Add fine-tuned reverse model
Browse files- adapter_model.safetensors +1 -1
- checkpoint-1200/adapter_model.safetensors +1 -1
- checkpoint-1200/optimizer.pt +1 -1
- checkpoint-1200/rng_state.pth +1 -1
- checkpoint-1200/trainer_state.json +302 -302
- checkpoint-1200/training_args.bin +1 -1
- checkpoint-1231/adapter_model.safetensors +1 -1
- checkpoint-1231/optimizer.pt +1 -1
- checkpoint-1231/rng_state.pth +1 -1
- checkpoint-1231/trainer_state.json +308 -308
- checkpoint-1231/training_args.bin +1 -1
- training_args.bin +1 -1
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8405472
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:785ee56c23465a10b4464b35d71e6e3a37ecea3efc9900199739964a182f8463
|
| 3 |
size 8405472
|
checkpoint-1200/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8405472
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c40325bc6b0d94db4ff1a7099dfde12ccef64d1d09393654f3f65dfc87b57038
|
| 3 |
size 8405472
|
checkpoint-1200/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4411194
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae87b3adf150d00d014ba15dca822ab4a737276032c9b389628c5d4c41f2e25f
|
| 3 |
size 4411194
|
checkpoint-1200/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c80e82a7639e588810bfa55efd6a4dda224b512193918b24627c05ad13a1a2b
|
| 3 |
size 14244
|
checkpoint-1200/trainer_state.json
CHANGED
|
@@ -10,842 +10,842 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004062563477554337,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 9.91876523151909e-05,
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.008125126955108674,
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
-
"learning_rate": 9.
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01218769043266301,
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
-
"learning_rate": 9.
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01625025391021735,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
-
"learning_rate": 9.
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.020312817387771683,
|
| 41 |
-
"grad_norm": 2.
|
| 42 |
-
"learning_rate": 9.
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02437538086532602,
|
| 48 |
-
"grad_norm": 1.
|
| 49 |
-
"learning_rate": 9.
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.028437944342880356,
|
| 55 |
-
"grad_norm": 2.
|
| 56 |
-
"learning_rate": 9.
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.0325005078204347,
|
| 62 |
-
"grad_norm": 1.
|
| 63 |
-
"learning_rate": 9.
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03656307129798903,
|
| 69 |
-
"grad_norm": 3.
|
| 70 |
-
"learning_rate": 9.
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.040625634775543366,
|
| 76 |
-
"grad_norm": 1.
|
| 77 |
-
"learning_rate": 9.
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.04468819825309771,
|
| 83 |
-
"grad_norm": 1.
|
| 84 |
-
"learning_rate": 9.
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.04875076173065204,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
-
"learning_rate": 9.
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05281332520820638,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
-
"learning_rate": 8.
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05687588868576071,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
-
"learning_rate": 8.
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06093845216331505,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
-
"learning_rate": 8.
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0650010156408694,
|
| 118 |
-
"grad_norm": 1.
|
| 119 |
-
"learning_rate": 8.
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.06906357911842373,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
-
"learning_rate": 8.
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07312614259597806,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate": 8.
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.0771887060735324,
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
-
"learning_rate": 8.
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08125126955108673,
|
| 146 |
-
"grad_norm": 3.
|
| 147 |
-
"learning_rate": 8.
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.08531383302864107,
|
| 153 |
-
"grad_norm": 1.
|
| 154 |
-
"learning_rate": 8.
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.08937639650619542,
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate": 8.
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.09343895998374975,
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate": 8.
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.09750152346130408,
|
| 174 |
-
"grad_norm": 1.
|
| 175 |
-
"learning_rate": 8.
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.10156408693885842,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate": 7.
|
| 183 |
-
"loss": 1.
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.10562665041641275,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate": 7.
|
| 190 |
-
"loss": 1.
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.10968921389396709,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate": 7.
|
| 197 |
-
"loss": 1.
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.11375177737152142,
|
| 202 |
-
"grad_norm": 1.
|
| 203 |
-
"learning_rate": 7.
|
| 204 |
-
"loss": 1.
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.11781434084907577,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate": 7.
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.1218769043266301,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate": 7.
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.12593946780418444,
|
| 223 |
-
"grad_norm": 1.
|
| 224 |
-
"learning_rate": 7.
|
| 225 |
-
"loss": 1.
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.1300020312817388,
|
| 230 |
-
"grad_norm": 2.
|
| 231 |
-
"learning_rate": 7.
|
| 232 |
-
"loss": 1.
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.1340645947592931,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate": 7.
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.13812715823684746,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate": 7.
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.14218972171440178,
|
| 251 |
-
"grad_norm": 3.
|
| 252 |
-
"learning_rate": 7.
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.14625228519195613,
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate": 7.
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.15031484866951045,
|
| 265 |
-
"grad_norm": 1.
|
| 266 |
-
"learning_rate": 7.
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1543774121470648,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate": 6.
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.15843997562461914,
|
| 279 |
-
"grad_norm": 1.
|
| 280 |
-
"learning_rate": 6.
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.16250253910217347,
|
| 286 |
-
"grad_norm": 1.
|
| 287 |
-
"learning_rate": 6.
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.16656510257972781,
|
| 293 |
-
"grad_norm": 1.
|
| 294 |
-
"learning_rate": 6.
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.17062766605728213,
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate": 6.
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.17469022953483648,
|
| 307 |
-
"grad_norm": 1.
|
| 308 |
-
"learning_rate": 6.
|
| 309 |
-
"loss": 1.
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.17875279301239083,
|
| 314 |
-
"grad_norm": 2.
|
| 315 |
-
"learning_rate": 6.
|
| 316 |
-
"loss": 1.
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.18281535648994515,
|
| 321 |
-
"grad_norm": 1.
|
| 322 |
-
"learning_rate": 6.
|
| 323 |
-
"loss": 1.
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.1868779199674995,
|
| 328 |
-
"grad_norm":
|
| 329 |
-
"learning_rate": 6.
|
| 330 |
-
"loss": 1.
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.19094048344505382,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate": 6.
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.19500304692260817,
|
| 342 |
-
"grad_norm": 3.
|
| 343 |
-
"learning_rate": 6.
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.1990656104001625,
|
| 349 |
-
"grad_norm":
|
| 350 |
-
"learning_rate": 6.
|
| 351 |
-
"loss": 1.
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.20312817387771684,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
-
"learning_rate": 5.
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.2071907373552712,
|
| 363 |
-
"grad_norm": 1.
|
| 364 |
-
"learning_rate": 5.
|
| 365 |
-
"loss": 1.
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.2112533008328255,
|
| 370 |
-
"grad_norm":
|
| 371 |
-
"learning_rate": 5.
|
| 372 |
-
"loss": 1.
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.21531586431037986,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
-
"learning_rate": 5.
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.21937842778793418,
|
| 384 |
-
"grad_norm": 1.
|
| 385 |
-
"learning_rate": 5.
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.22344099126548853,
|
| 391 |
-
"grad_norm":
|
| 392 |
-
"learning_rate": 5.
|
| 393 |
-
"loss": 1.
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.22750355474304285,
|
| 398 |
-
"grad_norm": 1.
|
| 399 |
-
"learning_rate": 5.
|
| 400 |
-
"loss": 1.
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2315661182205972,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate": 5.
|
| 407 |
-
"loss": 1.
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.23562868169815154,
|
| 412 |
-
"grad_norm": 1.
|
| 413 |
-
"learning_rate": 5.
|
| 414 |
-
"loss": 1.
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.23969124517570586,
|
| 419 |
-
"grad_norm": 1.
|
| 420 |
-
"learning_rate": 5.
|
| 421 |
-
"loss": 1.
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2437538086532602,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
-
"learning_rate": 5.
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.24781637213081453,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate": 5.
|
| 435 |
-
"loss": 1.
|
| 436 |
"step": 610
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.2518789356083689,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate": 4.
|
| 442 |
-
"loss": 1.
|
| 443 |
"step": 620
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.25594149908592323,
|
| 447 |
-
"grad_norm": 1.
|
| 448 |
-
"learning_rate": 4.
|
| 449 |
-
"loss": 1.
|
| 450 |
"step": 630
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.2600040625634776,
|
| 454 |
-
"grad_norm": 1.
|
| 455 |
-
"learning_rate": 4.
|
| 456 |
-
"loss": 1.
|
| 457 |
"step": 640
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.26406662604103187,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
"learning_rate": 4.735987002437044e-05,
|
| 463 |
-
"loss": 1.
|
| 464 |
"step": 650
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.2681291895185862,
|
| 468 |
-
"grad_norm": 2.
|
| 469 |
"learning_rate": 4.6547522339561334e-05,
|
| 470 |
-
"loss": 1.
|
| 471 |
"step": 660
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.27219175299614057,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
"learning_rate": 4.573517465475224e-05,
|
| 477 |
-
"loss": 1.
|
| 478 |
"step": 670
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2762543164736949,
|
| 482 |
-
"grad_norm": 1.
|
| 483 |
"learning_rate": 4.492282696994314e-05,
|
| 484 |
"loss": 1.3557,
|
| 485 |
"step": 680
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.28031687995124926,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
"learning_rate": 4.411047928513404e-05,
|
| 491 |
-
"loss": 1.
|
| 492 |
"step": 690
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.28437944342880356,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
"learning_rate": 4.329813160032494e-05,
|
| 498 |
-
"loss": 1.
|
| 499 |
"step": 700
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.2884420069063579,
|
| 503 |
-
"grad_norm": 1.
|
| 504 |
"learning_rate": 4.2485783915515845e-05,
|
| 505 |
-
"loss": 1.
|
| 506 |
"step": 710
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.29250457038391225,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
"learning_rate": 4.167343623070675e-05,
|
| 512 |
-
"loss": 1.
|
| 513 |
"step": 720
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.2965671338614666,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
"learning_rate": 4.0861088545897645e-05,
|
| 519 |
-
"loss": 1.
|
| 520 |
"step": 730
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.3006296973390209,
|
| 524 |
-
"grad_norm": 1.
|
| 525 |
"learning_rate": 4.004874086108855e-05,
|
| 526 |
-
"loss": 1.
|
| 527 |
"step": 740
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.30469226081657524,
|
| 531 |
-
"grad_norm":
|
| 532 |
"learning_rate": 3.923639317627945e-05,
|
| 533 |
-
"loss": 1.
|
| 534 |
"step": 750
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3087548242941296,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
"learning_rate": 3.842404549147035e-05,
|
| 540 |
-
"loss": 1.
|
| 541 |
"step": 760
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.31281738777168394,
|
| 545 |
-
"grad_norm": 1.
|
| 546 |
"learning_rate": 3.761169780666125e-05,
|
| 547 |
-
"loss": 1.
|
| 548 |
"step": 770
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.3168799512492383,
|
| 552 |
-
"grad_norm": 1.
|
| 553 |
"learning_rate": 3.6799350121852156e-05,
|
| 554 |
-
"loss": 1.
|
| 555 |
"step": 780
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.3209425147267926,
|
| 559 |
-
"grad_norm": 1.
|
| 560 |
"learning_rate": 3.598700243704305e-05,
|
| 561 |
-
"loss": 1.
|
| 562 |
"step": 790
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.32500507820434693,
|
| 566 |
-
"grad_norm":
|
| 567 |
"learning_rate": 3.517465475223396e-05,
|
| 568 |
-
"loss": 1.
|
| 569 |
"step": 800
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.3290676416819013,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
"learning_rate": 3.436230706742486e-05,
|
| 575 |
-
"loss": 1.
|
| 576 |
"step": 810
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.33313020515945563,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 3.3549959382615757e-05,
|
| 582 |
-
"loss": 1.
|
| 583 |
"step": 820
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.33719276863701,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
"learning_rate": 3.273761169780667e-05,
|
| 589 |
-
"loss": 1.
|
| 590 |
"step": 830
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.34125533211456427,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
"learning_rate": 3.1925264012997564e-05,
|
| 596 |
-
"loss": 1.
|
| 597 |
"step": 840
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.3453178955921186,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 3.111291632818847e-05,
|
| 603 |
-
"loss": 1.
|
| 604 |
"step": 850
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.34938045906967297,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 3.0300568643379367e-05,
|
| 610 |
-
"loss": 1.
|
| 611 |
"step": 860
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.3534430225472273,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 2.9488220958570267e-05,
|
| 617 |
-
"loss": 1.
|
| 618 |
"step": 870
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.35750558602478166,
|
| 622 |
-
"grad_norm": 1.
|
| 623 |
"learning_rate": 2.867587327376117e-05,
|
| 624 |
-
"loss": 1.
|
| 625 |
"step": 880
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.36156814950233596,
|
| 629 |
-
"grad_norm": 1.
|
| 630 |
"learning_rate": 2.786352558895207e-05,
|
| 631 |
-
"loss": 1.
|
| 632 |
"step": 890
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.3656307129798903,
|
| 636 |
-
"grad_norm": 2.
|
| 637 |
"learning_rate": 2.7051177904142978e-05,
|
| 638 |
-
"loss": 1.
|
| 639 |
"step": 900
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.36969327645744465,
|
| 643 |
-
"grad_norm": 1.
|
| 644 |
"learning_rate": 2.6238830219333875e-05,
|
| 645 |
-
"loss": 1.
|
| 646 |
"step": 910
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.373755839934999,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
"learning_rate": 2.5426482534524775e-05,
|
| 652 |
-
"loss": 1.
|
| 653 |
"step": 920
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.3778184034125533,
|
| 657 |
-
"grad_norm": 1.
|
| 658 |
"learning_rate": 2.461413484971568e-05,
|
| 659 |
-
"loss": 1.
|
| 660 |
"step": 930
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.38188096689010764,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
"learning_rate": 2.3801787164906582e-05,
|
| 666 |
-
"loss": 1.
|
| 667 |
"step": 940
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.385943530367662,
|
| 671 |
-
"grad_norm": 1.
|
| 672 |
"learning_rate": 2.2989439480097482e-05,
|
| 673 |
-
"loss": 1.
|
| 674 |
"step": 950
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.39000609384521634,
|
| 678 |
-
"grad_norm":
|
| 679 |
"learning_rate": 2.2177091795288386e-05,
|
| 680 |
-
"loss": 1.
|
| 681 |
"step": 960
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.3940686573227707,
|
| 685 |
-
"grad_norm": 2.
|
| 686 |
"learning_rate": 2.1364744110479286e-05,
|
| 687 |
-
"loss": 1.
|
| 688 |
"step": 970
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.398131220800325,
|
| 692 |
-
"grad_norm": 1.
|
| 693 |
"learning_rate": 2.0552396425670186e-05,
|
| 694 |
-
"loss": 1.
|
| 695 |
"step": 980
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.40219378427787933,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
"learning_rate": 1.974004874086109e-05,
|
| 701 |
-
"loss": 1.
|
| 702 |
"step": 990
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.4062563477554337,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
"learning_rate": 1.892770105605199e-05,
|
| 708 |
-
"loss": 1.
|
| 709 |
"step": 1000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.410318911232988,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
"learning_rate": 1.8115353371242893e-05,
|
| 715 |
-
"loss": 1.
|
| 716 |
"step": 1010
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.4143814747105424,
|
| 720 |
-
"grad_norm": 1.
|
| 721 |
"learning_rate": 1.7303005686433797e-05,
|
| 722 |
-
"loss": 1.
|
| 723 |
"step": 1020
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.41844403818809667,
|
| 727 |
-
"grad_norm": 1.
|
| 728 |
"learning_rate": 1.6490658001624697e-05,
|
| 729 |
-
"loss": 1.
|
| 730 |
"step": 1030
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.422506601665651,
|
| 734 |
-
"grad_norm":
|
| 735 |
"learning_rate": 1.5678310316815597e-05,
|
| 736 |
-
"loss": 1.
|
| 737 |
"step": 1040
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.42656916514320536,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
"learning_rate": 1.48659626320065e-05,
|
| 743 |
-
"loss": 1.
|
| 744 |
"step": 1050
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.4306317286207597,
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
"learning_rate": 1.4053614947197402e-05,
|
| 750 |
-
"loss": 1.
|
| 751 |
"step": 1060
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.43469429209831406,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
"learning_rate": 1.3241267262388301e-05,
|
| 757 |
-
"loss": 1.
|
| 758 |
"step": 1070
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.43875685557586835,
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
"learning_rate": 1.2428919577579204e-05,
|
| 764 |
-
"loss": 1.
|
| 765 |
"step": 1080
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.4428194190534227,
|
| 769 |
-
"grad_norm": 1.
|
| 770 |
"learning_rate": 1.1616571892770106e-05,
|
| 771 |
-
"loss": 1.
|
| 772 |
"step": 1090
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.44688198253097705,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
"learning_rate": 1.0804224207961008e-05,
|
| 778 |
-
"loss": 1.
|
| 779 |
"step": 1100
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.4509445460085314,
|
| 783 |
-
"grad_norm":
|
| 784 |
"learning_rate": 9.99187652315191e-06,
|
| 785 |
-
"loss": 1.
|
| 786 |
"step": 1110
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.4550071094860857,
|
| 790 |
-
"grad_norm": 1.
|
| 791 |
"learning_rate": 9.179528838342812e-06,
|
| 792 |
-
"loss": 1.
|
| 793 |
"step": 1120
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.45906967296364004,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
"learning_rate": 8.367181153533712e-06,
|
| 799 |
-
"loss": 1.
|
| 800 |
"step": 1130
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.4631322364411944,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
"learning_rate": 7.554833468724615e-06,
|
| 806 |
-
"loss": 1.
|
| 807 |
"step": 1140
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.46719479991874874,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
"learning_rate": 6.742485783915516e-06,
|
| 813 |
-
"loss": 1.
|
| 814 |
"step": 1150
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.4712573633963031,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
"learning_rate": 5.930138099106418e-06,
|
| 820 |
-
"loss": 1.
|
| 821 |
"step": 1160
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.4753199268738574,
|
| 825 |
-
"grad_norm": 1.
|
| 826 |
"learning_rate": 5.117790414297319e-06,
|
| 827 |
-
"loss": 1.
|
| 828 |
"step": 1170
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.4793824903514117,
|
| 832 |
-
"grad_norm": 1.
|
| 833 |
"learning_rate": 4.305442729488221e-06,
|
| 834 |
-
"loss": 1.
|
| 835 |
"step": 1180
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.4834450538289661,
|
| 839 |
-
"grad_norm":
|
| 840 |
"learning_rate": 3.4930950446791225e-06,
|
| 841 |
-
"loss": 1.
|
| 842 |
"step": 1190
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.4875076173065204,
|
| 846 |
-
"grad_norm": 2.
|
| 847 |
"learning_rate": 2.6807473598700244e-06,
|
| 848 |
-
"loss": 1.
|
| 849 |
"step": 1200
|
| 850 |
}
|
| 851 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004062563477554337,
|
| 13 |
+
"grad_norm": 1.032738447189331,
|
| 14 |
"learning_rate": 9.91876523151909e-05,
|
| 15 |
+
"loss": 2.0442,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.008125126955108674,
|
| 20 |
+
"grad_norm": 0.8217566609382629,
|
| 21 |
+
"learning_rate": 9.845653939886271e-05,
|
| 22 |
+
"loss": 1.9828,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01218769043266301,
|
| 27 |
+
"grad_norm": 0.6819207668304443,
|
| 28 |
+
"learning_rate": 9.764419171405362e-05,
|
| 29 |
+
"loss": 1.7187,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01625025391021735,
|
| 34 |
+
"grad_norm": 0.9606077075004578,
|
| 35 |
+
"learning_rate": 9.683184402924452e-05,
|
| 36 |
+
"loss": 1.6162,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.020312817387771683,
|
| 41 |
+
"grad_norm": 2.6619677543640137,
|
| 42 |
+
"learning_rate": 9.601949634443542e-05,
|
| 43 |
+
"loss": 1.6977,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02437538086532602,
|
| 48 |
+
"grad_norm": 1.5025429725646973,
|
| 49 |
+
"learning_rate": 9.520714865962633e-05,
|
| 50 |
+
"loss": 1.2874,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.028437944342880356,
|
| 55 |
+
"grad_norm": 2.7792210578918457,
|
| 56 |
+
"learning_rate": 9.439480097481722e-05,
|
| 57 |
+
"loss": 1.4551,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.0325005078204347,
|
| 62 |
+
"grad_norm": 1.557523250579834,
|
| 63 |
+
"learning_rate": 9.358245329000814e-05,
|
| 64 |
+
"loss": 1.4412,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03656307129798903,
|
| 69 |
+
"grad_norm": 3.4479055404663086,
|
| 70 |
+
"learning_rate": 9.277010560519903e-05,
|
| 71 |
+
"loss": 1.4806,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.040625634775543366,
|
| 76 |
+
"grad_norm": 1.282782793045044,
|
| 77 |
+
"learning_rate": 9.195775792038993e-05,
|
| 78 |
+
"loss": 1.3971,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.04468819825309771,
|
| 83 |
+
"grad_norm": 1.6942788362503052,
|
| 84 |
+
"learning_rate": 9.114541023558083e-05,
|
| 85 |
+
"loss": 1.543,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.04875076173065204,
|
| 90 |
+
"grad_norm": 0.7825320363044739,
|
| 91 |
+
"learning_rate": 9.033306255077174e-05,
|
| 92 |
+
"loss": 1.3638,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05281332520820638,
|
| 97 |
+
"grad_norm": 0.7435455918312073,
|
| 98 |
+
"learning_rate": 8.952071486596265e-05,
|
| 99 |
+
"loss": 1.4731,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05687588868576071,
|
| 104 |
+
"grad_norm": 0.5097799897193909,
|
| 105 |
+
"learning_rate": 8.870836718115354e-05,
|
| 106 |
+
"loss": 1.4218,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06093845216331505,
|
| 111 |
+
"grad_norm": 0.8880577087402344,
|
| 112 |
+
"learning_rate": 8.789601949634444e-05,
|
| 113 |
+
"loss": 1.3477,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0650010156408694,
|
| 118 |
+
"grad_norm": 1.1695178747177124,
|
| 119 |
+
"learning_rate": 8.708367181153534e-05,
|
| 120 |
+
"loss": 1.4521,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.06906357911842373,
|
| 125 |
+
"grad_norm": 0.7951045036315918,
|
| 126 |
+
"learning_rate": 8.627132412672623e-05,
|
| 127 |
+
"loss": 1.3417,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07312614259597806,
|
| 132 |
+
"grad_norm": 0.5302656888961792,
|
| 133 |
+
"learning_rate": 8.545897644191714e-05,
|
| 134 |
+
"loss": 1.5202,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.0771887060735324,
|
| 139 |
+
"grad_norm": 0.8830013275146484,
|
| 140 |
+
"learning_rate": 8.464662875710805e-05,
|
| 141 |
+
"loss": 1.4271,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08125126955108673,
|
| 146 |
+
"grad_norm": 3.1300947666168213,
|
| 147 |
+
"learning_rate": 8.383428107229895e-05,
|
| 148 |
+
"loss": 1.3743,
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.08531383302864107,
|
| 153 |
+
"grad_norm": 1.077304720878601,
|
| 154 |
+
"learning_rate": 8.302193338748985e-05,
|
| 155 |
+
"loss": 1.4244,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.08937639650619542,
|
| 160 |
+
"grad_norm": 1.6846965551376343,
|
| 161 |
+
"learning_rate": 8.220958570268074e-05,
|
| 162 |
+
"loss": 1.4471,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.09343895998374975,
|
| 167 |
+
"grad_norm": 1.066351056098938,
|
| 168 |
+
"learning_rate": 8.139723801787165e-05,
|
| 169 |
+
"loss": 1.3723,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.09750152346130408,
|
| 174 |
+
"grad_norm": 1.6999150514602661,
|
| 175 |
+
"learning_rate": 8.058489033306255e-05,
|
| 176 |
+
"loss": 1.368,
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.10156408693885842,
|
| 181 |
+
"grad_norm": 0.7261852622032166,
|
| 182 |
+
"learning_rate": 7.977254264825346e-05,
|
| 183 |
+
"loss": 1.3175,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.10562665041641275,
|
| 188 |
+
"grad_norm": 0.8671672344207764,
|
| 189 |
+
"learning_rate": 7.896019496344436e-05,
|
| 190 |
+
"loss": 1.4264,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.10968921389396709,
|
| 195 |
+
"grad_norm": 0.9500339031219482,
|
| 196 |
+
"learning_rate": 7.814784727863525e-05,
|
| 197 |
+
"loss": 1.2985,
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.11375177737152142,
|
| 202 |
+
"grad_norm": 1.071378469467163,
|
| 203 |
+
"learning_rate": 7.733549959382617e-05,
|
| 204 |
+
"loss": 1.447,
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.11781434084907577,
|
| 209 |
+
"grad_norm": 0.7448930144309998,
|
| 210 |
+
"learning_rate": 7.652315190901706e-05,
|
| 211 |
+
"loss": 1.374,
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.1218769043266301,
|
| 216 |
+
"grad_norm": 0.9604855179786682,
|
| 217 |
+
"learning_rate": 7.571080422420796e-05,
|
| 218 |
+
"loss": 1.4803,
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.12593946780418444,
|
| 223 |
+
"grad_norm": 1.9951509237289429,
|
| 224 |
+
"learning_rate": 7.489845653939887e-05,
|
| 225 |
+
"loss": 1.4477,
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.1300020312817388,
|
| 230 |
+
"grad_norm": 2.3266055583953857,
|
| 231 |
+
"learning_rate": 7.408610885458977e-05,
|
| 232 |
+
"loss": 1.6107,
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.1340645947592931,
|
| 237 |
+
"grad_norm": 0.878081738948822,
|
| 238 |
+
"learning_rate": 7.327376116978068e-05,
|
| 239 |
+
"loss": 1.295,
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.13812715823684746,
|
| 244 |
+
"grad_norm": 0.7642931938171387,
|
| 245 |
+
"learning_rate": 7.246141348497157e-05,
|
| 246 |
+
"loss": 1.4776,
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.14218972171440178,
|
| 251 |
+
"grad_norm": 3.2081470489501953,
|
| 252 |
+
"learning_rate": 7.164906580016247e-05,
|
| 253 |
+
"loss": 1.448,
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.14625228519195613,
|
| 258 |
+
"grad_norm": 0.9164071679115295,
|
| 259 |
+
"learning_rate": 7.083671811535338e-05,
|
| 260 |
+
"loss": 1.3792,
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.15031484866951045,
|
| 265 |
+
"grad_norm": 1.9146902561187744,
|
| 266 |
+
"learning_rate": 7.002437043054428e-05,
|
| 267 |
+
"loss": 1.239,
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1543774121470648,
|
| 272 |
+
"grad_norm": 0.6990465521812439,
|
| 273 |
+
"learning_rate": 6.921202274573519e-05,
|
| 274 |
+
"loss": 1.3211,
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.15843997562461914,
|
| 279 |
+
"grad_norm": 1.7084014415740967,
|
| 280 |
+
"learning_rate": 6.839967506092608e-05,
|
| 281 |
+
"loss": 1.3952,
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.16250253910217347,
|
| 286 |
+
"grad_norm": 1.0569523572921753,
|
| 287 |
+
"learning_rate": 6.758732737611698e-05,
|
| 288 |
+
"loss": 1.385,
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.16656510257972781,
|
| 293 |
+
"grad_norm": 1.4293252229690552,
|
| 294 |
+
"learning_rate": 6.677497969130788e-05,
|
| 295 |
+
"loss": 1.3736,
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.17062766605728213,
|
| 300 |
+
"grad_norm": 1.0318830013275146,
|
| 301 |
+
"learning_rate": 6.596263200649879e-05,
|
| 302 |
+
"loss": 1.2315,
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.17469022953483648,
|
| 307 |
+
"grad_norm": 1.0838662385940552,
|
| 308 |
+
"learning_rate": 6.515028432168968e-05,
|
| 309 |
+
"loss": 1.5313,
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.17875279301239083,
|
| 314 |
+
"grad_norm": 2.388718366622925,
|
| 315 |
+
"learning_rate": 6.43379366368806e-05,
|
| 316 |
+
"loss": 1.5047,
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.18281535648994515,
|
| 321 |
+
"grad_norm": 1.58635413646698,
|
| 322 |
+
"learning_rate": 6.352558895207149e-05,
|
| 323 |
+
"loss": 1.2847,
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.1868779199674995,
|
| 328 |
+
"grad_norm": 4.313031196594238,
|
| 329 |
+
"learning_rate": 6.271324126726239e-05,
|
| 330 |
+
"loss": 1.46,
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.19094048344505382,
|
| 335 |
+
"grad_norm": 0.5860478281974792,
|
| 336 |
+
"learning_rate": 6.190089358245328e-05,
|
| 337 |
+
"loss": 1.3614,
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.19500304692260817,
|
| 342 |
+
"grad_norm": 3.0871775150299072,
|
| 343 |
+
"learning_rate": 6.10885458976442e-05,
|
| 344 |
+
"loss": 1.3915,
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.1990656104001625,
|
| 349 |
+
"grad_norm": 2.025517463684082,
|
| 350 |
+
"learning_rate": 6.02761982128351e-05,
|
| 351 |
+
"loss": 1.3326,
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.20312817387771684,
|
| 356 |
+
"grad_norm": 0.8149346709251404,
|
| 357 |
+
"learning_rate": 5.9463850528026e-05,
|
| 358 |
+
"loss": 1.2553,
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.2071907373552712,
|
| 363 |
+
"grad_norm": 1.1703840494155884,
|
| 364 |
+
"learning_rate": 5.86515028432169e-05,
|
| 365 |
+
"loss": 1.4137,
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.2112533008328255,
|
| 370 |
+
"grad_norm": 2.0241522789001465,
|
| 371 |
+
"learning_rate": 5.7839155158407796e-05,
|
| 372 |
+
"loss": 1.3763,
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.21531586431037986,
|
| 377 |
+
"grad_norm": 0.5776875019073486,
|
| 378 |
+
"learning_rate": 5.70268074735987e-05,
|
| 379 |
+
"loss": 1.282,
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.21937842778793418,
|
| 384 |
+
"grad_norm": 1.644302487373352,
|
| 385 |
+
"learning_rate": 5.621445978878961e-05,
|
| 386 |
+
"loss": 1.3309,
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.22344099126548853,
|
| 391 |
+
"grad_norm": 1.0667213201522827,
|
| 392 |
+
"learning_rate": 5.5402112103980506e-05,
|
| 393 |
+
"loss": 1.4516,
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.22750355474304285,
|
| 398 |
+
"grad_norm": 1.1769509315490723,
|
| 399 |
+
"learning_rate": 5.458976441917141e-05,
|
| 400 |
+
"loss": 1.4305,
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2315661182205972,
|
| 405 |
+
"grad_norm": 0.9089193940162659,
|
| 406 |
+
"learning_rate": 5.377741673436231e-05,
|
| 407 |
+
"loss": 1.3263,
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.23562868169815154,
|
| 412 |
+
"grad_norm": 1.9467777013778687,
|
| 413 |
+
"learning_rate": 5.2965069049553203e-05,
|
| 414 |
+
"loss": 1.402,
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.23969124517570586,
|
| 419 |
+
"grad_norm": 1.313186526298523,
|
| 420 |
+
"learning_rate": 5.2152721364744114e-05,
|
| 421 |
+
"loss": 1.3604,
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2437538086532602,
|
| 426 |
+
"grad_norm": 0.48065802454948425,
|
| 427 |
+
"learning_rate": 5.134037367993502e-05,
|
| 428 |
+
"loss": 1.5186,
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.24781637213081453,
|
| 433 |
+
"grad_norm": 0.7379820942878723,
|
| 434 |
+
"learning_rate": 5.0528025995125914e-05,
|
| 435 |
+
"loss": 1.4457,
|
| 436 |
"step": 610
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.2518789356083689,
|
| 440 |
+
"grad_norm": 0.7363265156745911,
|
| 441 |
+
"learning_rate": 4.971567831031682e-05,
|
| 442 |
+
"loss": 1.3843,
|
| 443 |
"step": 620
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.25594149908592323,
|
| 447 |
+
"grad_norm": 1.349368691444397,
|
| 448 |
+
"learning_rate": 4.890333062550772e-05,
|
| 449 |
+
"loss": 1.355,
|
| 450 |
"step": 630
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.2600040625634776,
|
| 454 |
+
"grad_norm": 1.6113736629486084,
|
| 455 |
+
"learning_rate": 4.809098294069862e-05,
|
| 456 |
+
"loss": 1.5113,
|
| 457 |
"step": 640
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.26406662604103187,
|
| 461 |
+
"grad_norm": 0.7977277636528015,
|
| 462 |
"learning_rate": 4.735987002437044e-05,
|
| 463 |
+
"loss": 1.4413,
|
| 464 |
"step": 650
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.2681291895185862,
|
| 468 |
+
"grad_norm": 2.2527716159820557,
|
| 469 |
"learning_rate": 4.6547522339561334e-05,
|
| 470 |
+
"loss": 1.4792,
|
| 471 |
"step": 660
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.27219175299614057,
|
| 475 |
+
"grad_norm": 0.7665427923202515,
|
| 476 |
"learning_rate": 4.573517465475224e-05,
|
| 477 |
+
"loss": 1.2581,
|
| 478 |
"step": 670
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2762543164736949,
|
| 482 |
+
"grad_norm": 1.6001981496810913,
|
| 483 |
"learning_rate": 4.492282696994314e-05,
|
| 484 |
"loss": 1.3557,
|
| 485 |
"step": 680
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.28031687995124926,
|
| 489 |
+
"grad_norm": 0.6028335690498352,
|
| 490 |
"learning_rate": 4.411047928513404e-05,
|
| 491 |
+
"loss": 1.4625,
|
| 492 |
"step": 690
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.28437944342880356,
|
| 496 |
+
"grad_norm": 0.621457576751709,
|
| 497 |
"learning_rate": 4.329813160032494e-05,
|
| 498 |
+
"loss": 1.4177,
|
| 499 |
"step": 700
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.2884420069063579,
|
| 503 |
+
"grad_norm": 1.08269464969635,
|
| 504 |
"learning_rate": 4.2485783915515845e-05,
|
| 505 |
+
"loss": 1.5553,
|
| 506 |
"step": 710
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.29250457038391225,
|
| 510 |
+
"grad_norm": 0.7009453177452087,
|
| 511 |
"learning_rate": 4.167343623070675e-05,
|
| 512 |
+
"loss": 1.3478,
|
| 513 |
"step": 720
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.2965671338614666,
|
| 517 |
+
"grad_norm": 0.7686846256256104,
|
| 518 |
"learning_rate": 4.0861088545897645e-05,
|
| 519 |
+
"loss": 1.5388,
|
| 520 |
"step": 730
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.3006296973390209,
|
| 524 |
+
"grad_norm": 1.193659782409668,
|
| 525 |
"learning_rate": 4.004874086108855e-05,
|
| 526 |
+
"loss": 1.558,
|
| 527 |
"step": 740
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.30469226081657524,
|
| 531 |
+
"grad_norm": 0.9236335754394531,
|
| 532 |
"learning_rate": 3.923639317627945e-05,
|
| 533 |
+
"loss": 1.4143,
|
| 534 |
"step": 750
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3087548242941296,
|
| 538 |
+
"grad_norm": 0.7839071750640869,
|
| 539 |
"learning_rate": 3.842404549147035e-05,
|
| 540 |
+
"loss": 1.2117,
|
| 541 |
"step": 760
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.31281738777168394,
|
| 545 |
+
"grad_norm": 1.2941004037857056,
|
| 546 |
"learning_rate": 3.761169780666125e-05,
|
| 547 |
+
"loss": 1.4062,
|
| 548 |
"step": 770
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.3168799512492383,
|
| 552 |
+
"grad_norm": 1.426131010055542,
|
| 553 |
"learning_rate": 3.6799350121852156e-05,
|
| 554 |
+
"loss": 1.4909,
|
| 555 |
"step": 780
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.3209425147267926,
|
| 559 |
+
"grad_norm": 1.1472172737121582,
|
| 560 |
"learning_rate": 3.598700243704305e-05,
|
| 561 |
+
"loss": 1.2864,
|
| 562 |
"step": 790
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.32500507820434693,
|
| 566 |
+
"grad_norm": 2.1125993728637695,
|
| 567 |
"learning_rate": 3.517465475223396e-05,
|
| 568 |
+
"loss": 1.3456,
|
| 569 |
"step": 800
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.3290676416819013,
|
| 573 |
+
"grad_norm": 0.7743927240371704,
|
| 574 |
"learning_rate": 3.436230706742486e-05,
|
| 575 |
+
"loss": 1.3294,
|
| 576 |
"step": 810
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.33313020515945563,
|
| 580 |
+
"grad_norm": 0.8188801407814026,
|
| 581 |
"learning_rate": 3.3549959382615757e-05,
|
| 582 |
+
"loss": 1.2803,
|
| 583 |
"step": 820
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.33719276863701,
|
| 587 |
+
"grad_norm": 0.5288794636726379,
|
| 588 |
"learning_rate": 3.273761169780667e-05,
|
| 589 |
+
"loss": 1.4094,
|
| 590 |
"step": 830
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.34125533211456427,
|
| 594 |
+
"grad_norm": 0.6982353329658508,
|
| 595 |
"learning_rate": 3.1925264012997564e-05,
|
| 596 |
+
"loss": 1.1621,
|
| 597 |
"step": 840
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.3453178955921186,
|
| 601 |
+
"grad_norm": 0.7790868282318115,
|
| 602 |
"learning_rate": 3.111291632818847e-05,
|
| 603 |
+
"loss": 1.4534,
|
| 604 |
"step": 850
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.34938045906967297,
|
| 608 |
+
"grad_norm": 0.7421971559524536,
|
| 609 |
"learning_rate": 3.0300568643379367e-05,
|
| 610 |
+
"loss": 1.2951,
|
| 611 |
"step": 860
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.3534430225472273,
|
| 615 |
+
"grad_norm": 0.9682130813598633,
|
| 616 |
"learning_rate": 2.9488220958570267e-05,
|
| 617 |
+
"loss": 1.3827,
|
| 618 |
"step": 870
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.35750558602478166,
|
| 622 |
+
"grad_norm": 1.6505218744277954,
|
| 623 |
"learning_rate": 2.867587327376117e-05,
|
| 624 |
+
"loss": 1.3482,
|
| 625 |
"step": 880
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.36156814950233596,
|
| 629 |
+
"grad_norm": 1.013390064239502,
|
| 630 |
"learning_rate": 2.786352558895207e-05,
|
| 631 |
+
"loss": 1.4224,
|
| 632 |
"step": 890
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.3656307129798903,
|
| 636 |
+
"grad_norm": 2.4053940773010254,
|
| 637 |
"learning_rate": 2.7051177904142978e-05,
|
| 638 |
+
"loss": 1.4791,
|
| 639 |
"step": 900
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.36969327645744465,
|
| 643 |
+
"grad_norm": 1.7795937061309814,
|
| 644 |
"learning_rate": 2.6238830219333875e-05,
|
| 645 |
+
"loss": 1.472,
|
| 646 |
"step": 910
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.373755839934999,
|
| 650 |
+
"grad_norm": 0.7656849026679993,
|
| 651 |
"learning_rate": 2.5426482534524775e-05,
|
| 652 |
+
"loss": 1.3824,
|
| 653 |
"step": 920
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.3778184034125533,
|
| 657 |
+
"grad_norm": 1.1062685251235962,
|
| 658 |
"learning_rate": 2.461413484971568e-05,
|
| 659 |
+
"loss": 1.3462,
|
| 660 |
"step": 930
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.38188096689010764,
|
| 664 |
+
"grad_norm": 0.660727858543396,
|
| 665 |
"learning_rate": 2.3801787164906582e-05,
|
| 666 |
+
"loss": 1.3834,
|
| 667 |
"step": 940
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.385943530367662,
|
| 671 |
+
"grad_norm": 1.8191992044448853,
|
| 672 |
"learning_rate": 2.2989439480097482e-05,
|
| 673 |
+
"loss": 1.4556,
|
| 674 |
"step": 950
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.39000609384521634,
|
| 678 |
+
"grad_norm": 3.3901174068450928,
|
| 679 |
"learning_rate": 2.2177091795288386e-05,
|
| 680 |
+
"loss": 1.2865,
|
| 681 |
"step": 960
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.3940686573227707,
|
| 685 |
+
"grad_norm": 2.4866228103637695,
|
| 686 |
"learning_rate": 2.1364744110479286e-05,
|
| 687 |
+
"loss": 1.501,
|
| 688 |
"step": 970
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.398131220800325,
|
| 692 |
+
"grad_norm": 1.41251540184021,
|
| 693 |
"learning_rate": 2.0552396425670186e-05,
|
| 694 |
+
"loss": 1.3455,
|
| 695 |
"step": 980
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.40219378427787933,
|
| 699 |
+
"grad_norm": 0.645084798336029,
|
| 700 |
"learning_rate": 1.974004874086109e-05,
|
| 701 |
+
"loss": 1.3918,
|
| 702 |
"step": 990
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.4062563477554337,
|
| 706 |
+
"grad_norm": 0.8688336610794067,
|
| 707 |
"learning_rate": 1.892770105605199e-05,
|
| 708 |
+
"loss": 1.3823,
|
| 709 |
"step": 1000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.410318911232988,
|
| 713 |
+
"grad_norm": 0.7747243642807007,
|
| 714 |
"learning_rate": 1.8115353371242893e-05,
|
| 715 |
+
"loss": 1.2988,
|
| 716 |
"step": 1010
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.4143814747105424,
|
| 720 |
+
"grad_norm": 1.329567313194275,
|
| 721 |
"learning_rate": 1.7303005686433797e-05,
|
| 722 |
+
"loss": 1.2876,
|
| 723 |
"step": 1020
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.41844403818809667,
|
| 727 |
+
"grad_norm": 1.022557258605957,
|
| 728 |
"learning_rate": 1.6490658001624697e-05,
|
| 729 |
+
"loss": 1.3036,
|
| 730 |
"step": 1030
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.422506601665651,
|
| 734 |
+
"grad_norm": 0.9812535047531128,
|
| 735 |
"learning_rate": 1.5678310316815597e-05,
|
| 736 |
+
"loss": 1.3278,
|
| 737 |
"step": 1040
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.42656916514320536,
|
| 741 |
+
"grad_norm": 0.8079454898834229,
|
| 742 |
"learning_rate": 1.48659626320065e-05,
|
| 743 |
+
"loss": 1.2957,
|
| 744 |
"step": 1050
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.4306317286207597,
|
| 748 |
+
"grad_norm": 0.8280611038208008,
|
| 749 |
"learning_rate": 1.4053614947197402e-05,
|
| 750 |
+
"loss": 1.3747,
|
| 751 |
"step": 1060
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.43469429209831406,
|
| 755 |
+
"grad_norm": 0.85203617811203,
|
| 756 |
"learning_rate": 1.3241267262388301e-05,
|
| 757 |
+
"loss": 1.423,
|
| 758 |
"step": 1070
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.43875685557586835,
|
| 762 |
+
"grad_norm": 0.9492881298065186,
|
| 763 |
"learning_rate": 1.2428919577579204e-05,
|
| 764 |
+
"loss": 1.3845,
|
| 765 |
"step": 1080
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.4428194190534227,
|
| 769 |
+
"grad_norm": 1.6084877252578735,
|
| 770 |
"learning_rate": 1.1616571892770106e-05,
|
| 771 |
+
"loss": 1.2422,
|
| 772 |
"step": 1090
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.44688198253097705,
|
| 776 |
+
"grad_norm": 0.7060315012931824,
|
| 777 |
"learning_rate": 1.0804224207961008e-05,
|
| 778 |
+
"loss": 1.3758,
|
| 779 |
"step": 1100
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.4509445460085314,
|
| 783 |
+
"grad_norm": 2.0847179889678955,
|
| 784 |
"learning_rate": 9.99187652315191e-06,
|
| 785 |
+
"loss": 1.4202,
|
| 786 |
"step": 1110
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.4550071094860857,
|
| 790 |
+
"grad_norm": 1.1350940465927124,
|
| 791 |
"learning_rate": 9.179528838342812e-06,
|
| 792 |
+
"loss": 1.3273,
|
| 793 |
"step": 1120
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.45906967296364004,
|
| 797 |
+
"grad_norm": 0.841096043586731,
|
| 798 |
"learning_rate": 8.367181153533712e-06,
|
| 799 |
+
"loss": 1.3461,
|
| 800 |
"step": 1130
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.4631322364411944,
|
| 804 |
+
"grad_norm": 0.5926047563552856,
|
| 805 |
"learning_rate": 7.554833468724615e-06,
|
| 806 |
+
"loss": 1.2492,
|
| 807 |
"step": 1140
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.46719479991874874,
|
| 811 |
+
"grad_norm": 0.8289461135864258,
|
| 812 |
"learning_rate": 6.742485783915516e-06,
|
| 813 |
+
"loss": 1.4403,
|
| 814 |
"step": 1150
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.4712573633963031,
|
| 818 |
+
"grad_norm": 0.6718817353248596,
|
| 819 |
"learning_rate": 5.930138099106418e-06,
|
| 820 |
+
"loss": 1.2884,
|
| 821 |
"step": 1160
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.4753199268738574,
|
| 825 |
+
"grad_norm": 1.6381056308746338,
|
| 826 |
"learning_rate": 5.117790414297319e-06,
|
| 827 |
+
"loss": 1.4508,
|
| 828 |
"step": 1170
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.4793824903514117,
|
| 832 |
+
"grad_norm": 1.0323442220687866,
|
| 833 |
"learning_rate": 4.305442729488221e-06,
|
| 834 |
+
"loss": 1.4087,
|
| 835 |
"step": 1180
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.4834450538289661,
|
| 839 |
+
"grad_norm": 1.0119701623916626,
|
| 840 |
"learning_rate": 3.4930950446791225e-06,
|
| 841 |
+
"loss": 1.3611,
|
| 842 |
"step": 1190
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.4875076173065204,
|
| 846 |
+
"grad_norm": 2.113036870956421,
|
| 847 |
"learning_rate": 2.6807473598700244e-06,
|
| 848 |
+
"loss": 1.4355,
|
| 849 |
"step": 1200
|
| 850 |
}
|
| 851 |
],
|
checkpoint-1200/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8df8064174041f177505ba0d3a66e4fc8f5d15861f3d5535980ace500355480f
|
| 3 |
size 5176
|
checkpoint-1231/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 8405472
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:785ee56c23465a10b4464b35d71e6e3a37ecea3efc9900199739964a182f8463
|
| 3 |
size 8405472
|
checkpoint-1231/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4411194
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c30d9b34cc216e0661d95564283e7db79ccc4d29bdc8a770db5b3015e66be07
|
| 3 |
size 4411194
|
checkpoint-1231/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b39d6a8d0d352ee4e87b83934d135a11455b52740783a59a137aa6aadb237215
|
| 3 |
size 14244
|
checkpoint-1231/trainer_state.json
CHANGED
|
@@ -10,863 +10,863 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004062563477554337,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 9.91876523151909e-05,
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.008125126955108674,
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
-
"learning_rate": 9.
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01218769043266301,
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
-
"learning_rate": 9.
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01625025391021735,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
-
"learning_rate": 9.
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.020312817387771683,
|
| 41 |
-
"grad_norm": 2.
|
| 42 |
-
"learning_rate": 9.
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02437538086532602,
|
| 48 |
-
"grad_norm": 1.
|
| 49 |
-
"learning_rate": 9.
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.028437944342880356,
|
| 55 |
-
"grad_norm": 2.
|
| 56 |
-
"learning_rate": 9.
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.0325005078204347,
|
| 62 |
-
"grad_norm": 1.
|
| 63 |
-
"learning_rate": 9.
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03656307129798903,
|
| 69 |
-
"grad_norm": 3.
|
| 70 |
-
"learning_rate": 9.
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.040625634775543366,
|
| 76 |
-
"grad_norm": 1.
|
| 77 |
-
"learning_rate": 9.
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.04468819825309771,
|
| 83 |
-
"grad_norm": 1.
|
| 84 |
-
"learning_rate": 9.
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.04875076173065204,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
-
"learning_rate": 9.
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05281332520820638,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
-
"learning_rate": 8.
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05687588868576071,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
-
"learning_rate": 8.
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06093845216331505,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
-
"learning_rate": 8.
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0650010156408694,
|
| 118 |
-
"grad_norm": 1.
|
| 119 |
-
"learning_rate": 8.
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.06906357911842373,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
-
"learning_rate": 8.
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07312614259597806,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
-
"learning_rate": 8.
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.0771887060735324,
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
-
"learning_rate": 8.
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08125126955108673,
|
| 146 |
-
"grad_norm": 3.
|
| 147 |
-
"learning_rate": 8.
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.08531383302864107,
|
| 153 |
-
"grad_norm": 1.
|
| 154 |
-
"learning_rate": 8.
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.08937639650619542,
|
| 160 |
-
"grad_norm":
|
| 161 |
-
"learning_rate": 8.
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.09343895998374975,
|
| 167 |
-
"grad_norm":
|
| 168 |
-
"learning_rate": 8.
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.09750152346130408,
|
| 174 |
-
"grad_norm": 1.
|
| 175 |
-
"learning_rate": 8.
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.10156408693885842,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
-
"learning_rate": 7.
|
| 183 |
-
"loss": 1.
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.10562665041641275,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
-
"learning_rate": 7.
|
| 190 |
-
"loss": 1.
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.10968921389396709,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
-
"learning_rate": 7.
|
| 197 |
-
"loss": 1.
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.11375177737152142,
|
| 202 |
-
"grad_norm": 1.
|
| 203 |
-
"learning_rate": 7.
|
| 204 |
-
"loss": 1.
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.11781434084907577,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
-
"learning_rate": 7.
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.1218769043266301,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
-
"learning_rate": 7.
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.12593946780418444,
|
| 223 |
-
"grad_norm": 1.
|
| 224 |
-
"learning_rate": 7.
|
| 225 |
-
"loss": 1.
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.1300020312817388,
|
| 230 |
-
"grad_norm": 2.
|
| 231 |
-
"learning_rate": 7.
|
| 232 |
-
"loss": 1.
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.1340645947592931,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
-
"learning_rate": 7.
|
| 239 |
-
"loss": 1.
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.13812715823684746,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
-
"learning_rate": 7.
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.14218972171440178,
|
| 251 |
-
"grad_norm": 3.
|
| 252 |
-
"learning_rate": 7.
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.14625228519195613,
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate": 7.
|
| 260 |
-
"loss": 1.
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.15031484866951045,
|
| 265 |
-
"grad_norm": 1.
|
| 266 |
-
"learning_rate": 7.
|
| 267 |
-
"loss": 1.
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1543774121470648,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
-
"learning_rate": 6.
|
| 274 |
-
"loss": 1.
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.15843997562461914,
|
| 279 |
-
"grad_norm": 1.
|
| 280 |
-
"learning_rate": 6.
|
| 281 |
-
"loss": 1.
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.16250253910217347,
|
| 286 |
-
"grad_norm": 1.
|
| 287 |
-
"learning_rate": 6.
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.16656510257972781,
|
| 293 |
-
"grad_norm": 1.
|
| 294 |
-
"learning_rate": 6.
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.17062766605728213,
|
| 300 |
-
"grad_norm":
|
| 301 |
-
"learning_rate": 6.
|
| 302 |
-
"loss": 1.
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.17469022953483648,
|
| 307 |
-
"grad_norm": 1.
|
| 308 |
-
"learning_rate": 6.
|
| 309 |
-
"loss": 1.
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.17875279301239083,
|
| 314 |
-
"grad_norm": 2.
|
| 315 |
-
"learning_rate": 6.
|
| 316 |
-
"loss": 1.
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.18281535648994515,
|
| 321 |
-
"grad_norm": 1.
|
| 322 |
-
"learning_rate": 6.
|
| 323 |
-
"loss": 1.
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.1868779199674995,
|
| 328 |
-
"grad_norm":
|
| 329 |
-
"learning_rate": 6.
|
| 330 |
-
"loss": 1.
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.19094048344505382,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
-
"learning_rate": 6.
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.19500304692260817,
|
| 342 |
-
"grad_norm": 3.
|
| 343 |
-
"learning_rate": 6.
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.1990656104001625,
|
| 349 |
-
"grad_norm":
|
| 350 |
-
"learning_rate": 6.
|
| 351 |
-
"loss": 1.
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.20312817387771684,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
-
"learning_rate": 5.
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.2071907373552712,
|
| 363 |
-
"grad_norm": 1.
|
| 364 |
-
"learning_rate": 5.
|
| 365 |
-
"loss": 1.
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.2112533008328255,
|
| 370 |
-
"grad_norm":
|
| 371 |
-
"learning_rate": 5.
|
| 372 |
-
"loss": 1.
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.21531586431037986,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
-
"learning_rate": 5.
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.21937842778793418,
|
| 384 |
-
"grad_norm": 1.
|
| 385 |
-
"learning_rate": 5.
|
| 386 |
-
"loss": 1.
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.22344099126548853,
|
| 391 |
-
"grad_norm":
|
| 392 |
-
"learning_rate": 5.
|
| 393 |
-
"loss": 1.
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.22750355474304285,
|
| 398 |
-
"grad_norm": 1.
|
| 399 |
-
"learning_rate": 5.
|
| 400 |
-
"loss": 1.
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2315661182205972,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
-
"learning_rate": 5.
|
| 407 |
-
"loss": 1.
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.23562868169815154,
|
| 412 |
-
"grad_norm": 1.
|
| 413 |
-
"learning_rate": 5.
|
| 414 |
-
"loss": 1.
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.23969124517570586,
|
| 419 |
-
"grad_norm": 1.
|
| 420 |
-
"learning_rate": 5.
|
| 421 |
-
"loss": 1.
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2437538086532602,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
-
"learning_rate": 5.
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.24781637213081453,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
-
"learning_rate": 5.
|
| 435 |
-
"loss": 1.
|
| 436 |
"step": 610
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.2518789356083689,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
-
"learning_rate": 4.
|
| 442 |
-
"loss": 1.
|
| 443 |
"step": 620
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.25594149908592323,
|
| 447 |
-
"grad_norm": 1.
|
| 448 |
-
"learning_rate": 4.
|
| 449 |
-
"loss": 1.
|
| 450 |
"step": 630
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.2600040625634776,
|
| 454 |
-
"grad_norm": 1.
|
| 455 |
-
"learning_rate": 4.
|
| 456 |
-
"loss": 1.
|
| 457 |
"step": 640
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.26406662604103187,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
"learning_rate": 4.735987002437044e-05,
|
| 463 |
-
"loss": 1.
|
| 464 |
"step": 650
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.2681291895185862,
|
| 468 |
-
"grad_norm": 2.
|
| 469 |
"learning_rate": 4.6547522339561334e-05,
|
| 470 |
-
"loss": 1.
|
| 471 |
"step": 660
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.27219175299614057,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
"learning_rate": 4.573517465475224e-05,
|
| 477 |
-
"loss": 1.
|
| 478 |
"step": 670
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2762543164736949,
|
| 482 |
-
"grad_norm": 1.
|
| 483 |
"learning_rate": 4.492282696994314e-05,
|
| 484 |
"loss": 1.3557,
|
| 485 |
"step": 680
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.28031687995124926,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
"learning_rate": 4.411047928513404e-05,
|
| 491 |
-
"loss": 1.
|
| 492 |
"step": 690
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.28437944342880356,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
"learning_rate": 4.329813160032494e-05,
|
| 498 |
-
"loss": 1.
|
| 499 |
"step": 700
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.2884420069063579,
|
| 503 |
-
"grad_norm": 1.
|
| 504 |
"learning_rate": 4.2485783915515845e-05,
|
| 505 |
-
"loss": 1.
|
| 506 |
"step": 710
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.29250457038391225,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
"learning_rate": 4.167343623070675e-05,
|
| 512 |
-
"loss": 1.
|
| 513 |
"step": 720
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.2965671338614666,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
"learning_rate": 4.0861088545897645e-05,
|
| 519 |
-
"loss": 1.
|
| 520 |
"step": 730
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.3006296973390209,
|
| 524 |
-
"grad_norm": 1.
|
| 525 |
"learning_rate": 4.004874086108855e-05,
|
| 526 |
-
"loss": 1.
|
| 527 |
"step": 740
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.30469226081657524,
|
| 531 |
-
"grad_norm":
|
| 532 |
"learning_rate": 3.923639317627945e-05,
|
| 533 |
-
"loss": 1.
|
| 534 |
"step": 750
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3087548242941296,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
"learning_rate": 3.842404549147035e-05,
|
| 540 |
-
"loss": 1.
|
| 541 |
"step": 760
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.31281738777168394,
|
| 545 |
-
"grad_norm": 1.
|
| 546 |
"learning_rate": 3.761169780666125e-05,
|
| 547 |
-
"loss": 1.
|
| 548 |
"step": 770
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.3168799512492383,
|
| 552 |
-
"grad_norm": 1.
|
| 553 |
"learning_rate": 3.6799350121852156e-05,
|
| 554 |
-
"loss": 1.
|
| 555 |
"step": 780
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.3209425147267926,
|
| 559 |
-
"grad_norm": 1.
|
| 560 |
"learning_rate": 3.598700243704305e-05,
|
| 561 |
-
"loss": 1.
|
| 562 |
"step": 790
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.32500507820434693,
|
| 566 |
-
"grad_norm":
|
| 567 |
"learning_rate": 3.517465475223396e-05,
|
| 568 |
-
"loss": 1.
|
| 569 |
"step": 800
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.3290676416819013,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
"learning_rate": 3.436230706742486e-05,
|
| 575 |
-
"loss": 1.
|
| 576 |
"step": 810
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.33313020515945563,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 3.3549959382615757e-05,
|
| 582 |
-
"loss": 1.
|
| 583 |
"step": 820
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.33719276863701,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
"learning_rate": 3.273761169780667e-05,
|
| 589 |
-
"loss": 1.
|
| 590 |
"step": 830
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.34125533211456427,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
"learning_rate": 3.1925264012997564e-05,
|
| 596 |
-
"loss": 1.
|
| 597 |
"step": 840
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.3453178955921186,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 3.111291632818847e-05,
|
| 603 |
-
"loss": 1.
|
| 604 |
"step": 850
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.34938045906967297,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 3.0300568643379367e-05,
|
| 610 |
-
"loss": 1.
|
| 611 |
"step": 860
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.3534430225472273,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 2.9488220958570267e-05,
|
| 617 |
-
"loss": 1.
|
| 618 |
"step": 870
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.35750558602478166,
|
| 622 |
-
"grad_norm": 1.
|
| 623 |
"learning_rate": 2.867587327376117e-05,
|
| 624 |
-
"loss": 1.
|
| 625 |
"step": 880
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.36156814950233596,
|
| 629 |
-
"grad_norm": 1.
|
| 630 |
"learning_rate": 2.786352558895207e-05,
|
| 631 |
-
"loss": 1.
|
| 632 |
"step": 890
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.3656307129798903,
|
| 636 |
-
"grad_norm": 2.
|
| 637 |
"learning_rate": 2.7051177904142978e-05,
|
| 638 |
-
"loss": 1.
|
| 639 |
"step": 900
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.36969327645744465,
|
| 643 |
-
"grad_norm": 1.
|
| 644 |
"learning_rate": 2.6238830219333875e-05,
|
| 645 |
-
"loss": 1.
|
| 646 |
"step": 910
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.373755839934999,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
"learning_rate": 2.5426482534524775e-05,
|
| 652 |
-
"loss": 1.
|
| 653 |
"step": 920
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.3778184034125533,
|
| 657 |
-
"grad_norm": 1.
|
| 658 |
"learning_rate": 2.461413484971568e-05,
|
| 659 |
-
"loss": 1.
|
| 660 |
"step": 930
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.38188096689010764,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
"learning_rate": 2.3801787164906582e-05,
|
| 666 |
-
"loss": 1.
|
| 667 |
"step": 940
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.385943530367662,
|
| 671 |
-
"grad_norm": 1.
|
| 672 |
"learning_rate": 2.2989439480097482e-05,
|
| 673 |
-
"loss": 1.
|
| 674 |
"step": 950
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.39000609384521634,
|
| 678 |
-
"grad_norm":
|
| 679 |
"learning_rate": 2.2177091795288386e-05,
|
| 680 |
-
"loss": 1.
|
| 681 |
"step": 960
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.3940686573227707,
|
| 685 |
-
"grad_norm": 2.
|
| 686 |
"learning_rate": 2.1364744110479286e-05,
|
| 687 |
-
"loss": 1.
|
| 688 |
"step": 970
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.398131220800325,
|
| 692 |
-
"grad_norm": 1.
|
| 693 |
"learning_rate": 2.0552396425670186e-05,
|
| 694 |
-
"loss": 1.
|
| 695 |
"step": 980
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.40219378427787933,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
"learning_rate": 1.974004874086109e-05,
|
| 701 |
-
"loss": 1.
|
| 702 |
"step": 990
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.4062563477554337,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
"learning_rate": 1.892770105605199e-05,
|
| 708 |
-
"loss": 1.
|
| 709 |
"step": 1000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.410318911232988,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
"learning_rate": 1.8115353371242893e-05,
|
| 715 |
-
"loss": 1.
|
| 716 |
"step": 1010
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.4143814747105424,
|
| 720 |
-
"grad_norm": 1.
|
| 721 |
"learning_rate": 1.7303005686433797e-05,
|
| 722 |
-
"loss": 1.
|
| 723 |
"step": 1020
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.41844403818809667,
|
| 727 |
-
"grad_norm": 1.
|
| 728 |
"learning_rate": 1.6490658001624697e-05,
|
| 729 |
-
"loss": 1.
|
| 730 |
"step": 1030
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.422506601665651,
|
| 734 |
-
"grad_norm":
|
| 735 |
"learning_rate": 1.5678310316815597e-05,
|
| 736 |
-
"loss": 1.
|
| 737 |
"step": 1040
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.42656916514320536,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
"learning_rate": 1.48659626320065e-05,
|
| 743 |
-
"loss": 1.
|
| 744 |
"step": 1050
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.4306317286207597,
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
"learning_rate": 1.4053614947197402e-05,
|
| 750 |
-
"loss": 1.
|
| 751 |
"step": 1060
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.43469429209831406,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
"learning_rate": 1.3241267262388301e-05,
|
| 757 |
-
"loss": 1.
|
| 758 |
"step": 1070
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.43875685557586835,
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
"learning_rate": 1.2428919577579204e-05,
|
| 764 |
-
"loss": 1.
|
| 765 |
"step": 1080
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.4428194190534227,
|
| 769 |
-
"grad_norm": 1.
|
| 770 |
"learning_rate": 1.1616571892770106e-05,
|
| 771 |
-
"loss": 1.
|
| 772 |
"step": 1090
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.44688198253097705,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
"learning_rate": 1.0804224207961008e-05,
|
| 778 |
-
"loss": 1.
|
| 779 |
"step": 1100
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.4509445460085314,
|
| 783 |
-
"grad_norm":
|
| 784 |
"learning_rate": 9.99187652315191e-06,
|
| 785 |
-
"loss": 1.
|
| 786 |
"step": 1110
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.4550071094860857,
|
| 790 |
-
"grad_norm": 1.
|
| 791 |
"learning_rate": 9.179528838342812e-06,
|
| 792 |
-
"loss": 1.
|
| 793 |
"step": 1120
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.45906967296364004,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
"learning_rate": 8.367181153533712e-06,
|
| 799 |
-
"loss": 1.
|
| 800 |
"step": 1130
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.4631322364411944,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
"learning_rate": 7.554833468724615e-06,
|
| 806 |
-
"loss": 1.
|
| 807 |
"step": 1140
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.46719479991874874,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
"learning_rate": 6.742485783915516e-06,
|
| 813 |
-
"loss": 1.
|
| 814 |
"step": 1150
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.4712573633963031,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
"learning_rate": 5.930138099106418e-06,
|
| 820 |
-
"loss": 1.
|
| 821 |
"step": 1160
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.4753199268738574,
|
| 825 |
-
"grad_norm": 1.
|
| 826 |
"learning_rate": 5.117790414297319e-06,
|
| 827 |
-
"loss": 1.
|
| 828 |
"step": 1170
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.4793824903514117,
|
| 832 |
-
"grad_norm": 1.
|
| 833 |
"learning_rate": 4.305442729488221e-06,
|
| 834 |
-
"loss": 1.
|
| 835 |
"step": 1180
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.4834450538289661,
|
| 839 |
-
"grad_norm":
|
| 840 |
"learning_rate": 3.4930950446791225e-06,
|
| 841 |
-
"loss": 1.
|
| 842 |
"step": 1190
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.4875076173065204,
|
| 846 |
-
"grad_norm": 2.
|
| 847 |
"learning_rate": 2.6807473598700244e-06,
|
| 848 |
-
"loss": 1.
|
| 849 |
"step": 1200
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.4915701807840748,
|
| 853 |
-
"grad_norm": 1.
|
| 854 |
"learning_rate": 1.8683996750609262e-06,
|
| 855 |
-
"loss": 1.
|
| 856 |
"step": 1210
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.49563274426162907,
|
| 860 |
-
"grad_norm": 1.
|
| 861 |
"learning_rate": 1.0560519902518278e-06,
|
| 862 |
-
"loss": 1.
|
| 863 |
"step": 1220
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.4996953077391834,
|
| 867 |
-
"grad_norm": 2.
|
| 868 |
"learning_rate": 2.437043054427295e-07,
|
| 869 |
-
"loss": 1.
|
| 870 |
"step": 1230
|
| 871 |
}
|
| 872 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004062563477554337,
|
| 13 |
+
"grad_norm": 1.032738447189331,
|
| 14 |
"learning_rate": 9.91876523151909e-05,
|
| 15 |
+
"loss": 2.0442,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.008125126955108674,
|
| 20 |
+
"grad_norm": 0.8217566609382629,
|
| 21 |
+
"learning_rate": 9.845653939886271e-05,
|
| 22 |
+
"loss": 1.9828,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01218769043266301,
|
| 27 |
+
"grad_norm": 0.6819207668304443,
|
| 28 |
+
"learning_rate": 9.764419171405362e-05,
|
| 29 |
+
"loss": 1.7187,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01625025391021735,
|
| 34 |
+
"grad_norm": 0.9606077075004578,
|
| 35 |
+
"learning_rate": 9.683184402924452e-05,
|
| 36 |
+
"loss": 1.6162,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.020312817387771683,
|
| 41 |
+
"grad_norm": 2.6619677543640137,
|
| 42 |
+
"learning_rate": 9.601949634443542e-05,
|
| 43 |
+
"loss": 1.6977,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02437538086532602,
|
| 48 |
+
"grad_norm": 1.5025429725646973,
|
| 49 |
+
"learning_rate": 9.520714865962633e-05,
|
| 50 |
+
"loss": 1.2874,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.028437944342880356,
|
| 55 |
+
"grad_norm": 2.7792210578918457,
|
| 56 |
+
"learning_rate": 9.439480097481722e-05,
|
| 57 |
+
"loss": 1.4551,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.0325005078204347,
|
| 62 |
+
"grad_norm": 1.557523250579834,
|
| 63 |
+
"learning_rate": 9.358245329000814e-05,
|
| 64 |
+
"loss": 1.4412,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03656307129798903,
|
| 69 |
+
"grad_norm": 3.4479055404663086,
|
| 70 |
+
"learning_rate": 9.277010560519903e-05,
|
| 71 |
+
"loss": 1.4806,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.040625634775543366,
|
| 76 |
+
"grad_norm": 1.282782793045044,
|
| 77 |
+
"learning_rate": 9.195775792038993e-05,
|
| 78 |
+
"loss": 1.3971,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.04468819825309771,
|
| 83 |
+
"grad_norm": 1.6942788362503052,
|
| 84 |
+
"learning_rate": 9.114541023558083e-05,
|
| 85 |
+
"loss": 1.543,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.04875076173065204,
|
| 90 |
+
"grad_norm": 0.7825320363044739,
|
| 91 |
+
"learning_rate": 9.033306255077174e-05,
|
| 92 |
+
"loss": 1.3638,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05281332520820638,
|
| 97 |
+
"grad_norm": 0.7435455918312073,
|
| 98 |
+
"learning_rate": 8.952071486596265e-05,
|
| 99 |
+
"loss": 1.4731,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05687588868576071,
|
| 104 |
+
"grad_norm": 0.5097799897193909,
|
| 105 |
+
"learning_rate": 8.870836718115354e-05,
|
| 106 |
+
"loss": 1.4218,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06093845216331505,
|
| 111 |
+
"grad_norm": 0.8880577087402344,
|
| 112 |
+
"learning_rate": 8.789601949634444e-05,
|
| 113 |
+
"loss": 1.3477,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0650010156408694,
|
| 118 |
+
"grad_norm": 1.1695178747177124,
|
| 119 |
+
"learning_rate": 8.708367181153534e-05,
|
| 120 |
+
"loss": 1.4521,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.06906357911842373,
|
| 125 |
+
"grad_norm": 0.7951045036315918,
|
| 126 |
+
"learning_rate": 8.627132412672623e-05,
|
| 127 |
+
"loss": 1.3417,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07312614259597806,
|
| 132 |
+
"grad_norm": 0.5302656888961792,
|
| 133 |
+
"learning_rate": 8.545897644191714e-05,
|
| 134 |
+
"loss": 1.5202,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.0771887060735324,
|
| 139 |
+
"grad_norm": 0.8830013275146484,
|
| 140 |
+
"learning_rate": 8.464662875710805e-05,
|
| 141 |
+
"loss": 1.4271,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08125126955108673,
|
| 146 |
+
"grad_norm": 3.1300947666168213,
|
| 147 |
+
"learning_rate": 8.383428107229895e-05,
|
| 148 |
+
"loss": 1.3743,
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.08531383302864107,
|
| 153 |
+
"grad_norm": 1.077304720878601,
|
| 154 |
+
"learning_rate": 8.302193338748985e-05,
|
| 155 |
+
"loss": 1.4244,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.08937639650619542,
|
| 160 |
+
"grad_norm": 1.6846965551376343,
|
| 161 |
+
"learning_rate": 8.220958570268074e-05,
|
| 162 |
+
"loss": 1.4471,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.09343895998374975,
|
| 167 |
+
"grad_norm": 1.066351056098938,
|
| 168 |
+
"learning_rate": 8.139723801787165e-05,
|
| 169 |
+
"loss": 1.3723,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.09750152346130408,
|
| 174 |
+
"grad_norm": 1.6999150514602661,
|
| 175 |
+
"learning_rate": 8.058489033306255e-05,
|
| 176 |
+
"loss": 1.368,
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.10156408693885842,
|
| 181 |
+
"grad_norm": 0.7261852622032166,
|
| 182 |
+
"learning_rate": 7.977254264825346e-05,
|
| 183 |
+
"loss": 1.3175,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.10562665041641275,
|
| 188 |
+
"grad_norm": 0.8671672344207764,
|
| 189 |
+
"learning_rate": 7.896019496344436e-05,
|
| 190 |
+
"loss": 1.4264,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.10968921389396709,
|
| 195 |
+
"grad_norm": 0.9500339031219482,
|
| 196 |
+
"learning_rate": 7.814784727863525e-05,
|
| 197 |
+
"loss": 1.2985,
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.11375177737152142,
|
| 202 |
+
"grad_norm": 1.071378469467163,
|
| 203 |
+
"learning_rate": 7.733549959382617e-05,
|
| 204 |
+
"loss": 1.447,
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.11781434084907577,
|
| 209 |
+
"grad_norm": 0.7448930144309998,
|
| 210 |
+
"learning_rate": 7.652315190901706e-05,
|
| 211 |
+
"loss": 1.374,
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.1218769043266301,
|
| 216 |
+
"grad_norm": 0.9604855179786682,
|
| 217 |
+
"learning_rate": 7.571080422420796e-05,
|
| 218 |
+
"loss": 1.4803,
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.12593946780418444,
|
| 223 |
+
"grad_norm": 1.9951509237289429,
|
| 224 |
+
"learning_rate": 7.489845653939887e-05,
|
| 225 |
+
"loss": 1.4477,
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.1300020312817388,
|
| 230 |
+
"grad_norm": 2.3266055583953857,
|
| 231 |
+
"learning_rate": 7.408610885458977e-05,
|
| 232 |
+
"loss": 1.6107,
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.1340645947592931,
|
| 237 |
+
"grad_norm": 0.878081738948822,
|
| 238 |
+
"learning_rate": 7.327376116978068e-05,
|
| 239 |
+
"loss": 1.295,
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.13812715823684746,
|
| 244 |
+
"grad_norm": 0.7642931938171387,
|
| 245 |
+
"learning_rate": 7.246141348497157e-05,
|
| 246 |
+
"loss": 1.4776,
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.14218972171440178,
|
| 251 |
+
"grad_norm": 3.2081470489501953,
|
| 252 |
+
"learning_rate": 7.164906580016247e-05,
|
| 253 |
+
"loss": 1.448,
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.14625228519195613,
|
| 258 |
+
"grad_norm": 0.9164071679115295,
|
| 259 |
+
"learning_rate": 7.083671811535338e-05,
|
| 260 |
+
"loss": 1.3792,
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.15031484866951045,
|
| 265 |
+
"grad_norm": 1.9146902561187744,
|
| 266 |
+
"learning_rate": 7.002437043054428e-05,
|
| 267 |
+
"loss": 1.239,
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1543774121470648,
|
| 272 |
+
"grad_norm": 0.6990465521812439,
|
| 273 |
+
"learning_rate": 6.921202274573519e-05,
|
| 274 |
+
"loss": 1.3211,
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.15843997562461914,
|
| 279 |
+
"grad_norm": 1.7084014415740967,
|
| 280 |
+
"learning_rate": 6.839967506092608e-05,
|
| 281 |
+
"loss": 1.3952,
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.16250253910217347,
|
| 286 |
+
"grad_norm": 1.0569523572921753,
|
| 287 |
+
"learning_rate": 6.758732737611698e-05,
|
| 288 |
+
"loss": 1.385,
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.16656510257972781,
|
| 293 |
+
"grad_norm": 1.4293252229690552,
|
| 294 |
+
"learning_rate": 6.677497969130788e-05,
|
| 295 |
+
"loss": 1.3736,
|
| 296 |
"step": 410
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.17062766605728213,
|
| 300 |
+
"grad_norm": 1.0318830013275146,
|
| 301 |
+
"learning_rate": 6.596263200649879e-05,
|
| 302 |
+
"loss": 1.2315,
|
| 303 |
"step": 420
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.17469022953483648,
|
| 307 |
+
"grad_norm": 1.0838662385940552,
|
| 308 |
+
"learning_rate": 6.515028432168968e-05,
|
| 309 |
+
"loss": 1.5313,
|
| 310 |
"step": 430
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.17875279301239083,
|
| 314 |
+
"grad_norm": 2.388718366622925,
|
| 315 |
+
"learning_rate": 6.43379366368806e-05,
|
| 316 |
+
"loss": 1.5047,
|
| 317 |
"step": 440
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.18281535648994515,
|
| 321 |
+
"grad_norm": 1.58635413646698,
|
| 322 |
+
"learning_rate": 6.352558895207149e-05,
|
| 323 |
+
"loss": 1.2847,
|
| 324 |
"step": 450
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.1868779199674995,
|
| 328 |
+
"grad_norm": 4.313031196594238,
|
| 329 |
+
"learning_rate": 6.271324126726239e-05,
|
| 330 |
+
"loss": 1.46,
|
| 331 |
"step": 460
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.19094048344505382,
|
| 335 |
+
"grad_norm": 0.5860478281974792,
|
| 336 |
+
"learning_rate": 6.190089358245328e-05,
|
| 337 |
+
"loss": 1.3614,
|
| 338 |
"step": 470
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.19500304692260817,
|
| 342 |
+
"grad_norm": 3.0871775150299072,
|
| 343 |
+
"learning_rate": 6.10885458976442e-05,
|
| 344 |
+
"loss": 1.3915,
|
| 345 |
"step": 480
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.1990656104001625,
|
| 349 |
+
"grad_norm": 2.025517463684082,
|
| 350 |
+
"learning_rate": 6.02761982128351e-05,
|
| 351 |
+
"loss": 1.3326,
|
| 352 |
"step": 490
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.20312817387771684,
|
| 356 |
+
"grad_norm": 0.8149346709251404,
|
| 357 |
+
"learning_rate": 5.9463850528026e-05,
|
| 358 |
+
"loss": 1.2553,
|
| 359 |
"step": 500
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.2071907373552712,
|
| 363 |
+
"grad_norm": 1.1703840494155884,
|
| 364 |
+
"learning_rate": 5.86515028432169e-05,
|
| 365 |
+
"loss": 1.4137,
|
| 366 |
"step": 510
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.2112533008328255,
|
| 370 |
+
"grad_norm": 2.0241522789001465,
|
| 371 |
+
"learning_rate": 5.7839155158407796e-05,
|
| 372 |
+
"loss": 1.3763,
|
| 373 |
"step": 520
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.21531586431037986,
|
| 377 |
+
"grad_norm": 0.5776875019073486,
|
| 378 |
+
"learning_rate": 5.70268074735987e-05,
|
| 379 |
+
"loss": 1.282,
|
| 380 |
"step": 530
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.21937842778793418,
|
| 384 |
+
"grad_norm": 1.644302487373352,
|
| 385 |
+
"learning_rate": 5.621445978878961e-05,
|
| 386 |
+
"loss": 1.3309,
|
| 387 |
"step": 540
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.22344099126548853,
|
| 391 |
+
"grad_norm": 1.0667213201522827,
|
| 392 |
+
"learning_rate": 5.5402112103980506e-05,
|
| 393 |
+
"loss": 1.4516,
|
| 394 |
"step": 550
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.22750355474304285,
|
| 398 |
+
"grad_norm": 1.1769509315490723,
|
| 399 |
+
"learning_rate": 5.458976441917141e-05,
|
| 400 |
+
"loss": 1.4305,
|
| 401 |
"step": 560
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2315661182205972,
|
| 405 |
+
"grad_norm": 0.9089193940162659,
|
| 406 |
+
"learning_rate": 5.377741673436231e-05,
|
| 407 |
+
"loss": 1.3263,
|
| 408 |
"step": 570
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.23562868169815154,
|
| 412 |
+
"grad_norm": 1.9467777013778687,
|
| 413 |
+
"learning_rate": 5.2965069049553203e-05,
|
| 414 |
+
"loss": 1.402,
|
| 415 |
"step": 580
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.23969124517570586,
|
| 419 |
+
"grad_norm": 1.313186526298523,
|
| 420 |
+
"learning_rate": 5.2152721364744114e-05,
|
| 421 |
+
"loss": 1.3604,
|
| 422 |
"step": 590
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2437538086532602,
|
| 426 |
+
"grad_norm": 0.48065802454948425,
|
| 427 |
+
"learning_rate": 5.134037367993502e-05,
|
| 428 |
+
"loss": 1.5186,
|
| 429 |
"step": 600
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.24781637213081453,
|
| 433 |
+
"grad_norm": 0.7379820942878723,
|
| 434 |
+
"learning_rate": 5.0528025995125914e-05,
|
| 435 |
+
"loss": 1.4457,
|
| 436 |
"step": 610
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.2518789356083689,
|
| 440 |
+
"grad_norm": 0.7363265156745911,
|
| 441 |
+
"learning_rate": 4.971567831031682e-05,
|
| 442 |
+
"loss": 1.3843,
|
| 443 |
"step": 620
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.25594149908592323,
|
| 447 |
+
"grad_norm": 1.349368691444397,
|
| 448 |
+
"learning_rate": 4.890333062550772e-05,
|
| 449 |
+
"loss": 1.355,
|
| 450 |
"step": 630
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.2600040625634776,
|
| 454 |
+
"grad_norm": 1.6113736629486084,
|
| 455 |
+
"learning_rate": 4.809098294069862e-05,
|
| 456 |
+
"loss": 1.5113,
|
| 457 |
"step": 640
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.26406662604103187,
|
| 461 |
+
"grad_norm": 0.7977277636528015,
|
| 462 |
"learning_rate": 4.735987002437044e-05,
|
| 463 |
+
"loss": 1.4413,
|
| 464 |
"step": 650
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.2681291895185862,
|
| 468 |
+
"grad_norm": 2.2527716159820557,
|
| 469 |
"learning_rate": 4.6547522339561334e-05,
|
| 470 |
+
"loss": 1.4792,
|
| 471 |
"step": 660
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.27219175299614057,
|
| 475 |
+
"grad_norm": 0.7665427923202515,
|
| 476 |
"learning_rate": 4.573517465475224e-05,
|
| 477 |
+
"loss": 1.2581,
|
| 478 |
"step": 670
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2762543164736949,
|
| 482 |
+
"grad_norm": 1.6001981496810913,
|
| 483 |
"learning_rate": 4.492282696994314e-05,
|
| 484 |
"loss": 1.3557,
|
| 485 |
"step": 680
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.28031687995124926,
|
| 489 |
+
"grad_norm": 0.6028335690498352,
|
| 490 |
"learning_rate": 4.411047928513404e-05,
|
| 491 |
+
"loss": 1.4625,
|
| 492 |
"step": 690
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.28437944342880356,
|
| 496 |
+
"grad_norm": 0.621457576751709,
|
| 497 |
"learning_rate": 4.329813160032494e-05,
|
| 498 |
+
"loss": 1.4177,
|
| 499 |
"step": 700
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.2884420069063579,
|
| 503 |
+
"grad_norm": 1.08269464969635,
|
| 504 |
"learning_rate": 4.2485783915515845e-05,
|
| 505 |
+
"loss": 1.5553,
|
| 506 |
"step": 710
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.29250457038391225,
|
| 510 |
+
"grad_norm": 0.7009453177452087,
|
| 511 |
"learning_rate": 4.167343623070675e-05,
|
| 512 |
+
"loss": 1.3478,
|
| 513 |
"step": 720
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.2965671338614666,
|
| 517 |
+
"grad_norm": 0.7686846256256104,
|
| 518 |
"learning_rate": 4.0861088545897645e-05,
|
| 519 |
+
"loss": 1.5388,
|
| 520 |
"step": 730
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.3006296973390209,
|
| 524 |
+
"grad_norm": 1.193659782409668,
|
| 525 |
"learning_rate": 4.004874086108855e-05,
|
| 526 |
+
"loss": 1.558,
|
| 527 |
"step": 740
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.30469226081657524,
|
| 531 |
+
"grad_norm": 0.9236335754394531,
|
| 532 |
"learning_rate": 3.923639317627945e-05,
|
| 533 |
+
"loss": 1.4143,
|
| 534 |
"step": 750
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3087548242941296,
|
| 538 |
+
"grad_norm": 0.7839071750640869,
|
| 539 |
"learning_rate": 3.842404549147035e-05,
|
| 540 |
+
"loss": 1.2117,
|
| 541 |
"step": 760
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.31281738777168394,
|
| 545 |
+
"grad_norm": 1.2941004037857056,
|
| 546 |
"learning_rate": 3.761169780666125e-05,
|
| 547 |
+
"loss": 1.4062,
|
| 548 |
"step": 770
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.3168799512492383,
|
| 552 |
+
"grad_norm": 1.426131010055542,
|
| 553 |
"learning_rate": 3.6799350121852156e-05,
|
| 554 |
+
"loss": 1.4909,
|
| 555 |
"step": 780
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.3209425147267926,
|
| 559 |
+
"grad_norm": 1.1472172737121582,
|
| 560 |
"learning_rate": 3.598700243704305e-05,
|
| 561 |
+
"loss": 1.2864,
|
| 562 |
"step": 790
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.32500507820434693,
|
| 566 |
+
"grad_norm": 2.1125993728637695,
|
| 567 |
"learning_rate": 3.517465475223396e-05,
|
| 568 |
+
"loss": 1.3456,
|
| 569 |
"step": 800
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.3290676416819013,
|
| 573 |
+
"grad_norm": 0.7743927240371704,
|
| 574 |
"learning_rate": 3.436230706742486e-05,
|
| 575 |
+
"loss": 1.3294,
|
| 576 |
"step": 810
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.33313020515945563,
|
| 580 |
+
"grad_norm": 0.8188801407814026,
|
| 581 |
"learning_rate": 3.3549959382615757e-05,
|
| 582 |
+
"loss": 1.2803,
|
| 583 |
"step": 820
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.33719276863701,
|
| 587 |
+
"grad_norm": 0.5288794636726379,
|
| 588 |
"learning_rate": 3.273761169780667e-05,
|
| 589 |
+
"loss": 1.4094,
|
| 590 |
"step": 830
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.34125533211456427,
|
| 594 |
+
"grad_norm": 0.6982353329658508,
|
| 595 |
"learning_rate": 3.1925264012997564e-05,
|
| 596 |
+
"loss": 1.1621,
|
| 597 |
"step": 840
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.3453178955921186,
|
| 601 |
+
"grad_norm": 0.7790868282318115,
|
| 602 |
"learning_rate": 3.111291632818847e-05,
|
| 603 |
+
"loss": 1.4534,
|
| 604 |
"step": 850
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.34938045906967297,
|
| 608 |
+
"grad_norm": 0.7421971559524536,
|
| 609 |
"learning_rate": 3.0300568643379367e-05,
|
| 610 |
+
"loss": 1.2951,
|
| 611 |
"step": 860
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.3534430225472273,
|
| 615 |
+
"grad_norm": 0.9682130813598633,
|
| 616 |
"learning_rate": 2.9488220958570267e-05,
|
| 617 |
+
"loss": 1.3827,
|
| 618 |
"step": 870
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.35750558602478166,
|
| 622 |
+
"grad_norm": 1.6505218744277954,
|
| 623 |
"learning_rate": 2.867587327376117e-05,
|
| 624 |
+
"loss": 1.3482,
|
| 625 |
"step": 880
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.36156814950233596,
|
| 629 |
+
"grad_norm": 1.013390064239502,
|
| 630 |
"learning_rate": 2.786352558895207e-05,
|
| 631 |
+
"loss": 1.4224,
|
| 632 |
"step": 890
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.3656307129798903,
|
| 636 |
+
"grad_norm": 2.4053940773010254,
|
| 637 |
"learning_rate": 2.7051177904142978e-05,
|
| 638 |
+
"loss": 1.4791,
|
| 639 |
"step": 900
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.36969327645744465,
|
| 643 |
+
"grad_norm": 1.7795937061309814,
|
| 644 |
"learning_rate": 2.6238830219333875e-05,
|
| 645 |
+
"loss": 1.472,
|
| 646 |
"step": 910
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.373755839934999,
|
| 650 |
+
"grad_norm": 0.7656849026679993,
|
| 651 |
"learning_rate": 2.5426482534524775e-05,
|
| 652 |
+
"loss": 1.3824,
|
| 653 |
"step": 920
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.3778184034125533,
|
| 657 |
+
"grad_norm": 1.1062685251235962,
|
| 658 |
"learning_rate": 2.461413484971568e-05,
|
| 659 |
+
"loss": 1.3462,
|
| 660 |
"step": 930
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.38188096689010764,
|
| 664 |
+
"grad_norm": 0.660727858543396,
|
| 665 |
"learning_rate": 2.3801787164906582e-05,
|
| 666 |
+
"loss": 1.3834,
|
| 667 |
"step": 940
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.385943530367662,
|
| 671 |
+
"grad_norm": 1.8191992044448853,
|
| 672 |
"learning_rate": 2.2989439480097482e-05,
|
| 673 |
+
"loss": 1.4556,
|
| 674 |
"step": 950
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.39000609384521634,
|
| 678 |
+
"grad_norm": 3.3901174068450928,
|
| 679 |
"learning_rate": 2.2177091795288386e-05,
|
| 680 |
+
"loss": 1.2865,
|
| 681 |
"step": 960
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.3940686573227707,
|
| 685 |
+
"grad_norm": 2.4866228103637695,
|
| 686 |
"learning_rate": 2.1364744110479286e-05,
|
| 687 |
+
"loss": 1.501,
|
| 688 |
"step": 970
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.398131220800325,
|
| 692 |
+
"grad_norm": 1.41251540184021,
|
| 693 |
"learning_rate": 2.0552396425670186e-05,
|
| 694 |
+
"loss": 1.3455,
|
| 695 |
"step": 980
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.40219378427787933,
|
| 699 |
+
"grad_norm": 0.645084798336029,
|
| 700 |
"learning_rate": 1.974004874086109e-05,
|
| 701 |
+
"loss": 1.3918,
|
| 702 |
"step": 990
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.4062563477554337,
|
| 706 |
+
"grad_norm": 0.8688336610794067,
|
| 707 |
"learning_rate": 1.892770105605199e-05,
|
| 708 |
+
"loss": 1.3823,
|
| 709 |
"step": 1000
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.410318911232988,
|
| 713 |
+
"grad_norm": 0.7747243642807007,
|
| 714 |
"learning_rate": 1.8115353371242893e-05,
|
| 715 |
+
"loss": 1.2988,
|
| 716 |
"step": 1010
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.4143814747105424,
|
| 720 |
+
"grad_norm": 1.329567313194275,
|
| 721 |
"learning_rate": 1.7303005686433797e-05,
|
| 722 |
+
"loss": 1.2876,
|
| 723 |
"step": 1020
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.41844403818809667,
|
| 727 |
+
"grad_norm": 1.022557258605957,
|
| 728 |
"learning_rate": 1.6490658001624697e-05,
|
| 729 |
+
"loss": 1.3036,
|
| 730 |
"step": 1030
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.422506601665651,
|
| 734 |
+
"grad_norm": 0.9812535047531128,
|
| 735 |
"learning_rate": 1.5678310316815597e-05,
|
| 736 |
+
"loss": 1.3278,
|
| 737 |
"step": 1040
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.42656916514320536,
|
| 741 |
+
"grad_norm": 0.8079454898834229,
|
| 742 |
"learning_rate": 1.48659626320065e-05,
|
| 743 |
+
"loss": 1.2957,
|
| 744 |
"step": 1050
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.4306317286207597,
|
| 748 |
+
"grad_norm": 0.8280611038208008,
|
| 749 |
"learning_rate": 1.4053614947197402e-05,
|
| 750 |
+
"loss": 1.3747,
|
| 751 |
"step": 1060
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.43469429209831406,
|
| 755 |
+
"grad_norm": 0.85203617811203,
|
| 756 |
"learning_rate": 1.3241267262388301e-05,
|
| 757 |
+
"loss": 1.423,
|
| 758 |
"step": 1070
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.43875685557586835,
|
| 762 |
+
"grad_norm": 0.9492881298065186,
|
| 763 |
"learning_rate": 1.2428919577579204e-05,
|
| 764 |
+
"loss": 1.3845,
|
| 765 |
"step": 1080
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.4428194190534227,
|
| 769 |
+
"grad_norm": 1.6084877252578735,
|
| 770 |
"learning_rate": 1.1616571892770106e-05,
|
| 771 |
+
"loss": 1.2422,
|
| 772 |
"step": 1090
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.44688198253097705,
|
| 776 |
+
"grad_norm": 0.7060315012931824,
|
| 777 |
"learning_rate": 1.0804224207961008e-05,
|
| 778 |
+
"loss": 1.3758,
|
| 779 |
"step": 1100
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.4509445460085314,
|
| 783 |
+
"grad_norm": 2.0847179889678955,
|
| 784 |
"learning_rate": 9.99187652315191e-06,
|
| 785 |
+
"loss": 1.4202,
|
| 786 |
"step": 1110
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.4550071094860857,
|
| 790 |
+
"grad_norm": 1.1350940465927124,
|
| 791 |
"learning_rate": 9.179528838342812e-06,
|
| 792 |
+
"loss": 1.3273,
|
| 793 |
"step": 1120
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.45906967296364004,
|
| 797 |
+
"grad_norm": 0.841096043586731,
|
| 798 |
"learning_rate": 8.367181153533712e-06,
|
| 799 |
+
"loss": 1.3461,
|
| 800 |
"step": 1130
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.4631322364411944,
|
| 804 |
+
"grad_norm": 0.5926047563552856,
|
| 805 |
"learning_rate": 7.554833468724615e-06,
|
| 806 |
+
"loss": 1.2492,
|
| 807 |
"step": 1140
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.46719479991874874,
|
| 811 |
+
"grad_norm": 0.8289461135864258,
|
| 812 |
"learning_rate": 6.742485783915516e-06,
|
| 813 |
+
"loss": 1.4403,
|
| 814 |
"step": 1150
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.4712573633963031,
|
| 818 |
+
"grad_norm": 0.6718817353248596,
|
| 819 |
"learning_rate": 5.930138099106418e-06,
|
| 820 |
+
"loss": 1.2884,
|
| 821 |
"step": 1160
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.4753199268738574,
|
| 825 |
+
"grad_norm": 1.6381056308746338,
|
| 826 |
"learning_rate": 5.117790414297319e-06,
|
| 827 |
+
"loss": 1.4508,
|
| 828 |
"step": 1170
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.4793824903514117,
|
| 832 |
+
"grad_norm": 1.0323442220687866,
|
| 833 |
"learning_rate": 4.305442729488221e-06,
|
| 834 |
+
"loss": 1.4087,
|
| 835 |
"step": 1180
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.4834450538289661,
|
| 839 |
+
"grad_norm": 1.0119701623916626,
|
| 840 |
"learning_rate": 3.4930950446791225e-06,
|
| 841 |
+
"loss": 1.3611,
|
| 842 |
"step": 1190
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.4875076173065204,
|
| 846 |
+
"grad_norm": 2.113036870956421,
|
| 847 |
"learning_rate": 2.6807473598700244e-06,
|
| 848 |
+
"loss": 1.4355,
|
| 849 |
"step": 1200
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.4915701807840748,
|
| 853 |
+
"grad_norm": 1.257258415222168,
|
| 854 |
"learning_rate": 1.8683996750609262e-06,
|
| 855 |
+
"loss": 1.2809,
|
| 856 |
"step": 1210
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.49563274426162907,
|
| 860 |
+
"grad_norm": 1.6237741708755493,
|
| 861 |
"learning_rate": 1.0560519902518278e-06,
|
| 862 |
+
"loss": 1.2713,
|
| 863 |
"step": 1220
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.4996953077391834,
|
| 867 |
+
"grad_norm": 2.050067901611328,
|
| 868 |
"learning_rate": 2.437043054427295e-07,
|
| 869 |
+
"loss": 1.4235,
|
| 870 |
"step": 1230
|
| 871 |
}
|
| 872 |
],
|
checkpoint-1231/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8df8064174041f177505ba0d3a66e4fc8f5d15861f3d5535980ace500355480f
|
| 3 |
size 5176
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8df8064174041f177505ba0d3a66e4fc8f5d15861f3d5535980ace500355480f
|
| 3 |
size 5176
|