Delete checkpoint-1400/trainer_state.json with huggingface_hub
Browse files- checkpoint-1400/trainer_state.json +0 -1076
checkpoint-1400/trainer_state.json
DELETED
|
@@ -1,1076 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": 0.6909542083740234,
|
| 3 |
-
"best_model_checkpoint": "/workspace/rm/rm_adapter/checkpoint-1400",
|
| 4 |
-
"epoch": 2.121212121212121,
|
| 5 |
-
"eval_steps": 200,
|
| 6 |
-
"global_step": 1400,
|
| 7 |
-
"is_hyper_param_search": false,
|
| 8 |
-
"is_local_process_zero": true,
|
| 9 |
-
"is_world_process_zero": true,
|
| 10 |
-
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.015151515151515152,
|
| 13 |
-
"grad_norm": 3.026841640472412,
|
| 14 |
-
"learning_rate": 2e-05,
|
| 15 |
-
"loss": 0.6999,
|
| 16 |
-
"step": 10
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.030303030303030304,
|
| 20 |
-
"grad_norm": 2.487321138381958,
|
| 21 |
-
"learning_rate": 4e-05,
|
| 22 |
-
"loss": 0.6998,
|
| 23 |
-
"step": 20
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.045454545454545456,
|
| 27 |
-
"grad_norm": 2.6206929683685303,
|
| 28 |
-
"learning_rate": 6e-05,
|
| 29 |
-
"loss": 0.6957,
|
| 30 |
-
"step": 30
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.06060606060606061,
|
| 34 |
-
"grad_norm": 1.7623201608657837,
|
| 35 |
-
"learning_rate": 8e-05,
|
| 36 |
-
"loss": 0.6922,
|
| 37 |
-
"step": 40
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.07575757575757576,
|
| 41 |
-
"grad_norm": 1.821882724761963,
|
| 42 |
-
"learning_rate": 0.0001,
|
| 43 |
-
"loss": 0.695,
|
| 44 |
-
"step": 50
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.09090909090909091,
|
| 48 |
-
"grad_norm": 2.205653190612793,
|
| 49 |
-
"learning_rate": 9.99933760728612e-05,
|
| 50 |
-
"loss": 0.7181,
|
| 51 |
-
"step": 60
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.10606060606060606,
|
| 55 |
-
"grad_norm": 2.080017328262329,
|
| 56 |
-
"learning_rate": 9.997350604650123e-05,
|
| 57 |
-
"loss": 0.6952,
|
| 58 |
-
"step": 70
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.12121212121212122,
|
| 62 |
-
"grad_norm": 2.1524713039398193,
|
| 63 |
-
"learning_rate": 9.994039518562432e-05,
|
| 64 |
-
"loss": 0.6938,
|
| 65 |
-
"step": 80
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.13636363636363635,
|
| 69 |
-
"grad_norm": 2.467346668243408,
|
| 70 |
-
"learning_rate": 9.989405226318772e-05,
|
| 71 |
-
"loss": 0.6988,
|
| 72 |
-
"step": 90
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.15151515151515152,
|
| 76 |
-
"grad_norm": 2.5009102821350098,
|
| 77 |
-
"learning_rate": 9.983448955807708e-05,
|
| 78 |
-
"loss": 0.6819,
|
| 79 |
-
"step": 100
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.16666666666666666,
|
| 83 |
-
"grad_norm": 2.6343142986297607,
|
| 84 |
-
"learning_rate": 9.976172285185314e-05,
|
| 85 |
-
"loss": 0.6891,
|
| 86 |
-
"step": 110
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.18181818181818182,
|
| 90 |
-
"grad_norm": 2.9957756996154785,
|
| 91 |
-
"learning_rate": 9.967577142457032e-05,
|
| 92 |
-
"loss": 0.7024,
|
| 93 |
-
"step": 120
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.19696969696969696,
|
| 97 |
-
"grad_norm": 2.6430771350860596,
|
| 98 |
-
"learning_rate": 9.957665804966829e-05,
|
| 99 |
-
"loss": 0.7012,
|
| 100 |
-
"step": 130
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.21212121212121213,
|
| 104 |
-
"grad_norm": 2.6395034790039062,
|
| 105 |
-
"learning_rate": 9.946440898793801e-05,
|
| 106 |
-
"loss": 0.6826,
|
| 107 |
-
"step": 140
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.22727272727272727,
|
| 111 |
-
"grad_norm": 2.5003907680511475,
|
| 112 |
-
"learning_rate": 9.933905398056372e-05,
|
| 113 |
-
"loss": 0.7,
|
| 114 |
-
"step": 150
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.24242424242424243,
|
| 118 |
-
"grad_norm": 2.4217541217803955,
|
| 119 |
-
"learning_rate": 9.920062624124282e-05,
|
| 120 |
-
"loss": 0.6932,
|
| 121 |
-
"step": 160
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.25757575757575757,
|
| 125 |
-
"grad_norm": 3.947263240814209,
|
| 126 |
-
"learning_rate": 9.904916244738571e-05,
|
| 127 |
-
"loss": 0.7003,
|
| 128 |
-
"step": 170
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.2727272727272727,
|
| 132 |
-
"grad_norm": 3.162750244140625,
|
| 133 |
-
"learning_rate": 9.888470273039775e-05,
|
| 134 |
-
"loss": 0.6756,
|
| 135 |
-
"step": 180
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.2878787878787879,
|
| 139 |
-
"grad_norm": 2.7067689895629883,
|
| 140 |
-
"learning_rate": 9.870729066504629e-05,
|
| 141 |
-
"loss": 0.7031,
|
| 142 |
-
"step": 190
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.30303030303030304,
|
| 146 |
-
"grad_norm": 3.415238618850708,
|
| 147 |
-
"learning_rate": 9.851697325791505e-05,
|
| 148 |
-
"loss": 0.6888,
|
| 149 |
-
"step": 200
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.30303030303030304,
|
| 153 |
-
"eval_accuracy": 0.7482014388489209,
|
| 154 |
-
"eval_loss": 0.6937351822853088,
|
| 155 |
-
"eval_runtime": 19.5228,
|
| 156 |
-
"eval_samples_per_second": 14.24,
|
| 157 |
-
"eval_steps_per_second": 1.793,
|
| 158 |
-
"step": 200
|
| 159 |
-
},
|
| 160 |
-
{
|
| 161 |
-
"epoch": 0.3181818181818182,
|
| 162 |
-
"grad_norm": 3.3440141677856445,
|
| 163 |
-
"learning_rate": 9.831380093494957e-05,
|
| 164 |
-
"loss": 0.6784,
|
| 165 |
-
"step": 210
|
| 166 |
-
},
|
| 167 |
-
{
|
| 168 |
-
"epoch": 0.3333333333333333,
|
| 169 |
-
"grad_norm": 3.1289377212524414,
|
| 170 |
-
"learning_rate": 9.809782752809644e-05,
|
| 171 |
-
"loss": 0.7005,
|
| 172 |
-
"step": 220
|
| 173 |
-
},
|
| 174 |
-
{
|
| 175 |
-
"epoch": 0.3484848484848485,
|
| 176 |
-
"grad_norm": 4.689096927642822,
|
| 177 |
-
"learning_rate": 9.786911026104007e-05,
|
| 178 |
-
"loss": 0.7015,
|
| 179 |
-
"step": 230
|
| 180 |
-
},
|
| 181 |
-
{
|
| 182 |
-
"epoch": 0.36363636363636365,
|
| 183 |
-
"grad_norm": 3.6249144077301025,
|
| 184 |
-
"learning_rate": 9.762770973404094e-05,
|
| 185 |
-
"loss": 0.6818,
|
| 186 |
-
"step": 240
|
| 187 |
-
},
|
| 188 |
-
{
|
| 189 |
-
"epoch": 0.3787878787878788,
|
| 190 |
-
"grad_norm": 3.3653128147125244,
|
| 191 |
-
"learning_rate": 9.737368990787916e-05,
|
| 192 |
-
"loss": 0.706,
|
| 193 |
-
"step": 250
|
| 194 |
-
},
|
| 195 |
-
{
|
| 196 |
-
"epoch": 0.3939393939393939,
|
| 197 |
-
"grad_norm": 3.5432941913604736,
|
| 198 |
-
"learning_rate": 9.710711808690754e-05,
|
| 199 |
-
"loss": 0.7136,
|
| 200 |
-
"step": 260
|
| 201 |
-
},
|
| 202 |
-
{
|
| 203 |
-
"epoch": 0.4090909090909091,
|
| 204 |
-
"grad_norm": 3.5209529399871826,
|
| 205 |
-
"learning_rate": 9.682806490121885e-05,
|
| 206 |
-
"loss": 0.7134,
|
| 207 |
-
"step": 270
|
| 208 |
-
},
|
| 209 |
-
{
|
| 210 |
-
"epoch": 0.42424242424242425,
|
| 211 |
-
"grad_norm": 2.654764413833618,
|
| 212 |
-
"learning_rate": 9.653660428793188e-05,
|
| 213 |
-
"loss": 0.714,
|
| 214 |
-
"step": 280
|
| 215 |
-
},
|
| 216 |
-
{
|
| 217 |
-
"epoch": 0.4393939393939394,
|
| 218 |
-
"grad_norm": 6.175410747528076,
|
| 219 |
-
"learning_rate": 9.623281347160127e-05,
|
| 220 |
-
"loss": 0.7095,
|
| 221 |
-
"step": 290
|
| 222 |
-
},
|
| 223 |
-
{
|
| 224 |
-
"epoch": 0.45454545454545453,
|
| 225 |
-
"grad_norm": 2.3064019680023193,
|
| 226 |
-
"learning_rate": 9.591677294375636e-05,
|
| 227 |
-
"loss": 0.6922,
|
| 228 |
-
"step": 300
|
| 229 |
-
},
|
| 230 |
-
{
|
| 231 |
-
"epoch": 0.4696969696969697,
|
| 232 |
-
"grad_norm": 2.248598337173462,
|
| 233 |
-
"learning_rate": 9.558856644157432e-05,
|
| 234 |
-
"loss": 0.6999,
|
| 235 |
-
"step": 310
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.48484848484848486,
|
| 239 |
-
"grad_norm": 2.52163028717041,
|
| 240 |
-
"learning_rate": 9.52482809256934e-05,
|
| 241 |
-
"loss": 0.7185,
|
| 242 |
-
"step": 320
|
| 243 |
-
},
|
| 244 |
-
{
|
| 245 |
-
"epoch": 0.5,
|
| 246 |
-
"grad_norm": 1.9855085611343384,
|
| 247 |
-
"learning_rate": 9.489600655717217e-05,
|
| 248 |
-
"loss": 0.7,
|
| 249 |
-
"step": 330
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.5151515151515151,
|
| 253 |
-
"grad_norm": 2.6163477897644043,
|
| 254 |
-
"learning_rate": 9.453183667360062e-05,
|
| 255 |
-
"loss": 0.6893,
|
| 256 |
-
"step": 340
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 0.5303030303030303,
|
| 260 |
-
"grad_norm": 2.32633900642395,
|
| 261 |
-
"learning_rate": 9.415586776436973e-05,
|
| 262 |
-
"loss": 0.7029,
|
| 263 |
-
"step": 350
|
| 264 |
-
},
|
| 265 |
-
{
|
| 266 |
-
"epoch": 0.5454545454545454,
|
| 267 |
-
"grad_norm": 2.0239343643188477,
|
| 268 |
-
"learning_rate": 9.376819944510598e-05,
|
| 269 |
-
"loss": 0.6859,
|
| 270 |
-
"step": 360
|
| 271 |
-
},
|
| 272 |
-
{
|
| 273 |
-
"epoch": 0.5606060606060606,
|
| 274 |
-
"grad_norm": 2.0154311656951904,
|
| 275 |
-
"learning_rate": 9.336893443127738e-05,
|
| 276 |
-
"loss": 0.6942,
|
| 277 |
-
"step": 370
|
| 278 |
-
},
|
| 279 |
-
{
|
| 280 |
-
"epoch": 0.5757575757575758,
|
| 281 |
-
"grad_norm": 2.040499210357666,
|
| 282 |
-
"learning_rate": 9.295817851097837e-05,
|
| 283 |
-
"loss": 0.6988,
|
| 284 |
-
"step": 380
|
| 285 |
-
},
|
| 286 |
-
{
|
| 287 |
-
"epoch": 0.5909090909090909,
|
| 288 |
-
"grad_norm": 2.244065999984741,
|
| 289 |
-
"learning_rate": 9.253604051690046e-05,
|
| 290 |
-
"loss": 0.6984,
|
| 291 |
-
"step": 390
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.6060606060606061,
|
| 295 |
-
"grad_norm": 2.4374303817749023,
|
| 296 |
-
"learning_rate": 9.210263229749626e-05,
|
| 297 |
-
"loss": 0.6863,
|
| 298 |
-
"step": 400
|
| 299 |
-
},
|
| 300 |
-
{
|
| 301 |
-
"epoch": 0.6060606060606061,
|
| 302 |
-
"eval_accuracy": 0.697841726618705,
|
| 303 |
-
"eval_loss": 0.6917553544044495,
|
| 304 |
-
"eval_runtime": 19.5084,
|
| 305 |
-
"eval_samples_per_second": 14.25,
|
| 306 |
-
"eval_steps_per_second": 1.794,
|
| 307 |
-
"step": 400
|
| 308 |
-
},
|
| 309 |
-
{
|
| 310 |
-
"epoch": 0.6212121212121212,
|
| 311 |
-
"grad_norm": 2.305253505706787,
|
| 312 |
-
"learning_rate": 9.165806868734444e-05,
|
| 313 |
-
"loss": 0.7048,
|
| 314 |
-
"step": 410
|
| 315 |
-
},
|
| 316 |
-
{
|
| 317 |
-
"epoch": 0.6363636363636364,
|
| 318 |
-
"grad_norm": 2.419456720352173,
|
| 319 |
-
"learning_rate": 9.120246747672347e-05,
|
| 320 |
-
"loss": 0.7059,
|
| 321 |
-
"step": 420
|
| 322 |
-
},
|
| 323 |
-
{
|
| 324 |
-
"epoch": 0.6515151515151515,
|
| 325 |
-
"grad_norm": 2.2635741233825684,
|
| 326 |
-
"learning_rate": 9.073594938040231e-05,
|
| 327 |
-
"loss": 0.6961,
|
| 328 |
-
"step": 430
|
| 329 |
-
},
|
| 330 |
-
{
|
| 331 |
-
"epoch": 0.6666666666666666,
|
| 332 |
-
"grad_norm": 2.824849843978882,
|
| 333 |
-
"learning_rate": 9.025863800565613e-05,
|
| 334 |
-
"loss": 0.7004,
|
| 335 |
-
"step": 440
|
| 336 |
-
},
|
| 337 |
-
{
|
| 338 |
-
"epoch": 0.6818181818181818,
|
| 339 |
-
"grad_norm": 2.1357405185699463,
|
| 340 |
-
"learning_rate": 8.977065981951566e-05,
|
| 341 |
-
"loss": 0.6968,
|
| 342 |
-
"step": 450
|
| 343 |
-
},
|
| 344 |
-
{
|
| 345 |
-
"epoch": 0.696969696969697,
|
| 346 |
-
"grad_norm": 3.1660122871398926,
|
| 347 |
-
"learning_rate": 8.927214411525895e-05,
|
| 348 |
-
"loss": 0.702,
|
| 349 |
-
"step": 460
|
| 350 |
-
},
|
| 351 |
-
{
|
| 352 |
-
"epoch": 0.7121212121212122,
|
| 353 |
-
"grad_norm": 2.296741247177124,
|
| 354 |
-
"learning_rate": 8.876322297815405e-05,
|
| 355 |
-
"loss": 0.7001,
|
| 356 |
-
"step": 470
|
| 357 |
-
},
|
| 358 |
-
{
|
| 359 |
-
"epoch": 0.7272727272727273,
|
| 360 |
-
"grad_norm": 1.987774133682251,
|
| 361 |
-
"learning_rate": 8.824403125046225e-05,
|
| 362 |
-
"loss": 0.6947,
|
| 363 |
-
"step": 480
|
| 364 |
-
},
|
| 365 |
-
{
|
| 366 |
-
"epoch": 0.7424242424242424,
|
| 367 |
-
"grad_norm": 2.645599603652954,
|
| 368 |
-
"learning_rate": 8.771470649571056e-05,
|
| 369 |
-
"loss": 0.7026,
|
| 370 |
-
"step": 490
|
| 371 |
-
},
|
| 372 |
-
{
|
| 373 |
-
"epoch": 0.7575757575757576,
|
| 374 |
-
"grad_norm": 2.0273940563201904,
|
| 375 |
-
"learning_rate": 8.717538896224332e-05,
|
| 376 |
-
"loss": 0.703,
|
| 377 |
-
"step": 500
|
| 378 |
-
},
|
| 379 |
-
{
|
| 380 |
-
"epoch": 0.7727272727272727,
|
| 381 |
-
"grad_norm": 6.725697994232178,
|
| 382 |
-
"learning_rate": 8.662622154606237e-05,
|
| 383 |
-
"loss": 0.685,
|
| 384 |
-
"step": 510
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 0.7878787878787878,
|
| 388 |
-
"grad_norm": 1.968575358390808,
|
| 389 |
-
"learning_rate": 8.606734975296578e-05,
|
| 390 |
-
"loss": 0.7042,
|
| 391 |
-
"step": 520
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 0.803030303030303,
|
| 395 |
-
"grad_norm": 2.2325947284698486,
|
| 396 |
-
"learning_rate": 8.549892165999505e-05,
|
| 397 |
-
"loss": 0.69,
|
| 398 |
-
"step": 530
|
| 399 |
-
},
|
| 400 |
-
{
|
| 401 |
-
"epoch": 0.8181818181818182,
|
| 402 |
-
"grad_norm": 2.3587887287139893,
|
| 403 |
-
"learning_rate": 8.492108787620105e-05,
|
| 404 |
-
"loss": 0.6987,
|
| 405 |
-
"step": 540
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 0.8333333333333334,
|
| 409 |
-
"grad_norm": 2.0664846897125244,
|
| 410 |
-
"learning_rate": 8.433400150273906e-05,
|
| 411 |
-
"loss": 0.708,
|
| 412 |
-
"step": 550
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 0.8484848484848485,
|
| 416 |
-
"grad_norm": 4.4357147216796875,
|
| 417 |
-
"learning_rate": 8.373781809230355e-05,
|
| 418 |
-
"loss": 0.6981,
|
| 419 |
-
"step": 560
|
| 420 |
-
},
|
| 421 |
-
{
|
| 422 |
-
"epoch": 0.8636363636363636,
|
| 423 |
-
"grad_norm": 2.435680627822876,
|
| 424 |
-
"learning_rate": 8.313269560791342e-05,
|
| 425 |
-
"loss": 0.7105,
|
| 426 |
-
"step": 570
|
| 427 |
-
},
|
| 428 |
-
{
|
| 429 |
-
"epoch": 0.8787878787878788,
|
| 430 |
-
"grad_norm": 2.5096144676208496,
|
| 431 |
-
"learning_rate": 8.251879438105854e-05,
|
| 432 |
-
"loss": 0.6897,
|
| 433 |
-
"step": 580
|
| 434 |
-
},
|
| 435 |
-
{
|
| 436 |
-
"epoch": 0.8939393939393939,
|
| 437 |
-
"grad_norm": 1.9323264360427856,
|
| 438 |
-
"learning_rate": 8.189627706921877e-05,
|
| 439 |
-
"loss": 0.6746,
|
| 440 |
-
"step": 590
|
| 441 |
-
},
|
| 442 |
-
{
|
| 443 |
-
"epoch": 0.9090909090909091,
|
| 444 |
-
"grad_norm": 2.3687186241149902,
|
| 445 |
-
"learning_rate": 8.126530861276677e-05,
|
| 446 |
-
"loss": 0.6847,
|
| 447 |
-
"step": 600
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 0.9090909090909091,
|
| 451 |
-
"eval_accuracy": 0.5935251798561151,
|
| 452 |
-
"eval_loss": 0.6970531344413757,
|
| 453 |
-
"eval_runtime": 19.5244,
|
| 454 |
-
"eval_samples_per_second": 14.239,
|
| 455 |
-
"eval_steps_per_second": 1.793,
|
| 456 |
-
"step": 600
|
| 457 |
-
},
|
| 458 |
-
{
|
| 459 |
-
"epoch": 0.9242424242424242,
|
| 460 |
-
"grad_norm": 2.470337152481079,
|
| 461 |
-
"learning_rate": 8.062605619126584e-05,
|
| 462 |
-
"loss": 0.6787,
|
| 463 |
-
"step": 610
|
| 464 |
-
},
|
| 465 |
-
{
|
| 466 |
-
"epoch": 0.9393939393939394,
|
| 467 |
-
"grad_norm": 2.8660871982574463,
|
| 468 |
-
"learning_rate": 7.997868917917453e-05,
|
| 469 |
-
"loss": 0.6958,
|
| 470 |
-
"step": 620
|
| 471 |
-
},
|
| 472 |
-
{
|
| 473 |
-
"epoch": 0.9545454545454546,
|
| 474 |
-
"grad_norm": 3.2916390895843506,
|
| 475 |
-
"learning_rate": 7.932337910096961e-05,
|
| 476 |
-
"loss": 0.7024,
|
| 477 |
-
"step": 630
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"epoch": 0.9696969696969697,
|
| 481 |
-
"grad_norm": 2.203716278076172,
|
| 482 |
-
"learning_rate": 7.866029958569956e-05,
|
| 483 |
-
"loss": 0.6858,
|
| 484 |
-
"step": 640
|
| 485 |
-
},
|
| 486 |
-
{
|
| 487 |
-
"epoch": 0.9848484848484849,
|
| 488 |
-
"grad_norm": 2.21602463722229,
|
| 489 |
-
"learning_rate": 7.798962632098024e-05,
|
| 490 |
-
"loss": 0.7044,
|
| 491 |
-
"step": 650
|
| 492 |
-
},
|
| 493 |
-
{
|
| 494 |
-
"epoch": 1.0,
|
| 495 |
-
"grad_norm": 1.9984266757965088,
|
| 496 |
-
"learning_rate": 7.73115370064452e-05,
|
| 497 |
-
"loss": 0.696,
|
| 498 |
-
"step": 660
|
| 499 |
-
},
|
| 500 |
-
{
|
| 501 |
-
"epoch": 1.0151515151515151,
|
| 502 |
-
"grad_norm": 2.6973960399627686,
|
| 503 |
-
"learning_rate": 7.6626211306663e-05,
|
| 504 |
-
"loss": 0.6972,
|
| 505 |
-
"step": 670
|
| 506 |
-
},
|
| 507 |
-
{
|
| 508 |
-
"epoch": 1.0303030303030303,
|
| 509 |
-
"grad_norm": 2.1271116733551025,
|
| 510 |
-
"learning_rate": 7.59338308035337e-05,
|
| 511 |
-
"loss": 0.6862,
|
| 512 |
-
"step": 680
|
| 513 |
-
},
|
| 514 |
-
{
|
| 515 |
-
"epoch": 1.0454545454545454,
|
| 516 |
-
"grad_norm": 2.1646666526794434,
|
| 517 |
-
"learning_rate": 7.523457894817745e-05,
|
| 518 |
-
"loss": 0.6957,
|
| 519 |
-
"step": 690
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 1.0606060606060606,
|
| 523 |
-
"grad_norm": 2.122968912124634,
|
| 524 |
-
"learning_rate": 7.452864101232798e-05,
|
| 525 |
-
"loss": 0.6931,
|
| 526 |
-
"step": 700
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 1.0757575757575757,
|
| 530 |
-
"grad_norm": 2.318418502807617,
|
| 531 |
-
"learning_rate": 7.381620403924333e-05,
|
| 532 |
-
"loss": 0.6833,
|
| 533 |
-
"step": 710
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 1.0909090909090908,
|
| 537 |
-
"grad_norm": 2.31949782371521,
|
| 538 |
-
"learning_rate": 7.30974567941475e-05,
|
| 539 |
-
"loss": 0.6866,
|
| 540 |
-
"step": 720
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"epoch": 1.106060606060606,
|
| 544 |
-
"grad_norm": 2.116624355316162,
|
| 545 |
-
"learning_rate": 7.237258971421587e-05,
|
| 546 |
-
"loss": 0.6976,
|
| 547 |
-
"step": 730
|
| 548 |
-
},
|
| 549 |
-
{
|
| 550 |
-
"epoch": 1.121212121212121,
|
| 551 |
-
"grad_norm": 2.1707992553710938,
|
| 552 |
-
"learning_rate": 7.164179485811727e-05,
|
| 553 |
-
"loss": 0.6932,
|
| 554 |
-
"step": 740
|
| 555 |
-
},
|
| 556 |
-
{
|
| 557 |
-
"epoch": 1.1363636363636362,
|
| 558 |
-
"grad_norm": 2.0329554080963135,
|
| 559 |
-
"learning_rate": 7.090526585512696e-05,
|
| 560 |
-
"loss": 0.6872,
|
| 561 |
-
"step": 750
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 1.1515151515151516,
|
| 565 |
-
"grad_norm": 2.590719223022461,
|
| 566 |
-
"learning_rate": 7.016319785382296e-05,
|
| 567 |
-
"loss": 0.7154,
|
| 568 |
-
"step": 760
|
| 569 |
-
},
|
| 570 |
-
{
|
| 571 |
-
"epoch": 1.1666666666666667,
|
| 572 |
-
"grad_norm": 2.158923387527466,
|
| 573 |
-
"learning_rate": 6.941578747038023e-05,
|
| 574 |
-
"loss": 0.6906,
|
| 575 |
-
"step": 770
|
| 576 |
-
},
|
| 577 |
-
{
|
| 578 |
-
"epoch": 1.1818181818181819,
|
| 579 |
-
"grad_norm": 3.0064737796783447,
|
| 580 |
-
"learning_rate": 6.866323273647563e-05,
|
| 581 |
-
"loss": 0.6955,
|
| 582 |
-
"step": 780
|
| 583 |
-
},
|
| 584 |
-
{
|
| 585 |
-
"epoch": 1.196969696969697,
|
| 586 |
-
"grad_norm": 2.0017855167388916,
|
| 587 |
-
"learning_rate": 6.79057330468182e-05,
|
| 588 |
-
"loss": 0.6854,
|
| 589 |
-
"step": 790
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 1.2121212121212122,
|
| 593 |
-
"grad_norm": 3.6152710914611816,
|
| 594 |
-
"learning_rate": 6.7143489106318e-05,
|
| 595 |
-
"loss": 0.698,
|
| 596 |
-
"step": 800
|
| 597 |
-
},
|
| 598 |
-
{
|
| 599 |
-
"epoch": 1.2121212121212122,
|
| 600 |
-
"eval_accuracy": 0.5935251798561151,
|
| 601 |
-
"eval_loss": 0.6945483684539795,
|
| 602 |
-
"eval_runtime": 19.5034,
|
| 603 |
-
"eval_samples_per_second": 14.254,
|
| 604 |
-
"eval_steps_per_second": 1.795,
|
| 605 |
-
"step": 800
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 1.2272727272727273,
|
| 609 |
-
"grad_norm": 1.9689804315567017,
|
| 610 |
-
"learning_rate": 6.637670287690799e-05,
|
| 611 |
-
"loss": 0.7115,
|
| 612 |
-
"step": 810
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 1.2424242424242424,
|
| 616 |
-
"grad_norm": 4.077113151550293,
|
| 617 |
-
"learning_rate": 6.560557752403277e-05,
|
| 618 |
-
"loss": 0.6883,
|
| 619 |
-
"step": 820
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 1.2575757575757576,
|
| 623 |
-
"grad_norm": 2.2839391231536865,
|
| 624 |
-
"learning_rate": 6.483031736281843e-05,
|
| 625 |
-
"loss": 0.6972,
|
| 626 |
-
"step": 830
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 1.2727272727272727,
|
| 630 |
-
"grad_norm": 1.9423472881317139,
|
| 631 |
-
"learning_rate": 6.40511278039378e-05,
|
| 632 |
-
"loss": 0.701,
|
| 633 |
-
"step": 840
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 1.2878787878787878,
|
| 637 |
-
"grad_norm": 2.111860752105713,
|
| 638 |
-
"learning_rate": 6.326821529918553e-05,
|
| 639 |
-
"loss": 0.7023,
|
| 640 |
-
"step": 850
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 1.303030303030303,
|
| 644 |
-
"grad_norm": 1.9913779497146606,
|
| 645 |
-
"learning_rate": 6.248178728677711e-05,
|
| 646 |
-
"loss": 0.7014,
|
| 647 |
-
"step": 860
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 1.3181818181818181,
|
| 651 |
-
"grad_norm": 1.9043771028518677,
|
| 652 |
-
"learning_rate": 6.16920521363867e-05,
|
| 653 |
-
"loss": 0.6891,
|
| 654 |
-
"step": 870
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 1.3333333333333333,
|
| 658 |
-
"grad_norm": 2.018148899078369,
|
| 659 |
-
"learning_rate": 6.089921909393812e-05,
|
| 660 |
-
"loss": 0.6969,
|
| 661 |
-
"step": 880
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 1.3484848484848486,
|
| 665 |
-
"grad_norm": 1.8589205741882324,
|
| 666 |
-
"learning_rate": 6.0103498226163603e-05,
|
| 667 |
-
"loss": 0.6982,
|
| 668 |
-
"step": 890
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 1.3636363636363638,
|
| 672 |
-
"grad_norm": 1.8796669244766235,
|
| 673 |
-
"learning_rate": 5.93051003649452e-05,
|
| 674 |
-
"loss": 0.6787,
|
| 675 |
-
"step": 900
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 1.378787878787879,
|
| 679 |
-
"grad_norm": 2.0444390773773193,
|
| 680 |
-
"learning_rate": 5.850423705145334e-05,
|
| 681 |
-
"loss": 0.6934,
|
| 682 |
-
"step": 910
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"epoch": 1.393939393939394,
|
| 686 |
-
"grad_norm": 2.5297834873199463,
|
| 687 |
-
"learning_rate": 5.770112048009747e-05,
|
| 688 |
-
"loss": 0.7092,
|
| 689 |
-
"step": 920
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 1.4090909090909092,
|
| 693 |
-
"grad_norm": 1.9327147006988525,
|
| 694 |
-
"learning_rate": 5.68959634423037e-05,
|
| 695 |
-
"loss": 0.6951,
|
| 696 |
-
"step": 930
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 1.4242424242424243,
|
| 700 |
-
"grad_norm": 1.8969117403030396,
|
| 701 |
-
"learning_rate": 5.60889792701342e-05,
|
| 702 |
-
"loss": 0.7045,
|
| 703 |
-
"step": 940
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 1.4393939393939394,
|
| 707 |
-
"grad_norm": 2.584240198135376,
|
| 708 |
-
"learning_rate": 5.52803817797633e-05,
|
| 709 |
-
"loss": 0.6734,
|
| 710 |
-
"step": 950
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 1.4545454545454546,
|
| 714 |
-
"grad_norm": 2.54909086227417,
|
| 715 |
-
"learning_rate": 5.4470385214825416e-05,
|
| 716 |
-
"loss": 0.6727,
|
| 717 |
-
"step": 960
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 1.4696969696969697,
|
| 721 |
-
"grad_norm": 2.040754556655884,
|
| 722 |
-
"learning_rate": 5.365920418964973e-05,
|
| 723 |
-
"loss": 0.688,
|
| 724 |
-
"step": 970
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 1.4848484848484849,
|
| 728 |
-
"grad_norm": 2.1792185306549072,
|
| 729 |
-
"learning_rate": 5.28470536323965e-05,
|
| 730 |
-
"loss": 0.6911,
|
| 731 |
-
"step": 980
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 1.5,
|
| 735 |
-
"grad_norm": 2.115302562713623,
|
| 736 |
-
"learning_rate": 5.2034148728110424e-05,
|
| 737 |
-
"loss": 0.6942,
|
| 738 |
-
"step": 990
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 1.5151515151515151,
|
| 742 |
-
"grad_norm": 6.179975509643555,
|
| 743 |
-
"learning_rate": 5.1220704861705774e-05,
|
| 744 |
-
"loss": 0.6945,
|
| 745 |
-
"step": 1000
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 1.5151515151515151,
|
| 749 |
-
"eval_accuracy": 0.4460431654676259,
|
| 750 |
-
"eval_loss": 0.7005228996276855,
|
| 751 |
-
"eval_runtime": 19.5262,
|
| 752 |
-
"eval_samples_per_second": 14.237,
|
| 753 |
-
"eval_steps_per_second": 1.792,
|
| 754 |
-
"step": 1000
|
| 755 |
-
},
|
| 756 |
-
{
|
| 757 |
-
"epoch": 1.5303030303030303,
|
| 758 |
-
"grad_norm": 2.1656877994537354,
|
| 759 |
-
"learning_rate": 5.0406937560898646e-05,
|
| 760 |
-
"loss": 0.694,
|
| 761 |
-
"step": 1010
|
| 762 |
-
},
|
| 763 |
-
{
|
| 764 |
-
"epoch": 1.5454545454545454,
|
| 765 |
-
"grad_norm": 2.0086238384246826,
|
| 766 |
-
"learning_rate": 4.9593062439101365e-05,
|
| 767 |
-
"loss": 0.683,
|
| 768 |
-
"step": 1020
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 1.5606060606060606,
|
| 772 |
-
"grad_norm": 2.685886859893799,
|
| 773 |
-
"learning_rate": 4.877929513829424e-05,
|
| 774 |
-
"loss": 0.6927,
|
| 775 |
-
"step": 1030
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 1.5757575757575757,
|
| 779 |
-
"grad_norm": 2.2931978702545166,
|
| 780 |
-
"learning_rate": 4.796585127188958e-05,
|
| 781 |
-
"loss": 0.6943,
|
| 782 |
-
"step": 1040
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 1.5909090909090908,
|
| 786 |
-
"grad_norm": 1.9456837177276611,
|
| 787 |
-
"learning_rate": 4.715294636760352e-05,
|
| 788 |
-
"loss": 0.7238,
|
| 789 |
-
"step": 1050
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 1.606060606060606,
|
| 793 |
-
"grad_norm": 1.9456251859664917,
|
| 794 |
-
"learning_rate": 4.634079581035029e-05,
|
| 795 |
-
"loss": 0.6867,
|
| 796 |
-
"step": 1060
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 1.621212121212121,
|
| 800 |
-
"grad_norm": 1.8919247388839722,
|
| 801 |
-
"learning_rate": 4.55296147851746e-05,
|
| 802 |
-
"loss": 0.6679,
|
| 803 |
-
"step": 1070
|
| 804 |
-
},
|
| 805 |
-
{
|
| 806 |
-
"epoch": 1.6363636363636362,
|
| 807 |
-
"grad_norm": 2.3800299167633057,
|
| 808 |
-
"learning_rate": 4.471961822023671e-05,
|
| 809 |
-
"loss": 0.6932,
|
| 810 |
-
"step": 1080
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 1.6515151515151514,
|
| 814 |
-
"grad_norm": 2.1245882511138916,
|
| 815 |
-
"learning_rate": 4.391102072986581e-05,
|
| 816 |
-
"loss": 0.6823,
|
| 817 |
-
"step": 1090
|
| 818 |
-
},
|
| 819 |
-
{
|
| 820 |
-
"epoch": 1.6666666666666665,
|
| 821 |
-
"grad_norm": 2.484229326248169,
|
| 822 |
-
"learning_rate": 4.3104036557696295e-05,
|
| 823 |
-
"loss": 0.7094,
|
| 824 |
-
"step": 1100
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 1.6818181818181817,
|
| 828 |
-
"grad_norm": 2.081204652786255,
|
| 829 |
-
"learning_rate": 4.229887951990255e-05,
|
| 830 |
-
"loss": 0.6873,
|
| 831 |
-
"step": 1110
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 1.696969696969697,
|
| 835 |
-
"grad_norm": 2.008754253387451,
|
| 836 |
-
"learning_rate": 4.149576294854668e-05,
|
| 837 |
-
"loss": 0.6957,
|
| 838 |
-
"step": 1120
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 1.7121212121212122,
|
| 842 |
-
"grad_norm": 2.088444948196411,
|
| 843 |
-
"learning_rate": 4.069489963505482e-05,
|
| 844 |
-
"loss": 0.7134,
|
| 845 |
-
"step": 1130
|
| 846 |
-
},
|
| 847 |
-
{
|
| 848 |
-
"epoch": 1.7272727272727273,
|
| 849 |
-
"grad_norm": 2.0371851921081543,
|
| 850 |
-
"learning_rate": 3.98965017738364e-05,
|
| 851 |
-
"loss": 0.698,
|
| 852 |
-
"step": 1140
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 1.7424242424242424,
|
| 856 |
-
"grad_norm": 2.4857864379882812,
|
| 857 |
-
"learning_rate": 3.9100780906061896e-05,
|
| 858 |
-
"loss": 0.702,
|
| 859 |
-
"step": 1150
|
| 860 |
-
},
|
| 861 |
-
{
|
| 862 |
-
"epoch": 1.7575757575757576,
|
| 863 |
-
"grad_norm": 1.9480881690979004,
|
| 864 |
-
"learning_rate": 3.83079478636133e-05,
|
| 865 |
-
"loss": 0.6875,
|
| 866 |
-
"step": 1160
|
| 867 |
-
},
|
| 868 |
-
{
|
| 869 |
-
"epoch": 1.7727272727272727,
|
| 870 |
-
"grad_norm": 2.094036817550659,
|
| 871 |
-
"learning_rate": 3.7518212713222906e-05,
|
| 872 |
-
"loss": 0.7047,
|
| 873 |
-
"step": 1170
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 1.7878787878787878,
|
| 877 |
-
"grad_norm": 2.085651159286499,
|
| 878 |
-
"learning_rate": 3.673178470081448e-05,
|
| 879 |
-
"loss": 0.6933,
|
| 880 |
-
"step": 1180
|
| 881 |
-
},
|
| 882 |
-
{
|
| 883 |
-
"epoch": 1.803030303030303,
|
| 884 |
-
"grad_norm": 2.0030457973480225,
|
| 885 |
-
"learning_rate": 3.594887219606221e-05,
|
| 886 |
-
"loss": 0.6901,
|
| 887 |
-
"step": 1190
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 1.8181818181818183,
|
| 891 |
-
"grad_norm": 2.7572779655456543,
|
| 892 |
-
"learning_rate": 3.516968263718159e-05,
|
| 893 |
-
"loss": 0.6808,
|
| 894 |
-
"step": 1200
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 1.8181818181818183,
|
| 898 |
-
"eval_accuracy": 0.5755395683453237,
|
| 899 |
-
"eval_loss": 0.6959903836250305,
|
| 900 |
-
"eval_runtime": 19.4971,
|
| 901 |
-
"eval_samples_per_second": 14.259,
|
| 902 |
-
"eval_steps_per_second": 1.795,
|
| 903 |
-
"step": 1200
|
| 904 |
-
},
|
| 905 |
-
{
|
| 906 |
-
"epoch": 1.8333333333333335,
|
| 907 |
-
"grad_norm": 2.6810097694396973,
|
| 908 |
-
"learning_rate": 3.439442247596724e-05,
|
| 909 |
-
"loss": 0.7154,
|
| 910 |
-
"step": 1210
|
| 911 |
-
},
|
| 912 |
-
{
|
| 913 |
-
"epoch": 1.8484848484848486,
|
| 914 |
-
"grad_norm": 2.145552635192871,
|
| 915 |
-
"learning_rate": 3.3623297123092006e-05,
|
| 916 |
-
"loss": 0.7025,
|
| 917 |
-
"step": 1220
|
| 918 |
-
},
|
| 919 |
-
{
|
| 920 |
-
"epoch": 1.8636363636363638,
|
| 921 |
-
"grad_norm": 2.2261226177215576,
|
| 922 |
-
"learning_rate": 3.285651089368202e-05,
|
| 923 |
-
"loss": 0.6851,
|
| 924 |
-
"step": 1230
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 1.878787878787879,
|
| 928 |
-
"grad_norm": 2.2848427295684814,
|
| 929 |
-
"learning_rate": 3.209426695318182e-05,
|
| 930 |
-
"loss": 0.686,
|
| 931 |
-
"step": 1240
|
| 932 |
-
},
|
| 933 |
-
{
|
| 934 |
-
"epoch": 1.893939393939394,
|
| 935 |
-
"grad_norm": 2.1015465259552,
|
| 936 |
-
"learning_rate": 3.133676726352438e-05,
|
| 937 |
-
"loss": 0.6909,
|
| 938 |
-
"step": 1250
|
| 939 |
-
},
|
| 940 |
-
{
|
| 941 |
-
"epoch": 1.9090909090909092,
|
| 942 |
-
"grad_norm": 2.174621105194092,
|
| 943 |
-
"learning_rate": 3.0584212529619775e-05,
|
| 944 |
-
"loss": 0.6855,
|
| 945 |
-
"step": 1260
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 1.9242424242424243,
|
| 949 |
-
"grad_norm": 2.224524974822998,
|
| 950 |
-
"learning_rate": 2.9836802146177034e-05,
|
| 951 |
-
"loss": 0.7018,
|
| 952 |
-
"step": 1270
|
| 953 |
-
},
|
| 954 |
-
{
|
| 955 |
-
"epoch": 1.9393939393939394,
|
| 956 |
-
"grad_norm": 5.61815071105957,
|
| 957 |
-
"learning_rate": 2.9094734144873036e-05,
|
| 958 |
-
"loss": 0.6926,
|
| 959 |
-
"step": 1280
|
| 960 |
-
},
|
| 961 |
-
{
|
| 962 |
-
"epoch": 1.9545454545454546,
|
| 963 |
-
"grad_norm": 1.9148584604263306,
|
| 964 |
-
"learning_rate": 2.835820514188273e-05,
|
| 965 |
-
"loss": 0.6977,
|
| 966 |
-
"step": 1290
|
| 967 |
-
},
|
| 968 |
-
{
|
| 969 |
-
"epoch": 1.9696969696969697,
|
| 970 |
-
"grad_norm": 1.9495019912719727,
|
| 971 |
-
"learning_rate": 2.7627410285784163e-05,
|
| 972 |
-
"loss": 0.6976,
|
| 973 |
-
"step": 1300
|
| 974 |
-
},
|
| 975 |
-
{
|
| 976 |
-
"epoch": 1.9848484848484849,
|
| 977 |
-
"grad_norm": 2.185170888900757,
|
| 978 |
-
"learning_rate": 2.6902543205852492e-05,
|
| 979 |
-
"loss": 0.681,
|
| 980 |
-
"step": 1310
|
| 981 |
-
},
|
| 982 |
-
{
|
| 983 |
-
"epoch": 2.0,
|
| 984 |
-
"grad_norm": 2.0127036571502686,
|
| 985 |
-
"learning_rate": 2.618379596075668e-05,
|
| 986 |
-
"loss": 0.7196,
|
| 987 |
-
"step": 1320
|
| 988 |
-
},
|
| 989 |
-
{
|
| 990 |
-
"epoch": 2.015151515151515,
|
| 991 |
-
"grad_norm": 1.9384896755218506,
|
| 992 |
-
"learning_rate": 2.5471358987672017e-05,
|
| 993 |
-
"loss": 0.6875,
|
| 994 |
-
"step": 1330
|
| 995 |
-
},
|
| 996 |
-
{
|
| 997 |
-
"epoch": 2.0303030303030303,
|
| 998 |
-
"grad_norm": 1.8721387386322021,
|
| 999 |
-
"learning_rate": 2.476542105182254e-05,
|
| 1000 |
-
"loss": 0.7013,
|
| 1001 |
-
"step": 1340
|
| 1002 |
-
},
|
| 1003 |
-
{
|
| 1004 |
-
"epoch": 2.0454545454545454,
|
| 1005 |
-
"grad_norm": 1.915560245513916,
|
| 1006 |
-
"learning_rate": 2.4066169196466326e-05,
|
| 1007 |
-
"loss": 0.6822,
|
| 1008 |
-
"step": 1350
|
| 1009 |
-
},
|
| 1010 |
-
{
|
| 1011 |
-
"epoch": 2.0606060606060606,
|
| 1012 |
-
"grad_norm": 2.1030776500701904,
|
| 1013 |
-
"learning_rate": 2.3373788693337024e-05,
|
| 1014 |
-
"loss": 0.691,
|
| 1015 |
-
"step": 1360
|
| 1016 |
-
},
|
| 1017 |
-
{
|
| 1018 |
-
"epoch": 2.0757575757575757,
|
| 1019 |
-
"grad_norm": 2.210036277770996,
|
| 1020 |
-
"learning_rate": 2.268846299355481e-05,
|
| 1021 |
-
"loss": 0.6745,
|
| 1022 |
-
"step": 1370
|
| 1023 |
-
},
|
| 1024 |
-
{
|
| 1025 |
-
"epoch": 2.090909090909091,
|
| 1026 |
-
"grad_norm": 2.111227035522461,
|
| 1027 |
-
"learning_rate": 2.2010373679019776e-05,
|
| 1028 |
-
"loss": 0.6932,
|
| 1029 |
-
"step": 1380
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 2.106060606060606,
|
| 1033 |
-
"grad_norm": 2.6546616554260254,
|
| 1034 |
-
"learning_rate": 2.133970041430044e-05,
|
| 1035 |
-
"loss": 0.6984,
|
| 1036 |
-
"step": 1390
|
| 1037 |
-
},
|
| 1038 |
-
{
|
| 1039 |
-
"epoch": 2.121212121212121,
|
| 1040 |
-
"grad_norm": 2.6496689319610596,
|
| 1041 |
-
"learning_rate": 2.067662089903039e-05,
|
| 1042 |
-
"loss": 0.6879,
|
| 1043 |
-
"step": 1400
|
| 1044 |
-
},
|
| 1045 |
-
{
|
| 1046 |
-
"epoch": 2.121212121212121,
|
| 1047 |
-
"eval_accuracy": 0.7482014388489209,
|
| 1048 |
-
"eval_loss": 0.6909542083740234,
|
| 1049 |
-
"eval_runtime": 19.5229,
|
| 1050 |
-
"eval_samples_per_second": 14.24,
|
| 1051 |
-
"eval_steps_per_second": 1.793,
|
| 1052 |
-
"step": 1400
|
| 1053 |
-
}
|
| 1054 |
-
],
|
| 1055 |
-
"logging_steps": 10,
|
| 1056 |
-
"max_steps": 1980,
|
| 1057 |
-
"num_input_tokens_seen": 0,
|
| 1058 |
-
"num_train_epochs": 3,
|
| 1059 |
-
"save_steps": 200,
|
| 1060 |
-
"stateful_callbacks": {
|
| 1061 |
-
"TrainerControl": {
|
| 1062 |
-
"args": {
|
| 1063 |
-
"should_epoch_stop": false,
|
| 1064 |
-
"should_evaluate": false,
|
| 1065 |
-
"should_log": false,
|
| 1066 |
-
"should_save": true,
|
| 1067 |
-
"should_training_stop": false
|
| 1068 |
-
},
|
| 1069 |
-
"attributes": {}
|
| 1070 |
-
}
|
| 1071 |
-
},
|
| 1072 |
-
"total_flos": 0.0,
|
| 1073 |
-
"train_batch_size": 2,
|
| 1074 |
-
"trial_name": null,
|
| 1075 |
-
"trial_params": null
|
| 1076 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|