_norm": 0.
|
| 471 |
-
"learning_rate":
|
| 472 |
-
"loss": 0.
|
| 473 |
"step": 65
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 0.19555555555555557,
|
| 477 |
-
"grad_norm": 0.
|
| 478 |
-
"learning_rate":
|
| 479 |
-
"loss": 0.
|
| 480 |
"step": 66
|
| 481 |
},
|
| 482 |
{
|
| 483 |
"epoch": 0.1985185185185185,
|
| 484 |
-
"grad_norm": 0.
|
| 485 |
-
"learning_rate":
|
| 486 |
-
"loss": 0.
|
| 487 |
"step": 67
|
| 488 |
},
|
| 489 |
{
|
| 490 |
"epoch": 0.20148148148148148,
|
| 491 |
-
"grad_norm": 0.
|
| 492 |
-
"learning_rate":
|
| 493 |
-
"loss": 0.
|
| 494 |
"step": 68
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.20444444444444446,
|
| 498 |
-
"grad_norm": 0.
|
| 499 |
-
"learning_rate":
|
| 500 |
-
"loss": 0.
|
| 501 |
"step": 69
|
| 502 |
},
|
| 503 |
{
|
| 504 |
"epoch": 0.2074074074074074,
|
| 505 |
-
"grad_norm": 0.
|
| 506 |
-
"learning_rate":
|
| 507 |
-
"loss": 0.
|
| 508 |
"step": 70
|
| 509 |
},
|
| 510 |
{
|
| 511 |
"epoch": 0.21037037037037037,
|
| 512 |
-
"grad_norm": 0.
|
| 513 |
-
"learning_rate":
|
| 514 |
-
"loss": 0.
|
| 515 |
"step": 71
|
| 516 |
},
|
| 517 |
{
|
| 518 |
"epoch": 0.21333333333333335,
|
| 519 |
-
"grad_norm": 0.
|
| 520 |
-
"learning_rate":
|
| 521 |
-
"loss": 0.
|
| 522 |
"step": 72
|
| 523 |
},
|
| 524 |
{
|
| 525 |
"epoch": 0.2162962962962963,
|
| 526 |
-
"grad_norm": 0.
|
| 527 |
-
"learning_rate":
|
| 528 |
-
"loss": 0.
|
| 529 |
"step": 73
|
| 530 |
},
|
| 531 |
{
|
| 532 |
"epoch": 0.21925925925925926,
|
| 533 |
-
"grad_norm": 0.
|
| 534 |
-
"learning_rate":
|
| 535 |
-
"loss": 0.
|
| 536 |
"step": 74
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.2222222222222222,
|
| 540 |
-
"grad_norm": 0.
|
| 541 |
-
"learning_rate":
|
| 542 |
-
"loss": 0.
|
| 543 |
"step": 75
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 0.22518518518518518,
|
| 547 |
-
"grad_norm": 0.
|
| 548 |
-
"learning_rate":
|
| 549 |
-
"loss": 0.
|
| 550 |
"step": 76
|
| 551 |
},
|
| 552 |
{
|
| 553 |
"epoch": 0.22814814814814816,
|
| 554 |
-
"grad_norm": 0.
|
| 555 |
-
"learning_rate":
|
| 556 |
-
"loss": 0.
|
| 557 |
"step": 77
|
| 558 |
},
|
| 559 |
{
|
| 560 |
"epoch": 0.2311111111111111,
|
| 561 |
-
"grad_norm": 0.
|
| 562 |
-
"learning_rate":
|
| 563 |
-
"loss": 0.
|
| 564 |
"step": 78
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 0.23407407407407407,
|
| 568 |
-
"grad_norm": 0.
|
| 569 |
-
"learning_rate":
|
| 570 |
-
"loss": 0.
|
| 571 |
"step": 79
|
| 572 |
},
|
| 573 |
{
|
| 574 |
"epoch": 0.23703703703703705,
|
| 575 |
-
"grad_norm": 0.
|
| 576 |
-
"learning_rate":
|
| 577 |
-
"loss": 0.
|
| 578 |
"step": 80
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 0.24,
|
| 582 |
-
"grad_norm": 0.
|
| 583 |
-
"learning_rate":
|
| 584 |
-
"loss": 0.
|
| 585 |
"step": 81
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.24296296296296296,
|
| 589 |
-
"grad_norm": 0.
|
| 590 |
-
"learning_rate":
|
| 591 |
-
"loss": 0.
|
| 592 |
"step": 82
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 0.24592592592592594,
|
| 596 |
-
"grad_norm": 0.
|
| 597 |
-
"learning_rate":
|
| 598 |
-
"loss": 0.
|
| 599 |
"step": 83
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 0.24888888888888888,
|
| 603 |
-
"grad_norm": 0.
|
| 604 |
-
"learning_rate":
|
| 605 |
-
"loss": 0.
|
| 606 |
"step": 84
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 0.2518518518518518,
|
| 610 |
-
"grad_norm": 0.
|
| 611 |
-
"learning_rate":
|
| 612 |
-
"loss": 0.
|
| 613 |
"step": 85
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 0.2548148148148148,
|
| 617 |
-
"grad_norm": 0.
|
| 618 |
-
"learning_rate":
|
| 619 |
-
"loss": 0.
|
| 620 |
"step": 86
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 0.2577777777777778,
|
| 624 |
-
"grad_norm":
|
| 625 |
-
"learning_rate":
|
| 626 |
-
"loss": 0.
|
| 627 |
"step": 87
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 0.2607407407407407,
|
| 631 |
-
"grad_norm": 0.
|
| 632 |
-
"learning_rate":
|
| 633 |
-
"loss": 0.
|
| 634 |
"step": 88
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 0.2637037037037037,
|
| 638 |
-
"grad_norm": 0.
|
| 639 |
-
"learning_rate":
|
| 640 |
-
"loss": 0.
|
| 641 |
"step": 89
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 0.26666666666666666,
|
| 645 |
-
"grad_norm": 0.
|
| 646 |
-
"learning_rate": 2.
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 90
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 0.2696296296296296,
|
| 652 |
-
"grad_norm": 0.
|
| 653 |
-
"learning_rate": 2.
|
| 654 |
-
"loss": 0.
|
| 655 |
"step": 91
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 0.2725925925925926,
|
| 659 |
-
"grad_norm": 0.
|
| 660 |
-
"learning_rate": 2.
|
| 661 |
-
"loss": 0.
|
| 662 |
"step": 92
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 0.27555555555555555,
|
| 666 |
-
"grad_norm": 0.
|
| 667 |
-
"learning_rate": 2.
|
| 668 |
-
"loss": 0.
|
| 669 |
"step": 93
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 0.2785185185185185,
|
| 673 |
-
"grad_norm": 0.
|
| 674 |
-
"learning_rate": 2.
|
| 675 |
-
"loss": 0.
|
| 676 |
"step": 94
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 0.2814814814814815,
|
| 680 |
-
"grad_norm": 0.
|
| 681 |
-
"learning_rate": 2.
|
| 682 |
-
"loss": 0.
|
| 683 |
"step": 95
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 0.28444444444444444,
|
| 687 |
-
"grad_norm": 0.
|
| 688 |
-
"learning_rate": 2.
|
| 689 |
-
"loss": 0.
|
| 690 |
"step": 96
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 0.2874074074074074,
|
| 694 |
-
"grad_norm": 0.
|
| 695 |
-
"learning_rate": 2.
|
| 696 |
-
"loss": 0.
|
| 697 |
"step": 97
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.2903703703703704,
|
| 701 |
-
"grad_norm": 0.
|
| 702 |
-
"learning_rate": 2.
|
| 703 |
-
"loss": 0.
|
| 704 |
"step": 98
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 0.29333333333333333,
|
| 708 |
-
"grad_norm": 0.
|
| 709 |
-
"learning_rate": 2.
|
| 710 |
-
"loss": 0.
|
| 711 |
"step": 99
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 0.2962962962962963,
|
| 715 |
-
"grad_norm": 0.
|
| 716 |
-
"learning_rate": 2.
|
| 717 |
-
"loss": 0.
|
| 718 |
"step": 100
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 0.2962962962962963,
|
| 722 |
-
"eval_loss": 0.
|
| 723 |
-
"eval_runtime": 14.
|
| 724 |
-
"eval_samples_per_second": 42.
|
| 725 |
-
"eval_steps_per_second": 5.
|
| 726 |
"step": 100
|
| 727 |
},
|
| 728 |
{
|
| 729 |
"epoch": 0.2992592592592593,
|
| 730 |
-
"grad_norm": 0.
|
| 731 |
-
"learning_rate": 2.
|
| 732 |
-
"loss": 0.
|
| 733 |
"step": 101
|
| 734 |
},
|
| 735 |
{
|
| 736 |
"epoch": 0.3022222222222222,
|
| 737 |
-
"grad_norm": 0.
|
| 738 |
-
"learning_rate": 2.
|
| 739 |
-
"loss": 0.
|
| 740 |
"step": 102
|
| 741 |
},
|
| 742 |
{
|
| 743 |
"epoch": 0.30518518518518517,
|
| 744 |
-
"grad_norm": 0.
|
| 745 |
-
"learning_rate": 2.
|
| 746 |
-
"loss": 0.
|
| 747 |
"step": 103
|
| 748 |
},
|
| 749 |
{
|
| 750 |
"epoch": 0.30814814814814817,
|
| 751 |
-
"grad_norm": 0.
|
| 752 |
-
"learning_rate": 2.
|
| 753 |
-
"loss": 0.
|
| 754 |
"step": 104
|
| 755 |
},
|
| 756 |
{
|
| 757 |
"epoch": 0.3111111111111111,
|
| 758 |
-
"grad_norm": 0.
|
| 759 |
-
"learning_rate": 2.
|
| 760 |
-
"loss": 0.
|
| 761 |
"step": 105
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"epoch": 0.31407407407407406,
|
| 765 |
-
"grad_norm": 0.
|
| 766 |
-
"learning_rate": 2.
|
| 767 |
-
"loss": 0.
|
| 768 |
"step": 106
|
| 769 |
},
|
| 770 |
{
|
| 771 |
"epoch": 0.31703703703703706,
|
| 772 |
-
"grad_norm": 0.
|
| 773 |
-
"learning_rate": 2.
|
| 774 |
-
"loss": 0.
|
| 775 |
"step": 107
|
| 776 |
},
|
| 777 |
{
|
| 778 |
"epoch": 0.32,
|
| 779 |
-
"grad_norm": 0.
|
| 780 |
-
"learning_rate": 2.
|
| 781 |
-
"loss": 0.
|
| 782 |
"step": 108
|
| 783 |
},
|
| 784 |
{
|
| 785 |
"epoch": 0.32296296296296295,
|
| 786 |
-
"grad_norm": 0.
|
| 787 |
-
"learning_rate": 2.
|
| 788 |
-
"loss": 0.
|
| 789 |
"step": 109
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 0.32592592592592595,
|
| 793 |
-
"grad_norm": 0.
|
| 794 |
-
"learning_rate": 2.
|
| 795 |
-
"loss": 0.
|
| 796 |
"step": 110
|
| 797 |
},
|
| 798 |
{
|
| 799 |
"epoch": 0.3288888888888889,
|
| 800 |
-
"grad_norm": 0.
|
| 801 |
-
"learning_rate": 2.
|
| 802 |
-
"loss": 0.
|
| 803 |
"step": 111
|
| 804 |
},
|
| 805 |
{
|
| 806 |
"epoch": 0.33185185185185184,
|
| 807 |
-
"grad_norm": 0.
|
| 808 |
-
"learning_rate": 2.
|
| 809 |
-
"loss": 0.
|
| 810 |
"step": 112
|
| 811 |
},
|
| 812 |
{
|
| 813 |
"epoch": 0.3348148148148148,
|
| 814 |
-
"grad_norm": 0.
|
| 815 |
-
"learning_rate": 2.
|
| 816 |
-
"loss": 0.
|
| 817 |
"step": 113
|
| 818 |
},
|
| 819 |
{
|
| 820 |
"epoch": 0.3377777777777778,
|
| 821 |
-
"grad_norm": 0.
|
| 822 |
-
"learning_rate": 2.
|
| 823 |
-
"loss": 0.
|
| 824 |
"step": 114
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 0.34074074074074073,
|
| 828 |
-
"grad_norm": 0.
|
| 829 |
-
"learning_rate": 2.
|
| 830 |
"loss": 0.1768,
|
| 831 |
"step": 115
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 0.3437037037037037,
|
| 835 |
-
"grad_norm": 0.
|
| 836 |
-
"learning_rate": 2.
|
| 837 |
-
"loss": 0.
|
| 838 |
"step": 116
|
| 839 |
},
|
| 840 |
{
|
| 841 |
"epoch": 0.3466666666666667,
|
| 842 |
-
"grad_norm": 0.
|
| 843 |
-
"learning_rate":
|
| 844 |
-
"loss": 0.
|
| 845 |
"step": 117
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 0.3496296296296296,
|
| 849 |
-
"grad_norm": 0.
|
| 850 |
-
"learning_rate":
|
| 851 |
-
"loss": 0.
|
| 852 |
"step": 118
|
| 853 |
},
|
| 854 |
{
|
| 855 |
"epoch": 0.35259259259259257,
|
| 856 |
-
"grad_norm": 0.
|
| 857 |
-
"learning_rate":
|
| 858 |
-
"loss": 0.
|
| 859 |
"step": 119
|
| 860 |
},
|
| 861 |
{
|
| 862 |
"epoch": 0.35555555555555557,
|
| 863 |
-
"grad_norm": 0.
|
| 864 |
-
"learning_rate":
|
| 865 |
-
"loss": 0.
|
| 866 |
"step": 120
|
| 867 |
},
|
| 868 |
{
|
| 869 |
"epoch": 0.3585185185185185,
|
| 870 |
-
"grad_norm": 0.
|
| 871 |
-
"learning_rate":
|
| 872 |
-
"loss": 0.
|
| 873 |
"step": 121
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.36148148148148146,
|
| 877 |
-
"grad_norm": 0.
|
| 878 |
-
"learning_rate":
|
| 879 |
-
"loss": 0.
|
| 880 |
"step": 122
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 0.36444444444444446,
|
| 884 |
-
"grad_norm": 0.
|
| 885 |
-
"learning_rate":
|
| 886 |
-
"loss": 0.
|
| 887 |
"step": 123
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 0.3674074074074074,
|
| 891 |
-
"grad_norm": 0.
|
| 892 |
-
"learning_rate":
|
| 893 |
-
"loss": 0.
|
| 894 |
"step": 124
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 0.37037037037037035,
|
| 898 |
-
"grad_norm": 0.
|
| 899 |
-
"learning_rate":
|
| 900 |
-
"loss": 0.
|
| 901 |
"step": 125
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 0.37333333333333335,
|
| 905 |
-
"grad_norm": 0.
|
| 906 |
-
"learning_rate":
|
| 907 |
-
"loss": 0.
|
| 908 |
"step": 126
|
| 909 |
},
|
| 910 |
{
|
| 911 |
"epoch": 0.3762962962962963,
|
| 912 |
-
"grad_norm": 0.
|
| 913 |
-
"learning_rate":
|
| 914 |
-
"loss": 0.
|
| 915 |
"step": 127
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 0.37925925925925924,
|
| 919 |
-
"grad_norm": 0.
|
| 920 |
-
"learning_rate":
|
| 921 |
-
"loss": 0.
|
| 922 |
"step": 128
|
| 923 |
},
|
| 924 |
{
|
| 925 |
"epoch": 0.38222222222222224,
|
| 926 |
-
"grad_norm": 0.
|
| 927 |
-
"learning_rate":
|
| 928 |
-
"loss": 0.
|
| 929 |
"step": 129
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 0.3851851851851852,
|
| 933 |
-
"grad_norm": 0.
|
| 934 |
-
"learning_rate":
|
| 935 |
-
"loss": 0.
|
| 936 |
"step": 130
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 0.38814814814814813,
|
| 940 |
-
"grad_norm": 0.
|
| 941 |
-
"learning_rate":
|
| 942 |
-
"loss": 0.
|
| 943 |
"step": 131
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 0.39111111111111113,
|
| 947 |
-
"grad_norm": 0.
|
| 948 |
-
"learning_rate":
|
| 949 |
-
"loss": 0.
|
| 950 |
"step": 132
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 0.3940740740740741,
|
| 954 |
-
"grad_norm": 0.
|
| 955 |
-
"learning_rate":
|
| 956 |
-
"loss": 0.
|
| 957 |
"step": 133
|
| 958 |
},
|
| 959 |
{
|
| 960 |
"epoch": 0.397037037037037,
|
| 961 |
-
"grad_norm": 0.
|
| 962 |
-
"learning_rate":
|
| 963 |
-
"loss": 0.
|
| 964 |
"step": 134
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 0.4,
|
| 968 |
-
"grad_norm": 0.
|
| 969 |
-
"learning_rate":
|
| 970 |
-
"loss": 0.
|
| 971 |
"step": 135
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 0.40296296296296297,
|
| 975 |
-
"grad_norm": 0.
|
| 976 |
-
"learning_rate":
|
| 977 |
-
"loss": 0.
|
| 978 |
"step": 136
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 0.4059259259259259,
|
| 982 |
-
"grad_norm": 0.
|
| 983 |
-
"learning_rate":
|
| 984 |
-
"loss": 0.
|
| 985 |
"step": 137
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 0.4088888888888889,
|
| 989 |
-
"grad_norm": 0.
|
| 990 |
-
"learning_rate":
|
| 991 |
-
"loss": 0.
|
| 992 |
"step": 138
|
| 993 |
},
|
| 994 |
{
|
| 995 |
"epoch": 0.41185185185185186,
|
| 996 |
-
"grad_norm": 0.
|
| 997 |
-
"learning_rate":
|
| 998 |
"loss": 0.1839,
|
| 999 |
"step": 139
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 0.4148148148148148,
|
| 1003 |
-
"grad_norm":
|
| 1004 |
-
"learning_rate":
|
| 1005 |
-
"loss": 0.
|
| 1006 |
"step": 140
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
"epoch": 0.4177777777777778,
|
| 1010 |
-
"grad_norm": 0.
|
| 1011 |
-
"learning_rate":
|
| 1012 |
-
"loss": 0.
|
| 1013 |
"step": 141
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
"epoch": 0.42074074074074075,
|
| 1017 |
-
"grad_norm": 0.
|
| 1018 |
-
"learning_rate":
|
| 1019 |
-
"loss": 0.
|
| 1020 |
"step": 142
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
"epoch": 0.4237037037037037,
|
| 1024 |
-
"grad_norm": 0.
|
| 1025 |
-
"learning_rate":
|
| 1026 |
-
"loss": 0.
|
| 1027 |
"step": 143
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
"epoch": 0.4266666666666667,
|
| 1031 |
-
"grad_norm": 0.
|
| 1032 |
-
"learning_rate":
|
| 1033 |
-
"loss": 0.
|
| 1034 |
"step": 144
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 0.42962962962962964,
|
| 1038 |
-
"grad_norm": 0.
|
| 1039 |
-
"learning_rate":
|
| 1040 |
-
"loss": 0.
|
| 1041 |
"step": 145
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 0.4325925925925926,
|
| 1045 |
-
"grad_norm": 0.
|
| 1046 |
-
"learning_rate":
|
| 1047 |
-
"loss": 0.
|
| 1048 |
"step": 146
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
"epoch": 0.43555555555555553,
|
| 1052 |
-
"grad_norm": 0.
|
| 1053 |
-
"learning_rate":
|
| 1054 |
-
"loss": 0.
|
| 1055 |
"step": 147
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
"epoch": 0.43851851851851853,
|
| 1059 |
-
"grad_norm": 0.
|
| 1060 |
-
"learning_rate":
|
| 1061 |
-
"loss": 0.
|
| 1062 |
"step": 148
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 0.4414814814814815,
|
| 1066 |
-
"grad_norm": 0.
|
| 1067 |
-
"learning_rate":
|
| 1068 |
-
"loss": 0.
|
| 1069 |
"step": 149
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
"epoch": 0.4444444444444444,
|
| 1073 |
-
"grad_norm": 0.
|
| 1074 |
-
"learning_rate":
|
| 1075 |
-
"loss": 0.
|
| 1076 |
"step": 150
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
"epoch": 0.4444444444444444,
|
| 1080 |
-
"eval_loss": 0.
|
| 1081 |
-
"eval_runtime": 14.
|
| 1082 |
-
"eval_samples_per_second": 42.
|
| 1083 |
-
"eval_steps_per_second": 5.
|
| 1084 |
"step": 150
|
| 1085 |
},
|
| 1086 |
{
|
| 1087 |
"epoch": 0.4474074074074074,
|
| 1088 |
-
"grad_norm": 0.
|
| 1089 |
-
"learning_rate":
|
| 1090 |
-
"loss": 0.
|
| 1091 |
"step": 151
|
| 1092 |
},
|
| 1093 |
{
|
| 1094 |
"epoch": 0.45037037037037037,
|
| 1095 |
-
"grad_norm": 0.
|
| 1096 |
-
"learning_rate":
|
| 1097 |
-
"loss": 0.
|
| 1098 |
"step": 152
|
| 1099 |
},
|
| 1100 |
{
|
| 1101 |
"epoch": 0.4533333333333333,
|
| 1102 |
-
"grad_norm": 0.
|
| 1103 |
-
"learning_rate":
|
| 1104 |
-
"loss": 0.
|
| 1105 |
"step": 153
|
| 1106 |
},
|
| 1107 |
{
|
| 1108 |
"epoch": 0.4562962962962963,
|
| 1109 |
-
"grad_norm": 0.
|
| 1110 |
-
"learning_rate":
|
| 1111 |
-
"loss": 0.
|
| 1112 |
"step": 154
|
| 1113 |
},
|
| 1114 |
{
|
| 1115 |
"epoch": 0.45925925925925926,
|
| 1116 |
-
"grad_norm": 0.
|
| 1117 |
-
"learning_rate":
|
| 1118 |
-
"loss": 0.
|
| 1119 |
"step": 155
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
"epoch": 0.4622222222222222,
|
| 1123 |
-
"grad_norm": 0.
|
| 1124 |
-
"learning_rate":
|
| 1125 |
-
"loss": 0.
|
| 1126 |
"step": 156
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
"epoch": 0.4651851851851852,
|
| 1130 |
-
"grad_norm": 0.
|
| 1131 |
-
"learning_rate":
|
| 1132 |
-
"loss": 0.
|
| 1133 |
"step": 157
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
"epoch": 0.46814814814814815,
|
| 1137 |
-
"grad_norm": 0.
|
| 1138 |
-
"learning_rate":
|
| 1139 |
-
"loss": 0.
|
| 1140 |
"step": 158
|
| 1141 |
},
|
| 1142 |
{
|
| 1143 |
"epoch": 0.4711111111111111,
|
| 1144 |
-
"grad_norm": 0.
|
| 1145 |
-
"learning_rate":
|
| 1146 |
-
"loss": 0.
|
| 1147 |
"step": 159
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
"epoch": 0.4740740740740741,
|
| 1151 |
-
"grad_norm": 0.
|
| 1152 |
-
"learning_rate":
|
| 1153 |
-
"loss": 0.
|
| 1154 |
"step": 160
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
"epoch": 0.47703703703703704,
|
| 1158 |
-
"grad_norm": 0.
|
| 1159 |
-
"learning_rate":
|
| 1160 |
-
"loss": 0.
|
| 1161 |
"step": 161
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.48,
|
| 1165 |
-
"grad_norm": 0.
|
| 1166 |
-
"learning_rate":
|
| 1167 |
-
"loss": 0.
|
| 1168 |
"step": 162
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 0.482962962962963,
|
| 1172 |
-
"grad_norm": 0.
|
| 1173 |
-
"learning_rate":
|
| 1174 |
-
"loss": 0.
|
| 1175 |
"step": 163
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 0.48592592592592593,
|
| 1179 |
-
"grad_norm": 0.
|
| 1180 |
-
"learning_rate":
|
| 1181 |
-
"loss": 0.
|
| 1182 |
"step": 164
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 0.4888888888888889,
|
| 1186 |
-
"grad_norm": 0.
|
| 1187 |
-
"learning_rate":
|
| 1188 |
-
"loss": 0.
|
| 1189 |
"step": 165
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 0.4918518518518519,
|
| 1193 |
-
"grad_norm": 0.
|
| 1194 |
-
"learning_rate":
|
| 1195 |
-
"loss": 0.
|
| 1196 |
"step": 166
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 0.4948148148148148,
|
| 1200 |
-
"grad_norm": 0.
|
| 1201 |
-
"learning_rate":
|
| 1202 |
-
"loss": 0.
|
| 1203 |
"step": 167
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 0.49777777777777776,
|
| 1207 |
-
"grad_norm": 0.
|
| 1208 |
-
"learning_rate":
|
| 1209 |
-
"loss": 0.
|
| 1210 |
"step": 168
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 0.5007407407407407,
|
| 1214 |
-
"grad_norm": 0.
|
| 1215 |
-
"learning_rate":
|
| 1216 |
-
"loss": 0.
|
| 1217 |
"step": 169
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 0.5037037037037037,
|
| 1221 |
-
"grad_norm": 0.
|
| 1222 |
-
"learning_rate":
|
| 1223 |
-
"loss": 0.
|
| 1224 |
"step": 170
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 0.5066666666666667,
|
| 1228 |
-
"grad_norm": 0.
|
| 1229 |
-
"learning_rate":
|
| 1230 |
-
"loss": 0.
|
| 1231 |
"step": 171
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 0.5096296296296297,
|
| 1235 |
-
"grad_norm": 0.
|
| 1236 |
-
"learning_rate":
|
| 1237 |
-
"loss": 0.
|
| 1238 |
"step": 172
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 0.5125925925925926,
|
| 1242 |
-
"grad_norm": 0.
|
| 1243 |
-
"learning_rate": 1.
|
| 1244 |
-
"loss": 0.
|
| 1245 |
"step": 173
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 0.5155555555555555,
|
| 1249 |
-
"grad_norm": 0.
|
| 1250 |
-
"learning_rate": 1.
|
| 1251 |
-
"loss": 0.
|
| 1252 |
"step": 174
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 0.5185185185185185,
|
| 1256 |
-
"grad_norm": 0.
|
| 1257 |
-
"learning_rate": 1.
|
| 1258 |
-
"loss": 0.
|
| 1259 |
"step": 175
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 0.5214814814814814,
|
| 1263 |
-
"grad_norm": 0.
|
| 1264 |
-
"learning_rate": 1.
|
| 1265 |
-
"loss": 0.
|
| 1266 |
"step": 176
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 0.5244444444444445,
|
| 1270 |
-
"grad_norm": 0.
|
| 1271 |
-
"learning_rate": 1.
|
| 1272 |
-
"loss": 0.
|
| 1273 |
"step": 177
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 0.5274074074074074,
|
| 1277 |
-
"grad_norm": 0.
|
| 1278 |
-
"learning_rate": 1.
|
| 1279 |
-
"loss": 0.
|
| 1280 |
"step": 178
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 0.5303703703703704,
|
| 1284 |
-
"grad_norm": 0.
|
| 1285 |
-
"learning_rate": 1.
|
| 1286 |
-
"loss": 0.
|
| 1287 |
"step": 179
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 0.5333333333333333,
|
| 1291 |
-
"grad_norm": 0.
|
| 1292 |
-
"learning_rate": 1.
|
| 1293 |
-
"loss": 0.
|
| 1294 |
"step": 180
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 0.5362962962962963,
|
| 1298 |
-
"grad_norm": 0.
|
| 1299 |
-
"learning_rate": 1.
|
| 1300 |
-
"loss": 0.
|
| 1301 |
"step": 181
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 0.5392592592592592,
|
| 1305 |
-
"grad_norm": 0.
|
| 1306 |
-
"learning_rate": 1.
|
| 1307 |
-
"loss": 0.
|
| 1308 |
"step": 182
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 0.5422222222222223,
|
| 1312 |
-
"grad_norm": 0.
|
| 1313 |
-
"learning_rate": 1.
|
| 1314 |
-
"loss": 0.
|
| 1315 |
"step": 183
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 0.5451851851851852,
|
| 1319 |
-
"grad_norm": 0.
|
| 1320 |
-
"learning_rate": 1.
|
| 1321 |
-
"loss": 0.
|
| 1322 |
"step": 184
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 0.5481481481481482,
|
| 1326 |
-
"grad_norm": 0.
|
| 1327 |
-
"learning_rate": 1.
|
| 1328 |
-
"loss": 0.
|
| 1329 |
"step": 185
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 0.5511111111111111,
|
| 1333 |
-
"grad_norm": 0.
|
| 1334 |
-
"learning_rate": 1.
|
| 1335 |
-
"loss": 0.
|
| 1336 |
"step": 186
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 0.554074074074074,
|
| 1340 |
-
"grad_norm": 0.
|
| 1341 |
-
"learning_rate": 1.
|
| 1342 |
-
"loss": 0.
|
| 1343 |
"step": 187
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 0.557037037037037,
|
| 1347 |
-
"grad_norm": 0.
|
| 1348 |
-
"learning_rate": 1.
|
| 1349 |
-
"loss": 0.
|
| 1350 |
"step": 188
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 0.56,
|
| 1354 |
-
"grad_norm": 0.
|
| 1355 |
-
"learning_rate": 1.
|
| 1356 |
-
"loss": 0.
|
| 1357 |
"step": 189
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 0.562962962962963,
|
| 1361 |
-
"grad_norm": 0.
|
| 1362 |
-
"learning_rate": 1.
|
| 1363 |
-
"loss": 0.
|
| 1364 |
"step": 190
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 0.5659259259259259,
|
| 1368 |
-
"grad_norm": 0.
|
| 1369 |
-
"learning_rate": 1.
|
| 1370 |
-
"loss": 0.
|
| 1371 |
"step": 191
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 0.5688888888888889,
|
| 1375 |
-
"grad_norm": 0.
|
| 1376 |
-
"learning_rate": 1.
|
| 1377 |
-
"loss": 0.
|
| 1378 |
"step": 192
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 0.5718518518518518,
|
| 1382 |
-
"grad_norm": 0.
|
| 1383 |
-
"learning_rate": 1.
|
| 1384 |
-
"loss": 0.
|
| 1385 |
"step": 193
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 0.5748148148148148,
|
| 1389 |
-
"grad_norm": 0.
|
| 1390 |
-
"learning_rate": 1.
|
| 1391 |
-
"loss": 0.
|
| 1392 |
"step": 194
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 0.5777777777777777,
|
| 1396 |
-
"grad_norm": 0.
|
| 1397 |
-
"learning_rate": 1.
|
| 1398 |
-
"loss": 0.
|
| 1399 |
"step": 195
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 0.5807407407407408,
|
| 1403 |
-
"grad_norm": 0.
|
| 1404 |
-
"learning_rate": 1.
|
| 1405 |
-
"loss": 0.
|
| 1406 |
"step": 196
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 0.5837037037037037,
|
| 1410 |
-
"grad_norm": 0.
|
| 1411 |
-
"learning_rate": 1.
|
| 1412 |
-
"loss": 0.
|
| 1413 |
"step": 197
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 0.5866666666666667,
|
| 1417 |
-
"grad_norm": 0.
|
| 1418 |
-
"learning_rate": 1.
|
| 1419 |
-
"loss": 0.
|
| 1420 |
"step": 198
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 0.5896296296296296,
|
| 1424 |
-
"grad_norm": 0.
|
| 1425 |
-
"learning_rate": 1.
|
| 1426 |
-
"loss": 0.
|
| 1427 |
"step": 199
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 0.5925925925925926,
|
| 1431 |
-
"grad_norm": 0.
|
| 1432 |
-
"learning_rate": 1.
|
| 1433 |
-
"loss": 0.
|
| 1434 |
"step": 200
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 0.5925925925925926,
|
| 1438 |
-
"eval_loss": 0.
|
| 1439 |
-
"eval_runtime": 14.
|
| 1440 |
-
"eval_samples_per_second": 42.
|
| 1441 |
-
"eval_steps_per_second": 5.
|
| 1442 |
"step": 200
|
| 1443 |
},
|
| 1444 |
{
|
| 1445 |
"epoch": 0.5955555555555555,
|
| 1446 |
-
"grad_norm": 0.
|
| 1447 |
-
"learning_rate": 1.
|
| 1448 |
-
"loss": 0.
|
| 1449 |
"step": 201
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.5985185185185186,
|
| 1453 |
-
"grad_norm": 0.
|
| 1454 |
-
"learning_rate": 1.
|
| 1455 |
-
"loss": 0.
|
| 1456 |
"step": 202
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 0.6014814814814815,
|
| 1460 |
-
"grad_norm": 0.
|
| 1461 |
-
"learning_rate": 1.
|
| 1462 |
-
"loss": 0.
|
| 1463 |
"step": 203
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 0.6044444444444445,
|
| 1467 |
-
"grad_norm": 0.
|
| 1468 |
-
"learning_rate": 1.
|
| 1469 |
-
"loss": 0.
|
| 1470 |
"step": 204
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 0.6074074074074074,
|
| 1474 |
-
"grad_norm": 0.
|
| 1475 |
-
"learning_rate": 1.
|
| 1476 |
-
"loss": 0.
|
| 1477 |
"step": 205
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 0.6103703703703703,
|
| 1481 |
-
"grad_norm": 0.
|
| 1482 |
-
"learning_rate": 1.
|
| 1483 |
-
"loss": 0.
|
| 1484 |
"step": 206
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 0.6133333333333333,
|
| 1488 |
-
"grad_norm": 0.
|
| 1489 |
-
"learning_rate": 1.
|
| 1490 |
-
"loss": 0.
|
| 1491 |
"step": 207
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 0.6162962962962963,
|
| 1495 |
-
"grad_norm": 0.
|
| 1496 |
-
"learning_rate": 1.
|
| 1497 |
-
"loss": 0.
|
| 1498 |
"step": 208
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 0.6192592592592593,
|
| 1502 |
-
"grad_norm": 0.
|
| 1503 |
-
"learning_rate": 1.
|
| 1504 |
-
"loss": 0.
|
| 1505 |
"step": 209
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 0.6222222222222222,
|
| 1509 |
-
"grad_norm": 0.
|
| 1510 |
-
"learning_rate": 1.
|
| 1511 |
-
"loss": 0.
|
| 1512 |
"step": 210
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 0.6251851851851852,
|
| 1516 |
-
"grad_norm": 0.
|
| 1517 |
-
"learning_rate": 1.
|
| 1518 |
-
"loss": 0.
|
| 1519 |
"step": 211
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 0.6281481481481481,
|
| 1523 |
-
"grad_norm": 0.
|
| 1524 |
-
"learning_rate": 1.
|
| 1525 |
-
"loss": 0.
|
| 1526 |
"step": 212
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 0.6311111111111111,
|
| 1530 |
-
"grad_norm": 0.
|
| 1531 |
-
"learning_rate": 1.
|
| 1532 |
-
"loss": 0.
|
| 1533 |
"step": 213
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 0.6340740740740741,
|
| 1537 |
-
"grad_norm": 0.
|
| 1538 |
-
"learning_rate": 1.
|
| 1539 |
-
"loss": 0.
|
| 1540 |
"step": 214
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 0.6370370370370371,
|
| 1544 |
-
"grad_norm": 0.
|
| 1545 |
-
"learning_rate": 1.
|
| 1546 |
-
"loss": 0.
|
| 1547 |
"step": 215
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 0.64,
|
| 1551 |
-
"grad_norm": 0.
|
| 1552 |
-
"learning_rate": 1.
|
| 1553 |
-
"loss": 0.
|
| 1554 |
"step": 216
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 0.642962962962963,
|
| 1558 |
-
"grad_norm": 0.
|
| 1559 |
-
"learning_rate": 1.
|
| 1560 |
-
"loss": 0.
|
| 1561 |
"step": 217
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 0.6459259259259259,
|
| 1565 |
-
"grad_norm": 0.
|
| 1566 |
-
"learning_rate": 1.
|
| 1567 |
-
"loss": 0.
|
| 1568 |
"step": 218
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 0.6488888888888888,
|
| 1572 |
-
"grad_norm": 0.
|
| 1573 |
-
"learning_rate": 1.
|
| 1574 |
-
"loss": 0.
|
| 1575 |
"step": 219
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 0.6518518518518519,
|
| 1579 |
-
"grad_norm": 0.
|
| 1580 |
-
"learning_rate": 1.
|
| 1581 |
-
"loss": 0.
|
| 1582 |
"step": 220
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 0.6548148148148148,
|
| 1586 |
-
"grad_norm": 0.
|
| 1587 |
-
"learning_rate": 1.
|
| 1588 |
-
"loss": 0.
|
| 1589 |
"step": 221
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 0.6577777777777778,
|
| 1593 |
-
"grad_norm": 0.
|
| 1594 |
-
"learning_rate": 1.
|
| 1595 |
-
"loss": 0.
|
| 1596 |
"step": 222
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 0.6607407407407407,
|
| 1600 |
-
"grad_norm":
|
| 1601 |
-
"learning_rate": 1.
|
| 1602 |
-
"loss": 0.
|
| 1603 |
"step": 223
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 0.6637037037037037,
|
| 1607 |
-
"grad_norm": 0.
|
| 1608 |
-
"learning_rate": 1.
|
| 1609 |
-
"loss": 0.
|
| 1610 |
"step": 224
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 0.6666666666666666,
|
| 1614 |
-
"grad_norm": 0.
|
| 1615 |
-
"learning_rate": 1.
|
| 1616 |
-
"loss": 0.
|
| 1617 |
"step": 225
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 0.6696296296296296,
|
| 1621 |
-
"grad_norm": 0.
|
| 1622 |
-
"learning_rate": 1.
|
| 1623 |
-
"loss": 0.
|
| 1624 |
"step": 226
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 0.6725925925925926,
|
| 1628 |
-
"grad_norm": 0.
|
| 1629 |
-
"learning_rate": 1.
|
| 1630 |
-
"loss": 0.
|
| 1631 |
"step": 227
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 0.6755555555555556,
|
| 1635 |
-
"grad_norm": 0.
|
| 1636 |
-
"learning_rate":
|
| 1637 |
-
"loss": 0.
|
| 1638 |
"step": 228
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 0.6785185185185185,
|
| 1642 |
-
"grad_norm": 0.
|
| 1643 |
-
"learning_rate":
|
| 1644 |
-
"loss": 0.
|
| 1645 |
"step": 229
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
"epoch": 0.6814814814814815,
|
| 1649 |
-
"grad_norm": 0.
|
| 1650 |
-
"learning_rate":
|
| 1651 |
-
"loss": 0.
|
| 1652 |
"step": 230
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
"epoch": 0.6844444444444444,
|
| 1656 |
-
"grad_norm": 0.
|
| 1657 |
-
"learning_rate":
|
| 1658 |
-
"loss": 0.
|
| 1659 |
"step": 231
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
"epoch": 0.6874074074074074,
|
| 1663 |
-
"grad_norm": 0.
|
| 1664 |
-
"learning_rate":
|
| 1665 |
-
"loss": 0.
|
| 1666 |
"step": 232
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
"epoch": 0.6903703703703704,
|
| 1670 |
-
"grad_norm": 0.
|
| 1671 |
-
"learning_rate":
|
| 1672 |
-
"loss": 0.
|
| 1673 |
"step": 233
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
"epoch": 0.6933333333333334,
|
| 1677 |
-
"grad_norm": 0.
|
| 1678 |
-
"learning_rate":
|
| 1679 |
-
"loss": 0.
|
| 1680 |
"step": 234
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
"epoch": 0.6962962962962963,
|
| 1684 |
-
"grad_norm": 0.
|
| 1685 |
-
"learning_rate":
|
| 1686 |
-
"loss": 0.
|
| 1687 |
"step": 235
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
"epoch": 0.6992592592592592,
|
| 1691 |
-
"grad_norm": 0.
|
| 1692 |
-
"learning_rate":
|
| 1693 |
-
"loss": 0.
|
| 1694 |
"step": 236
|
| 1695 |
},
|
| 1696 |
{
|
| 1697 |
"epoch": 0.7022222222222222,
|
| 1698 |
-
"grad_norm": 0.
|
| 1699 |
-
"learning_rate":
|
| 1700 |
-
"loss": 0.
|
| 1701 |
"step": 237
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 0.7051851851851851,
|
| 1705 |
-
"grad_norm": 0.
|
| 1706 |
-
"learning_rate":
|
| 1707 |
-
"loss": 0.
|
| 1708 |
"step": 238
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
"epoch": 0.7081481481481482,
|
| 1712 |
-
"grad_norm": 0.
|
| 1713 |
-
"learning_rate":
|
| 1714 |
-
"loss": 0.
|
| 1715 |
"step": 239
|
| 1716 |
},
|
| 1717 |
{
|
| 1718 |
"epoch": 0.7111111111111111,
|
| 1719 |
-
"grad_norm": 0.
|
| 1720 |
-
"learning_rate":
|
| 1721 |
-
"loss": 0.
|
| 1722 |
"step": 240
|
| 1723 |
},
|
| 1724 |
{
|
| 1725 |
"epoch": 0.7140740740740741,
|
| 1726 |
-
"grad_norm": 0.
|
| 1727 |
-
"learning_rate":
|
| 1728 |
-
"loss": 0.
|
| 1729 |
"step": 241
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
"epoch": 0.717037037037037,
|
| 1733 |
-
"grad_norm": 0.
|
| 1734 |
-
"learning_rate":
|
| 1735 |
-
"loss": 0.
|
| 1736 |
"step": 242
|
| 1737 |
},
|
| 1738 |
{
|
| 1739 |
"epoch": 0.72,
|
| 1740 |
-
"grad_norm": 0.
|
| 1741 |
-
"learning_rate":
|
| 1742 |
-
"loss": 0.
|
| 1743 |
"step": 243
|
| 1744 |
},
|
| 1745 |
{
|
| 1746 |
"epoch": 0.7229629629629629,
|
| 1747 |
-
"grad_norm": 0.
|
| 1748 |
-
"learning_rate":
|
| 1749 |
-
"loss": 0.
|
| 1750 |
"step": 244
|
| 1751 |
},
|
| 1752 |
{
|
| 1753 |
"epoch": 0.725925925925926,
|
| 1754 |
-
"grad_norm": 0.
|
| 1755 |
-
"learning_rate":
|
| 1756 |
-
"loss": 0.
|
| 1757 |
"step": 245
|
| 1758 |
},
|
| 1759 |
{
|
| 1760 |
"epoch": 0.7288888888888889,
|
| 1761 |
-
"grad_norm": 0.
|
| 1762 |
-
"learning_rate":
|
| 1763 |
-
"loss": 0.
|
| 1764 |
"step": 246
|
| 1765 |
},
|
| 1766 |
{
|
| 1767 |
"epoch": 0.7318518518518519,
|
| 1768 |
-
"grad_norm": 0.
|
| 1769 |
-
"learning_rate":
|
| 1770 |
-
"loss": 0.
|
| 1771 |
"step": 247
|
| 1772 |
},
|
| 1773 |
{
|
| 1774 |
"epoch": 0.7348148148148148,
|
| 1775 |
-
"grad_norm": 0.
|
| 1776 |
-
"learning_rate":
|
| 1777 |
-
"loss": 0.
|
| 1778 |
"step": 248
|
| 1779 |
},
|
| 1780 |
{
|
| 1781 |
"epoch": 0.7377777777777778,
|
| 1782 |
-
"grad_norm": 0.
|
| 1783 |
-
"learning_rate":
|
| 1784 |
-
"loss": 0.
|
| 1785 |
"step": 249
|
| 1786 |
},
|
| 1787 |
{
|
| 1788 |
"epoch": 0.7407407407407407,
|
| 1789 |
-
"grad_norm": 0.
|
| 1790 |
-
"learning_rate":
|
| 1791 |
-
"loss": 0.
|
| 1792 |
"step": 250
|
| 1793 |
},
|
| 1794 |
{
|
| 1795 |
"epoch": 0.7407407407407407,
|
| 1796 |
-
"eval_loss": 0.
|
| 1797 |
-
"eval_runtime": 14.
|
| 1798 |
-
"eval_samples_per_second": 42.
|
| 1799 |
-
"eval_steps_per_second": 5.
|
| 1800 |
"step": 250
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 0.7437037037037038,
|
| 1804 |
-
"grad_norm": 0.
|
| 1805 |
-
"learning_rate":
|
| 1806 |
-
"loss": 0.
|
| 1807 |
"step": 251
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 0.7466666666666667,
|
| 1811 |
-
"grad_norm": 0.
|
| 1812 |
-
"learning_rate":
|
| 1813 |
-
"loss": 0.
|
| 1814 |
"step": 252
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 0.7496296296296296,
|
| 1818 |
-
"grad_norm": 0.
|
| 1819 |
-
"learning_rate":
|
| 1820 |
-
"loss": 0.
|
| 1821 |
"step": 253
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 0.7525925925925926,
|
| 1825 |
-
"grad_norm": 0.
|
| 1826 |
-
"learning_rate":
|
| 1827 |
"loss": 0.1413,
|
| 1828 |
"step": 254
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 0.7555555555555555,
|
| 1832 |
-
"grad_norm": 0.
|
| 1833 |
-
"learning_rate":
|
| 1834 |
-
"loss": 0.
|
| 1835 |
"step": 255
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 0.7585185185185185,
|
| 1839 |
-
"grad_norm": 0.
|
| 1840 |
-
"learning_rate":
|
| 1841 |
-
"loss": 0.
|
| 1842 |
"step": 256
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 0.7614814814814815,
|
| 1846 |
-
"grad_norm": 0.
|
| 1847 |
-
"learning_rate":
|
| 1848 |
-
"loss": 0.
|
| 1849 |
"step": 257
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 0.7644444444444445,
|
| 1853 |
-
"grad_norm": 0.
|
| 1854 |
-
"learning_rate":
|
| 1855 |
-
"loss": 0.
|
| 1856 |
"step": 258
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 0.7674074074074074,
|
| 1860 |
-
"grad_norm": 0.
|
| 1861 |
-
"learning_rate":
|
| 1862 |
-
"loss": 0.
|
| 1863 |
"step": 259
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 0.7703703703703704,
|
| 1867 |
-
"grad_norm": 0.
|
| 1868 |
-
"learning_rate":
|
| 1869 |
-
"loss": 0.
|
| 1870 |
"step": 260
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 0.7733333333333333,
|
| 1874 |
-
"grad_norm": 0.
|
| 1875 |
-
"learning_rate":
|
| 1876 |
-
"loss": 0.
|
| 1877 |
"step": 261
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 0.7762962962962963,
|
| 1881 |
-
"grad_norm": 0.
|
| 1882 |
-
"learning_rate":
|
| 1883 |
-
"loss": 0.
|
| 1884 |
"step": 262
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 0.7792592592592592,
|
| 1888 |
-
"grad_norm": 0.
|
| 1889 |
-
"learning_rate":
|
| 1890 |
-
"loss": 0.
|
| 1891 |
"step": 263
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 0.7822222222222223,
|
| 1895 |
-
"grad_norm": 0.
|
| 1896 |
-
"learning_rate":
|
| 1897 |
-
"loss": 0.
|
| 1898 |
"step": 264
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 0.7851851851851852,
|
| 1902 |
-
"grad_norm": 0.
|
| 1903 |
-
"learning_rate":
|
| 1904 |
-
"loss": 0.
|
| 1905 |
"step": 265
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 0.7881481481481482,
|
| 1909 |
-
"grad_norm": 0.
|
| 1910 |
-
"learning_rate":
|
| 1911 |
-
"loss": 0.
|
| 1912 |
"step": 266
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 0.7911111111111111,
|
| 1916 |
-
"grad_norm": 0.
|
| 1917 |
-
"learning_rate":
|
| 1918 |
-
"loss": 0.
|
| 1919 |
"step": 267
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 0.794074074074074,
|
| 1923 |
-
"grad_norm": 0.
|
| 1924 |
-
"learning_rate":
|
| 1925 |
-
"loss": 0.
|
| 1926 |
"step": 268
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 0.797037037037037,
|
| 1930 |
-
"grad_norm": 0.
|
| 1931 |
-
"learning_rate":
|
| 1932 |
-
"loss": 0.
|
| 1933 |
"step": 269
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 0.8,
|
| 1937 |
-
"grad_norm": 0.
|
| 1938 |
-
"learning_rate":
|
| 1939 |
-
"loss": 0.
|
| 1940 |
"step": 270
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 0.802962962962963,
|
| 1944 |
-
"grad_norm": 0.
|
| 1945 |
-
"learning_rate":
|
| 1946 |
-
"loss": 0.
|
| 1947 |
"step": 271
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 0.8059259259259259,
|
| 1951 |
-
"grad_norm": 0.
|
| 1952 |
-
"learning_rate":
|
| 1953 |
-
"loss": 0.
|
| 1954 |
"step": 272
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 0.8088888888888889,
|
| 1958 |
-
"grad_norm": 0.
|
| 1959 |
-
"learning_rate":
|
| 1960 |
-
"loss": 0.
|
| 1961 |
"step": 273
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 0.8118518518518518,
|
| 1965 |
-
"grad_norm": 0.
|
| 1966 |
-
"learning_rate":
|
| 1967 |
-
"loss": 0.
|
| 1968 |
"step": 274
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 0.8148148148148148,
|
| 1972 |
-
"grad_norm": 0.
|
| 1973 |
-
"learning_rate":
|
| 1974 |
-
"loss": 0.
|
| 1975 |
"step": 275
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 0.8177777777777778,
|
| 1979 |
-
"grad_norm": 0.
|
| 1980 |
-
"learning_rate":
|
| 1981 |
-
"loss": 0.
|
| 1982 |
"step": 276
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 0.8207407407407408,
|
| 1986 |
-
"grad_norm": 0.
|
| 1987 |
-
"learning_rate":
|
| 1988 |
-
"loss": 0.
|
| 1989 |
"step": 277
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 0.8237037037037037,
|
| 1993 |
-
"grad_norm": 0.
|
| 1994 |
-
"learning_rate":
|
| 1995 |
-
"loss": 0.
|
| 1996 |
"step": 278
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 0.8266666666666667,
|
| 2000 |
-
"grad_norm": 0.
|
| 2001 |
-
"learning_rate":
|
| 2002 |
-
"loss": 0.
|
| 2003 |
"step": 279
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 0.8296296296296296,
|
| 2007 |
-
"grad_norm": 0.
|
| 2008 |
-
"learning_rate":
|
| 2009 |
-
"loss": 0.
|
| 2010 |
"step": 280
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 0.8325925925925926,
|
| 2014 |
-
"grad_norm": 0.
|
| 2015 |
-
"learning_rate":
|
| 2016 |
-
"loss": 0.
|
| 2017 |
"step": 281
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 0.8355555555555556,
|
| 2021 |
-
"grad_norm": 0.
|
| 2022 |
-
"learning_rate":
|
| 2023 |
-
"loss": 0.
|
| 2024 |
"step": 282
|
| 2025 |
},
|
| 2026 |
{
|
| 2027 |
"epoch": 0.8385185185185186,
|
| 2028 |
-
"grad_norm": 0.
|
| 2029 |
-
"learning_rate":
|
| 2030 |
-
"loss": 0.
|
| 2031 |
"step": 283
|
| 2032 |
},
|
| 2033 |
{
|
| 2034 |
"epoch": 0.8414814814814815,
|
| 2035 |
-
"grad_norm": 0.
|
| 2036 |
-
"learning_rate":
|
| 2037 |
-
"loss": 0.
|
| 2038 |
"step": 284
|
| 2039 |
},
|
| 2040 |
{
|
| 2041 |
"epoch": 0.8444444444444444,
|
| 2042 |
-
"grad_norm": 0.
|
| 2043 |
-
"learning_rate":
|
| 2044 |
-
"loss": 0.
|
| 2045 |
"step": 285
|
| 2046 |
},
|
| 2047 |
{
|
| 2048 |
"epoch": 0.8474074074074074,
|
| 2049 |
-
"grad_norm": 0.
|
| 2050 |
-
"learning_rate":
|
| 2051 |
-
"loss": 0.
|
| 2052 |
"step": 286
|
| 2053 |
},
|
| 2054 |
{
|
| 2055 |
"epoch": 0.8503703703703703,
|
| 2056 |
-
"grad_norm": 0.
|
| 2057 |
-
"learning_rate":
|
| 2058 |
"loss": 0.1248,
|
| 2059 |
"step": 287
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
"epoch": 0.8533333333333334,
|
| 2063 |
-
"grad_norm": 0.
|
| 2064 |
-
"learning_rate":
|
| 2065 |
-
"loss": 0.
|
| 2066 |
"step": 288
|
| 2067 |
},
|
| 2068 |
{
|
| 2069 |
"epoch": 0.8562962962962963,
|
| 2070 |
-
"grad_norm": 0.
|
| 2071 |
-
"learning_rate":
|
| 2072 |
-
"loss": 0.
|
| 2073 |
"step": 289
|
| 2074 |
},
|
| 2075 |
{
|
| 2076 |
"epoch": 0.8592592592592593,
|
| 2077 |
-
"grad_norm": 0.
|
| 2078 |
-
"learning_rate":
|
| 2079 |
-
"loss": 0.
|
| 2080 |
"step": 290
|
| 2081 |
},
|
| 2082 |
{
|
| 2083 |
"epoch": 0.8622222222222222,
|
| 2084 |
-
"grad_norm": 0.
|
| 2085 |
-
"learning_rate":
|
| 2086 |
-
"loss": 0.
|
| 2087 |
"step": 291
|
| 2088 |
},
|
| 2089 |
{
|
| 2090 |
"epoch": 0.8651851851851852,
|
| 2091 |
-
"grad_norm": 0.
|
| 2092 |
-
"learning_rate":
|
| 2093 |
"loss": 0.1234,
|
| 2094 |
"step": 292
|
| 2095 |
},
|
| 2096 |
{
|
| 2097 |
"epoch": 0.8681481481481481,
|
| 2098 |
-
"grad_norm": 0.
|
| 2099 |
-
"learning_rate":
|
| 2100 |
-
"loss": 0.
|
| 2101 |
"step": 293
|
| 2102 |
},
|
| 2103 |
{
|
| 2104 |
"epoch": 0.8711111111111111,
|
| 2105 |
-
"grad_norm": 0.
|
| 2106 |
-
"learning_rate":
|
| 2107 |
-
"loss": 0.
|
| 2108 |
"step": 294
|
| 2109 |
},
|
| 2110 |
{
|
| 2111 |
"epoch": 0.8740740740740741,
|
| 2112 |
-
"grad_norm": 0.
|
| 2113 |
-
"learning_rate":
|
| 2114 |
-
"loss": 0.
|
| 2115 |
"step": 295
|
| 2116 |
},
|
| 2117 |
{
|
| 2118 |
"epoch": 0.8770370370370371,
|
| 2119 |
-
"grad_norm": 0.
|
| 2120 |
-
"learning_rate":
|
| 2121 |
-
"loss": 0.
|
| 2122 |
"step": 296
|
| 2123 |
},
|
| 2124 |
{
|
| 2125 |
"epoch": 0.88,
|
| 2126 |
-
"grad_norm": 0.
|
| 2127 |
-
"learning_rate":
|
| 2128 |
-
"loss": 0.
|
| 2129 |
"step": 297
|
| 2130 |
},
|
| 2131 |
{
|
| 2132 |
"epoch": 0.882962962962963,
|
| 2133 |
-
"grad_norm": 0.
|
| 2134 |
-
"learning_rate":
|
| 2135 |
-
"loss": 0.
|
| 2136 |
"step": 298
|
| 2137 |
},
|
| 2138 |
{
|
| 2139 |
"epoch": 0.8859259259259259,
|
| 2140 |
-
"grad_norm": 0.
|
| 2141 |
-
"learning_rate":
|
| 2142 |
-
"loss": 0.
|
| 2143 |
"step": 299
|
| 2144 |
},
|
| 2145 |
{
|
| 2146 |
"epoch": 0.8888888888888888,
|
| 2147 |
-
"grad_norm": 0.
|
| 2148 |
-
"learning_rate":
|
| 2149 |
-
"loss": 0.
|
| 2150 |
"step": 300
|
| 2151 |
},
|
| 2152 |
{
|
| 2153 |
"epoch": 0.8888888888888888,
|
| 2154 |
-
"eval_loss": 0.
|
| 2155 |
-
"eval_runtime": 14.
|
| 2156 |
-
"eval_samples_per_second": 42.
|
| 2157 |
-
"eval_steps_per_second": 5.
|
| 2158 |
"step": 300
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
"epoch": 0.8918518518518519,
|
| 2162 |
-
"grad_norm": 0.
|
| 2163 |
-
"learning_rate":
|
| 2164 |
-
"loss": 0.
|
| 2165 |
"step": 301
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
"epoch": 0.8948148148148148,
|
| 2169 |
-
"grad_norm": 0.
|
| 2170 |
-
"learning_rate":
|
| 2171 |
-
"loss": 0.
|
| 2172 |
"step": 302
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
"epoch": 0.8977777777777778,
|
| 2176 |
-
"grad_norm": 0.
|
| 2177 |
-
"learning_rate":
|
| 2178 |
-
"loss": 0.
|
| 2179 |
"step": 303
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
"epoch": 0.9007407407407407,
|
| 2183 |
-
"grad_norm": 0.
|
| 2184 |
-
"learning_rate":
|
| 2185 |
-
"loss": 0.
|
| 2186 |
"step": 304
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
"epoch": 0.9037037037037037,
|
| 2190 |
-
"grad_norm": 0.
|
| 2191 |
-
"learning_rate":
|
| 2192 |
-
"loss": 0.
|
| 2193 |
"step": 305
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
"epoch": 0.9066666666666666,
|
| 2197 |
-
"grad_norm": 0.
|
| 2198 |
-
"learning_rate":
|
| 2199 |
-
"loss": 0.
|
| 2200 |
"step": 306
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
"epoch": 0.9096296296296297,
|
| 2204 |
-
"grad_norm": 0.
|
| 2205 |
-
"learning_rate":
|
| 2206 |
-
"loss": 0.
|
| 2207 |
"step": 307
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
"epoch": 0.9125925925925926,
|
| 2211 |
-
"grad_norm": 0.
|
| 2212 |
-
"learning_rate":
|
| 2213 |
-
"loss": 0.
|
| 2214 |
"step": 308
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 0.9155555555555556,
|
| 2218 |
-
"grad_norm": 0.
|
| 2219 |
-
"learning_rate":
|
| 2220 |
-
"loss": 0.
|
| 2221 |
"step": 309
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
"epoch": 0.9185185185185185,
|
| 2225 |
-
"grad_norm": 0.
|
| 2226 |
-
"learning_rate":
|
| 2227 |
-
"loss": 0.
|
| 2228 |
"step": 310
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
"epoch": 0.9214814814814815,
|
| 2232 |
-
"grad_norm": 0.
|
| 2233 |
-
"learning_rate":
|
| 2234 |
-
"loss": 0.
|
| 2235 |
"step": 311
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
"epoch": 0.9244444444444444,
|
| 2239 |
-
"grad_norm": 0.
|
| 2240 |
-
"learning_rate":
|
| 2241 |
-
"loss": 0.
|
| 2242 |
"step": 312
|
| 2243 |
},
|
| 2244 |
{
|
| 2245 |
"epoch": 0.9274074074074075,
|
| 2246 |
-
"grad_norm": 0.
|
| 2247 |
-
"learning_rate":
|
| 2248 |
-
"loss": 0.
|
| 2249 |
"step": 313
|
| 2250 |
},
|
| 2251 |
{
|
| 2252 |
"epoch": 0.9303703703703704,
|
| 2253 |
-
"grad_norm": 0.
|
| 2254 |
-
"learning_rate":
|
| 2255 |
-
"loss": 0.
|
| 2256 |
"step": 314
|
| 2257 |
},
|
| 2258 |
{
|
| 2259 |
"epoch": 0.9333333333333333,
|
| 2260 |
-
"grad_norm": 0.
|
| 2261 |
-
"learning_rate": 2.
|
| 2262 |
-
"loss": 0.
|
| 2263 |
"step": 315
|
| 2264 |
},
|
| 2265 |
{
|
| 2266 |
"epoch": 0.9362962962962963,
|
| 2267 |
-
"grad_norm": 0.
|
| 2268 |
-
"learning_rate": 2.
|
| 2269 |
-
"loss": 0.
|
| 2270 |
"step": 316
|
| 2271 |
},
|
| 2272 |
{
|
| 2273 |
"epoch": 0.9392592592592592,
|
| 2274 |
-
"grad_norm": 0.
|
| 2275 |
-
"learning_rate":
|
| 2276 |
-
"loss": 0.
|
| 2277 |
"step": 317
|
| 2278 |
},
|
| 2279 |
{
|
| 2280 |
"epoch": 0.9422222222222222,
|
| 2281 |
-
"grad_norm": 0.
|
| 2282 |
-
"learning_rate":
|
| 2283 |
-
"loss": 0.
|
| 2284 |
"step": 318
|
| 2285 |
},
|
| 2286 |
{
|
| 2287 |
"epoch": 0.9451851851851852,
|
| 2288 |
-
"grad_norm": 0.
|
| 2289 |
-
"learning_rate":
|
| 2290 |
-
"loss": 0.
|
| 2291 |
"step": 319
|
| 2292 |
},
|
| 2293 |
{
|
| 2294 |
"epoch": 0.9481481481481482,
|
| 2295 |
-
"grad_norm": 0.
|
| 2296 |
-
"learning_rate":
|
| 2297 |
-
"loss": 0.
|
| 2298 |
"step": 320
|
| 2299 |
},
|
| 2300 |
{
|
| 2301 |
"epoch": 0.9511111111111111,
|
| 2302 |
-
"grad_norm": 0.
|
| 2303 |
-
"learning_rate":
|
| 2304 |
-
"loss": 0.
|
| 2305 |
"step": 321
|
| 2306 |
},
|
| 2307 |
{
|
| 2308 |
"epoch": 0.9540740740740741,
|
| 2309 |
-
"grad_norm": 0.
|
| 2310 |
-
"learning_rate":
|
| 2311 |
-
"loss": 0.
|
| 2312 |
"step": 322
|
| 2313 |
},
|
| 2314 |
{
|
| 2315 |
"epoch": 0.957037037037037,
|
| 2316 |
-
"grad_norm": 0.
|
| 2317 |
-
"learning_rate": 1.
|
| 2318 |
-
"loss": 0.
|
| 2319 |
"step": 323
|
| 2320 |
},
|
| 2321 |
{
|
| 2322 |
"epoch": 0.96,
|
| 2323 |
-
"grad_norm": 0.
|
| 2324 |
-
"learning_rate": 1.
|
| 2325 |
-
"loss": 0.
|
| 2326 |
"step": 324
|
| 2327 |
},
|
| 2328 |
{
|
| 2329 |
"epoch": 0.9629629629629629,
|
| 2330 |
-
"grad_norm": 0.
|
| 2331 |
-
"learning_rate": 1.
|
| 2332 |
-
"loss": 0.
|
| 2333 |
"step": 325
|
| 2334 |
},
|
| 2335 |
{
|
| 2336 |
"epoch": 0.965925925925926,
|
| 2337 |
-
"grad_norm": 0.
|
| 2338 |
-
"learning_rate": 1.
|
| 2339 |
-
"loss": 0.
|
| 2340 |
"step": 326
|
| 2341 |
},
|
| 2342 |
{
|
| 2343 |
"epoch": 0.9688888888888889,
|
| 2344 |
-
"grad_norm": 0.
|
| 2345 |
-
"learning_rate": 1.
|
| 2346 |
-
"loss": 0.
|
| 2347 |
"step": 327
|
| 2348 |
},
|
| 2349 |
{
|
| 2350 |
"epoch": 0.9718518518518519,
|
| 2351 |
-
"grad_norm": 0.
|
| 2352 |
-
"learning_rate":
|
| 2353 |
-
"loss": 0.
|
| 2354 |
"step": 328
|
| 2355 |
},
|
| 2356 |
{
|
| 2357 |
"epoch": 0.9748148148148148,
|
| 2358 |
-
"grad_norm": 0.
|
| 2359 |
-
"learning_rate":
|
| 2360 |
-
"loss": 0.
|
| 2361 |
"step": 329
|
| 2362 |
},
|
| 2363 |
{
|
| 2364 |
"epoch": 0.9777777777777777,
|
| 2365 |
-
"grad_norm": 0.
|
| 2366 |
-
"learning_rate":
|
| 2367 |
-
"loss": 0.
|
| 2368 |
"step": 330
|
| 2369 |
},
|
| 2370 |
{
|
| 2371 |
"epoch": 0.9807407407407407,
|
| 2372 |
-
"grad_norm": 0.
|
| 2373 |
-
"learning_rate":
|
| 2374 |
-
"loss": 0.
|
| 2375 |
"step": 331
|
| 2376 |
},
|
| 2377 |
{
|
| 2378 |
"epoch": 0.9837037037037037,
|
| 2379 |
-
"grad_norm": 0.
|
| 2380 |
-
"learning_rate":
|
| 2381 |
-
"loss": 0.
|
| 2382 |
"step": 332
|
| 2383 |
},
|
| 2384 |
{
|
| 2385 |
"epoch": 0.9866666666666667,
|
| 2386 |
-
"grad_norm": 0.
|
| 2387 |
-
"learning_rate":
|
| 2388 |
-
"loss": 0.
|
| 2389 |
"step": 333
|
| 2390 |
},
|
| 2391 |
{
|
| 2392 |
"epoch": 0.9896296296296296,
|
| 2393 |
-
"grad_norm": 0.
|
| 2394 |
-
"learning_rate":
|
| 2395 |
-
"loss": 0.
|
| 2396 |
"step": 334
|
| 2397 |
},
|
| 2398 |
{
|
| 2399 |
"epoch": 0.9925925925925926,
|
| 2400 |
-
"grad_norm": 0.
|
| 2401 |
-
"learning_rate":
|
| 2402 |
-
"loss": 0.
|
| 2403 |
"step": 335
|
| 2404 |
},
|
| 2405 |
{
|
| 2406 |
"epoch": 0.9955555555555555,
|
| 2407 |
-
"grad_norm": 0.
|
| 2408 |
-
"learning_rate":
|
| 2409 |
-
"loss": 0.
|
| 2410 |
"step": 336
|
| 2411 |
},
|
| 2412 |
{
|
| 2413 |
"epoch": 0.9985185185185185,
|
| 2414 |
-
"grad_norm": 0.
|
| 2415 |
-
"learning_rate":
|
| 2416 |
-
"loss": 0.
|
| 2417 |
"step": 337
|
| 2418 |
},
|
| 2419 |
{
|
| 2420 |
"epoch": 1.0,
|
| 2421 |
-
"grad_norm": 0.
|
| 2422 |
-
"learning_rate":
|
| 2423 |
-
"loss": 0.
|
| 2424 |
"step": 338
|
| 2425 |
}
|
| 2426 |
],
|
|
|
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.002962962962962963,
|
| 14 |
+
"grad_norm": 11.746185302734375,
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
"loss": 1.4603,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.005925925925925926,
|
| 21 |
+
"grad_norm": 12.557947158813477,
|
| 22 |
+
"learning_rate": 6e-06,
|
| 23 |
"loss": 1.662,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.008888888888888889,
|
| 28 |
+
"grad_norm": 10.80350112915039,
|
| 29 |
+
"learning_rate": 1.2e-05,
|
| 30 |
+
"loss": 1.482,
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.011851851851851851,
|
| 35 |
+
"grad_norm": 8.621623992919922,
|
| 36 |
+
"learning_rate": 1.8e-05,
|
| 37 |
+
"loss": 1.1961,
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.014814814814814815,
|
| 42 |
+
"grad_norm": 5.72680139541626,
|
| 43 |
+
"learning_rate": 2.4e-05,
|
| 44 |
+
"loss": 0.7526,
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.017777777777777778,
|
| 49 |
+
"grad_norm": 8.962793350219727,
|
| 50 |
+
"learning_rate": 3e-05,
|
| 51 |
+
"loss": 0.7239,
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.02074074074074074,
|
| 56 |
+
"grad_norm": 4.835477828979492,
|
| 57 |
+
"learning_rate": 2.9909909909909908e-05,
|
| 58 |
+
"loss": 0.4284,
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.023703703703703703,
|
| 63 |
+
"grad_norm": 2.002763032913208,
|
| 64 |
+
"learning_rate": 2.9819819819819822e-05,
|
| 65 |
+
"loss": 0.3711,
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.02666666666666667,
|
| 70 |
+
"grad_norm": 1.270501971244812,
|
| 71 |
+
"learning_rate": 2.972972972972973e-05,
|
| 72 |
+
"loss": 0.285,
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.02962962962962963,
|
| 77 |
+
"grad_norm": 1.124094843864441,
|
| 78 |
+
"learning_rate": 2.963963963963964e-05,
|
| 79 |
+
"loss": 0.2939,
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.03259259259259259,
|
| 84 |
+
"grad_norm": 0.8928890824317932,
|
| 85 |
+
"learning_rate": 2.954954954954955e-05,
|
| 86 |
+
"loss": 0.2322,
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.035555555555555556,
|
| 91 |
+
"grad_norm": 1.1215291023254395,
|
| 92 |
+
"learning_rate": 2.945945945945946e-05,
|
| 93 |
+
"loss": 0.3284,
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.03851851851851852,
|
| 98 |
+
"grad_norm": 0.6896376609802246,
|
| 99 |
+
"learning_rate": 2.936936936936937e-05,
|
| 100 |
+
"loss": 0.1688,
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.04148148148148148,
|
| 105 |
+
"grad_norm": 0.9925211668014526,
|
| 106 |
+
"learning_rate": 2.927927927927928e-05,
|
| 107 |
+
"loss": 0.3048,
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.044444444444444446,
|
| 112 |
+
"grad_norm": 0.7719022631645203,
|
| 113 |
+
"learning_rate": 2.918918918918919e-05,
|
| 114 |
+
"loss": 0.1955,
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.047407407407407405,
|
| 119 |
+
"grad_norm": 0.887997031211853,
|
| 120 |
+
"learning_rate": 2.90990990990991e-05,
|
| 121 |
+
"loss": 0.2717,
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.05037037037037037,
|
| 126 |
+
"grad_norm": 0.8488534092903137,
|
| 127 |
+
"learning_rate": 2.900900900900901e-05,
|
| 128 |
+
"loss": 0.2596,
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.05333333333333334,
|
| 133 |
+
"grad_norm": 0.7735264897346497,
|
| 134 |
+
"learning_rate": 2.891891891891892e-05,
|
| 135 |
+
"loss": 0.1438,
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.056296296296296296,
|
| 140 |
+
"grad_norm": 0.9446486234664917,
|
| 141 |
+
"learning_rate": 2.8828828828828828e-05,
|
| 142 |
+
"loss": 0.2588,
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.05925925925925926,
|
| 147 |
+
"grad_norm": 1.0172632932662964,
|
| 148 |
+
"learning_rate": 2.873873873873874e-05,
|
| 149 |
+
"loss": 0.3043,
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.06222222222222222,
|
| 154 |
+
"grad_norm": 0.9288851618766785,
|
| 155 |
+
"learning_rate": 2.864864864864865e-05,
|
| 156 |
+
"loss": 0.178,
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.06518518518518518,
|
| 161 |
+
"grad_norm": 1.0964614152908325,
|
| 162 |
+
"learning_rate": 2.855855855855856e-05,
|
| 163 |
+
"loss": 0.2958,
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.06814814814814815,
|
| 168 |
+
"grad_norm": 0.925119936466217,
|
| 169 |
+
"learning_rate": 2.846846846846847e-05,
|
| 170 |
+
"loss": 0.3181,
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.07111111111111111,
|
| 175 |
+
"grad_norm": 1.1452863216400146,
|
| 176 |
+
"learning_rate": 2.8378378378378378e-05,
|
| 177 |
+
"loss": 0.3221,
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.07407407407407407,
|
| 182 |
+
"grad_norm": 0.737485408782959,
|
| 183 |
+
"learning_rate": 2.828828828828829e-05,
|
| 184 |
+
"loss": 0.191,
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.07703703703703704,
|
| 189 |
+
"grad_norm": 0.8521791696548462,
|
| 190 |
+
"learning_rate": 2.81981981981982e-05,
|
| 191 |
+
"loss": 0.2746,
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.08,
|
| 196 |
+
"grad_norm": 1.0057052373886108,
|
| 197 |
+
"learning_rate": 2.8108108108108107e-05,
|
| 198 |
+
"loss": 0.3188,
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.08296296296296296,
|
| 203 |
+
"grad_norm": 0.9865684509277344,
|
| 204 |
+
"learning_rate": 2.801801801801802e-05,
|
| 205 |
+
"loss": 0.3238,
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.08592592592592592,
|
| 210 |
+
"grad_norm": 0.7946491837501526,
|
| 211 |
+
"learning_rate": 2.792792792792793e-05,
|
| 212 |
+
"loss": 0.2466,
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.08888888888888889,
|
| 217 |
+
"grad_norm": 0.650829553604126,
|
| 218 |
+
"learning_rate": 2.7837837837837836e-05,
|
| 219 |
+
"loss": 0.1926,
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.09185185185185185,
|
| 224 |
+
"grad_norm": 0.646949291229248,
|
| 225 |
+
"learning_rate": 2.774774774774775e-05,
|
| 226 |
+
"loss": 0.1657,
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.09481481481481481,
|
| 231 |
+
"grad_norm": 0.43284088373184204,
|
| 232 |
+
"learning_rate": 2.7657657657657658e-05,
|
| 233 |
+
"loss": 0.0852,
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.09777777777777778,
|
| 238 |
+
"grad_norm": 0.597963809967041,
|
| 239 |
+
"learning_rate": 2.756756756756757e-05,
|
| 240 |
+
"loss": 0.1536,
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.10074074074074074,
|
| 245 |
+
"grad_norm": 0.706259548664093,
|
| 246 |
+
"learning_rate": 2.747747747747748e-05,
|
| 247 |
"loss": 0.1996,
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.1037037037037037,
|
| 252 |
+
"grad_norm": 0.9260814189910889,
|
| 253 |
+
"learning_rate": 2.7387387387387387e-05,
|
| 254 |
+
"loss": 0.291,
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.10666666666666667,
|
| 259 |
+
"grad_norm": 0.7700441479682922,
|
| 260 |
+
"learning_rate": 2.7297297297297298e-05,
|
| 261 |
+
"loss": 0.1879,
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.10962962962962963,
|
| 266 |
+
"grad_norm": 0.7443490624427795,
|
| 267 |
+
"learning_rate": 2.720720720720721e-05,
|
| 268 |
+
"loss": 0.2275,
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.11259259259259259,
|
| 273 |
+
"grad_norm": 0.7297588586807251,
|
| 274 |
+
"learning_rate": 2.7117117117117116e-05,
|
| 275 |
+
"loss": 0.2054,
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.11555555555555555,
|
| 280 |
+
"grad_norm": 0.7750630974769592,
|
| 281 |
+
"learning_rate": 2.7027027027027027e-05,
|
| 282 |
+
"loss": 0.1496,
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.11851851851851852,
|
| 287 |
+
"grad_norm": 0.8763120770454407,
|
| 288 |
+
"learning_rate": 2.6936936936936938e-05,
|
| 289 |
+
"loss": 0.171,
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.12148148148148148,
|
| 294 |
+
"grad_norm": 0.9478317499160767,
|
| 295 |
+
"learning_rate": 2.684684684684685e-05,
|
| 296 |
+
"loss": 0.2252,
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.12444444444444444,
|
| 301 |
+
"grad_norm": 1.028319239616394,
|
| 302 |
+
"learning_rate": 2.6756756756756756e-05,
|
| 303 |
+
"loss": 0.3022,
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.1274074074074074,
|
| 308 |
+
"grad_norm": 0.5855357050895691,
|
| 309 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 310 |
+
"loss": 0.1266,
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.13037037037037036,
|
| 315 |
+
"grad_norm": 0.7112776637077332,
|
| 316 |
+
"learning_rate": 2.6576576576576577e-05,
|
| 317 |
+
"loss": 0.1682,
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.13333333333333333,
|
| 322 |
+
"grad_norm": 0.7943480610847473,
|
| 323 |
+
"learning_rate": 2.6486486486486485e-05,
|
| 324 |
+
"loss": 0.2211,
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.1362962962962963,
|
| 329 |
+
"grad_norm": 0.8550866842269897,
|
| 330 |
+
"learning_rate": 2.63963963963964e-05,
|
| 331 |
+
"loss": 0.2552,
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.13925925925925925,
|
| 336 |
+
"grad_norm": 1.2403337955474854,
|
| 337 |
+
"learning_rate": 2.6306306306306306e-05,
|
| 338 |
+
"loss": 0.3602,
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.14222222222222222,
|
| 343 |
+
"grad_norm": 0.5785015821456909,
|
| 344 |
+
"learning_rate": 2.6216216216216217e-05,
|
| 345 |
+
"loss": 0.1206,
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.1451851851851852,
|
| 350 |
+
"grad_norm": 0.8272493481636047,
|
| 351 |
+
"learning_rate": 2.6126126126126128e-05,
|
| 352 |
+
"loss": 0.1927,
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.14814814814814814,
|
| 357 |
+
"grad_norm": 0.6352198719978333,
|
| 358 |
+
"learning_rate": 2.6036036036036035e-05,
|
| 359 |
+
"loss": 0.1525,
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.14814814814814814,
|
| 364 |
+
"eval_loss": 0.18184001743793488,
|
| 365 |
+
"eval_runtime": 14.3322,
|
| 366 |
+
"eval_samples_per_second": 41.864,
|
| 367 |
+
"eval_steps_per_second": 5.233,
|
| 368 |
"step": 50
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.1511111111111111,
|
| 372 |
+
"grad_norm": 0.9342156648635864,
|
| 373 |
+
"learning_rate": 2.594594594594595e-05,
|
| 374 |
+
"loss": 0.2809,
|
| 375 |
"step": 51
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 0.15407407407407409,
|
| 379 |
+
"grad_norm": 0.8072832822799683,
|
| 380 |
+
"learning_rate": 2.5855855855855857e-05,
|
| 381 |
+
"loss": 0.2345,
|
| 382 |
"step": 52
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 0.15703703703703703,
|
| 386 |
+
"grad_norm": 0.8556937575340271,
|
| 387 |
+
"learning_rate": 2.5765765765765764e-05,
|
| 388 |
+
"loss": 0.2537,
|
| 389 |
"step": 53
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 0.16,
|
| 393 |
+
"grad_norm": 0.8798084855079651,
|
| 394 |
+
"learning_rate": 2.567567567567568e-05,
|
| 395 |
+
"loss": 0.2227,
|
| 396 |
"step": 54
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.16296296296296298,
|
| 400 |
+
"grad_norm": 0.5890814661979675,
|
| 401 |
+
"learning_rate": 2.5585585585585586e-05,
|
| 402 |
+
"loss": 0.1339,
|
| 403 |
"step": 55
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 0.16592592592592592,
|
| 407 |
+
"grad_norm": 0.7949811816215515,
|
| 408 |
+
"learning_rate": 2.5495495495495497e-05,
|
| 409 |
+
"loss": 0.1882,
|
| 410 |
"step": 56
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.1688888888888889,
|
| 414 |
+
"grad_norm": 0.6133240461349487,
|
| 415 |
+
"learning_rate": 2.5405405405405408e-05,
|
| 416 |
+
"loss": 0.1554,
|
| 417 |
"step": 57
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.17185185185185184,
|
| 421 |
+
"grad_norm": 0.508789598941803,
|
| 422 |
+
"learning_rate": 2.5315315315315315e-05,
|
| 423 |
+
"loss": 0.1072,
|
| 424 |
"step": 58
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 0.1748148148148148,
|
| 428 |
+
"grad_norm": 0.7848381996154785,
|
| 429 |
+
"learning_rate": 2.5225225225225226e-05,
|
| 430 |
+
"loss": 0.2298,
|
| 431 |
"step": 59
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 0.17777777777777778,
|
| 435 |
+
"grad_norm": 0.6376358270645142,
|
| 436 |
+
"learning_rate": 2.5135135135135137e-05,
|
| 437 |
+
"loss": 0.1922,
|
| 438 |
"step": 60
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 0.18074074074074073,
|
| 442 |
+
"grad_norm": 0.5989100337028503,
|
| 443 |
+
"learning_rate": 2.5045045045045044e-05,
|
| 444 |
+
"loss": 0.1713,
|
| 445 |
"step": 61
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 0.1837037037037037,
|
| 449 |
+
"grad_norm": 0.5262016654014587,
|
| 450 |
+
"learning_rate": 2.4954954954954955e-05,
|
| 451 |
+
"loss": 0.1425,
|
| 452 |
"step": 62
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.18666666666666668,
|
| 456 |
+
"grad_norm": 0.8312467336654663,
|
| 457 |
+
"learning_rate": 2.4864864864864866e-05,
|
| 458 |
+
"loss": 0.2618,
|
| 459 |
"step": 63
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 0.18962962962962962,
|
| 463 |
+
"grad_norm": 0.6138928532600403,
|
| 464 |
+
"learning_rate": 2.4774774774774777e-05,
|
| 465 |
+
"loss": 0.1872,
|
| 466 |
"step": 64
|
| 467 |
},
|
| 468 |
{
|
| 469 |
"epoch": 0.1925925925925926,
|
| 470 |
+
"grad_norm": 0.7495738863945007,
|
| 471 |
+
"learning_rate": 2.4684684684684684e-05,
|
| 472 |
+
"loss": 0.2121,
|
| 473 |
"step": 65
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 0.19555555555555557,
|
| 477 |
+
"grad_norm": 0.7085468769073486,
|
| 478 |
+
"learning_rate": 2.4594594594594595e-05,
|
| 479 |
+
"loss": 0.1641,
|
| 480 |
"step": 66
|
| 481 |
},
|
| 482 |
{
|
| 483 |
"epoch": 0.1985185185185185,
|
| 484 |
+
"grad_norm": 0.7193159461021423,
|
| 485 |
+
"learning_rate": 2.4504504504504506e-05,
|
| 486 |
+
"loss": 0.2518,
|
| 487 |
"step": 67
|
| 488 |
},
|
| 489 |
{
|
| 490 |
"epoch": 0.20148148148148148,
|
| 491 |
+
"grad_norm": 0.964632511138916,
|
| 492 |
+
"learning_rate": 2.4414414414414413e-05,
|
| 493 |
+
"loss": 0.2928,
|
| 494 |
"step": 68
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.20444444444444446,
|
| 498 |
+
"grad_norm": 0.7328157424926758,
|
| 499 |
+
"learning_rate": 2.4324324324324327e-05,
|
| 500 |
+
"loss": 0.2096,
|
| 501 |
"step": 69
|
| 502 |
},
|
| 503 |
{
|
| 504 |
"epoch": 0.2074074074074074,
|
| 505 |
+
"grad_norm": 0.5567067861557007,
|
| 506 |
+
"learning_rate": 2.4234234234234235e-05,
|
| 507 |
+
"loss": 0.118,
|
| 508 |
"step": 70
|
| 509 |
},
|
| 510 |
{
|
| 511 |
"epoch": 0.21037037037037037,
|
| 512 |
+
"grad_norm": 0.754827618598938,
|
| 513 |
+
"learning_rate": 2.4144144144144142e-05,
|
| 514 |
+
"loss": 0.2237,
|
| 515 |
"step": 71
|
| 516 |
},
|
| 517 |
{
|
| 518 |
"epoch": 0.21333333333333335,
|
| 519 |
+
"grad_norm": 0.8856765031814575,
|
| 520 |
+
"learning_rate": 2.4054054054054056e-05,
|
| 521 |
+
"loss": 0.2326,
|
| 522 |
"step": 72
|
| 523 |
},
|
| 524 |
{
|
| 525 |
"epoch": 0.2162962962962963,
|
| 526 |
+
"grad_norm": 0.6610215902328491,
|
| 527 |
+
"learning_rate": 2.3963963963963964e-05,
|
| 528 |
+
"loss": 0.1446,
|
| 529 |
"step": 73
|
| 530 |
},
|
| 531 |
{
|
| 532 |
"epoch": 0.21925925925925926,
|
| 533 |
+
"grad_norm": 0.5447685122489929,
|
| 534 |
+
"learning_rate": 2.3873873873873874e-05,
|
| 535 |
+
"loss": 0.1208,
|
| 536 |
"step": 74
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.2222222222222222,
|
| 540 |
+
"grad_norm": 0.692625105381012,
|
| 541 |
+
"learning_rate": 2.3783783783783785e-05,
|
| 542 |
+
"loss": 0.1657,
|
| 543 |
"step": 75
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 0.22518518518518518,
|
| 547 |
+
"grad_norm": 0.6857089996337891,
|
| 548 |
+
"learning_rate": 2.3693693693693693e-05,
|
| 549 |
+
"loss": 0.1906,
|
| 550 |
"step": 76
|
| 551 |
},
|
| 552 |
{
|
| 553 |
"epoch": 0.22814814814814816,
|
| 554 |
+
"grad_norm": 0.815542995929718,
|
| 555 |
+
"learning_rate": 2.3603603603603607e-05,
|
| 556 |
+
"loss": 0.2234,
|
| 557 |
"step": 77
|
| 558 |
},
|
| 559 |
{
|
| 560 |
"epoch": 0.2311111111111111,
|
| 561 |
+
"grad_norm": 0.8306949138641357,
|
| 562 |
+
"learning_rate": 2.3513513513513514e-05,
|
| 563 |
+
"loss": 0.227,
|
| 564 |
"step": 78
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 0.23407407407407407,
|
| 568 |
+
"grad_norm": 0.575930118560791,
|
| 569 |
+
"learning_rate": 2.342342342342342e-05,
|
| 570 |
+
"loss": 0.1329,
|
| 571 |
"step": 79
|
| 572 |
},
|
| 573 |
{
|
| 574 |
"epoch": 0.23703703703703705,
|
| 575 |
+
"grad_norm": 0.7562434673309326,
|
| 576 |
+
"learning_rate": 2.3333333333333336e-05,
|
| 577 |
+
"loss": 0.2114,
|
| 578 |
"step": 80
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 0.24,
|
| 582 |
+
"grad_norm": 0.6087313294410706,
|
| 583 |
+
"learning_rate": 2.3243243243243243e-05,
|
| 584 |
+
"loss": 0.1878,
|
| 585 |
"step": 81
|
| 586 |
},
|
| 587 |
{
|
| 588 |
"epoch": 0.24296296296296296,
|
| 589 |
+
"grad_norm": 0.6526697874069214,
|
| 590 |
+
"learning_rate": 2.3153153153153154e-05,
|
| 591 |
+
"loss": 0.1527,
|
| 592 |
"step": 82
|
| 593 |
},
|
| 594 |
{
|
| 595 |
"epoch": 0.24592592592592594,
|
| 596 |
+
"grad_norm": 0.5339913964271545,
|
| 597 |
+
"learning_rate": 2.3063063063063065e-05,
|
| 598 |
+
"loss": 0.1332,
|
| 599 |
"step": 83
|
| 600 |
},
|
| 601 |
{
|
| 602 |
"epoch": 0.24888888888888888,
|
| 603 |
+
"grad_norm": 0.7385029196739197,
|
| 604 |
+
"learning_rate": 2.2972972972972972e-05,
|
| 605 |
+
"loss": 0.205,
|
| 606 |
"step": 84
|
| 607 |
},
|
| 608 |
{
|
| 609 |
"epoch": 0.2518518518518518,
|
| 610 |
+
"grad_norm": 0.721841037273407,
|
| 611 |
+
"learning_rate": 2.2882882882882883e-05,
|
| 612 |
+
"loss": 0.2236,
|
| 613 |
"step": 85
|
| 614 |
},
|
| 615 |
{
|
| 616 |
"epoch": 0.2548148148148148,
|
| 617 |
+
"grad_norm": 0.5965892672538757,
|
| 618 |
+
"learning_rate": 2.2792792792792794e-05,
|
| 619 |
+
"loss": 0.1677,
|
| 620 |
"step": 86
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 0.2577777777777778,
|
| 624 |
+
"grad_norm": 1.0791587829589844,
|
| 625 |
+
"learning_rate": 2.2702702702702705e-05,
|
| 626 |
+
"loss": 0.3187,
|
| 627 |
"step": 87
|
| 628 |
},
|
| 629 |
{
|
| 630 |
"epoch": 0.2607407407407407,
|
| 631 |
+
"grad_norm": 0.7455445528030396,
|
| 632 |
+
"learning_rate": 2.2612612612612612e-05,
|
| 633 |
+
"loss": 0.2389,
|
| 634 |
"step": 88
|
| 635 |
},
|
| 636 |
{
|
| 637 |
"epoch": 0.2637037037037037,
|
| 638 |
+
"grad_norm": 0.5387834310531616,
|
| 639 |
+
"learning_rate": 2.2522522522522523e-05,
|
| 640 |
+
"loss": 0.148,
|
| 641 |
"step": 89
|
| 642 |
},
|
| 643 |
{
|
| 644 |
"epoch": 0.26666666666666666,
|
| 645 |
+
"grad_norm": 0.6185374855995178,
|
| 646 |
+
"learning_rate": 2.2432432432432434e-05,
|
| 647 |
+
"loss": 0.1763,
|
| 648 |
"step": 90
|
| 649 |
},
|
| 650 |
{
|
| 651 |
"epoch": 0.2696296296296296,
|
| 652 |
+
"grad_norm": 0.6860742568969727,
|
| 653 |
+
"learning_rate": 2.234234234234234e-05,
|
| 654 |
+
"loss": 0.2481,
|
| 655 |
"step": 91
|
| 656 |
},
|
| 657 |
{
|
| 658 |
"epoch": 0.2725925925925926,
|
| 659 |
+
"grad_norm": 0.5713387131690979,
|
| 660 |
+
"learning_rate": 2.2252252252252255e-05,
|
| 661 |
+
"loss": 0.1486,
|
| 662 |
"step": 92
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 0.27555555555555555,
|
| 666 |
+
"grad_norm": 0.4434517025947571,
|
| 667 |
+
"learning_rate": 2.2162162162162163e-05,
|
| 668 |
+
"loss": 0.1053,
|
| 669 |
"step": 93
|
| 670 |
},
|
| 671 |
{
|
| 672 |
"epoch": 0.2785185185185185,
|
| 673 |
+
"grad_norm": 0.7606307864189148,
|
| 674 |
+
"learning_rate": 2.207207207207207e-05,
|
| 675 |
+
"loss": 0.2736,
|
| 676 |
"step": 94
|
| 677 |
},
|
| 678 |
{
|
| 679 |
"epoch": 0.2814814814814815,
|
| 680 |
+
"grad_norm": 0.6613526940345764,
|
| 681 |
+
"learning_rate": 2.1981981981981984e-05,
|
| 682 |
+
"loss": 0.2177,
|
| 683 |
"step": 95
|
| 684 |
},
|
| 685 |
{
|
| 686 |
"epoch": 0.28444444444444444,
|
| 687 |
+
"grad_norm": 0.7913828492164612,
|
| 688 |
+
"learning_rate": 2.1891891891891892e-05,
|
| 689 |
+
"loss": 0.2088,
|
| 690 |
"step": 96
|
| 691 |
},
|
| 692 |
{
|
| 693 |
"epoch": 0.2874074074074074,
|
| 694 |
+
"grad_norm": 0.6992068886756897,
|
| 695 |
+
"learning_rate": 2.1801801801801803e-05,
|
| 696 |
+
"loss": 0.2376,
|
| 697 |
"step": 97
|
| 698 |
},
|
| 699 |
{
|
| 700 |
"epoch": 0.2903703703703704,
|
| 701 |
+
"grad_norm": 0.6808008551597595,
|
| 702 |
+
"learning_rate": 2.1711711711711713e-05,
|
| 703 |
+
"loss": 0.1901,
|
| 704 |
"step": 98
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 0.29333333333333333,
|
| 708 |
+
"grad_norm": 0.8115521669387817,
|
| 709 |
+
"learning_rate": 2.162162162162162e-05,
|
| 710 |
+
"loss": 0.3181,
|
| 711 |
"step": 99
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 0.2962962962962963,
|
| 715 |
+
"grad_norm": 0.6143046021461487,
|
| 716 |
+
"learning_rate": 2.153153153153153e-05,
|
| 717 |
+
"loss": 0.2197,
|
| 718 |
"step": 100
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 0.2962962962962963,
|
| 722 |
+
"eval_loss": 0.1685572862625122,
|
| 723 |
+
"eval_runtime": 14.0722,
|
| 724 |
+
"eval_samples_per_second": 42.637,
|
| 725 |
+
"eval_steps_per_second": 5.33,
|
| 726 |
"step": 100
|
| 727 |
},
|
| 728 |
{
|
| 729 |
"epoch": 0.2992592592592593,
|
| 730 |
+
"grad_norm": 0.5645914673805237,
|
| 731 |
+
"learning_rate": 2.1441441441441442e-05,
|
| 732 |
+
"loss": 0.1305,
|
| 733 |
"step": 101
|
| 734 |
},
|
| 735 |
{
|
| 736 |
"epoch": 0.3022222222222222,
|
| 737 |
+
"grad_norm": 0.6650404334068298,
|
| 738 |
+
"learning_rate": 2.135135135135135e-05,
|
| 739 |
+
"loss": 0.186,
|
| 740 |
"step": 102
|
| 741 |
},
|
| 742 |
{
|
| 743 |
"epoch": 0.30518518518518517,
|
| 744 |
+
"grad_norm": 0.8169294595718384,
|
| 745 |
+
"learning_rate": 2.126126126126126e-05,
|
| 746 |
+
"loss": 0.2747,
|
| 747 |
"step": 103
|
| 748 |
},
|
| 749 |
{
|
| 750 |
"epoch": 0.30814814814814817,
|
| 751 |
+
"grad_norm": 0.6059281826019287,
|
| 752 |
+
"learning_rate": 2.117117117117117e-05,
|
| 753 |
+
"loss": 0.1654,
|
| 754 |
"step": 104
|
| 755 |
},
|
| 756 |
{
|
| 757 |
"epoch": 0.3111111111111111,
|
| 758 |
+
"grad_norm": 0.7298572659492493,
|
| 759 |
+
"learning_rate": 2.1081081081081082e-05,
|
| 760 |
+
"loss": 0.2615,
|
| 761 |
"step": 105
|
| 762 |
},
|
| 763 |
{
|
| 764 |
"epoch": 0.31407407407407406,
|
| 765 |
+
"grad_norm": 0.47728487849235535,
|
| 766 |
+
"learning_rate": 2.0990990990990993e-05,
|
| 767 |
+
"loss": 0.1333,
|
| 768 |
"step": 106
|
| 769 |
},
|
| 770 |
{
|
| 771 |
"epoch": 0.31703703703703706,
|
| 772 |
+
"grad_norm": 0.7812051773071289,
|
| 773 |
+
"learning_rate": 2.09009009009009e-05,
|
| 774 |
+
"loss": 0.1698,
|
| 775 |
"step": 107
|
| 776 |
},
|
| 777 |
{
|
| 778 |
"epoch": 0.32,
|
| 779 |
+
"grad_norm": 0.604000985622406,
|
| 780 |
+
"learning_rate": 2.081081081081081e-05,
|
| 781 |
+
"loss": 0.1674,
|
| 782 |
"step": 108
|
| 783 |
},
|
| 784 |
{
|
| 785 |
"epoch": 0.32296296296296295,
|
| 786 |
+
"grad_norm": 0.7859566807746887,
|
| 787 |
+
"learning_rate": 2.0720720720720722e-05,
|
| 788 |
+
"loss": 0.2189,
|
| 789 |
"step": 109
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 0.32592592592592595,
|
| 793 |
+
"grad_norm": 0.6453425288200378,
|
| 794 |
+
"learning_rate": 2.0630630630630633e-05,
|
| 795 |
+
"loss": 0.2027,
|
| 796 |
"step": 110
|
| 797 |
},
|
| 798 |
{
|
| 799 |
"epoch": 0.3288888888888889,
|
| 800 |
+
"grad_norm": 0.8944854140281677,
|
| 801 |
+
"learning_rate": 2.054054054054054e-05,
|
| 802 |
+
"loss": 0.2321,
|
| 803 |
"step": 111
|
| 804 |
},
|
| 805 |
{
|
| 806 |
"epoch": 0.33185185185185184,
|
| 807 |
+
"grad_norm": 0.6334295868873596,
|
| 808 |
+
"learning_rate": 2.045045045045045e-05,
|
| 809 |
+
"loss": 0.1806,
|
| 810 |
"step": 112
|
| 811 |
},
|
| 812 |
{
|
| 813 |
"epoch": 0.3348148148148148,
|
| 814 |
+
"grad_norm": 0.6126565933227539,
|
| 815 |
+
"learning_rate": 2.0360360360360362e-05,
|
| 816 |
+
"loss": 0.1625,
|
| 817 |
"step": 113
|
| 818 |
},
|
| 819 |
{
|
| 820 |
"epoch": 0.3377777777777778,
|
| 821 |
+
"grad_norm": 0.7241846323013306,
|
| 822 |
+
"learning_rate": 2.027027027027027e-05,
|
| 823 |
+
"loss": 0.2333,
|
| 824 |
"step": 114
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 0.34074074074074073,
|
| 828 |
+
"grad_norm": 0.6099463701248169,
|
| 829 |
+
"learning_rate": 2.0180180180180183e-05,
|
| 830 |
"loss": 0.1768,
|
| 831 |
"step": 115
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 0.3437037037037037,
|
| 835 |
+
"grad_norm": 0.951642632484436,
|
| 836 |
+
"learning_rate": 2.009009009009009e-05,
|
| 837 |
+
"loss": 0.2856,
|
| 838 |
"step": 116
|
| 839 |
},
|
| 840 |
{
|
| 841 |
"epoch": 0.3466666666666667,
|
| 842 |
+
"grad_norm": 0.733674943447113,
|
| 843 |
+
"learning_rate": 1.9999999999999998e-05,
|
| 844 |
+
"loss": 0.1836,
|
| 845 |
"step": 117
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 0.3496296296296296,
|
| 849 |
+
"grad_norm": 0.7276878356933594,
|
| 850 |
+
"learning_rate": 1.9909909909909912e-05,
|
| 851 |
+
"loss": 0.2382,
|
| 852 |
"step": 118
|
| 853 |
},
|
| 854 |
{
|
| 855 |
"epoch": 0.35259259259259257,
|
| 856 |
+
"grad_norm": 0.756348729133606,
|
| 857 |
+
"learning_rate": 1.981981981981982e-05,
|
| 858 |
+
"loss": 0.2648,
|
| 859 |
"step": 119
|
| 860 |
},
|
| 861 |
{
|
| 862 |
"epoch": 0.35555555555555557,
|
| 863 |
+
"grad_norm": 0.6031218767166138,
|
| 864 |
+
"learning_rate": 1.972972972972973e-05,
|
| 865 |
+
"loss": 0.1861,
|
| 866 |
"step": 120
|
| 867 |
},
|
| 868 |
{
|
| 869 |
"epoch": 0.3585185185185185,
|
| 870 |
+
"grad_norm": 0.6297987103462219,
|
| 871 |
+
"learning_rate": 1.963963963963964e-05,
|
| 872 |
+
"loss": 0.2095,
|
| 873 |
"step": 121
|
| 874 |
},
|
| 875 |
{
|
| 876 |
"epoch": 0.36148148148148146,
|
| 877 |
+
"grad_norm": 0.5252201557159424,
|
| 878 |
+
"learning_rate": 1.954954954954955e-05,
|
| 879 |
+
"loss": 0.1721,
|
| 880 |
"step": 122
|
| 881 |
},
|
| 882 |
{
|
| 883 |
"epoch": 0.36444444444444446,
|
| 884 |
+
"grad_norm": 0.6074128746986389,
|
| 885 |
+
"learning_rate": 1.945945945945946e-05,
|
| 886 |
+
"loss": 0.1657,
|
| 887 |
"step": 123
|
| 888 |
},
|
| 889 |
{
|
| 890 |
"epoch": 0.3674074074074074,
|
| 891 |
+
"grad_norm": 0.8106815218925476,
|
| 892 |
+
"learning_rate": 1.936936936936937e-05,
|
| 893 |
+
"loss": 0.261,
|
| 894 |
"step": 124
|
| 895 |
},
|
| 896 |
{
|
| 897 |
"epoch": 0.37037037037037035,
|
| 898 |
+
"grad_norm": 0.7447537779808044,
|
| 899 |
+
"learning_rate": 1.9279279279279278e-05,
|
| 900 |
+
"loss": 0.2073,
|
| 901 |
"step": 125
|
| 902 |
},
|
| 903 |
{
|
| 904 |
"epoch": 0.37333333333333335,
|
| 905 |
+
"grad_norm": 0.7889114022254944,
|
| 906 |
+
"learning_rate": 1.918918918918919e-05,
|
| 907 |
+
"loss": 0.2474,
|
| 908 |
"step": 126
|
| 909 |
},
|
| 910 |
{
|
| 911 |
"epoch": 0.3762962962962963,
|
| 912 |
+
"grad_norm": 0.8324449062347412,
|
| 913 |
+
"learning_rate": 1.90990990990991e-05,
|
| 914 |
+
"loss": 0.1576,
|
| 915 |
"step": 127
|
| 916 |
},
|
| 917 |
{
|
| 918 |
"epoch": 0.37925925925925924,
|
| 919 |
+
"grad_norm": 0.5501788854598999,
|
| 920 |
+
"learning_rate": 1.900900900900901e-05,
|
| 921 |
+
"loss": 0.1469,
|
| 922 |
"step": 128
|
| 923 |
},
|
| 924 |
{
|
| 925 |
"epoch": 0.38222222222222224,
|
| 926 |
+
"grad_norm": 0.6907466650009155,
|
| 927 |
+
"learning_rate": 1.8918918918918918e-05,
|
| 928 |
+
"loss": 0.16,
|
| 929 |
"step": 129
|
| 930 |
},
|
| 931 |
{
|
| 932 |
"epoch": 0.3851851851851852,
|
| 933 |
+
"grad_norm": 0.6078192591667175,
|
| 934 |
+
"learning_rate": 1.882882882882883e-05,
|
| 935 |
+
"loss": 0.1342,
|
| 936 |
"step": 130
|
| 937 |
},
|
| 938 |
{
|
| 939 |
"epoch": 0.38814814814814813,
|
| 940 |
+
"grad_norm": 0.7098512649536133,
|
| 941 |
+
"learning_rate": 1.873873873873874e-05,
|
| 942 |
+
"loss": 0.1977,
|
| 943 |
"step": 131
|
| 944 |
},
|
| 945 |
{
|
| 946 |
"epoch": 0.39111111111111113,
|
| 947 |
+
"grad_norm": 0.7604762315750122,
|
| 948 |
+
"learning_rate": 1.864864864864865e-05,
|
| 949 |
+
"loss": 0.1754,
|
| 950 |
"step": 132
|
| 951 |
},
|
| 952 |
{
|
| 953 |
"epoch": 0.3940740740740741,
|
| 954 |
+
"grad_norm": 0.750038206577301,
|
| 955 |
+
"learning_rate": 1.855855855855856e-05,
|
| 956 |
+
"loss": 0.2389,
|
| 957 |
"step": 133
|
| 958 |
},
|
| 959 |
{
|
| 960 |
"epoch": 0.397037037037037,
|
| 961 |
+
"grad_norm": 0.7395908832550049,
|
| 962 |
+
"learning_rate": 1.846846846846847e-05,
|
| 963 |
+
"loss": 0.218,
|
| 964 |
"step": 134
|
| 965 |
},
|
| 966 |
{
|
| 967 |
"epoch": 0.4,
|
| 968 |
+
"grad_norm": 0.6242744326591492,
|
| 969 |
+
"learning_rate": 1.837837837837838e-05,
|
| 970 |
+
"loss": 0.1861,
|
| 971 |
"step": 135
|
| 972 |
},
|
| 973 |
{
|
| 974 |
"epoch": 0.40296296296296297,
|
| 975 |
+
"grad_norm": 0.5683760046958923,
|
| 976 |
+
"learning_rate": 1.828828828828829e-05,
|
| 977 |
+
"loss": 0.124,
|
| 978 |
"step": 136
|
| 979 |
},
|
| 980 |
{
|
| 981 |
"epoch": 0.4059259259259259,
|
| 982 |
+
"grad_norm": 0.6790345311164856,
|
| 983 |
+
"learning_rate": 1.8198198198198197e-05,
|
| 984 |
+
"loss": 0.1422,
|
| 985 |
"step": 137
|
| 986 |
},
|
| 987 |
{
|
| 988 |
"epoch": 0.4088888888888889,
|
| 989 |
+
"grad_norm": 0.658165693283081,
|
| 990 |
+
"learning_rate": 1.810810810810811e-05,
|
| 991 |
+
"loss": 0.1611,
|
| 992 |
"step": 138
|
| 993 |
},
|
| 994 |
{
|
| 995 |
"epoch": 0.41185185185185186,
|
| 996 |
+
"grad_norm": 0.6879218220710754,
|
| 997 |
+
"learning_rate": 1.801801801801802e-05,
|
| 998 |
"loss": 0.1839,
|
| 999 |
"step": 139
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
"epoch": 0.4148148148148148,
|
| 1003 |
+
"grad_norm": 1.0073219537734985,
|
| 1004 |
+
"learning_rate": 1.7927927927927926e-05,
|
| 1005 |
+
"loss": 0.3063,
|
| 1006 |
"step": 140
|
| 1007 |
},
|
| 1008 |
{
|
| 1009 |
"epoch": 0.4177777777777778,
|
| 1010 |
+
"grad_norm": 0.780751645565033,
|
| 1011 |
+
"learning_rate": 1.783783783783784e-05,
|
| 1012 |
+
"loss": 0.2481,
|
| 1013 |
"step": 141
|
| 1014 |
},
|
| 1015 |
{
|
| 1016 |
"epoch": 0.42074074074074075,
|
| 1017 |
+
"grad_norm": 0.6793040037155151,
|
| 1018 |
+
"learning_rate": 1.7747747747747748e-05,
|
| 1019 |
+
"loss": 0.1434,
|
| 1020 |
"step": 142
|
| 1021 |
},
|
| 1022 |
{
|
| 1023 |
"epoch": 0.4237037037037037,
|
| 1024 |
+
"grad_norm": 0.7167030572891235,
|
| 1025 |
+
"learning_rate": 1.7657657657657655e-05,
|
| 1026 |
+
"loss": 0.2328,
|
| 1027 |
"step": 143
|
| 1028 |
},
|
| 1029 |
{
|
| 1030 |
"epoch": 0.4266666666666667,
|
| 1031 |
+
"grad_norm": 0.7468717694282532,
|
| 1032 |
+
"learning_rate": 1.756756756756757e-05,
|
| 1033 |
+
"loss": 0.2056,
|
| 1034 |
"step": 144
|
| 1035 |
},
|
| 1036 |
{
|
| 1037 |
"epoch": 0.42962962962962964,
|
| 1038 |
+
"grad_norm": 0.6910548210144043,
|
| 1039 |
+
"learning_rate": 1.7477477477477477e-05,
|
| 1040 |
+
"loss": 0.1464,
|
| 1041 |
"step": 145
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
"epoch": 0.4325925925925926,
|
| 1045 |
+
"grad_norm": 0.4421556293964386,
|
| 1046 |
+
"learning_rate": 1.7387387387387388e-05,
|
| 1047 |
+
"loss": 0.0964,
|
| 1048 |
"step": 146
|
| 1049 |
},
|
| 1050 |
{
|
| 1051 |
"epoch": 0.43555555555555553,
|
| 1052 |
+
"grad_norm": 0.5973996520042419,
|
| 1053 |
+
"learning_rate": 1.72972972972973e-05,
|
| 1054 |
+
"loss": 0.1431,
|
| 1055 |
"step": 147
|
| 1056 |
},
|
| 1057 |
{
|
| 1058 |
"epoch": 0.43851851851851853,
|
| 1059 |
+
"grad_norm": 0.6073348522186279,
|
| 1060 |
+
"learning_rate": 1.7207207207207206e-05,
|
| 1061 |
+
"loss": 0.1837,
|
| 1062 |
"step": 148
|
| 1063 |
},
|
| 1064 |
{
|
| 1065 |
"epoch": 0.4414814814814815,
|
| 1066 |
+
"grad_norm": 0.6484878063201904,
|
| 1067 |
+
"learning_rate": 1.7117117117117117e-05,
|
| 1068 |
+
"loss": 0.1853,
|
| 1069 |
"step": 149
|
| 1070 |
},
|
| 1071 |
{
|
| 1072 |
"epoch": 0.4444444444444444,
|
| 1073 |
+
"grad_norm": 0.7600032687187195,
|
| 1074 |
+
"learning_rate": 1.7027027027027028e-05,
|
| 1075 |
+
"loss": 0.2407,
|
| 1076 |
"step": 150
|
| 1077 |
},
|
| 1078 |
{
|
| 1079 |
"epoch": 0.4444444444444444,
|
| 1080 |
+
"eval_loss": 0.1618269383907318,
|
| 1081 |
+
"eval_runtime": 14.1337,
|
| 1082 |
+
"eval_samples_per_second": 42.452,
|
| 1083 |
+
"eval_steps_per_second": 5.306,
|
| 1084 |
"step": 150
|
| 1085 |
},
|
| 1086 |
{
|
| 1087 |
"epoch": 0.4474074074074074,
|
| 1088 |
+
"grad_norm": 0.8500089049339294,
|
| 1089 |
+
"learning_rate": 1.693693693693694e-05,
|
| 1090 |
+
"loss": 0.248,
|
| 1091 |
"step": 151
|
| 1092 |
},
|
| 1093 |
{
|
| 1094 |
"epoch": 0.45037037037037037,
|
| 1095 |
+
"grad_norm": 0.48620492219924927,
|
| 1096 |
+
"learning_rate": 1.6846846846846846e-05,
|
| 1097 |
+
"loss": 0.1389,
|
| 1098 |
"step": 152
|
| 1099 |
},
|
| 1100 |
{
|
| 1101 |
"epoch": 0.4533333333333333,
|
| 1102 |
+
"grad_norm": 0.6419157385826111,
|
| 1103 |
+
"learning_rate": 1.6756756756756757e-05,
|
| 1104 |
+
"loss": 0.1489,
|
| 1105 |
"step": 153
|
| 1106 |
},
|
| 1107 |
{
|
| 1108 |
"epoch": 0.4562962962962963,
|
| 1109 |
+
"grad_norm": 0.7250919938087463,
|
| 1110 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1111 |
+
"loss": 0.1896,
|
| 1112 |
"step": 154
|
| 1113 |
},
|
| 1114 |
{
|
| 1115 |
"epoch": 0.45925925925925926,
|
| 1116 |
+
"grad_norm": 0.5173049569129944,
|
| 1117 |
+
"learning_rate": 1.6576576576576575e-05,
|
| 1118 |
+
"loss": 0.1172,
|
| 1119 |
"step": 155
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
"epoch": 0.4622222222222222,
|
| 1123 |
+
"grad_norm": 0.338739275932312,
|
| 1124 |
+
"learning_rate": 1.648648648648649e-05,
|
| 1125 |
+
"loss": 0.0475,
|
| 1126 |
"step": 156
|
| 1127 |
},
|
| 1128 |
{
|
| 1129 |
"epoch": 0.4651851851851852,
|
| 1130 |
+
"grad_norm": 0.7956584095954895,
|
| 1131 |
+
"learning_rate": 1.6396396396396396e-05,
|
| 1132 |
+
"loss": 0.2128,
|
| 1133 |
"step": 157
|
| 1134 |
},
|
| 1135 |
{
|
| 1136 |
"epoch": 0.46814814814814815,
|
| 1137 |
+
"grad_norm": 0.5871915221214294,
|
| 1138 |
+
"learning_rate": 1.6306306306306304e-05,
|
| 1139 |
+
"loss": 0.1504,
|
| 1140 |
"step": 158
|
| 1141 |
},
|
| 1142 |
{
|
| 1143 |
"epoch": 0.4711111111111111,
|
| 1144 |
+
"grad_norm": 0.7355495691299438,
|
| 1145 |
+
"learning_rate": 1.6216216216216218e-05,
|
| 1146 |
+
"loss": 0.2194,
|
| 1147 |
"step": 159
|
| 1148 |
},
|
| 1149 |
{
|
| 1150 |
"epoch": 0.4740740740740741,
|
| 1151 |
+
"grad_norm": 0.5841131210327148,
|
| 1152 |
+
"learning_rate": 1.6126126126126126e-05,
|
| 1153 |
+
"loss": 0.1257,
|
| 1154 |
"step": 160
|
| 1155 |
},
|
| 1156 |
{
|
| 1157 |
"epoch": 0.47703703703703704,
|
| 1158 |
+
"grad_norm": 0.6331678628921509,
|
| 1159 |
+
"learning_rate": 1.603603603603604e-05,
|
| 1160 |
+
"loss": 0.1862,
|
| 1161 |
"step": 161
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
"epoch": 0.48,
|
| 1165 |
+
"grad_norm": 0.6405560970306396,
|
| 1166 |
+
"learning_rate": 1.5945945945945947e-05,
|
| 1167 |
+
"loss": 0.1326,
|
| 1168 |
"step": 162
|
| 1169 |
},
|
| 1170 |
{
|
| 1171 |
"epoch": 0.482962962962963,
|
| 1172 |
+
"grad_norm": 0.7067241668701172,
|
| 1173 |
+
"learning_rate": 1.5855855855855855e-05,
|
| 1174 |
+
"loss": 0.1998,
|
| 1175 |
"step": 163
|
| 1176 |
},
|
| 1177 |
{
|
| 1178 |
"epoch": 0.48592592592592593,
|
| 1179 |
+
"grad_norm": 0.5503882765769958,
|
| 1180 |
+
"learning_rate": 1.576576576576577e-05,
|
| 1181 |
+
"loss": 0.1112,
|
| 1182 |
"step": 164
|
| 1183 |
},
|
| 1184 |
{
|
| 1185 |
"epoch": 0.4888888888888889,
|
| 1186 |
+
"grad_norm": 0.6191250085830688,
|
| 1187 |
+
"learning_rate": 1.5675675675675676e-05,
|
| 1188 |
+
"loss": 0.1862,
|
| 1189 |
"step": 165
|
| 1190 |
},
|
| 1191 |
{
|
| 1192 |
"epoch": 0.4918518518518519,
|
| 1193 |
+
"grad_norm": 0.5445926189422607,
|
| 1194 |
+
"learning_rate": 1.5585585585585584e-05,
|
| 1195 |
+
"loss": 0.129,
|
| 1196 |
"step": 166
|
| 1197 |
},
|
| 1198 |
{
|
| 1199 |
"epoch": 0.4948148148148148,
|
| 1200 |
+
"grad_norm": 0.6146531701087952,
|
| 1201 |
+
"learning_rate": 1.5495495495495498e-05,
|
| 1202 |
+
"loss": 0.1442,
|
| 1203 |
"step": 167
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
"epoch": 0.49777777777777776,
|
| 1207 |
+
"grad_norm": 0.8105955719947815,
|
| 1208 |
+
"learning_rate": 1.5405405405405405e-05,
|
| 1209 |
+
"loss": 0.2753,
|
| 1210 |
"step": 168
|
| 1211 |
},
|
| 1212 |
{
|
| 1213 |
"epoch": 0.5007407407407407,
|
| 1214 |
+
"grad_norm": 0.7763640880584717,
|
| 1215 |
+
"learning_rate": 1.5315315315315316e-05,
|
| 1216 |
+
"loss": 0.1617,
|
| 1217 |
"step": 169
|
| 1218 |
},
|
| 1219 |
{
|
| 1220 |
"epoch": 0.5037037037037037,
|
| 1221 |
+
"grad_norm": 0.7471247315406799,
|
| 1222 |
+
"learning_rate": 1.5225225225225225e-05,
|
| 1223 |
+
"loss": 0.2223,
|
| 1224 |
"step": 170
|
| 1225 |
},
|
| 1226 |
{
|
| 1227 |
"epoch": 0.5066666666666667,
|
| 1228 |
+
"grad_norm": 0.7017563581466675,
|
| 1229 |
+
"learning_rate": 1.5135135135135134e-05,
|
| 1230 |
+
"loss": 0.2089,
|
| 1231 |
"step": 171
|
| 1232 |
},
|
| 1233 |
{
|
| 1234 |
"epoch": 0.5096296296296297,
|
| 1235 |
+
"grad_norm": 0.8412315249443054,
|
| 1236 |
+
"learning_rate": 1.5045045045045047e-05,
|
| 1237 |
+
"loss": 0.2379,
|
| 1238 |
"step": 172
|
| 1239 |
},
|
| 1240 |
{
|
| 1241 |
"epoch": 0.5125925925925926,
|
| 1242 |
+
"grad_norm": 0.687402606010437,
|
| 1243 |
+
"learning_rate": 1.4954954954954954e-05,
|
| 1244 |
+
"loss": 0.2149,
|
| 1245 |
"step": 173
|
| 1246 |
},
|
| 1247 |
{
|
| 1248 |
"epoch": 0.5155555555555555,
|
| 1249 |
+
"grad_norm": 0.754208505153656,
|
| 1250 |
+
"learning_rate": 1.4864864864864865e-05,
|
| 1251 |
+
"loss": 0.1982,
|
| 1252 |
"step": 174
|
| 1253 |
},
|
| 1254 |
{
|
| 1255 |
"epoch": 0.5185185185185185,
|
| 1256 |
+
"grad_norm": 0.6880926489830017,
|
| 1257 |
+
"learning_rate": 1.4774774774774776e-05,
|
| 1258 |
+
"loss": 0.186,
|
| 1259 |
"step": 175
|
| 1260 |
},
|
| 1261 |
{
|
| 1262 |
"epoch": 0.5214814814814814,
|
| 1263 |
+
"grad_norm": 0.7988137006759644,
|
| 1264 |
+
"learning_rate": 1.4684684684684685e-05,
|
| 1265 |
+
"loss": 0.2355,
|
| 1266 |
"step": 176
|
| 1267 |
},
|
| 1268 |
{
|
| 1269 |
"epoch": 0.5244444444444445,
|
| 1270 |
+
"grad_norm": 0.5612587332725525,
|
| 1271 |
+
"learning_rate": 1.4594594594594596e-05,
|
| 1272 |
+
"loss": 0.1292,
|
| 1273 |
"step": 177
|
| 1274 |
},
|
| 1275 |
{
|
| 1276 |
"epoch": 0.5274074074074074,
|
| 1277 |
+
"grad_norm": 0.6964252591133118,
|
| 1278 |
+
"learning_rate": 1.4504504504504505e-05,
|
| 1279 |
+
"loss": 0.1765,
|
| 1280 |
"step": 178
|
| 1281 |
},
|
| 1282 |
{
|
| 1283 |
"epoch": 0.5303703703703704,
|
| 1284 |
+
"grad_norm": 0.6944383382797241,
|
| 1285 |
+
"learning_rate": 1.4414414414414414e-05,
|
| 1286 |
+
"loss": 0.1968,
|
| 1287 |
"step": 179
|
| 1288 |
},
|
| 1289 |
{
|
| 1290 |
"epoch": 0.5333333333333333,
|
| 1291 |
+
"grad_norm": 0.4856385290622711,
|
| 1292 |
+
"learning_rate": 1.4324324324324325e-05,
|
| 1293 |
+
"loss": 0.1477,
|
| 1294 |
"step": 180
|
| 1295 |
},
|
| 1296 |
{
|
| 1297 |
"epoch": 0.5362962962962963,
|
| 1298 |
+
"grad_norm": 0.6305952072143555,
|
| 1299 |
+
"learning_rate": 1.4234234234234235e-05,
|
| 1300 |
+
"loss": 0.1756,
|
| 1301 |
"step": 181
|
| 1302 |
},
|
| 1303 |
{
|
| 1304 |
"epoch": 0.5392592592592592,
|
| 1305 |
+
"grad_norm": 0.6567600965499878,
|
| 1306 |
+
"learning_rate": 1.4144144144144145e-05,
|
| 1307 |
+
"loss": 0.1663,
|
| 1308 |
"step": 182
|
| 1309 |
},
|
| 1310 |
{
|
| 1311 |
"epoch": 0.5422222222222223,
|
| 1312 |
+
"grad_norm": 0.5791081190109253,
|
| 1313 |
+
"learning_rate": 1.4054054054054054e-05,
|
| 1314 |
+
"loss": 0.1343,
|
| 1315 |
"step": 183
|
| 1316 |
},
|
| 1317 |
{
|
| 1318 |
"epoch": 0.5451851851851852,
|
| 1319 |
+
"grad_norm": 0.868181586265564,
|
| 1320 |
+
"learning_rate": 1.3963963963963964e-05,
|
| 1321 |
+
"loss": 0.2596,
|
| 1322 |
"step": 184
|
| 1323 |
},
|
| 1324 |
{
|
| 1325 |
"epoch": 0.5481481481481482,
|
| 1326 |
+
"grad_norm": 0.5308585166931152,
|
| 1327 |
+
"learning_rate": 1.3873873873873875e-05,
|
| 1328 |
+
"loss": 0.113,
|
| 1329 |
"step": 185
|
| 1330 |
},
|
| 1331 |
{
|
| 1332 |
"epoch": 0.5511111111111111,
|
| 1333 |
+
"grad_norm": 0.7182576656341553,
|
| 1334 |
+
"learning_rate": 1.3783783783783784e-05,
|
| 1335 |
+
"loss": 0.2106,
|
| 1336 |
"step": 186
|
| 1337 |
},
|
| 1338 |
{
|
| 1339 |
"epoch": 0.554074074074074,
|
| 1340 |
+
"grad_norm": 0.8075292110443115,
|
| 1341 |
+
"learning_rate": 1.3693693693693693e-05,
|
| 1342 |
+
"loss": 0.2567,
|
| 1343 |
"step": 187
|
| 1344 |
},
|
| 1345 |
{
|
| 1346 |
"epoch": 0.557037037037037,
|
| 1347 |
+
"grad_norm": 0.6146810054779053,
|
| 1348 |
+
"learning_rate": 1.3603603603603604e-05,
|
| 1349 |
+
"loss": 0.1772,
|
| 1350 |
"step": 188
|
| 1351 |
},
|
| 1352 |
{
|
| 1353 |
"epoch": 0.56,
|
| 1354 |
+
"grad_norm": 0.7381985187530518,
|
| 1355 |
+
"learning_rate": 1.3513513513513513e-05,
|
| 1356 |
+
"loss": 0.21,
|
| 1357 |
"step": 189
|
| 1358 |
},
|
| 1359 |
{
|
| 1360 |
"epoch": 0.562962962962963,
|
| 1361 |
+
"grad_norm": 0.7753574848175049,
|
| 1362 |
+
"learning_rate": 1.3423423423423424e-05,
|
| 1363 |
+
"loss": 0.2079,
|
| 1364 |
"step": 190
|
| 1365 |
},
|
| 1366 |
{
|
| 1367 |
"epoch": 0.5659259259259259,
|
| 1368 |
+
"grad_norm": 0.605355441570282,
|
| 1369 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 1370 |
+
"loss": 0.1797,
|
| 1371 |
"step": 191
|
| 1372 |
},
|
| 1373 |
{
|
| 1374 |
"epoch": 0.5688888888888889,
|
| 1375 |
+
"grad_norm": 0.5721647143363953,
|
| 1376 |
+
"learning_rate": 1.3243243243243242e-05,
|
| 1377 |
+
"loss": 0.1413,
|
| 1378 |
"step": 192
|
| 1379 |
},
|
| 1380 |
{
|
| 1381 |
"epoch": 0.5718518518518518,
|
| 1382 |
+
"grad_norm": 0.6950632333755493,
|
| 1383 |
+
"learning_rate": 1.3153153153153153e-05,
|
| 1384 |
+
"loss": 0.2411,
|
| 1385 |
"step": 193
|
| 1386 |
},
|
| 1387 |
{
|
| 1388 |
"epoch": 0.5748148148148148,
|
| 1389 |
+
"grad_norm": 0.6648383140563965,
|
| 1390 |
+
"learning_rate": 1.3063063063063064e-05,
|
| 1391 |
+
"loss": 0.2191,
|
| 1392 |
"step": 194
|
| 1393 |
},
|
| 1394 |
{
|
| 1395 |
"epoch": 0.5777777777777777,
|
| 1396 |
+
"grad_norm": 0.6028621792793274,
|
| 1397 |
+
"learning_rate": 1.2972972972972975e-05,
|
| 1398 |
+
"loss": 0.1432,
|
| 1399 |
"step": 195
|
| 1400 |
},
|
| 1401 |
{
|
| 1402 |
"epoch": 0.5807407407407408,
|
| 1403 |
+
"grad_norm": 0.5522150993347168,
|
| 1404 |
+
"learning_rate": 1.2882882882882882e-05,
|
| 1405 |
+
"loss": 0.148,
|
| 1406 |
"step": 196
|
| 1407 |
},
|
| 1408 |
{
|
| 1409 |
"epoch": 0.5837037037037037,
|
| 1410 |
+
"grad_norm": 0.6221995949745178,
|
| 1411 |
+
"learning_rate": 1.2792792792792793e-05,
|
| 1412 |
+
"loss": 0.1969,
|
| 1413 |
"step": 197
|
| 1414 |
},
|
| 1415 |
{
|
| 1416 |
"epoch": 0.5866666666666667,
|
| 1417 |
+
"grad_norm": 0.6506313681602478,
|
| 1418 |
+
"learning_rate": 1.2702702702702704e-05,
|
| 1419 |
+
"loss": 0.1873,
|
| 1420 |
"step": 198
|
| 1421 |
},
|
| 1422 |
{
|
| 1423 |
"epoch": 0.5896296296296296,
|
| 1424 |
+
"grad_norm": 0.5992417335510254,
|
| 1425 |
+
"learning_rate": 1.2612612612612613e-05,
|
| 1426 |
+
"loss": 0.1549,
|
| 1427 |
"step": 199
|
| 1428 |
},
|
| 1429 |
{
|
| 1430 |
"epoch": 0.5925925925925926,
|
| 1431 |
+
"grad_norm": 0.5438278913497925,
|
| 1432 |
+
"learning_rate": 1.2522522522522522e-05,
|
| 1433 |
+
"loss": 0.1301,
|
| 1434 |
"step": 200
|
| 1435 |
},
|
| 1436 |
{
|
| 1437 |
"epoch": 0.5925925925925926,
|
| 1438 |
+
"eval_loss": 0.15750615298748016,
|
| 1439 |
+
"eval_runtime": 14.0501,
|
| 1440 |
+
"eval_samples_per_second": 42.704,
|
| 1441 |
+
"eval_steps_per_second": 5.338,
|
| 1442 |
"step": 200
|
| 1443 |
},
|
| 1444 |
{
|
| 1445 |
"epoch": 0.5955555555555555,
|
| 1446 |
+
"grad_norm": 0.9591789841651917,
|
| 1447 |
+
"learning_rate": 1.2432432432432433e-05,
|
| 1448 |
+
"loss": 0.3016,
|
| 1449 |
"step": 201
|
| 1450 |
},
|
| 1451 |
{
|
| 1452 |
"epoch": 0.5985185185185186,
|
| 1453 |
+
"grad_norm": 0.6065961718559265,
|
| 1454 |
+
"learning_rate": 1.2342342342342342e-05,
|
| 1455 |
+
"loss": 0.1659,
|
| 1456 |
"step": 202
|
| 1457 |
},
|
| 1458 |
{
|
| 1459 |
"epoch": 0.6014814814814815,
|
| 1460 |
+
"grad_norm": 0.5144144296646118,
|
| 1461 |
+
"learning_rate": 1.2252252252252253e-05,
|
| 1462 |
+
"loss": 0.1059,
|
| 1463 |
"step": 203
|
| 1464 |
},
|
| 1465 |
{
|
| 1466 |
"epoch": 0.6044444444444445,
|
| 1467 |
+
"grad_norm": 0.8562734723091125,
|
| 1468 |
+
"learning_rate": 1.2162162162162164e-05,
|
| 1469 |
+
"loss": 0.2553,
|
| 1470 |
"step": 204
|
| 1471 |
},
|
| 1472 |
{
|
| 1473 |
"epoch": 0.6074074074074074,
|
| 1474 |
+
"grad_norm": 0.8882193565368652,
|
| 1475 |
+
"learning_rate": 1.2072072072072071e-05,
|
| 1476 |
+
"loss": 0.2379,
|
| 1477 |
"step": 205
|
| 1478 |
},
|
| 1479 |
{
|
| 1480 |
"epoch": 0.6103703703703703,
|
| 1481 |
+
"grad_norm": 0.5086420774459839,
|
| 1482 |
+
"learning_rate": 1.1981981981981982e-05,
|
| 1483 |
+
"loss": 0.1328,
|
| 1484 |
"step": 206
|
| 1485 |
},
|
| 1486 |
{
|
| 1487 |
"epoch": 0.6133333333333333,
|
| 1488 |
+
"grad_norm": 0.4841505289077759,
|
| 1489 |
+
"learning_rate": 1.1891891891891893e-05,
|
| 1490 |
+
"loss": 0.0902,
|
| 1491 |
"step": 207
|
| 1492 |
},
|
| 1493 |
{
|
| 1494 |
"epoch": 0.6162962962962963,
|
| 1495 |
+
"grad_norm": 0.5195430517196655,
|
| 1496 |
+
"learning_rate": 1.1801801801801803e-05,
|
| 1497 |
+
"loss": 0.1214,
|
| 1498 |
"step": 208
|
| 1499 |
},
|
| 1500 |
{
|
| 1501 |
"epoch": 0.6192592592592593,
|
| 1502 |
+
"grad_norm": 0.5387447476387024,
|
| 1503 |
+
"learning_rate": 1.171171171171171e-05,
|
| 1504 |
+
"loss": 0.1749,
|
| 1505 |
"step": 209
|
| 1506 |
},
|
| 1507 |
{
|
| 1508 |
"epoch": 0.6222222222222222,
|
| 1509 |
+
"grad_norm": 0.6126540899276733,
|
| 1510 |
+
"learning_rate": 1.1621621621621622e-05,
|
| 1511 |
+
"loss": 0.1853,
|
| 1512 |
"step": 210
|
| 1513 |
},
|
| 1514 |
{
|
| 1515 |
"epoch": 0.6251851851851852,
|
| 1516 |
+
"grad_norm": 0.6670091152191162,
|
| 1517 |
+
"learning_rate": 1.1531531531531532e-05,
|
| 1518 |
+
"loss": 0.2141,
|
| 1519 |
"step": 211
|
| 1520 |
},
|
| 1521 |
{
|
| 1522 |
"epoch": 0.6281481481481481,
|
| 1523 |
+
"grad_norm": 0.6592415571212769,
|
| 1524 |
+
"learning_rate": 1.1441441441441442e-05,
|
| 1525 |
+
"loss": 0.1911,
|
| 1526 |
"step": 212
|
| 1527 |
},
|
| 1528 |
{
|
| 1529 |
"epoch": 0.6311111111111111,
|
| 1530 |
+
"grad_norm": 0.6606993675231934,
|
| 1531 |
+
"learning_rate": 1.1351351351351352e-05,
|
| 1532 |
+
"loss": 0.193,
|
| 1533 |
"step": 213
|
| 1534 |
},
|
| 1535 |
{
|
| 1536 |
"epoch": 0.6340740740740741,
|
| 1537 |
+
"grad_norm": 0.5082506537437439,
|
| 1538 |
+
"learning_rate": 1.1261261261261261e-05,
|
| 1539 |
+
"loss": 0.1445,
|
| 1540 |
"step": 214
|
| 1541 |
},
|
| 1542 |
{
|
| 1543 |
"epoch": 0.6370370370370371,
|
| 1544 |
+
"grad_norm": 0.5250320434570312,
|
| 1545 |
+
"learning_rate": 1.117117117117117e-05,
|
| 1546 |
+
"loss": 0.1229,
|
| 1547 |
"step": 215
|
| 1548 |
},
|
| 1549 |
{
|
| 1550 |
"epoch": 0.64,
|
| 1551 |
+
"grad_norm": 0.573753833770752,
|
| 1552 |
+
"learning_rate": 1.1081081081081081e-05,
|
| 1553 |
+
"loss": 0.1637,
|
| 1554 |
"step": 216
|
| 1555 |
},
|
| 1556 |
{
|
| 1557 |
"epoch": 0.642962962962963,
|
| 1558 |
+
"grad_norm": 0.5497123599052429,
|
| 1559 |
+
"learning_rate": 1.0990990990990992e-05,
|
| 1560 |
+
"loss": 0.1632,
|
| 1561 |
"step": 217
|
| 1562 |
},
|
| 1563 |
{
|
| 1564 |
"epoch": 0.6459259259259259,
|
| 1565 |
+
"grad_norm": 0.5139077305793762,
|
| 1566 |
+
"learning_rate": 1.0900900900900901e-05,
|
| 1567 |
+
"loss": 0.1234,
|
| 1568 |
"step": 218
|
| 1569 |
},
|
| 1570 |
{
|
| 1571 |
"epoch": 0.6488888888888888,
|
| 1572 |
+
"grad_norm": 0.7593399286270142,
|
| 1573 |
+
"learning_rate": 1.081081081081081e-05,
|
| 1574 |
+
"loss": 0.2258,
|
| 1575 |
"step": 219
|
| 1576 |
},
|
| 1577 |
{
|
| 1578 |
"epoch": 0.6518518518518519,
|
| 1579 |
+
"grad_norm": 0.6373380422592163,
|
| 1580 |
+
"learning_rate": 1.0720720720720721e-05,
|
| 1581 |
+
"loss": 0.1533,
|
| 1582 |
"step": 220
|
| 1583 |
},
|
| 1584 |
{
|
| 1585 |
"epoch": 0.6548148148148148,
|
| 1586 |
+
"grad_norm": 0.6002466678619385,
|
| 1587 |
+
"learning_rate": 1.063063063063063e-05,
|
| 1588 |
+
"loss": 0.1754,
|
| 1589 |
"step": 221
|
| 1590 |
},
|
| 1591 |
{
|
| 1592 |
"epoch": 0.6577777777777778,
|
| 1593 |
+
"grad_norm": 0.727178156375885,
|
| 1594 |
+
"learning_rate": 1.0540540540540541e-05,
|
| 1595 |
+
"loss": 0.1875,
|
| 1596 |
"step": 222
|
| 1597 |
},
|
| 1598 |
{
|
| 1599 |
"epoch": 0.6607407407407407,
|
| 1600 |
+
"grad_norm": 0.5689584612846375,
|
| 1601 |
+
"learning_rate": 1.045045045045045e-05,
|
| 1602 |
+
"loss": 0.1414,
|
| 1603 |
"step": 223
|
| 1604 |
},
|
| 1605 |
{
|
| 1606 |
"epoch": 0.6637037037037037,
|
| 1607 |
+
"grad_norm": 0.5667452216148376,
|
| 1608 |
+
"learning_rate": 1.0360360360360361e-05,
|
| 1609 |
+
"loss": 0.107,
|
| 1610 |
"step": 224
|
| 1611 |
},
|
| 1612 |
{
|
| 1613 |
"epoch": 0.6666666666666666,
|
| 1614 |
+
"grad_norm": 0.487768292427063,
|
| 1615 |
+
"learning_rate": 1.027027027027027e-05,
|
| 1616 |
+
"loss": 0.1156,
|
| 1617 |
"step": 225
|
| 1618 |
},
|
| 1619 |
{
|
| 1620 |
"epoch": 0.6696296296296296,
|
| 1621 |
+
"grad_norm": 0.6705539226531982,
|
| 1622 |
+
"learning_rate": 1.0180180180180181e-05,
|
| 1623 |
+
"loss": 0.1421,
|
| 1624 |
"step": 226
|
| 1625 |
},
|
| 1626 |
{
|
| 1627 |
"epoch": 0.6725925925925926,
|
| 1628 |
+
"grad_norm": 0.9306380748748779,
|
| 1629 |
+
"learning_rate": 1.0090090090090092e-05,
|
| 1630 |
+
"loss": 0.272,
|
| 1631 |
"step": 227
|
| 1632 |
},
|
| 1633 |
{
|
| 1634 |
"epoch": 0.6755555555555556,
|
| 1635 |
+
"grad_norm": 0.4459589123725891,
|
| 1636 |
+
"learning_rate": 9.999999999999999e-06,
|
| 1637 |
+
"loss": 0.0993,
|
| 1638 |
"step": 228
|
| 1639 |
},
|
| 1640 |
{
|
| 1641 |
"epoch": 0.6785185185185185,
|
| 1642 |
+
"grad_norm": 0.4910501539707184,
|
| 1643 |
+
"learning_rate": 9.90990990990991e-06,
|
| 1644 |
+
"loss": 0.1039,
|
| 1645 |
"step": 229
|
| 1646 |
},
|
| 1647 |
{
|
| 1648 |
"epoch": 0.6814814814814815,
|
| 1649 |
+
"grad_norm": 0.5006682872772217,
|
| 1650 |
+
"learning_rate": 9.81981981981982e-06,
|
| 1651 |
+
"loss": 0.1261,
|
| 1652 |
"step": 230
|
| 1653 |
},
|
| 1654 |
{
|
| 1655 |
"epoch": 0.6844444444444444,
|
| 1656 |
+
"grad_norm": 0.7891579866409302,
|
| 1657 |
+
"learning_rate": 9.72972972972973e-06,
|
| 1658 |
+
"loss": 0.2181,
|
| 1659 |
"step": 231
|
| 1660 |
},
|
| 1661 |
{
|
| 1662 |
"epoch": 0.6874074074074074,
|
| 1663 |
+
"grad_norm": 0.5958547592163086,
|
| 1664 |
+
"learning_rate": 9.639639639639639e-06,
|
| 1665 |
+
"loss": 0.1632,
|
| 1666 |
"step": 232
|
| 1667 |
},
|
| 1668 |
{
|
| 1669 |
"epoch": 0.6903703703703704,
|
| 1670 |
+
"grad_norm": 0.5740420818328857,
|
| 1671 |
+
"learning_rate": 9.54954954954955e-06,
|
| 1672 |
+
"loss": 0.141,
|
| 1673 |
"step": 233
|
| 1674 |
},
|
| 1675 |
{
|
| 1676 |
"epoch": 0.6933333333333334,
|
| 1677 |
+
"grad_norm": 0.5734637975692749,
|
| 1678 |
+
"learning_rate": 9.459459459459459e-06,
|
| 1679 |
+
"loss": 0.1005,
|
| 1680 |
"step": 234
|
| 1681 |
},
|
| 1682 |
{
|
| 1683 |
"epoch": 0.6962962962962963,
|
| 1684 |
+
"grad_norm": 0.8963714241981506,
|
| 1685 |
+
"learning_rate": 9.36936936936937e-06,
|
| 1686 |
+
"loss": 0.1689,
|
| 1687 |
"step": 235
|
| 1688 |
},
|
| 1689 |
{
|
| 1690 |
"epoch": 0.6992592592592592,
|
| 1691 |
+
"grad_norm": 0.7850510478019714,
|
| 1692 |
+
"learning_rate": 9.27927927927928e-06,
|
| 1693 |
+
"loss": 0.1931,
|
| 1694 |
"step": 236
|
| 1695 |
},
|
| 1696 |
{
|
| 1697 |
"epoch": 0.7022222222222222,
|
| 1698 |
+
"grad_norm": 0.6909759044647217,
|
| 1699 |
+
"learning_rate": 9.18918918918919e-06,
|
| 1700 |
+
"loss": 0.1874,
|
| 1701 |
"step": 237
|
| 1702 |
},
|
| 1703 |
{
|
| 1704 |
"epoch": 0.7051851851851851,
|
| 1705 |
+
"grad_norm": 0.6632979512214661,
|
| 1706 |
+
"learning_rate": 9.099099099099099e-06,
|
| 1707 |
+
"loss": 0.1576,
|
| 1708 |
"step": 238
|
| 1709 |
},
|
| 1710 |
{
|
| 1711 |
"epoch": 0.7081481481481482,
|
| 1712 |
+
"grad_norm": 0.5669253468513489,
|
| 1713 |
+
"learning_rate": 9.00900900900901e-06,
|
| 1714 |
+
"loss": 0.1293,
|
| 1715 |
"step": 239
|
| 1716 |
},
|
| 1717 |
{
|
| 1718 |
"epoch": 0.7111111111111111,
|
| 1719 |
+
"grad_norm": 0.807396650314331,
|
| 1720 |
+
"learning_rate": 8.91891891891892e-06,
|
| 1721 |
+
"loss": 0.2094,
|
| 1722 |
"step": 240
|
| 1723 |
},
|
| 1724 |
{
|
| 1725 |
"epoch": 0.7140740740740741,
|
| 1726 |
+
"grad_norm": 0.739194929599762,
|
| 1727 |
+
"learning_rate": 8.828828828828828e-06,
|
| 1728 |
+
"loss": 0.1901,
|
| 1729 |
"step": 241
|
| 1730 |
},
|
| 1731 |
{
|
| 1732 |
"epoch": 0.717037037037037,
|
| 1733 |
+
"grad_norm": 0.6328135132789612,
|
| 1734 |
+
"learning_rate": 8.738738738738739e-06,
|
| 1735 |
+
"loss": 0.1622,
|
| 1736 |
"step": 242
|
| 1737 |
},
|
| 1738 |
{
|
| 1739 |
"epoch": 0.72,
|
| 1740 |
+
"grad_norm": 0.7799888849258423,
|
| 1741 |
+
"learning_rate": 8.64864864864865e-06,
|
| 1742 |
+
"loss": 0.246,
|
| 1743 |
"step": 243
|
| 1744 |
},
|
| 1745 |
{
|
| 1746 |
"epoch": 0.7229629629629629,
|
| 1747 |
+
"grad_norm": 0.6167428493499756,
|
| 1748 |
+
"learning_rate": 8.558558558558558e-06,
|
| 1749 |
+
"loss": 0.1321,
|
| 1750 |
"step": 244
|
| 1751 |
},
|
| 1752 |
{
|
| 1753 |
"epoch": 0.725925925925926,
|
| 1754 |
+
"grad_norm": 0.5111340880393982,
|
| 1755 |
+
"learning_rate": 8.46846846846847e-06,
|
| 1756 |
+
"loss": 0.1061,
|
| 1757 |
"step": 245
|
| 1758 |
},
|
| 1759 |
{
|
| 1760 |
"epoch": 0.7288888888888889,
|
| 1761 |
+
"grad_norm": 0.714773952960968,
|
| 1762 |
+
"learning_rate": 8.378378378378378e-06,
|
| 1763 |
+
"loss": 0.2025,
|
| 1764 |
"step": 246
|
| 1765 |
},
|
| 1766 |
{
|
| 1767 |
"epoch": 0.7318518518518519,
|
| 1768 |
+
"grad_norm": 0.5915043950080872,
|
| 1769 |
+
"learning_rate": 8.288288288288287e-06,
|
| 1770 |
+
"loss": 0.1148,
|
| 1771 |
"step": 247
|
| 1772 |
},
|
| 1773 |
{
|
| 1774 |
"epoch": 0.7348148148148148,
|
| 1775 |
+
"grad_norm": 0.6125442981719971,
|
| 1776 |
+
"learning_rate": 8.198198198198198e-06,
|
| 1777 |
+
"loss": 0.1343,
|
| 1778 |
"step": 248
|
| 1779 |
},
|
| 1780 |
{
|
| 1781 |
"epoch": 0.7377777777777778,
|
| 1782 |
+
"grad_norm": 0.8247624635696411,
|
| 1783 |
+
"learning_rate": 8.108108108108109e-06,
|
| 1784 |
+
"loss": 0.2321,
|
| 1785 |
"step": 249
|
| 1786 |
},
|
| 1787 |
{
|
| 1788 |
"epoch": 0.7407407407407407,
|
| 1789 |
+
"grad_norm": 0.6755908131599426,
|
| 1790 |
+
"learning_rate": 8.01801801801802e-06,
|
| 1791 |
+
"loss": 0.1827,
|
| 1792 |
"step": 250
|
| 1793 |
},
|
| 1794 |
{
|
| 1795 |
"epoch": 0.7407407407407407,
|
| 1796 |
+
"eval_loss": 0.15461350977420807,
|
| 1797 |
+
"eval_runtime": 14.1271,
|
| 1798 |
+
"eval_samples_per_second": 42.472,
|
| 1799 |
+
"eval_steps_per_second": 5.309,
|
| 1800 |
"step": 250
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 0.7437037037037038,
|
| 1804 |
+
"grad_norm": 0.63921058177948,
|
| 1805 |
+
"learning_rate": 7.927927927927927e-06,
|
| 1806 |
+
"loss": 0.1599,
|
| 1807 |
"step": 251
|
| 1808 |
},
|
| 1809 |
{
|
| 1810 |
"epoch": 0.7466666666666667,
|
| 1811 |
+
"grad_norm": 0.547126054763794,
|
| 1812 |
+
"learning_rate": 7.837837837837838e-06,
|
| 1813 |
+
"loss": 0.142,
|
| 1814 |
"step": 252
|
| 1815 |
},
|
| 1816 |
{
|
| 1817 |
"epoch": 0.7496296296296296,
|
| 1818 |
+
"grad_norm": 0.7663922905921936,
|
| 1819 |
+
"learning_rate": 7.747747747747749e-06,
|
| 1820 |
+
"loss": 0.1981,
|
| 1821 |
"step": 253
|
| 1822 |
},
|
| 1823 |
{
|
| 1824 |
"epoch": 0.7525925925925926,
|
| 1825 |
+
"grad_norm": 0.6325225234031677,
|
| 1826 |
+
"learning_rate": 7.657657657657658e-06,
|
| 1827 |
"loss": 0.1413,
|
| 1828 |
"step": 254
|
| 1829 |
},
|
| 1830 |
{
|
| 1831 |
"epoch": 0.7555555555555555,
|
| 1832 |
+
"grad_norm": 0.7257774472236633,
|
| 1833 |
+
"learning_rate": 7.567567567567567e-06,
|
| 1834 |
+
"loss": 0.2066,
|
| 1835 |
"step": 255
|
| 1836 |
},
|
| 1837 |
{
|
| 1838 |
"epoch": 0.7585185185185185,
|
| 1839 |
+
"grad_norm": 0.6610811352729797,
|
| 1840 |
+
"learning_rate": 7.477477477477477e-06,
|
| 1841 |
+
"loss": 0.1755,
|
| 1842 |
"step": 256
|
| 1843 |
},
|
| 1844 |
{
|
| 1845 |
"epoch": 0.7614814814814815,
|
| 1846 |
+
"grad_norm": 0.6078842282295227,
|
| 1847 |
+
"learning_rate": 7.387387387387388e-06,
|
| 1848 |
+
"loss": 0.2016,
|
| 1849 |
"step": 257
|
| 1850 |
},
|
| 1851 |
{
|
| 1852 |
"epoch": 0.7644444444444445,
|
| 1853 |
+
"grad_norm": 0.6328191161155701,
|
| 1854 |
+
"learning_rate": 7.297297297297298e-06,
|
| 1855 |
+
"loss": 0.1884,
|
| 1856 |
"step": 258
|
| 1857 |
},
|
| 1858 |
{
|
| 1859 |
"epoch": 0.7674074074074074,
|
| 1860 |
+
"grad_norm": 0.7390186190605164,
|
| 1861 |
+
"learning_rate": 7.207207207207207e-06,
|
| 1862 |
+
"loss": 0.1898,
|
| 1863 |
"step": 259
|
| 1864 |
},
|
| 1865 |
{
|
| 1866 |
"epoch": 0.7703703703703704,
|
| 1867 |
+
"grad_norm": 0.6354762315750122,
|
| 1868 |
+
"learning_rate": 7.117117117117118e-06,
|
| 1869 |
+
"loss": 0.1649,
|
| 1870 |
"step": 260
|
| 1871 |
},
|
| 1872 |
{
|
| 1873 |
"epoch": 0.7733333333333333,
|
| 1874 |
+
"grad_norm": 0.8199350237846375,
|
| 1875 |
+
"learning_rate": 7.027027027027027e-06,
|
| 1876 |
+
"loss": 0.2304,
|
| 1877 |
"step": 261
|
| 1878 |
},
|
| 1879 |
{
|
| 1880 |
"epoch": 0.7762962962962963,
|
| 1881 |
+
"grad_norm": 0.8470929861068726,
|
| 1882 |
+
"learning_rate": 6.936936936936938e-06,
|
| 1883 |
+
"loss": 0.2142,
|
| 1884 |
"step": 262
|
| 1885 |
},
|
| 1886 |
{
|
| 1887 |
"epoch": 0.7792592592592592,
|
| 1888 |
+
"grad_norm": 0.8667031526565552,
|
| 1889 |
+
"learning_rate": 6.846846846846847e-06,
|
| 1890 |
+
"loss": 0.248,
|
| 1891 |
"step": 263
|
| 1892 |
},
|
| 1893 |
{
|
| 1894 |
"epoch": 0.7822222222222223,
|
| 1895 |
+
"grad_norm": 0.5739657282829285,
|
| 1896 |
+
"learning_rate": 6.756756756756757e-06,
|
| 1897 |
+
"loss": 0.1179,
|
| 1898 |
"step": 264
|
| 1899 |
},
|
| 1900 |
{
|
| 1901 |
"epoch": 0.7851851851851852,
|
| 1902 |
+
"grad_norm": 0.5817047953605652,
|
| 1903 |
+
"learning_rate": 6.666666666666667e-06,
|
| 1904 |
+
"loss": 0.1191,
|
| 1905 |
"step": 265
|
| 1906 |
},
|
| 1907 |
{
|
| 1908 |
"epoch": 0.7881481481481482,
|
| 1909 |
+
"grad_norm": 0.5076894164085388,
|
| 1910 |
+
"learning_rate": 6.576576576576577e-06,
|
| 1911 |
+
"loss": 0.1247,
|
| 1912 |
"step": 266
|
| 1913 |
},
|
| 1914 |
{
|
| 1915 |
"epoch": 0.7911111111111111,
|
| 1916 |
+
"grad_norm": 0.6611981987953186,
|
| 1917 |
+
"learning_rate": 6.486486486486487e-06,
|
| 1918 |
+
"loss": 0.1937,
|
| 1919 |
"step": 267
|
| 1920 |
},
|
| 1921 |
{
|
| 1922 |
"epoch": 0.794074074074074,
|
| 1923 |
+
"grad_norm": 0.7290340065956116,
|
| 1924 |
+
"learning_rate": 6.3963963963963965e-06,
|
| 1925 |
+
"loss": 0.1726,
|
| 1926 |
"step": 268
|
| 1927 |
},
|
| 1928 |
{
|
| 1929 |
"epoch": 0.797037037037037,
|
| 1930 |
+
"grad_norm": 0.7665138244628906,
|
| 1931 |
+
"learning_rate": 6.3063063063063065e-06,
|
| 1932 |
+
"loss": 0.1934,
|
| 1933 |
"step": 269
|
| 1934 |
},
|
| 1935 |
{
|
| 1936 |
"epoch": 0.8,
|
| 1937 |
+
"grad_norm": 0.8837843537330627,
|
| 1938 |
+
"learning_rate": 6.2162162162162164e-06,
|
| 1939 |
+
"loss": 0.2664,
|
| 1940 |
"step": 270
|
| 1941 |
},
|
| 1942 |
{
|
| 1943 |
"epoch": 0.802962962962963,
|
| 1944 |
+
"grad_norm": 0.6573258638381958,
|
| 1945 |
+
"learning_rate": 6.126126126126126e-06,
|
| 1946 |
+
"loss": 0.1608,
|
| 1947 |
"step": 271
|
| 1948 |
},
|
| 1949 |
{
|
| 1950 |
"epoch": 0.8059259259259259,
|
| 1951 |
+
"grad_norm": 0.8149768710136414,
|
| 1952 |
+
"learning_rate": 6.0360360360360355e-06,
|
| 1953 |
+
"loss": 0.2045,
|
| 1954 |
"step": 272
|
| 1955 |
},
|
| 1956 |
{
|
| 1957 |
"epoch": 0.8088888888888889,
|
| 1958 |
+
"grad_norm": 0.8355318307876587,
|
| 1959 |
+
"learning_rate": 5.945945945945946e-06,
|
| 1960 |
+
"loss": 0.2532,
|
| 1961 |
"step": 273
|
| 1962 |
},
|
| 1963 |
{
|
| 1964 |
"epoch": 0.8118518518518518,
|
| 1965 |
+
"grad_norm": 0.45708897709846497,
|
| 1966 |
+
"learning_rate": 5.855855855855855e-06,
|
| 1967 |
+
"loss": 0.0903,
|
| 1968 |
"step": 274
|
| 1969 |
},
|
| 1970 |
{
|
| 1971 |
"epoch": 0.8148148148148148,
|
| 1972 |
+
"grad_norm": 0.7013540267944336,
|
| 1973 |
+
"learning_rate": 5.765765765765766e-06,
|
| 1974 |
+
"loss": 0.2129,
|
| 1975 |
"step": 275
|
| 1976 |
},
|
| 1977 |
{
|
| 1978 |
"epoch": 0.8177777777777778,
|
| 1979 |
+
"grad_norm": 0.8260526061058044,
|
| 1980 |
+
"learning_rate": 5.675675675675676e-06,
|
| 1981 |
+
"loss": 0.2359,
|
| 1982 |
"step": 276
|
| 1983 |
},
|
| 1984 |
{
|
| 1985 |
"epoch": 0.8207407407407408,
|
| 1986 |
+
"grad_norm": 0.6625615954399109,
|
| 1987 |
+
"learning_rate": 5.585585585585585e-06,
|
| 1988 |
+
"loss": 0.1719,
|
| 1989 |
"step": 277
|
| 1990 |
},
|
| 1991 |
{
|
| 1992 |
"epoch": 0.8237037037037037,
|
| 1993 |
+
"grad_norm": 0.48971474170684814,
|
| 1994 |
+
"learning_rate": 5.495495495495496e-06,
|
| 1995 |
+
"loss": 0.0954,
|
| 1996 |
"step": 278
|
| 1997 |
},
|
| 1998 |
{
|
| 1999 |
"epoch": 0.8266666666666667,
|
| 2000 |
+
"grad_norm": 0.6506958603858948,
|
| 2001 |
+
"learning_rate": 5.405405405405405e-06,
|
| 2002 |
+
"loss": 0.1524,
|
| 2003 |
"step": 279
|
| 2004 |
},
|
| 2005 |
{
|
| 2006 |
"epoch": 0.8296296296296296,
|
| 2007 |
+
"grad_norm": 0.8318085074424744,
|
| 2008 |
+
"learning_rate": 5.315315315315315e-06,
|
| 2009 |
+
"loss": 0.2372,
|
| 2010 |
"step": 280
|
| 2011 |
},
|
| 2012 |
{
|
| 2013 |
"epoch": 0.8325925925925926,
|
| 2014 |
+
"grad_norm": 0.5079582929611206,
|
| 2015 |
+
"learning_rate": 5.225225225225225e-06,
|
| 2016 |
+
"loss": 0.1051,
|
| 2017 |
"step": 281
|
| 2018 |
},
|
| 2019 |
{
|
| 2020 |
"epoch": 0.8355555555555556,
|
| 2021 |
+
"grad_norm": 0.7515102624893188,
|
| 2022 |
+
"learning_rate": 5.135135135135135e-06,
|
| 2023 |
+
"loss": 0.2023,
|
| 2024 |
"step": 282
|
| 2025 |
},
|
| 2026 |
{
|
| 2027 |
"epoch": 0.8385185185185186,
|
| 2028 |
+
"grad_norm": 0.5988075733184814,
|
| 2029 |
+
"learning_rate": 5.045045045045046e-06,
|
| 2030 |
+
"loss": 0.1527,
|
| 2031 |
"step": 283
|
| 2032 |
},
|
| 2033 |
{
|
| 2034 |
"epoch": 0.8414814814814815,
|
| 2035 |
+
"grad_norm": 0.6875013113021851,
|
| 2036 |
+
"learning_rate": 4.954954954954955e-06,
|
| 2037 |
+
"loss": 0.1658,
|
| 2038 |
"step": 284
|
| 2039 |
},
|
| 2040 |
{
|
| 2041 |
"epoch": 0.8444444444444444,
|
| 2042 |
+
"grad_norm": 0.6339918375015259,
|
| 2043 |
+
"learning_rate": 4.864864864864865e-06,
|
| 2044 |
+
"loss": 0.1247,
|
| 2045 |
"step": 285
|
| 2046 |
},
|
| 2047 |
{
|
| 2048 |
"epoch": 0.8474074074074074,
|
| 2049 |
+
"grad_norm": 0.5002294778823853,
|
| 2050 |
+
"learning_rate": 4.774774774774775e-06,
|
| 2051 |
+
"loss": 0.1244,
|
| 2052 |
"step": 286
|
| 2053 |
},
|
| 2054 |
{
|
| 2055 |
"epoch": 0.8503703703703703,
|
| 2056 |
+
"grad_norm": 0.6315642595291138,
|
| 2057 |
+
"learning_rate": 4.684684684684685e-06,
|
| 2058 |
"loss": 0.1248,
|
| 2059 |
"step": 287
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
"epoch": 0.8533333333333334,
|
| 2063 |
+
"grad_norm": 0.5210931897163391,
|
| 2064 |
+
"learning_rate": 4.594594594594595e-06,
|
| 2065 |
+
"loss": 0.112,
|
| 2066 |
"step": 288
|
| 2067 |
},
|
| 2068 |
{
|
| 2069 |
"epoch": 0.8562962962962963,
|
| 2070 |
+
"grad_norm": 0.6035313606262207,
|
| 2071 |
+
"learning_rate": 4.504504504504505e-06,
|
| 2072 |
+
"loss": 0.164,
|
| 2073 |
"step": 289
|
| 2074 |
},
|
| 2075 |
{
|
| 2076 |
"epoch": 0.8592592592592593,
|
| 2077 |
+
"grad_norm": 0.5063045620918274,
|
| 2078 |
+
"learning_rate": 4.414414414414414e-06,
|
| 2079 |
+
"loss": 0.0946,
|
| 2080 |
"step": 290
|
| 2081 |
},
|
| 2082 |
{
|
| 2083 |
"epoch": 0.8622222222222222,
|
| 2084 |
+
"grad_norm": 0.5500133633613586,
|
| 2085 |
+
"learning_rate": 4.324324324324325e-06,
|
| 2086 |
+
"loss": 0.1246,
|
| 2087 |
"step": 291
|
| 2088 |
},
|
| 2089 |
{
|
| 2090 |
"epoch": 0.8651851851851852,
|
| 2091 |
+
"grad_norm": 0.5283143520355225,
|
| 2092 |
+
"learning_rate": 4.234234234234235e-06,
|
| 2093 |
"loss": 0.1234,
|
| 2094 |
"step": 292
|
| 2095 |
},
|
| 2096 |
{
|
| 2097 |
"epoch": 0.8681481481481481,
|
| 2098 |
+
"grad_norm": 0.5906702280044556,
|
| 2099 |
+
"learning_rate": 4.144144144144144e-06,
|
| 2100 |
+
"loss": 0.1139,
|
| 2101 |
"step": 293
|
| 2102 |
},
|
| 2103 |
{
|
| 2104 |
"epoch": 0.8711111111111111,
|
| 2105 |
+
"grad_norm": 0.7998249530792236,
|
| 2106 |
+
"learning_rate": 4.0540540540540545e-06,
|
| 2107 |
+
"loss": 0.2293,
|
| 2108 |
"step": 294
|
| 2109 |
},
|
| 2110 |
{
|
| 2111 |
"epoch": 0.8740740740740741,
|
| 2112 |
+
"grad_norm": 0.477461040019989,
|
| 2113 |
+
"learning_rate": 3.963963963963964e-06,
|
| 2114 |
+
"loss": 0.1225,
|
| 2115 |
"step": 295
|
| 2116 |
},
|
| 2117 |
{
|
| 2118 |
"epoch": 0.8770370370370371,
|
| 2119 |
+
"grad_norm": 0.718265175819397,
|
| 2120 |
+
"learning_rate": 3.8738738738738744e-06,
|
| 2121 |
+
"loss": 0.1591,
|
| 2122 |
"step": 296
|
| 2123 |
},
|
| 2124 |
{
|
| 2125 |
"epoch": 0.88,
|
| 2126 |
+
"grad_norm": 0.5837140679359436,
|
| 2127 |
+
"learning_rate": 3.7837837837837835e-06,
|
| 2128 |
+
"loss": 0.1375,
|
| 2129 |
"step": 297
|
| 2130 |
},
|
| 2131 |
{
|
| 2132 |
"epoch": 0.882962962962963,
|
| 2133 |
+
"grad_norm": 0.6149828433990479,
|
| 2134 |
+
"learning_rate": 3.693693693693694e-06,
|
| 2135 |
+
"loss": 0.1791,
|
| 2136 |
"step": 298
|
| 2137 |
},
|
| 2138 |
{
|
| 2139 |
"epoch": 0.8859259259259259,
|
| 2140 |
+
"grad_norm": 0.5193979740142822,
|
| 2141 |
+
"learning_rate": 3.6036036036036035e-06,
|
| 2142 |
+
"loss": 0.1331,
|
| 2143 |
"step": 299
|
| 2144 |
},
|
| 2145 |
{
|
| 2146 |
"epoch": 0.8888888888888888,
|
| 2147 |
+
"grad_norm": 0.5128735899925232,
|
| 2148 |
+
"learning_rate": 3.5135135135135134e-06,
|
| 2149 |
+
"loss": 0.1142,
|
| 2150 |
"step": 300
|
| 2151 |
},
|
| 2152 |
{
|
| 2153 |
"epoch": 0.8888888888888888,
|
| 2154 |
+
"eval_loss": 0.1521201729774475,
|
| 2155 |
+
"eval_runtime": 14.109,
|
| 2156 |
+
"eval_samples_per_second": 42.526,
|
| 2157 |
+
"eval_steps_per_second": 5.316,
|
| 2158 |
"step": 300
|
| 2159 |
},
|
| 2160 |
{
|
| 2161 |
"epoch": 0.8918518518518519,
|
| 2162 |
+
"grad_norm": 0.6757302284240723,
|
| 2163 |
+
"learning_rate": 3.4234234234234234e-06,
|
| 2164 |
+
"loss": 0.1897,
|
| 2165 |
"step": 301
|
| 2166 |
},
|
| 2167 |
{
|
| 2168 |
"epoch": 0.8948148148148148,
|
| 2169 |
+
"grad_norm": 0.7718675136566162,
|
| 2170 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 2171 |
+
"loss": 0.1665,
|
| 2172 |
"step": 302
|
| 2173 |
},
|
| 2174 |
{
|
| 2175 |
"epoch": 0.8977777777777778,
|
| 2176 |
+
"grad_norm": 0.482134073972702,
|
| 2177 |
+
"learning_rate": 3.2432432432432437e-06,
|
| 2178 |
+
"loss": 0.0866,
|
| 2179 |
"step": 303
|
| 2180 |
},
|
| 2181 |
{
|
| 2182 |
"epoch": 0.9007407407407407,
|
| 2183 |
+
"grad_norm": 0.6804472804069519,
|
| 2184 |
+
"learning_rate": 3.1531531531531532e-06,
|
| 2185 |
+
"loss": 0.208,
|
| 2186 |
"step": 304
|
| 2187 |
},
|
| 2188 |
{
|
| 2189 |
"epoch": 0.9037037037037037,
|
| 2190 |
+
"grad_norm": 0.7419773936271667,
|
| 2191 |
+
"learning_rate": 3.063063063063063e-06,
|
| 2192 |
+
"loss": 0.1108,
|
| 2193 |
"step": 305
|
| 2194 |
},
|
| 2195 |
{
|
| 2196 |
"epoch": 0.9066666666666666,
|
| 2197 |
+
"grad_norm": 0.689755380153656,
|
| 2198 |
+
"learning_rate": 2.972972972972973e-06,
|
| 2199 |
+
"loss": 0.1296,
|
| 2200 |
"step": 306
|
| 2201 |
},
|
| 2202 |
{
|
| 2203 |
"epoch": 0.9096296296296297,
|
| 2204 |
+
"grad_norm": 0.49612197279930115,
|
| 2205 |
+
"learning_rate": 2.882882882882883e-06,
|
| 2206 |
+
"loss": 0.1145,
|
| 2207 |
"step": 307
|
| 2208 |
},
|
| 2209 |
{
|
| 2210 |
"epoch": 0.9125925925925926,
|
| 2211 |
+
"grad_norm": 0.73726487159729,
|
| 2212 |
+
"learning_rate": 2.7927927927927926e-06,
|
| 2213 |
+
"loss": 0.2119,
|
| 2214 |
"step": 308
|
| 2215 |
},
|
| 2216 |
{
|
| 2217 |
"epoch": 0.9155555555555556,
|
| 2218 |
+
"grad_norm": 0.7747570276260376,
|
| 2219 |
+
"learning_rate": 2.7027027027027026e-06,
|
| 2220 |
+
"loss": 0.2202,
|
| 2221 |
"step": 309
|
| 2222 |
},
|
| 2223 |
{
|
| 2224 |
"epoch": 0.9185185185185185,
|
| 2225 |
+
"grad_norm": 0.7008584141731262,
|
| 2226 |
+
"learning_rate": 2.6126126126126125e-06,
|
| 2227 |
+
"loss": 0.1677,
|
| 2228 |
"step": 310
|
| 2229 |
},
|
| 2230 |
{
|
| 2231 |
"epoch": 0.9214814814814815,
|
| 2232 |
+
"grad_norm": 0.4077196717262268,
|
| 2233 |
+
"learning_rate": 2.522522522522523e-06,
|
| 2234 |
+
"loss": 0.0912,
|
| 2235 |
"step": 311
|
| 2236 |
},
|
| 2237 |
{
|
| 2238 |
"epoch": 0.9244444444444444,
|
| 2239 |
+
"grad_norm": 0.5494208931922913,
|
| 2240 |
+
"learning_rate": 2.4324324324324325e-06,
|
| 2241 |
+
"loss": 0.1325,
|
| 2242 |
"step": 312
|
| 2243 |
},
|
| 2244 |
{
|
| 2245 |
"epoch": 0.9274074074074075,
|
| 2246 |
+
"grad_norm": 0.7233456373214722,
|
| 2247 |
+
"learning_rate": 2.3423423423423424e-06,
|
| 2248 |
+
"loss": 0.1924,
|
| 2249 |
"step": 313
|
| 2250 |
},
|
| 2251 |
{
|
| 2252 |
"epoch": 0.9303703703703704,
|
| 2253 |
+
"grad_norm": 0.5116931200027466,
|
| 2254 |
+
"learning_rate": 2.2522522522522524e-06,
|
| 2255 |
+
"loss": 0.0999,
|
| 2256 |
"step": 314
|
| 2257 |
},
|
| 2258 |
{
|
| 2259 |
"epoch": 0.9333333333333333,
|
| 2260 |
+
"grad_norm": 0.6791926026344299,
|
| 2261 |
+
"learning_rate": 2.1621621621621623e-06,
|
| 2262 |
+
"loss": 0.1724,
|
| 2263 |
"step": 315
|
| 2264 |
},
|
| 2265 |
{
|
| 2266 |
"epoch": 0.9362962962962963,
|
| 2267 |
+
"grad_norm": 0.6117425560951233,
|
| 2268 |
+
"learning_rate": 2.072072072072072e-06,
|
| 2269 |
+
"loss": 0.1936,
|
| 2270 |
"step": 316
|
| 2271 |
},
|
| 2272 |
{
|
| 2273 |
"epoch": 0.9392592592592592,
|
| 2274 |
+
"grad_norm": 0.6541153192520142,
|
| 2275 |
+
"learning_rate": 1.981981981981982e-06,
|
| 2276 |
+
"loss": 0.1502,
|
| 2277 |
"step": 317
|
| 2278 |
},
|
| 2279 |
{
|
| 2280 |
"epoch": 0.9422222222222222,
|
| 2281 |
+
"grad_norm": 0.6360153555870056,
|
| 2282 |
+
"learning_rate": 1.8918918918918918e-06,
|
| 2283 |
+
"loss": 0.1556,
|
| 2284 |
"step": 318
|
| 2285 |
},
|
| 2286 |
{
|
| 2287 |
"epoch": 0.9451851851851852,
|
| 2288 |
+
"grad_norm": 0.5484958291053772,
|
| 2289 |
+
"learning_rate": 1.8018018018018017e-06,
|
| 2290 |
+
"loss": 0.1567,
|
| 2291 |
"step": 319
|
| 2292 |
},
|
| 2293 |
{
|
| 2294 |
"epoch": 0.9481481481481482,
|
| 2295 |
+
"grad_norm": 0.6073195934295654,
|
| 2296 |
+
"learning_rate": 1.7117117117117117e-06,
|
| 2297 |
+
"loss": 0.1388,
|
| 2298 |
"step": 320
|
| 2299 |
},
|
| 2300 |
{
|
| 2301 |
"epoch": 0.9511111111111111,
|
| 2302 |
+
"grad_norm": 0.6052750945091248,
|
| 2303 |
+
"learning_rate": 1.6216216216216219e-06,
|
| 2304 |
+
"loss": 0.1427,
|
| 2305 |
"step": 321
|
| 2306 |
},
|
| 2307 |
{
|
| 2308 |
"epoch": 0.9540740740740741,
|
| 2309 |
+
"grad_norm": 0.5835985541343689,
|
| 2310 |
+
"learning_rate": 1.5315315315315316e-06,
|
| 2311 |
+
"loss": 0.1611,
|
| 2312 |
"step": 322
|
| 2313 |
},
|
| 2314 |
{
|
| 2315 |
"epoch": 0.957037037037037,
|
| 2316 |
+
"grad_norm": 0.5069778561592102,
|
| 2317 |
+
"learning_rate": 1.4414414414414416e-06,
|
| 2318 |
+
"loss": 0.1398,
|
| 2319 |
"step": 323
|
| 2320 |
},
|
| 2321 |
{
|
| 2322 |
"epoch": 0.96,
|
| 2323 |
+
"grad_norm": 0.5789016485214233,
|
| 2324 |
+
"learning_rate": 1.3513513513513513e-06,
|
| 2325 |
+
"loss": 0.1085,
|
| 2326 |
"step": 324
|
| 2327 |
},
|
| 2328 |
{
|
| 2329 |
"epoch": 0.9629629629629629,
|
| 2330 |
+
"grad_norm": 0.4746359586715698,
|
| 2331 |
+
"learning_rate": 1.2612612612612615e-06,
|
| 2332 |
+
"loss": 0.109,
|
| 2333 |
"step": 325
|
| 2334 |
},
|
| 2335 |
{
|
| 2336 |
"epoch": 0.965925925925926,
|
| 2337 |
+
"grad_norm": 0.7349570989608765,
|
| 2338 |
+
"learning_rate": 1.1711711711711712e-06,
|
| 2339 |
+
"loss": 0.2082,
|
| 2340 |
"step": 326
|
| 2341 |
},
|
| 2342 |
{
|
| 2343 |
"epoch": 0.9688888888888889,
|
| 2344 |
+
"grad_norm": 0.4968010187149048,
|
| 2345 |
+
"learning_rate": 1.0810810810810812e-06,
|
| 2346 |
+
"loss": 0.1097,
|
| 2347 |
"step": 327
|
| 2348 |
},
|
| 2349 |
{
|
| 2350 |
"epoch": 0.9718518518518519,
|
| 2351 |
+
"grad_norm": 0.6664372086524963,
|
| 2352 |
+
"learning_rate": 9.90990990990991e-07,
|
| 2353 |
+
"loss": 0.1765,
|
| 2354 |
"step": 328
|
| 2355 |
},
|
| 2356 |
{
|
| 2357 |
"epoch": 0.9748148148148148,
|
| 2358 |
+
"grad_norm": 0.7878815531730652,
|
| 2359 |
+
"learning_rate": 9.009009009009009e-07,
|
| 2360 |
+
"loss": 0.1887,
|
| 2361 |
"step": 329
|
| 2362 |
},
|
| 2363 |
{
|
| 2364 |
"epoch": 0.9777777777777777,
|
| 2365 |
+
"grad_norm": 0.6833536028862,
|
| 2366 |
+
"learning_rate": 8.108108108108109e-07,
|
| 2367 |
+
"loss": 0.1135,
|
| 2368 |
"step": 330
|
| 2369 |
},
|
| 2370 |
{
|
| 2371 |
"epoch": 0.9807407407407407,
|
| 2372 |
+
"grad_norm": 0.7881770730018616,
|
| 2373 |
+
"learning_rate": 7.207207207207208e-07,
|
| 2374 |
+
"loss": 0.2267,
|
| 2375 |
"step": 331
|
| 2376 |
},
|
| 2377 |
{
|
| 2378 |
"epoch": 0.9837037037037037,
|
| 2379 |
+
"grad_norm": 0.5758517384529114,
|
| 2380 |
+
"learning_rate": 6.306306306306307e-07,
|
| 2381 |
+
"loss": 0.163,
|
| 2382 |
"step": 332
|
| 2383 |
},
|
| 2384 |
{
|
| 2385 |
"epoch": 0.9866666666666667,
|
| 2386 |
+
"grad_norm": 0.5272687077522278,
|
| 2387 |
+
"learning_rate": 5.405405405405406e-07,
|
| 2388 |
+
"loss": 0.123,
|
| 2389 |
"step": 333
|
| 2390 |
},
|
| 2391 |
{
|
| 2392 |
"epoch": 0.9896296296296296,
|
| 2393 |
+
"grad_norm": 0.5003986358642578,
|
| 2394 |
+
"learning_rate": 4.5045045045045043e-07,
|
| 2395 |
+
"loss": 0.1161,
|
| 2396 |
"step": 334
|
| 2397 |
},
|
| 2398 |
{
|
| 2399 |
"epoch": 0.9925925925925926,
|
| 2400 |
+
"grad_norm": 0.6778259873390198,
|
| 2401 |
+
"learning_rate": 3.603603603603604e-07,
|
| 2402 |
+
"loss": 0.1845,
|
| 2403 |
"step": 335
|
| 2404 |
},
|
| 2405 |
{
|
| 2406 |
"epoch": 0.9955555555555555,
|
| 2407 |
+
"grad_norm": 0.7316866517066956,
|
| 2408 |
+
"learning_rate": 2.702702702702703e-07,
|
| 2409 |
+
"loss": 0.201,
|
| 2410 |
"step": 336
|
| 2411 |
},
|
| 2412 |
{
|
| 2413 |
"epoch": 0.9985185185185185,
|
| 2414 |
+
"grad_norm": 0.7482271790504456,
|
| 2415 |
+
"learning_rate": 1.801801801801802e-07,
|
| 2416 |
+
"loss": 0.1817,
|
| 2417 |
"step": 337
|
| 2418 |
},
|
| 2419 |
{
|
| 2420 |
"epoch": 1.0,
|
| 2421 |
+
"grad_norm": 0.9517431855201721,
|
| 2422 |
+
"learning_rate": 9.00900900900901e-08,
|
| 2423 |
+
"loss": 0.1744,
|
| 2424 |
"step": 338
|
| 2425 |
}
|
| 2426 |
],
|
checkpoint-338/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6225
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3886491748a57dcafa54bd22de46a1c52118c309f1ea7cd3a4afc50d85843e8
|
| 3 |
size 6225
|