Zheng Li
commited on
End of training
Browse files- README.md +1 -0
- all_results.json +10 -10
- eval_results.json +5 -5
- runs/Apr08_16-30-28_cs-Precision-7960-Tower/events.out.tfevents.1744144935.cs-Precision-7960-Tower.26085.1 +3 -0
- train_results.json +5 -5
- trainer_state.json +592 -592
README.md
CHANGED
|
@@ -3,6 +3,7 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: facebook/wav2vec2-base
|
| 5 |
tags:
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- superb
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: facebook/wav2vec2-base
|
| 5 |
tags:
|
| 6 |
+
- audio-classification
|
| 7 |
- generated_from_trainer
|
| 8 |
datasets:
|
| 9 |
- superb
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.989355040701315,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos": 2.
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 3.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.989355040701315,
|
| 3 |
+
"eval_accuracy": 0.9814651368049426,
|
| 4 |
+
"eval_loss": 0.09822726994752884,
|
| 5 |
+
"eval_runtime": 5.0668,
|
| 6 |
+
"eval_samples_per_second": 1341.675,
|
| 7 |
+
"eval_steps_per_second": 42.038,
|
| 8 |
+
"total_flos": 2.357895379209216e+18,
|
| 9 |
+
"train_loss": 0.5734889231528854,
|
| 10 |
+
"train_runtime": 654.1726,
|
| 11 |
+
"train_samples_per_second": 390.524,
|
| 12 |
+
"train_steps_per_second": 3.05
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.989355040701315,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime":
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.989355040701315,
|
| 3 |
+
"eval_accuracy": 0.9814651368049426,
|
| 4 |
+
"eval_loss": 0.09822726994752884,
|
| 5 |
+
"eval_runtime": 5.0668,
|
| 6 |
+
"eval_samples_per_second": 1341.675,
|
| 7 |
+
"eval_steps_per_second": 42.038
|
| 8 |
}
|
runs/Apr08_16-30-28_cs-Precision-7960-Tower/events.out.tfevents.1744144935.cs-Precision-7960-Tower.26085.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:480df884b4c8da556dd20cd791e8396fab73f9de77d3d0accb84af0b143feecd
|
| 3 |
+
size 411
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.989355040701315,
|
| 3 |
-
"total_flos": 2.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 3.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.989355040701315,
|
| 3 |
+
"total_flos": 2.357895379209216e+18,
|
| 4 |
+
"train_loss": 0.5734889231528854,
|
| 5 |
+
"train_runtime": 654.1726,
|
| 6 |
+
"train_samples_per_second": 390.524,
|
| 7 |
+
"train_steps_per_second": 3.05
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-1995",
|
| 4 |
"epoch": 4.989355040701315,
|
| 5 |
"eval_steps": 500,
|
|
@@ -10,1450 +10,1450 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.025046963055729492,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 1.5e-06,
|
| 15 |
-
"loss":
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.050093926111458985,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 3e-06,
|
| 22 |
-
"loss":
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.07514088916718847,
|
| 27 |
-
"grad_norm":
|
| 28 |
"learning_rate": 4.5e-06,
|
| 29 |
-
"loss":
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.10018785222291797,
|
| 34 |
-
"grad_norm":
|
| 35 |
"learning_rate": 6e-06,
|
| 36 |
-
"loss":
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.12523481527864747,
|
| 41 |
-
"grad_norm":
|
| 42 |
"learning_rate": 7.5e-06,
|
| 43 |
-
"loss":
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.15028177833437695,
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 9e-06,
|
| 50 |
-
"loss":
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.17532874139010646,
|
| 55 |
-
"grad_norm":
|
| 56 |
"learning_rate": 1.05e-05,
|
| 57 |
-
"loss":
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.20037570444583594,
|
| 62 |
-
"grad_norm":
|
| 63 |
"learning_rate": 1.2e-05,
|
| 64 |
-
"loss":
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.22542266750156542,
|
| 69 |
-
"grad_norm":
|
| 70 |
"learning_rate": 1.3500000000000001e-05,
|
| 71 |
-
"loss":
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.25046963055729493,
|
| 76 |
-
"grad_norm":
|
| 77 |
"learning_rate": 1.5e-05,
|
| 78 |
-
"loss":
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.27551659361302444,
|
| 83 |
-
"grad_norm":
|
| 84 |
"learning_rate": 1.65e-05,
|
| 85 |
-
"loss":
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.3005635566687539,
|
| 90 |
-
"grad_norm":
|
| 91 |
"learning_rate": 1.8e-05,
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.3256105197244834,
|
| 97 |
-
"grad_norm":
|
| 98 |
"learning_rate": 1.95e-05,
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.3506574827802129,
|
| 104 |
-
"grad_norm": 1.
|
| 105 |
"learning_rate": 2.1e-05,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.37570444583594237,
|
| 111 |
-
"grad_norm": 1.
|
| 112 |
"learning_rate": 2.25e-05,
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.4007514088916719,
|
| 118 |
-
"grad_norm":
|
| 119 |
"learning_rate": 2.4e-05,
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.4257983719474014,
|
| 125 |
-
"grad_norm":
|
| 126 |
"learning_rate": 2.55e-05,
|
| 127 |
-
"loss": 1.
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.45084533500313084,
|
| 132 |
-
"grad_norm": 2.
|
| 133 |
"learning_rate": 2.7000000000000002e-05,
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.47589229805886035,
|
| 139 |
-
"grad_norm":
|
| 140 |
"learning_rate": 2.8499999999999998e-05,
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.5009392611145899,
|
| 146 |
-
"grad_norm":
|
| 147 |
"learning_rate": 3e-05,
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.5259862241703194,
|
| 153 |
-
"grad_norm":
|
| 154 |
"learning_rate": 2.9832869080779945e-05,
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.5510331872260489,
|
| 160 |
-
"grad_norm":
|
| 161 |
"learning_rate": 2.9665738161559886e-05,
|
| 162 |
-
"loss": 1.
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.5760801502817783,
|
| 167 |
-
"grad_norm": 2.
|
| 168 |
"learning_rate": 2.9498607242339834e-05,
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.6011271133375078,
|
| 174 |
-
"grad_norm":
|
| 175 |
"learning_rate": 2.933147632311978e-05,
|
| 176 |
-
"loss": 1.
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.6261740763932373,
|
| 181 |
-
"grad_norm": 3.
|
| 182 |
"learning_rate": 2.916434540389972e-05,
|
| 183 |
-
"loss":
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.6512210394489668,
|
| 188 |
-
"grad_norm":
|
| 189 |
"learning_rate": 2.8997214484679665e-05,
|
| 190 |
-
"loss":
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.6762680025046963,
|
| 195 |
-
"grad_norm":
|
| 196 |
"learning_rate": 2.8830083565459613e-05,
|
| 197 |
-
"loss":
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.7013149655604258,
|
| 202 |
-
"grad_norm": 3.
|
| 203 |
"learning_rate": 2.8662952646239554e-05,
|
| 204 |
-
"loss":
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.7263619286161553,
|
| 209 |
-
"grad_norm":
|
| 210 |
"learning_rate": 2.84958217270195e-05,
|
| 211 |
-
"loss":
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.7514088916718847,
|
| 216 |
-
"grad_norm":
|
| 217 |
"learning_rate": 2.8328690807799443e-05,
|
| 218 |
-
"loss":
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.7764558547276142,
|
| 223 |
-
"grad_norm":
|
| 224 |
"learning_rate": 2.8161559888579388e-05,
|
| 225 |
-
"loss":
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.8015028177833438,
|
| 230 |
-
"grad_norm":
|
| 231 |
"learning_rate": 2.7994428969359332e-05,
|
| 232 |
-
"loss":
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.8265497808390733,
|
| 237 |
-
"grad_norm":
|
| 238 |
"learning_rate": 2.7827298050139277e-05,
|
| 239 |
-
"loss":
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.8515967438948028,
|
| 244 |
-
"grad_norm":
|
| 245 |
"learning_rate": 2.7660167130919218e-05,
|
| 246 |
-
"loss":
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.8766437069505323,
|
| 251 |
-
"grad_norm":
|
| 252 |
"learning_rate": 2.7493036211699166e-05,
|
| 253 |
-
"loss":
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.9016906700062617,
|
| 258 |
-
"grad_norm":
|
| 259 |
"learning_rate": 2.732590529247911e-05,
|
| 260 |
-
"loss":
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.9267376330619912,
|
| 265 |
-
"grad_norm":
|
| 266 |
"learning_rate": 2.7158774373259055e-05,
|
| 267 |
-
"loss":
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.9517845961177207,
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate": 2.
|
| 274 |
-
"loss":
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.9768315591734502,
|
| 279 |
-
"grad_norm": 2.
|
| 280 |
-
"learning_rate": 2.
|
| 281 |
-
"loss":
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 1.0,
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate": 2.
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
-
"eval_accuracy": 0.
|
| 294 |
-
"eval_loss": 0.
|
| 295 |
-
"eval_runtime": 4.
|
| 296 |
-
"eval_samples_per_second":
|
| 297 |
-
"eval_steps_per_second":
|
| 298 |
"step": 400
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 1.0250469630557295,
|
| 302 |
-
"grad_norm":
|
| 303 |
-
"learning_rate": 2.
|
| 304 |
-
"loss": 0.
|
| 305 |
"step": 410
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 1.050093926111459,
|
| 309 |
-
"grad_norm":
|
| 310 |
-
"learning_rate": 2.
|
| 311 |
-
"loss": 0.
|
| 312 |
"step": 420
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 1.0751408891671885,
|
| 316 |
-
"grad_norm":
|
| 317 |
-
"learning_rate": 2.
|
| 318 |
-
"loss": 0.
|
| 319 |
"step": 430
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 1.100187852222918,
|
| 323 |
-
"grad_norm":
|
| 324 |
-
"learning_rate": 2.
|
| 325 |
-
"loss": 0.
|
| 326 |
"step": 440
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 1.1252348152786475,
|
| 330 |
-
"grad_norm":
|
| 331 |
-
"learning_rate": 2.
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 450
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 1.150281778334377,
|
| 337 |
-
"grad_norm":
|
| 338 |
-
"learning_rate": 2.
|
| 339 |
-
"loss": 0.
|
| 340 |
"step": 460
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 1.1753287413901066,
|
| 344 |
-
"grad_norm":
|
| 345 |
-
"learning_rate": 2.
|
| 346 |
-
"loss": 0.
|
| 347 |
"step": 470
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 1.2003757044458359,
|
| 351 |
-
"grad_norm":
|
| 352 |
-
"learning_rate": 2.
|
| 353 |
-
"loss": 0.
|
| 354 |
"step": 480
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 1.2254226675015654,
|
| 358 |
-
"grad_norm":
|
| 359 |
-
"learning_rate": 2.
|
| 360 |
-
"loss": 0.
|
| 361 |
"step": 490
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 1.2504696305572949,
|
| 365 |
-
"grad_norm":
|
| 366 |
-
"learning_rate": 2.
|
| 367 |
-
"loss": 0.
|
| 368 |
"step": 500
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 1.2755165936130244,
|
| 372 |
-
"grad_norm":
|
| 373 |
-
"learning_rate": 2.
|
| 374 |
-
"loss": 0.
|
| 375 |
"step": 510
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 1.300563556668754,
|
| 379 |
-
"grad_norm":
|
| 380 |
-
"learning_rate": 2.
|
| 381 |
-
"loss": 0.
|
| 382 |
"step": 520
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 1.3256105197244834,
|
| 386 |
-
"grad_norm":
|
| 387 |
-
"learning_rate": 2.
|
| 388 |
-
"loss": 0.
|
| 389 |
"step": 530
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 1.350657482780213,
|
| 393 |
-
"grad_norm":
|
| 394 |
-
"learning_rate": 2.
|
| 395 |
-
"loss": 0.
|
| 396 |
"step": 540
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 1.3757044458359424,
|
| 400 |
-
"grad_norm":
|
| 401 |
-
"learning_rate": 2.
|
| 402 |
-
"loss": 0.
|
| 403 |
"step": 550
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 1.400751408891672,
|
| 407 |
-
"grad_norm":
|
| 408 |
-
"learning_rate": 2.
|
| 409 |
-
"loss": 0.
|
| 410 |
"step": 560
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 1.4257983719474014,
|
| 414 |
-
"grad_norm": 5.
|
| 415 |
-
"learning_rate": 2.
|
| 416 |
-
"loss": 0.
|
| 417 |
"step": 570
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 1.4508453350031307,
|
| 421 |
-
"grad_norm":
|
| 422 |
-
"learning_rate": 2.
|
| 423 |
-
"loss": 0.
|
| 424 |
"step": 580
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 1.4758922980588602,
|
| 428 |
-
"grad_norm":
|
| 429 |
-
"learning_rate": 2.
|
| 430 |
-
"loss": 0.
|
| 431 |
"step": 590
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 1.5009392611145898,
|
| 435 |
-
"grad_norm":
|
| 436 |
-
"learning_rate": 2.
|
| 437 |
-
"loss": 0.
|
| 438 |
"step": 600
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 1.5259862241703193,
|
| 442 |
-
"grad_norm":
|
| 443 |
-
"learning_rate": 2.
|
| 444 |
-
"loss": 0.
|
| 445 |
"step": 610
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 1.5510331872260488,
|
| 449 |
-
"grad_norm":
|
| 450 |
-
"learning_rate": 2.
|
| 451 |
-
"loss": 0.
|
| 452 |
"step": 620
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 1.5760801502817783,
|
| 456 |
-
"grad_norm":
|
| 457 |
-
"learning_rate": 2.
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 630
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 1.6011271133375078,
|
| 463 |
-
"grad_norm":
|
| 464 |
-
"learning_rate": 2.
|
| 465 |
-
"loss": 0.
|
| 466 |
"step": 640
|
| 467 |
},
|
| 468 |
{
|
| 469 |
"epoch": 1.6261740763932373,
|
| 470 |
-
"grad_norm": 2.
|
| 471 |
-
"learning_rate": 2.
|
| 472 |
-
"loss": 0.
|
| 473 |
"step": 650
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 1.6512210394489668,
|
| 477 |
-
"grad_norm":
|
| 478 |
-
"learning_rate": 2.
|
| 479 |
-
"loss": 0.
|
| 480 |
"step": 660
|
| 481 |
},
|
| 482 |
{
|
| 483 |
"epoch": 1.6762680025046963,
|
| 484 |
-
"grad_norm":
|
| 485 |
-
"learning_rate": 2.
|
| 486 |
-
"loss": 0.
|
| 487 |
"step": 670
|
| 488 |
},
|
| 489 |
{
|
| 490 |
"epoch": 1.7013149655604258,
|
| 491 |
-
"grad_norm":
|
| 492 |
-
"learning_rate": 2.
|
| 493 |
-
"loss": 0.
|
| 494 |
"step": 680
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 1.7263619286161553,
|
| 498 |
-
"grad_norm":
|
| 499 |
-
"learning_rate": 2.
|
| 500 |
-
"loss": 0.
|
| 501 |
"step": 690
|
| 502 |
},
|
| 503 |
{
|
| 504 |
"epoch": 1.7514088916718848,
|
| 505 |
-
"grad_norm":
|
| 506 |
-
"learning_rate": 2.
|
| 507 |
-
"loss": 0.
|
| 508 |
"step": 700
|
| 509 |
},
|
| 510 |
{
|
| 511 |
"epoch": 1.7764558547276144,
|
| 512 |
-
"grad_norm": 4.
|
| 513 |
-
"learning_rate": 2.
|
| 514 |
-
"loss": 0.
|
| 515 |
"step": 710
|
| 516 |
},
|
| 517 |
{
|
| 518 |
"epoch": 1.8015028177833439,
|
| 519 |
-
"grad_norm":
|
| 520 |
-
"learning_rate": 2.
|
| 521 |
-
"loss": 0.
|
| 522 |
"step": 720
|
| 523 |
},
|
| 524 |
{
|
| 525 |
"epoch": 1.8265497808390734,
|
| 526 |
-
"grad_norm":
|
| 527 |
-
"learning_rate": 2.
|
| 528 |
-
"loss": 0.
|
| 529 |
"step": 730
|
| 530 |
},
|
| 531 |
{
|
| 532 |
"epoch": 1.8515967438948029,
|
| 533 |
-
"grad_norm":
|
| 534 |
-
"learning_rate": 2.
|
| 535 |
-
"loss": 0.
|
| 536 |
"step": 740
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 1.8766437069505324,
|
| 540 |
-
"grad_norm":
|
| 541 |
-
"learning_rate": 2.
|
| 542 |
-
"loss": 0.
|
| 543 |
"step": 750
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 1.9016906700062617,
|
| 547 |
-
"grad_norm":
|
| 548 |
-
"learning_rate": 2.
|
| 549 |
-
"loss": 0.
|
| 550 |
"step": 760
|
| 551 |
},
|
| 552 |
{
|
| 553 |
"epoch": 1.9267376330619912,
|
| 554 |
-
"grad_norm":
|
| 555 |
-
"learning_rate": 2.
|
| 556 |
-
"loss": 0.
|
| 557 |
"step": 770
|
| 558 |
},
|
| 559 |
{
|
| 560 |
"epoch": 1.9517845961177207,
|
| 561 |
-
"grad_norm":
|
| 562 |
-
"learning_rate": 2.
|
| 563 |
-
"loss": 0.
|
| 564 |
"step": 780
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 1.9768315591734502,
|
| 568 |
-
"grad_norm":
|
| 569 |
-
"learning_rate": 2.
|
| 570 |
-
"loss": 0.
|
| 571 |
"step": 790
|
| 572 |
},
|
| 573 |
{
|
| 574 |
"epoch": 2.0,
|
| 575 |
-
"grad_norm": 0.
|
| 576 |
-
"learning_rate": 1.
|
| 577 |
-
"loss": 0.
|
| 578 |
"step": 800
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 2.0,
|
| 582 |
-
"eval_accuracy": 0.
|
| 583 |
-
"eval_loss": 0.
|
| 584 |
-
"eval_runtime": 4.
|
| 585 |
-
"eval_samples_per_second":
|
| 586 |
-
"eval_steps_per_second": 48.
|
| 587 |
"step": 800
|
| 588 |
},
|
| 589 |
{
|
| 590 |
"epoch": 2.0250469630557295,
|
| 591 |
-
"grad_norm":
|
| 592 |
-
"learning_rate": 1.
|
| 593 |
-
"loss": 0.
|
| 594 |
"step": 810
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 2.050093926111459,
|
| 598 |
-
"grad_norm": 4.
|
| 599 |
-
"learning_rate": 1.
|
| 600 |
-
"loss": 0.
|
| 601 |
"step": 820
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"epoch": 2.0751408891671885,
|
| 605 |
-
"grad_norm":
|
| 606 |
-
"learning_rate": 1.
|
| 607 |
-
"loss": 0.
|
| 608 |
"step": 830
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 2.100187852222918,
|
| 612 |
-
"grad_norm":
|
| 613 |
-
"learning_rate": 1.
|
| 614 |
-
"loss": 0.
|
| 615 |
"step": 840
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 2.1252348152786475,
|
| 619 |
-
"grad_norm":
|
| 620 |
-
"learning_rate": 1.
|
| 621 |
-
"loss": 0.
|
| 622 |
"step": 850
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 2.150281778334377,
|
| 626 |
-
"grad_norm":
|
| 627 |
-
"learning_rate": 1.
|
| 628 |
-
"loss": 0.
|
| 629 |
"step": 860
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 2.1753287413901066,
|
| 633 |
-
"grad_norm": 4.
|
| 634 |
-
"learning_rate": 1.
|
| 635 |
-
"loss": 0.
|
| 636 |
"step": 870
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 2.200375704445836,
|
| 640 |
-
"grad_norm":
|
| 641 |
-
"learning_rate": 1.
|
| 642 |
-
"loss": 0.
|
| 643 |
"step": 880
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 2.2254226675015656,
|
| 647 |
-
"grad_norm": 1.
|
| 648 |
-
"learning_rate": 1.
|
| 649 |
-
"loss": 0.
|
| 650 |
"step": 890
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 2.250469630557295,
|
| 654 |
-
"grad_norm":
|
| 655 |
-
"learning_rate": 1.
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 900
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 2.2755165936130246,
|
| 661 |
-
"grad_norm": 3.
|
| 662 |
-
"learning_rate": 1.
|
| 663 |
-
"loss": 0.
|
| 664 |
"step": 910
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 2.300563556668754,
|
| 668 |
-
"grad_norm":
|
| 669 |
-
"learning_rate": 1.
|
| 670 |
-
"loss": 0.
|
| 671 |
"step": 920
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 2.325610519724483,
|
| 675 |
-
"grad_norm":
|
| 676 |
-
"learning_rate": 1.
|
| 677 |
-
"loss": 0.
|
| 678 |
"step": 930
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 2.350657482780213,
|
| 682 |
-
"grad_norm": 3.
|
| 683 |
-
"learning_rate": 1.
|
| 684 |
-
"loss": 0.
|
| 685 |
"step": 940
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 2.375704445835942,
|
| 689 |
-
"grad_norm": 1.
|
| 690 |
-
"learning_rate": 1.
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 950
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 2.4007514088916717,
|
| 696 |
-
"grad_norm":
|
| 697 |
-
"learning_rate": 1.
|
| 698 |
-
"loss": 0.
|
| 699 |
"step": 960
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 2.425798371947401,
|
| 703 |
-
"grad_norm":
|
| 704 |
-
"learning_rate": 1.
|
| 705 |
-
"loss": 0.
|
| 706 |
"step": 970
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 2.4508453350031307,
|
| 710 |
-
"grad_norm":
|
| 711 |
-
"learning_rate": 1.
|
| 712 |
-
"loss": 0.
|
| 713 |
"step": 980
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 2.4758922980588602,
|
| 717 |
-
"grad_norm":
|
| 718 |
-
"learning_rate": 1.
|
| 719 |
-
"loss": 0.
|
| 720 |
"step": 990
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 2.5009392611145898,
|
| 724 |
-
"grad_norm":
|
| 725 |
-
"learning_rate": 1.
|
| 726 |
-
"loss": 0.
|
| 727 |
"step": 1000
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 2.5259862241703193,
|
| 731 |
-
"grad_norm":
|
| 732 |
-
"learning_rate": 1.
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 1010
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 2.5510331872260488,
|
| 738 |
-
"grad_norm":
|
| 739 |
-
"learning_rate": 1.
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 1020
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 2.5760801502817783,
|
| 745 |
-
"grad_norm":
|
| 746 |
-
"learning_rate": 1.
|
| 747 |
-
"loss": 0.
|
| 748 |
"step": 1030
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 2.601127113337508,
|
| 752 |
-
"grad_norm":
|
| 753 |
-
"learning_rate": 1.
|
| 754 |
-
"loss": 0.
|
| 755 |
"step": 1040
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 2.6261740763932373,
|
| 759 |
-
"grad_norm":
|
| 760 |
-
"learning_rate": 1.
|
| 761 |
-
"loss": 0.
|
| 762 |
"step": 1050
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 2.651221039448967,
|
| 766 |
-
"grad_norm":
|
| 767 |
-
"learning_rate": 1.
|
| 768 |
-
"loss": 0.
|
| 769 |
"step": 1060
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 2.6762680025046963,
|
| 773 |
-
"grad_norm":
|
| 774 |
-
"learning_rate": 1.
|
| 775 |
-
"loss": 0.
|
| 776 |
"step": 1070
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 2.701314965560426,
|
| 780 |
-
"grad_norm": 3.
|
| 781 |
-
"learning_rate": 1.
|
| 782 |
-
"loss": 0.
|
| 783 |
"step": 1080
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 2.7263619286161553,
|
| 787 |
-
"grad_norm":
|
| 788 |
-
"learning_rate": 1.
|
| 789 |
-
"loss": 0.
|
| 790 |
"step": 1090
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 2.751408891671885,
|
| 794 |
-
"grad_norm":
|
| 795 |
-
"learning_rate": 1.
|
| 796 |
-
"loss": 0.
|
| 797 |
"step": 1100
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 2.7764558547276144,
|
| 801 |
-
"grad_norm": 4.
|
| 802 |
-
"learning_rate": 1.
|
| 803 |
-
"loss": 0.
|
| 804 |
"step": 1110
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 2.801502817783344,
|
| 808 |
-
"grad_norm":
|
| 809 |
-
"learning_rate": 1.
|
| 810 |
-
"loss": 0.
|
| 811 |
"step": 1120
|
| 812 |
},
|
| 813 |
{
|
| 814 |
"epoch": 2.8265497808390734,
|
| 815 |
-
"grad_norm":
|
| 816 |
-
"learning_rate": 1.
|
| 817 |
-
"loss": 0.
|
| 818 |
"step": 1130
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 2.851596743894803,
|
| 822 |
-
"grad_norm":
|
| 823 |
-
"learning_rate": 1.
|
| 824 |
-
"loss": 0.
|
| 825 |
"step": 1140
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 2.8766437069505324,
|
| 829 |
-
"grad_norm":
|
| 830 |
-
"learning_rate": 1.
|
| 831 |
-
"loss": 0.
|
| 832 |
"step": 1150
|
| 833 |
},
|
| 834 |
{
|
| 835 |
"epoch": 2.9016906700062615,
|
| 836 |
-
"grad_norm":
|
| 837 |
-
"learning_rate": 1.
|
| 838 |
-
"loss": 0.
|
| 839 |
"step": 1160
|
| 840 |
},
|
| 841 |
{
|
| 842 |
"epoch": 2.9267376330619914,
|
| 843 |
-
"grad_norm": 3.
|
| 844 |
-
"learning_rate": 1.
|
| 845 |
-
"loss": 0.
|
| 846 |
"step": 1170
|
| 847 |
},
|
| 848 |
{
|
| 849 |
"epoch": 2.9517845961177205,
|
| 850 |
-
"grad_norm":
|
| 851 |
-
"learning_rate": 1.
|
| 852 |
-
"loss": 0.
|
| 853 |
"step": 1180
|
| 854 |
},
|
| 855 |
{
|
| 856 |
"epoch": 2.9768315591734504,
|
| 857 |
-
"grad_norm":
|
| 858 |
-
"learning_rate": 1.
|
| 859 |
-
"loss": 0.
|
| 860 |
"step": 1190
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 3.0,
|
| 864 |
-
"grad_norm":
|
| 865 |
-
"learning_rate": 1.
|
| 866 |
-
"loss": 0.
|
| 867 |
"step": 1200
|
| 868 |
},
|
| 869 |
{
|
| 870 |
"epoch": 3.0,
|
| 871 |
-
"eval_accuracy": 0.
|
| 872 |
-
"eval_loss": 0.
|
| 873 |
-
"eval_runtime": 4.
|
| 874 |
-
"eval_samples_per_second":
|
| 875 |
-
"eval_steps_per_second":
|
| 876 |
"step": 1200
|
| 877 |
},
|
| 878 |
{
|
| 879 |
"epoch": 3.0250469630557295,
|
| 880 |
-
"grad_norm":
|
| 881 |
-
"learning_rate": 1.
|
| 882 |
-
"loss": 0.
|
| 883 |
"step": 1210
|
| 884 |
},
|
| 885 |
{
|
| 886 |
"epoch": 3.050093926111459,
|
| 887 |
-
"grad_norm":
|
| 888 |
-
"learning_rate": 1.
|
| 889 |
-
"loss": 0.
|
| 890 |
"step": 1220
|
| 891 |
},
|
| 892 |
{
|
| 893 |
"epoch": 3.0751408891671885,
|
| 894 |
-
"grad_norm":
|
| 895 |
-
"learning_rate": 1.
|
| 896 |
-
"loss": 0.
|
| 897 |
"step": 1230
|
| 898 |
},
|
| 899 |
{
|
| 900 |
"epoch": 3.100187852222918,
|
| 901 |
-
"grad_norm":
|
| 902 |
-
"learning_rate": 1.
|
| 903 |
-
"loss": 0.
|
| 904 |
"step": 1240
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 3.1252348152786475,
|
| 908 |
-
"grad_norm":
|
| 909 |
-
"learning_rate": 1.
|
| 910 |
-
"loss": 0.
|
| 911 |
"step": 1250
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 3.150281778334377,
|
| 915 |
-
"grad_norm":
|
| 916 |
-
"learning_rate": 1.
|
| 917 |
-
"loss": 0.
|
| 918 |
"step": 1260
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 3.1753287413901066,
|
| 922 |
-
"grad_norm":
|
| 923 |
-
"learning_rate": 1.
|
| 924 |
-
"loss": 0.
|
| 925 |
"step": 1270
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 3.200375704445836,
|
| 929 |
-
"grad_norm":
|
| 930 |
-
"learning_rate": 1.
|
| 931 |
-
"loss": 0.
|
| 932 |
"step": 1280
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 3.2254226675015656,
|
| 936 |
-
"grad_norm":
|
| 937 |
-
"learning_rate": 1.
|
| 938 |
-
"loss": 0.
|
| 939 |
"step": 1290
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 3.250469630557295,
|
| 943 |
-
"grad_norm":
|
| 944 |
-
"learning_rate": 1.
|
| 945 |
-
"loss": 0.
|
| 946 |
"step": 1300
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 3.2755165936130246,
|
| 950 |
-
"grad_norm": 2.
|
| 951 |
-
"learning_rate": 1.
|
| 952 |
-
"loss": 0.
|
| 953 |
"step": 1310
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 3.300563556668754,
|
| 957 |
-
"grad_norm":
|
| 958 |
-
"learning_rate": 1.
|
| 959 |
-
"loss": 0.
|
| 960 |
"step": 1320
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 3.325610519724483,
|
| 964 |
-
"grad_norm": 2.
|
| 965 |
-
"learning_rate": 1.
|
| 966 |
-
"loss": 0.
|
| 967 |
"step": 1330
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 3.350657482780213,
|
| 971 |
-
"grad_norm": 2.
|
| 972 |
-
"learning_rate": 1.
|
| 973 |
-
"loss": 0.
|
| 974 |
"step": 1340
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 3.375704445835942,
|
| 978 |
-
"grad_norm": 5.
|
| 979 |
-
"learning_rate": 1.
|
| 980 |
-
"loss": 0.
|
| 981 |
"step": 1350
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 3.4007514088916717,
|
| 985 |
-
"grad_norm": 3.
|
| 986 |
-
"learning_rate": 1.
|
| 987 |
-
"loss": 0.
|
| 988 |
"step": 1360
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 3.425798371947401,
|
| 992 |
-
"grad_norm":
|
| 993 |
-
"learning_rate": 1.
|
| 994 |
-
"loss": 0.
|
| 995 |
"step": 1370
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 3.4508453350031307,
|
| 999 |
-
"grad_norm":
|
| 1000 |
-
"learning_rate": 1.
|
| 1001 |
-
"loss": 0.
|
| 1002 |
"step": 1380
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 3.4758922980588602,
|
| 1006 |
-
"grad_norm":
|
| 1007 |
-
"learning_rate": 1.
|
| 1008 |
-
"loss": 0.
|
| 1009 |
"step": 1390
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 3.5009392611145898,
|
| 1013 |
-
"grad_norm":
|
| 1014 |
-
"learning_rate": 9.
|
| 1015 |
-
"loss": 0.
|
| 1016 |
"step": 1400
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 3.5259862241703193,
|
| 1020 |
-
"grad_norm":
|
| 1021 |
-
"learning_rate": 9.
|
| 1022 |
-
"loss": 0.
|
| 1023 |
"step": 1410
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 3.5510331872260488,
|
| 1027 |
-
"grad_norm": 1.
|
| 1028 |
-
"learning_rate": 9.
|
| 1029 |
-
"loss": 0.
|
| 1030 |
"step": 1420
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 3.5760801502817783,
|
| 1034 |
-
"grad_norm": 3.
|
| 1035 |
-
"learning_rate": 9.
|
| 1036 |
-
"loss": 0.
|
| 1037 |
"step": 1430
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 3.601127113337508,
|
| 1041 |
-
"grad_norm":
|
| 1042 |
-
"learning_rate": 9.
|
| 1043 |
-
"loss": 0.
|
| 1044 |
"step": 1440
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 3.6261740763932373,
|
| 1048 |
-
"grad_norm":
|
| 1049 |
-
"learning_rate": 9.
|
| 1050 |
-
"loss": 0.
|
| 1051 |
"step": 1450
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
"epoch": 3.651221039448967,
|
| 1055 |
-
"grad_norm":
|
| 1056 |
-
"learning_rate": 8.
|
| 1057 |
-
"loss": 0.
|
| 1058 |
"step": 1460
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
"epoch": 3.6762680025046963,
|
| 1062 |
-
"grad_norm": 1.
|
| 1063 |
-
"learning_rate": 8.
|
| 1064 |
-
"loss": 0.
|
| 1065 |
"step": 1470
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 3.701314965560426,
|
| 1069 |
-
"grad_norm":
|
| 1070 |
-
"learning_rate": 8.
|
| 1071 |
-
"loss": 0.
|
| 1072 |
"step": 1480
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
"epoch": 3.7263619286161553,
|
| 1076 |
-
"grad_norm": 3.
|
| 1077 |
-
"learning_rate": 8.
|
| 1078 |
-
"loss": 0.
|
| 1079 |
"step": 1490
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
"epoch": 3.751408891671885,
|
| 1083 |
-
"grad_norm": 2.
|
| 1084 |
-
"learning_rate": 8.
|
| 1085 |
-
"loss": 0.
|
| 1086 |
"step": 1500
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 3.7764558547276144,
|
| 1090 |
-
"grad_norm":
|
| 1091 |
-
"learning_rate": 8.
|
| 1092 |
-
"loss": 0.
|
| 1093 |
"step": 1510
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
"epoch": 3.801502817783344,
|
| 1097 |
-
"grad_norm":
|
| 1098 |
-
"learning_rate": 7.
|
| 1099 |
-
"loss": 0.
|
| 1100 |
"step": 1520
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
"epoch": 3.8265497808390734,
|
| 1104 |
-
"grad_norm":
|
| 1105 |
-
"learning_rate": 7.
|
| 1106 |
-
"loss": 0.
|
| 1107 |
"step": 1530
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
"epoch": 3.851596743894803,
|
| 1111 |
-
"grad_norm":
|
| 1112 |
-
"learning_rate": 7.
|
| 1113 |
-
"loss": 0.
|
| 1114 |
"step": 1540
|
| 1115 |
},
|
| 1116 |
{
|
| 1117 |
"epoch": 3.8766437069505324,
|
| 1118 |
-
"grad_norm": 2.
|
| 1119 |
-
"learning_rate": 7.
|
| 1120 |
-
"loss": 0.
|
| 1121 |
"step": 1550
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
"epoch": 3.9016906700062615,
|
| 1125 |
-
"grad_norm":
|
| 1126 |
-
"learning_rate": 7.
|
| 1127 |
-
"loss": 0.
|
| 1128 |
"step": 1560
|
| 1129 |
},
|
| 1130 |
{
|
| 1131 |
"epoch": 3.9267376330619914,
|
| 1132 |
-
"grad_norm":
|
| 1133 |
-
"learning_rate": 7.
|
| 1134 |
-
"loss": 0.
|
| 1135 |
"step": 1570
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
"epoch": 3.9517845961177205,
|
| 1139 |
-
"grad_norm": 2.
|
| 1140 |
-
"learning_rate": 6.
|
| 1141 |
-
"loss": 0.
|
| 1142 |
"step": 1580
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
"epoch": 3.9768315591734504,
|
| 1146 |
-
"grad_norm":
|
| 1147 |
-
"learning_rate": 6.
|
| 1148 |
-
"loss": 0.
|
| 1149 |
"step": 1590
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
"epoch": 4.0,
|
| 1153 |
-
"grad_norm":
|
| 1154 |
-
"learning_rate": 6.
|
| 1155 |
-
"loss": 0.
|
| 1156 |
"step": 1600
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
"epoch": 4.0,
|
| 1160 |
-
"eval_accuracy": 0.
|
| 1161 |
-
"eval_loss": 0.
|
| 1162 |
-
"eval_runtime": 4.
|
| 1163 |
-
"eval_samples_per_second":
|
| 1164 |
-
"eval_steps_per_second":
|
| 1165 |
"step": 1600
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 4.025046963055729,
|
| 1169 |
-
"grad_norm": 3.
|
| 1170 |
-
"learning_rate": 6.
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 1610
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 4.050093926111459,
|
| 1176 |
-
"grad_norm":
|
| 1177 |
-
"learning_rate": 6.
|
| 1178 |
-
"loss": 0.
|
| 1179 |
"step": 1620
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 4.075140889167188,
|
| 1183 |
-
"grad_norm":
|
| 1184 |
-
"learning_rate": 6.
|
| 1185 |
-
"loss": 0.
|
| 1186 |
"step": 1630
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 4.100187852222918,
|
| 1190 |
-
"grad_norm":
|
| 1191 |
-
"learning_rate": 5.
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 1640
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 4.125234815278647,
|
| 1197 |
-
"grad_norm":
|
| 1198 |
-
"learning_rate": 5.
|
| 1199 |
-
"loss": 0.
|
| 1200 |
"step": 1650
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 4.150281778334377,
|
| 1204 |
-
"grad_norm":
|
| 1205 |
-
"learning_rate": 5.
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 1660
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 4.175328741390106,
|
| 1211 |
-
"grad_norm":
|
| 1212 |
-
"learning_rate": 5.
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 1670
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 4.200375704445836,
|
| 1218 |
-
"grad_norm":
|
| 1219 |
-
"learning_rate": 5.
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 1680
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 4.225422667501565,
|
| 1225 |
-
"grad_norm": 2.
|
| 1226 |
-
"learning_rate": 5.
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 1690
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 4.250469630557295,
|
| 1232 |
-
"grad_norm":
|
| 1233 |
-
"learning_rate": 4.
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 1700
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 4.275516593613024,
|
| 1239 |
-
"grad_norm":
|
| 1240 |
-
"learning_rate": 4.
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 1710
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 4.300563556668754,
|
| 1246 |
-
"grad_norm":
|
| 1247 |
-
"learning_rate": 4.
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 1720
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 4.325610519724483,
|
| 1253 |
-
"grad_norm": 4.
|
| 1254 |
-
"learning_rate": 4.
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 1730
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 4.350657482780213,
|
| 1260 |
-
"grad_norm":
|
| 1261 |
-
"learning_rate": 4.
|
| 1262 |
-
"loss": 0.
|
| 1263 |
"step": 1740
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 4.375704445835942,
|
| 1267 |
-
"grad_norm":
|
| 1268 |
-
"learning_rate": 4.
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 1750
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 4.400751408891672,
|
| 1274 |
-
"grad_norm":
|
| 1275 |
-
"learning_rate": 3.
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 1760
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 4.425798371947401,
|
| 1281 |
-
"grad_norm": 3.
|
| 1282 |
-
"learning_rate": 3.
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 1770
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 4.450845335003131,
|
| 1288 |
-
"grad_norm":
|
| 1289 |
-
"learning_rate": 3.
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 1780
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 4.47589229805886,
|
| 1295 |
-
"grad_norm":
|
| 1296 |
-
"learning_rate": 3.
|
| 1297 |
-
"loss": 0.
|
| 1298 |
"step": 1790
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 4.50093926111459,
|
| 1302 |
-
"grad_norm": 3.
|
| 1303 |
-
"learning_rate": 3.
|
| 1304 |
-
"loss": 0.
|
| 1305 |
"step": 1800
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 4.525986224170319,
|
| 1309 |
-
"grad_norm":
|
| 1310 |
-
"learning_rate": 3.
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 1810
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 4.551033187226049,
|
| 1316 |
-
"grad_norm":
|
| 1317 |
-
"learning_rate": 2.
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 1820
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 4.576080150281778,
|
| 1323 |
-
"grad_norm":
|
| 1324 |
-
"learning_rate": 2.
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 1830
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 4.601127113337508,
|
| 1330 |
-
"grad_norm":
|
| 1331 |
-
"learning_rate": 2.
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 1840
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 4.626174076393237,
|
| 1337 |
-
"grad_norm":
|
| 1338 |
-
"learning_rate": 2.
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 1850
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 4.651221039448966,
|
| 1344 |
-
"grad_norm": 3.
|
| 1345 |
-
"learning_rate": 2.
|
| 1346 |
-
"loss": 0.
|
| 1347 |
"step": 1860
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 4.676268002504696,
|
| 1351 |
-
"grad_norm":
|
| 1352 |
-
"learning_rate": 2.
|
| 1353 |
-
"loss": 0.
|
| 1354 |
"step": 1870
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 4.701314965560426,
|
| 1358 |
-
"grad_norm": 0.
|
| 1359 |
-
"learning_rate": 1.
|
| 1360 |
-
"loss": 0.
|
| 1361 |
"step": 1880
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 4.726361928616155,
|
| 1365 |
-
"grad_norm":
|
| 1366 |
-
"learning_rate": 1.
|
| 1367 |
-
"loss": 0.
|
| 1368 |
"step": 1890
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 4.751408891671884,
|
| 1372 |
-
"grad_norm": 2.
|
| 1373 |
-
"learning_rate": 1.
|
| 1374 |
-
"loss": 0.
|
| 1375 |
"step": 1900
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 4.776455854727614,
|
| 1379 |
-
"grad_norm":
|
| 1380 |
-
"learning_rate": 1.
|
| 1381 |
-
"loss": 0.
|
| 1382 |
"step": 1910
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 4.801502817783343,
|
| 1386 |
-
"grad_norm":
|
| 1387 |
-
"learning_rate": 1.
|
| 1388 |
-
"loss": 0.
|
| 1389 |
"step": 1920
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 4.826549780839073,
|
| 1393 |
-
"grad_norm":
|
| 1394 |
-
"learning_rate": 1.
|
| 1395 |
-
"loss": 0.
|
| 1396 |
"step": 1930
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 4.851596743894802,
|
| 1400 |
-
"grad_norm": 4.
|
| 1401 |
-
"learning_rate": 9.
|
| 1402 |
-
"loss": 0.
|
| 1403 |
"step": 1940
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 4.876643706950532,
|
| 1407 |
-
"grad_norm":
|
| 1408 |
-
"learning_rate": 7.
|
| 1409 |
-
"loss": 0.
|
| 1410 |
"step": 1950
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 4.9016906700062615,
|
| 1414 |
-
"grad_norm":
|
| 1415 |
-
"learning_rate":
|
| 1416 |
-
"loss": 0.
|
| 1417 |
"step": 1960
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 4.926737633061991,
|
| 1421 |
-
"grad_norm": 1.
|
| 1422 |
-
"learning_rate": 4.
|
| 1423 |
-
"loss": 0.
|
| 1424 |
"step": 1970
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
"epoch": 4.9517845961177205,
|
| 1428 |
-
"grad_norm":
|
| 1429 |
-
"learning_rate": 2.
|
| 1430 |
-
"loss": 0.
|
| 1431 |
"step": 1980
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 4.97683155917345,
|
| 1435 |
-
"grad_norm":
|
| 1436 |
-
"learning_rate":
|
| 1437 |
-
"loss": 0.
|
| 1438 |
"step": 1990
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
"epoch": 4.989355040701315,
|
| 1442 |
-
"eval_accuracy": 0.
|
| 1443 |
-
"eval_loss": 0.
|
| 1444 |
-
"eval_runtime": 4.
|
| 1445 |
-
"eval_samples_per_second":
|
| 1446 |
-
"eval_steps_per_second":
|
| 1447 |
"step": 1995
|
| 1448 |
},
|
| 1449 |
{
|
| 1450 |
"epoch": 4.989355040701315,
|
| 1451 |
"step": 1995,
|
| 1452 |
-
"total_flos": 2.
|
| 1453 |
-
"train_loss": 0.
|
| 1454 |
-
"train_runtime":
|
| 1455 |
-
"train_samples_per_second":
|
| 1456 |
-
"train_steps_per_second": 3.
|
| 1457 |
}
|
| 1458 |
],
|
| 1459 |
"logging_steps": 10,
|
|
@@ -1473,7 +1473,7 @@
|
|
| 1473 |
"attributes": {}
|
| 1474 |
}
|
| 1475 |
},
|
| 1476 |
-
"total_flos": 2.
|
| 1477 |
"train_batch_size": 32,
|
| 1478 |
"trial_name": null,
|
| 1479 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.9814651368049426,
|
| 3 |
"best_model_checkpoint": "wav2vec2-base-ft-keyword-spotting/checkpoint-1995",
|
| 4 |
"epoch": 4.989355040701315,
|
| 5 |
"eval_steps": 500,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.025046963055729492,
|
| 13 |
+
"grad_norm": 2.144172430038452,
|
| 14 |
"learning_rate": 1.5e-06,
|
| 15 |
+
"loss": 3.8317,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.050093926111458985,
|
| 20 |
+
"grad_norm": 3.0447957515716553,
|
| 21 |
"learning_rate": 3e-06,
|
| 22 |
+
"loss": 4.1331,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.07514088916718847,
|
| 27 |
+
"grad_norm": 3.1870126724243164,
|
| 28 |
"learning_rate": 4.5e-06,
|
| 29 |
+
"loss": 4.0889,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.10018785222291797,
|
| 34 |
+
"grad_norm": 4.074451923370361,
|
| 35 |
"learning_rate": 6e-06,
|
| 36 |
+
"loss": 3.9025,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.12523481527864747,
|
| 41 |
+
"grad_norm": 5.182351112365723,
|
| 42 |
"learning_rate": 7.5e-06,
|
| 43 |
+
"loss": 3.6201,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.15028177833437695,
|
| 48 |
+
"grad_norm": 5.756130218505859,
|
| 49 |
"learning_rate": 9e-06,
|
| 50 |
+
"loss": 3.1977,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.17532874139010646,
|
| 55 |
+
"grad_norm": 5.65469217300415,
|
| 56 |
"learning_rate": 1.05e-05,
|
| 57 |
+
"loss": 2.7121,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.20037570444583594,
|
| 62 |
+
"grad_norm": 5.120871067047119,
|
| 63 |
"learning_rate": 1.2e-05,
|
| 64 |
+
"loss": 2.4593,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.22542266750156542,
|
| 69 |
+
"grad_norm": 4.952624320983887,
|
| 70 |
"learning_rate": 1.3500000000000001e-05,
|
| 71 |
+
"loss": 2.2216,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.25046963055729493,
|
| 76 |
+
"grad_norm": 4.202530384063721,
|
| 77 |
"learning_rate": 1.5e-05,
|
| 78 |
+
"loss": 2.0977,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.27551659361302444,
|
| 83 |
+
"grad_norm": 3.235758066177368,
|
| 84 |
"learning_rate": 1.65e-05,
|
| 85 |
+
"loss": 2.0442,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.3005635566687539,
|
| 90 |
+
"grad_norm": 3.456002712249756,
|
| 91 |
"learning_rate": 1.8e-05,
|
| 92 |
+
"loss": 1.8601,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.3256105197244834,
|
| 97 |
+
"grad_norm": 1.5474969148635864,
|
| 98 |
"learning_rate": 1.95e-05,
|
| 99 |
+
"loss": 1.7964,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.3506574827802129,
|
| 104 |
+
"grad_norm": 1.7901959419250488,
|
| 105 |
"learning_rate": 2.1e-05,
|
| 106 |
+
"loss": 1.8098,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.37570444583594237,
|
| 111 |
+
"grad_norm": 1.2036372423171997,
|
| 112 |
"learning_rate": 2.25e-05,
|
| 113 |
+
"loss": 1.7429,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.4007514088916719,
|
| 118 |
+
"grad_norm": 0.6613264083862305,
|
| 119 |
"learning_rate": 2.4e-05,
|
| 120 |
+
"loss": 1.7313,
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.4257983719474014,
|
| 125 |
+
"grad_norm": 1.468883991241455,
|
| 126 |
"learning_rate": 2.55e-05,
|
| 127 |
+
"loss": 1.7528,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.45084533500313084,
|
| 132 |
+
"grad_norm": 2.4236254692077637,
|
| 133 |
"learning_rate": 2.7000000000000002e-05,
|
| 134 |
+
"loss": 1.642,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.47589229805886035,
|
| 139 |
+
"grad_norm": 10.500153541564941,
|
| 140 |
"learning_rate": 2.8499999999999998e-05,
|
| 141 |
+
"loss": 1.6462,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.5009392611145899,
|
| 146 |
+
"grad_norm": 8.173652648925781,
|
| 147 |
"learning_rate": 3e-05,
|
| 148 |
+
"loss": 1.6049,
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.5259862241703194,
|
| 153 |
+
"grad_norm": 4.767404556274414,
|
| 154 |
"learning_rate": 2.9832869080779945e-05,
|
| 155 |
+
"loss": 1.5754,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.5510331872260489,
|
| 160 |
+
"grad_norm": 2.0099925994873047,
|
| 161 |
"learning_rate": 2.9665738161559886e-05,
|
| 162 |
+
"loss": 1.5001,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.5760801502817783,
|
| 167 |
+
"grad_norm": 2.7814888954162598,
|
| 168 |
"learning_rate": 2.9498607242339834e-05,
|
| 169 |
+
"loss": 1.4049,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.6011271133375078,
|
| 174 |
+
"grad_norm": 4.7790846824646,
|
| 175 |
"learning_rate": 2.933147632311978e-05,
|
| 176 |
+
"loss": 1.5321,
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.6261740763932373,
|
| 181 |
+
"grad_norm": 3.758580207824707,
|
| 182 |
"learning_rate": 2.916434540389972e-05,
|
| 183 |
+
"loss": 1.4002,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.6512210394489668,
|
| 188 |
+
"grad_norm": 5.455554008483887,
|
| 189 |
"learning_rate": 2.8997214484679665e-05,
|
| 190 |
+
"loss": 1.4013,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.6762680025046963,
|
| 195 |
+
"grad_norm": 5.183338165283203,
|
| 196 |
"learning_rate": 2.8830083565459613e-05,
|
| 197 |
+
"loss": 1.2016,
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.7013149655604258,
|
| 202 |
+
"grad_norm": 3.5465261936187744,
|
| 203 |
"learning_rate": 2.8662952646239554e-05,
|
| 204 |
+
"loss": 1.1541,
|
| 205 |
"step": 280
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.7263619286161553,
|
| 209 |
+
"grad_norm": 8.002464294433594,
|
| 210 |
"learning_rate": 2.84958217270195e-05,
|
| 211 |
+
"loss": 1.1754,
|
| 212 |
"step": 290
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.7514088916718847,
|
| 216 |
+
"grad_norm": 4.145716190338135,
|
| 217 |
"learning_rate": 2.8328690807799443e-05,
|
| 218 |
+
"loss": 1.1651,
|
| 219 |
"step": 300
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.7764558547276142,
|
| 223 |
+
"grad_norm": 8.420044898986816,
|
| 224 |
"learning_rate": 2.8161559888579388e-05,
|
| 225 |
+
"loss": 1.2086,
|
| 226 |
"step": 310
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.8015028177833438,
|
| 230 |
+
"grad_norm": 2.530792713165283,
|
| 231 |
"learning_rate": 2.7994428969359332e-05,
|
| 232 |
+
"loss": 1.1337,
|
| 233 |
"step": 320
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.8265497808390733,
|
| 237 |
+
"grad_norm": 3.45489501953125,
|
| 238 |
"learning_rate": 2.7827298050139277e-05,
|
| 239 |
+
"loss": 1.0499,
|
| 240 |
"step": 330
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.8515967438948028,
|
| 244 |
+
"grad_norm": 5.169933795928955,
|
| 245 |
"learning_rate": 2.7660167130919218e-05,
|
| 246 |
+
"loss": 1.1298,
|
| 247 |
"step": 340
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.8766437069505323,
|
| 251 |
+
"grad_norm": 5.91841459274292,
|
| 252 |
"learning_rate": 2.7493036211699166e-05,
|
| 253 |
+
"loss": 1.1476,
|
| 254 |
"step": 350
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.9016906700062617,
|
| 258 |
+
"grad_norm": 4.016351699829102,
|
| 259 |
"learning_rate": 2.732590529247911e-05,
|
| 260 |
+
"loss": 1.0944,
|
| 261 |
"step": 360
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.9267376330619912,
|
| 265 |
+
"grad_norm": 5.553752899169922,
|
| 266 |
"learning_rate": 2.7158774373259055e-05,
|
| 267 |
+
"loss": 1.0822,
|
| 268 |
"step": 370
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.9517845961177207,
|
| 272 |
+
"grad_norm": 4.519126892089844,
|
| 273 |
+
"learning_rate": 2.7008356545961002e-05,
|
| 274 |
+
"loss": 1.0057,
|
| 275 |
"step": 380
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.9768315591734502,
|
| 279 |
+
"grad_norm": 2.7625792026519775,
|
| 280 |
+
"learning_rate": 2.6841225626740946e-05,
|
| 281 |
+
"loss": 1.009,
|
| 282 |
"step": 390
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 1.0,
|
| 286 |
+
"grad_norm": 1.0362297296524048,
|
| 287 |
+
"learning_rate": 2.6674094707520894e-05,
|
| 288 |
+
"loss": 0.8747,
|
| 289 |
"step": 400
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 1.0,
|
| 293 |
+
"eval_accuracy": 0.9340982641953516,
|
| 294 |
+
"eval_loss": 0.806678831577301,
|
| 295 |
+
"eval_runtime": 4.4599,
|
| 296 |
+
"eval_samples_per_second": 1524.249,
|
| 297 |
+
"eval_steps_per_second": 47.759,
|
| 298 |
"step": 400
|
| 299 |
},
|
| 300 |
{
|
| 301 |
"epoch": 1.0250469630557295,
|
| 302 |
+
"grad_norm": 7.034424304962158,
|
| 303 |
+
"learning_rate": 2.6506963788300836e-05,
|
| 304 |
+
"loss": 0.8779,
|
| 305 |
"step": 410
|
| 306 |
},
|
| 307 |
{
|
| 308 |
"epoch": 1.050093926111459,
|
| 309 |
+
"grad_norm": 5.3810014724731445,
|
| 310 |
+
"learning_rate": 2.633983286908078e-05,
|
| 311 |
+
"loss": 0.7878,
|
| 312 |
"step": 420
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 1.0751408891671885,
|
| 316 |
+
"grad_norm": 3.344200372695923,
|
| 317 |
+
"learning_rate": 2.6172701949860725e-05,
|
| 318 |
+
"loss": 0.762,
|
| 319 |
"step": 430
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 1.100187852222918,
|
| 323 |
+
"grad_norm": 6.402628421783447,
|
| 324 |
+
"learning_rate": 2.600557103064067e-05,
|
| 325 |
+
"loss": 0.7631,
|
| 326 |
"step": 440
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 1.1252348152786475,
|
| 330 |
+
"grad_norm": 4.863397121429443,
|
| 331 |
+
"learning_rate": 2.5838440111420614e-05,
|
| 332 |
+
"loss": 0.6772,
|
| 333 |
"step": 450
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 1.150281778334377,
|
| 337 |
+
"grad_norm": 6.484178066253662,
|
| 338 |
+
"learning_rate": 2.567130919220056e-05,
|
| 339 |
+
"loss": 0.6461,
|
| 340 |
"step": 460
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 1.1753287413901066,
|
| 344 |
+
"grad_norm": 3.2635133266448975,
|
| 345 |
+
"learning_rate": 2.55041782729805e-05,
|
| 346 |
+
"loss": 0.5984,
|
| 347 |
"step": 470
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 1.2003757044458359,
|
| 351 |
+
"grad_norm": 2.40271258354187,
|
| 352 |
+
"learning_rate": 2.5337047353760448e-05,
|
| 353 |
+
"loss": 0.606,
|
| 354 |
"step": 480
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 1.2254226675015654,
|
| 358 |
+
"grad_norm": 2.1505284309387207,
|
| 359 |
+
"learning_rate": 2.5169916434540392e-05,
|
| 360 |
+
"loss": 0.5553,
|
| 361 |
"step": 490
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 1.2504696305572949,
|
| 365 |
+
"grad_norm": 2.476496934890747,
|
| 366 |
+
"learning_rate": 2.5002785515320333e-05,
|
| 367 |
+
"loss": 0.5238,
|
| 368 |
"step": 500
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 1.2755165936130244,
|
| 372 |
+
"grad_norm": 4.137519359588623,
|
| 373 |
+
"learning_rate": 2.4835654596100278e-05,
|
| 374 |
+
"loss": 0.5148,
|
| 375 |
"step": 510
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 1.300563556668754,
|
| 379 |
+
"grad_norm": 5.227903842926025,
|
| 380 |
+
"learning_rate": 2.4668523676880226e-05,
|
| 381 |
+
"loss": 0.4863,
|
| 382 |
"step": 520
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 1.3256105197244834,
|
| 386 |
+
"grad_norm": 4.056149482727051,
|
| 387 |
+
"learning_rate": 2.4501392757660167e-05,
|
| 388 |
+
"loss": 0.49,
|
| 389 |
"step": 530
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 1.350657482780213,
|
| 393 |
+
"grad_norm": 6.162842273712158,
|
| 394 |
+
"learning_rate": 2.433426183844011e-05,
|
| 395 |
+
"loss": 0.4264,
|
| 396 |
"step": 540
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 1.3757044458359424,
|
| 400 |
+
"grad_norm": 4.439515113830566,
|
| 401 |
+
"learning_rate": 2.4167130919220056e-05,
|
| 402 |
+
"loss": 0.4449,
|
| 403 |
"step": 550
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 1.400751408891672,
|
| 407 |
+
"grad_norm": 4.906720161437988,
|
| 408 |
+
"learning_rate": 2.4e-05,
|
| 409 |
+
"loss": 0.408,
|
| 410 |
"step": 560
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 1.4257983719474014,
|
| 414 |
+
"grad_norm": 5.718549728393555,
|
| 415 |
+
"learning_rate": 2.3832869080779945e-05,
|
| 416 |
+
"loss": 0.4143,
|
| 417 |
"step": 570
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 1.4508453350031307,
|
| 421 |
+
"grad_norm": 3.9077138900756836,
|
| 422 |
+
"learning_rate": 2.366573816155989e-05,
|
| 423 |
+
"loss": 0.3716,
|
| 424 |
"step": 580
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 1.4758922980588602,
|
| 428 |
+
"grad_norm": 3.5204200744628906,
|
| 429 |
+
"learning_rate": 2.349860724233983e-05,
|
| 430 |
+
"loss": 0.3932,
|
| 431 |
"step": 590
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 1.5009392611145898,
|
| 435 |
+
"grad_norm": 5.870133399963379,
|
| 436 |
+
"learning_rate": 2.333147632311978e-05,
|
| 437 |
+
"loss": 0.3932,
|
| 438 |
"step": 600
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 1.5259862241703193,
|
| 442 |
+
"grad_norm": 5.287498950958252,
|
| 443 |
+
"learning_rate": 2.3164345403899723e-05,
|
| 444 |
+
"loss": 0.4031,
|
| 445 |
"step": 610
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 1.5510331872260488,
|
| 449 |
+
"grad_norm": 5.271251201629639,
|
| 450 |
+
"learning_rate": 2.2997214484679665e-05,
|
| 451 |
+
"loss": 0.3467,
|
| 452 |
"step": 620
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 1.5760801502817783,
|
| 456 |
+
"grad_norm": 5.845817565917969,
|
| 457 |
+
"learning_rate": 2.283008356545961e-05,
|
| 458 |
+
"loss": 0.3593,
|
| 459 |
"step": 630
|
| 460 |
},
|
| 461 |
{
|
| 462 |
"epoch": 1.6011271133375078,
|
| 463 |
+
"grad_norm": 3.02872896194458,
|
| 464 |
+
"learning_rate": 2.2662952646239557e-05,
|
| 465 |
+
"loss": 0.3535,
|
| 466 |
"step": 640
|
| 467 |
},
|
| 468 |
{
|
| 469 |
"epoch": 1.6261740763932373,
|
| 470 |
+
"grad_norm": 2.3705966472625732,
|
| 471 |
+
"learning_rate": 2.2495821727019498e-05,
|
| 472 |
+
"loss": 0.3313,
|
| 473 |
"step": 650
|
| 474 |
},
|
| 475 |
{
|
| 476 |
"epoch": 1.6512210394489668,
|
| 477 |
+
"grad_norm": 3.9336166381835938,
|
| 478 |
+
"learning_rate": 2.2328690807799443e-05,
|
| 479 |
+
"loss": 0.374,
|
| 480 |
"step": 660
|
| 481 |
},
|
| 482 |
{
|
| 483 |
"epoch": 1.6762680025046963,
|
| 484 |
+
"grad_norm": 6.896333694458008,
|
| 485 |
+
"learning_rate": 2.2161559888579387e-05,
|
| 486 |
+
"loss": 0.359,
|
| 487 |
"step": 670
|
| 488 |
},
|
| 489 |
{
|
| 490 |
"epoch": 1.7013149655604258,
|
| 491 |
+
"grad_norm": 3.1803808212280273,
|
| 492 |
+
"learning_rate": 2.1994428969359335e-05,
|
| 493 |
+
"loss": 0.3215,
|
| 494 |
"step": 680
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 1.7263619286161553,
|
| 498 |
+
"grad_norm": 3.6727025508880615,
|
| 499 |
+
"learning_rate": 2.1827298050139276e-05,
|
| 500 |
+
"loss": 0.3313,
|
| 501 |
"step": 690
|
| 502 |
},
|
| 503 |
{
|
| 504 |
"epoch": 1.7514088916718848,
|
| 505 |
+
"grad_norm": 4.481452941894531,
|
| 506 |
+
"learning_rate": 2.166016713091922e-05,
|
| 507 |
+
"loss": 0.3075,
|
| 508 |
"step": 700
|
| 509 |
},
|
| 510 |
{
|
| 511 |
"epoch": 1.7764558547276144,
|
| 512 |
+
"grad_norm": 4.977258205413818,
|
| 513 |
+
"learning_rate": 2.1493036211699166e-05,
|
| 514 |
+
"loss": 0.2814,
|
| 515 |
"step": 710
|
| 516 |
},
|
| 517 |
{
|
| 518 |
"epoch": 1.8015028177833439,
|
| 519 |
+
"grad_norm": 4.018652439117432,
|
| 520 |
+
"learning_rate": 2.1325905292479107e-05,
|
| 521 |
+
"loss": 0.2902,
|
| 522 |
"step": 720
|
| 523 |
},
|
| 524 |
{
|
| 525 |
"epoch": 1.8265497808390734,
|
| 526 |
+
"grad_norm": 6.154052734375,
|
| 527 |
+
"learning_rate": 2.1158774373259055e-05,
|
| 528 |
+
"loss": 0.3123,
|
| 529 |
"step": 730
|
| 530 |
},
|
| 531 |
{
|
| 532 |
"epoch": 1.8515967438948029,
|
| 533 |
+
"grad_norm": 3.7956981658935547,
|
| 534 |
+
"learning_rate": 2.0991643454039e-05,
|
| 535 |
+
"loss": 0.3119,
|
| 536 |
"step": 740
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 1.8766437069505324,
|
| 540 |
+
"grad_norm": 6.320951461791992,
|
| 541 |
+
"learning_rate": 2.082451253481894e-05,
|
| 542 |
+
"loss": 0.2874,
|
| 543 |
"step": 750
|
| 544 |
},
|
| 545 |
{
|
| 546 |
"epoch": 1.9016906700062617,
|
| 547 |
+
"grad_norm": 2.6970086097717285,
|
| 548 |
+
"learning_rate": 2.0657381615598885e-05,
|
| 549 |
+
"loss": 0.2889,
|
| 550 |
"step": 760
|
| 551 |
},
|
| 552 |
{
|
| 553 |
"epoch": 1.9267376330619912,
|
| 554 |
+
"grad_norm": 4.386446952819824,
|
| 555 |
+
"learning_rate": 2.0490250696378833e-05,
|
| 556 |
+
"loss": 0.2529,
|
| 557 |
"step": 770
|
| 558 |
},
|
| 559 |
{
|
| 560 |
"epoch": 1.9517845961177207,
|
| 561 |
+
"grad_norm": 5.870710372924805,
|
| 562 |
+
"learning_rate": 2.0323119777158774e-05,
|
| 563 |
+
"loss": 0.2819,
|
| 564 |
"step": 780
|
| 565 |
},
|
| 566 |
{
|
| 567 |
"epoch": 1.9768315591734502,
|
| 568 |
+
"grad_norm": 5.930877685546875,
|
| 569 |
+
"learning_rate": 2.015598885793872e-05,
|
| 570 |
+
"loss": 0.2725,
|
| 571 |
"step": 790
|
| 572 |
},
|
| 573 |
{
|
| 574 |
"epoch": 2.0,
|
| 575 |
+
"grad_norm": 0.3753320872783661,
|
| 576 |
+
"learning_rate": 1.9988857938718663e-05,
|
| 577 |
+
"loss": 0.2332,
|
| 578 |
"step": 800
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 2.0,
|
| 582 |
+
"eval_accuracy": 0.9745513386290086,
|
| 583 |
+
"eval_loss": 0.17788007855415344,
|
| 584 |
+
"eval_runtime": 4.4101,
|
| 585 |
+
"eval_samples_per_second": 1541.463,
|
| 586 |
+
"eval_steps_per_second": 48.298,
|
| 587 |
"step": 800
|
| 588 |
},
|
| 589 |
{
|
| 590 |
"epoch": 2.0250469630557295,
|
| 591 |
+
"grad_norm": 6.34937047958374,
|
| 592 |
+
"learning_rate": 1.9821727019498608e-05,
|
| 593 |
+
"loss": 0.2545,
|
| 594 |
"step": 810
|
| 595 |
},
|
| 596 |
{
|
| 597 |
"epoch": 2.050093926111459,
|
| 598 |
+
"grad_norm": 4.1271138191223145,
|
| 599 |
+
"learning_rate": 1.9654596100278552e-05,
|
| 600 |
+
"loss": 0.2497,
|
| 601 |
"step": 820
|
| 602 |
},
|
| 603 |
{
|
| 604 |
"epoch": 2.0751408891671885,
|
| 605 |
+
"grad_norm": 5.419626712799072,
|
| 606 |
+
"learning_rate": 1.9487465181058497e-05,
|
| 607 |
+
"loss": 0.2502,
|
| 608 |
"step": 830
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 2.100187852222918,
|
| 612 |
+
"grad_norm": 6.219632148742676,
|
| 613 |
+
"learning_rate": 1.9320334261838438e-05,
|
| 614 |
+
"loss": 0.2492,
|
| 615 |
"step": 840
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 2.1252348152786475,
|
| 619 |
+
"grad_norm": 6.462090492248535,
|
| 620 |
+
"learning_rate": 1.9153203342618386e-05,
|
| 621 |
+
"loss": 0.2517,
|
| 622 |
"step": 850
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 2.150281778334377,
|
| 626 |
+
"grad_norm": 3.6385743618011475,
|
| 627 |
+
"learning_rate": 1.898607242339833e-05,
|
| 628 |
+
"loss": 0.2393,
|
| 629 |
"step": 860
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 2.1753287413901066,
|
| 633 |
+
"grad_norm": 4.627376556396484,
|
| 634 |
+
"learning_rate": 1.8818941504178272e-05,
|
| 635 |
+
"loss": 0.2414,
|
| 636 |
"step": 870
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 2.200375704445836,
|
| 640 |
+
"grad_norm": 5.165160179138184,
|
| 641 |
+
"learning_rate": 1.8651810584958216e-05,
|
| 642 |
+
"loss": 0.2357,
|
| 643 |
"step": 880
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 2.2254226675015656,
|
| 647 |
+
"grad_norm": 1.4684484004974365,
|
| 648 |
+
"learning_rate": 1.8484679665738164e-05,
|
| 649 |
+
"loss": 0.2548,
|
| 650 |
"step": 890
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 2.250469630557295,
|
| 654 |
+
"grad_norm": 3.5594701766967773,
|
| 655 |
+
"learning_rate": 1.8317548746518105e-05,
|
| 656 |
+
"loss": 0.2403,
|
| 657 |
"step": 900
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 2.2755165936130246,
|
| 661 |
+
"grad_norm": 3.314188003540039,
|
| 662 |
+
"learning_rate": 1.815041782729805e-05,
|
| 663 |
+
"loss": 0.2788,
|
| 664 |
"step": 910
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 2.300563556668754,
|
| 668 |
+
"grad_norm": 2.3456945419311523,
|
| 669 |
+
"learning_rate": 1.7983286908077995e-05,
|
| 670 |
+
"loss": 0.2707,
|
| 671 |
"step": 920
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 2.325610519724483,
|
| 675 |
+
"grad_norm": 3.4486682415008545,
|
| 676 |
+
"learning_rate": 1.781615598885794e-05,
|
| 677 |
+
"loss": 0.2304,
|
| 678 |
"step": 930
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 2.350657482780213,
|
| 682 |
+
"grad_norm": 3.3779501914978027,
|
| 683 |
+
"learning_rate": 1.7649025069637884e-05,
|
| 684 |
+
"loss": 0.2318,
|
| 685 |
"step": 940
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 2.375704445835942,
|
| 689 |
+
"grad_norm": 1.7540189027786255,
|
| 690 |
+
"learning_rate": 1.7481894150417828e-05,
|
| 691 |
+
"loss": 0.2125,
|
| 692 |
"step": 950
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 2.4007514088916717,
|
| 696 |
+
"grad_norm": 8.057242393493652,
|
| 697 |
+
"learning_rate": 1.731476323119777e-05,
|
| 698 |
+
"loss": 0.2407,
|
| 699 |
"step": 960
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 2.425798371947401,
|
| 703 |
+
"grad_norm": 3.5279042720794678,
|
| 704 |
+
"learning_rate": 1.7147632311977717e-05,
|
| 705 |
+
"loss": 0.2432,
|
| 706 |
"step": 970
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 2.4508453350031307,
|
| 710 |
+
"grad_norm": 4.324340343475342,
|
| 711 |
+
"learning_rate": 1.6980501392757662e-05,
|
| 712 |
+
"loss": 0.2316,
|
| 713 |
"step": 980
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 2.4758922980588602,
|
| 717 |
+
"grad_norm": 3.12505841255188,
|
| 718 |
+
"learning_rate": 1.6813370473537606e-05,
|
| 719 |
+
"loss": 0.2226,
|
| 720 |
"step": 990
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 2.5009392611145898,
|
| 724 |
+
"grad_norm": 3.6332638263702393,
|
| 725 |
+
"learning_rate": 1.6646239554317548e-05,
|
| 726 |
+
"loss": 0.2305,
|
| 727 |
"step": 1000
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 2.5259862241703193,
|
| 731 |
+
"grad_norm": 3.011505126953125,
|
| 732 |
+
"learning_rate": 1.6479108635097496e-05,
|
| 733 |
+
"loss": 0.2361,
|
| 734 |
"step": 1010
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 2.5510331872260488,
|
| 738 |
+
"grad_norm": 3.6663644313812256,
|
| 739 |
+
"learning_rate": 1.631197771587744e-05,
|
| 740 |
+
"loss": 0.2398,
|
| 741 |
"step": 1020
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 2.5760801502817783,
|
| 745 |
+
"grad_norm": 3.6134729385375977,
|
| 746 |
+
"learning_rate": 1.614484679665738e-05,
|
| 747 |
+
"loss": 0.1777,
|
| 748 |
"step": 1030
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 2.601127113337508,
|
| 752 |
+
"grad_norm": 3.180910587310791,
|
| 753 |
+
"learning_rate": 1.5977715877437326e-05,
|
| 754 |
+
"loss": 0.2304,
|
| 755 |
"step": 1040
|
| 756 |
},
|
| 757 |
{
|
| 758 |
"epoch": 2.6261740763932373,
|
| 759 |
+
"grad_norm": 2.667623281478882,
|
| 760 |
+
"learning_rate": 1.581058495821727e-05,
|
| 761 |
+
"loss": 0.1787,
|
| 762 |
"step": 1050
|
| 763 |
},
|
| 764 |
{
|
| 765 |
"epoch": 2.651221039448967,
|
| 766 |
+
"grad_norm": 6.972463130950928,
|
| 767 |
+
"learning_rate": 1.5643454038997215e-05,
|
| 768 |
+
"loss": 0.2026,
|
| 769 |
"step": 1060
|
| 770 |
},
|
| 771 |
{
|
| 772 |
"epoch": 2.6762680025046963,
|
| 773 |
+
"grad_norm": 1.4821382761001587,
|
| 774 |
+
"learning_rate": 1.547632311977716e-05,
|
| 775 |
+
"loss": 0.2295,
|
| 776 |
"step": 1070
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 2.701314965560426,
|
| 780 |
+
"grad_norm": 3.339320421218872,
|
| 781 |
+
"learning_rate": 1.5309192200557104e-05,
|
| 782 |
+
"loss": 0.2234,
|
| 783 |
"step": 1080
|
| 784 |
},
|
| 785 |
{
|
| 786 |
"epoch": 2.7263619286161553,
|
| 787 |
+
"grad_norm": 3.3274917602539062,
|
| 788 |
+
"learning_rate": 1.5142061281337047e-05,
|
| 789 |
+
"loss": 0.1899,
|
| 790 |
"step": 1090
|
| 791 |
},
|
| 792 |
{
|
| 793 |
"epoch": 2.751408891671885,
|
| 794 |
+
"grad_norm": 3.077637195587158,
|
| 795 |
+
"learning_rate": 1.4974930362116992e-05,
|
| 796 |
+
"loss": 0.1747,
|
| 797 |
"step": 1100
|
| 798 |
},
|
| 799 |
{
|
| 800 |
"epoch": 2.7764558547276144,
|
| 801 |
+
"grad_norm": 4.980368137359619,
|
| 802 |
+
"learning_rate": 1.4807799442896936e-05,
|
| 803 |
+
"loss": 0.2359,
|
| 804 |
"step": 1110
|
| 805 |
},
|
| 806 |
{
|
| 807 |
"epoch": 2.801502817783344,
|
| 808 |
+
"grad_norm": 3.46724534034729,
|
| 809 |
+
"learning_rate": 1.464066852367688e-05,
|
| 810 |
+
"loss": 0.1945,
|
| 811 |
"step": 1120
|
| 812 |
},
|
| 813 |
{
|
| 814 |
"epoch": 2.8265497808390734,
|
| 815 |
+
"grad_norm": 6.3585710525512695,
|
| 816 |
+
"learning_rate": 1.4473537604456825e-05,
|
| 817 |
+
"loss": 0.2356,
|
| 818 |
"step": 1130
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 2.851596743894803,
|
| 822 |
+
"grad_norm": 3.6611506938934326,
|
| 823 |
+
"learning_rate": 1.4306406685236768e-05,
|
| 824 |
+
"loss": 0.223,
|
| 825 |
"step": 1140
|
| 826 |
},
|
| 827 |
{
|
| 828 |
"epoch": 2.8766437069505324,
|
| 829 |
+
"grad_norm": 3.0758209228515625,
|
| 830 |
+
"learning_rate": 1.4139275766016714e-05,
|
| 831 |
+
"loss": 0.2675,
|
| 832 |
"step": 1150
|
| 833 |
},
|
| 834 |
{
|
| 835 |
"epoch": 2.9016906700062615,
|
| 836 |
+
"grad_norm": 3.0930421352386475,
|
| 837 |
+
"learning_rate": 1.3972144846796657e-05,
|
| 838 |
+
"loss": 0.2342,
|
| 839 |
"step": 1160
|
| 840 |
},
|
| 841 |
{
|
| 842 |
"epoch": 2.9267376330619914,
|
| 843 |
+
"grad_norm": 3.915057897567749,
|
| 844 |
+
"learning_rate": 1.3805013927576602e-05,
|
| 845 |
+
"loss": 0.2089,
|
| 846 |
"step": 1170
|
| 847 |
},
|
| 848 |
{
|
| 849 |
"epoch": 2.9517845961177205,
|
| 850 |
+
"grad_norm": 3.755885601043701,
|
| 851 |
+
"learning_rate": 1.3637883008356546e-05,
|
| 852 |
+
"loss": 0.2251,
|
| 853 |
"step": 1180
|
| 854 |
},
|
| 855 |
{
|
| 856 |
"epoch": 2.9768315591734504,
|
| 857 |
+
"grad_norm": 3.1619045734405518,
|
| 858 |
+
"learning_rate": 1.3470752089136491e-05,
|
| 859 |
+
"loss": 0.1927,
|
| 860 |
"step": 1190
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 3.0,
|
| 864 |
+
"grad_norm": 3.681858539581299,
|
| 865 |
+
"learning_rate": 1.3303621169916434e-05,
|
| 866 |
+
"loss": 0.217,
|
| 867 |
"step": 1200
|
| 868 |
},
|
| 869 |
{
|
| 870 |
"epoch": 3.0,
|
| 871 |
+
"eval_accuracy": 0.9766107678729038,
|
| 872 |
+
"eval_loss": 0.1263045072555542,
|
| 873 |
+
"eval_runtime": 4.7057,
|
| 874 |
+
"eval_samples_per_second": 1444.62,
|
| 875 |
+
"eval_steps_per_second": 45.264,
|
| 876 |
"step": 1200
|
| 877 |
},
|
| 878 |
{
|
| 879 |
"epoch": 3.0250469630557295,
|
| 880 |
+
"grad_norm": 4.34127140045166,
|
| 881 |
+
"learning_rate": 1.313649025069638e-05,
|
| 882 |
+
"loss": 0.2222,
|
| 883 |
"step": 1210
|
| 884 |
},
|
| 885 |
{
|
| 886 |
"epoch": 3.050093926111459,
|
| 887 |
+
"grad_norm": 2.2270660400390625,
|
| 888 |
+
"learning_rate": 1.2969359331476323e-05,
|
| 889 |
+
"loss": 0.1822,
|
| 890 |
"step": 1220
|
| 891 |
},
|
| 892 |
{
|
| 893 |
"epoch": 3.0751408891671885,
|
| 894 |
+
"grad_norm": 3.879969358444214,
|
| 895 |
+
"learning_rate": 1.2802228412256267e-05,
|
| 896 |
+
"loss": 0.1947,
|
| 897 |
"step": 1230
|
| 898 |
},
|
| 899 |
{
|
| 900 |
"epoch": 3.100187852222918,
|
| 901 |
+
"grad_norm": 4.284245491027832,
|
| 902 |
+
"learning_rate": 1.2635097493036212e-05,
|
| 903 |
+
"loss": 0.2126,
|
| 904 |
"step": 1240
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 3.1252348152786475,
|
| 908 |
+
"grad_norm": 3.0802762508392334,
|
| 909 |
+
"learning_rate": 1.2467966573816157e-05,
|
| 910 |
+
"loss": 0.1954,
|
| 911 |
"step": 1250
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 3.150281778334377,
|
| 915 |
+
"grad_norm": 4.699860095977783,
|
| 916 |
+
"learning_rate": 1.2300835654596101e-05,
|
| 917 |
+
"loss": 0.2189,
|
| 918 |
"step": 1260
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 3.1753287413901066,
|
| 922 |
+
"grad_norm": 4.755823612213135,
|
| 923 |
+
"learning_rate": 1.2133704735376046e-05,
|
| 924 |
+
"loss": 0.2019,
|
| 925 |
"step": 1270
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 3.200375704445836,
|
| 929 |
+
"grad_norm": 4.778765678405762,
|
| 930 |
+
"learning_rate": 1.1966573816155989e-05,
|
| 931 |
+
"loss": 0.1987,
|
| 932 |
"step": 1280
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 3.2254226675015656,
|
| 936 |
+
"grad_norm": 4.719218730926514,
|
| 937 |
+
"learning_rate": 1.1799442896935935e-05,
|
| 938 |
+
"loss": 0.1947,
|
| 939 |
"step": 1290
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 3.250469630557295,
|
| 943 |
+
"grad_norm": 4.547497272491455,
|
| 944 |
+
"learning_rate": 1.1632311977715878e-05,
|
| 945 |
+
"loss": 0.2097,
|
| 946 |
"step": 1300
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 3.2755165936130246,
|
| 950 |
+
"grad_norm": 2.1130096912384033,
|
| 951 |
+
"learning_rate": 1.1465181058495822e-05,
|
| 952 |
+
"loss": 0.1327,
|
| 953 |
"step": 1310
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 3.300563556668754,
|
| 957 |
+
"grad_norm": 4.512012958526611,
|
| 958 |
+
"learning_rate": 1.1298050139275767e-05,
|
| 959 |
+
"loss": 0.178,
|
| 960 |
"step": 1320
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 3.325610519724483,
|
| 964 |
+
"grad_norm": 2.9694018363952637,
|
| 965 |
+
"learning_rate": 1.1130919220055711e-05,
|
| 966 |
+
"loss": 0.2077,
|
| 967 |
"step": 1330
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 3.350657482780213,
|
| 971 |
+
"grad_norm": 2.5430564880371094,
|
| 972 |
+
"learning_rate": 1.0963788300835654e-05,
|
| 973 |
+
"loss": 0.1774,
|
| 974 |
"step": 1340
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 3.375704445835942,
|
| 978 |
+
"grad_norm": 5.131649971008301,
|
| 979 |
+
"learning_rate": 1.0796657381615599e-05,
|
| 980 |
+
"loss": 0.2013,
|
| 981 |
"step": 1350
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 3.4007514088916717,
|
| 985 |
+
"grad_norm": 3.391754627227783,
|
| 986 |
+
"learning_rate": 1.0629526462395543e-05,
|
| 987 |
+
"loss": 0.1772,
|
| 988 |
"step": 1360
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 3.425798371947401,
|
| 992 |
+
"grad_norm": 3.4772632122039795,
|
| 993 |
+
"learning_rate": 1.0462395543175486e-05,
|
| 994 |
+
"loss": 0.1597,
|
| 995 |
"step": 1370
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 3.4508453350031307,
|
| 999 |
+
"grad_norm": 3.5944013595581055,
|
| 1000 |
+
"learning_rate": 1.0295264623955432e-05,
|
| 1001 |
+
"loss": 0.2192,
|
| 1002 |
"step": 1380
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 3.4758922980588602,
|
| 1006 |
+
"grad_norm": 2.526901960372925,
|
| 1007 |
+
"learning_rate": 1.0128133704735375e-05,
|
| 1008 |
+
"loss": 0.168,
|
| 1009 |
"step": 1390
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 3.5009392611145898,
|
| 1013 |
+
"grad_norm": 3.234485626220703,
|
| 1014 |
+
"learning_rate": 9.961002785515322e-06,
|
| 1015 |
+
"loss": 0.2007,
|
| 1016 |
"step": 1400
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 3.5259862241703193,
|
| 1020 |
+
"grad_norm": 4.3917622566223145,
|
| 1021 |
+
"learning_rate": 9.793871866295264e-06,
|
| 1022 |
+
"loss": 0.1766,
|
| 1023 |
"step": 1410
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 3.5510331872260488,
|
| 1027 |
+
"grad_norm": 1.4780800342559814,
|
| 1028 |
+
"learning_rate": 9.626740947075209e-06,
|
| 1029 |
+
"loss": 0.1738,
|
| 1030 |
"step": 1420
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 3.5760801502817783,
|
| 1034 |
+
"grad_norm": 3.670740842819214,
|
| 1035 |
+
"learning_rate": 9.459610027855154e-06,
|
| 1036 |
+
"loss": 0.2033,
|
| 1037 |
"step": 1430
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 3.601127113337508,
|
| 1041 |
+
"grad_norm": 4.08475923538208,
|
| 1042 |
+
"learning_rate": 9.292479108635098e-06,
|
| 1043 |
+
"loss": 0.1638,
|
| 1044 |
"step": 1440
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 3.6261740763932373,
|
| 1048 |
+
"grad_norm": 3.445945978164673,
|
| 1049 |
+
"learning_rate": 9.125348189415041e-06,
|
| 1050 |
+
"loss": 0.1854,
|
| 1051 |
"step": 1450
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
"epoch": 3.651221039448967,
|
| 1055 |
+
"grad_norm": 3.1884312629699707,
|
| 1056 |
+
"learning_rate": 8.958217270194987e-06,
|
| 1057 |
+
"loss": 0.1967,
|
| 1058 |
"step": 1460
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
"epoch": 3.6762680025046963,
|
| 1062 |
+
"grad_norm": 1.9130624532699585,
|
| 1063 |
+
"learning_rate": 8.79108635097493e-06,
|
| 1064 |
+
"loss": 0.1618,
|
| 1065 |
"step": 1470
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 3.701314965560426,
|
| 1069 |
+
"grad_norm": 1.0646212100982666,
|
| 1070 |
+
"learning_rate": 8.623955431754875e-06,
|
| 1071 |
+
"loss": 0.1816,
|
| 1072 |
"step": 1480
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
"epoch": 3.7263619286161553,
|
| 1076 |
+
"grad_norm": 3.629429817199707,
|
| 1077 |
+
"learning_rate": 8.45682451253482e-06,
|
| 1078 |
+
"loss": 0.1432,
|
| 1079 |
"step": 1490
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
"epoch": 3.751408891671885,
|
| 1083 |
+
"grad_norm": 2.1418120861053467,
|
| 1084 |
+
"learning_rate": 8.289693593314764e-06,
|
| 1085 |
+
"loss": 0.1662,
|
| 1086 |
"step": 1500
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 3.7764558547276144,
|
| 1090 |
+
"grad_norm": 3.682490825653076,
|
| 1091 |
+
"learning_rate": 8.122562674094707e-06,
|
| 1092 |
+
"loss": 0.1819,
|
| 1093 |
"step": 1510
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
"epoch": 3.801502817783344,
|
| 1097 |
+
"grad_norm": 2.9112191200256348,
|
| 1098 |
+
"learning_rate": 7.955431754874653e-06,
|
| 1099 |
+
"loss": 0.1785,
|
| 1100 |
"step": 1520
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
"epoch": 3.8265497808390734,
|
| 1104 |
+
"grad_norm": 3.727522134780884,
|
| 1105 |
+
"learning_rate": 7.788300835654596e-06,
|
| 1106 |
+
"loss": 0.1476,
|
| 1107 |
"step": 1530
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
"epoch": 3.851596743894803,
|
| 1111 |
+
"grad_norm": 4.77044153213501,
|
| 1112 |
+
"learning_rate": 7.621169916434541e-06,
|
| 1113 |
+
"loss": 0.1668,
|
| 1114 |
"step": 1540
|
| 1115 |
},
|
| 1116 |
{
|
| 1117 |
"epoch": 3.8766437069505324,
|
| 1118 |
+
"grad_norm": 2.953248977661133,
|
| 1119 |
+
"learning_rate": 7.454038997214485e-06,
|
| 1120 |
+
"loss": 0.1712,
|
| 1121 |
"step": 1550
|
| 1122 |
},
|
| 1123 |
{
|
| 1124 |
"epoch": 3.9016906700062615,
|
| 1125 |
+
"grad_norm": 4.06650972366333,
|
| 1126 |
+
"learning_rate": 7.2869080779944286e-06,
|
| 1127 |
+
"loss": 0.1621,
|
| 1128 |
"step": 1560
|
| 1129 |
},
|
| 1130 |
{
|
| 1131 |
"epoch": 3.9267376330619914,
|
| 1132 |
+
"grad_norm": 4.628715991973877,
|
| 1133 |
+
"learning_rate": 7.119777158774373e-06,
|
| 1134 |
+
"loss": 0.1513,
|
| 1135 |
"step": 1570
|
| 1136 |
},
|
| 1137 |
{
|
| 1138 |
"epoch": 3.9517845961177205,
|
| 1139 |
+
"grad_norm": 2.5671701431274414,
|
| 1140 |
+
"learning_rate": 6.952646239554318e-06,
|
| 1141 |
+
"loss": 0.1918,
|
| 1142 |
"step": 1580
|
| 1143 |
},
|
| 1144 |
{
|
| 1145 |
"epoch": 3.9768315591734504,
|
| 1146 |
+
"grad_norm": 4.511129379272461,
|
| 1147 |
+
"learning_rate": 6.785515320334261e-06,
|
| 1148 |
+
"loss": 0.1957,
|
| 1149 |
"step": 1590
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
"epoch": 4.0,
|
| 1153 |
+
"grad_norm": 1.4342707395553589,
|
| 1154 |
+
"learning_rate": 6.618384401114206e-06,
|
| 1155 |
+
"loss": 0.1529,
|
| 1156 |
"step": 1600
|
| 1157 |
},
|
| 1158 |
{
|
| 1159 |
"epoch": 4.0,
|
| 1160 |
+
"eval_accuracy": 0.9805825242718447,
|
| 1161 |
+
"eval_loss": 0.10449180752038956,
|
| 1162 |
+
"eval_runtime": 4.637,
|
| 1163 |
+
"eval_samples_per_second": 1466.037,
|
| 1164 |
+
"eval_steps_per_second": 45.935,
|
| 1165 |
"step": 1600
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 4.025046963055729,
|
| 1169 |
+
"grad_norm": 3.8492510318756104,
|
| 1170 |
+
"learning_rate": 6.4512534818941505e-06,
|
| 1171 |
+
"loss": 0.1797,
|
| 1172 |
"step": 1610
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 4.050093926111459,
|
| 1176 |
+
"grad_norm": 4.300637245178223,
|
| 1177 |
+
"learning_rate": 6.284122562674095e-06,
|
| 1178 |
+
"loss": 0.1598,
|
| 1179 |
"step": 1620
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 4.075140889167188,
|
| 1183 |
+
"grad_norm": 1.9370712041854858,
|
| 1184 |
+
"learning_rate": 6.116991643454039e-06,
|
| 1185 |
+
"loss": 0.1746,
|
| 1186 |
"step": 1630
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 4.100187852222918,
|
| 1190 |
+
"grad_norm": 2.7867672443389893,
|
| 1191 |
+
"learning_rate": 5.949860724233983e-06,
|
| 1192 |
+
"loss": 0.1913,
|
| 1193 |
"step": 1640
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 4.125234815278647,
|
| 1197 |
+
"grad_norm": 3.536440372467041,
|
| 1198 |
+
"learning_rate": 5.782729805013928e-06,
|
| 1199 |
+
"loss": 0.1731,
|
| 1200 |
"step": 1650
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 4.150281778334377,
|
| 1204 |
+
"grad_norm": 3.2297356128692627,
|
| 1205 |
+
"learning_rate": 5.615598885793872e-06,
|
| 1206 |
+
"loss": 0.2019,
|
| 1207 |
"step": 1660
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 4.175328741390106,
|
| 1211 |
+
"grad_norm": 5.139032363891602,
|
| 1212 |
+
"learning_rate": 5.448467966573816e-06,
|
| 1213 |
+
"loss": 0.1793,
|
| 1214 |
"step": 1670
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 4.200375704445836,
|
| 1218 |
+
"grad_norm": 3.031764030456543,
|
| 1219 |
+
"learning_rate": 5.281337047353761e-06,
|
| 1220 |
+
"loss": 0.1836,
|
| 1221 |
"step": 1680
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 4.225422667501565,
|
| 1225 |
+
"grad_norm": 2.6612586975097656,
|
| 1226 |
+
"learning_rate": 5.114206128133705e-06,
|
| 1227 |
+
"loss": 0.1629,
|
| 1228 |
"step": 1690
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 4.250469630557295,
|
| 1232 |
+
"grad_norm": 2.8296072483062744,
|
| 1233 |
+
"learning_rate": 4.947075208913649e-06,
|
| 1234 |
+
"loss": 0.1691,
|
| 1235 |
"step": 1700
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 4.275516593613024,
|
| 1239 |
+
"grad_norm": 4.683578968048096,
|
| 1240 |
+
"learning_rate": 4.7799442896935936e-06,
|
| 1241 |
+
"loss": 0.1489,
|
| 1242 |
"step": 1710
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 4.300563556668754,
|
| 1246 |
+
"grad_norm": 3.7361319065093994,
|
| 1247 |
+
"learning_rate": 4.612813370473538e-06,
|
| 1248 |
+
"loss": 0.1453,
|
| 1249 |
"step": 1720
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 4.325610519724483,
|
| 1253 |
+
"grad_norm": 4.293661117553711,
|
| 1254 |
+
"learning_rate": 4.445682451253482e-06,
|
| 1255 |
+
"loss": 0.1445,
|
| 1256 |
"step": 1730
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 4.350657482780213,
|
| 1260 |
+
"grad_norm": 3.1756207942962646,
|
| 1261 |
+
"learning_rate": 4.278551532033426e-06,
|
| 1262 |
+
"loss": 0.1523,
|
| 1263 |
"step": 1740
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 4.375704445835942,
|
| 1267 |
+
"grad_norm": 3.921405792236328,
|
| 1268 |
+
"learning_rate": 4.111420612813371e-06,
|
| 1269 |
+
"loss": 0.1603,
|
| 1270 |
"step": 1750
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 4.400751408891672,
|
| 1274 |
+
"grad_norm": 2.8336334228515625,
|
| 1275 |
+
"learning_rate": 3.944289693593315e-06,
|
| 1276 |
+
"loss": 0.1764,
|
| 1277 |
"step": 1760
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 4.425798371947401,
|
| 1281 |
+
"grad_norm": 3.7013275623321533,
|
| 1282 |
+
"learning_rate": 3.7771587743732592e-06,
|
| 1283 |
+
"loss": 0.1481,
|
| 1284 |
"step": 1770
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 4.450845335003131,
|
| 1288 |
+
"grad_norm": 1.9302759170532227,
|
| 1289 |
+
"learning_rate": 3.6100278551532034e-06,
|
| 1290 |
+
"loss": 0.1526,
|
| 1291 |
"step": 1780
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 4.47589229805886,
|
| 1295 |
+
"grad_norm": 4.772688865661621,
|
| 1296 |
+
"learning_rate": 3.4428969359331475e-06,
|
| 1297 |
+
"loss": 0.1595,
|
| 1298 |
"step": 1790
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 4.50093926111459,
|
| 1302 |
+
"grad_norm": 3.1509013175964355,
|
| 1303 |
+
"learning_rate": 3.275766016713092e-06,
|
| 1304 |
+
"loss": 0.1384,
|
| 1305 |
"step": 1800
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 4.525986224170319,
|
| 1309 |
+
"grad_norm": 2.2213551998138428,
|
| 1310 |
+
"learning_rate": 3.108635097493036e-06,
|
| 1311 |
+
"loss": 0.1542,
|
| 1312 |
"step": 1810
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 4.551033187226049,
|
| 1316 |
+
"grad_norm": 2.7349748611450195,
|
| 1317 |
+
"learning_rate": 2.9415041782729803e-06,
|
| 1318 |
+
"loss": 0.1739,
|
| 1319 |
"step": 1820
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 4.576080150281778,
|
| 1323 |
+
"grad_norm": 5.009521961212158,
|
| 1324 |
+
"learning_rate": 2.774373259052925e-06,
|
| 1325 |
+
"loss": 0.1355,
|
| 1326 |
"step": 1830
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 4.601127113337508,
|
| 1330 |
+
"grad_norm": 3.5362050533294678,
|
| 1331 |
+
"learning_rate": 2.607242339832869e-06,
|
| 1332 |
+
"loss": 0.1335,
|
| 1333 |
"step": 1840
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 4.626174076393237,
|
| 1337 |
+
"grad_norm": 4.31157922744751,
|
| 1338 |
+
"learning_rate": 2.4401114206128136e-06,
|
| 1339 |
+
"loss": 0.1848,
|
| 1340 |
"step": 1850
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 4.651221039448966,
|
| 1344 |
+
"grad_norm": 3.6075448989868164,
|
| 1345 |
+
"learning_rate": 2.2729805013927577e-06,
|
| 1346 |
+
"loss": 0.1763,
|
| 1347 |
"step": 1860
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 4.676268002504696,
|
| 1351 |
+
"grad_norm": 3.1636250019073486,
|
| 1352 |
+
"learning_rate": 2.1058495821727023e-06,
|
| 1353 |
+
"loss": 0.1672,
|
| 1354 |
"step": 1870
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 4.701314965560426,
|
| 1358 |
+
"grad_norm": 0.7837923169136047,
|
| 1359 |
+
"learning_rate": 1.9387186629526464e-06,
|
| 1360 |
+
"loss": 0.1223,
|
| 1361 |
"step": 1880
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 4.726361928616155,
|
| 1365 |
+
"grad_norm": 3.2040278911590576,
|
| 1366 |
+
"learning_rate": 1.7715877437325906e-06,
|
| 1367 |
+
"loss": 0.1676,
|
| 1368 |
"step": 1890
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 4.751408891671884,
|
| 1372 |
+
"grad_norm": 2.6808199882507324,
|
| 1373 |
+
"learning_rate": 1.604456824512535e-06,
|
| 1374 |
+
"loss": 0.1606,
|
| 1375 |
"step": 1900
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 4.776455854727614,
|
| 1379 |
+
"grad_norm": 2.356374979019165,
|
| 1380 |
+
"learning_rate": 1.437325905292479e-06,
|
| 1381 |
+
"loss": 0.1655,
|
| 1382 |
"step": 1910
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 4.801502817783343,
|
| 1386 |
+
"grad_norm": 3.077162504196167,
|
| 1387 |
+
"learning_rate": 1.2701949860724234e-06,
|
| 1388 |
+
"loss": 0.1332,
|
| 1389 |
"step": 1920
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 4.826549780839073,
|
| 1393 |
+
"grad_norm": 3.580504894256592,
|
| 1394 |
+
"learning_rate": 1.1030640668523677e-06,
|
| 1395 |
+
"loss": 0.1687,
|
| 1396 |
"step": 1930
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 4.851596743894802,
|
| 1400 |
+
"grad_norm": 4.193363189697266,
|
| 1401 |
+
"learning_rate": 9.35933147632312e-07,
|
| 1402 |
+
"loss": 0.163,
|
| 1403 |
"step": 1940
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 4.876643706950532,
|
| 1407 |
+
"grad_norm": 3.2785427570343018,
|
| 1408 |
+
"learning_rate": 7.688022284122563e-07,
|
| 1409 |
+
"loss": 0.1546,
|
| 1410 |
"step": 1950
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 4.9016906700062615,
|
| 1414 |
+
"grad_norm": 2.1263206005096436,
|
| 1415 |
+
"learning_rate": 6.016713091922006e-07,
|
| 1416 |
+
"loss": 0.1445,
|
| 1417 |
"step": 1960
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 4.926737633061991,
|
| 1421 |
+
"grad_norm": 1.8054914474487305,
|
| 1422 |
+
"learning_rate": 4.3454038997214486e-07,
|
| 1423 |
+
"loss": 0.1489,
|
| 1424 |
"step": 1970
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
"epoch": 4.9517845961177205,
|
| 1428 |
+
"grad_norm": 2.179539680480957,
|
| 1429 |
+
"learning_rate": 2.6740947075208915e-07,
|
| 1430 |
+
"loss": 0.1632,
|
| 1431 |
"step": 1980
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 4.97683155917345,
|
| 1435 |
+
"grad_norm": 2.8142571449279785,
|
| 1436 |
+
"learning_rate": 1.0027855153203343e-07,
|
| 1437 |
+
"loss": 0.15,
|
| 1438 |
"step": 1990
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
"epoch": 4.989355040701315,
|
| 1442 |
+
"eval_accuracy": 0.9814651368049426,
|
| 1443 |
+
"eval_loss": 0.09822726994752884,
|
| 1444 |
+
"eval_runtime": 4.7671,
|
| 1445 |
+
"eval_samples_per_second": 1426.026,
|
| 1446 |
+
"eval_steps_per_second": 44.681,
|
| 1447 |
"step": 1995
|
| 1448 |
},
|
| 1449 |
{
|
| 1450 |
"epoch": 4.989355040701315,
|
| 1451 |
"step": 1995,
|
| 1452 |
+
"total_flos": 2.357895379209216e+18,
|
| 1453 |
+
"train_loss": 0.5734889231528854,
|
| 1454 |
+
"train_runtime": 654.1726,
|
| 1455 |
+
"train_samples_per_second": 390.524,
|
| 1456 |
+
"train_steps_per_second": 3.05
|
| 1457 |
}
|
| 1458 |
],
|
| 1459 |
"logging_steps": 10,
|
|
|
|
| 1473 |
"attributes": {}
|
| 1474 |
}
|
| 1475 |
},
|
| 1476 |
+
"total_flos": 2.357895379209216e+18,
|
| 1477 |
"train_batch_size": 32,
|
| 1478 |
"trial_name": null,
|
| 1479 |
"trial_params": null
|