Upload ./trainer_state.json with huggingface_hub
Browse files- trainer_state.json +166 -2805
trainer_state.json
CHANGED
|
@@ -1,2881 +1,242 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "./output/training_results/
|
| 4 |
"epoch": 4.0,
|
| 5 |
-
"eval_steps":
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 2.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 1.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 2.
|
| 23 |
-
"step":
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss":
|
| 30 |
-
"step":
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
"
|
| 37 |
-
"
|
|
|
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm": 1.
|
| 42 |
-
"learning_rate": 1.
|
| 43 |
-
"loss": 2.058,
|
| 44 |
-
"step": 20
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.05040322580645161,
|
| 48 |
-
"grad_norm": 2.12580198675208,
|
| 49 |
-
"learning_rate": 1.47507132047542e-05,
|
| 50 |
-
"loss": 2.0324,
|
| 51 |
-
"step": 25
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.06048387096774194,
|
| 55 |
-
"grad_norm": 1.4836280728067701,
|
| 56 |
-
"learning_rate": 1.4424209058252117e-05,
|
| 57 |
-
"loss": 1.9982,
|
| 58 |
-
"step": 30
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.07056451612903226,
|
| 62 |
-
"grad_norm": 1.6623583118858967,
|
| 63 |
-
"learning_rate": 1.4025354123166567e-05,
|
| 64 |
-
"loss": 1.8959,
|
| 65 |
-
"step": 35
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.08064516129032258,
|
| 69 |
-
"grad_norm": 1.5916601844779366,
|
| 70 |
-
"learning_rate": 1.3636585978219431e-05,
|
| 71 |
-
"loss": 2.0492,
|
| 72 |
-
"step": 40
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.0907258064516129,
|
| 76 |
-
"grad_norm": 1.7125711612090788,
|
| 77 |
-
"learning_rate": 1.3257674460153073e-05,
|
| 78 |
-
"loss": 1.9655,
|
| 79 |
-
"step": 45
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.10080645161290322,
|
| 83 |
-
"grad_norm": 1.632995352775119,
|
| 84 |
-
"learning_rate": 1.288839408598928e-05,
|
| 85 |
-
"loss": 2.0569,
|
| 86 |
-
"step": 50
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.11088709677419355,
|
| 90 |
-
"grad_norm": 1.9174014047156116,
|
| 91 |
-
"learning_rate": 1.2528523969541332e-05,
|
| 92 |
-
"loss": 1.9287,
|
| 93 |
-
"step": 55
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.12096774193548387,
|
| 97 |
-
"grad_norm": 1.867558573747783,
|
| 98 |
-
"learning_rate": 1.2177847739205852e-05,
|
| 99 |
-
"loss": 1.9894,
|
| 100 |
-
"step": 60
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.1310483870967742,
|
| 104 |
-
"grad_norm": 1.567626378627341,
|
| 105 |
-
"learning_rate": 1.1836153457018077e-05,
|
| 106 |
-
"loss": 2.0005,
|
| 107 |
-
"step": 65
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.14112903225806453,
|
| 111 |
-
"grad_norm": 1.3955216806991564,
|
| 112 |
-
"learning_rate": 1.1569125393187758e-05,
|
| 113 |
-
"loss": 2.0943,
|
| 114 |
-
"step": 70
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.15120967741935484,
|
| 118 |
-
"grad_norm": 1.4220083911016554,
|
| 119 |
-
"learning_rate": 1.1243078381650268e-05,
|
| 120 |
-
"loss": 1.9521,
|
| 121 |
-
"step": 75
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.16129032258064516,
|
| 125 |
-
"grad_norm": 1.5790411664333717,
|
| 126 |
-
"learning_rate": 1.0925442805236739e-05,
|
| 127 |
-
"loss": 2.0206,
|
| 128 |
-
"step": 80
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.17137096774193547,
|
| 132 |
-
"grad_norm": 1.3900088055150563,
|
| 133 |
-
"learning_rate": 1.0616022857385877e-05,
|
| 134 |
-
"loss": 1.9713,
|
| 135 |
-
"step": 85
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.1814516129032258,
|
| 139 |
-
"grad_norm": 1.3384719686058493,
|
| 140 |
-
"learning_rate": 1.0314626793719932e-05,
|
| 141 |
-
"loss": 1.925,
|
| 142 |
-
"step": 90
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.19153225806451613,
|
| 146 |
-
"grad_norm": 1.6023250792595878,
|
| 147 |
-
"learning_rate": 1.0021066858113034e-05,
|
| 148 |
-
"loss": 2.0494,
|
| 149 |
-
"step": 95
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.20161290322580644,
|
| 153 |
-
"grad_norm": 1.4194138427481093,
|
| 154 |
-
"learning_rate": 9.735159209915871e-06,
|
| 155 |
-
"loss": 2.0098,
|
| 156 |
-
"step": 100
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.21169354838709678,
|
| 160 |
-
"grad_norm": 1.2578428133053623,
|
| 161 |
-
"learning_rate": 9.456723852321558e-06,
|
| 162 |
-
"loss": 1.9048,
|
| 163 |
-
"step": 105
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.2217741935483871,
|
| 167 |
-
"grad_norm": 1.2538874670770739,
|
| 168 |
-
"learning_rate": 9.239237068300538e-06,
|
| 169 |
-
"loss": 1.9968,
|
| 170 |
-
"step": 110
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.2318548387096774,
|
| 174 |
-
"grad_norm": 1.39475982271136,
|
| 175 |
-
"learning_rate": 8.973810241991637e-06,
|
| 176 |
-
"loss": 2.0145,
|
| 177 |
-
"step": 115
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.24193548387096775,
|
| 181 |
-
"grad_norm": 1.2474176211653318,
|
| 182 |
-
"learning_rate": 8.715371568554332e-06,
|
| 183 |
-
"loss": 2.0293,
|
| 184 |
-
"step": 120
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.25201612903225806,
|
| 188 |
-
"grad_norm": 1.4147919193342449,
|
| 189 |
-
"learning_rate": 8.463755024156938e-06,
|
| 190 |
-
"loss": 1.9515,
|
| 191 |
-
"step": 125
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.2620967741935484,
|
| 195 |
-
"grad_norm": 1.2583008452280033,
|
| 196 |
-
"learning_rate": 8.218798100385521e-06,
|
| 197 |
-
"loss": 1.9602,
|
| 198 |
-
"step": 130
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.2721774193548387,
|
| 202 |
-
"grad_norm": 1.4165902883907349,
|
| 203 |
-
"learning_rate": 7.980341738939008e-06,
|
| 204 |
-
"loss": 1.985,
|
| 205 |
-
"step": 135
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.28225806451612906,
|
| 209 |
-
"grad_norm": 1.2768405135809222,
|
| 210 |
-
"learning_rate": 7.748230267366966e-06,
|
| 211 |
-
"loss": 1.9847,
|
| 212 |
-
"step": 140
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.2923387096774194,
|
| 216 |
-
"grad_norm": 1.4092721850037662,
|
| 217 |
-
"learning_rate": 7.522311335836008e-06,
|
| 218 |
-
"loss": 1.876,
|
| 219 |
-
"step": 145
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.3024193548387097,
|
| 223 |
-
"grad_norm": 1.307344104572624,
|
| 224 |
-
"learning_rate": 7.302435854911248e-06,
|
| 225 |
-
"loss": 1.9837,
|
| 226 |
-
"step": 150
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.3125,
|
| 230 |
-
"grad_norm": 1.1706462667165936,
|
| 231 |
-
"learning_rate": 7.088457934339014e-06,
|
| 232 |
-
"loss": 1.9666,
|
| 233 |
-
"step": 155
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.3225806451612903,
|
| 237 |
-
"grad_norm": 1.4451121840176004,
|
| 238 |
-
"learning_rate": 6.880234822817464e-06,
|
| 239 |
-
"loss": 1.9447,
|
| 240 |
-
"step": 160
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.3326612903225806,
|
| 244 |
-
"grad_norm": 1.5334691388320456,
|
| 245 |
-
"learning_rate": 6.677626848741738e-06,
|
| 246 |
-
"loss": 1.9565,
|
| 247 |
-
"step": 165
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.34274193548387094,
|
| 251 |
-
"grad_norm": 1.3638980480173026,
|
| 252 |
-
"learning_rate": 6.480497361910502e-06,
|
| 253 |
-
"loss": 1.8949,
|
| 254 |
-
"step": 170
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.3528225806451613,
|
| 258 |
-
"grad_norm": 1.1960476927370378,
|
| 259 |
-
"learning_rate": 6.288712676180814e-06,
|
| 260 |
-
"loss": 1.8283,
|
| 261 |
-
"step": 175
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.3629032258064516,
|
| 265 |
-
"grad_norm": 1.2534954529871944,
|
| 266 |
-
"learning_rate": 6.102142013058452e-06,
|
| 267 |
-
"loss": 1.9803,
|
| 268 |
-
"step": 180
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.37298387096774194,
|
| 272 |
-
"grad_norm": 1.371419907969094,
|
| 273 |
-
"learning_rate": 5.92065744621091e-06,
|
| 274 |
-
"loss": 1.966,
|
| 275 |
-
"step": 185
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.38306451612903225,
|
| 279 |
-
"grad_norm": 1.2633124617977696,
|
| 280 |
-
"learning_rate": 5.744133846890542e-06,
|
| 281 |
-
"loss": 1.9811,
|
| 282 |
-
"step": 190
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 0.39314516129032256,
|
| 286 |
-
"grad_norm": 1.164537124006197,
|
| 287 |
-
"learning_rate": 5.572448830255249e-06,
|
| 288 |
-
"loss": 1.9782,
|
| 289 |
-
"step": 195
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.4032258064516129,
|
| 293 |
-
"grad_norm": 1.2462948785459638,
|
| 294 |
-
"learning_rate": 5.405482702574484e-06,
|
| 295 |
-
"loss": 1.9087,
|
| 296 |
-
"step": 200
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.4032258064516129,
|
| 300 |
-
"eval_loss": 1.971670150756836,
|
| 301 |
-
"eval_runtime": 90.0126,
|
| 302 |
-
"eval_samples_per_second": 19.564,
|
| 303 |
-
"eval_steps_per_second": 0.311,
|
| 304 |
-
"step": 200
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.41330645161290325,
|
| 308 |
-
"grad_norm": 1.3766492549691904,
|
| 309 |
-
"learning_rate": 5.243118409308317e-06,
|
| 310 |
-
"loss": 2.0798,
|
| 311 |
-
"step": 205
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.42338709677419356,
|
| 315 |
-
"grad_norm": 1.2423997076500177,
|
| 316 |
-
"learning_rate": 5.085241484047513e-06,
|
| 317 |
-
"loss": 1.8739,
|
| 318 |
-
"step": 210
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.4334677419354839,
|
| 322 |
-
"grad_norm": 1.1300210393690464,
|
| 323 |
-
"learning_rate": 4.931739998302693e-06,
|
| 324 |
-
"loss": 1.8057,
|
| 325 |
-
"step": 215
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.4435483870967742,
|
| 329 |
-
"grad_norm": 1.4140153817527936,
|
| 330 |
-
"learning_rate": 4.782504512130776e-06,
|
| 331 |
-
"loss": 1.9858,
|
| 332 |
-
"step": 220
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.4536290322580645,
|
| 336 |
-
"grad_norm": 1.2444950990794887,
|
| 337 |
-
"learning_rate": 4.637428025587017e-06,
|
| 338 |
"loss": 1.9789,
|
| 339 |
-
"step":
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.4637096774193548,
|
| 343 |
-
"grad_norm": 1.2052674239904773,
|
| 344 |
-
"learning_rate": 4.4964059309911556e-06,
|
| 345 |
-
"loss": 1.9367,
|
| 346 |
-
"step": 230
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.4737903225806452,
|
| 350 |
-
"grad_norm": 1.277871648485073,
|
| 351 |
-
"learning_rate": 4.3593359659961716e-06,
|
| 352 |
-
"loss": 1.9616,
|
| 353 |
-
"step": 235
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.4838709677419355,
|
| 357 |
-
"grad_norm": 1.3535390602470023,
|
| 358 |
-
"learning_rate": 4.226118167448512e-06,
|
| 359 |
-
"loss": 1.9542,
|
| 360 |
-
"step": 240
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.4939516129032258,
|
| 364 |
-
"grad_norm": 1.222656117941944,
|
| 365 |
-
"learning_rate": 4.096654826028485e-06,
|
| 366 |
-
"loss": 1.9179,
|
| 367 |
-
"step": 245
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.5040322580645161,
|
| 371 |
-
"grad_norm": 1.2370551856793075,
|
| 372 |
-
"learning_rate": 3.97085044165989e-06,
|
| 373 |
-
"loss": 1.8558,
|
| 374 |
-
"step": 250
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.5141129032258065,
|
| 378 |
-
"grad_norm": 1.2889675743730276,
|
| 379 |
-
"learning_rate": 3.848611679677959e-06,
|
| 380 |
-
"loss": 1.9233,
|
| 381 |
-
"step": 255
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 0.
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 1.
|
| 388 |
-
"step":
|
| 389 |
},
|
| 390 |
{
|
| 391 |
-
"epoch": 0.
|
| 392 |
-
"
|
| 393 |
-
"
|
| 394 |
-
"
|
| 395 |
-
"
|
|
|
|
| 396 |
},
|
| 397 |
{
|
| 398 |
-
"epoch":
|
| 399 |
-
"grad_norm": 1.
|
| 400 |
-
"learning_rate": 3.
|
| 401 |
-
"loss": 1.
|
| 402 |
-
"step":
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch":
|
| 406 |
-
"grad_norm": 1.
|
| 407 |
-
"learning_rate":
|
| 408 |
-
"loss": 1.
|
| 409 |
-
"step":
|
| 410 |
},
|
| 411 |
{
|
| 412 |
-
"epoch":
|
| 413 |
-
"
|
| 414 |
-
"
|
| 415 |
-
"
|
| 416 |
-
"
|
|
|
|
| 417 |
},
|
| 418 |
{
|
| 419 |
-
"epoch":
|
| 420 |
-
"grad_norm": 1.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss": 1.
|
| 423 |
-
"step":
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"grad_norm": 1.
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss": 1.
|
| 430 |
-
"step":
|
| 431 |
},
|
| 432 |
{
|
| 433 |
-
"epoch":
|
| 434 |
-
"
|
| 435 |
-
"
|
| 436 |
-
"
|
| 437 |
-
"
|
|
|
|
| 438 |
},
|
| 439 |
{
|
| 440 |
-
"epoch":
|
| 441 |
-
"grad_norm": 1.
|
| 442 |
-
"learning_rate":
|
| 443 |
-
"loss": 1.
|
| 444 |
-
"step":
|
| 445 |
},
|
| 446 |
{
|
| 447 |
-
"epoch":
|
| 448 |
-
"grad_norm": 1.
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"loss": 1.
|
| 451 |
-
"step":
|
| 452 |
},
|
| 453 |
{
|
| 454 |
-
"epoch":
|
| 455 |
-
"
|
| 456 |
-
"
|
| 457 |
-
"
|
| 458 |
-
"
|
|
|
|
| 459 |
},
|
| 460 |
{
|
| 461 |
-
"epoch":
|
| 462 |
-
"grad_norm": 1.
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 1.
|
| 465 |
-
"step":
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"grad_norm": 1.
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss":
|
| 472 |
-
"step":
|
| 473 |
},
|
| 474 |
{
|
| 475 |
-
"epoch":
|
| 476 |
-
"
|
| 477 |
-
"
|
| 478 |
-
"
|
| 479 |
-
"
|
|
|
|
| 480 |
},
|
| 481 |
{
|
| 482 |
-
"epoch":
|
| 483 |
-
"grad_norm": 1.
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss": 1.
|
| 486 |
-
"step":
|
| 487 |
},
|
| 488 |
{
|
| 489 |
-
"epoch":
|
| 490 |
-
"grad_norm": 1.
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss": 1.
|
| 493 |
-
"step":
|
| 494 |
},
|
| 495 |
{
|
| 496 |
-
"epoch":
|
| 497 |
-
"
|
| 498 |
-
"
|
| 499 |
-
"
|
| 500 |
-
"
|
|
|
|
| 501 |
},
|
| 502 |
{
|
| 503 |
-
"epoch":
|
| 504 |
-
"grad_norm": 1.
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss": 1.8891,
|
| 507 |
-
"step": 345
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.7056451612903226,
|
| 511 |
-
"grad_norm": 1.1893154239696047,
|
| 512 |
-
"learning_rate": 2.093870446430792e-06,
|
| 513 |
-
"loss": 1.9165,
|
| 514 |
-
"step": 350
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.7157258064516129,
|
| 518 |
-
"grad_norm": 1.3456137263979548,
|
| 519 |
-
"learning_rate": 2.026320833737434e-06,
|
| 520 |
-
"loss": 1.9668,
|
| 521 |
-
"step": 355
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.7258064516129032,
|
| 525 |
-
"grad_norm": 1.237670919084104,
|
| 526 |
-
"learning_rate": 1.9608069053895406e-06,
|
| 527 |
"loss": 1.8764,
|
| 528 |
-
"step":
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.7358870967741935,
|
| 532 |
-
"grad_norm": 1.2467674896007754,
|
| 533 |
-
"learning_rate": 1.8972732810927808e-06,
|
| 534 |
-
"loss": 1.8413,
|
| 535 |
-
"step": 365
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.7459677419354839,
|
| 539 |
-
"grad_norm": 1.2641168575246482,
|
| 540 |
-
"learning_rate": 1.8356659238320686e-06,
|
| 541 |
-
"loss": 1.8496,
|
| 542 |
-
"step": 370
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.7560483870967742,
|
| 546 |
-
"grad_norm": 1.3760663460506264,
|
| 547 |
-
"learning_rate": 1.7759321112753693e-06,
|
| 548 |
-
"loss": 1.8394,
|
| 549 |
-
"step": 375
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.7661290322580645,
|
| 553 |
-
"grad_norm": 1.1549697025197416,
|
| 554 |
-
"learning_rate": 1.718020407700906e-06,
|
| 555 |
-
"loss": 1.8772,
|
| 556 |
-
"step": 380
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.7762096774193549,
|
| 560 |
-
"grad_norm": 1.173935455479261,
|
| 561 |
-
"learning_rate": 1.6618806364397543e-06,
|
| 562 |
-
"loss": 1.873,
|
| 563 |
-
"step": 385
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.7862903225806451,
|
| 567 |
-
"grad_norm": 1.1687914205281313,
|
| 568 |
-
"learning_rate": 1.6074638528259137e-06,
|
| 569 |
-
"loss": 1.7951,
|
| 570 |
-
"step": 390
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.7963709677419355,
|
| 574 |
-
"grad_norm": 1.2122222438808594,
|
| 575 |
-
"learning_rate": 1.5547223176460478e-06,
|
| 576 |
-
"loss": 1.9303,
|
| 577 |
-
"step": 395
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.8064516129032258,
|
| 581 |
-
"grad_norm": 1.2659166954554237,
|
| 582 |
-
"learning_rate": 1.503609471081174e-06,
|
| 583 |
-
"loss": 1.8752,
|
| 584 |
-
"step": 400
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.8064516129032258,
|
| 588 |
-
"eval_loss": 1.9417628049850464,
|
| 589 |
-
"eval_runtime": 89.9499,
|
| 590 |
-
"eval_samples_per_second": 19.578,
|
| 591 |
-
"eval_steps_per_second": 0.311,
|
| 592 |
-
"step": 400
|
| 593 |
-
},
|
| 594 |
-
{
|
| 595 |
-
"epoch": 0.8165322580645161,
|
| 596 |
-
"grad_norm": 1.3170841848170978,
|
| 597 |
-
"learning_rate": 1.4540799071326888e-06,
|
| 598 |
-
"loss": 1.916,
|
| 599 |
-
"step": 405
|
| 600 |
-
},
|
| 601 |
-
{
|
| 602 |
-
"epoch": 0.8266129032258065,
|
| 603 |
-
"grad_norm": 1.2106133937481183,
|
| 604 |
-
"learning_rate": 1.4060893485251885e-06,
|
| 605 |
-
"loss": 1.9357,
|
| 606 |
-
"step": 410
|
| 607 |
-
},
|
| 608 |
-
{
|
| 609 |
-
"epoch": 0.8366935483870968,
|
| 610 |
-
"grad_norm": 1.1609644269427515,
|
| 611 |
-
"learning_rate": 1.3595946220786867e-06,
|
| 612 |
-
"loss": 1.9729,
|
| 613 |
-
"step": 415
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 0.8467741935483871,
|
| 617 |
-
"grad_norm": 1.255134988378285,
|
| 618 |
-
"learning_rate": 1.3145536345428467e-06,
|
| 619 |
-
"loss": 1.8483,
|
| 620 |
-
"step": 420
|
| 621 |
-
},
|
| 622 |
-
{
|
| 623 |
-
"epoch": 0.8568548387096774,
|
| 624 |
-
"grad_norm": 1.2271496366595076,
|
| 625 |
-
"learning_rate": 1.2709253488860179e-06,
|
| 626 |
-
"loss": 1.9611,
|
| 627 |
-
"step": 425
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 0.8669354838709677,
|
| 631 |
-
"grad_norm": 1.2112912625332515,
|
| 632 |
-
"learning_rate": 1.228669761031903e-06,
|
| 633 |
-
"loss": 1.8703,
|
| 634 |
-
"step": 430
|
| 635 |
-
},
|
| 636 |
-
{
|
| 637 |
-
"epoch": 0.8770161290322581,
|
| 638 |
-
"grad_norm": 1.1796106189216746,
|
| 639 |
-
"learning_rate": 1.1877478770367983e-06,
|
| 640 |
-
"loss": 1.8931,
|
| 641 |
-
"step": 435
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch": 0.8870967741935484,
|
| 645 |
-
"grad_norm": 1.2743988647789708,
|
| 646 |
-
"learning_rate": 1.1481216907004316e-06,
|
| 647 |
-
"loss": 2.0008,
|
| 648 |
-
"step": 440
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 0.8971774193548387,
|
| 652 |
-
"grad_norm": 1.109580598205878,
|
| 653 |
-
"learning_rate": 1.109754161603501e-06,
|
| 654 |
-
"loss": 1.8754,
|
| 655 |
-
"step": 445
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 0.907258064516129,
|
| 659 |
-
"grad_norm": 1.3265747651562068,
|
| 660 |
-
"learning_rate": 1.07260919356514e-06,
|
| 661 |
-
"loss": 1.937,
|
| 662 |
-
"step": 450
|
| 663 |
-
},
|
| 664 |
-
{
|
| 665 |
-
"epoch": 0.9173387096774194,
|
| 666 |
-
"grad_norm": 1.2255846935811405,
|
| 667 |
-
"learning_rate": 1.036651613513567e-06,
|
| 668 |
-
"loss": 1.839,
|
| 669 |
-
"step": 455
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.9274193548387096,
|
| 673 |
-
"grad_norm": 1.1637588001750867,
|
| 674 |
-
"learning_rate": 1.0018471507633137e-06,
|
| 675 |
-
"loss": 1.8875,
|
| 676 |
-
"step": 460
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 0.9375,
|
| 680 |
-
"grad_norm": 1.2461604557758548,
|
| 681 |
-
"learning_rate": 9.681624166924873e-07,
|
| 682 |
-
"loss": 2.0349,
|
| 683 |
-
"step": 465
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 0.9475806451612904,
|
| 687 |
-
"grad_norm": 1.106959630092776,
|
| 688 |
-
"learning_rate": 9.355648848136057e-07,
|
| 689 |
-
"loss": 1.7721,
|
| 690 |
-
"step": 470
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 0.9576612903225806,
|
| 694 |
-
"grad_norm": 1.1845513909345546,
|
| 695 |
-
"learning_rate": 9.040228712316404e-07,
|
| 696 |
-
"loss": 1.9764,
|
| 697 |
-
"step": 475
|
| 698 |
-
},
|
| 699 |
-
{
|
| 700 |
-
"epoch": 0.967741935483871,
|
| 701 |
-
"grad_norm": 1.1157403262103818,
|
| 702 |
-
"learning_rate": 8.735055154829718e-07,
|
| 703 |
-
"loss": 1.951,
|
| 704 |
-
"step": 480
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 0.9778225806451613,
|
| 708 |
-
"grad_norm": 1.184174473692059,
|
| 709 |
-
"learning_rate": 8.439827617490521e-07,
|
| 710 |
-
"loss": 1.9679,
|
| 711 |
-
"step": 485
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 0.9879032258064516,
|
| 715 |
-
"grad_norm": 1.2877898930931635,
|
| 716 |
-
"learning_rate": 8.154253404386484e-07,
|
| 717 |
-
"loss": 1.9346,
|
| 718 |
-
"step": 490
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 0.9979838709677419,
|
| 722 |
-
"grad_norm": 1.1827480417909397,
|
| 723 |
-
"learning_rate": 7.878047501326227e-07,
|
| 724 |
-
"loss": 1.9182,
|
| 725 |
-
"step": 495
|
| 726 |
-
},
|
| 727 |
-
{
|
| 728 |
-
"epoch": 1.0080645161290323,
|
| 729 |
-
"grad_norm": 1.3196758723195976,
|
| 730 |
-
"learning_rate": 7.61093239885272e-07,
|
| 731 |
-
"loss": 1.7384,
|
| 732 |
-
"step": 500
|
| 733 |
-
},
|
| 734 |
-
{
|
| 735 |
-
"epoch": 1.0181451612903225,
|
| 736 |
-
"grad_norm": 1.2675698548410592,
|
| 737 |
-
"learning_rate": 7.35263791876364e-07,
|
| 738 |
-
"loss": 1.7799,
|
| 739 |
-
"step": 505
|
| 740 |
-
},
|
| 741 |
-
{
|
| 742 |
-
"epoch": 1.028225806451613,
|
| 743 |
-
"grad_norm": 1.3275129553040272,
|
| 744 |
-
"learning_rate": 7.102901044080282e-07,
|
| 745 |
-
"loss": 1.7003,
|
| 746 |
-
"step": 510
|
| 747 |
-
},
|
| 748 |
-
{
|
| 749 |
-
"epoch": 1.0383064516129032,
|
| 750 |
-
"grad_norm": 1.4098711468145062,
|
| 751 |
-
"learning_rate": 6.861465752407945e-07,
|
| 752 |
-
"loss": 1.7438,
|
| 753 |
-
"step": 515
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 1.0483870967741935,
|
| 757 |
-
"grad_norm": 1.46840354164103,
|
| 758 |
-
"learning_rate": 6.628082852631131e-07,
|
| 759 |
-
"loss": 1.8117,
|
| 760 |
-
"step": 520
|
| 761 |
-
},
|
| 762 |
-
{
|
| 763 |
-
"epoch": 1.0584677419354838,
|
| 764 |
-
"grad_norm": 1.2413756391084652,
|
| 765 |
-
"learning_rate": 6.402509824887728e-07,
|
| 766 |
-
"loss": 1.7275,
|
| 767 |
-
"step": 525
|
| 768 |
-
},
|
| 769 |
-
{
|
| 770 |
-
"epoch": 1.0685483870967742,
|
| 771 |
-
"grad_norm": 1.1966512738085646,
|
| 772 |
-
"learning_rate": 6.184510663767266e-07,
|
| 773 |
-
"loss": 1.7607,
|
| 774 |
-
"step": 530
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 1.0786290322580645,
|
| 778 |
-
"grad_norm": 1.1871529707057311,
|
| 779 |
-
"learning_rate": 5.973855724678734e-07,
|
| 780 |
-
"loss": 1.7508,
|
| 781 |
-
"step": 535
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 1.0887096774193548,
|
| 785 |
-
"grad_norm": 1.2178547151060322,
|
| 786 |
-
"learning_rate": 5.770321573334551e-07,
|
| 787 |
-
"loss": 1.7386,
|
| 788 |
-
"step": 540
|
| 789 |
-
},
|
| 790 |
-
{
|
| 791 |
-
"epoch": 1.0987903225806452,
|
| 792 |
-
"grad_norm": 1.215474948218683,
|
| 793 |
-
"learning_rate": 5.573690838297752e-07,
|
| 794 |
-
"loss": 1.7264,
|
| 795 |
-
"step": 545
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 1.1088709677419355,
|
| 799 |
-
"grad_norm": 1.3022255779670897,
|
| 800 |
-
"learning_rate": 5.383752066540304e-07,
|
| 801 |
-
"loss": 1.7367,
|
| 802 |
-
"step": 550
|
| 803 |
-
},
|
| 804 |
-
{
|
| 805 |
-
"epoch": 1.1189516129032258,
|
| 806 |
-
"grad_norm": 1.2677022586640083,
|
| 807 |
-
"learning_rate": 5.200299581961105e-07,
|
| 808 |
-
"loss": 1.8039,
|
| 809 |
-
"step": 555
|
| 810 |
-
},
|
| 811 |
-
{
|
| 812 |
-
"epoch": 1.129032258064516,
|
| 813 |
-
"grad_norm": 1.2498635415826649,
|
| 814 |
-
"learning_rate": 5.023133346812977e-07,
|
| 815 |
-
"loss": 1.7811,
|
| 816 |
-
"step": 560
|
| 817 |
-
},
|
| 818 |
-
{
|
| 819 |
-
"epoch": 1.1391129032258065,
|
| 820 |
-
"grad_norm": 1.1813135414245424,
|
| 821 |
-
"learning_rate": 4.85205882598862e-07,
|
| 822 |
-
"loss": 1.7935,
|
| 823 |
-
"step": 565
|
| 824 |
-
},
|
| 825 |
-
{
|
| 826 |
-
"epoch": 1.1491935483870968,
|
| 827 |
-
"grad_norm": 1.283761351053989,
|
| 828 |
-
"learning_rate": 4.686886854116219e-07,
|
| 829 |
-
"loss": 1.7428,
|
| 830 |
-
"step": 570
|
| 831 |
-
},
|
| 832 |
-
{
|
| 833 |
-
"epoch": 1.159274193548387,
|
| 834 |
-
"grad_norm": 1.181148159521782,
|
| 835 |
-
"learning_rate": 4.5274335054160525e-07,
|
| 836 |
-
"loss": 1.7405,
|
| 837 |
-
"step": 575
|
| 838 |
-
},
|
| 839 |
-
{
|
| 840 |
-
"epoch": 1.1693548387096775,
|
| 841 |
-
"grad_norm": 1.2540563610260587,
|
| 842 |
-
"learning_rate": 4.3735199662701383e-07,
|
| 843 |
-
"loss": 1.7017,
|
| 844 |
-
"step": 580
|
| 845 |
-
},
|
| 846 |
-
{
|
| 847 |
-
"epoch": 1.1794354838709677,
|
| 848 |
-
"grad_norm": 1.26673278964677,
|
| 849 |
-
"learning_rate": 4.2249724104576337e-07,
|
| 850 |
-
"loss": 1.7258,
|
| 851 |
-
"step": 585
|
| 852 |
-
},
|
| 853 |
-
{
|
| 854 |
-
"epoch": 1.189516129032258,
|
| 855 |
-
"grad_norm": 1.2483740436576478,
|
| 856 |
-
"learning_rate": 4.081621877009305e-07,
|
| 857 |
-
"loss": 1.7842,
|
| 858 |
-
"step": 590
|
| 859 |
-
},
|
| 860 |
-
{
|
| 861 |
-
"epoch": 1.1995967741935485,
|
| 862 |
-
"grad_norm": 1.2338327909706366,
|
| 863 |
-
"learning_rate": 3.9433041506352036e-07,
|
| 864 |
-
"loss": 1.7602,
|
| 865 |
-
"step": 595
|
| 866 |
-
},
|
| 867 |
-
{
|
| 868 |
-
"epoch": 1.2096774193548387,
|
| 869 |
-
"grad_norm": 1.2070627326429186,
|
| 870 |
-
"learning_rate": 3.8098596446800587e-07,
|
| 871 |
-
"loss": 1.6383,
|
| 872 |
-
"step": 600
|
| 873 |
-
},
|
| 874 |
-
{
|
| 875 |
-
"epoch": 1.2096774193548387,
|
| 876 |
-
"eval_loss": 1.9439929723739624,
|
| 877 |
-
"eval_runtime": 89.7597,
|
| 878 |
-
"eval_samples_per_second": 19.619,
|
| 879 |
-
"eval_steps_per_second": 0.312,
|
| 880 |
-
"step": 600
|
| 881 |
-
},
|
| 882 |
-
{
|
| 883 |
-
"epoch": 1.219758064516129,
|
| 884 |
-
"grad_norm": 1.2154229290189673,
|
| 885 |
-
"learning_rate": 3.681133286561828e-07,
|
| 886 |
-
"loss": 1.8134,
|
| 887 |
-
"step": 605
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 1.2298387096774193,
|
| 891 |
-
"grad_norm": 1.2295658806251533,
|
| 892 |
-
"learning_rate": 3.556974405649289e-07,
|
| 893 |
-
"loss": 1.725,
|
| 894 |
-
"step": 610
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 1.2399193548387097,
|
| 898 |
-
"grad_norm": 1.2791124588407303,
|
| 899 |
-
"learning_rate": 3.437236623535232e-07,
|
| 900 |
-
"loss": 1.7526,
|
| 901 |
-
"step": 615
|
| 902 |
-
},
|
| 903 |
-
{
|
| 904 |
-
"epoch": 1.25,
|
| 905 |
-
"grad_norm": 1.1899209284700905,
|
| 906 |
-
"learning_rate": 3.3217777466625205e-07,
|
| 907 |
-
"loss": 1.7313,
|
| 908 |
-
"step": 620
|
| 909 |
-
},
|
| 910 |
-
{
|
| 911 |
-
"epoch": 1.2600806451612903,
|
| 912 |
-
"grad_norm": 1.3442383601969903,
|
| 913 |
-
"learning_rate": 3.2104596612607314e-07,
|
| 914 |
-
"loss": 1.8358,
|
| 915 |
-
"step": 625
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 1.2701612903225805,
|
| 919 |
-
"grad_norm": 1.2622913229084678,
|
| 920 |
-
"learning_rate": 3.1031482305518425e-07,
|
| 921 |
-
"loss": 1.7504,
|
| 922 |
-
"step": 630
|
| 923 |
-
},
|
| 924 |
-
{
|
| 925 |
-
"epoch": 1.280241935483871,
|
| 926 |
-
"grad_norm": 1.1574713957890075,
|
| 927 |
-
"learning_rate": 2.9997131941839894e-07,
|
| 928 |
-
"loss": 1.7646,
|
| 929 |
-
"step": 635
|
| 930 |
-
},
|
| 931 |
-
{
|
| 932 |
-
"epoch": 1.2903225806451613,
|
| 933 |
-
"grad_norm": 1.346661773742281,
|
| 934 |
-
"learning_rate": 2.9000280698528666e-07,
|
| 935 |
-
"loss": 1.7116,
|
| 936 |
-
"step": 640
|
| 937 |
-
},
|
| 938 |
-
{
|
| 939 |
-
"epoch": 1.3004032258064515,
|
| 940 |
-
"grad_norm": 1.3332483111343922,
|
| 941 |
-
"learning_rate": 2.8039700570709997e-07,
|
| 942 |
-
"loss": 1.8109,
|
| 943 |
-
"step": 645
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 1.310483870967742,
|
| 947 |
-
"grad_norm": 1.2333462778908426,
|
| 948 |
-
"learning_rate": 2.7114199430456553e-07,
|
| 949 |
-
"loss": 1.7683,
|
| 950 |
-
"step": 650
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 1.3205645161290323,
|
| 954 |
-
"grad_norm": 1.1938385640085034,
|
| 955 |
-
"learning_rate": 2.6222620106267406e-07,
|
| 956 |
-
"loss": 1.6945,
|
| 957 |
-
"step": 655
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 1.3306451612903225,
|
| 961 |
-
"grad_norm": 1.4010824213914208,
|
| 962 |
-
"learning_rate": 2.536383948286623e-07,
|
| 963 |
-
"loss": 1.7537,
|
| 964 |
-
"step": 660
|
| 965 |
-
},
|
| 966 |
-
{
|
| 967 |
-
"epoch": 1.340725806451613,
|
| 968 |
-
"grad_norm": 1.3881248076295551,
|
| 969 |
-
"learning_rate": 2.4536767620943495e-07,
|
| 970 |
-
"loss": 1.7843,
|
| 971 |
-
"step": 665
|
| 972 |
-
},
|
| 973 |
-
{
|
| 974 |
-
"epoch": 1.3508064516129032,
|
| 975 |
-
"grad_norm": 1.2168315394832683,
|
| 976 |
-
"learning_rate": 2.3740346896473077e-07,
|
| 977 |
-
"loss": 1.7928,
|
| 978 |
-
"step": 670
|
| 979 |
-
},
|
| 980 |
-
{
|
| 981 |
-
"epoch": 1.3608870967741935,
|
| 982 |
-
"grad_norm": 1.4475784692183353,
|
| 983 |
-
"learning_rate": 2.2973551159239254e-07,
|
| 984 |
-
"loss": 1.7755,
|
| 985 |
-
"step": 675
|
| 986 |
-
},
|
| 987 |
-
{
|
| 988 |
-
"epoch": 1.370967741935484,
|
| 989 |
-
"grad_norm": 1.182910838845923,
|
| 990 |
-
"learning_rate": 2.223538491021522e-07,
|
| 991 |
-
"loss": 1.7251,
|
| 992 |
-
"step": 680
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 1.3810483870967742,
|
| 996 |
-
"grad_norm": 1.226170490302768,
|
| 997 |
-
"learning_rate": 2.1524882497440346e-07,
|
| 998 |
-
"loss": 1.7814,
|
| 999 |
-
"step": 685
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 1.3911290322580645,
|
| 1003 |
-
"grad_norm": 1.1753601198052888,
|
| 1004 |
-
"learning_rate": 2.084110733004757e-07,
|
| 1005 |
-
"loss": 1.7355,
|
| 1006 |
-
"step": 690
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 1.4012096774193548,
|
| 1010 |
-
"grad_norm": 1.176444951923013,
|
| 1011 |
-
"learning_rate": 2.0183151110098748e-07,
|
| 1012 |
-
"loss": 1.7006,
|
| 1013 |
-
"step": 695
|
| 1014 |
-
},
|
| 1015 |
-
{
|
| 1016 |
-
"epoch": 1.4112903225806452,
|
| 1017 |
-
"grad_norm": 1.2575906929525473,
|
| 1018 |
-
"learning_rate": 1.955013308189006e-07,
|
| 1019 |
-
"loss": 1.6631,
|
| 1020 |
-
"step": 700
|
| 1021 |
-
},
|
| 1022 |
-
{
|
| 1023 |
-
"epoch": 1.4213709677419355,
|
| 1024 |
-
"grad_norm": 1.3222267696078929,
|
| 1025 |
-
"learning_rate": 1.8941199298395633e-07,
|
| 1026 |
-
"loss": 1.7313,
|
| 1027 |
-
"step": 705
|
| 1028 |
-
},
|
| 1029 |
-
{
|
| 1030 |
-
"epoch": 1.4314516129032258,
|
| 1031 |
-
"grad_norm": 1.3554981903288013,
|
| 1032 |
-
"learning_rate": 1.8355521904521423e-07,
|
| 1033 |
-
"loss": 1.7183,
|
| 1034 |
-
"step": 710
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 1.441532258064516,
|
| 1038 |
-
"grad_norm": 1.2333423736099849,
|
| 1039 |
-
"learning_rate": 1.7792298436847624e-07,
|
| 1040 |
-
"loss": 1.7057,
|
| 1041 |
-
"step": 715
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 1.4516129032258065,
|
| 1045 |
-
"grad_norm": 1.2791350480965469,
|
| 1046 |
-
"learning_rate": 1.7250751139542018e-07,
|
| 1047 |
-
"loss": 1.6957,
|
| 1048 |
-
"step": 720
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 1.4616935483870968,
|
| 1052 |
-
"grad_norm": 1.5651823236688753,
|
| 1053 |
-
"learning_rate": 1.6730126296131982e-07,
|
| 1054 |
-
"loss": 1.808,
|
| 1055 |
-
"step": 725
|
| 1056 |
-
},
|
| 1057 |
-
{
|
| 1058 |
-
"epoch": 1.471774193548387,
|
| 1059 |
-
"grad_norm": 1.3083961982147532,
|
| 1060 |
-
"learning_rate": 1.622969357682752e-07,
|
| 1061 |
-
"loss": 1.7913,
|
| 1062 |
-
"step": 730
|
| 1063 |
-
},
|
| 1064 |
-
{
|
| 1065 |
-
"epoch": 1.4818548387096775,
|
| 1066 |
-
"grad_norm": 1.2078318183364274,
|
| 1067 |
-
"learning_rate": 1.574874540109262e-07,
|
| 1068 |
-
"loss": 1.7273,
|
| 1069 |
-
"step": 735
|
| 1070 |
-
},
|
| 1071 |
-
{
|
| 1072 |
-
"epoch": 1.4919354838709677,
|
| 1073 |
-
"grad_norm": 1.2904657184612536,
|
| 1074 |
-
"learning_rate": 1.5286596315166833e-07,
|
| 1075 |
-
"loss": 1.6646,
|
| 1076 |
-
"step": 740
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 1.502016129032258,
|
| 1080 |
-
"grad_norm": 1.240861344603422,
|
| 1081 |
-
"learning_rate": 1.4842582384243795e-07,
|
| 1082 |
-
"loss": 1.6495,
|
| 1083 |
-
"step": 745
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 1.5120967741935485,
|
| 1087 |
-
"grad_norm": 1.2574003293432507,
|
| 1088 |
-
"learning_rate": 1.4416060599017888e-07,
|
| 1089 |
-
"loss": 1.7356,
|
| 1090 |
-
"step": 750
|
| 1091 |
-
},
|
| 1092 |
-
{
|
| 1093 |
-
"epoch": 1.5221774193548387,
|
| 1094 |
-
"grad_norm": 1.4067959486898025,
|
| 1095 |
-
"learning_rate": 1.4006408296314945e-07,
|
| 1096 |
-
"loss": 1.7802,
|
| 1097 |
-
"step": 755
|
| 1098 |
-
},
|
| 1099 |
-
{
|
| 1100 |
-
"epoch": 1.532258064516129,
|
| 1101 |
-
"grad_norm": 1.3011200023092266,
|
| 1102 |
-
"learning_rate": 1.3613022593527299e-07,
|
| 1103 |
-
"loss": 1.7435,
|
| 1104 |
-
"step": 760
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 1.5423387096774195,
|
| 1108 |
-
"grad_norm": 1.5064955060135545,
|
| 1109 |
-
"learning_rate": 1.3235319836577956e-07,
|
| 1110 |
-
"loss": 1.8153,
|
| 1111 |
-
"step": 765
|
| 1112 |
-
},
|
| 1113 |
-
{
|
| 1114 |
-
"epoch": 1.5524193548387095,
|
| 1115 |
-
"grad_norm": 1.2545915727253294,
|
| 1116 |
-
"learning_rate": 1.287273506114331e-07,
|
| 1117 |
-
"loss": 1.7333,
|
| 1118 |
-
"step": 770
|
| 1119 |
-
},
|
| 1120 |
-
{
|
| 1121 |
-
"epoch": 1.5625,
|
| 1122 |
-
"grad_norm": 1.210516142345582,
|
| 1123 |
-
"learning_rate": 1.2524721466867644e-07,
|
| 1124 |
-
"loss": 1.8145,
|
| 1125 |
-
"step": 775
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 1.5725806451612905,
|
| 1129 |
-
"grad_norm": 1.2428363984374335,
|
| 1130 |
-
"learning_rate": 1.2190749904307564e-07,
|
| 1131 |
-
"loss": 1.7466,
|
| 1132 |
-
"step": 780
|
| 1133 |
-
},
|
| 1134 |
-
{
|
| 1135 |
-
"epoch": 1.5826612903225805,
|
| 1136 |
-
"grad_norm": 1.2372984868597403,
|
| 1137 |
-
"learning_rate": 1.187030837434837e-07,
|
| 1138 |
-
"loss": 1.7707,
|
| 1139 |
-
"step": 785
|
| 1140 |
-
},
|
| 1141 |
-
{
|
| 1142 |
-
"epoch": 1.592741935483871,
|
| 1143 |
-
"grad_norm": 1.223197065220676,
|
| 1144 |
-
"learning_rate": 1.1562901539838627e-07,
|
| 1145 |
-
"loss": 1.7515,
|
| 1146 |
-
"step": 790
|
| 1147 |
-
},
|
| 1148 |
-
{
|
| 1149 |
-
"epoch": 1.6028225806451613,
|
| 1150 |
-
"grad_norm": 1.360461478384606,
|
| 1151 |
-
"learning_rate": 1.1268050249193624e-07,
|
| 1152 |
-
"loss": 1.7033,
|
| 1153 |
-
"step": 795
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 1.6129032258064515,
|
| 1157 |
-
"grad_norm": 1.3523444473608008,
|
| 1158 |
-
"learning_rate": 1.0985291071721938e-07,
|
| 1159 |
-
"loss": 1.7073,
|
| 1160 |
-
"step": 800
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 1.6129032258064515,
|
| 1164 |
-
"eval_loss": 1.9434727430343628,
|
| 1165 |
-
"eval_runtime": 89.696,
|
| 1166 |
-
"eval_samples_per_second": 19.633,
|
| 1167 |
-
"eval_steps_per_second": 0.312,
|
| 1168 |
-
"step": 800
|
| 1169 |
-
},
|
| 1170 |
-
{
|
| 1171 |
-
"epoch": 1.622983870967742,
|
| 1172 |
-
"grad_norm": 1.1908153349621502,
|
| 1173 |
-
"learning_rate": 1.0714175844433986e-07,
|
| 1174 |
-
"loss": 1.7606,
|
| 1175 |
-
"step": 805
|
| 1176 |
-
},
|
| 1177 |
-
{
|
| 1178 |
-
"epoch": 1.6330645161290323,
|
| 1179 |
-
"grad_norm": 1.1559466036906028,
|
| 1180 |
-
"learning_rate": 1.0454271230095008e-07,
|
| 1181 |
-
"loss": 1.6855,
|
| 1182 |
-
"step": 810
|
| 1183 |
-
},
|
| 1184 |
-
{
|
| 1185 |
-
"epoch": 1.6431451612903225,
|
| 1186 |
-
"grad_norm": 1.2177765725388214,
|
| 1187 |
-
"learning_rate": 1.0205158286289079e-07,
|
| 1188 |
-
"loss": 1.6938,
|
| 1189 |
-
"step": 815
|
| 1190 |
-
},
|
| 1191 |
-
{
|
| 1192 |
-
"epoch": 1.653225806451613,
|
| 1193 |
-
"grad_norm": 1.2381331316689286,
|
| 1194 |
-
"learning_rate": 9.966432045264543e-08,
|
| 1195 |
-
"loss": 1.7047,
|
| 1196 |
-
"step": 820
|
| 1197 |
-
},
|
| 1198 |
-
{
|
| 1199 |
-
"epoch": 1.6633064516129032,
|
| 1200 |
-
"grad_norm": 1.2662439850688683,
|
| 1201 |
-
"learning_rate": 9.73770110433516e-08,
|
| 1202 |
-
"loss": 1.8619,
|
| 1203 |
-
"step": 825
|
| 1204 |
-
},
|
| 1205 |
-
{
|
| 1206 |
-
"epoch": 1.6733870967741935,
|
| 1207 |
-
"grad_norm": 1.3616392797318724,
|
| 1208 |
-
"learning_rate": 9.518587226615035e-08,
|
| 1209 |
-
"loss": 1.8544,
|
| 1210 |
-
"step": 830
|
| 1211 |
-
},
|
| 1212 |
-
{
|
| 1213 |
-
"epoch": 1.683467741935484,
|
| 1214 |
-
"grad_norm": 1.2521663767495246,
|
| 1215 |
-
"learning_rate": 9.30872495186915e-08,
|
| 1216 |
-
"loss": 1.7568,
|
| 1217 |
-
"step": 835
|
| 1218 |
-
},
|
| 1219 |
-
{
|
| 1220 |
-
"epoch": 1.6935483870967742,
|
| 1221 |
-
"grad_norm": 1.2504361983829193,
|
| 1222 |
-
"learning_rate": 9.107761217265013e-08,
|
| 1223 |
-
"loss": 1.721,
|
| 1224 |
-
"step": 840
|
| 1225 |
-
},
|
| 1226 |
-
{
|
| 1227 |
-
"epoch": 1.7036290322580645,
|
| 1228 |
-
"grad_norm": 1.2617228722420664,
|
| 1229 |
-
"learning_rate": 8.915354987814596e-08,
|
| 1230 |
-
"loss": 1.7788,
|
| 1231 |
-
"step": 845
|
| 1232 |
-
},
|
| 1233 |
-
{
|
| 1234 |
-
"epoch": 1.713709677419355,
|
| 1235 |
-
"grad_norm": 1.2060296867893439,
|
| 1236 |
-
"learning_rate": 8.731176896299368e-08,
|
| 1237 |
-
"loss": 1.7307,
|
| 1238 |
-
"step": 850
|
| 1239 |
-
},
|
| 1240 |
-
{
|
| 1241 |
-
"epoch": 1.723790322580645,
|
| 1242 |
-
"grad_norm": 1.2261412385974246,
|
| 1243 |
-
"learning_rate": 8.554908892474695e-08,
|
| 1244 |
-
"loss": 1.6492,
|
| 1245 |
-
"step": 855
|
| 1246 |
-
},
|
| 1247 |
-
{
|
| 1248 |
-
"epoch": 1.7338709677419355,
|
| 1249 |
-
"grad_norm": 1.1985063418448514,
|
| 1250 |
-
"learning_rate": 8.38624390135359e-08,
|
| 1251 |
-
"loss": 1.8062,
|
| 1252 |
-
"step": 860
|
| 1253 |
-
},
|
| 1254 |
-
{
|
| 1255 |
-
"epoch": 1.7439516129032258,
|
| 1256 |
-
"grad_norm": 1.1687433855869847,
|
| 1257 |
-
"learning_rate": 8.224885490372989e-08,
|
| 1258 |
-
"loss": 1.6903,
|
| 1259 |
-
"step": 865
|
| 1260 |
-
},
|
| 1261 |
-
{
|
| 1262 |
-
"epoch": 1.754032258064516,
|
| 1263 |
-
"grad_norm": 1.2094813386232657,
|
| 1264 |
-
"learning_rate": 8.070547545249396e-08,
|
| 1265 |
-
"loss": 1.7466,
|
| 1266 |
-
"step": 870
|
| 1267 |
-
},
|
| 1268 |
-
{
|
| 1269 |
-
"epoch": 1.7641129032258065,
|
| 1270 |
-
"grad_norm": 1.5715777085972131,
|
| 1271 |
-
"learning_rate": 7.922953954333971e-08,
|
| 1272 |
-
"loss": 1.7321,
|
| 1273 |
-
"step": 875
|
| 1274 |
-
},
|
| 1275 |
-
{
|
| 1276 |
-
"epoch": 1.7741935483870968,
|
| 1277 |
-
"grad_norm": 1.232854137306794,
|
| 1278 |
-
"learning_rate": 7.781838301280604e-08,
|
| 1279 |
-
"loss": 1.7854,
|
| 1280 |
-
"step": 880
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 1.784274193548387,
|
| 1284 |
-
"grad_norm": 1.2306505361743405,
|
| 1285 |
-
"learning_rate": 7.646943565843591e-08,
|
| 1286 |
-
"loss": 1.7035,
|
| 1287 |
-
"step": 885
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 1.7943548387096775,
|
| 1291 |
-
"grad_norm": 1.4038527293019962,
|
| 1292 |
-
"learning_rate": 7.518021832625021e-08,
|
| 1293 |
-
"loss": 1.7898,
|
| 1294 |
-
"step": 890
|
| 1295 |
-
},
|
| 1296 |
-
{
|
| 1297 |
-
"epoch": 1.8044354838709677,
|
| 1298 |
-
"grad_norm": 1.1701930640429434,
|
| 1299 |
-
"learning_rate": 7.39483400759502e-08,
|
| 1300 |
-
"loss": 1.7473,
|
| 1301 |
-
"step": 895
|
| 1302 |
-
},
|
| 1303 |
-
{
|
| 1304 |
-
"epoch": 1.814516129032258,
|
| 1305 |
-
"grad_norm": 1.165650133692535,
|
| 1306 |
-
"learning_rate": 7.277149542211219e-08,
|
| 1307 |
-
"loss": 1.7392,
|
| 1308 |
-
"step": 900
|
| 1309 |
-
},
|
| 1310 |
-
{
|
| 1311 |
-
"epoch": 1.8245967741935485,
|
| 1312 |
-
"grad_norm": 1.242160120226514,
|
| 1313 |
-
"learning_rate": 7.164746164966925e-08,
|
| 1314 |
-
"loss": 1.691,
|
| 1315 |
-
"step": 905
|
| 1316 |
-
},
|
| 1317 |
-
{
|
| 1318 |
-
"epoch": 1.8346774193548387,
|
| 1319 |
-
"grad_norm": 1.3662669550498554,
|
| 1320 |
-
"learning_rate": 7.057409620200535e-08,
|
| 1321 |
-
"loss": 1.7456,
|
| 1322 |
-
"step": 910
|
| 1323 |
-
},
|
| 1324 |
-
{
|
| 1325 |
-
"epoch": 1.844758064516129,
|
| 1326 |
-
"grad_norm": 1.2736994338683274,
|
| 1327 |
-
"learning_rate": 6.954933414001788e-08,
|
| 1328 |
-
"loss": 1.9068,
|
| 1329 |
-
"step": 915
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 1.8548387096774195,
|
| 1333 |
-
"grad_norm": 1.3755517276213882,
|
| 1334 |
-
"learning_rate": 6.857118567053417e-08,
|
| 1335 |
-
"loss": 1.7568,
|
| 1336 |
-
"step": 920
|
| 1337 |
-
},
|
| 1338 |
-
{
|
| 1339 |
-
"epoch": 1.8649193548387095,
|
| 1340 |
-
"grad_norm": 1.1731257131372528,
|
| 1341 |
-
"learning_rate": 6.763773374249751e-08,
|
| 1342 |
-
"loss": 1.7815,
|
| 1343 |
-
"step": 925
|
| 1344 |
-
},
|
| 1345 |
-
{
|
| 1346 |
-
"epoch": 1.875,
|
| 1347 |
-
"grad_norm": 1.2300272384037476,
|
| 1348 |
-
"learning_rate": 6.674713170936664e-08,
|
| 1349 |
-
"loss": 1.7488,
|
| 1350 |
-
"step": 930
|
| 1351 |
-
},
|
| 1352 |
-
{
|
| 1353 |
-
"epoch": 1.8850806451612905,
|
| 1354 |
-
"grad_norm": 1.2420985179509971,
|
| 1355 |
-
"learning_rate": 6.589760105620217e-08,
|
| 1356 |
-
"loss": 1.7614,
|
| 1357 |
-
"step": 935
|
| 1358 |
-
},
|
| 1359 |
-
{
|
| 1360 |
-
"epoch": 1.8951612903225805,
|
| 1361 |
-
"grad_norm": 1.7224910659058157,
|
| 1362 |
-
"learning_rate": 6.508742918994098e-08,
|
| 1363 |
-
"loss": 1.7496,
|
| 1364 |
-
"step": 940
|
| 1365 |
-
},
|
| 1366 |
-
{
|
| 1367 |
-
"epoch": 1.905241935483871,
|
| 1368 |
-
"grad_norm": 1.3013159637595342,
|
| 1369 |
-
"learning_rate": 6.431496729138832e-08,
|
| 1370 |
-
"loss": 1.7375,
|
| 1371 |
-
"step": 945
|
| 1372 |
-
},
|
| 1373 |
-
{
|
| 1374 |
-
"epoch": 1.9153225806451613,
|
| 1375 |
-
"grad_norm": 1.2702415604664459,
|
| 1376 |
-
"learning_rate": 6.357862822748362e-08,
|
| 1377 |
-
"loss": 1.7668,
|
| 1378 |
-
"step": 950
|
| 1379 |
-
},
|
| 1380 |
-
{
|
| 1381 |
-
"epoch": 1.9254032258064515,
|
| 1382 |
-
"grad_norm": 1.168954401665018,
|
| 1383 |
-
"learning_rate": 6.287688452242478e-08,
|
| 1384 |
-
"loss": 1.7489,
|
| 1385 |
-
"step": 955
|
| 1386 |
-
},
|
| 1387 |
-
{
|
| 1388 |
-
"epoch": 1.935483870967742,
|
| 1389 |
-
"grad_norm": 1.3375224905639664,
|
| 1390 |
-
"learning_rate": 6.220826638626108e-08,
|
| 1391 |
-
"loss": 1.6308,
|
| 1392 |
-
"step": 960
|
| 1393 |
-
},
|
| 1394 |
-
{
|
| 1395 |
-
"epoch": 1.9455645161290323,
|
| 1396 |
-
"grad_norm": 1.2577572377495214,
|
| 1397 |
-
"learning_rate": 6.15713597995916e-08,
|
| 1398 |
-
"loss": 1.7912,
|
| 1399 |
-
"step": 965
|
| 1400 |
-
},
|
| 1401 |
-
{
|
| 1402 |
-
"epoch": 1.9556451612903225,
|
| 1403 |
-
"grad_norm": 1.2336415332120372,
|
| 1404 |
-
"learning_rate": 6.096480465303296e-08,
|
| 1405 |
-
"loss": 1.7562,
|
| 1406 |
-
"step": 970
|
| 1407 |
-
},
|
| 1408 |
-
{
|
| 1409 |
-
"epoch": 1.965725806451613,
|
| 1410 |
-
"grad_norm": 1.3102010598064548,
|
| 1411 |
-
"learning_rate": 6.038729294014376e-08,
|
| 1412 |
-
"loss": 1.6953,
|
| 1413 |
-
"step": 975
|
| 1414 |
-
},
|
| 1415 |
-
{
|
| 1416 |
-
"epoch": 1.9758064516129032,
|
| 1417 |
-
"grad_norm": 1.3132413812361838,
|
| 1418 |
-
"learning_rate": 5.983756700252073e-08,
|
| 1419 |
-
"loss": 1.7719,
|
| 1420 |
-
"step": 980
|
| 1421 |
-
},
|
| 1422 |
-
{
|
| 1423 |
-
"epoch": 1.9858870967741935,
|
| 1424 |
-
"grad_norm": 1.2310128233562452,
|
| 1425 |
-
"learning_rate": 5.9314417825804556e-08,
|
| 1426 |
-
"loss": 1.7212,
|
| 1427 |
-
"step": 985
|
| 1428 |
-
},
|
| 1429 |
-
{
|
| 1430 |
-
"epoch": 1.995967741935484,
|
| 1431 |
-
"grad_norm": 1.2006191684102008,
|
| 1432 |
-
"learning_rate": 5.8816683385358974e-08,
|
| 1433 |
-
"loss": 1.7667,
|
| 1434 |
-
"step": 990
|
| 1435 |
-
},
|
| 1436 |
-
{
|
| 1437 |
-
"epoch": 2.006048387096774,
|
| 1438 |
-
"grad_norm": 1.2035537415937343,
|
| 1439 |
-
"learning_rate": 5.834324704041041e-08,
|
| 1440 |
-
"loss": 1.7686,
|
| 1441 |
-
"step": 995
|
| 1442 |
-
},
|
| 1443 |
-
{
|
| 1444 |
-
"epoch": 2.0161290322580645,
|
| 1445 |
-
"grad_norm": 1.2316714422775488,
|
| 1446 |
-
"learning_rate": 5.789303597545942e-08,
|
| 1447 |
-
"loss": 1.6699,
|
| 1448 |
-
"step": 1000
|
| 1449 |
-
},
|
| 1450 |
-
{
|
| 1451 |
-
"epoch": 2.0161290322580645,
|
| 1452 |
-
"eval_loss": 1.9428125619888306,
|
| 1453 |
-
"eval_runtime": 89.8315,
|
| 1454 |
-
"eval_samples_per_second": 19.603,
|
| 1455 |
-
"eval_steps_per_second": 0.312,
|
| 1456 |
-
"step": 1000
|
| 1457 |
-
},
|
| 1458 |
-
{
|
| 1459 |
-
"epoch": 2.026209677419355,
|
| 1460 |
-
"grad_norm": 1.2569040983124893,
|
| 1461 |
-
"learning_rate": 5.746501968779849e-08,
|
| 1462 |
-
"loss": 1.7248,
|
| 1463 |
-
"step": 1005
|
| 1464 |
-
},
|
| 1465 |
-
{
|
| 1466 |
-
"epoch": 2.036290322580645,
|
| 1467 |
-
"grad_norm": 1.2073156251056785,
|
| 1468 |
-
"learning_rate": 5.7058208519993964e-08,
|
| 1469 |
-
"loss": 1.5994,
|
| 1470 |
-
"step": 1010
|
| 1471 |
-
},
|
| 1472 |
-
{
|
| 1473 |
-
"epoch": 2.0463709677419355,
|
| 1474 |
-
"grad_norm": 1.250476166765303,
|
| 1475 |
-
"learning_rate": 5.667165223621266e-08,
|
| 1476 |
-
"loss": 1.7618,
|
| 1477 |
-
"step": 1015
|
| 1478 |
-
},
|
| 1479 |
-
{
|
| 1480 |
-
"epoch": 2.056451612903226,
|
| 1481 |
-
"grad_norm": 1.1771740320614876,
|
| 1482 |
-
"learning_rate": 5.630443864129586e-08,
|
| 1483 |
-
"loss": 1.7109,
|
| 1484 |
-
"step": 1020
|
| 1485 |
-
},
|
| 1486 |
-
{
|
| 1487 |
-
"epoch": 2.066532258064516,
|
| 1488 |
-
"grad_norm": 1.2423874053675774,
|
| 1489 |
-
"learning_rate": 5.595569224150602e-08,
|
| 1490 |
-
"loss": 1.688,
|
| 1491 |
-
"step": 1025
|
| 1492 |
-
},
|
| 1493 |
-
{
|
| 1494 |
-
"epoch": 2.0766129032258065,
|
| 1495 |
-
"grad_norm": 1.1913045210590958,
|
| 1496 |
-
"learning_rate": 5.562457294589246e-08,
|
| 1497 |
-
"loss": 1.7122,
|
| 1498 |
-
"step": 1030
|
| 1499 |
-
},
|
| 1500 |
-
{
|
| 1501 |
-
"epoch": 2.086693548387097,
|
| 1502 |
-
"grad_norm": 1.5628319901302377,
|
| 1503 |
-
"learning_rate": 5.53102748072445e-08,
|
| 1504 |
-
"loss": 1.8032,
|
| 1505 |
-
"step": 1035
|
| 1506 |
-
},
|
| 1507 |
-
{
|
| 1508 |
-
"epoch": 2.096774193548387,
|
| 1509 |
-
"grad_norm": 1.2186459343954958,
|
| 1510 |
-
"learning_rate": 5.501202480162106e-08,
|
| 1511 |
-
"loss": 1.706,
|
| 1512 |
-
"step": 1040
|
| 1513 |
-
},
|
| 1514 |
-
{
|
| 1515 |
-
"epoch": 2.1068548387096775,
|
| 1516 |
-
"grad_norm": 1.5980051096787016,
|
| 1517 |
-
"learning_rate": 5.472908164546672e-08,
|
| 1518 |
-
"loss": 1.6859,
|
| 1519 |
-
"step": 1045
|
| 1520 |
-
},
|
| 1521 |
-
{
|
| 1522 |
-
"epoch": 2.1169354838709675,
|
| 1523 |
-
"grad_norm": 1.2032746720272727,
|
| 1524 |
-
"learning_rate": 5.446073464934468e-08,
|
| 1525 |
-
"loss": 1.6677,
|
| 1526 |
-
"step": 1050
|
| 1527 |
-
},
|
| 1528 |
-
{
|
| 1529 |
-
"epoch": 2.127016129032258,
|
| 1530 |
-
"grad_norm": 1.2566967045401873,
|
| 1531 |
-
"learning_rate": 5.420630260733693e-08,
|
| 1532 |
-
"loss": 1.7704,
|
| 1533 |
-
"step": 1055
|
| 1534 |
-
},
|
| 1535 |
-
{
|
| 1536 |
-
"epoch": 2.1370967741935485,
|
| 1537 |
-
"grad_norm": 1.1883212669149759,
|
| 1538 |
-
"learning_rate": 5.396513272118224e-08,
|
| 1539 |
-
"loss": 1.7207,
|
| 1540 |
-
"step": 1060
|
| 1541 |
-
},
|
| 1542 |
-
{
|
| 1543 |
-
"epoch": 2.1471774193548385,
|
| 1544 |
-
"grad_norm": 1.4452669028801322,
|
| 1545 |
-
"learning_rate": 5.373659955824137e-08,
|
| 1546 |
-
"loss": 1.7838,
|
| 1547 |
-
"step": 1065
|
| 1548 |
-
},
|
| 1549 |
-
{
|
| 1550 |
-
"epoch": 2.157258064516129,
|
| 1551 |
-
"grad_norm": 1.2860799707706096,
|
| 1552 |
-
"learning_rate": 5.3520104042398855e-08,
|
| 1553 |
-
"loss": 1.7416,
|
| 1554 |
-
"step": 1070
|
| 1555 |
-
},
|
| 1556 |
-
{
|
| 1557 |
-
"epoch": 2.1673387096774195,
|
| 1558 |
-
"grad_norm": 1.1967308129229457,
|
| 1559 |
-
"learning_rate": 5.331507247702889e-08,
|
| 1560 |
-
"loss": 1.6755,
|
| 1561 |
-
"step": 1075
|
| 1562 |
-
},
|
| 1563 |
-
{
|
| 1564 |
-
"epoch": 2.1774193548387095,
|
| 1565 |
-
"grad_norm": 1.2537902733229633,
|
| 1566 |
-
"learning_rate": 5.3120955599171935e-08,
|
| 1567 |
-
"loss": 1.7548,
|
| 1568 |
-
"step": 1080
|
| 1569 |
-
},
|
| 1570 |
-
{
|
| 1571 |
-
"epoch": 2.1875,
|
| 1572 |
-
"grad_norm": 1.2447529415534258,
|
| 1573 |
-
"learning_rate": 5.293722766408659e-08,
|
| 1574 |
-
"loss": 1.7471,
|
| 1575 |
-
"step": 1085
|
| 1576 |
-
},
|
| 1577 |
-
{
|
| 1578 |
-
"epoch": 2.1975806451612905,
|
| 1579 |
-
"grad_norm": 1.1991864633094496,
|
| 1580 |
-
"learning_rate": 5.276338555935963e-08,
|
| 1581 |
-
"loss": 1.7606,
|
| 1582 |
-
"step": 1090
|
| 1583 |
-
},
|
| 1584 |
-
{
|
| 1585 |
-
"epoch": 2.2076612903225805,
|
| 1586 |
-
"grad_norm": 1.219389195740527,
|
| 1587 |
-
"learning_rate": 5.259894794777423e-08,
|
| 1588 |
-
"loss": 1.7225,
|
| 1589 |
-
"step": 1095
|
| 1590 |
-
},
|
| 1591 |
-
{
|
| 1592 |
-
"epoch": 2.217741935483871,
|
| 1593 |
-
"grad_norm": 1.2339738380016207,
|
| 1594 |
-
"learning_rate": 5.2443454438154534e-08,
|
| 1595 |
-
"loss": 1.7114,
|
| 1596 |
-
"step": 1100
|
| 1597 |
-
},
|
| 1598 |
-
{
|
| 1599 |
-
"epoch": 2.2278225806451615,
|
| 1600 |
-
"grad_norm": 1.1802337869081512,
|
| 1601 |
-
"learning_rate": 5.229646478342111e-08,
|
| 1602 |
-
"loss": 1.694,
|
| 1603 |
-
"step": 1105
|
| 1604 |
-
},
|
| 1605 |
-
{
|
| 1606 |
-
"epoch": 2.2379032258064515,
|
| 1607 |
-
"grad_norm": 1.2158683240460215,
|
| 1608 |
-
"learning_rate": 5.2157558105109226e-08,
|
| 1609 |
-
"loss": 1.7462,
|
| 1610 |
-
"step": 1110
|
| 1611 |
-
},
|
| 1612 |
-
{
|
| 1613 |
-
"epoch": 2.247983870967742,
|
| 1614 |
-
"grad_norm": 1.174384875714619,
|
| 1615 |
-
"learning_rate": 5.202633214361811e-08,
|
| 1616 |
-
"loss": 1.695,
|
| 1617 |
-
"step": 1115
|
| 1618 |
-
},
|
| 1619 |
-
{
|
| 1620 |
-
"epoch": 2.258064516129032,
|
| 1621 |
-
"grad_norm": 1.2247069091252694,
|
| 1622 |
-
"learning_rate": 5.190240253347578e-08,
|
| 1623 |
-
"loss": 1.7591,
|
| 1624 |
-
"step": 1120
|
| 1625 |
-
},
|
| 1626 |
-
{
|
| 1627 |
-
"epoch": 2.2681451612903225,
|
| 1628 |
-
"grad_norm": 1.3095787384406332,
|
| 1629 |
-
"learning_rate": 5.178540210291998e-08,
|
| 1630 |
-
"loss": 1.7217,
|
| 1631 |
-
"step": 1125
|
| 1632 |
-
},
|
| 1633 |
-
{
|
| 1634 |
-
"epoch": 2.278225806451613,
|
| 1635 |
-
"grad_norm": 1.2493489539759923,
|
| 1636 |
-
"learning_rate": 5.167498019711146e-08,
|
| 1637 |
-
"loss": 1.7889,
|
| 1638 |
-
"step": 1130
|
| 1639 |
-
},
|
| 1640 |
-
{
|
| 1641 |
-
"epoch": 2.288306451612903,
|
| 1642 |
-
"grad_norm": 1.2374059562548885,
|
| 1643 |
-
"learning_rate": 5.157080202431146e-08,
|
| 1644 |
-
"loss": 1.6853,
|
| 1645 |
-
"step": 1135
|
| 1646 |
-
},
|
| 1647 |
-
{
|
| 1648 |
-
"epoch": 2.2983870967741935,
|
| 1649 |
-
"grad_norm": 1.2417228451011928,
|
| 1650 |
-
"learning_rate": 5.1472548024370265e-08,
|
| 1651 |
-
"loss": 1.7252,
|
| 1652 |
-
"step": 1140
|
| 1653 |
-
},
|
| 1654 |
-
{
|
| 1655 |
-
"epoch": 2.308467741935484,
|
| 1656 |
-
"grad_norm": 1.2382899521975939,
|
| 1657 |
-
"learning_rate": 5.137991325888869e-08,
|
| 1658 |
-
"loss": 1.6828,
|
| 1659 |
-
"step": 1145
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 2.318548387096774,
|
| 1663 |
-
"grad_norm": 1.2406554761874888,
|
| 1664 |
-
"learning_rate": 5.129260682242921e-08,
|
| 1665 |
-
"loss": 1.7995,
|
| 1666 |
-
"step": 1150
|
| 1667 |
-
},
|
| 1668 |
-
{
|
| 1669 |
-
"epoch": 2.3286290322580645,
|
| 1670 |
-
"grad_norm": 1.2857931147752715,
|
| 1671 |
-
"learning_rate": 5.121035127416763e-08,
|
| 1672 |
-
"loss": 1.6909,
|
| 1673 |
-
"step": 1155
|
| 1674 |
-
},
|
| 1675 |
-
{
|
| 1676 |
-
"epoch": 2.338709677419355,
|
| 1677 |
-
"grad_norm": 1.2227137154473942,
|
| 1678 |
-
"learning_rate": 5.1132882089390505e-08,
|
| 1679 |
-
"loss": 1.8297,
|
| 1680 |
-
"step": 1160
|
| 1681 |
-
},
|
| 1682 |
-
{
|
| 1683 |
-
"epoch": 2.348790322580645,
|
| 1684 |
-
"grad_norm": 1.2092856119101558,
|
| 1685 |
-
"learning_rate": 5.105994713025749e-08,
|
| 1686 |
-
"loss": 1.7084,
|
| 1687 |
-
"step": 1165
|
| 1688 |
-
},
|
| 1689 |
-
{
|
| 1690 |
-
"epoch": 2.3588709677419355,
|
| 1691 |
-
"grad_norm": 1.2553731796947716,
|
| 1692 |
-
"learning_rate": 5.099130613526128e-08,
|
| 1693 |
-
"loss": 1.7698,
|
| 1694 |
-
"step": 1170
|
| 1695 |
-
},
|
| 1696 |
-
{
|
| 1697 |
-
"epoch": 2.368951612903226,
|
| 1698 |
-
"grad_norm": 1.2347295734439494,
|
| 1699 |
-
"learning_rate": 5.092673022683157e-08,
|
| 1700 |
-
"loss": 1.7176,
|
| 1701 |
-
"step": 1175
|
| 1702 |
-
},
|
| 1703 |
-
{
|
| 1704 |
-
"epoch": 2.379032258064516,
|
| 1705 |
-
"grad_norm": 1.2236546996683366,
|
| 1706 |
-
"learning_rate": 5.086600143654227e-08,
|
| 1707 |
-
"loss": 1.7972,
|
| 1708 |
-
"step": 1180
|
| 1709 |
-
},
|
| 1710 |
-
{
|
| 1711 |
-
"epoch": 2.3891129032258065,
|
| 1712 |
-
"grad_norm": 1.1892980030528701,
|
| 1713 |
-
"learning_rate": 5.080891224739457e-08,
|
| 1714 |
-
"loss": 1.7075,
|
| 1715 |
-
"step": 1185
|
| 1716 |
-
},
|
| 1717 |
-
{
|
| 1718 |
-
"epoch": 2.399193548387097,
|
| 1719 |
-
"grad_norm": 1.2693570071261813,
|
| 1720 |
-
"learning_rate": 5.075526515266076e-08,
|
| 1721 |
-
"loss": 1.762,
|
| 1722 |
-
"step": 1190
|
| 1723 |
-
},
|
| 1724 |
-
{
|
| 1725 |
-
"epoch": 2.409274193548387,
|
| 1726 |
-
"grad_norm": 1.2598507176307956,
|
| 1727 |
-
"learning_rate": 5.07048722307867e-08,
|
| 1728 |
-
"loss": 1.7596,
|
| 1729 |
-
"step": 1195
|
| 1730 |
-
},
|
| 1731 |
-
{
|
| 1732 |
-
"epoch": 2.4193548387096775,
|
| 1733 |
-
"grad_norm": 1.2086250967210856,
|
| 1734 |
-
"learning_rate": 5.065755473586252e-08,
|
| 1735 |
-
"loss": 1.7212,
|
| 1736 |
-
"step": 1200
|
| 1737 |
-
},
|
| 1738 |
-
{
|
| 1739 |
-
"epoch": 2.4193548387096775,
|
| 1740 |
-
"eval_loss": 1.9444948434829712,
|
| 1741 |
-
"eval_runtime": 89.768,
|
| 1742 |
-
"eval_samples_per_second": 19.617,
|
| 1743 |
-
"eval_steps_per_second": 0.312,
|
| 1744 |
-
"step": 1200
|
| 1745 |
-
},
|
| 1746 |
-
{
|
| 1747 |
-
"epoch": 2.429435483870968,
|
| 1748 |
-
"grad_norm": 1.2270017834004545,
|
| 1749 |
-
"learning_rate": 5.061314270318372e-08,
|
| 1750 |
-
"loss": 1.7673,
|
| 1751 |
-
"step": 1205
|
| 1752 |
-
},
|
| 1753 |
-
{
|
| 1754 |
-
"epoch": 2.439516129032258,
|
| 1755 |
-
"grad_norm": 1.2758334557746702,
|
| 1756 |
-
"learning_rate": 5.057147456943636e-08,
|
| 1757 |
-
"loss": 1.7733,
|
| 1758 |
-
"step": 1210
|
| 1759 |
-
},
|
| 1760 |
-
{
|
| 1761 |
-
"epoch": 2.4495967741935485,
|
| 1762 |
-
"grad_norm": 1.2671506994834478,
|
| 1763 |
-
"learning_rate": 5.053239680705157e-08,
|
| 1764 |
-
"loss": 1.685,
|
| 1765 |
-
"step": 1215
|
| 1766 |
-
},
|
| 1767 |
-
{
|
| 1768 |
-
"epoch": 2.4596774193548385,
|
| 1769 |
-
"grad_norm": 1.3082114423304811,
|
| 1770 |
-
"learning_rate": 5.049576357228625e-08,
|
| 1771 |
-
"loss": 1.7673,
|
| 1772 |
-
"step": 1220
|
| 1773 |
-
},
|
| 1774 |
-
{
|
| 1775 |
-
"epoch": 2.469758064516129,
|
| 1776 |
-
"grad_norm": 1.2410498924874347,
|
| 1777 |
-
"learning_rate": 5.0461436366597737e-08,
|
| 1778 |
-
"loss": 1.6777,
|
| 1779 |
-
"step": 1225
|
| 1780 |
-
},
|
| 1781 |
-
{
|
| 1782 |
-
"epoch": 2.4798387096774195,
|
| 1783 |
-
"grad_norm": 1.2225089432500613,
|
| 1784 |
-
"learning_rate": 5.0429283710891395e-08,
|
| 1785 |
-
"loss": 1.7561,
|
| 1786 |
-
"step": 1230
|
| 1787 |
-
},
|
| 1788 |
-
{
|
| 1789 |
-
"epoch": 2.4899193548387095,
|
| 1790 |
-
"grad_norm": 1.3411537670367075,
|
| 1791 |
-
"learning_rate": 5.0399180832230535e-08,
|
| 1792 |
-
"loss": 1.7278,
|
| 1793 |
-
"step": 1235
|
| 1794 |
-
},
|
| 1795 |
-
{
|
| 1796 |
-
"epoch": 2.5,
|
| 1797 |
-
"grad_norm": 1.2608235504861707,
|
| 1798 |
-
"learning_rate": 5.0371009362608996e-08,
|
| 1799 |
-
"loss": 1.6774,
|
| 1800 |
-
"step": 1240
|
| 1801 |
-
},
|
| 1802 |
-
{
|
| 1803 |
-
"epoch": 2.5100806451612905,
|
| 1804 |
-
"grad_norm": 1.2549696696310126,
|
| 1805 |
-
"learning_rate": 5.0344657049396644e-08,
|
| 1806 |
-
"loss": 1.6711,
|
| 1807 |
-
"step": 1245
|
| 1808 |
-
},
|
| 1809 |
-
{
|
| 1810 |
-
"epoch": 2.5201612903225805,
|
| 1811 |
-
"grad_norm": 1.2520913106199976,
|
| 1812 |
-
"learning_rate": 5.032001747707849e-08,
|
| 1813 |
-
"loss": 1.7873,
|
| 1814 |
-
"step": 1250
|
| 1815 |
-
},
|
| 1816 |
-
{
|
| 1817 |
-
"epoch": 2.530241935483871,
|
| 1818 |
-
"grad_norm": 1.2195192727589492,
|
| 1819 |
-
"learning_rate": 5.029698979991788e-08,
|
| 1820 |
-
"loss": 1.7961,
|
| 1821 |
-
"step": 1255
|
| 1822 |
-
},
|
| 1823 |
-
{
|
| 1824 |
-
"epoch": 2.540322580645161,
|
| 1825 |
-
"grad_norm": 1.2296378280476938,
|
| 1826 |
-
"learning_rate": 5.0275478485184044e-08,
|
| 1827 |
-
"loss": 1.7105,
|
| 1828 |
-
"step": 1260
|
| 1829 |
-
},
|
| 1830 |
-
{
|
| 1831 |
-
"epoch": 2.5504032258064515,
|
| 1832 |
-
"grad_norm": 1.2672207613505957,
|
| 1833 |
-
"learning_rate": 5.025539306659372e-08,
|
| 1834 |
-
"loss": 1.7319,
|
| 1835 |
-
"step": 1265
|
| 1836 |
-
},
|
| 1837 |
-
{
|
| 1838 |
-
"epoch": 2.560483870967742,
|
| 1839 |
-
"grad_norm": 1.33348278916816,
|
| 1840 |
-
"learning_rate": 5.0236647907626076e-08,
|
| 1841 |
-
"loss": 1.7581,
|
| 1842 |
-
"step": 1270
|
| 1843 |
-
},
|
| 1844 |
-
{
|
| 1845 |
-
"epoch": 2.570564516129032,
|
| 1846 |
-
"grad_norm": 1.2147181184983937,
|
| 1847 |
-
"learning_rate": 5.021916197437918e-08,
|
| 1848 |
-
"loss": 1.7909,
|
| 1849 |
-
"step": 1275
|
| 1850 |
-
},
|
| 1851 |
-
{
|
| 1852 |
-
"epoch": 2.5806451612903225,
|
| 1853 |
-
"grad_norm": 1.2006424516658476,
|
| 1854 |
-
"learning_rate": 5.020285861764526e-08,
|
| 1855 |
-
"loss": 1.7094,
|
| 1856 |
-
"step": 1280
|
| 1857 |
-
},
|
| 1858 |
-
{
|
| 1859 |
-
"epoch": 2.590725806451613,
|
| 1860 |
-
"grad_norm": 1.304898262044129,
|
| 1861 |
-
"learning_rate": 5.018766536389094e-08,
|
| 1862 |
-
"loss": 1.7601,
|
| 1863 |
-
"step": 1285
|
| 1864 |
-
},
|
| 1865 |
-
{
|
| 1866 |
-
"epoch": 2.600806451612903,
|
| 1867 |
-
"grad_norm": 1.2541227980327716,
|
| 1868 |
-
"learning_rate": 5.017351371483695e-08,
|
| 1869 |
-
"loss": 1.6886,
|
| 1870 |
-
"step": 1290
|
| 1871 |
-
},
|
| 1872 |
-
{
|
| 1873 |
-
"epoch": 2.6108870967741935,
|
| 1874 |
-
"grad_norm": 1.2252553715485226,
|
| 1875 |
-
"learning_rate": 5.01603389553406e-08,
|
| 1876 |
-
"loss": 1.7265,
|
| 1877 |
-
"step": 1295
|
| 1878 |
-
},
|
| 1879 |
-
{
|
| 1880 |
-
"epoch": 2.620967741935484,
|
| 1881 |
-
"grad_norm": 1.249302955897003,
|
| 1882 |
-
"learning_rate": 5.0148079969292275e-08,
|
| 1883 |
-
"loss": 1.6571,
|
| 1884 |
-
"step": 1300
|
| 1885 |
-
},
|
| 1886 |
-
{
|
| 1887 |
-
"epoch": 2.631048387096774,
|
| 1888 |
-
"grad_norm": 1.1957123027809284,
|
| 1889 |
-
"learning_rate": 5.0136679063245423e-08,
|
| 1890 |
-
"loss": 1.7532,
|
| 1891 |
-
"step": 1305
|
| 1892 |
-
},
|
| 1893 |
-
{
|
| 1894 |
-
"epoch": 2.6411290322580645,
|
| 1895 |
-
"grad_norm": 1.201048445138404,
|
| 1896 |
-
"learning_rate": 5.012608179750745e-08,
|
| 1897 |
-
"loss": 1.7326,
|
| 1898 |
-
"step": 1310
|
| 1899 |
-
},
|
| 1900 |
-
{
|
| 1901 |
-
"epoch": 2.651209677419355,
|
| 1902 |
-
"grad_norm": 1.2081583082273468,
|
| 1903 |
-
"learning_rate": 5.011623682442653e-08,
|
| 1904 |
-
"loss": 1.6896,
|
| 1905 |
-
"step": 1315
|
| 1906 |
-
},
|
| 1907 |
-
{
|
| 1908 |
-
"epoch": 2.661290322580645,
|
| 1909 |
-
"grad_norm": 1.2165377024812585,
|
| 1910 |
-
"learning_rate": 5.010709573361717e-08,
|
| 1911 |
-
"loss": 1.7138,
|
| 1912 |
-
"step": 1320
|
| 1913 |
-
},
|
| 1914 |
-
{
|
| 1915 |
-
"epoch": 2.6713709677419355,
|
| 1916 |
-
"grad_norm": 1.2512113015598476,
|
| 1917 |
-
"learning_rate": 5.0098612903874436e-08,
|
| 1918 |
-
"loss": 1.7417,
|
| 1919 |
-
"step": 1325
|
| 1920 |
-
},
|
| 1921 |
-
{
|
| 1922 |
-
"epoch": 2.681451612903226,
|
| 1923 |
-
"grad_norm": 1.3393843580639568,
|
| 1924 |
-
"learning_rate": 5.009074536153439e-08,
|
| 1925 |
-
"loss": 1.6794,
|
| 1926 |
-
"step": 1330
|
| 1927 |
-
},
|
| 1928 |
-
{
|
| 1929 |
-
"epoch": 2.691532258064516,
|
| 1930 |
-
"grad_norm": 1.244775300561878,
|
| 1931 |
-
"learning_rate": 5.008345264504506e-08,
|
| 1932 |
-
"loss": 1.6785,
|
| 1933 |
-
"step": 1335
|
| 1934 |
-
},
|
| 1935 |
-
{
|
| 1936 |
-
"epoch": 2.7016129032258065,
|
| 1937 |
-
"grad_norm": 1.2309880190717277,
|
| 1938 |
-
"learning_rate": 5.007669667551931e-08,
|
| 1939 |
-
"loss": 1.7386,
|
| 1940 |
-
"step": 1340
|
| 1941 |
-
},
|
| 1942 |
-
{
|
| 1943 |
-
"epoch": 2.711693548387097,
|
| 1944 |
-
"grad_norm": 1.2646430684773242,
|
| 1945 |
-
"learning_rate": 5.007044163304792e-08,
|
| 1946 |
-
"loss": 1.7271,
|
| 1947 |
-
"step": 1345
|
| 1948 |
-
},
|
| 1949 |
-
{
|
| 1950 |
-
"epoch": 2.721774193548387,
|
| 1951 |
-
"grad_norm": 1.3573014483042332,
|
| 1952 |
-
"learning_rate": 5.0064653838557546e-08,
|
| 1953 |
-
"loss": 1.8166,
|
| 1954 |
-
"step": 1350
|
| 1955 |
-
},
|
| 1956 |
-
{
|
| 1957 |
-
"epoch": 2.7318548387096775,
|
| 1958 |
-
"grad_norm": 1.260061171998465,
|
| 1959 |
-
"learning_rate": 5.0059301641004814e-08,
|
| 1960 |
-
"loss": 1.7825,
|
| 1961 |
-
"step": 1355
|
| 1962 |
-
},
|
| 1963 |
-
{
|
| 1964 |
-
"epoch": 2.741935483870968,
|
| 1965 |
-
"grad_norm": 1.2930142933374957,
|
| 1966 |
-
"learning_rate": 5.0054355309704326e-08,
|
| 1967 |
-
"loss": 1.7362,
|
| 1968 |
-
"step": 1360
|
| 1969 |
-
},
|
| 1970 |
-
{
|
| 1971 |
-
"epoch": 2.752016129032258,
|
| 1972 |
-
"grad_norm": 1.1917500201859248,
|
| 1973 |
-
"learning_rate": 5.004978693159421e-08,
|
| 1974 |
-
"loss": 1.7862,
|
| 1975 |
-
"step": 1365
|
| 1976 |
-
},
|
| 1977 |
-
{
|
| 1978 |
-
"epoch": 2.7620967741935485,
|
| 1979 |
-
"grad_norm": 1.309363529031979,
|
| 1980 |
-
"learning_rate": 5.004557031324913e-08,
|
| 1981 |
-
"loss": 1.766,
|
| 1982 |
-
"step": 1370
|
| 1983 |
-
},
|
| 1984 |
-
{
|
| 1985 |
-
"epoch": 2.772177419354839,
|
| 1986 |
-
"grad_norm": 1.260854474561999,
|
| 1987 |
-
"learning_rate": 5.00416808874565e-08,
|
| 1988 |
-
"loss": 1.7798,
|
| 1989 |
-
"step": 1375
|
| 1990 |
-
},
|
| 1991 |
-
{
|
| 1992 |
-
"epoch": 2.782258064516129,
|
| 1993 |
-
"grad_norm": 1.2133948555271834,
|
| 1994 |
-
"learning_rate": 5.003809562417745e-08,
|
| 1995 |
-
"loss": 1.7702,
|
| 1996 |
-
"step": 1380
|
| 1997 |
-
},
|
| 1998 |
-
{
|
| 1999 |
-
"epoch": 2.7923387096774195,
|
| 2000 |
-
"grad_norm": 1.20285403043514,
|
| 2001 |
-
"learning_rate": 5.00347929457196e-08,
|
| 2002 |
-
"loss": 1.7629,
|
| 2003 |
-
"step": 1385
|
| 2004 |
-
},
|
| 2005 |
-
{
|
| 2006 |
-
"epoch": 2.8024193548387095,
|
| 2007 |
-
"grad_norm": 1.2055338414881442,
|
| 2008 |
-
"learning_rate": 5.003175264595438e-08,
|
| 2009 |
-
"loss": 1.6825,
|
| 2010 |
-
"step": 1390
|
| 2011 |
-
},
|
| 2012 |
-
{
|
| 2013 |
-
"epoch": 2.8125,
|
| 2014 |
-
"grad_norm": 1.2692380612645446,
|
| 2015 |
-
"learning_rate": 5.002895581341684e-08,
|
| 2016 |
-
"loss": 1.7922,
|
| 2017 |
-
"step": 1395
|
| 2018 |
-
},
|
| 2019 |
-
{
|
| 2020 |
-
"epoch": 2.8225806451612905,
|
| 2021 |
-
"grad_norm": 1.197125036474829,
|
| 2022 |
-
"learning_rate": 5.002638475813119e-08,
|
| 2023 |
-
"loss": 1.7346,
|
| 2024 |
-
"step": 1400
|
| 2025 |
-
},
|
| 2026 |
-
{
|
| 2027 |
-
"epoch": 2.8225806451612905,
|
| 2028 |
-
"eval_loss": 1.9443275928497314,
|
| 2029 |
-
"eval_runtime": 89.8463,
|
| 2030 |
-
"eval_samples_per_second": 19.6,
|
| 2031 |
-
"eval_steps_per_second": 0.312,
|
| 2032 |
-
"step": 1400
|
| 2033 |
-
},
|
| 2034 |
-
{
|
| 2035 |
-
"epoch": 2.8326612903225805,
|
| 2036 |
-
"grad_norm": 1.3129336482449783,
|
| 2037 |
-
"learning_rate": 5.0024022942010466e-08,
|
| 2038 |
-
"loss": 1.7541,
|
| 2039 |
-
"step": 1405
|
| 2040 |
-
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 2.842741935483871,
|
| 2043 |
-
"grad_norm": 1.1872922962263166,
|
| 2044 |
-
"learning_rate": 5.002185491268355e-08,
|
| 2045 |
-
"loss": 1.6476,
|
| 2046 |
-
"step": 1410
|
| 2047 |
-
},
|
| 2048 |
-
{
|
| 2049 |
-
"epoch": 2.852822580645161,
|
| 2050 |
-
"grad_norm": 1.3099170695792106,
|
| 2051 |
-
"learning_rate": 5.0019866240607806e-08,
|
| 2052 |
-
"loss": 1.6647,
|
| 2053 |
-
"step": 1415
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 2.8629032258064515,
|
| 2057 |
-
"grad_norm": 1.3296060602873512,
|
| 2058 |
-
"learning_rate": 5.001804345933021e-08,
|
| 2059 |
-
"loss": 1.7079,
|
| 2060 |
-
"step": 1420
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 2.872983870967742,
|
| 2064 |
-
"grad_norm": 1.2481767829563293,
|
| 2065 |
-
"learning_rate": 5.001637400876446e-08,
|
| 2066 |
-
"loss": 1.6979,
|
| 2067 |
-
"step": 1425
|
| 2068 |
-
},
|
| 2069 |
-
{
|
| 2070 |
-
"epoch": 2.883064516129032,
|
| 2071 |
-
"grad_norm": 1.3774040469968931,
|
| 2072 |
-
"learning_rate": 5.001484618135618e-08,
|
| 2073 |
-
"loss": 1.716,
|
| 2074 |
-
"step": 1430
|
| 2075 |
-
},
|
| 2076 |
-
{
|
| 2077 |
-
"epoch": 2.8931451612903225,
|
| 2078 |
-
"grad_norm": 1.3300791138778125,
|
| 2079 |
-
"learning_rate": 5.0013449071012536e-08,
|
| 2080 |
-
"loss": 1.8281,
|
| 2081 |
-
"step": 1435
|
| 2082 |
-
},
|
| 2083 |
-
{
|
| 2084 |
-
"epoch": 2.903225806451613,
|
| 2085 |
-
"grad_norm": 1.2648870887474146,
|
| 2086 |
-
"learning_rate": 5.0012172524676996e-08,
|
| 2087 |
-
"loss": 1.7511,
|
| 2088 |
-
"step": 1440
|
| 2089 |
-
},
|
| 2090 |
-
{
|
| 2091 |
-
"epoch": 2.913306451612903,
|
| 2092 |
-
"grad_norm": 1.2324711482162307,
|
| 2093 |
-
"learning_rate": 5.001100709643403e-08,
|
| 2094 |
-
"loss": 1.7665,
|
| 2095 |
-
"step": 1445
|
| 2096 |
-
},
|
| 2097 |
-
{
|
| 2098 |
-
"epoch": 2.9233870967741935,
|
| 2099 |
-
"grad_norm": 1.1884469998607576,
|
| 2100 |
-
"learning_rate": 5.0009944004032616e-08,
|
| 2101 |
-
"loss": 1.7515,
|
| 2102 |
-
"step": 1450
|
| 2103 |
-
},
|
| 2104 |
-
{
|
| 2105 |
-
"epoch": 2.933467741935484,
|
| 2106 |
-
"grad_norm": 1.2242938574470985,
|
| 2107 |
-
"learning_rate": 5.000897508772145e-08,
|
| 2108 |
-
"loss": 1.779,
|
| 2109 |
-
"step": 1455
|
| 2110 |
-
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 2.943548387096774,
|
| 2113 |
-
"grad_norm": 1.1767587628025173,
|
| 2114 |
-
"learning_rate": 5.000809277129238e-08,
|
| 2115 |
-
"loss": 1.7013,
|
| 2116 |
-
"step": 1460
|
| 2117 |
-
},
|
| 2118 |
-
{
|
| 2119 |
-
"epoch": 2.9536290322580645,
|
| 2120 |
-
"grad_norm": 1.1876105747863372,
|
| 2121 |
-
"learning_rate": 5.000729002523258e-08,
|
| 2122 |
-
"loss": 1.7691,
|
| 2123 |
-
"step": 1465
|
| 2124 |
-
},
|
| 2125 |
-
{
|
| 2126 |
-
"epoch": 2.963709677419355,
|
| 2127 |
-
"grad_norm": 1.2804911094505236,
|
| 2128 |
-
"learning_rate": 5.0006560331889364e-08,
|
| 2129 |
-
"loss": 1.7952,
|
| 2130 |
-
"step": 1470
|
| 2131 |
-
},
|
| 2132 |
-
{
|
| 2133 |
-
"epoch": 2.973790322580645,
|
| 2134 |
-
"grad_norm": 1.2634767897263572,
|
| 2135 |
-
"learning_rate": 5.0005897652555256e-08,
|
| 2136 |
-
"loss": 1.8265,
|
| 2137 |
-
"step": 1475
|
| 2138 |
-
},
|
| 2139 |
-
{
|
| 2140 |
-
"epoch": 2.9838709677419355,
|
| 2141 |
-
"grad_norm": 1.376696085159001,
|
| 2142 |
-
"learning_rate": 5.000529639638423e-08,
|
| 2143 |
-
"loss": 1.7974,
|
| 2144 |
-
"step": 1480
|
| 2145 |
-
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 2.993951612903226,
|
| 2148 |
-
"grad_norm": 1.1674051041442062,
|
| 2149 |
-
"learning_rate": 5.000475139105343e-08,
|
| 2150 |
-
"loss": 1.6758,
|
| 2151 |
-
"step": 1485
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 3.004032258064516,
|
| 2155 |
-
"grad_norm": 1.2278436300934317,
|
| 2156 |
-
"learning_rate": 5.000425785508798e-08,
|
| 2157 |
-
"loss": 1.7076,
|
| 2158 |
-
"step": 1490
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 3.0141129032258065,
|
| 2162 |
-
"grad_norm": 1.208822859914408,
|
| 2163 |
-
"learning_rate": 5.0003811371769436e-08,
|
| 2164 |
-
"loss": 1.6804,
|
| 2165 |
-
"step": 1495
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 3.024193548387097,
|
| 2169 |
-
"grad_norm": 1.2607881481569827,
|
| 2170 |
-
"learning_rate": 5.000340786455173e-08,
|
| 2171 |
-
"loss": 1.7044,
|
| 2172 |
-
"step": 1500
|
| 2173 |
-
},
|
| 2174 |
-
{
|
| 2175 |
-
"epoch": 3.034274193548387,
|
| 2176 |
-
"grad_norm": 1.2481468783314629,
|
| 2177 |
-
"learning_rate": 5.00030435739113e-08,
|
| 2178 |
-
"loss": 1.7133,
|
| 2179 |
-
"step": 1505
|
| 2180 |
-
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 3.0443548387096775,
|
| 2183 |
-
"grad_norm": 1.2625933161048073,
|
| 2184 |
-
"learning_rate": 5.0002715035560964e-08,
|
| 2185 |
-
"loss": 1.7043,
|
| 2186 |
-
"step": 1510
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 3.0544354838709675,
|
| 2190 |
-
"grad_norm": 1.2211349887353309,
|
| 2191 |
-
"learning_rate": 5.000241905995991e-08,
|
| 2192 |
-
"loss": 1.7931,
|
| 2193 |
-
"step": 1515
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 3.064516129032258,
|
| 2197 |
-
"grad_norm": 1.2628888420799884,
|
| 2198 |
-
"learning_rate": 5.000215271305494e-08,
|
| 2199 |
-
"loss": 1.7648,
|
| 2200 |
-
"step": 1520
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 3.0745967741935485,
|
| 2204 |
-
"grad_norm": 1.2349226985107031,
|
| 2205 |
-
"learning_rate": 5.000191329819063e-08,
|
| 2206 |
-
"loss": 1.72,
|
| 2207 |
-
"step": 1525
|
| 2208 |
-
},
|
| 2209 |
-
{
|
| 2210 |
-
"epoch": 3.0846774193548385,
|
| 2211 |
-
"grad_norm": 1.2993139941453224,
|
| 2212 |
-
"learning_rate": 5.0001698339128735e-08,
|
| 2213 |
-
"loss": 1.7493,
|
| 2214 |
-
"step": 1530
|
| 2215 |
-
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 3.094758064516129,
|
| 2218 |
-
"grad_norm": 1.3021805987677582,
|
| 2219 |
-
"learning_rate": 5.0001505564119494e-08,
|
| 2220 |
-
"loss": 1.7801,
|
| 2221 |
-
"step": 1535
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 3.1048387096774195,
|
| 2225 |
-
"grad_norm": 1.2177089098452982,
|
| 2226 |
-
"learning_rate": 5.000133289097e-08,
|
| 2227 |
-
"loss": 1.7488,
|
| 2228 |
-
"step": 1540
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 3.1149193548387095,
|
| 2232 |
-
"grad_norm": 1.2475923991723818,
|
| 2233 |
-
"learning_rate": 5.0001178413057135e-08,
|
| 2234 |
-
"loss": 1.8035,
|
| 2235 |
-
"step": 1545
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 3.125,
|
| 2239 |
-
"grad_norm": 1.2750062841943113,
|
| 2240 |
-
"learning_rate": 5.0001040386234535e-08,
|
| 2241 |
-
"loss": 1.6927,
|
| 2242 |
-
"step": 1550
|
| 2243 |
-
},
|
| 2244 |
-
{
|
| 2245 |
-
"epoch": 3.1350806451612905,
|
| 2246 |
-
"grad_norm": 1.2521384201728667,
|
| 2247 |
-
"learning_rate": 5.000091721658573e-08,
|
| 2248 |
-
"loss": 1.7867,
|
| 2249 |
-
"step": 1555
|
| 2250 |
-
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 3.1451612903225805,
|
| 2253 |
-
"grad_norm": 1.2788389170185523,
|
| 2254 |
-
"learning_rate": 5.0000807448977166e-08,
|
| 2255 |
-
"loss": 1.7463,
|
| 2256 |
-
"step": 1560
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 3.155241935483871,
|
| 2260 |
-
"grad_norm": 1.1932742831754912,
|
| 2261 |
-
"learning_rate": 5.000070975636714e-08,
|
| 2262 |
-
"loss": 1.6646,
|
| 2263 |
-
"step": 1565
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 3.1653225806451615,
|
| 2267 |
-
"grad_norm": 1.2026279996869336,
|
| 2268 |
-
"learning_rate": 5.0000622929828655e-08,
|
| 2269 |
-
"loss": 1.7426,
|
| 2270 |
-
"step": 1570
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 3.1754032258064515,
|
| 2274 |
-
"grad_norm": 1.2658970944840857,
|
| 2275 |
-
"learning_rate": 5.000054586924596e-08,
|
| 2276 |
-
"loss": 1.6139,
|
| 2277 |
-
"step": 1575
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 3.185483870967742,
|
| 2281 |
-
"grad_norm": 1.2582895003633043,
|
| 2282 |
-
"learning_rate": 5.0000477574646376e-08,
|
| 2283 |
-
"loss": 1.6655,
|
| 2284 |
-
"step": 1580
|
| 2285 |
-
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 3.195564516129032,
|
| 2288 |
-
"grad_norm": 1.286929029486268,
|
| 2289 |
-
"learning_rate": 5.000041713813098e-08,
|
| 2290 |
-
"loss": 1.8082,
|
| 2291 |
-
"step": 1585
|
| 2292 |
-
},
|
| 2293 |
-
{
|
| 2294 |
-
"epoch": 3.2056451612903225,
|
| 2295 |
-
"grad_norm": 1.5242979315943919,
|
| 2296 |
-
"learning_rate": 5.0000363736369096e-08,
|
| 2297 |
-
"loss": 1.7129,
|
| 2298 |
-
"step": 1590
|
| 2299 |
-
},
|
| 2300 |
-
{
|
| 2301 |
-
"epoch": 3.215725806451613,
|
| 2302 |
-
"grad_norm": 1.1974095695410052,
|
| 2303 |
-
"learning_rate": 5.000031662362353e-08,
|
| 2304 |
-
"loss": 1.7858,
|
| 2305 |
-
"step": 1595
|
| 2306 |
-
},
|
| 2307 |
-
{
|
| 2308 |
-
"epoch": 3.225806451612903,
|
| 2309 |
-
"grad_norm": 1.2334951186835896,
|
| 2310 |
-
"learning_rate": 5.000027512527476e-08,
|
| 2311 |
-
"loss": 1.7028,
|
| 2312 |
-
"step": 1600
|
| 2313 |
-
},
|
| 2314 |
-
{
|
| 2315 |
-
"epoch": 3.225806451612903,
|
| 2316 |
-
"eval_loss": 1.9448145627975464,
|
| 2317 |
-
"eval_runtime": 89.9094,
|
| 2318 |
-
"eval_samples_per_second": 19.586,
|
| 2319 |
-
"eval_steps_per_second": 0.311,
|
| 2320 |
-
"step": 1600
|
| 2321 |
-
},
|
| 2322 |
-
{
|
| 2323 |
-
"epoch": 3.2358870967741935,
|
| 2324 |
-
"grad_norm": 1.2639640178197116,
|
| 2325 |
-
"learning_rate": 5.0000238631814066e-08,
|
| 2326 |
-
"loss": 1.8326,
|
| 2327 |
-
"step": 1605
|
| 2328 |
-
},
|
| 2329 |
-
{
|
| 2330 |
-
"epoch": 3.245967741935484,
|
| 2331 |
-
"grad_norm": 1.2369747127772903,
|
| 2332 |
-
"learning_rate": 5.000020659327692e-08,
|
| 2333 |
-
"loss": 1.6435,
|
| 2334 |
-
"step": 1610
|
| 2335 |
-
},
|
| 2336 |
-
{
|
| 2337 |
-
"epoch": 3.256048387096774,
|
| 2338 |
-
"grad_norm": 1.2345433192530466,
|
| 2339 |
-
"learning_rate": 5.000017851408939e-08,
|
| 2340 |
-
"loss": 1.7267,
|
| 2341 |
-
"step": 1615
|
| 2342 |
-
},
|
| 2343 |
-
{
|
| 2344 |
-
"epoch": 3.2661290322580645,
|
| 2345 |
-
"grad_norm": 1.2504468827628226,
|
| 2346 |
-
"learning_rate": 5.0000153948301755e-08,
|
| 2347 |
-
"loss": 1.749,
|
| 2348 |
-
"step": 1620
|
| 2349 |
-
},
|
| 2350 |
-
{
|
| 2351 |
-
"epoch": 3.276209677419355,
|
| 2352 |
-
"grad_norm": 1.2484879937766102,
|
| 2353 |
-
"learning_rate": 5.000013249518468e-08,
|
| 2354 |
-
"loss": 1.7147,
|
| 2355 |
-
"step": 1625
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 3.286290322580645,
|
| 2359 |
-
"grad_norm": 1.2765939466692027,
|
| 2360 |
-
"learning_rate": 5.000011379516478e-08,
|
| 2361 |
-
"loss": 1.708,
|
| 2362 |
-
"step": 1630
|
| 2363 |
-
},
|
| 2364 |
-
{
|
| 2365 |
-
"epoch": 3.2963709677419355,
|
| 2366 |
-
"grad_norm": 1.2548830946050018,
|
| 2367 |
-
"learning_rate": 5.000009752607744e-08,
|
| 2368 |
-
"loss": 1.6576,
|
| 2369 |
-
"step": 1635
|
| 2370 |
-
},
|
| 2371 |
-
{
|
| 2372 |
-
"epoch": 3.306451612903226,
|
| 2373 |
-
"grad_norm": 1.3380090814755392,
|
| 2374 |
-
"learning_rate": 5.00000833997159e-08,
|
| 2375 |
-
"loss": 1.7899,
|
| 2376 |
-
"step": 1640
|
| 2377 |
-
},
|
| 2378 |
-
{
|
| 2379 |
-
"epoch": 3.316532258064516,
|
| 2380 |
-
"grad_norm": 1.2589567647915303,
|
| 2381 |
-
"learning_rate": 5.000007115865701e-08,
|
| 2382 |
-
"loss": 1.7283,
|
| 2383 |
-
"step": 1645
|
| 2384 |
-
},
|
| 2385 |
-
{
|
| 2386 |
-
"epoch": 3.3266129032258065,
|
| 2387 |
-
"grad_norm": 1.330891571705861,
|
| 2388 |
-
"learning_rate": 5.00000605733447e-08,
|
| 2389 |
-
"loss": 1.7638,
|
| 2390 |
-
"step": 1650
|
| 2391 |
-
},
|
| 2392 |
-
{
|
| 2393 |
-
"epoch": 3.336693548387097,
|
| 2394 |
-
"grad_norm": 1.2968430137554552,
|
| 2395 |
-
"learning_rate": 5.0000051439413665e-08,
|
| 2396 |
-
"loss": 1.7894,
|
| 2397 |
-
"step": 1655
|
| 2398 |
-
},
|
| 2399 |
-
{
|
| 2400 |
-
"epoch": 3.346774193548387,
|
| 2401 |
-
"grad_norm": 1.26655914538691,
|
| 2402 |
-
"learning_rate": 5.000004357523643e-08,
|
| 2403 |
-
"loss": 1.6794,
|
| 2404 |
-
"step": 1660
|
| 2405 |
-
},
|
| 2406 |
-
{
|
| 2407 |
-
"epoch": 3.3568548387096775,
|
| 2408 |
-
"grad_norm": 1.3255247557489993,
|
| 2409 |
-
"learning_rate": 5.000003681967807e-08,
|
| 2410 |
-
"loss": 1.7458,
|
| 2411 |
-
"step": 1665
|
| 2412 |
-
},
|
| 2413 |
-
{
|
| 2414 |
-
"epoch": 3.366935483870968,
|
| 2415 |
-
"grad_norm": 1.4616051241449233,
|
| 2416 |
-
"learning_rate": 5.000003103004378e-08,
|
| 2417 |
-
"loss": 1.7682,
|
| 2418 |
-
"step": 1670
|
| 2419 |
-
},
|
| 2420 |
-
{
|
| 2421 |
-
"epoch": 3.377016129032258,
|
| 2422 |
-
"grad_norm": 1.2909185531222274,
|
| 2423 |
-
"learning_rate": 5.000002608020516e-08,
|
| 2424 |
-
"loss": 1.7676,
|
| 2425 |
-
"step": 1675
|
| 2426 |
-
},
|
| 2427 |
-
{
|
| 2428 |
-
"epoch": 3.3870967741935485,
|
| 2429 |
-
"grad_norm": 1.5116117360689252,
|
| 2430 |
-
"learning_rate": 5.0000021858892204e-08,
|
| 2431 |
-
"loss": 1.8041,
|
| 2432 |
-
"step": 1680
|
| 2433 |
-
},
|
| 2434 |
-
{
|
| 2435 |
-
"epoch": 3.3971774193548385,
|
| 2436 |
-
"grad_norm": 1.2019942003042767,
|
| 2437 |
-
"learning_rate": 5.000001826813845e-08,
|
| 2438 |
-
"loss": 1.6975,
|
| 2439 |
-
"step": 1685
|
| 2440 |
-
},
|
| 2441 |
-
{
|
| 2442 |
-
"epoch": 3.407258064516129,
|
| 2443 |
-
"grad_norm": 1.2423004849326869,
|
| 2444 |
-
"learning_rate": 5.000001522186781e-08,
|
| 2445 |
-
"loss": 1.7856,
|
| 2446 |
-
"step": 1690
|
| 2447 |
-
},
|
| 2448 |
-
{
|
| 2449 |
-
"epoch": 3.4173387096774195,
|
| 2450 |
-
"grad_norm": 1.1833866067645467,
|
| 2451 |
-
"learning_rate": 5.0000012644612046e-08,
|
| 2452 |
-
"loss": 1.6801,
|
| 2453 |
-
"step": 1695
|
| 2454 |
-
},
|
| 2455 |
-
{
|
| 2456 |
-
"epoch": 3.4274193548387095,
|
| 2457 |
-
"grad_norm": 1.2527324283211838,
|
| 2458 |
-
"learning_rate": 5.000001047034869e-08,
|
| 2459 |
-
"loss": 1.7829,
|
| 2460 |
-
"step": 1700
|
| 2461 |
-
},
|
| 2462 |
-
{
|
| 2463 |
-
"epoch": 3.4375,
|
| 2464 |
-
"grad_norm": 1.2021893152164589,
|
| 2465 |
-
"learning_rate": 5.000000864144987e-08,
|
| 2466 |
-
"loss": 1.7365,
|
| 2467 |
-
"step": 1705
|
| 2468 |
-
},
|
| 2469 |
-
{
|
| 2470 |
-
"epoch": 3.4475806451612905,
|
| 2471 |
-
"grad_norm": 1.2533733399310467,
|
| 2472 |
-
"learning_rate": 5.0000007107733085e-08,
|
| 2473 |
-
"loss": 1.7684,
|
| 2474 |
-
"step": 1710
|
| 2475 |
-
},
|
| 2476 |
-
{
|
| 2477 |
-
"epoch": 3.4576612903225805,
|
| 2478 |
-
"grad_norm": 1.2808992378124462,
|
| 2479 |
-
"learning_rate": 5.000000582560542e-08,
|
| 2480 |
-
"loss": 1.7195,
|
| 2481 |
-
"step": 1715
|
| 2482 |
-
},
|
| 2483 |
-
{
|
| 2484 |
-
"epoch": 3.467741935483871,
|
| 2485 |
-
"grad_norm": 1.2244209413905693,
|
| 2486 |
-
"learning_rate": 5.00000047572936e-08,
|
| 2487 |
-
"loss": 1.7485,
|
| 2488 |
-
"step": 1720
|
| 2489 |
-
},
|
| 2490 |
-
{
|
| 2491 |
-
"epoch": 3.4778225806451615,
|
| 2492 |
-
"grad_norm": 1.2540825243078901,
|
| 2493 |
-
"learning_rate": 5.000000387015242e-08,
|
| 2494 |
-
"loss": 1.7303,
|
| 2495 |
-
"step": 1725
|
| 2496 |
-
},
|
| 2497 |
-
{
|
| 2498 |
-
"epoch": 3.4879032258064515,
|
| 2499 |
-
"grad_norm": 1.2847488818819852,
|
| 2500 |
-
"learning_rate": 5.0000003136044955e-08,
|
| 2501 |
-
"loss": 1.8182,
|
| 2502 |
-
"step": 1730
|
| 2503 |
-
},
|
| 2504 |
-
{
|
| 2505 |
-
"epoch": 3.497983870967742,
|
| 2506 |
-
"grad_norm": 1.247272423872035,
|
| 2507 |
-
"learning_rate": 5.000000253078795e-08,
|
| 2508 |
-
"loss": 1.7646,
|
| 2509 |
-
"step": 1735
|
| 2510 |
-
},
|
| 2511 |
-
{
|
| 2512 |
-
"epoch": 3.508064516129032,
|
| 2513 |
-
"grad_norm": 1.1616622600983573,
|
| 2514 |
-
"learning_rate": 5.000000203365694e-08,
|
| 2515 |
-
"loss": 1.61,
|
| 2516 |
-
"step": 1740
|
| 2517 |
-
},
|
| 2518 |
-
{
|
| 2519 |
-
"epoch": 3.5181451612903225,
|
| 2520 |
-
"grad_norm": 1.2602897978426342,
|
| 2521 |
-
"learning_rate": 5.000000162694535e-08,
|
| 2522 |
-
"loss": 1.7119,
|
| 2523 |
-
"step": 1745
|
| 2524 |
-
},
|
| 2525 |
-
{
|
| 2526 |
-
"epoch": 3.528225806451613,
|
| 2527 |
-
"grad_norm": 1.2351778503899846,
|
| 2528 |
-
"learning_rate": 5.000000129557269e-08,
|
| 2529 |
-
"loss": 1.7027,
|
| 2530 |
-
"step": 1750
|
| 2531 |
-
},
|
| 2532 |
-
{
|
| 2533 |
-
"epoch": 3.538306451612903,
|
| 2534 |
-
"grad_norm": 1.2344538732881385,
|
| 2535 |
-
"learning_rate": 5.000000102673722e-08,
|
| 2536 |
-
"loss": 1.7228,
|
| 2537 |
-
"step": 1755
|
| 2538 |
-
},
|
| 2539 |
-
{
|
| 2540 |
-
"epoch": 3.5483870967741935,
|
| 2541 |
-
"grad_norm": 1.2335083557737399,
|
| 2542 |
-
"learning_rate": 5.0000000809608727e-08,
|
| 2543 |
-
"loss": 1.7107,
|
| 2544 |
-
"step": 1760
|
| 2545 |
-
},
|
| 2546 |
-
{
|
| 2547 |
-
"epoch": 3.558467741935484,
|
| 2548 |
-
"grad_norm": 1.2491739462049536,
|
| 2549 |
-
"learning_rate": 5.000000063505754e-08,
|
| 2550 |
-
"loss": 1.6427,
|
| 2551 |
-
"step": 1765
|
| 2552 |
-
},
|
| 2553 |
-
{
|
| 2554 |
-
"epoch": 3.568548387096774,
|
| 2555 |
-
"grad_norm": 1.2486316865372293,
|
| 2556 |
-
"learning_rate": 5.000000049541612e-08,
|
| 2557 |
-
"loss": 1.7365,
|
| 2558 |
-
"step": 1770
|
| 2559 |
-
},
|
| 2560 |
-
{
|
| 2561 |
-
"epoch": 3.5786290322580645,
|
| 2562 |
-
"grad_norm": 1.3051302158136038,
|
| 2563 |
-
"learning_rate": 5.000000038426991e-08,
|
| 2564 |
-
"loss": 1.7311,
|
| 2565 |
-
"step": 1775
|
| 2566 |
-
},
|
| 2567 |
-
{
|
| 2568 |
-
"epoch": 3.588709677419355,
|
| 2569 |
-
"grad_norm": 1.23592969040822,
|
| 2570 |
-
"learning_rate": 5.000000029627433e-08,
|
| 2571 |
-
"loss": 1.7672,
|
| 2572 |
-
"step": 1780
|
| 2573 |
-
},
|
| 2574 |
-
{
|
| 2575 |
-
"epoch": 3.598790322580645,
|
| 2576 |
-
"grad_norm": 1.1961157739658335,
|
| 2577 |
-
"learning_rate": 5.000000022699522e-08,
|
| 2578 |
-
"loss": 1.6726,
|
| 2579 |
-
"step": 1785
|
| 2580 |
-
},
|
| 2581 |
-
{
|
| 2582 |
-
"epoch": 3.6088709677419355,
|
| 2583 |
-
"grad_norm": 1.2487972857881837,
|
| 2584 |
-
"learning_rate": 5.000000017277015e-08,
|
| 2585 |
-
"loss": 1.823,
|
| 2586 |
-
"step": 1790
|
| 2587 |
-
},
|
| 2588 |
-
{
|
| 2589 |
-
"epoch": 3.618951612903226,
|
| 2590 |
-
"grad_norm": 1.2187655485543294,
|
| 2591 |
-
"learning_rate": 5.0000000130588186e-08,
|
| 2592 |
-
"loss": 1.7657,
|
| 2593 |
-
"step": 1795
|
| 2594 |
-
},
|
| 2595 |
-
{
|
| 2596 |
-
"epoch": 3.629032258064516,
|
| 2597 |
-
"grad_norm": 1.196141679163838,
|
| 2598 |
-
"learning_rate": 5.000000009798615e-08,
|
| 2599 |
-
"loss": 1.7383,
|
| 2600 |
-
"step": 1800
|
| 2601 |
-
},
|
| 2602 |
-
{
|
| 2603 |
-
"epoch": 3.629032258064516,
|
| 2604 |
-
"eval_loss": 1.9450066089630127,
|
| 2605 |
-
"eval_runtime": 89.836,
|
| 2606 |
-
"eval_samples_per_second": 19.602,
|
| 2607 |
-
"eval_steps_per_second": 0.312,
|
| 2608 |
-
"step": 1800
|
| 2609 |
-
},
|
| 2610 |
-
{
|
| 2611 |
-
"epoch": 3.6391129032258065,
|
| 2612 |
-
"grad_norm": 1.2843403341661654,
|
| 2613 |
-
"learning_rate": 5.0000000072959346e-08,
|
| 2614 |
-
"loss": 1.6978,
|
| 2615 |
-
"step": 1805
|
| 2616 |
-
},
|
| 2617 |
-
{
|
| 2618 |
-
"epoch": 3.649193548387097,
|
| 2619 |
-
"grad_norm": 1.2729441010270168,
|
| 2620 |
-
"learning_rate": 5.000000005388504e-08,
|
| 2621 |
-
"loss": 1.7236,
|
| 2622 |
-
"step": 1810
|
| 2623 |
-
},
|
| 2624 |
-
{
|
| 2625 |
-
"epoch": 3.659274193548387,
|
| 2626 |
-
"grad_norm": 1.278805125989498,
|
| 2627 |
-
"learning_rate": 5.000000003945723e-08,
|
| 2628 |
-
"loss": 1.7535,
|
| 2629 |
-
"step": 1815
|
| 2630 |
-
},
|
| 2631 |
-
{
|
| 2632 |
-
"epoch": 3.6693548387096775,
|
| 2633 |
-
"grad_norm": 1.3166466197064246,
|
| 2634 |
-
"learning_rate": 5.0000000028631134e-08,
|
| 2635 |
-
"loss": 1.7694,
|
| 2636 |
-
"step": 1820
|
| 2637 |
-
},
|
| 2638 |
-
{
|
| 2639 |
-
"epoch": 3.679435483870968,
|
| 2640 |
-
"grad_norm": 1.2114793447075247,
|
| 2641 |
-
"learning_rate": 5.0000000020576294e-08,
|
| 2642 |
-
"loss": 1.7647,
|
| 2643 |
-
"step": 1825
|
| 2644 |
-
},
|
| 2645 |
-
{
|
| 2646 |
-
"epoch": 3.689516129032258,
|
| 2647 |
-
"grad_norm": 1.2395966902447142,
|
| 2648 |
-
"learning_rate": 5.000000001463704e-08,
|
| 2649 |
-
"loss": 1.7112,
|
| 2650 |
-
"step": 1830
|
| 2651 |
-
},
|
| 2652 |
-
{
|
| 2653 |
-
"epoch": 3.6995967741935485,
|
| 2654 |
-
"grad_norm": 1.2716522150792229,
|
| 2655 |
-
"learning_rate": 5.0000000010299415e-08,
|
| 2656 |
-
"loss": 1.7674,
|
| 2657 |
-
"step": 1835
|
| 2658 |
-
},
|
| 2659 |
-
{
|
| 2660 |
-
"epoch": 3.709677419354839,
|
| 2661 |
-
"grad_norm": 1.2205432657380169,
|
| 2662 |
-
"learning_rate": 5.000000000716364e-08,
|
| 2663 |
-
"loss": 1.6942,
|
| 2664 |
-
"step": 1840
|
| 2665 |
-
},
|
| 2666 |
-
{
|
| 2667 |
-
"epoch": 3.719758064516129,
|
| 2668 |
-
"grad_norm": 1.3306099794060742,
|
| 2669 |
-
"learning_rate": 5.000000000492121e-08,
|
| 2670 |
-
"loss": 1.6627,
|
| 2671 |
-
"step": 1845
|
| 2672 |
-
},
|
| 2673 |
-
{
|
| 2674 |
-
"epoch": 3.7298387096774195,
|
| 2675 |
-
"grad_norm": 1.2101658596755476,
|
| 2676 |
-
"learning_rate": 5.000000000333616e-08,
|
| 2677 |
-
"loss": 1.719,
|
| 2678 |
-
"step": 1850
|
| 2679 |
-
},
|
| 2680 |
-
{
|
| 2681 |
-
"epoch": 3.7399193548387095,
|
| 2682 |
-
"grad_norm": 1.2500275205533455,
|
| 2683 |
-
"learning_rate": 5.000000000222965e-08,
|
| 2684 |
-
"loss": 1.7493,
|
| 2685 |
-
"step": 1855
|
| 2686 |
-
},
|
| 2687 |
-
{
|
| 2688 |
-
"epoch": 3.75,
|
| 2689 |
-
"grad_norm": 1.2312257846231576,
|
| 2690 |
-
"learning_rate": 5.0000000001467476e-08,
|
| 2691 |
-
"loss": 1.7789,
|
| 2692 |
-
"step": 1860
|
| 2693 |
-
},
|
| 2694 |
-
{
|
| 2695 |
-
"epoch": 3.7600806451612905,
|
| 2696 |
-
"grad_norm": 1.2152299183587771,
|
| 2697 |
-
"learning_rate": 5.000000000095e-08,
|
| 2698 |
-
"loss": 1.7211,
|
| 2699 |
-
"step": 1865
|
| 2700 |
-
},
|
| 2701 |
-
{
|
| 2702 |
-
"epoch": 3.7701612903225805,
|
| 2703 |
-
"grad_norm": 1.1996585863358138,
|
| 2704 |
-
"learning_rate": 5.0000000000604084e-08,
|
| 2705 |
-
"loss": 1.7994,
|
| 2706 |
-
"step": 1870
|
| 2707 |
-
},
|
| 2708 |
-
{
|
| 2709 |
-
"epoch": 3.780241935483871,
|
| 2710 |
-
"grad_norm": 1.3230860176481571,
|
| 2711 |
-
"learning_rate": 5.0000000000376735e-08,
|
| 2712 |
-
"loss": 1.7701,
|
| 2713 |
-
"step": 1875
|
| 2714 |
-
},
|
| 2715 |
-
{
|
| 2716 |
-
"epoch": 3.790322580645161,
|
| 2717 |
-
"grad_norm": 1.2748379623445445,
|
| 2718 |
-
"learning_rate": 5.000000000023003e-08,
|
| 2719 |
-
"loss": 1.7793,
|
| 2720 |
-
"step": 1880
|
| 2721 |
-
},
|
| 2722 |
-
{
|
| 2723 |
-
"epoch": 3.8004032258064515,
|
| 2724 |
-
"grad_norm": 1.2509667612393764,
|
| 2725 |
-
"learning_rate": 5.000000000013723e-08,
|
| 2726 |
-
"loss": 1.7706,
|
| 2727 |
-
"step": 1885
|
| 2728 |
-
},
|
| 2729 |
-
{
|
| 2730 |
-
"epoch": 3.810483870967742,
|
| 2731 |
-
"grad_norm": 1.2694885460408405,
|
| 2732 |
-
"learning_rate": 5.000000000007981e-08,
|
| 2733 |
-
"loss": 1.7397,
|
| 2734 |
-
"step": 1890
|
| 2735 |
-
},
|
| 2736 |
-
{
|
| 2737 |
-
"epoch": 3.820564516129032,
|
| 2738 |
-
"grad_norm": 1.3091515145607215,
|
| 2739 |
-
"learning_rate": 5.000000000004513e-08,
|
| 2740 |
-
"loss": 1.7211,
|
| 2741 |
-
"step": 1895
|
| 2742 |
-
},
|
| 2743 |
-
{
|
| 2744 |
-
"epoch": 3.8306451612903225,
|
| 2745 |
-
"grad_norm": 1.247922300219581,
|
| 2746 |
-
"learning_rate": 5.0000000000024734e-08,
|
| 2747 |
-
"loss": 1.774,
|
| 2748 |
-
"step": 1900
|
| 2749 |
-
},
|
| 2750 |
-
{
|
| 2751 |
-
"epoch": 3.840725806451613,
|
| 2752 |
-
"grad_norm": 1.2523345160639314,
|
| 2753 |
-
"learning_rate": 5.0000000000013094e-08,
|
| 2754 |
-
"loss": 1.6928,
|
| 2755 |
-
"step": 1905
|
| 2756 |
-
},
|
| 2757 |
-
{
|
| 2758 |
-
"epoch": 3.850806451612903,
|
| 2759 |
-
"grad_norm": 1.1986290945949785,
|
| 2760 |
-
"learning_rate": 5.000000000000667e-08,
|
| 2761 |
-
"loss": 1.7161,
|
| 2762 |
-
"step": 1910
|
| 2763 |
-
},
|
| 2764 |
-
{
|
| 2765 |
-
"epoch": 3.8608870967741935,
|
| 2766 |
-
"grad_norm": 1.2245878014293898,
|
| 2767 |
-
"learning_rate": 5.000000000000325e-08,
|
| 2768 |
-
"loss": 1.686,
|
| 2769 |
-
"step": 1915
|
| 2770 |
-
},
|
| 2771 |
-
{
|
| 2772 |
-
"epoch": 3.870967741935484,
|
| 2773 |
-
"grad_norm": 1.2249127134236595,
|
| 2774 |
-
"learning_rate": 5.00000000000015e-08,
|
| 2775 |
-
"loss": 1.7206,
|
| 2776 |
-
"step": 1920
|
| 2777 |
-
},
|
| 2778 |
-
{
|
| 2779 |
-
"epoch": 3.881048387096774,
|
| 2780 |
-
"grad_norm": 1.3558345618343015,
|
| 2781 |
-
"learning_rate": 5.000000000000065e-08,
|
| 2782 |
-
"loss": 1.68,
|
| 2783 |
-
"step": 1925
|
| 2784 |
-
},
|
| 2785 |
-
{
|
| 2786 |
-
"epoch": 3.8911290322580645,
|
| 2787 |
-
"grad_norm": 1.3236906608919645,
|
| 2788 |
-
"learning_rate": 5.000000000000027e-08,
|
| 2789 |
-
"loss": 1.6661,
|
| 2790 |
-
"step": 1930
|
| 2791 |
-
},
|
| 2792 |
-
{
|
| 2793 |
-
"epoch": 3.901209677419355,
|
| 2794 |
-
"grad_norm": 1.183426329240457,
|
| 2795 |
-
"learning_rate": 5.00000000000001e-08,
|
| 2796 |
-
"loss": 1.727,
|
| 2797 |
-
"step": 1935
|
| 2798 |
-
},
|
| 2799 |
-
{
|
| 2800 |
-
"epoch": 3.911290322580645,
|
| 2801 |
-
"grad_norm": 1.2348171369390457,
|
| 2802 |
-
"learning_rate": 5.000000000000003e-08,
|
| 2803 |
-
"loss": 1.6334,
|
| 2804 |
-
"step": 1940
|
| 2805 |
-
},
|
| 2806 |
-
{
|
| 2807 |
-
"epoch": 3.9213709677419355,
|
| 2808 |
-
"grad_norm": 1.2911302775087936,
|
| 2809 |
-
"learning_rate": 5.000000000000001e-08,
|
| 2810 |
-
"loss": 1.8057,
|
| 2811 |
-
"step": 1945
|
| 2812 |
-
},
|
| 2813 |
-
{
|
| 2814 |
-
"epoch": 3.931451612903226,
|
| 2815 |
-
"grad_norm": 1.1863076460974697,
|
| 2816 |
-
"learning_rate": 5e-08,
|
| 2817 |
-
"loss": 1.6921,
|
| 2818 |
-
"step": 1950
|
| 2819 |
},
|
| 2820 |
{
|
| 2821 |
-
"epoch": 3.
|
| 2822 |
-
"grad_norm": 1.
|
| 2823 |
-
"learning_rate":
|
| 2824 |
-
"loss": 1.
|
| 2825 |
-
"step":
|
| 2826 |
},
|
| 2827 |
{
|
| 2828 |
-
"epoch": 3.
|
| 2829 |
-
"
|
| 2830 |
-
"
|
| 2831 |
-
"
|
| 2832 |
-
"
|
|
|
|
| 2833 |
},
|
| 2834 |
{
|
| 2835 |
-
"epoch": 3.
|
| 2836 |
-
"grad_norm": 1.
|
| 2837 |
-
"learning_rate":
|
| 2838 |
-
"loss": 1.
|
| 2839 |
-
"step":
|
| 2840 |
},
|
| 2841 |
{
|
| 2842 |
-
"epoch": 3.
|
| 2843 |
-
"grad_norm": 1.
|
| 2844 |
-
"learning_rate":
|
| 2845 |
-
"loss": 1.
|
| 2846 |
-
"step":
|
| 2847 |
},
|
| 2848 |
{
|
| 2849 |
-
"epoch": 3.
|
| 2850 |
-
"
|
| 2851 |
-
"
|
| 2852 |
-
"
|
| 2853 |
-
"
|
|
|
|
| 2854 |
},
|
| 2855 |
{
|
| 2856 |
-
"epoch": 3.
|
| 2857 |
-
"grad_norm": 1.
|
| 2858 |
-
"learning_rate":
|
| 2859 |
-
"loss": 1.
|
| 2860 |
-
"step":
|
| 2861 |
},
|
| 2862 |
{
|
| 2863 |
"epoch": 4.0,
|
| 2864 |
-
"step":
|
| 2865 |
-
"total_flos":
|
| 2866 |
-
"train_loss": 1.
|
| 2867 |
-
"train_runtime":
|
| 2868 |
-
"train_samples_per_second":
|
| 2869 |
-
"train_steps_per_second": 0.
|
| 2870 |
}
|
| 2871 |
],
|
| 2872 |
-
"logging_steps":
|
| 2873 |
-
"max_steps":
|
| 2874 |
"num_input_tokens_seen": 0,
|
| 2875 |
"num_train_epochs": 4,
|
| 2876 |
-
"save_steps":
|
| 2877 |
-
"total_flos":
|
| 2878 |
-
"train_batch_size":
|
| 2879 |
"trial_name": null,
|
| 2880 |
"trial_params": null
|
| 2881 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.9378653764724731,
|
| 3 |
+
"best_model_checkpoint": "./output/training_results/C020_Meta-Llama-3-8B_pretrain_20240726_033210/checkpoint-32778",
|
| 4 |
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 3642,
|
| 6 |
+
"global_step": 36412,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.00010985389432055367,
|
| 13 |
+
"grad_norm": 0.0,
|
| 14 |
+
"learning_rate": 0.0,
|
| 15 |
+
"loss": 2.0915,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.20004394155772823,
|
| 20 |
+
"grad_norm": 1.880800447163735,
|
| 21 |
+
"learning_rate": 1.9904796777737092e-06,
|
| 22 |
+
"loss": 2.0663,
|
| 23 |
+
"step": 1821
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.40008788311545646,
|
| 27 |
+
"grad_norm": 1.91582729652135,
|
| 28 |
+
"learning_rate": 2.239936089625888e-06,
|
| 29 |
+
"loss": 2.007,
|
| 30 |
+
"step": 3642
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.40008788311545646,
|
| 34 |
+
"eval_loss": 1.9830611944198608,
|
| 35 |
+
"eval_runtime": 322.5516,
|
| 36 |
+
"eval_samples_per_second": 200.684,
|
| 37 |
+
"eval_steps_per_second": 1.569,
|
| 38 |
+
"step": 3642
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.6001318246731847,
|
| 42 |
+
"grad_norm": 1.8799884350488187,
|
| 43 |
+
"learning_rate": 1.2178786981582618e-06,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
"loss": 1.9789,
|
| 45 |
+
"step": 5463
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.8001757662309129,
|
| 49 |
+
"grad_norm": 2.167713398694089,
|
| 50 |
+
"learning_rate": 6.502057578924368e-07,
|
| 51 |
+
"loss": 1.9635,
|
| 52 |
+
"step": 7284
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.8001757662309129,
|
| 56 |
+
"eval_loss": 1.9486902952194214,
|
| 57 |
+
"eval_runtime": 315.8318,
|
| 58 |
+
"eval_samples_per_second": 204.954,
|
| 59 |
+
"eval_steps_per_second": 1.602,
|
| 60 |
+
"step": 7284
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"epoch": 1.000219707788641,
|
| 64 |
+
"grad_norm": 1.9185025050401976,
|
| 65 |
+
"learning_rate": 3.4539250265217177e-07,
|
| 66 |
+
"loss": 1.9518,
|
| 67 |
+
"step": 9105
|
| 68 |
},
|
| 69 |
{
|
| 70 |
+
"epoch": 1.2002636493463694,
|
| 71 |
+
"grad_norm": 1.8537277337784435,
|
| 72 |
+
"learning_rate": 1.8838570953925226e-07,
|
| 73 |
+
"loss": 1.8917,
|
| 74 |
+
"step": 10926
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 1.2002636493463694,
|
| 78 |
+
"eval_loss": 1.9423131942749023,
|
| 79 |
+
"eval_runtime": 315.8691,
|
| 80 |
+
"eval_samples_per_second": 204.93,
|
| 81 |
+
"eval_steps_per_second": 1.602,
|
| 82 |
+
"step": 10926
|
| 83 |
},
|
| 84 |
{
|
| 85 |
+
"epoch": 1.4003075909040976,
|
| 86 |
+
"grad_norm": 1.8472997131423734,
|
| 87 |
+
"learning_rate": 1.113186862393777e-07,
|
| 88 |
+
"loss": 1.8886,
|
| 89 |
+
"step": 12747
|
| 90 |
},
|
| 91 |
{
|
| 92 |
+
"epoch": 1.6003515324618258,
|
| 93 |
+
"grad_norm": 1.895662682121472,
|
| 94 |
+
"learning_rate": 7.547123756350748e-08,
|
| 95 |
+
"loss": 1.8884,
|
| 96 |
+
"step": 14568
|
| 97 |
},
|
| 98 |
{
|
| 99 |
+
"epoch": 1.6003515324618258,
|
| 100 |
+
"eval_loss": 1.9401631355285645,
|
| 101 |
+
"eval_runtime": 315.8338,
|
| 102 |
+
"eval_samples_per_second": 204.953,
|
| 103 |
+
"eval_steps_per_second": 1.602,
|
| 104 |
+
"step": 14568
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 1.800395474019554,
|
| 108 |
+
"grad_norm": 1.8596122602328744,
|
| 109 |
+
"learning_rate": 5.97978448731285e-08,
|
| 110 |
+
"loss": 1.891,
|
| 111 |
+
"step": 16389
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 2.000439415577282,
|
| 115 |
+
"grad_norm": 1.752487380981475,
|
| 116 |
+
"learning_rate": 5.343917594361068e-08,
|
| 117 |
+
"loss": 1.8872,
|
| 118 |
+
"step": 18210
|
| 119 |
},
|
| 120 |
{
|
| 121 |
+
"epoch": 2.000439415577282,
|
| 122 |
+
"eval_loss": 1.9390411376953125,
|
| 123 |
+
"eval_runtime": 316.1594,
|
| 124 |
+
"eval_samples_per_second": 204.742,
|
| 125 |
+
"eval_steps_per_second": 1.6,
|
| 126 |
+
"step": 18210
|
| 127 |
},
|
| 128 |
{
|
| 129 |
+
"epoch": 2.2004833571350106,
|
| 130 |
+
"grad_norm": 1.8941258580557327,
|
| 131 |
+
"learning_rate": 5.108265610728981e-08,
|
| 132 |
+
"loss": 1.8788,
|
| 133 |
+
"step": 20031
|
| 134 |
},
|
| 135 |
{
|
| 136 |
+
"epoch": 2.400527298692739,
|
| 137 |
+
"grad_norm": 1.8532314600998243,
|
| 138 |
+
"learning_rate": 5.0297076317689476e-08,
|
| 139 |
+
"loss": 1.8811,
|
| 140 |
+
"step": 21852
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 2.400527298692739,
|
| 144 |
+
"eval_loss": 1.9393320083618164,
|
| 145 |
+
"eval_runtime": 316.5674,
|
| 146 |
+
"eval_samples_per_second": 204.478,
|
| 147 |
+
"eval_steps_per_second": 1.598,
|
| 148 |
+
"step": 21852
|
| 149 |
},
|
| 150 |
{
|
| 151 |
+
"epoch": 2.600571240250467,
|
| 152 |
+
"grad_norm": 1.856636820022415,
|
| 153 |
+
"learning_rate": 5.006877574024932e-08,
|
| 154 |
+
"loss": 1.8807,
|
| 155 |
+
"step": 23673
|
| 156 |
},
|
| 157 |
{
|
| 158 |
+
"epoch": 2.8006151818081952,
|
| 159 |
+
"grad_norm": 1.8975919928137213,
|
| 160 |
+
"learning_rate": 5.001268969632882e-08,
|
| 161 |
+
"loss": 1.8782,
|
| 162 |
+
"step": 25494
|
| 163 |
},
|
| 164 |
{
|
| 165 |
+
"epoch": 2.8006151818081952,
|
| 166 |
+
"eval_loss": 1.938640832901001,
|
| 167 |
+
"eval_runtime": 316.1355,
|
| 168 |
+
"eval_samples_per_second": 204.757,
|
| 169 |
+
"eval_steps_per_second": 1.601,
|
| 170 |
+
"step": 25494
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 3.0006591233659234,
|
| 174 |
+
"grad_norm": 1.8638851197626498,
|
| 175 |
+
"learning_rate": 5.000172130703981e-08,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 176 |
"loss": 1.8764,
|
| 177 |
+
"step": 27315
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 3.2007030649236516,
|
| 181 |
+
"grad_norm": 1.9182088121869934,
|
| 182 |
+
"learning_rate": 5.000014937976813e-08,
|
| 183 |
+
"loss": 1.8742,
|
| 184 |
+
"step": 29136
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 3.2007030649236516,
|
| 188 |
+
"eval_loss": 1.9384987354278564,
|
| 189 |
+
"eval_runtime": 316.0484,
|
| 190 |
+
"eval_samples_per_second": 204.814,
|
| 191 |
+
"eval_steps_per_second": 1.601,
|
| 192 |
+
"step": 29136
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 3.40074700648138,
|
| 196 |
+
"grad_norm": 1.9601402086429087,
|
| 197 |
+
"learning_rate": 5.000000644319432e-08,
|
| 198 |
+
"loss": 1.8775,
|
| 199 |
+
"step": 30957
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 3.600790948039108,
|
| 203 |
+
"grad_norm": 1.8139760695838174,
|
| 204 |
+
"learning_rate": 5.000000007721787e-08,
|
| 205 |
+
"loss": 1.8756,
|
| 206 |
+
"step": 32778
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 3.600790948039108,
|
| 210 |
+
"eval_loss": 1.9378653764724731,
|
| 211 |
+
"eval_runtime": 316.2121,
|
| 212 |
+
"eval_samples_per_second": 204.708,
|
| 213 |
+
"eval_steps_per_second": 1.6,
|
| 214 |
+
"step": 32778
|
| 215 |
},
|
| 216 |
{
|
| 217 |
+
"epoch": 3.8008348895968362,
|
| 218 |
+
"grad_norm": 1.9502862375669288,
|
| 219 |
+
"learning_rate": 5.000000000004247e-08,
|
| 220 |
+
"loss": 1.871,
|
| 221 |
+
"step": 34599
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 4.0,
|
| 225 |
+
"step": 36412,
|
| 226 |
+
"total_flos": 3807150279229440.0,
|
| 227 |
+
"train_loss": 1.9091134535458327,
|
| 228 |
+
"train_runtime": 51834.1311,
|
| 229 |
+
"train_samples_per_second": 44.957,
|
| 230 |
+
"train_steps_per_second": 0.702
|
| 231 |
}
|
| 232 |
],
|
| 233 |
+
"logging_steps": 1821,
|
| 234 |
+
"max_steps": 36412,
|
| 235 |
"num_input_tokens_seen": 0,
|
| 236 |
"num_train_epochs": 4,
|
| 237 |
+
"save_steps": 3642,
|
| 238 |
+
"total_flos": 3807150279229440.0,
|
| 239 |
+
"train_batch_size": 8,
|
| 240 |
"trial_name": null,
|
| 241 |
"trial_params": null
|
| 242 |
}
|