Training in progress, step 3000, checkpoint
Browse files- checkpoint-3000/config.json +1 -1
- checkpoint-3000/generation_config.json +1 -1
- checkpoint-3000/model.safetensors +1 -1
- checkpoint-3000/optimizer.pt +1 -1
- checkpoint-3000/rng_state.pth +1 -1
- checkpoint-3000/scheduler.pt +1 -1
- checkpoint-3000/trainer_state.json +494 -494
- checkpoint-3000/training_args.bin +1 -1
checkpoint-3000/config.json
CHANGED
|
@@ -84,7 +84,7 @@
|
|
| 84 |
"speech_decoder_prenet_layers": 2,
|
| 85 |
"speech_decoder_prenet_units": 256,
|
| 86 |
"tie_word_embeddings": true,
|
| 87 |
-
"transformers_version": "5.19.0
|
| 88 |
"use_cache": false,
|
| 89 |
"use_guided_attention_loss": false,
|
| 90 |
"vocab_size": 81
|
|
|
|
| 84 |
"speech_decoder_prenet_layers": 2,
|
| 85 |
"speech_decoder_prenet_units": 256,
|
| 86 |
"tie_word_embeddings": true,
|
| 87 |
+
"transformers_version": "5.19.0",
|
| 88 |
"use_cache": false,
|
| 89 |
"use_guided_attention_loss": false,
|
| 90 |
"vocab_size": 81
|
checkpoint-3000/generation_config.json
CHANGED
|
@@ -6,6 +6,6 @@
|
|
| 6 |
"output_attentions": false,
|
| 7 |
"output_hidden_states": false,
|
| 8 |
"pad_token_id": 1,
|
| 9 |
-
"transformers_version": "5.19.0
|
| 10 |
"use_cache": true
|
| 11 |
}
|
|
|
|
| 6 |
"output_attentions": false,
|
| 7 |
"output_hidden_states": false,
|
| 8 |
"pad_token_id": 1,
|
| 9 |
+
"transformers_version": "5.19.0",
|
| 10 |
"use_cache": true
|
| 11 |
}
|
checkpoint-3000/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 577789320
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ca1f78d5daea325ee751c38cb40790aa70b61a798917511e359cc3cfbdd4080
|
| 3 |
size 577789320
|
checkpoint-3000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1155772634
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6df66634e28ef330a0b6aa10bcc0b78ee4926d39a52c19120a0a69dffe67ccd4
|
| 3 |
size 1155772634
|
checkpoint-3000/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b44a4193110a25b149145e19959bfbb591cb0728106d6471fe02dd1923b19258
|
| 3 |
size 14645
|
checkpoint-3000/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ef0503b0da6fb98529d4d87ebb3de74b69fec1e34562a525c2c3c825bff66bc6
|
| 3 |
size 1465
|
checkpoint-3000/trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": 3000,
|
| 3 |
-
"best_metric": 0.
|
| 4 |
"best_model_checkpoint": "/kaggle/working/tts_processed/checkpoint-3000",
|
| 5 |
-
"epoch": 14.
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 3000,
|
| 8 |
"is_hyper_param_search": false,
|
|
@@ -10,891 +10,891 @@
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
"learning_rate": 4.8e-08,
|
| 16 |
-
"loss":
|
| 17 |
"step": 25
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm":
|
| 22 |
"learning_rate": 9.8e-08,
|
| 23 |
-
"loss":
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
"learning_rate": 1.4799999999999998e-07,
|
| 30 |
-
"loss":
|
| 31 |
"step": 75
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm":
|
| 36 |
"learning_rate": 1.98e-07,
|
| 37 |
-
"loss":
|
| 38 |
"step": 100
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm":
|
| 43 |
"learning_rate": 2.48e-07,
|
| 44 |
-
"loss":
|
| 45 |
"step": 125
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm":
|
| 50 |
"learning_rate": 2.98e-07,
|
| 51 |
-
"loss":
|
| 52 |
"step": 150
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm":
|
| 57 |
"learning_rate": 3.4799999999999994e-07,
|
| 58 |
-
"loss":
|
| 59 |
"step": 175
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm":
|
| 64 |
"learning_rate": 3.98e-07,
|
| 65 |
-
"loss":
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 1.
|
| 70 |
-
"grad_norm":
|
| 71 |
"learning_rate": 4.48e-07,
|
| 72 |
-
"loss":
|
| 73 |
"step": 225
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 1.
|
| 77 |
-
"grad_norm":
|
| 78 |
"learning_rate": 4.979999999999999e-07,
|
| 79 |
-
"loss":
|
| 80 |
"step": 250
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 1.
|
| 84 |
-
"grad_norm":
|
| 85 |
"learning_rate": 5.48e-07,
|
| 86 |
-
"loss":
|
| 87 |
"step": 275
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 1.
|
| 91 |
-
"grad_norm":
|
| 92 |
"learning_rate": 5.979999999999999e-07,
|
| 93 |
-
"loss":
|
| 94 |
"step": 300
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 1.
|
| 98 |
-
"grad_norm":
|
| 99 |
"learning_rate": 6.48e-07,
|
| 100 |
-
"loss":
|
| 101 |
"step": 325
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch": 1.
|
| 105 |
-
"grad_norm":
|
| 106 |
"learning_rate": 6.979999999999999e-07,
|
| 107 |
-
"loss":
|
| 108 |
"step": 350
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch": 1.
|
| 112 |
-
"grad_norm":
|
| 113 |
"learning_rate": 7.48e-07,
|
| 114 |
-
"loss":
|
| 115 |
"step": 375
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch": 1.
|
| 119 |
-
"grad_norm":
|
| 120 |
"learning_rate": 7.98e-07,
|
| 121 |
-
"loss":
|
| 122 |
"step": 400
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 2.
|
| 126 |
-
"grad_norm":
|
| 127 |
"learning_rate": 8.48e-07,
|
| 128 |
-
"loss":
|
| 129 |
"step": 425
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 2.
|
| 133 |
-
"grad_norm":
|
| 134 |
"learning_rate": 8.98e-07,
|
| 135 |
-
"loss":
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch": 2.
|
| 140 |
-
"grad_norm":
|
| 141 |
"learning_rate": 9.479999999999999e-07,
|
| 142 |
-
"loss":
|
| 143 |
"step": 475
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch": 2.
|
| 147 |
-
"grad_norm":
|
| 148 |
"learning_rate": 9.98e-07,
|
| 149 |
-
"loss":
|
| 150 |
"step": 500
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch": 2.
|
| 154 |
-
"eval_loss": 1.
|
| 155 |
-
"eval_runtime":
|
| 156 |
-
"eval_samples_per_second":
|
| 157 |
-
"eval_steps_per_second":
|
| 158 |
"step": 500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch": 2.
|
| 162 |
-
"grad_norm":
|
| 163 |
-
"learning_rate": 9.
|
| 164 |
-
"loss":
|
| 165 |
"step": 525
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 2.
|
| 169 |
-
"grad_norm":
|
| 170 |
-
"learning_rate": 9.
|
| 171 |
-
"loss":
|
| 172 |
"step": 550
|
| 173 |
},
|
| 174 |
{
|
| 175 |
-
"epoch": 2.
|
| 176 |
-
"grad_norm":
|
| 177 |
-
"learning_rate": 9.
|
| 178 |
-
"loss":
|
| 179 |
"step": 575
|
| 180 |
},
|
| 181 |
{
|
| 182 |
-
"epoch": 2.
|
| 183 |
-
"grad_norm":
|
| 184 |
-
"learning_rate": 9.
|
| 185 |
-
"loss":
|
| 186 |
"step": 600
|
| 187 |
},
|
| 188 |
{
|
| 189 |
-
"epoch": 3.
|
| 190 |
-
"grad_norm":
|
| 191 |
-
"learning_rate": 9.
|
| 192 |
-
"loss":
|
| 193 |
"step": 625
|
| 194 |
},
|
| 195 |
{
|
| 196 |
-
"epoch": 3.
|
| 197 |
-
"grad_norm":
|
| 198 |
-
"learning_rate": 9.
|
| 199 |
-
"loss":
|
| 200 |
"step": 650
|
| 201 |
},
|
| 202 |
{
|
| 203 |
-
"epoch": 3.
|
| 204 |
-
"grad_norm":
|
| 205 |
-
"learning_rate": 9.
|
| 206 |
-
"loss":
|
| 207 |
"step": 675
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch": 3.
|
| 211 |
-
"grad_norm":
|
| 212 |
-
"learning_rate": 9.
|
| 213 |
-
"loss":
|
| 214 |
"step": 700
|
| 215 |
},
|
| 216 |
{
|
| 217 |
-
"epoch": 3.
|
| 218 |
-
"grad_norm":
|
| 219 |
-
"learning_rate": 9.
|
| 220 |
-
"loss":
|
| 221 |
"step": 725
|
| 222 |
},
|
| 223 |
{
|
| 224 |
-
"epoch": 3.
|
| 225 |
-
"grad_norm":
|
| 226 |
-
"learning_rate": 9.
|
| 227 |
-
"loss":
|
| 228 |
"step": 750
|
| 229 |
},
|
| 230 |
{
|
| 231 |
-
"epoch": 3.
|
| 232 |
-
"grad_norm":
|
| 233 |
-
"learning_rate": 9.
|
| 234 |
-
"loss":
|
| 235 |
"step": 775
|
| 236 |
},
|
| 237 |
{
|
| 238 |
-
"epoch": 3.
|
| 239 |
-
"grad_norm":
|
| 240 |
-
"learning_rate": 9.
|
| 241 |
-
"loss":
|
| 242 |
"step": 800
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch": 4.
|
| 246 |
-
"grad_norm":
|
| 247 |
-
"learning_rate": 9.
|
| 248 |
-
"loss":
|
| 249 |
"step": 825
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch": 4.
|
| 253 |
-
"grad_norm":
|
| 254 |
-
"learning_rate": 9.
|
| 255 |
-
"loss":
|
| 256 |
"step": 850
|
| 257 |
},
|
| 258 |
{
|
| 259 |
-
"epoch": 4.
|
| 260 |
-
"grad_norm":
|
| 261 |
-
"learning_rate": 9.
|
| 262 |
-
"loss":
|
| 263 |
"step": 875
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch": 4.
|
| 267 |
-
"grad_norm":
|
| 268 |
-
"learning_rate": 9.
|
| 269 |
-
"loss":
|
| 270 |
"step": 900
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch": 4.
|
| 274 |
-
"grad_norm":
|
| 275 |
-
"learning_rate": 9.
|
| 276 |
-
"loss":
|
| 277 |
"step": 925
|
| 278 |
},
|
| 279 |
{
|
| 280 |
-
"epoch": 4.
|
| 281 |
-
"grad_norm":
|
| 282 |
-
"learning_rate": 9.
|
| 283 |
-
"loss":
|
| 284 |
"step": 950
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch": 4.
|
| 288 |
-
"grad_norm":
|
| 289 |
-
"learning_rate": 9.
|
| 290 |
-
"loss":
|
| 291 |
"step": 975
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch": 4.
|
| 295 |
-
"grad_norm":
|
| 296 |
-
"learning_rate": 9.
|
| 297 |
-
"loss":
|
| 298 |
"step": 1000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
-
"epoch": 4.
|
| 302 |
-
"eval_loss": 1.
|
| 303 |
-
"eval_runtime":
|
| 304 |
-
"eval_samples_per_second":
|
| 305 |
-
"eval_steps_per_second":
|
| 306 |
"step": 1000
|
| 307 |
},
|
| 308 |
{
|
| 309 |
-
"epoch": 5.
|
| 310 |
-
"grad_norm":
|
| 311 |
-
"learning_rate": 9.
|
| 312 |
-
"loss":
|
| 313 |
"step": 1025
|
| 314 |
},
|
| 315 |
{
|
| 316 |
-
"epoch": 5.
|
| 317 |
-
"grad_norm":
|
| 318 |
-
"learning_rate": 9.
|
| 319 |
-
"loss":
|
| 320 |
"step": 1050
|
| 321 |
},
|
| 322 |
{
|
| 323 |
-
"epoch": 5.
|
| 324 |
-
"grad_norm":
|
| 325 |
-
"learning_rate": 9.
|
| 326 |
-
"loss":
|
| 327 |
"step": 1075
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch": 5.
|
| 331 |
-
"grad_norm":
|
| 332 |
-
"learning_rate": 9.
|
| 333 |
-
"loss":
|
| 334 |
"step": 1100
|
| 335 |
},
|
| 336 |
{
|
| 337 |
-
"epoch": 5.
|
| 338 |
-
"grad_norm":
|
| 339 |
-
"learning_rate": 9.
|
| 340 |
-
"loss":
|
| 341 |
"step": 1125
|
| 342 |
},
|
| 343 |
{
|
| 344 |
-
"epoch": 5.
|
| 345 |
-
"grad_norm":
|
| 346 |
-
"learning_rate": 9.
|
| 347 |
-
"loss":
|
| 348 |
"step": 1150
|
| 349 |
},
|
| 350 |
{
|
| 351 |
-
"epoch": 5.
|
| 352 |
-
"grad_norm":
|
| 353 |
-
"learning_rate": 9.
|
| 354 |
-
"loss":
|
| 355 |
"step": 1175
|
| 356 |
},
|
| 357 |
{
|
| 358 |
-
"epoch": 5.
|
| 359 |
-
"grad_norm":
|
| 360 |
-
"learning_rate": 9.
|
| 361 |
-
"loss":
|
| 362 |
"step": 1200
|
| 363 |
},
|
| 364 |
{
|
| 365 |
-
"epoch": 6.
|
| 366 |
-
"grad_norm":
|
| 367 |
-
"learning_rate": 9.
|
| 368 |
-
"loss":
|
| 369 |
"step": 1225
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 6.
|
| 373 |
-
"grad_norm":
|
| 374 |
-
"learning_rate": 9.
|
| 375 |
-
"loss":
|
| 376 |
"step": 1250
|
| 377 |
},
|
| 378 |
{
|
| 379 |
-
"epoch": 6.
|
| 380 |
-
"grad_norm":
|
| 381 |
-
"learning_rate": 9.
|
| 382 |
-
"loss":
|
| 383 |
"step": 1275
|
| 384 |
},
|
| 385 |
{
|
| 386 |
-
"epoch": 6.
|
| 387 |
-
"grad_norm":
|
| 388 |
-
"learning_rate": 9.
|
| 389 |
-
"loss":
|
| 390 |
"step": 1300
|
| 391 |
},
|
| 392 |
{
|
| 393 |
-
"epoch": 6.
|
| 394 |
-
"grad_norm":
|
| 395 |
-
"learning_rate": 9.
|
| 396 |
-
"loss":
|
| 397 |
"step": 1325
|
| 398 |
},
|
| 399 |
{
|
| 400 |
-
"epoch": 6.
|
| 401 |
-
"grad_norm":
|
| 402 |
-
"learning_rate": 9.
|
| 403 |
-
"loss":
|
| 404 |
"step": 1350
|
| 405 |
},
|
| 406 |
{
|
| 407 |
-
"epoch": 6.
|
| 408 |
-
"grad_norm":
|
| 409 |
-
"learning_rate": 9.
|
| 410 |
-
"loss":
|
| 411 |
"step": 1375
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch": 6.
|
| 415 |
-
"grad_norm":
|
| 416 |
-
"learning_rate": 9.
|
| 417 |
-
"loss":
|
| 418 |
"step": 1400
|
| 419 |
},
|
| 420 |
{
|
| 421 |
-
"epoch": 7.
|
| 422 |
-
"grad_norm":
|
| 423 |
-
"learning_rate": 9.
|
| 424 |
-
"loss":
|
| 425 |
"step": 1425
|
| 426 |
},
|
| 427 |
{
|
| 428 |
-
"epoch": 7.
|
| 429 |
-
"grad_norm":
|
| 430 |
-
"learning_rate": 9.
|
| 431 |
-
"loss":
|
| 432 |
"step": 1450
|
| 433 |
},
|
| 434 |
{
|
| 435 |
-
"epoch": 7.
|
| 436 |
-
"grad_norm":
|
| 437 |
-
"learning_rate": 9.
|
| 438 |
-
"loss":
|
| 439 |
"step": 1475
|
| 440 |
},
|
| 441 |
{
|
| 442 |
-
"epoch": 7.
|
| 443 |
-
"grad_norm":
|
| 444 |
-
"learning_rate": 9.
|
| 445 |
-
"loss":
|
| 446 |
"step": 1500
|
| 447 |
},
|
| 448 |
{
|
| 449 |
-
"epoch": 7.
|
| 450 |
-
"eval_loss": 0.
|
| 451 |
-
"eval_runtime":
|
| 452 |
-
"eval_samples_per_second":
|
| 453 |
-
"eval_steps_per_second":
|
| 454 |
"step": 1500
|
| 455 |
},
|
| 456 |
{
|
| 457 |
-
"epoch": 7.
|
| 458 |
-
"grad_norm":
|
| 459 |
-
"learning_rate": 9.
|
| 460 |
-
"loss":
|
| 461 |
"step": 1525
|
| 462 |
},
|
| 463 |
{
|
| 464 |
-
"epoch": 7.
|
| 465 |
-
"grad_norm":
|
| 466 |
-
"learning_rate": 9.
|
| 467 |
-
"loss":
|
| 468 |
"step": 1550
|
| 469 |
},
|
| 470 |
{
|
| 471 |
-
"epoch": 7.
|
| 472 |
-
"grad_norm":
|
| 473 |
-
"learning_rate": 9.
|
| 474 |
-
"loss":
|
| 475 |
"step": 1575
|
| 476 |
},
|
| 477 |
{
|
| 478 |
-
"epoch": 7.
|
| 479 |
-
"grad_norm":
|
| 480 |
-
"learning_rate": 9.
|
| 481 |
-
"loss":
|
| 482 |
"step": 1600
|
| 483 |
},
|
| 484 |
{
|
| 485 |
-
"epoch": 8.
|
| 486 |
-
"grad_norm":
|
| 487 |
-
"learning_rate": 9.
|
| 488 |
-
"loss":
|
| 489 |
"step": 1625
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch": 8.
|
| 493 |
-
"grad_norm":
|
| 494 |
-
"learning_rate": 9.
|
| 495 |
-
"loss":
|
| 496 |
"step": 1650
|
| 497 |
},
|
| 498 |
{
|
| 499 |
-
"epoch": 8.
|
| 500 |
-
"grad_norm":
|
| 501 |
-
"learning_rate": 9.
|
| 502 |
-
"loss":
|
| 503 |
"step": 1675
|
| 504 |
},
|
| 505 |
{
|
| 506 |
-
"epoch": 8.
|
| 507 |
-
"grad_norm":
|
| 508 |
-
"learning_rate": 9.
|
| 509 |
-
"loss":
|
| 510 |
"step": 1700
|
| 511 |
},
|
| 512 |
{
|
| 513 |
-
"epoch": 8.
|
| 514 |
-
"grad_norm":
|
| 515 |
-
"learning_rate": 9.
|
| 516 |
-
"loss":
|
| 517 |
"step": 1725
|
| 518 |
},
|
| 519 |
{
|
| 520 |
-
"epoch": 8.
|
| 521 |
-
"grad_norm":
|
| 522 |
-
"learning_rate": 9.
|
| 523 |
-
"loss":
|
| 524 |
"step": 1750
|
| 525 |
},
|
| 526 |
{
|
| 527 |
-
"epoch": 8.
|
| 528 |
-
"grad_norm":
|
| 529 |
-
"learning_rate": 9.
|
| 530 |
-
"loss":
|
| 531 |
"step": 1775
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch": 8.
|
| 535 |
-
"grad_norm":
|
| 536 |
-
"learning_rate": 9.
|
| 537 |
-
"loss":
|
| 538 |
"step": 1800
|
| 539 |
},
|
| 540 |
{
|
| 541 |
-
"epoch": 9.
|
| 542 |
-
"grad_norm":
|
| 543 |
-
"learning_rate": 9.
|
| 544 |
-
"loss":
|
| 545 |
"step": 1825
|
| 546 |
},
|
| 547 |
{
|
| 548 |
-
"epoch": 9.
|
| 549 |
-
"grad_norm":
|
| 550 |
-
"learning_rate": 9.
|
| 551 |
-
"loss":
|
| 552 |
"step": 1850
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"epoch": 9.
|
| 556 |
-
"grad_norm":
|
| 557 |
-
"learning_rate": 9.
|
| 558 |
-
"loss":
|
| 559 |
"step": 1875
|
| 560 |
},
|
| 561 |
{
|
| 562 |
-
"epoch": 9.
|
| 563 |
-
"grad_norm":
|
| 564 |
-
"learning_rate": 9.
|
| 565 |
-
"loss":
|
| 566 |
"step": 1900
|
| 567 |
},
|
| 568 |
{
|
| 569 |
-
"epoch": 9.
|
| 570 |
-
"grad_norm":
|
| 571 |
-
"learning_rate": 9.
|
| 572 |
-
"loss":
|
| 573 |
"step": 1925
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch": 9.
|
| 577 |
-
"grad_norm":
|
| 578 |
-
"learning_rate": 9.
|
| 579 |
-
"loss":
|
| 580 |
"step": 1950
|
| 581 |
},
|
| 582 |
{
|
| 583 |
-
"epoch": 9.
|
| 584 |
-
"grad_norm":
|
| 585 |
-
"learning_rate": 9.
|
| 586 |
-
"loss":
|
| 587 |
"step": 1975
|
| 588 |
},
|
| 589 |
{
|
| 590 |
-
"epoch": 9.
|
| 591 |
-
"grad_norm":
|
| 592 |
-
"learning_rate": 9.
|
| 593 |
-
"loss":
|
| 594 |
"step": 2000
|
| 595 |
},
|
| 596 |
{
|
| 597 |
-
"epoch": 9.
|
| 598 |
-
"eval_loss": 0.
|
| 599 |
-
"eval_runtime":
|
| 600 |
-
"eval_samples_per_second":
|
| 601 |
-
"eval_steps_per_second":
|
| 602 |
"step": 2000
|
| 603 |
},
|
| 604 |
{
|
| 605 |
-
"epoch": 10.
|
| 606 |
-
"grad_norm":
|
| 607 |
-
"learning_rate": 9.
|
| 608 |
-
"loss":
|
| 609 |
"step": 2025
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch": 10.
|
| 613 |
-
"grad_norm":
|
| 614 |
-
"learning_rate": 9.
|
| 615 |
-
"loss":
|
| 616 |
"step": 2050
|
| 617 |
},
|
| 618 |
{
|
| 619 |
-
"epoch": 10.
|
| 620 |
-
"grad_norm":
|
| 621 |
-
"learning_rate": 9.
|
| 622 |
-
"loss":
|
| 623 |
"step": 2075
|
| 624 |
},
|
| 625 |
{
|
| 626 |
-
"epoch": 10.
|
| 627 |
-
"grad_norm":
|
| 628 |
-
"learning_rate": 9.
|
| 629 |
-
"loss":
|
| 630 |
"step": 2100
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch": 10.
|
| 634 |
-
"grad_norm":
|
| 635 |
-
"learning_rate": 9.
|
| 636 |
-
"loss":
|
| 637 |
"step": 2125
|
| 638 |
},
|
| 639 |
{
|
| 640 |
-
"epoch": 10.
|
| 641 |
-
"grad_norm":
|
| 642 |
-
"learning_rate": 9.
|
| 643 |
-
"loss":
|
| 644 |
"step": 2150
|
| 645 |
},
|
| 646 |
{
|
| 647 |
-
"epoch": 10.
|
| 648 |
-
"grad_norm":
|
| 649 |
-
"learning_rate": 9.
|
| 650 |
-
"loss":
|
| 651 |
"step": 2175
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch": 10.
|
| 655 |
-
"grad_norm":
|
| 656 |
-
"learning_rate": 9.
|
| 657 |
-
"loss":
|
| 658 |
"step": 2200
|
| 659 |
},
|
| 660 |
{
|
| 661 |
-
"epoch": 11.
|
| 662 |
-
"grad_norm":
|
| 663 |
-
"learning_rate": 9.
|
| 664 |
-
"loss":
|
| 665 |
"step": 2225
|
| 666 |
},
|
| 667 |
{
|
| 668 |
-
"epoch": 11.
|
| 669 |
-
"grad_norm":
|
| 670 |
-
"learning_rate": 9.
|
| 671 |
-
"loss":
|
| 672 |
"step": 2250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
-
"epoch": 11.
|
| 676 |
-
"grad_norm":
|
| 677 |
-
"learning_rate":
|
| 678 |
-
"loss":
|
| 679 |
"step": 2275
|
| 680 |
},
|
| 681 |
{
|
| 682 |
-
"epoch": 11.
|
| 683 |
-
"grad_norm":
|
| 684 |
-
"learning_rate":
|
| 685 |
-
"loss":
|
| 686 |
"step": 2300
|
| 687 |
},
|
| 688 |
{
|
| 689 |
-
"epoch": 11.
|
| 690 |
-
"grad_norm":
|
| 691 |
-
"learning_rate":
|
| 692 |
-
"loss":
|
| 693 |
"step": 2325
|
| 694 |
},
|
| 695 |
{
|
| 696 |
-
"epoch": 11.
|
| 697 |
-
"grad_norm":
|
| 698 |
-
"learning_rate":
|
| 699 |
-
"loss":
|
| 700 |
"step": 2350
|
| 701 |
},
|
| 702 |
{
|
| 703 |
-
"epoch": 11.
|
| 704 |
-
"grad_norm":
|
| 705 |
-
"learning_rate":
|
| 706 |
-
"loss":
|
| 707 |
"step": 2375
|
| 708 |
},
|
| 709 |
{
|
| 710 |
-
"epoch": 11.
|
| 711 |
-
"grad_norm":
|
| 712 |
-
"learning_rate":
|
| 713 |
-
"loss":
|
| 714 |
"step": 2400
|
| 715 |
},
|
| 716 |
{
|
| 717 |
-
"epoch": 12.
|
| 718 |
-
"grad_norm":
|
| 719 |
-
"learning_rate":
|
| 720 |
-
"loss":
|
| 721 |
"step": 2425
|
| 722 |
},
|
| 723 |
{
|
| 724 |
-
"epoch": 12.
|
| 725 |
-
"grad_norm":
|
| 726 |
-
"learning_rate":
|
| 727 |
-
"loss":
|
| 728 |
"step": 2450
|
| 729 |
},
|
| 730 |
{
|
| 731 |
-
"epoch": 12.
|
| 732 |
-
"grad_norm":
|
| 733 |
-
"learning_rate":
|
| 734 |
-
"loss":
|
| 735 |
"step": 2475
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch": 12.
|
| 739 |
-
"grad_norm":
|
| 740 |
-
"learning_rate":
|
| 741 |
-
"loss":
|
| 742 |
"step": 2500
|
| 743 |
},
|
| 744 |
{
|
| 745 |
-
"epoch": 12.
|
| 746 |
-
"eval_loss": 0.
|
| 747 |
-
"eval_runtime":
|
| 748 |
-
"eval_samples_per_second":
|
| 749 |
-
"eval_steps_per_second":
|
| 750 |
"step": 2500
|
| 751 |
},
|
| 752 |
{
|
| 753 |
-
"epoch": 12.
|
| 754 |
-
"grad_norm":
|
| 755 |
-
"learning_rate":
|
| 756 |
-
"loss":
|
| 757 |
"step": 2525
|
| 758 |
},
|
| 759 |
{
|
| 760 |
-
"epoch": 12.
|
| 761 |
-
"grad_norm":
|
| 762 |
-
"learning_rate":
|
| 763 |
-
"loss":
|
| 764 |
"step": 2550
|
| 765 |
},
|
| 766 |
{
|
| 767 |
-
"epoch": 12.
|
| 768 |
-
"grad_norm":
|
| 769 |
-
"learning_rate":
|
| 770 |
-
"loss":
|
| 771 |
"step": 2575
|
| 772 |
},
|
| 773 |
{
|
| 774 |
-
"epoch": 12.
|
| 775 |
-
"grad_norm":
|
| 776 |
-
"learning_rate":
|
| 777 |
-
"loss":
|
| 778 |
"step": 2600
|
| 779 |
},
|
| 780 |
{
|
| 781 |
-
"epoch": 12.
|
| 782 |
-
"grad_norm":
|
| 783 |
-
"learning_rate":
|
| 784 |
-
"loss":
|
| 785 |
"step": 2625
|
| 786 |
},
|
| 787 |
{
|
| 788 |
-
"epoch": 13.
|
| 789 |
-
"grad_norm":
|
| 790 |
-
"learning_rate":
|
| 791 |
-
"loss":
|
| 792 |
"step": 2650
|
| 793 |
},
|
| 794 |
{
|
| 795 |
-
"epoch": 13.
|
| 796 |
-
"grad_norm":
|
| 797 |
-
"learning_rate":
|
| 798 |
-
"loss":
|
| 799 |
"step": 2675
|
| 800 |
},
|
| 801 |
{
|
| 802 |
-
"epoch": 13.
|
| 803 |
-
"grad_norm":
|
| 804 |
-
"learning_rate":
|
| 805 |
-
"loss":
|
| 806 |
"step": 2700
|
| 807 |
},
|
| 808 |
{
|
| 809 |
-
"epoch": 13.
|
| 810 |
-
"grad_norm":
|
| 811 |
-
"learning_rate":
|
| 812 |
-
"loss":
|
| 813 |
"step": 2725
|
| 814 |
},
|
| 815 |
{
|
| 816 |
-
"epoch": 13.
|
| 817 |
-
"grad_norm":
|
| 818 |
-
"learning_rate":
|
| 819 |
-
"loss":
|
| 820 |
"step": 2750
|
| 821 |
},
|
| 822 |
{
|
| 823 |
-
"epoch": 13.
|
| 824 |
-
"grad_norm":
|
| 825 |
-
"learning_rate":
|
| 826 |
-
"loss":
|
| 827 |
"step": 2775
|
| 828 |
},
|
| 829 |
{
|
| 830 |
-
"epoch": 13.
|
| 831 |
-
"grad_norm":
|
| 832 |
-
"learning_rate":
|
| 833 |
-
"loss":
|
| 834 |
"step": 2800
|
| 835 |
},
|
| 836 |
{
|
| 837 |
-
"epoch": 13.
|
| 838 |
-
"grad_norm":
|
| 839 |
-
"learning_rate":
|
| 840 |
-
"loss":
|
| 841 |
"step": 2825
|
| 842 |
},
|
| 843 |
{
|
| 844 |
-
"epoch": 14.
|
| 845 |
-
"grad_norm":
|
| 846 |
-
"learning_rate":
|
| 847 |
-
"loss":
|
| 848 |
"step": 2850
|
| 849 |
},
|
| 850 |
{
|
| 851 |
-
"epoch": 14.
|
| 852 |
-
"grad_norm":
|
| 853 |
-
"learning_rate":
|
| 854 |
-
"loss":
|
| 855 |
"step": 2875
|
| 856 |
},
|
| 857 |
{
|
| 858 |
-
"epoch": 14.
|
| 859 |
-
"grad_norm":
|
| 860 |
-
"learning_rate":
|
| 861 |
-
"loss":
|
| 862 |
"step": 2900
|
| 863 |
},
|
| 864 |
{
|
| 865 |
-
"epoch": 14.
|
| 866 |
-
"grad_norm":
|
| 867 |
-
"learning_rate":
|
| 868 |
-
"loss":
|
| 869 |
"step": 2925
|
| 870 |
},
|
| 871 |
{
|
| 872 |
-
"epoch": 14.
|
| 873 |
-
"grad_norm":
|
| 874 |
-
"learning_rate":
|
| 875 |
-
"loss":
|
| 876 |
"step": 2950
|
| 877 |
},
|
| 878 |
{
|
| 879 |
-
"epoch": 14.
|
| 880 |
-
"grad_norm":
|
| 881 |
-
"learning_rate":
|
| 882 |
-
"loss":
|
| 883 |
"step": 2975
|
| 884 |
},
|
| 885 |
{
|
| 886 |
-
"epoch": 14.
|
| 887 |
-
"grad_norm":
|
| 888 |
-
"learning_rate":
|
| 889 |
-
"loss":
|
| 890 |
"step": 3000
|
| 891 |
},
|
| 892 |
{
|
| 893 |
-
"epoch": 14.
|
| 894 |
-
"eval_loss": 0.
|
| 895 |
-
"eval_runtime":
|
| 896 |
-
"eval_samples_per_second":
|
| 897 |
-
"eval_steps_per_second":
|
| 898 |
"step": 3000
|
| 899 |
}
|
| 900 |
],
|
|
@@ -924,8 +924,8 @@
|
|
| 924 |
"attributes": {}
|
| 925 |
}
|
| 926 |
},
|
| 927 |
-
"total_flos": 1.
|
| 928 |
-
"train_batch_size":
|
| 929 |
"trial_name": null,
|
| 930 |
"trial_params": null
|
| 931 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": 3000,
|
| 3 |
+
"best_metric": 0.920803427696228,
|
| 4 |
"best_model_checkpoint": "/kaggle/working/tts_processed/checkpoint-3000",
|
| 5 |
+
"epoch": 14.851485148514852,
|
| 6 |
"eval_steps": 500,
|
| 7 |
"global_step": 3000,
|
| 8 |
"is_hyper_param_search": false,
|
|
|
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.12376237623762376,
|
| 14 |
+
"grad_norm": 363.43377685546875,
|
| 15 |
"learning_rate": 4.8e-08,
|
| 16 |
+
"loss": 10.89256103515625,
|
| 17 |
"step": 25
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.24752475247524752,
|
| 21 |
+
"grad_norm": 336.6938171386719,
|
| 22 |
"learning_rate": 9.8e-08,
|
| 23 |
+
"loss": 10.47541259765625,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.3712871287128713,
|
| 28 |
+
"grad_norm": 255.0070343017578,
|
| 29 |
"learning_rate": 1.4799999999999998e-07,
|
| 30 |
+
"loss": 10.08507568359375,
|
| 31 |
"step": 75
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.49504950495049505,
|
| 35 |
+
"grad_norm": 247.54541015625,
|
| 36 |
"learning_rate": 1.98e-07,
|
| 37 |
+
"loss": 9.282888793945313,
|
| 38 |
"step": 100
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.6188118811881188,
|
| 42 |
+
"grad_norm": 124.208740234375,
|
| 43 |
"learning_rate": 2.48e-07,
|
| 44 |
+
"loss": 8.717689208984375,
|
| 45 |
"step": 125
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.7425742574257426,
|
| 49 |
+
"grad_norm": 194.5498809814453,
|
| 50 |
"learning_rate": 2.98e-07,
|
| 51 |
+
"loss": 8.212156372070313,
|
| 52 |
"step": 150
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.8663366336633663,
|
| 56 |
+
"grad_norm": 143.2777557373047,
|
| 57 |
"learning_rate": 3.4799999999999994e-07,
|
| 58 |
+
"loss": 7.8293896484375,
|
| 59 |
"step": 175
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.9900990099009901,
|
| 63 |
+
"grad_norm": 55.50076675415039,
|
| 64 |
"learning_rate": 3.98e-07,
|
| 65 |
+
"loss": 7.3809716796875,
|
| 66 |
"step": 200
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 1.113861386138614,
|
| 70 |
+
"grad_norm": 69.43047332763672,
|
| 71 |
"learning_rate": 4.48e-07,
|
| 72 |
+
"loss": 6.937587890625,
|
| 73 |
"step": 225
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 1.2376237623762376,
|
| 77 |
+
"grad_norm": 49.69321060180664,
|
| 78 |
"learning_rate": 4.979999999999999e-07,
|
| 79 |
+
"loss": 6.8110888671875,
|
| 80 |
"step": 250
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 1.3613861386138613,
|
| 84 |
+
"grad_norm": 26.72427749633789,
|
| 85 |
"learning_rate": 5.48e-07,
|
| 86 |
+
"loss": 6.522671508789062,
|
| 87 |
"step": 275
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 1.4851485148514851,
|
| 91 |
+
"grad_norm": 51.36345672607422,
|
| 92 |
"learning_rate": 5.979999999999999e-07,
|
| 93 |
+
"loss": 6.398966064453125,
|
| 94 |
"step": 300
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 1.608910891089109,
|
| 98 |
+
"grad_norm": 38.85439682006836,
|
| 99 |
"learning_rate": 6.48e-07,
|
| 100 |
+
"loss": 6.198662109375,
|
| 101 |
"step": 325
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 1.7326732673267327,
|
| 105 |
+
"grad_norm": 27.508665084838867,
|
| 106 |
"learning_rate": 6.979999999999999e-07,
|
| 107 |
+
"loss": 6.075740356445312,
|
| 108 |
"step": 350
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 1.8564356435643563,
|
| 112 |
+
"grad_norm": 36.8642463684082,
|
| 113 |
"learning_rate": 7.48e-07,
|
| 114 |
+
"loss": 6.0834722900390625,
|
| 115 |
"step": 375
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 1.9801980198019802,
|
| 119 |
+
"grad_norm": 27.07192039489746,
|
| 120 |
"learning_rate": 7.98e-07,
|
| 121 |
+
"loss": 5.9166015625,
|
| 122 |
"step": 400
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 2.103960396039604,
|
| 126 |
+
"grad_norm": 22.465463638305664,
|
| 127 |
"learning_rate": 8.48e-07,
|
| 128 |
+
"loss": 5.795056762695313,
|
| 129 |
"step": 425
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 2.227722772277228,
|
| 133 |
+
"grad_norm": 34.86723709106445,
|
| 134 |
"learning_rate": 8.98e-07,
|
| 135 |
+
"loss": 5.741965942382812,
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 2.3514851485148514,
|
| 140 |
+
"grad_norm": 22.37932777404785,
|
| 141 |
"learning_rate": 9.479999999999999e-07,
|
| 142 |
+
"loss": 5.716455688476563,
|
| 143 |
"step": 475
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 2.4752475247524752,
|
| 147 |
+
"grad_norm": 29.54213523864746,
|
| 148 |
"learning_rate": 9.98e-07,
|
| 149 |
+
"loss": 5.599214477539062,
|
| 150 |
"step": 500
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 2.4752475247524752,
|
| 154 |
+
"eval_loss": 1.2534884214401245,
|
| 155 |
+
"eval_runtime": 82.4141,
|
| 156 |
+
"eval_samples_per_second": 19.608,
|
| 157 |
+
"eval_steps_per_second": 4.902,
|
| 158 |
"step": 500
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 2.599009900990099,
|
| 162 |
+
"grad_norm": 21.54619026184082,
|
| 163 |
+
"learning_rate": 9.999953592789052e-07,
|
| 164 |
+
"loss": 5.526549072265625,
|
| 165 |
"step": 525
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 2.7227722772277225,
|
| 169 |
+
"grad_norm": 31.9982967376709,
|
| 170 |
+
"learning_rate": 9.999806557001091e-07,
|
| 171 |
+
"loss": 5.38673828125,
|
| 172 |
"step": 550
|
| 173 |
},
|
| 174 |
{
|
| 175 |
+
"epoch": 2.8465346534653464,
|
| 176 |
+
"grad_norm": 19.33993148803711,
|
| 177 |
+
"learning_rate": 9.999558815029513e-07,
|
| 178 |
+
"loss": 5.291504516601562,
|
| 179 |
"step": 575
|
| 180 |
},
|
| 181 |
{
|
| 182 |
+
"epoch": 2.9702970297029703,
|
| 183 |
+
"grad_norm": 18.838518142700195,
|
| 184 |
+
"learning_rate": 9.999210371864342e-07,
|
| 185 |
+
"loss": 5.345247192382812,
|
| 186 |
"step": 600
|
| 187 |
},
|
| 188 |
{
|
| 189 |
+
"epoch": 3.094059405940594,
|
| 190 |
+
"grad_norm": 19.941740036010742,
|
| 191 |
+
"learning_rate": 9.998761234523925e-07,
|
| 192 |
+
"loss": 5.081980285644531,
|
| 193 |
"step": 625
|
| 194 |
},
|
| 195 |
{
|
| 196 |
+
"epoch": 3.217821782178218,
|
| 197 |
+
"grad_norm": 24.84575080871582,
|
| 198 |
+
"learning_rate": 9.998211412054793e-07,
|
| 199 |
+
"loss": 5.025005493164063,
|
| 200 |
"step": 650
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"epoch": 3.3415841584158414,
|
| 204 |
+
"grad_norm": 24.532045364379883,
|
| 205 |
+
"learning_rate": 9.99756091553148e-07,
|
| 206 |
+
"loss": 5.004798889160156,
|
| 207 |
"step": 675
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 3.4653465346534653,
|
| 211 |
+
"grad_norm": 21.20513916015625,
|
| 212 |
+
"learning_rate": 9.996809758056298e-07,
|
| 213 |
+
"loss": 4.870877685546875,
|
| 214 |
"step": 700
|
| 215 |
},
|
| 216 |
{
|
| 217 |
+
"epoch": 3.589108910891089,
|
| 218 |
+
"grad_norm": 16.589248657226562,
|
| 219 |
+
"learning_rate": 9.995957954759072e-07,
|
| 220 |
+
"loss": 4.95923095703125,
|
| 221 |
"step": 725
|
| 222 |
},
|
| 223 |
{
|
| 224 |
+
"epoch": 3.7128712871287126,
|
| 225 |
+
"grad_norm": 26.863059997558594,
|
| 226 |
+
"learning_rate": 9.995005522796837e-07,
|
| 227 |
+
"loss": 4.81072265625,
|
| 228 |
"step": 750
|
| 229 |
},
|
| 230 |
{
|
| 231 |
+
"epoch": 3.8366336633663365,
|
| 232 |
+
"grad_norm": 14.973981857299805,
|
| 233 |
+
"learning_rate": 9.993952481353493e-07,
|
| 234 |
+
"loss": 4.8048876953125,
|
| 235 |
"step": 775
|
| 236 |
},
|
| 237 |
{
|
| 238 |
+
"epoch": 3.9603960396039604,
|
| 239 |
+
"grad_norm": 18.018110275268555,
|
| 240 |
+
"learning_rate": 9.992798851639418e-07,
|
| 241 |
+
"loss": 4.770014343261718,
|
| 242 |
"step": 800
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 4.084158415841584,
|
| 246 |
+
"grad_norm": 15.32919979095459,
|
| 247 |
+
"learning_rate": 9.991544656891043e-07,
|
| 248 |
+
"loss": 4.663901672363282,
|
| 249 |
"step": 825
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 4.207920792079208,
|
| 253 |
+
"grad_norm": 14.925799369812012,
|
| 254 |
+
"learning_rate": 9.990189922370373e-07,
|
| 255 |
+
"loss": 4.614352111816406,
|
| 256 |
"step": 850
|
| 257 |
},
|
| 258 |
{
|
| 259 |
+
"epoch": 4.3316831683168315,
|
| 260 |
+
"grad_norm": 14.5320405960083,
|
| 261 |
+
"learning_rate": 9.988734675364496e-07,
|
| 262 |
+
"loss": 4.607881774902344,
|
| 263 |
"step": 875
|
| 264 |
},
|
| 265 |
{
|
| 266 |
+
"epoch": 4.455445544554456,
|
| 267 |
+
"grad_norm": 17.06329345703125,
|
| 268 |
+
"learning_rate": 9.987178945185019e-07,
|
| 269 |
+
"loss": 4.582254028320312,
|
| 270 |
"step": 900
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 4.579207920792079,
|
| 274 |
+
"grad_norm": 16.860803604125977,
|
| 275 |
+
"learning_rate": 9.985522763167486e-07,
|
| 276 |
+
"loss": 4.6025848388671875,
|
| 277 |
"step": 925
|
| 278 |
},
|
| 279 |
{
|
| 280 |
+
"epoch": 4.702970297029703,
|
| 281 |
+
"grad_norm": 12.743483543395996,
|
| 282 |
+
"learning_rate": 9.983766162670739e-07,
|
| 283 |
+
"loss": 4.603981628417968,
|
| 284 |
"step": 950
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 4.826732673267327,
|
| 288 |
+
"grad_norm": 11.648216247558594,
|
| 289 |
+
"learning_rate": 9.981909179076255e-07,
|
| 290 |
+
"loss": 4.5333544921875,
|
| 291 |
"step": 975
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 4.9504950495049505,
|
| 295 |
+
"grad_norm": 12.38879108428955,
|
| 296 |
+
"learning_rate": 9.979951849787429e-07,
|
| 297 |
+
"loss": 4.538429565429688,
|
| 298 |
"step": 1000
|
| 299 |
},
|
| 300 |
{
|
| 301 |
+
"epoch": 4.9504950495049505,
|
| 302 |
+
"eval_loss": 1.0112695693969727,
|
| 303 |
+
"eval_runtime": 82.132,
|
| 304 |
+
"eval_samples_per_second": 19.676,
|
| 305 |
+
"eval_steps_per_second": 4.919,
|
| 306 |
"step": 1000
|
| 307 |
},
|
| 308 |
{
|
| 309 |
+
"epoch": 5.074257425742574,
|
| 310 |
+
"grad_norm": 39.850196838378906,
|
| 311 |
+
"learning_rate": 9.977894214228815e-07,
|
| 312 |
+
"loss": 4.55607666015625,
|
| 313 |
"step": 1025
|
| 314 |
},
|
| 315 |
{
|
| 316 |
+
"epoch": 5.198019801980198,
|
| 317 |
+
"grad_norm": 13.423649787902832,
|
| 318 |
+
"learning_rate": 9.975736313845345e-07,
|
| 319 |
+
"loss": 4.43293212890625,
|
| 320 |
"step": 1050
|
| 321 |
},
|
| 322 |
{
|
| 323 |
+
"epoch": 5.321782178217822,
|
| 324 |
+
"grad_norm": 15.34720516204834,
|
| 325 |
+
"learning_rate": 9.97347819210148e-07,
|
| 326 |
+
"loss": 4.4848849487304685,
|
| 327 |
"step": 1075
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 5.445544554455446,
|
| 331 |
+
"grad_norm": 10.895915031433105,
|
| 332 |
+
"learning_rate": 9.971119894480344e-07,
|
| 333 |
+
"loss": 4.418946228027344,
|
| 334 |
"step": 1100
|
| 335 |
},
|
| 336 |
{
|
| 337 |
+
"epoch": 5.569306930693069,
|
| 338 |
+
"grad_norm": 13.77745532989502,
|
| 339 |
+
"learning_rate": 9.96866146848281e-07,
|
| 340 |
+
"loss": 4.675699768066406,
|
| 341 |
"step": 1125
|
| 342 |
},
|
| 343 |
{
|
| 344 |
+
"epoch": 5.693069306930693,
|
| 345 |
+
"grad_norm": 21.53974723815918,
|
| 346 |
+
"learning_rate": 9.96610296362653e-07,
|
| 347 |
+
"loss": 4.483356628417969,
|
| 348 |
"step": 1150
|
| 349 |
},
|
| 350 |
{
|
| 351 |
+
"epoch": 5.816831683168317,
|
| 352 |
+
"grad_norm": 14.47573184967041,
|
| 353 |
+
"learning_rate": 9.963444431444948e-07,
|
| 354 |
+
"loss": 4.555756530761719,
|
| 355 |
"step": 1175
|
| 356 |
},
|
| 357 |
{
|
| 358 |
+
"epoch": 5.9405940594059405,
|
| 359 |
+
"grad_norm": 11.701032638549805,
|
| 360 |
+
"learning_rate": 9.960685925486265e-07,
|
| 361 |
+
"loss": 4.419072265625,
|
| 362 |
"step": 1200
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 6.064356435643564,
|
| 366 |
+
"grad_norm": 17.197729110717773,
|
| 367 |
+
"learning_rate": 9.957827501312352e-07,
|
| 368 |
+
"loss": 4.375239562988281,
|
| 369 |
"step": 1225
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 6.188118811881188,
|
| 373 |
+
"grad_norm": 12.936083793640137,
|
| 374 |
+
"learning_rate": 9.954869216497633e-07,
|
| 375 |
+
"loss": 4.3699853515625,
|
| 376 |
"step": 1250
|
| 377 |
},
|
| 378 |
{
|
| 379 |
+
"epoch": 6.311881188118812,
|
| 380 |
+
"grad_norm": 15.114112854003906,
|
| 381 |
+
"learning_rate": 9.95181113062793e-07,
|
| 382 |
+
"loss": 4.41375732421875,
|
| 383 |
"step": 1275
|
| 384 |
},
|
| 385 |
{
|
| 386 |
+
"epoch": 6.435643564356436,
|
| 387 |
+
"grad_norm": 17.72418785095215,
|
| 388 |
+
"learning_rate": 9.948653305299256e-07,
|
| 389 |
+
"loss": 4.430754089355469,
|
| 390 |
"step": 1300
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 6.5594059405940595,
|
| 394 |
+
"grad_norm": 18.342178344726562,
|
| 395 |
+
"learning_rate": 9.945395804116578e-07,
|
| 396 |
+
"loss": 4.4040380859375,
|
| 397 |
"step": 1325
|
| 398 |
},
|
| 399 |
{
|
| 400 |
+
"epoch": 6.683168316831683,
|
| 401 |
+
"grad_norm": 12.649271965026855,
|
| 402 |
+
"learning_rate": 9.94203869269254e-07,
|
| 403 |
+
"loss": 4.3536474609375,
|
| 404 |
"step": 1350
|
| 405 |
},
|
| 406 |
{
|
| 407 |
+
"epoch": 6.806930693069307,
|
| 408 |
+
"grad_norm": 32.92485809326172,
|
| 409 |
+
"learning_rate": 9.938582038646131e-07,
|
| 410 |
+
"loss": 4.349139709472656,
|
| 411 |
"step": 1375
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 6.930693069306931,
|
| 415 |
+
"grad_norm": 11.99206829071045,
|
| 416 |
+
"learning_rate": 9.935025911601334e-07,
|
| 417 |
+
"loss": 4.317200317382812,
|
| 418 |
"step": 1400
|
| 419 |
},
|
| 420 |
{
|
| 421 |
+
"epoch": 7.054455445544554,
|
| 422 |
+
"grad_norm": 15.798395156860352,
|
| 423 |
+
"learning_rate": 9.931370383185716e-07,
|
| 424 |
+
"loss": 4.390576782226563,
|
| 425 |
"step": 1425
|
| 426 |
},
|
| 427 |
{
|
| 428 |
+
"epoch": 7.178217821782178,
|
| 429 |
+
"grad_norm": 12.485000610351562,
|
| 430 |
+
"learning_rate": 9.92761552702899e-07,
|
| 431 |
+
"loss": 4.415244750976562,
|
| 432 |
"step": 1450
|
| 433 |
},
|
| 434 |
{
|
| 435 |
+
"epoch": 7.301980198019802,
|
| 436 |
+
"grad_norm": 14.873079299926758,
|
| 437 |
+
"learning_rate": 9.923761418761526e-07,
|
| 438 |
+
"loss": 4.3524868774414065,
|
| 439 |
"step": 1475
|
| 440 |
},
|
| 441 |
{
|
| 442 |
+
"epoch": 7.425742574257426,
|
| 443 |
+
"grad_norm": 13.601946830749512,
|
| 444 |
+
"learning_rate": 9.919808136012837e-07,
|
| 445 |
+
"loss": 4.294336547851563,
|
| 446 |
"step": 1500
|
| 447 |
},
|
| 448 |
{
|
| 449 |
+
"epoch": 7.425742574257426,
|
| 450 |
+
"eval_loss": 0.9704130291938782,
|
| 451 |
+
"eval_runtime": 86.4979,
|
| 452 |
+
"eval_samples_per_second": 18.683,
|
| 453 |
+
"eval_steps_per_second": 4.671,
|
| 454 |
"step": 1500
|
| 455 |
},
|
| 456 |
{
|
| 457 |
+
"epoch": 7.5495049504950495,
|
| 458 |
+
"grad_norm": 15.53011417388916,
|
| 459 |
+
"learning_rate": 9.915755758410008e-07,
|
| 460 |
+
"loss": 4.254639587402344,
|
| 461 |
"step": 1525
|
| 462 |
},
|
| 463 |
{
|
| 464 |
+
"epoch": 7.673267326732673,
|
| 465 |
+
"grad_norm": 11.795248985290527,
|
| 466 |
+
"learning_rate": 9.91160436757609e-07,
|
| 467 |
+
"loss": 4.307816467285156,
|
| 468 |
"step": 1550
|
| 469 |
},
|
| 470 |
{
|
| 471 |
+
"epoch": 7.797029702970297,
|
| 472 |
+
"grad_norm": 15.00704574584961,
|
| 473 |
+
"learning_rate": 9.907354047128464e-07,
|
| 474 |
+
"loss": 4.349948120117188,
|
| 475 |
"step": 1575
|
| 476 |
},
|
| 477 |
{
|
| 478 |
+
"epoch": 7.920792079207921,
|
| 479 |
+
"grad_norm": 13.17168140411377,
|
| 480 |
+
"learning_rate": 9.903004882677155e-07,
|
| 481 |
+
"loss": 4.2257516479492185,
|
| 482 |
"step": 1600
|
| 483 |
},
|
| 484 |
{
|
| 485 |
+
"epoch": 8.044554455445544,
|
| 486 |
+
"grad_norm": 13.770243644714355,
|
| 487 |
+
"learning_rate": 9.898556961823103e-07,
|
| 488 |
+
"loss": 4.310787048339844,
|
| 489 |
"step": 1625
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 8.168316831683168,
|
| 493 |
+
"grad_norm": 13.734089851379395,
|
| 494 |
+
"learning_rate": 9.894010374156398e-07,
|
| 495 |
+
"loss": 4.148223266601563,
|
| 496 |
"step": 1650
|
| 497 |
},
|
| 498 |
{
|
| 499 |
+
"epoch": 8.292079207920793,
|
| 500 |
+
"grad_norm": 15.90095329284668,
|
| 501 |
+
"learning_rate": 9.889365211254485e-07,
|
| 502 |
+
"loss": 4.3510107421875,
|
| 503 |
"step": 1675
|
| 504 |
},
|
| 505 |
{
|
| 506 |
+
"epoch": 8.415841584158416,
|
| 507 |
+
"grad_norm": 12.133323669433594,
|
| 508 |
+
"learning_rate": 9.884621566680306e-07,
|
| 509 |
+
"loss": 4.337423400878906,
|
| 510 |
"step": 1700
|
| 511 |
},
|
| 512 |
{
|
| 513 |
+
"epoch": 8.53960396039604,
|
| 514 |
+
"grad_norm": 50.35639190673828,
|
| 515 |
+
"learning_rate": 9.87977953598043e-07,
|
| 516 |
+
"loss": 4.3044869995117185,
|
| 517 |
"step": 1725
|
| 518 |
},
|
| 519 |
{
|
| 520 |
+
"epoch": 8.663366336633663,
|
| 521 |
+
"grad_norm": 12.457736015319824,
|
| 522 |
+
"learning_rate": 9.874839216683112e-07,
|
| 523 |
+
"loss": 4.3085205078125,
|
| 524 |
"step": 1750
|
| 525 |
},
|
| 526 |
{
|
| 527 |
+
"epoch": 8.787128712871286,
|
| 528 |
+
"grad_norm": 14.483731269836426,
|
| 529 |
+
"learning_rate": 9.869800708296346e-07,
|
| 530 |
+
"loss": 4.1831143188476565,
|
| 531 |
"step": 1775
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 8.910891089108912,
|
| 535 |
+
"grad_norm": 11.174741744995117,
|
| 536 |
+
"learning_rate": 9.864664112305846e-07,
|
| 537 |
+
"loss": 4.162017822265625,
|
| 538 |
"step": 1800
|
| 539 |
},
|
| 540 |
{
|
| 541 |
+
"epoch": 9.034653465346535,
|
| 542 |
+
"grad_norm": 10.312405586242676,
|
| 543 |
+
"learning_rate": 9.859429532173012e-07,
|
| 544 |
+
"loss": 4.320256958007812,
|
| 545 |
"step": 1825
|
| 546 |
},
|
| 547 |
{
|
| 548 |
+
"epoch": 9.158415841584159,
|
| 549 |
+
"grad_norm": 13.814125061035156,
|
| 550 |
+
"learning_rate": 9.85409707333284e-07,
|
| 551 |
+
"loss": 4.314511413574219,
|
| 552 |
"step": 1850
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"epoch": 9.282178217821782,
|
| 556 |
+
"grad_norm": 11.276887893676758,
|
| 557 |
+
"learning_rate": 9.848666843191804e-07,
|
| 558 |
+
"loss": 4.327219543457031,
|
| 559 |
"step": 1875
|
| 560 |
},
|
| 561 |
{
|
| 562 |
+
"epoch": 9.405940594059405,
|
| 563 |
+
"grad_norm": 13.48475170135498,
|
| 564 |
+
"learning_rate": 9.84313895112568e-07,
|
| 565 |
+
"loss": 4.2580517578125,
|
| 566 |
"step": 1900
|
| 567 |
},
|
| 568 |
{
|
| 569 |
+
"epoch": 9.52970297029703,
|
| 570 |
+
"grad_norm": 13.70500659942627,
|
| 571 |
+
"learning_rate": 9.837513508477365e-07,
|
| 572 |
+
"loss": 4.269812316894531,
|
| 573 |
"step": 1925
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 9.653465346534654,
|
| 577 |
+
"grad_norm": 12.701525688171387,
|
| 578 |
+
"learning_rate": 9.831790628554612e-07,
|
| 579 |
+
"loss": 4.257364196777344,
|
| 580 |
"step": 1950
|
| 581 |
},
|
| 582 |
{
|
| 583 |
+
"epoch": 9.777227722772277,
|
| 584 |
+
"grad_norm": 11.086860656738281,
|
| 585 |
+
"learning_rate": 9.82597042662776e-07,
|
| 586 |
+
"loss": 4.120067443847656,
|
| 587 |
"step": 1975
|
| 588 |
},
|
| 589 |
{
|
| 590 |
+
"epoch": 9.900990099009901,
|
| 591 |
+
"grad_norm": 14.811179161071777,
|
| 592 |
+
"learning_rate": 9.820053019927408e-07,
|
| 593 |
+
"loss": 4.13207275390625,
|
| 594 |
"step": 2000
|
| 595 |
},
|
| 596 |
{
|
| 597 |
+
"epoch": 9.900990099009901,
|
| 598 |
+
"eval_loss": 0.9480290412902832,
|
| 599 |
+
"eval_runtime": 81.9519,
|
| 600 |
+
"eval_samples_per_second": 19.719,
|
| 601 |
+
"eval_steps_per_second": 4.93,
|
| 602 |
"step": 2000
|
| 603 |
},
|
| 604 |
{
|
| 605 |
+
"epoch": 10.024752475247524,
|
| 606 |
+
"grad_norm": 13.266257286071777,
|
| 607 |
+
"learning_rate": 9.814038527642059e-07,
|
| 608 |
+
"loss": 4.281947021484375,
|
| 609 |
"step": 2025
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 10.148514851485148,
|
| 613 |
+
"grad_norm": 13.618546485900879,
|
| 614 |
+
"learning_rate": 9.80792707091571e-07,
|
| 615 |
+
"loss": 4.166450500488281,
|
| 616 |
"step": 2050
|
| 617 |
},
|
| 618 |
{
|
| 619 |
+
"epoch": 10.272277227722773,
|
| 620 |
+
"grad_norm": 12.208344459533691,
|
| 621 |
+
"learning_rate": 9.801718772845422e-07,
|
| 622 |
+
"loss": 4.151670227050781,
|
| 623 |
"step": 2075
|
| 624 |
},
|
| 625 |
{
|
| 626 |
+
"epoch": 10.396039603960396,
|
| 627 |
+
"grad_norm": 12.220300674438477,
|
| 628 |
+
"learning_rate": 9.795413758478835e-07,
|
| 629 |
+
"loss": 4.250581359863281,
|
| 630 |
"step": 2100
|
| 631 |
},
|
| 632 |
{
|
| 633 |
+
"epoch": 10.51980198019802,
|
| 634 |
+
"grad_norm": 10.937910079956055,
|
| 635 |
+
"learning_rate": 9.789012154811646e-07,
|
| 636 |
+
"loss": 4.219160461425782,
|
| 637 |
"step": 2125
|
| 638 |
},
|
| 639 |
{
|
| 640 |
+
"epoch": 10.643564356435643,
|
| 641 |
+
"grad_norm": 12.94577407836914,
|
| 642 |
+
"learning_rate": 9.78251409078506e-07,
|
| 643 |
+
"loss": 4.256142272949218,
|
| 644 |
"step": 2150
|
| 645 |
},
|
| 646 |
{
|
| 647 |
+
"epoch": 10.767326732673267,
|
| 648 |
+
"grad_norm": 12.101743698120117,
|
| 649 |
+
"learning_rate": 9.77591969728319e-07,
|
| 650 |
+
"loss": 4.145942077636719,
|
| 651 |
"step": 2175
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 10.891089108910892,
|
| 655 |
+
"grad_norm": 15.079996109008789,
|
| 656 |
+
"learning_rate": 9.769229107130408e-07,
|
| 657 |
+
"loss": 4.2266848754882815,
|
| 658 |
"step": 2200
|
| 659 |
},
|
| 660 |
{
|
| 661 |
+
"epoch": 11.014851485148515,
|
| 662 |
+
"grad_norm": 13.785998344421387,
|
| 663 |
+
"learning_rate": 9.762442455088694e-07,
|
| 664 |
+
"loss": 4.173150634765625,
|
| 665 |
"step": 2225
|
| 666 |
},
|
| 667 |
{
|
| 668 |
+
"epoch": 11.138613861386139,
|
| 669 |
+
"grad_norm": 33.18384552001953,
|
| 670 |
+
"learning_rate": 9.7555598778549e-07,
|
| 671 |
+
"loss": 4.193008728027344,
|
| 672 |
"step": 2250
|
| 673 |
},
|
| 674 |
{
|
| 675 |
+
"epoch": 11.262376237623762,
|
| 676 |
+
"grad_norm": 17.419424057006836,
|
| 677 |
+
"learning_rate": 9.748581514058005e-07,
|
| 678 |
+
"loss": 4.254207153320312,
|
| 679 |
"step": 2275
|
| 680 |
},
|
| 681 |
{
|
| 682 |
+
"epoch": 11.386138613861386,
|
| 683 |
+
"grad_norm": 11.025071144104004,
|
| 684 |
+
"learning_rate": 9.741507504256326e-07,
|
| 685 |
+
"loss": 4.130965270996094,
|
| 686 |
"step": 2300
|
| 687 |
},
|
| 688 |
{
|
| 689 |
+
"epoch": 11.509900990099009,
|
| 690 |
+
"grad_norm": 11.500884056091309,
|
| 691 |
+
"learning_rate": 9.734337990934681e-07,
|
| 692 |
+
"loss": 4.12019775390625,
|
| 693 |
"step": 2325
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 11.633663366336634,
|
| 697 |
+
"grad_norm": 13.27956771850586,
|
| 698 |
+
"learning_rate": 9.727073118501523e-07,
|
| 699 |
+
"loss": 4.216612854003906,
|
| 700 |
"step": 2350
|
| 701 |
},
|
| 702 |
{
|
| 703 |
+
"epoch": 11.757425742574258,
|
| 704 |
+
"grad_norm": 16.143634796142578,
|
| 705 |
+
"learning_rate": 9.719713033286028e-07,
|
| 706 |
+
"loss": 4.163922424316406,
|
| 707 |
"step": 2375
|
| 708 |
},
|
| 709 |
{
|
| 710 |
+
"epoch": 11.881188118811881,
|
| 711 |
+
"grad_norm": 14.004642486572266,
|
| 712 |
+
"learning_rate": 9.712257883535152e-07,
|
| 713 |
+
"loss": 4.1760952758789065,
|
| 714 |
"step": 2400
|
| 715 |
},
|
| 716 |
{
|
| 717 |
+
"epoch": 12.004950495049505,
|
| 718 |
+
"grad_norm": 11.351595878601074,
|
| 719 |
+
"learning_rate": 9.704707819410634e-07,
|
| 720 |
+
"loss": 4.229373168945313,
|
| 721 |
"step": 2425
|
| 722 |
},
|
| 723 |
{
|
| 724 |
+
"epoch": 12.128712871287128,
|
| 725 |
+
"grad_norm": 9.855143547058105,
|
| 726 |
+
"learning_rate": 9.697062992985993e-07,
|
| 727 |
+
"loss": 4.096666259765625,
|
| 728 |
"step": 2450
|
| 729 |
},
|
| 730 |
{
|
| 731 |
+
"epoch": 12.252475247524753,
|
| 732 |
+
"grad_norm": 11.369580268859863,
|
| 733 |
+
"learning_rate": 9.689323558243444e-07,
|
| 734 |
+
"loss": 4.218254089355469,
|
| 735 |
"step": 2475
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 12.376237623762377,
|
| 739 |
+
"grad_norm": 9.929821014404297,
|
| 740 |
+
"learning_rate": 9.681489671070802e-07,
|
| 741 |
+
"loss": 4.054186401367187,
|
| 742 |
"step": 2500
|
| 743 |
},
|
| 744 |
{
|
| 745 |
+
"epoch": 12.376237623762377,
|
| 746 |
+
"eval_loss": 0.9306696653366089,
|
| 747 |
+
"eval_runtime": 82.6001,
|
| 748 |
+
"eval_samples_per_second": 19.564,
|
| 749 |
+
"eval_steps_per_second": 4.891,
|
| 750 |
"step": 2500
|
| 751 |
},
|
| 752 |
{
|
| 753 |
+
"epoch": 12.5,
|
| 754 |
+
"grad_norm": 23.94037628173828,
|
| 755 |
+
"learning_rate": 9.673561489258352e-07,
|
| 756 |
+
"loss": 4.108751220703125,
|
| 757 |
"step": 2525
|
| 758 |
},
|
| 759 |
{
|
| 760 |
+
"epoch": 12.623762376237623,
|
| 761 |
+
"grad_norm": 10.571653366088867,
|
| 762 |
+
"learning_rate": 9.665539172495656e-07,
|
| 763 |
+
"loss": 4.077666320800781,
|
| 764 |
"step": 2550
|
| 765 |
},
|
| 766 |
{
|
| 767 |
+
"epoch": 12.747524752475247,
|
| 768 |
+
"grad_norm": 12.341602325439453,
|
| 769 |
+
"learning_rate": 9.657422882368347e-07,
|
| 770 |
+
"loss": 4.187429504394531,
|
| 771 |
"step": 2575
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 12.871287128712872,
|
| 775 |
+
"grad_norm": 9.840188026428223,
|
| 776 |
+
"learning_rate": 9.64921278235487e-07,
|
| 777 |
+
"loss": 4.135216369628906,
|
| 778 |
"step": 2600
|
| 779 |
},
|
| 780 |
{
|
| 781 |
+
"epoch": 12.995049504950495,
|
| 782 |
+
"grad_norm": 11.124446868896484,
|
| 783 |
+
"learning_rate": 9.640909037823194e-07,
|
| 784 |
+
"loss": 4.087601318359375,
|
| 785 |
"step": 2625
|
| 786 |
},
|
| 787 |
{
|
| 788 |
+
"epoch": 13.118811881188119,
|
| 789 |
+
"grad_norm": 11.180383682250977,
|
| 790 |
+
"learning_rate": 9.63251181602747e-07,
|
| 791 |
+
"loss": 4.0938818359375,
|
| 792 |
"step": 2650
|
| 793 |
},
|
| 794 |
{
|
| 795 |
+
"epoch": 13.242574257425742,
|
| 796 |
+
"grad_norm": 12.539691925048828,
|
| 797 |
+
"learning_rate": 9.624021286104674e-07,
|
| 798 |
+
"loss": 4.085841369628906,
|
| 799 |
"step": 2675
|
| 800 |
},
|
| 801 |
{
|
| 802 |
+
"epoch": 13.366336633663366,
|
| 803 |
+
"grad_norm": 11.089624404907227,
|
| 804 |
+
"learning_rate": 9.615437619071198e-07,
|
| 805 |
+
"loss": 4.057616577148438,
|
| 806 |
"step": 2700
|
| 807 |
},
|
| 808 |
{
|
| 809 |
+
"epoch": 13.490099009900991,
|
| 810 |
+
"grad_norm": 10.118108749389648,
|
| 811 |
+
"learning_rate": 9.6067609878194e-07,
|
| 812 |
+
"loss": 4.123738403320313,
|
| 813 |
"step": 2725
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 13.613861386138614,
|
| 817 |
+
"grad_norm": 13.25112247467041,
|
| 818 |
+
"learning_rate": 9.597991567114124e-07,
|
| 819 |
+
"loss": 4.035686645507813,
|
| 820 |
"step": 2750
|
| 821 |
},
|
| 822 |
{
|
| 823 |
+
"epoch": 13.737623762376238,
|
| 824 |
+
"grad_norm": 10.743168830871582,
|
| 825 |
+
"learning_rate": 9.589129533589185e-07,
|
| 826 |
+
"loss": 4.089259643554687,
|
| 827 |
"step": 2775
|
| 828 |
},
|
| 829 |
{
|
| 830 |
+
"epoch": 13.861386138613861,
|
| 831 |
+
"grad_norm": 13.900023460388184,
|
| 832 |
+
"learning_rate": 9.580175065743798e-07,
|
| 833 |
+
"loss": 4.086239929199219,
|
| 834 |
"step": 2800
|
| 835 |
},
|
| 836 |
{
|
| 837 |
+
"epoch": 13.985148514851485,
|
| 838 |
+
"grad_norm": 14.214149475097656,
|
| 839 |
+
"learning_rate": 9.571128343939005e-07,
|
| 840 |
+
"loss": 4.133528747558594,
|
| 841 |
"step": 2825
|
| 842 |
},
|
| 843 |
{
|
| 844 |
+
"epoch": 14.108910891089108,
|
| 845 |
+
"grad_norm": 13.158830642700195,
|
| 846 |
+
"learning_rate": 9.561989550394016e-07,
|
| 847 |
+
"loss": 4.137518005371094,
|
| 848 |
"step": 2850
|
| 849 |
},
|
| 850 |
{
|
| 851 |
+
"epoch": 14.232673267326733,
|
| 852 |
+
"grad_norm": 13.256293296813965,
|
| 853 |
+
"learning_rate": 9.552758869182557e-07,
|
| 854 |
+
"loss": 4.049294128417968,
|
| 855 |
"step": 2875
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 14.356435643564357,
|
| 859 |
+
"grad_norm": 10.833374977111816,
|
| 860 |
+
"learning_rate": 9.543436486229161e-07,
|
| 861 |
+
"loss": 4.0441677856445315,
|
| 862 |
"step": 2900
|
| 863 |
},
|
| 864 |
{
|
| 865 |
+
"epoch": 14.48019801980198,
|
| 866 |
+
"grad_norm": 12.035900115966797,
|
| 867 |
+
"learning_rate": 9.534022589305411e-07,
|
| 868 |
+
"loss": 4.099019775390625,
|
| 869 |
"step": 2925
|
| 870 |
},
|
| 871 |
{
|
| 872 |
+
"epoch": 14.603960396039604,
|
| 873 |
+
"grad_norm": 11.711991310119629,
|
| 874 |
+
"learning_rate": 9.524517368026168e-07,
|
| 875 |
+
"loss": 4.064933166503907,
|
| 876 |
"step": 2950
|
| 877 |
},
|
| 878 |
{
|
| 879 |
+
"epoch": 14.727722772277227,
|
| 880 |
+
"grad_norm": 20.470388412475586,
|
| 881 |
+
"learning_rate": 9.514921013845754e-07,
|
| 882 |
+
"loss": 4.110359802246093,
|
| 883 |
"step": 2975
|
| 884 |
},
|
| 885 |
{
|
| 886 |
+
"epoch": 14.851485148514852,
|
| 887 |
+
"grad_norm": 12.352797508239746,
|
| 888 |
+
"learning_rate": 9.505233720054085e-07,
|
| 889 |
+
"loss": 4.122251892089844,
|
| 890 |
"step": 3000
|
| 891 |
},
|
| 892 |
{
|
| 893 |
+
"epoch": 14.851485148514852,
|
| 894 |
+
"eval_loss": 0.920803427696228,
|
| 895 |
+
"eval_runtime": 86.0706,
|
| 896 |
+
"eval_samples_per_second": 18.775,
|
| 897 |
+
"eval_steps_per_second": 4.694,
|
| 898 |
"step": 3000
|
| 899 |
}
|
| 900 |
],
|
|
|
|
| 924 |
"attributes": {}
|
| 925 |
}
|
| 926 |
},
|
| 927 |
+
"total_flos": 1.2872276681413248e+16,
|
| 928 |
+
"train_batch_size": 8,
|
| 929 |
"trial_name": null,
|
| 930 |
"trial_params": null
|
| 931 |
}
|
checkpoint-3000/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5393
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d34645042a62af83117180e7d8c3a51dc32127e71ed6c053e7766610460e0b1
|
| 3 |
size 5393
|