Training in progress, step 400, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 723674912
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dcb42890fd3e3733df15325188e71ea98cd125dad14aa982eb9d9229b15a8bdc
|
| 3 |
size 723674912
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 735625626
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7aaaf6afb1371a4f19d7057b3f5ca8fca65fcc7a584da182d9f36fb7032085bb
|
| 3 |
size 735625626
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9feae33b2fec0a6229240e7adaee6ecc8f5cfdf1a8bd0e827b1d8a241424e3c0
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a673aaf85c0fe6b6c29cb8f3e7dbd829eef637110e4ad9a775f3fcf001c92591
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -11,1490 +11,306 @@
|
|
| 11 |
{
|
| 12 |
"epoch": 0.0002959455460195324,
|
| 13 |
"eval_loss": 3.0371296405792236,
|
| 14 |
-
"eval_runtime":
|
| 15 |
-
"eval_samples_per_second":
|
| 16 |
-
"eval_steps_per_second":
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.002959455460195324,
|
| 21 |
"grad_norm": 2.84375,
|
| 22 |
"learning_rate": 1.6000000000000003e-05,
|
| 23 |
-
"loss": 2.
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.005918910920390648,
|
| 28 |
-
"grad_norm": 4.
|
| 29 |
"learning_rate": 3.2000000000000005e-05,
|
| 30 |
"loss": 2.6895,
|
| 31 |
"step": 20
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.008878366380585973,
|
| 35 |
-
"grad_norm": 5.
|
| 36 |
"learning_rate": 4.8e-05,
|
| 37 |
-
"loss": 2.
|
| 38 |
"step": 30
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.011837821840781295,
|
| 42 |
-
"grad_norm": 7.
|
| 43 |
"learning_rate": 6.400000000000001e-05,
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 40
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.01479727730097662,
|
| 49 |
"grad_norm": 25.375,
|
| 50 |
"learning_rate": 8e-05,
|
| 51 |
-
"loss": 3.
|
| 52 |
"step": 50
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.017756732761171946,
|
| 56 |
-
"grad_norm": 3.
|
| 57 |
"learning_rate": 9.6e-05,
|
| 58 |
-
"loss": 2.
|
| 59 |
"step": 60
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.020716188221367268,
|
| 63 |
-
"grad_norm": 2.
|
| 64 |
"learning_rate": 0.00011200000000000001,
|
| 65 |
-
"loss": 2.
|
| 66 |
"step": 70
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.02367564368156259,
|
| 70 |
"grad_norm": 4.34375,
|
| 71 |
"learning_rate": 0.00012800000000000002,
|
| 72 |
-
"loss": 2.
|
| 73 |
"step": 80
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.026635099141757917,
|
| 77 |
"grad_norm": 5.5,
|
| 78 |
"learning_rate": 0.000144,
|
| 79 |
-
"loss": 2.
|
| 80 |
"step": 90
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.02959455460195324,
|
| 84 |
-
"grad_norm": 22.
|
| 85 |
"learning_rate": 0.00016,
|
| 86 |
-
"loss": 3.
|
| 87 |
"step": 100
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.032554010062148565,
|
| 91 |
-
"grad_norm": 3.
|
| 92 |
"learning_rate": 0.00017600000000000002,
|
| 93 |
-
"loss": 2.
|
| 94 |
"step": 110
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.03551346552234389,
|
| 98 |
-
"grad_norm": 2.
|
| 99 |
"learning_rate": 0.000192,
|
| 100 |
-
"loss": 2.
|
| 101 |
"step": 120
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.03847292098253921,
|
| 105 |
-
"grad_norm": 2.
|
| 106 |
"learning_rate": 0.0001999978128380225,
|
| 107 |
-
"loss": 2.
|
| 108 |
"step": 130
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.041432376442734536,
|
| 112 |
-
"grad_norm": 7.
|
| 113 |
"learning_rate": 0.0001999803161162393,
|
| 114 |
"loss": 2.3626,
|
| 115 |
"step": 140
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.04439183190292986,
|
| 119 |
-
"grad_norm":
|
| 120 |
"learning_rate": 0.00019994532573409262,
|
| 121 |
-
"loss": 2.
|
| 122 |
"step": 150
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.04735128736312518,
|
| 126 |
"grad_norm": 5.6875,
|
| 127 |
"learning_rate": 0.00019989284781388617,
|
| 128 |
-
"loss": 2.
|
| 129 |
"step": 160
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.05031074282332051,
|
| 133 |
-
"grad_norm": 2.
|
| 134 |
"learning_rate": 0.00019982289153773646,
|
| 135 |
-
"loss": 2.
|
| 136 |
"step": 170
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.053270198283515834,
|
| 140 |
-
"grad_norm": 4.
|
| 141 |
"learning_rate": 0.00019973546914596623,
|
| 142 |
-
"loss": 2.
|
| 143 |
"step": 180
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.05622965374371116,
|
| 147 |
"grad_norm": 6.90625,
|
| 148 |
"learning_rate": 0.00019963059593496268,
|
| 149 |
-
"loss": 2.
|
| 150 |
"step": 190
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.05918910920390648,
|
| 154 |
-
"grad_norm":
|
| 155 |
"learning_rate": 0.00019950829025450114,
|
| 156 |
-
"loss": 2.
|
| 157 |
"step": 200
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.05918910920390648,
|
| 161 |
-
"eval_loss": 2.
|
| 162 |
-
"eval_runtime":
|
| 163 |
-
"eval_samples_per_second":
|
| 164 |
-
"eval_steps_per_second":
|
| 165 |
"step": 200
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.062148564664101805,
|
| 169 |
-
"grad_norm": 2.
|
| 170 |
"learning_rate": 0.0001993685735045343,
|
| 171 |
-
"loss": 2.
|
| 172 |
"step": 210
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.06510802012429713,
|
| 176 |
"grad_norm": 2.890625,
|
| 177 |
"learning_rate": 0.0001992114701314478,
|
| 178 |
-
"loss": 2.
|
| 179 |
"step": 220
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.06806747558449246,
|
| 183 |
-
"grad_norm": 3.
|
| 184 |
"learning_rate": 0.000199037007623783,
|
| 185 |
-
"loss": 2.
|
| 186 |
"step": 230
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 0.07102693104468778,
|
| 190 |
"grad_norm": 6.65625,
|
| 191 |
"learning_rate": 0.00019884521650742715,
|
| 192 |
-
"loss": 2.
|
| 193 |
"step": 240
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.0739863865048831,
|
| 197 |
-
"grad_norm": 39.
|
| 198 |
"learning_rate": 0.00019863613034027224,
|
| 199 |
-
"loss": 2.
|
| 200 |
"step": 250
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.07694584196507842,
|
| 204 |
-
"grad_norm": 5.
|
| 205 |
"learning_rate": 0.0001984097857063434,
|
| 206 |
-
"loss": 2.
|
| 207 |
"step": 260
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.07990529742527375,
|
| 211 |
-
"grad_norm": 2.
|
| 212 |
"learning_rate": 0.0001981662222093976,
|
| 213 |
-
"loss": 2.
|
| 214 |
"step": 270
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 0.08286475288546907,
|
| 218 |
-
"grad_norm": 3.
|
| 219 |
"learning_rate": 0.00019790548246599447,
|
| 220 |
-
"loss": 2.
|
| 221 |
"step": 280
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 0.0858242083456644,
|
| 225 |
-
"grad_norm": 5.
|
| 226 |
"learning_rate": 0.00019762761209803927,
|
| 227 |
-
"loss": 2.
|
| 228 |
"step": 290
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 0.08878366380585972,
|
| 232 |
-
"grad_norm": 15.
|
| 233 |
"learning_rate": 0.0001973326597248006,
|
| 234 |
-
"loss": 2.
|
| 235 |
"step": 300
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 0.09174311926605505,
|
| 239 |
"grad_norm": 3.234375,
|
| 240 |
"learning_rate": 0.00019702067695440332,
|
| 241 |
-
"loss": 2.
|
| 242 |
"step": 310
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.09470257472625036,
|
| 246 |
-
"grad_norm": 2.
|
| 247 |
"learning_rate": 0.00019669171837479873,
|
| 248 |
-
"loss": 2.
|
| 249 |
"step": 320
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.09766203018644569,
|
| 253 |
"grad_norm": 3.0,
|
| 254 |
"learning_rate": 0.00019634584154421317,
|
| 255 |
-
"loss": 2.
|
| 256 |
"step": 330
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 0.10062148564664102,
|
| 260 |
-
"grad_norm": 5.
|
| 261 |
"learning_rate": 0.00019598310698107702,
|
| 262 |
-
"loss": 2.
|
| 263 |
"step": 340
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.10358094110683634,
|
| 267 |
-
"grad_norm": 23.
|
| 268 |
"learning_rate": 0.00019560357815343577,
|
| 269 |
-
"loss": 1.
|
| 270 |
"step": 350
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 0.10654039656703167,
|
| 274 |
"grad_norm": 2.515625,
|
| 275 |
"learning_rate": 0.00019520732146784491,
|
| 276 |
-
"loss": 2.
|
| 277 |
"step": 360
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 0.109499852027227,
|
| 281 |
"grad_norm": 3.09375,
|
| 282 |
"learning_rate": 0.0001947944062577507,
|
| 283 |
-
"loss": 2.
|
| 284 |
"step": 370
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.11245930748742232,
|
| 288 |
"grad_norm": 4.125,
|
| 289 |
"learning_rate": 0.00019436490477135878,
|
| 290 |
-
"loss": 2.
|
| 291 |
"step": 380
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 0.11541876294761765,
|
| 295 |
-
"grad_norm": 5.
|
| 296 |
"learning_rate": 0.00019391889215899299,
|
| 297 |
-
"loss": 1.
|
| 298 |
"step": 390
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 0.11837821840781296,
|
| 302 |
-
"grad_norm": 25.
|
| 303 |
"learning_rate": 0.0001934564464599461,
|
| 304 |
-
"loss": 1.
|
| 305 |
"step": 400
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 0.11837821840781296,
|
| 309 |
-
"eval_loss": 2.
|
| 310 |
-
"eval_runtime":
|
| 311 |
-
"eval_samples_per_second":
|
| 312 |
-
"eval_steps_per_second":
|
| 313 |
"step": 400
|
| 314 |
-
},
|
| 315 |
-
{
|
| 316 |
-
"epoch": 0.12133767386800828,
|
| 317 |
-
"grad_norm": 1.9296875,
|
| 318 |
-
"learning_rate": 0.00019297764858882514,
|
| 319 |
-
"loss": 2.4536,
|
| 320 |
-
"step": 410
|
| 321 |
-
},
|
| 322 |
-
{
|
| 323 |
-
"epoch": 0.12429712932820361,
|
| 324 |
-
"grad_norm": 3.796875,
|
| 325 |
-
"learning_rate": 0.00019248258232139388,
|
| 326 |
-
"loss": 2.2639,
|
| 327 |
-
"step": 420
|
| 328 |
-
},
|
| 329 |
-
{
|
| 330 |
-
"epoch": 0.12725658478839894,
|
| 331 |
-
"grad_norm": 4.21875,
|
| 332 |
-
"learning_rate": 0.00019197133427991436,
|
| 333 |
-
"loss": 2.1831,
|
| 334 |
-
"step": 430
|
| 335 |
-
},
|
| 336 |
-
{
|
| 337 |
-
"epoch": 0.13021604024859426,
|
| 338 |
-
"grad_norm": 3.78125,
|
| 339 |
-
"learning_rate": 0.00019144399391799043,
|
| 340 |
-
"loss": 2.0126,
|
| 341 |
-
"step": 440
|
| 342 |
-
},
|
| 343 |
-
{
|
| 344 |
-
"epoch": 0.1331754957087896,
|
| 345 |
-
"grad_norm": 26.875,
|
| 346 |
-
"learning_rate": 0.00019090065350491626,
|
| 347 |
-
"loss": 2.449,
|
| 348 |
-
"step": 450
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 0.1361349511689849,
|
| 352 |
-
"grad_norm": 2.0625,
|
| 353 |
-
"learning_rate": 0.0001903414081095315,
|
| 354 |
-
"loss": 2.3837,
|
| 355 |
-
"step": 460
|
| 356 |
-
},
|
| 357 |
-
{
|
| 358 |
-
"epoch": 0.13909440662918024,
|
| 359 |
-
"grad_norm": 2.265625,
|
| 360 |
-
"learning_rate": 0.00018976635558358722,
|
| 361 |
-
"loss": 2.121,
|
| 362 |
-
"step": 470
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 0.14205386208937557,
|
| 366 |
-
"grad_norm": 3.953125,
|
| 367 |
-
"learning_rate": 0.00018917559654462474,
|
| 368 |
-
"loss": 2.1884,
|
| 369 |
-
"step": 480
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.1450133175495709,
|
| 373 |
-
"grad_norm": 5.21875,
|
| 374 |
-
"learning_rate": 0.00018856923435837022,
|
| 375 |
-
"loss": 2.1975,
|
| 376 |
-
"step": 490
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 0.1479727730097662,
|
| 380 |
-
"grad_norm": 19.75,
|
| 381 |
-
"learning_rate": 0.0001879473751206489,
|
| 382 |
-
"loss": 1.6031,
|
| 383 |
-
"step": 500
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 0.15093222846996152,
|
| 387 |
-
"grad_norm": 2.109375,
|
| 388 |
-
"learning_rate": 0.00018731012763882133,
|
| 389 |
-
"loss": 2.4093,
|
| 390 |
-
"step": 510
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 0.15389168393015684,
|
| 394 |
-
"grad_norm": 2.8125,
|
| 395 |
-
"learning_rate": 0.00018665760341274505,
|
| 396 |
-
"loss": 2.199,
|
| 397 |
-
"step": 520
|
| 398 |
-
},
|
| 399 |
-
{
|
| 400 |
-
"epoch": 0.15685113939035217,
|
| 401 |
-
"grad_norm": 3.765625,
|
| 402 |
-
"learning_rate": 0.00018598991661526572,
|
| 403 |
-
"loss": 2.2808,
|
| 404 |
-
"step": 530
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 0.1598105948505475,
|
| 408 |
-
"grad_norm": 3.796875,
|
| 409 |
-
"learning_rate": 0.00018530718407223974,
|
| 410 |
-
"loss": 2.221,
|
| 411 |
-
"step": 540
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 0.16277005031074282,
|
| 415 |
-
"grad_norm": 18.875,
|
| 416 |
-
"learning_rate": 0.00018460952524209355,
|
| 417 |
-
"loss": 1.9977,
|
| 418 |
-
"step": 550
|
| 419 |
-
},
|
| 420 |
-
{
|
| 421 |
-
"epoch": 0.16572950577093815,
|
| 422 |
-
"grad_norm": 2.84375,
|
| 423 |
-
"learning_rate": 0.00018389706219492147,
|
| 424 |
-
"loss": 2.2811,
|
| 425 |
-
"step": 560
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 0.16868896123113347,
|
| 429 |
-
"grad_norm": 2.3125,
|
| 430 |
-
"learning_rate": 0.00018316991959112716,
|
| 431 |
-
"loss": 2.3694,
|
| 432 |
-
"step": 570
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"epoch": 0.1716484166913288,
|
| 436 |
-
"grad_norm": 2.796875,
|
| 437 |
-
"learning_rate": 0.00018242822465961176,
|
| 438 |
-
"loss": 1.8728,
|
| 439 |
-
"step": 580
|
| 440 |
-
},
|
| 441 |
-
{
|
| 442 |
-
"epoch": 0.17460787215152412,
|
| 443 |
-
"grad_norm": 8.0625,
|
| 444 |
-
"learning_rate": 0.00018167210717551224,
|
| 445 |
-
"loss": 2.0789,
|
| 446 |
-
"step": 590
|
| 447 |
-
},
|
| 448 |
-
{
|
| 449 |
-
"epoch": 0.17756732761171945,
|
| 450 |
-
"grad_norm": 17.375,
|
| 451 |
-
"learning_rate": 0.00018090169943749476,
|
| 452 |
-
"loss": 1.9051,
|
| 453 |
-
"step": 600
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 0.17756732761171945,
|
| 457 |
-
"eval_loss": 2.1772165298461914,
|
| 458 |
-
"eval_runtime": 38.2926,
|
| 459 |
-
"eval_samples_per_second": 39.224,
|
| 460 |
-
"eval_steps_per_second": 9.819,
|
| 461 |
-
"step": 600
|
| 462 |
-
},
|
| 463 |
-
{
|
| 464 |
-
"epoch": 0.18052678307191478,
|
| 465 |
-
"grad_norm": 2.75,
|
| 466 |
-
"learning_rate": 0.00018011713624460608,
|
| 467 |
-
"loss": 2.0598,
|
| 468 |
-
"step": 610
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 0.1834862385321101,
|
| 472 |
-
"grad_norm": 2.203125,
|
| 473 |
-
"learning_rate": 0.00017931855487268782,
|
| 474 |
-
"loss": 2.0458,
|
| 475 |
-
"step": 620
|
| 476 |
-
},
|
| 477 |
-
{
|
| 478 |
-
"epoch": 0.18644569399230543,
|
| 479 |
-
"grad_norm": 3.046875,
|
| 480 |
-
"learning_rate": 0.0001785060950503568,
|
| 481 |
-
"loss": 2.4327,
|
| 482 |
-
"step": 630
|
| 483 |
-
},
|
| 484 |
-
{
|
| 485 |
-
"epoch": 0.18940514945250073,
|
| 486 |
-
"grad_norm": 10.5,
|
| 487 |
-
"learning_rate": 0.00017767989893455698,
|
| 488 |
-
"loss": 2.2315,
|
| 489 |
-
"step": 640
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 0.19236460491269605,
|
| 493 |
-
"grad_norm": 19.125,
|
| 494 |
-
"learning_rate": 0.00017684011108568592,
|
| 495 |
-
"loss": 2.2792,
|
| 496 |
-
"step": 650
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 0.19532406037289138,
|
| 500 |
-
"grad_norm": 2.140625,
|
| 501 |
-
"learning_rate": 0.00017598687844230088,
|
| 502 |
-
"loss": 2.4382,
|
| 503 |
-
"step": 660
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 0.1982835158330867,
|
| 507 |
-
"grad_norm": 2.78125,
|
| 508 |
-
"learning_rate": 0.00017512035029540885,
|
| 509 |
-
"loss": 2.1791,
|
| 510 |
-
"step": 670
|
| 511 |
-
},
|
| 512 |
-
{
|
| 513 |
-
"epoch": 0.20124297129328203,
|
| 514 |
-
"grad_norm": 5.1875,
|
| 515 |
-
"learning_rate": 0.000174240678262345,
|
| 516 |
-
"loss": 2.2385,
|
| 517 |
-
"step": 680
|
| 518 |
-
},
|
| 519 |
-
{
|
| 520 |
-
"epoch": 0.20420242675347736,
|
| 521 |
-
"grad_norm": 6.25,
|
| 522 |
-
"learning_rate": 0.000173348016260244,
|
| 523 |
-
"loss": 1.9492,
|
| 524 |
-
"step": 690
|
| 525 |
-
},
|
| 526 |
-
{
|
| 527 |
-
"epoch": 0.20716188221367268,
|
| 528 |
-
"grad_norm": 27.125,
|
| 529 |
-
"learning_rate": 0.00017244252047910892,
|
| 530 |
-
"loss": 1.8527,
|
| 531 |
-
"step": 700
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 0.210121337673868,
|
| 535 |
-
"grad_norm": 2.15625,
|
| 536 |
-
"learning_rate": 0.00017152434935448256,
|
| 537 |
-
"loss": 2.2149,
|
| 538 |
-
"step": 710
|
| 539 |
-
},
|
| 540 |
-
{
|
| 541 |
-
"epoch": 0.21308079313406333,
|
| 542 |
-
"grad_norm": 2.46875,
|
| 543 |
-
"learning_rate": 0.0001705936635397259,
|
| 544 |
-
"loss": 2.3124,
|
| 545 |
-
"step": 720
|
| 546 |
-
},
|
| 547 |
-
{
|
| 548 |
-
"epoch": 0.21604024859425866,
|
| 549 |
-
"grad_norm": 3.5625,
|
| 550 |
-
"learning_rate": 0.00016965062587790823,
|
| 551 |
-
"loss": 2.109,
|
| 552 |
-
"step": 730
|
| 553 |
-
},
|
| 554 |
-
{
|
| 555 |
-
"epoch": 0.218999704054454,
|
| 556 |
-
"grad_norm": 2.703125,
|
| 557 |
-
"learning_rate": 0.00016869540137331445,
|
| 558 |
-
"loss": 1.9086,
|
| 559 |
-
"step": 740
|
| 560 |
-
},
|
| 561 |
-
{
|
| 562 |
-
"epoch": 0.2219591595146493,
|
| 563 |
-
"grad_norm": 10.0625,
|
| 564 |
-
"learning_rate": 0.00016772815716257412,
|
| 565 |
-
"loss": 1.9308,
|
| 566 |
-
"step": 750
|
| 567 |
-
},
|
| 568 |
-
{
|
| 569 |
-
"epoch": 0.22491861497484464,
|
| 570 |
-
"grad_norm": 1.78125,
|
| 571 |
-
"learning_rate": 0.00016674906248541726,
|
| 572 |
-
"loss": 2.3023,
|
| 573 |
-
"step": 760
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 0.22787807043503996,
|
| 577 |
-
"grad_norm": 3.78125,
|
| 578 |
-
"learning_rate": 0.00016575828865506245,
|
| 579 |
-
"loss": 2.1113,
|
| 580 |
-
"step": 770
|
| 581 |
-
},
|
| 582 |
-
{
|
| 583 |
-
"epoch": 0.2308375258952353,
|
| 584 |
-
"grad_norm": 4.125,
|
| 585 |
-
"learning_rate": 0.0001647560090282419,
|
| 586 |
-
"loss": 2.0416,
|
| 587 |
-
"step": 780
|
| 588 |
-
},
|
| 589 |
-
{
|
| 590 |
-
"epoch": 0.2337969813554306,
|
| 591 |
-
"grad_norm": 6.5625,
|
| 592 |
-
"learning_rate": 0.000163742398974869,
|
| 593 |
-
"loss": 1.9142,
|
| 594 |
-
"step": 790
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 0.23675643681562591,
|
| 598 |
-
"grad_norm": 43.0,
|
| 599 |
-
"learning_rate": 0.0001627176358473537,
|
| 600 |
-
"loss": 2.1002,
|
| 601 |
-
"step": 800
|
| 602 |
-
},
|
| 603 |
-
{
|
| 604 |
-
"epoch": 0.23675643681562591,
|
| 605 |
-
"eval_loss": 2.160681962966919,
|
| 606 |
-
"eval_runtime": 38.0783,
|
| 607 |
-
"eval_samples_per_second": 39.445,
|
| 608 |
-
"eval_steps_per_second": 9.874,
|
| 609 |
-
"step": 800
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 0.23971589227582124,
|
| 613 |
-
"grad_norm": 2.390625,
|
| 614 |
-
"learning_rate": 0.0001616818989495711,
|
| 615 |
-
"loss": 2.4747,
|
| 616 |
-
"step": 810
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 0.24267534773601657,
|
| 620 |
-
"grad_norm": 3.59375,
|
| 621 |
-
"learning_rate": 0.00016063536950548826,
|
| 622 |
-
"loss": 2.1356,
|
| 623 |
-
"step": 820
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 0.2456348031962119,
|
| 627 |
-
"grad_norm": 2.484375,
|
| 628 |
-
"learning_rate": 0.0001595782306274553,
|
| 629 |
-
"loss": 2.0635,
|
| 630 |
-
"step": 830
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 0.24859425865640722,
|
| 634 |
-
"grad_norm": 5.625,
|
| 635 |
-
"learning_rate": 0.00015851066728416618,
|
| 636 |
-
"loss": 2.122,
|
| 637 |
-
"step": 840
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 0.25155371411660254,
|
| 641 |
-
"grad_norm": 27.75,
|
| 642 |
-
"learning_rate": 0.00015743286626829437,
|
| 643 |
-
"loss": 2.1348,
|
| 644 |
-
"step": 850
|
| 645 |
-
},
|
| 646 |
-
{
|
| 647 |
-
"epoch": 0.25451316957679787,
|
| 648 |
-
"grad_norm": 1.875,
|
| 649 |
-
"learning_rate": 0.00015634501616380967,
|
| 650 |
-
"loss": 2.2736,
|
| 651 |
-
"step": 860
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 0.2574726250369932,
|
| 655 |
-
"grad_norm": 3.34375,
|
| 656 |
-
"learning_rate": 0.00015524730731298134,
|
| 657 |
-
"loss": 2.1229,
|
| 658 |
-
"step": 870
|
| 659 |
-
},
|
| 660 |
-
{
|
| 661 |
-
"epoch": 0.2604320804971885,
|
| 662 |
-
"grad_norm": 3.640625,
|
| 663 |
-
"learning_rate": 0.0001541399317830738,
|
| 664 |
-
"loss": 2.3073,
|
| 665 |
-
"step": 880
|
| 666 |
-
},
|
| 667 |
-
{
|
| 668 |
-
"epoch": 0.26339153595738385,
|
| 669 |
-
"grad_norm": 4.4375,
|
| 670 |
-
"learning_rate": 0.0001530230833327405,
|
| 671 |
-
"loss": 1.9859,
|
| 672 |
-
"step": 890
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 0.2663509914175792,
|
| 676 |
-
"grad_norm": 12.25,
|
| 677 |
-
"learning_rate": 0.00015189695737812152,
|
| 678 |
-
"loss": 2.0202,
|
| 679 |
-
"step": 900
|
| 680 |
-
},
|
| 681 |
-
{
|
| 682 |
-
"epoch": 0.2693104468777745,
|
| 683 |
-
"grad_norm": 2.1875,
|
| 684 |
-
"learning_rate": 0.0001507617509586517,
|
| 685 |
-
"loss": 2.2262,
|
| 686 |
-
"step": 910
|
| 687 |
-
},
|
| 688 |
-
{
|
| 689 |
-
"epoch": 0.2722699023379698,
|
| 690 |
-
"grad_norm": 2.125,
|
| 691 |
-
"learning_rate": 0.00014961766270258422,
|
| 692 |
-
"loss": 2.2005,
|
| 693 |
-
"step": 920
|
| 694 |
-
},
|
| 695 |
-
{
|
| 696 |
-
"epoch": 0.27522935779816515,
|
| 697 |
-
"grad_norm": 2.734375,
|
| 698 |
-
"learning_rate": 0.00014846489279223652,
|
| 699 |
-
"loss": 1.8805,
|
| 700 |
-
"step": 930
|
| 701 |
-
},
|
| 702 |
-
{
|
| 703 |
-
"epoch": 0.2781888132583605,
|
| 704 |
-
"grad_norm": 4.5,
|
| 705 |
-
"learning_rate": 0.0001473036429289641,
|
| 706 |
-
"loss": 2.272,
|
| 707 |
-
"step": 940
|
| 708 |
-
},
|
| 709 |
-
{
|
| 710 |
-
"epoch": 0.2811482687185558,
|
| 711 |
-
"grad_norm": 23.5,
|
| 712 |
-
"learning_rate": 0.0001461341162978688,
|
| 713 |
-
"loss": 1.926,
|
| 714 |
-
"step": 950
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 0.28410772417875113,
|
| 718 |
-
"grad_norm": 2.234375,
|
| 719 |
-
"learning_rate": 0.00014495651753224705,
|
| 720 |
-
"loss": 2.264,
|
| 721 |
-
"step": 960
|
| 722 |
-
},
|
| 723 |
-
{
|
| 724 |
-
"epoch": 0.28706717963894646,
|
| 725 |
-
"grad_norm": 2.28125,
|
| 726 |
-
"learning_rate": 0.00014377105267778518,
|
| 727 |
-
"loss": 2.3195,
|
| 728 |
-
"step": 970
|
| 729 |
-
},
|
| 730 |
-
{
|
| 731 |
-
"epoch": 0.2900266350991418,
|
| 732 |
-
"grad_norm": 2.859375,
|
| 733 |
-
"learning_rate": 0.00014257792915650728,
|
| 734 |
-
"loss": 2.1811,
|
| 735 |
-
"step": 980
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 0.2929860905593371,
|
| 739 |
-
"grad_norm": 5.34375,
|
| 740 |
-
"learning_rate": 0.00014137735573048233,
|
| 741 |
-
"loss": 2.2588,
|
| 742 |
-
"step": 990
|
| 743 |
-
},
|
| 744 |
-
{
|
| 745 |
-
"epoch": 0.2959455460195324,
|
| 746 |
-
"grad_norm": 37.75,
|
| 747 |
-
"learning_rate": 0.00014016954246529696,
|
| 748 |
-
"loss": 1.901,
|
| 749 |
-
"step": 1000
|
| 750 |
-
},
|
| 751 |
-
{
|
| 752 |
-
"epoch": 0.2959455460195324,
|
| 753 |
-
"eval_loss": 2.1415603160858154,
|
| 754 |
-
"eval_runtime": 37.3621,
|
| 755 |
-
"eval_samples_per_second": 40.201,
|
| 756 |
-
"eval_steps_per_second": 10.064,
|
| 757 |
-
"step": 1000
|
| 758 |
-
},
|
| 759 |
-
{
|
| 760 |
-
"epoch": 0.2989050014797277,
|
| 761 |
-
"grad_norm": 2.15625,
|
| 762 |
-
"learning_rate": 0.00013895470069330004,
|
| 763 |
-
"loss": 2.4456,
|
| 764 |
-
"step": 1010
|
| 765 |
-
},
|
| 766 |
-
{
|
| 767 |
-
"epoch": 0.30186445693992303,
|
| 768 |
-
"grad_norm": 2.984375,
|
| 769 |
-
"learning_rate": 0.00013773304297662559,
|
| 770 |
-
"loss": 2.0006,
|
| 771 |
-
"step": 1020
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 0.30482391240011836,
|
| 775 |
-
"grad_norm": 3.5625,
|
| 776 |
-
"learning_rate": 0.00013650478307000057,
|
| 777 |
-
"loss": 2.1085,
|
| 778 |
-
"step": 1030
|
| 779 |
-
},
|
| 780 |
-
{
|
| 781 |
-
"epoch": 0.3077833678603137,
|
| 782 |
-
"grad_norm": 9.5,
|
| 783 |
-
"learning_rate": 0.00013527013588334415,
|
| 784 |
-
"loss": 2.2559,
|
| 785 |
-
"step": 1040
|
| 786 |
-
},
|
| 787 |
-
{
|
| 788 |
-
"epoch": 0.310742823320509,
|
| 789 |
-
"grad_norm": 19.375,
|
| 790 |
-
"learning_rate": 0.00013402931744416433,
|
| 791 |
-
"loss": 2.0562,
|
| 792 |
-
"step": 1050
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"epoch": 0.31370227878070434,
|
| 796 |
-
"grad_norm": 2.109375,
|
| 797 |
-
"learning_rate": 0.00013278254485975976,
|
| 798 |
-
"loss": 2.2395,
|
| 799 |
-
"step": 1060
|
| 800 |
-
},
|
| 801 |
-
{
|
| 802 |
-
"epoch": 0.31666173424089966,
|
| 803 |
-
"grad_norm": 2.578125,
|
| 804 |
-
"learning_rate": 0.00013153003627923218,
|
| 805 |
-
"loss": 2.3515,
|
| 806 |
-
"step": 1070
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 0.319621189701095,
|
| 810 |
-
"grad_norm": 3.890625,
|
| 811 |
-
"learning_rate": 0.00013027201085531634,
|
| 812 |
-
"loss": 2.3772,
|
| 813 |
-
"step": 1080
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 0.3225806451612903,
|
| 817 |
-
"grad_norm": 3.390625,
|
| 818 |
-
"learning_rate": 0.00012900868870603503,
|
| 819 |
-
"loss": 2.0078,
|
| 820 |
-
"step": 1090
|
| 821 |
-
},
|
| 822 |
-
{
|
| 823 |
-
"epoch": 0.32554010062148564,
|
| 824 |
-
"grad_norm": 17.75,
|
| 825 |
-
"learning_rate": 0.00012774029087618446,
|
| 826 |
-
"loss": 2.3486,
|
| 827 |
-
"step": 1100
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"epoch": 0.32849955608168097,
|
| 831 |
-
"grad_norm": 2.0,
|
| 832 |
-
"learning_rate": 0.00012646703929865817,
|
| 833 |
-
"loss": 2.1657,
|
| 834 |
-
"step": 1110
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 0.3314590115418763,
|
| 838 |
-
"grad_norm": 2.578125,
|
| 839 |
-
"learning_rate": 0.00012518915675561483,
|
| 840 |
-
"loss": 2.3418,
|
| 841 |
-
"step": 1120
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 0.3344184670020716,
|
| 845 |
-
"grad_norm": 3.390625,
|
| 846 |
-
"learning_rate": 0.00012390686683949798,
|
| 847 |
-
"loss": 2.0369,
|
| 848 |
-
"step": 1130
|
| 849 |
-
},
|
| 850 |
-
{
|
| 851 |
-
"epoch": 0.33737792246226694,
|
| 852 |
-
"grad_norm": 7.90625,
|
| 853 |
-
"learning_rate": 0.00012262039391391404,
|
| 854 |
-
"loss": 2.0235,
|
| 855 |
-
"step": 1140
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 0.34033737792246227,
|
| 859 |
-
"grad_norm": 31.75,
|
| 860 |
-
"learning_rate": 0.0001213299630743747,
|
| 861 |
-
"loss": 2.1905,
|
| 862 |
-
"step": 1150
|
| 863 |
-
},
|
| 864 |
-
{
|
| 865 |
-
"epoch": 0.3432968333826576,
|
| 866 |
-
"grad_norm": 2.421875,
|
| 867 |
-
"learning_rate": 0.00012003580010891213,
|
| 868 |
-
"loss": 2.4817,
|
| 869 |
-
"step": 1160
|
| 870 |
-
},
|
| 871 |
-
{
|
| 872 |
-
"epoch": 0.3462562888428529,
|
| 873 |
-
"grad_norm": 2.25,
|
| 874 |
-
"learning_rate": 0.00011873813145857249,
|
| 875 |
-
"loss": 2.195,
|
| 876 |
-
"step": 1170
|
| 877 |
-
},
|
| 878 |
-
{
|
| 879 |
-
"epoch": 0.34921574430304825,
|
| 880 |
-
"grad_norm": 4.0,
|
| 881 |
-
"learning_rate": 0.00011743718417779517,
|
| 882 |
-
"loss": 2.307,
|
| 883 |
-
"step": 1180
|
| 884 |
-
},
|
| 885 |
-
{
|
| 886 |
-
"epoch": 0.3521751997632436,
|
| 887 |
-
"grad_norm": 4.59375,
|
| 888 |
-
"learning_rate": 0.00011613318589468511,
|
| 889 |
-
"loss": 1.9753,
|
| 890 |
-
"step": 1190
|
| 891 |
-
},
|
| 892 |
-
{
|
| 893 |
-
"epoch": 0.3551346552234389,
|
| 894 |
-
"grad_norm": 11.3125,
|
| 895 |
-
"learning_rate": 0.0001148263647711842,
|
| 896 |
-
"loss": 1.53,
|
| 897 |
-
"step": 1200
|
| 898 |
-
},
|
| 899 |
-
{
|
| 900 |
-
"epoch": 0.3551346552234389,
|
| 901 |
-
"eval_loss": 2.1453933715820312,
|
| 902 |
-
"eval_runtime": 37.6799,
|
| 903 |
-
"eval_samples_per_second": 39.862,
|
| 904 |
-
"eval_steps_per_second": 9.979,
|
| 905 |
-
"step": 1200
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 0.3580941106836342,
|
| 909 |
-
"grad_norm": 2.46875,
|
| 910 |
-
"learning_rate": 0.0001135169494631497,
|
| 911 |
-
"loss": 2.2422,
|
| 912 |
-
"step": 1210
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 0.36105356614382955,
|
| 916 |
-
"grad_norm": 2.46875,
|
| 917 |
-
"learning_rate": 0.00011220516908034601,
|
| 918 |
-
"loss": 2.1766,
|
| 919 |
-
"step": 1220
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 0.3640130216040249,
|
| 923 |
-
"grad_norm": 4.40625,
|
| 924 |
-
"learning_rate": 0.00011089125314635726,
|
| 925 |
-
"loss": 2.2057,
|
| 926 |
-
"step": 1230
|
| 927 |
-
},
|
| 928 |
-
{
|
| 929 |
-
"epoch": 0.3669724770642202,
|
| 930 |
-
"grad_norm": 5.09375,
|
| 931 |
-
"learning_rate": 0.00010957543155842702,
|
| 932 |
-
"loss": 1.7583,
|
| 933 |
-
"step": 1240
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 0.36993193252441553,
|
| 937 |
-
"grad_norm": 16.125,
|
| 938 |
-
"learning_rate": 0.00010825793454723325,
|
| 939 |
-
"loss": 1.9914,
|
| 940 |
-
"step": 1250
|
| 941 |
-
},
|
| 942 |
-
{
|
| 943 |
-
"epoch": 0.37289138798461086,
|
| 944 |
-
"grad_norm": 2.046875,
|
| 945 |
-
"learning_rate": 0.00010693899263660441,
|
| 946 |
-
"loss": 2.4002,
|
| 947 |
-
"step": 1260
|
| 948 |
-
},
|
| 949 |
-
{
|
| 950 |
-
"epoch": 0.3758508434448062,
|
| 951 |
-
"grad_norm": 3.0,
|
| 952 |
-
"learning_rate": 0.00010561883660318455,
|
| 953 |
-
"loss": 2.2105,
|
| 954 |
-
"step": 1270
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 0.37881029890500145,
|
| 958 |
-
"grad_norm": 3.953125,
|
| 959 |
-
"learning_rate": 0.00010429769743605407,
|
| 960 |
-
"loss": 1.9458,
|
| 961 |
-
"step": 1280
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 0.3817697543651968,
|
| 965 |
-
"grad_norm": 3.953125,
|
| 966 |
-
"learning_rate": 0.00010297580629631325,
|
| 967 |
-
"loss": 1.7066,
|
| 968 |
-
"step": 1290
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 0.3847292098253921,
|
| 972 |
-
"grad_norm": 20.5,
|
| 973 |
-
"learning_rate": 0.00010165339447663587,
|
| 974 |
-
"loss": 1.9038,
|
| 975 |
-
"step": 1300
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 0.38768866528558743,
|
| 979 |
-
"grad_norm": 2.109375,
|
| 980 |
-
"learning_rate": 0.00010033069336079952,
|
| 981 |
-
"loss": 2.2051,
|
| 982 |
-
"step": 1310
|
| 983 |
-
},
|
| 984 |
-
{
|
| 985 |
-
"epoch": 0.39064812074578276,
|
| 986 |
-
"grad_norm": 4.15625,
|
| 987 |
-
"learning_rate": 9.900793438320037e-05,
|
| 988 |
-
"loss": 2.137,
|
| 989 |
-
"step": 1320
|
| 990 |
-
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 0.3936075762059781,
|
| 993 |
-
"grad_norm": 3.40625,
|
| 994 |
-
"learning_rate": 9.768534898835862e-05,
|
| 995 |
-
"loss": 2.072,
|
| 996 |
-
"step": 1330
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 0.3965670316661734,
|
| 1000 |
-
"grad_norm": 4.125,
|
| 1001 |
-
"learning_rate": 9.636316859042259e-05,
|
| 1002 |
-
"loss": 2.2351,
|
| 1003 |
-
"step": 1340
|
| 1004 |
-
},
|
| 1005 |
-
{
|
| 1006 |
-
"epoch": 0.39952648712636873,
|
| 1007 |
-
"grad_norm": 12.5625,
|
| 1008 |
-
"learning_rate": 9.504162453267777e-05,
|
| 1009 |
-
"loss": 2.0439,
|
| 1010 |
-
"step": 1350
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 0.40248594258656406,
|
| 1014 |
-
"grad_norm": 2.875,
|
| 1015 |
-
"learning_rate": 9.372094804706867e-05,
|
| 1016 |
-
"loss": 2.5911,
|
| 1017 |
-
"step": 1360
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 0.4054453980467594,
|
| 1021 |
-
"grad_norm": 3.75,
|
| 1022 |
-
"learning_rate": 9.24013702137397e-05,
|
| 1023 |
-
"loss": 2.1565,
|
| 1024 |
-
"step": 1370
|
| 1025 |
-
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 0.4084048535069547,
|
| 1028 |
-
"grad_norm": 4.0,
|
| 1029 |
-
"learning_rate": 9.108312192060298e-05,
|
| 1030 |
-
"loss": 2.2613,
|
| 1031 |
-
"step": 1380
|
| 1032 |
-
},
|
| 1033 |
-
{
|
| 1034 |
-
"epoch": 0.41136430896715004,
|
| 1035 |
-
"grad_norm": 4.875,
|
| 1036 |
-
"learning_rate": 8.97664338229395e-05,
|
| 1037 |
-
"loss": 1.8462,
|
| 1038 |
-
"step": 1390
|
| 1039 |
-
},
|
| 1040 |
-
{
|
| 1041 |
-
"epoch": 0.41432376442734536,
|
| 1042 |
-
"grad_norm": 24.375,
|
| 1043 |
-
"learning_rate": 8.845153630304139e-05,
|
| 1044 |
-
"loss": 2.1576,
|
| 1045 |
-
"step": 1400
|
| 1046 |
-
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 0.41432376442734536,
|
| 1049 |
-
"eval_loss": 2.132373571395874,
|
| 1050 |
-
"eval_runtime": 38.4807,
|
| 1051 |
-
"eval_samples_per_second": 39.033,
|
| 1052 |
-
"eval_steps_per_second": 9.771,
|
| 1053 |
-
"step": 1400
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 0.4172832198875407,
|
| 1057 |
-
"grad_norm": 2.046875,
|
| 1058 |
-
"learning_rate": 8.713865942990141e-05,
|
| 1059 |
-
"loss": 2.4016,
|
| 1060 |
-
"step": 1410
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 0.420242675347736,
|
| 1064 |
-
"grad_norm": 3.515625,
|
| 1065 |
-
"learning_rate": 8.582803291895758e-05,
|
| 1066 |
-
"loss": 2.3257,
|
| 1067 |
-
"step": 1420
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 0.42320213080793134,
|
| 1071 |
-
"grad_norm": 6.3125,
|
| 1072 |
-
"learning_rate": 8.451988609189987e-05,
|
| 1073 |
-
"loss": 2.0979,
|
| 1074 |
-
"step": 1430
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 0.42616158626812667,
|
| 1078 |
-
"grad_norm": 4.25,
|
| 1079 |
-
"learning_rate": 8.321444783654524e-05,
|
| 1080 |
-
"loss": 1.8707,
|
| 1081 |
-
"step": 1440
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 0.429121041728322,
|
| 1085 |
-
"grad_norm": 13.4375,
|
| 1086 |
-
"learning_rate": 8.191194656678904e-05,
|
| 1087 |
-
"loss": 2.05,
|
| 1088 |
-
"step": 1450
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.4320804971885173,
|
| 1092 |
-
"grad_norm": 1.8671875,
|
| 1093 |
-
"learning_rate": 8.061261018263919e-05,
|
| 1094 |
-
"loss": 2.3737,
|
| 1095 |
-
"step": 1460
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 0.43503995264871265,
|
| 1099 |
-
"grad_norm": 3.015625,
|
| 1100 |
-
"learning_rate": 7.931666603034033e-05,
|
| 1101 |
-
"loss": 2.3148,
|
| 1102 |
-
"step": 1470
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 0.437999408108908,
|
| 1106 |
-
"grad_norm": 3.296875,
|
| 1107 |
-
"learning_rate": 7.80243408625947e-05,
|
| 1108 |
-
"loss": 2.3303,
|
| 1109 |
-
"step": 1480
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 0.4409588635691033,
|
| 1113 |
-
"grad_norm": 4.15625,
|
| 1114 |
-
"learning_rate": 7.673586079888698e-05,
|
| 1115 |
-
"loss": 2.2395,
|
| 1116 |
-
"step": 1490
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 0.4439183190292986,
|
| 1120 |
-
"grad_norm": 18.375,
|
| 1121 |
-
"learning_rate": 7.54514512859201e-05,
|
| 1122 |
-
"loss": 2.1118,
|
| 1123 |
-
"step": 1500
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 0.44687777448949395,
|
| 1127 |
-
"grad_norm": 1.78125,
|
| 1128 |
-
"learning_rate": 7.417133705816837e-05,
|
| 1129 |
-
"loss": 2.2514,
|
| 1130 |
-
"step": 1510
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.4498372299496893,
|
| 1134 |
-
"grad_norm": 2.359375,
|
| 1135 |
-
"learning_rate": 7.289574209855559e-05,
|
| 1136 |
-
"loss": 2.3813,
|
| 1137 |
-
"step": 1520
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 0.4527966854098846,
|
| 1141 |
-
"grad_norm": 3.25,
|
| 1142 |
-
"learning_rate": 7.16248895992645e-05,
|
| 1143 |
-
"loss": 2.0406,
|
| 1144 |
-
"step": 1530
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 0.45575614087007993,
|
| 1148 |
-
"grad_norm": 6.15625,
|
| 1149 |
-
"learning_rate": 7.035900192268464e-05,
|
| 1150 |
-
"loss": 1.8527,
|
| 1151 |
-
"step": 1540
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 0.45871559633027525,
|
| 1155 |
-
"grad_norm": 24.125,
|
| 1156 |
-
"learning_rate": 6.909830056250527e-05,
|
| 1157 |
-
"loss": 1.8146,
|
| 1158 |
-
"step": 1550
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 0.4616750517904706,
|
| 1162 |
-
"grad_norm": 1.9453125,
|
| 1163 |
-
"learning_rate": 6.784300610496048e-05,
|
| 1164 |
-
"loss": 2.4665,
|
| 1165 |
-
"step": 1560
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 0.46463450725066585,
|
| 1169 |
-
"grad_norm": 2.859375,
|
| 1170 |
-
"learning_rate": 6.65933381902329e-05,
|
| 1171 |
-
"loss": 2.2625,
|
| 1172 |
-
"step": 1570
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.4675939627108612,
|
| 1176 |
-
"grad_norm": 3.609375,
|
| 1177 |
-
"learning_rate": 6.534951547402322e-05,
|
| 1178 |
-
"loss": 2.2715,
|
| 1179 |
-
"step": 1580
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.4705534181710565,
|
| 1183 |
-
"grad_norm": 4.5625,
|
| 1184 |
-
"learning_rate": 6.411175558929152e-05,
|
| 1185 |
-
"loss": 1.9711,
|
| 1186 |
-
"step": 1590
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 0.47351287363125183,
|
| 1190 |
-
"grad_norm": 11.125,
|
| 1191 |
-
"learning_rate": 6.28802751081779e-05,
|
| 1192 |
-
"loss": 2.1731,
|
| 1193 |
-
"step": 1600
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 0.47351287363125183,
|
| 1197 |
-
"eval_loss": 2.12680983543396,
|
| 1198 |
-
"eval_runtime": 39.4155,
|
| 1199 |
-
"eval_samples_per_second": 38.107,
|
| 1200 |
-
"eval_steps_per_second": 9.539,
|
| 1201 |
-
"step": 1600
|
| 1202 |
-
},
|
| 1203 |
-
{
|
| 1204 |
-
"epoch": 0.47647232909144716,
|
| 1205 |
-
"grad_norm": 1.96875,
|
| 1206 |
-
"learning_rate": 6.165528950410884e-05,
|
| 1207 |
-
"loss": 2.1656,
|
| 1208 |
-
"step": 1610
|
| 1209 |
-
},
|
| 1210 |
-
{
|
| 1211 |
-
"epoch": 0.4794317845516425,
|
| 1212 |
-
"grad_norm": 3.1875,
|
| 1213 |
-
"learning_rate": 6.0437013114095195e-05,
|
| 1214 |
-
"loss": 2.2137,
|
| 1215 |
-
"step": 1620
|
| 1216 |
-
},
|
| 1217 |
-
{
|
| 1218 |
-
"epoch": 0.4823912400118378,
|
| 1219 |
-
"grad_norm": 3.71875,
|
| 1220 |
-
"learning_rate": 5.922565910122967e-05,
|
| 1221 |
-
"loss": 2.253,
|
| 1222 |
-
"step": 1630
|
| 1223 |
-
},
|
| 1224 |
-
{
|
| 1225 |
-
"epoch": 0.48535069547203313,
|
| 1226 |
-
"grad_norm": 7.21875,
|
| 1227 |
-
"learning_rate": 5.8021439417389444e-05,
|
| 1228 |
-
"loss": 2.077,
|
| 1229 |
-
"step": 1640
|
| 1230 |
-
},
|
| 1231 |
-
{
|
| 1232 |
-
"epoch": 0.48831015093222846,
|
| 1233 |
-
"grad_norm": 18.375,
|
| 1234 |
-
"learning_rate": 5.6824564766150726e-05,
|
| 1235 |
-
"loss": 1.8215,
|
| 1236 |
-
"step": 1650
|
| 1237 |
-
},
|
| 1238 |
-
{
|
| 1239 |
-
"epoch": 0.4912696063924238,
|
| 1240 |
-
"grad_norm": 1.953125,
|
| 1241 |
-
"learning_rate": 5.563524456592163e-05,
|
| 1242 |
-
"loss": 2.2989,
|
| 1243 |
-
"step": 1660
|
| 1244 |
-
},
|
| 1245 |
-
{
|
| 1246 |
-
"epoch": 0.4942290618526191,
|
| 1247 |
-
"grad_norm": 3.0625,
|
| 1248 |
-
"learning_rate": 5.4453686913300074e-05,
|
| 1249 |
-
"loss": 2.1334,
|
| 1250 |
-
"step": 1670
|
| 1251 |
-
},
|
| 1252 |
-
{
|
| 1253 |
-
"epoch": 0.49718851731281444,
|
| 1254 |
-
"grad_norm": 4.125,
|
| 1255 |
-
"learning_rate": 5.328009854666303e-05,
|
| 1256 |
-
"loss": 2.3074,
|
| 1257 |
-
"step": 1680
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 0.5001479727730098,
|
| 1261 |
-
"grad_norm": 4.3125,
|
| 1262 |
-
"learning_rate": 5.2114684809993044e-05,
|
| 1263 |
-
"loss": 2.1466,
|
| 1264 |
-
"step": 1690
|
| 1265 |
-
},
|
| 1266 |
-
{
|
| 1267 |
-
"epoch": 0.5031074282332051,
|
| 1268 |
-
"grad_norm": 21.375,
|
| 1269 |
-
"learning_rate": 5.095764961694922e-05,
|
| 1270 |
-
"loss": 1.9702,
|
| 1271 |
-
"step": 1700
|
| 1272 |
-
},
|
| 1273 |
-
{
|
| 1274 |
-
"epoch": 0.5060668836934004,
|
| 1275 |
-
"grad_norm": 2.671875,
|
| 1276 |
-
"learning_rate": 4.980919541518796e-05,
|
| 1277 |
-
"loss": 2.3311,
|
| 1278 |
-
"step": 1710
|
| 1279 |
-
},
|
| 1280 |
-
{
|
| 1281 |
-
"epoch": 0.5090263391535957,
|
| 1282 |
-
"grad_norm": 3.546875,
|
| 1283 |
-
"learning_rate": 4.866952315094088e-05,
|
| 1284 |
-
"loss": 2.1625,
|
| 1285 |
-
"step": 1720
|
| 1286 |
-
},
|
| 1287 |
-
{
|
| 1288 |
-
"epoch": 0.511985794613791,
|
| 1289 |
-
"grad_norm": 4.9375,
|
| 1290 |
-
"learning_rate": 4.753883223385467e-05,
|
| 1291 |
-
"loss": 2.1047,
|
| 1292 |
-
"step": 1730
|
| 1293 |
-
},
|
| 1294 |
-
{
|
| 1295 |
-
"epoch": 0.5149452500739864,
|
| 1296 |
-
"grad_norm": 5.40625,
|
| 1297 |
-
"learning_rate": 4.6417320502100316e-05,
|
| 1298 |
-
"loss": 1.9118,
|
| 1299 |
-
"step": 1740
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 0.5179047055341817,
|
| 1303 |
-
"grad_norm": 22.5,
|
| 1304 |
-
"learning_rate": 4.530518418775733e-05,
|
| 1305 |
-
"loss": 2.3183,
|
| 1306 |
-
"step": 1750
|
| 1307 |
-
},
|
| 1308 |
-
{
|
| 1309 |
-
"epoch": 0.520864160994377,
|
| 1310 |
-
"grad_norm": 2.09375,
|
| 1311 |
-
"learning_rate": 4.4202617882478405e-05,
|
| 1312 |
-
"loss": 2.3082,
|
| 1313 |
-
"step": 1760
|
| 1314 |
-
},
|
| 1315 |
-
{
|
| 1316 |
-
"epoch": 0.5238236164545723,
|
| 1317 |
-
"grad_norm": 2.46875,
|
| 1318 |
-
"learning_rate": 4.310981450344189e-05,
|
| 1319 |
-
"loss": 2.1393,
|
| 1320 |
-
"step": 1770
|
| 1321 |
-
},
|
| 1322 |
-
{
|
| 1323 |
-
"epoch": 0.5267830719147677,
|
| 1324 |
-
"grad_norm": 3.109375,
|
| 1325 |
-
"learning_rate": 4.2026965259596666e-05,
|
| 1326 |
-
"loss": 2.0223,
|
| 1327 |
-
"step": 1780
|
| 1328 |
-
},
|
| 1329 |
-
{
|
| 1330 |
-
"epoch": 0.529742527374963,
|
| 1331 |
-
"grad_norm": 7.875,
|
| 1332 |
-
"learning_rate": 4.0954259618206295e-05,
|
| 1333 |
-
"loss": 1.9122,
|
| 1334 |
-
"step": 1790
|
| 1335 |
-
},
|
| 1336 |
-
{
|
| 1337 |
-
"epoch": 0.5327019828351583,
|
| 1338 |
-
"grad_norm": 38.25,
|
| 1339 |
-
"learning_rate": 3.9891885271697496e-05,
|
| 1340 |
-
"loss": 2.149,
|
| 1341 |
-
"step": 1800
|
| 1342 |
-
},
|
| 1343 |
-
{
|
| 1344 |
-
"epoch": 0.5327019828351583,
|
| 1345 |
-
"eval_loss": 2.1312239170074463,
|
| 1346 |
-
"eval_runtime": 37.8217,
|
| 1347 |
-
"eval_samples_per_second": 39.713,
|
| 1348 |
-
"eval_steps_per_second": 9.941,
|
| 1349 |
-
"step": 1800
|
| 1350 |
-
},
|
| 1351 |
-
{
|
| 1352 |
-
"epoch": 0.5356614382953536,
|
| 1353 |
-
"grad_norm": 1.7890625,
|
| 1354 |
-
"learning_rate": 3.884002810481958e-05,
|
| 1355 |
-
"loss": 2.3824,
|
| 1356 |
-
"step": 1810
|
| 1357 |
-
},
|
| 1358 |
-
{
|
| 1359 |
-
"epoch": 0.538620893755549,
|
| 1360 |
-
"grad_norm": 2.859375,
|
| 1361 |
-
"learning_rate": 3.779887216211995e-05,
|
| 1362 |
-
"loss": 2.1831,
|
| 1363 |
-
"step": 1820
|
| 1364 |
-
},
|
| 1365 |
-
{
|
| 1366 |
-
"epoch": 0.5415803492157443,
|
| 1367 |
-
"grad_norm": 4.25,
|
| 1368 |
-
"learning_rate": 3.676859961574162e-05,
|
| 1369 |
-
"loss": 2.0926,
|
| 1370 |
-
"step": 1830
|
| 1371 |
-
},
|
| 1372 |
-
{
|
| 1373 |
-
"epoch": 0.5445398046759397,
|
| 1374 |
-
"grad_norm": 5.4375,
|
| 1375 |
-
"learning_rate": 3.574939073354838e-05,
|
| 1376 |
-
"loss": 2.107,
|
| 1377 |
-
"step": 1840
|
| 1378 |
-
},
|
| 1379 |
-
{
|
| 1380 |
-
"epoch": 0.5474992601361349,
|
| 1381 |
-
"grad_norm": 22.75,
|
| 1382 |
-
"learning_rate": 3.4741423847583134e-05,
|
| 1383 |
-
"loss": 2.1866,
|
| 1384 |
-
"step": 1850
|
| 1385 |
-
},
|
| 1386 |
-
{
|
| 1387 |
-
"epoch": 0.5504587155963303,
|
| 1388 |
-
"grad_norm": 2.046875,
|
| 1389 |
-
"learning_rate": 3.3744875322865034e-05,
|
| 1390 |
-
"loss": 2.3345,
|
| 1391 |
-
"step": 1860
|
| 1392 |
-
},
|
| 1393 |
-
{
|
| 1394 |
-
"epoch": 0.5534181710565256,
|
| 1395 |
-
"grad_norm": 2.734375,
|
| 1396 |
-
"learning_rate": 3.275991952653054e-05,
|
| 1397 |
-
"loss": 2.2518,
|
| 1398 |
-
"step": 1870
|
| 1399 |
-
},
|
| 1400 |
-
{
|
| 1401 |
-
"epoch": 0.556377626516721,
|
| 1402 |
-
"grad_norm": 6.0,
|
| 1403 |
-
"learning_rate": 3.178672879732435e-05,
|
| 1404 |
-
"loss": 2.0092,
|
| 1405 |
-
"step": 1880
|
| 1406 |
-
},
|
| 1407 |
-
{
|
| 1408 |
-
"epoch": 0.5593370819769162,
|
| 1409 |
-
"grad_norm": 7.0625,
|
| 1410 |
-
"learning_rate": 3.0825473415445074e-05,
|
| 1411 |
-
"loss": 1.7693,
|
| 1412 |
-
"step": 1890
|
| 1413 |
-
},
|
| 1414 |
-
{
|
| 1415 |
-
"epoch": 0.5622965374371116,
|
| 1416 |
-
"grad_norm": 34.25,
|
| 1417 |
-
"learning_rate": 2.9876321572751144e-05,
|
| 1418 |
-
"loss": 2.3584,
|
| 1419 |
-
"step": 1900
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 0.5652559928973069,
|
| 1423 |
-
"grad_norm": 1.8828125,
|
| 1424 |
-
"learning_rate": 2.8939439343332086e-05,
|
| 1425 |
-
"loss": 2.4875,
|
| 1426 |
-
"step": 1910
|
| 1427 |
-
},
|
| 1428 |
-
{
|
| 1429 |
-
"epoch": 0.5682154483575023,
|
| 1430 |
-
"grad_norm": 2.625,
|
| 1431 |
-
"learning_rate": 2.8014990654450325e-05,
|
| 1432 |
-
"loss": 2.3285,
|
| 1433 |
-
"step": 1920
|
| 1434 |
-
},
|
| 1435 |
-
{
|
| 1436 |
-
"epoch": 0.5711749038176975,
|
| 1437 |
-
"grad_norm": 3.5,
|
| 1438 |
-
"learning_rate": 2.7103137257858868e-05,
|
| 1439 |
-
"loss": 2.2535,
|
| 1440 |
-
"step": 1930
|
| 1441 |
-
},
|
| 1442 |
-
{
|
| 1443 |
-
"epoch": 0.5741343592778929,
|
| 1444 |
-
"grad_norm": 5.875,
|
| 1445 |
-
"learning_rate": 2.6204038701499056e-05,
|
| 1446 |
-
"loss": 2.0563,
|
| 1447 |
-
"step": 1940
|
| 1448 |
-
},
|
| 1449 |
-
{
|
| 1450 |
-
"epoch": 0.5770938147380882,
|
| 1451 |
-
"grad_norm": 16.5,
|
| 1452 |
-
"learning_rate": 2.5317852301584643e-05,
|
| 1453 |
-
"loss": 1.8811,
|
| 1454 |
-
"step": 1950
|
| 1455 |
-
},
|
| 1456 |
-
{
|
| 1457 |
-
"epoch": 0.5800532701982836,
|
| 1458 |
-
"grad_norm": 1.796875,
|
| 1459 |
-
"learning_rate": 2.4444733115075823e-05,
|
| 1460 |
-
"loss": 2.3993,
|
| 1461 |
-
"step": 1960
|
| 1462 |
-
},
|
| 1463 |
-
{
|
| 1464 |
-
"epoch": 0.5830127256584788,
|
| 1465 |
-
"grad_norm": 2.359375,
|
| 1466 |
-
"learning_rate": 2.3584833912548888e-05,
|
| 1467 |
-
"loss": 2.1668,
|
| 1468 |
-
"step": 1970
|
| 1469 |
-
},
|
| 1470 |
-
{
|
| 1471 |
-
"epoch": 0.5859721811186742,
|
| 1472 |
-
"grad_norm": 3.71875,
|
| 1473 |
-
"learning_rate": 2.2738305151465645e-05,
|
| 1474 |
-
"loss": 2.1702,
|
| 1475 |
-
"step": 1980
|
| 1476 |
-
},
|
| 1477 |
-
{
|
| 1478 |
-
"epoch": 0.5889316365788695,
|
| 1479 |
-
"grad_norm": 6.53125,
|
| 1480 |
-
"learning_rate": 2.190529494984782e-05,
|
| 1481 |
-
"loss": 2.2042,
|
| 1482 |
-
"step": 1990
|
| 1483 |
-
},
|
| 1484 |
-
{
|
| 1485 |
-
"epoch": 0.5918910920390648,
|
| 1486 |
-
"grad_norm": 18.375,
|
| 1487 |
-
"learning_rate": 2.1085949060360654e-05,
|
| 1488 |
-
"loss": 2.7263,
|
| 1489 |
-
"step": 2000
|
| 1490 |
-
},
|
| 1491 |
-
{
|
| 1492 |
-
"epoch": 0.5918910920390648,
|
| 1493 |
-
"eval_loss": 2.1315317153930664,
|
| 1494 |
-
"eval_runtime": 38.3028,
|
| 1495 |
-
"eval_samples_per_second": 39.214,
|
| 1496 |
-
"eval_steps_per_second": 9.817,
|
| 1497 |
-
"step": 2000
|
| 1498 |
}
|
| 1499 |
],
|
| 1500 |
"logging_steps": 10,
|
|
@@ -1514,7 +330,7 @@
|
|
| 1514 |
"attributes": {}
|
| 1515 |
}
|
| 1516 |
},
|
| 1517 |
-
"total_flos":
|
| 1518 |
"train_batch_size": 4,
|
| 1519 |
"trial_name": null,
|
| 1520 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.11837821840781296,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.0002959455460195324,
|
| 13 |
"eval_loss": 3.0371296405792236,
|
| 14 |
+
"eval_runtime": 24.5776,
|
| 15 |
+
"eval_samples_per_second": 61.112,
|
| 16 |
+
"eval_steps_per_second": 15.298,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.002959455460195324,
|
| 21 |
"grad_norm": 2.84375,
|
| 22 |
"learning_rate": 1.6000000000000003e-05,
|
| 23 |
+
"loss": 2.5729,
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.005918910920390648,
|
| 28 |
+
"grad_norm": 4.4375,
|
| 29 |
"learning_rate": 3.2000000000000005e-05,
|
| 30 |
"loss": 2.6895,
|
| 31 |
"step": 20
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.008878366380585973,
|
| 35 |
+
"grad_norm": 5.3125,
|
| 36 |
"learning_rate": 4.8e-05,
|
| 37 |
+
"loss": 2.5333,
|
| 38 |
"step": 30
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.011837821840781295,
|
| 42 |
+
"grad_norm": 7.9375,
|
| 43 |
"learning_rate": 6.400000000000001e-05,
|
| 44 |
+
"loss": 2.9663,
|
| 45 |
"step": 40
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.01479727730097662,
|
| 49 |
"grad_norm": 25.375,
|
| 50 |
"learning_rate": 8e-05,
|
| 51 |
+
"loss": 3.4106,
|
| 52 |
"step": 50
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.017756732761171946,
|
| 56 |
+
"grad_norm": 3.125,
|
| 57 |
"learning_rate": 9.6e-05,
|
| 58 |
+
"loss": 2.3132,
|
| 59 |
"step": 60
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.020716188221367268,
|
| 63 |
+
"grad_norm": 2.9375,
|
| 64 |
"learning_rate": 0.00011200000000000001,
|
| 65 |
+
"loss": 2.3039,
|
| 66 |
"step": 70
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.02367564368156259,
|
| 70 |
"grad_norm": 4.34375,
|
| 71 |
"learning_rate": 0.00012800000000000002,
|
| 72 |
+
"loss": 2.4118,
|
| 73 |
"step": 80
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.026635099141757917,
|
| 77 |
"grad_norm": 5.5,
|
| 78 |
"learning_rate": 0.000144,
|
| 79 |
+
"loss": 2.3272,
|
| 80 |
"step": 90
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.02959455460195324,
|
| 84 |
+
"grad_norm": 22.0,
|
| 85 |
"learning_rate": 0.00016,
|
| 86 |
+
"loss": 3.3301,
|
| 87 |
"step": 100
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.032554010062148565,
|
| 91 |
+
"grad_norm": 3.890625,
|
| 92 |
"learning_rate": 0.00017600000000000002,
|
| 93 |
+
"loss": 2.4817,
|
| 94 |
"step": 110
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.03551346552234389,
|
| 98 |
+
"grad_norm": 2.921875,
|
| 99 |
"learning_rate": 0.000192,
|
| 100 |
+
"loss": 2.2621,
|
| 101 |
"step": 120
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.03847292098253921,
|
| 105 |
+
"grad_norm": 2.859375,
|
| 106 |
"learning_rate": 0.0001999978128380225,
|
| 107 |
+
"loss": 2.3509,
|
| 108 |
"step": 130
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.041432376442734536,
|
| 112 |
+
"grad_norm": 7.125,
|
| 113 |
"learning_rate": 0.0001999803161162393,
|
| 114 |
"loss": 2.3626,
|
| 115 |
"step": 140
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.04439183190292986,
|
| 119 |
+
"grad_norm": 32.5,
|
| 120 |
"learning_rate": 0.00019994532573409262,
|
| 121 |
+
"loss": 2.2664,
|
| 122 |
"step": 150
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.04735128736312518,
|
| 126 |
"grad_norm": 5.6875,
|
| 127 |
"learning_rate": 0.00019989284781388617,
|
| 128 |
+
"loss": 2.378,
|
| 129 |
"step": 160
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.05031074282332051,
|
| 133 |
+
"grad_norm": 2.609375,
|
| 134 |
"learning_rate": 0.00019982289153773646,
|
| 135 |
+
"loss": 2.3008,
|
| 136 |
"step": 170
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.053270198283515834,
|
| 140 |
+
"grad_norm": 4.40625,
|
| 141 |
"learning_rate": 0.00019973546914596623,
|
| 142 |
+
"loss": 2.4329,
|
| 143 |
"step": 180
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.05622965374371116,
|
| 147 |
"grad_norm": 6.90625,
|
| 148 |
"learning_rate": 0.00019963059593496268,
|
| 149 |
+
"loss": 2.2273,
|
| 150 |
"step": 190
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.05918910920390648,
|
| 154 |
+
"grad_norm": 26.125,
|
| 155 |
"learning_rate": 0.00019950829025450114,
|
| 156 |
+
"loss": 2.5059,
|
| 157 |
"step": 200
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.05918910920390648,
|
| 161 |
+
"eval_loss": 2.298793077468872,
|
| 162 |
+
"eval_runtime": 23.8828,
|
| 163 |
+
"eval_samples_per_second": 62.891,
|
| 164 |
+
"eval_steps_per_second": 15.744,
|
| 165 |
"step": 200
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.062148564664101805,
|
| 169 |
+
"grad_norm": 2.53125,
|
| 170 |
"learning_rate": 0.0001993685735045343,
|
| 171 |
+
"loss": 2.4118,
|
| 172 |
"step": 210
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.06510802012429713,
|
| 176 |
"grad_norm": 2.890625,
|
| 177 |
"learning_rate": 0.0001992114701314478,
|
| 178 |
+
"loss": 2.3572,
|
| 179 |
"step": 220
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.06806747558449246,
|
| 183 |
+
"grad_norm": 3.046875,
|
| 184 |
"learning_rate": 0.000199037007623783,
|
| 185 |
+
"loss": 2.2494,
|
| 186 |
"step": 230
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 0.07102693104468778,
|
| 190 |
"grad_norm": 6.65625,
|
| 191 |
"learning_rate": 0.00019884521650742715,
|
| 192 |
+
"loss": 2.2517,
|
| 193 |
"step": 240
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.0739863865048831,
|
| 197 |
+
"grad_norm": 39.75,
|
| 198 |
"learning_rate": 0.00019863613034027224,
|
| 199 |
+
"loss": 2.7168,
|
| 200 |
"step": 250
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.07694584196507842,
|
| 204 |
+
"grad_norm": 5.09375,
|
| 205 |
"learning_rate": 0.0001984097857063434,
|
| 206 |
+
"loss": 2.1898,
|
| 207 |
"step": 260
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 0.07990529742527375,
|
| 211 |
+
"grad_norm": 2.75,
|
| 212 |
"learning_rate": 0.0001981662222093976,
|
| 213 |
+
"loss": 2.409,
|
| 214 |
"step": 270
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 0.08286475288546907,
|
| 218 |
+
"grad_norm": 3.15625,
|
| 219 |
"learning_rate": 0.00019790548246599447,
|
| 220 |
+
"loss": 2.2657,
|
| 221 |
"step": 280
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 0.0858242083456644,
|
| 225 |
+
"grad_norm": 5.375,
|
| 226 |
"learning_rate": 0.00019762761209803927,
|
| 227 |
+
"loss": 2.0511,
|
| 228 |
"step": 290
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 0.08878366380585972,
|
| 232 |
+
"grad_norm": 15.3125,
|
| 233 |
"learning_rate": 0.0001973326597248006,
|
| 234 |
+
"loss": 2.2033,
|
| 235 |
"step": 300
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 0.09174311926605505,
|
| 239 |
"grad_norm": 3.234375,
|
| 240 |
"learning_rate": 0.00019702067695440332,
|
| 241 |
+
"loss": 2.4281,
|
| 242 |
"step": 310
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.09470257472625036,
|
| 246 |
+
"grad_norm": 2.890625,
|
| 247 |
"learning_rate": 0.00019669171837479873,
|
| 248 |
+
"loss": 2.0234,
|
| 249 |
"step": 320
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 0.09766203018644569,
|
| 253 |
"grad_norm": 3.0,
|
| 254 |
"learning_rate": 0.00019634584154421317,
|
| 255 |
+
"loss": 2.2922,
|
| 256 |
"step": 330
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 0.10062148564664102,
|
| 260 |
+
"grad_norm": 5.5,
|
| 261 |
"learning_rate": 0.00019598310698107702,
|
| 262 |
+
"loss": 2.1127,
|
| 263 |
"step": 340
|
| 264 |
},
|
| 265 |
{
|
| 266 |
"epoch": 0.10358094110683634,
|
| 267 |
+
"grad_norm": 23.125,
|
| 268 |
"learning_rate": 0.00019560357815343577,
|
| 269 |
+
"loss": 1.9014,
|
| 270 |
"step": 350
|
| 271 |
},
|
| 272 |
{
|
| 273 |
"epoch": 0.10654039656703167,
|
| 274 |
"grad_norm": 2.515625,
|
| 275 |
"learning_rate": 0.00019520732146784491,
|
| 276 |
+
"loss": 2.2815,
|
| 277 |
"step": 360
|
| 278 |
},
|
| 279 |
{
|
| 280 |
"epoch": 0.109499852027227,
|
| 281 |
"grad_norm": 3.09375,
|
| 282 |
"learning_rate": 0.0001947944062577507,
|
| 283 |
+
"loss": 2.0484,
|
| 284 |
"step": 370
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.11245930748742232,
|
| 288 |
"grad_norm": 4.125,
|
| 289 |
"learning_rate": 0.00019436490477135878,
|
| 290 |
+
"loss": 2.2128,
|
| 291 |
"step": 380
|
| 292 |
},
|
| 293 |
{
|
| 294 |
"epoch": 0.11541876294761765,
|
| 295 |
+
"grad_norm": 5.5625,
|
| 296 |
"learning_rate": 0.00019391889215899299,
|
| 297 |
+
"loss": 1.9424,
|
| 298 |
"step": 390
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 0.11837821840781296,
|
| 302 |
+
"grad_norm": 25.5,
|
| 303 |
"learning_rate": 0.0001934564464599461,
|
| 304 |
+
"loss": 1.8624,
|
| 305 |
"step": 400
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 0.11837821840781296,
|
| 309 |
+
"eval_loss": 2.2062976360321045,
|
| 310 |
+
"eval_runtime": 23.7998,
|
| 311 |
+
"eval_samples_per_second": 63.11,
|
| 312 |
+
"eval_steps_per_second": 15.798,
|
| 313 |
"step": 400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
}
|
| 315 |
],
|
| 316 |
"logging_steps": 10,
|
|
|
|
| 330 |
"attributes": {}
|
| 331 |
}
|
| 332 |
},
|
| 333 |
+
"total_flos": 6216909638860800.0,
|
| 334 |
"train_batch_size": 4,
|
| 335 |
"trial_name": null,
|
| 336 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6840
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c005ee5b7b542fe5c31ab145aa5cb83513af587ae4027e5fb31ac777061c2a2
|
| 3 |
size 6840
|