Training in progress, epoch 1, checkpoint
Browse files
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1980859973
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a9dab3c237778abcdc6020386d5d6aa7d91bf8f2f12d9072e55f02ac241a9d8f
|
| 3 |
size 1980859973
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 990408885
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f7fa3bbd5872f4ad1210653805e664a05fa2a4d3c7b54fcc57b41c643262594
|
| 3 |
size 990408885
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f886a97068c2df9888daf16e7cbd0ea9a12de504b206548c880a0b888a429bcb
|
| 3 |
size 14575
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73a4b9df454edb9352836894520e4dbf710c154432a1dc2d702b1ac3e1210897
|
| 3 |
size 627
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,643 +1,74 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 2.
|
| 3 |
-
"best_model_checkpoint": "dq158/morbius/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 3.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 3.
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 3.
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 3.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 3.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 4.
|
| 44 |
-
"loss":
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.46,
|
| 49 |
-
"learning_rate": 4.770521898767375e-05,
|
| 50 |
-
"loss": 3.0594,
|
| 51 |
-
"step": 3500
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.52,
|
| 55 |
-
"learning_rate": 4.737739312877e-05,
|
| 56 |
-
"loss": 3.0052,
|
| 57 |
-
"step": 4000
|
| 58 |
-
},
|
| 59 |
-
{
|
| 60 |
-
"epoch": 0.59,
|
| 61 |
-
"learning_rate": 4.704956726986625e-05,
|
| 62 |
-
"loss": 2.9479,
|
| 63 |
-
"step": 4500
|
| 64 |
-
},
|
| 65 |
-
{
|
| 66 |
-
"epoch": 0.66,
|
| 67 |
-
"learning_rate": 4.67217414109625e-05,
|
| 68 |
-
"loss": 2.911,
|
| 69 |
-
"step": 5000
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.72,
|
| 73 |
-
"learning_rate": 4.639391555205875e-05,
|
| 74 |
-
"loss": 2.911,
|
| 75 |
-
"step": 5500
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.79,
|
| 79 |
-
"learning_rate": 4.6066089693155e-05,
|
| 80 |
-
"loss": 2.8922,
|
| 81 |
-
"step": 6000
|
| 82 |
-
},
|
| 83 |
-
{
|
| 84 |
-
"epoch": 0.85,
|
| 85 |
-
"learning_rate": 4.573826383425125e-05,
|
| 86 |
-
"loss": 2.852,
|
| 87 |
-
"step": 6500
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.92,
|
| 91 |
-
"learning_rate": 4.54104379753475e-05,
|
| 92 |
-
"loss": 2.9084,
|
| 93 |
-
"step": 7000
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.98,
|
| 97 |
-
"learning_rate": 4.508261211644375e-05,
|
| 98 |
-
"loss": 2.8791,
|
| 99 |
-
"step": 7500
|
| 100 |
-
},
|
| 101 |
{
|
| 102 |
"epoch": 1.0,
|
| 103 |
-
"
|
| 104 |
-
"
|
| 105 |
-
"
|
| 106 |
-
"
|
| 107 |
-
"
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
"
|
| 116 |
-
"
|
| 117 |
-
"
|
| 118 |
-
"step":
|
| 119 |
-
},
|
| 120 |
-
{
|
| 121 |
-
"epoch": 1.11,
|
| 122 |
-
"learning_rate": 4.442696039863625e-05,
|
| 123 |
-
"loss": 2.7981,
|
| 124 |
-
"step": 8500
|
| 125 |
-
},
|
| 126 |
-
{
|
| 127 |
-
"epoch": 1.18,
|
| 128 |
-
"learning_rate": 4.4099134539732497e-05,
|
| 129 |
-
"loss": 2.7356,
|
| 130 |
-
"step": 9000
|
| 131 |
-
},
|
| 132 |
-
{
|
| 133 |
-
"epoch": 1.25,
|
| 134 |
-
"learning_rate": 4.3771308680828746e-05,
|
| 135 |
-
"loss": 2.8206,
|
| 136 |
-
"step": 9500
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 1.31,
|
| 140 |
-
"learning_rate": 4.3443482821924996e-05,
|
| 141 |
-
"loss": 2.7265,
|
| 142 |
-
"step": 10000
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 1.38,
|
| 146 |
-
"learning_rate": 4.3115656963021246e-05,
|
| 147 |
-
"loss": 2.7106,
|
| 148 |
-
"step": 10500
|
| 149 |
-
},
|
| 150 |
-
{
|
| 151 |
-
"epoch": 1.44,
|
| 152 |
-
"learning_rate": 4.2787831104117495e-05,
|
| 153 |
-
"loss": 2.7254,
|
| 154 |
-
"step": 11000
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 1.51,
|
| 158 |
-
"learning_rate": 4.2460005245213745e-05,
|
| 159 |
-
"loss": 2.6474,
|
| 160 |
-
"step": 11500
|
| 161 |
-
},
|
| 162 |
-
{
|
| 163 |
-
"epoch": 1.57,
|
| 164 |
-
"learning_rate": 4.2132179386309995e-05,
|
| 165 |
-
"loss": 2.6853,
|
| 166 |
-
"step": 12000
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 1.64,
|
| 170 |
-
"learning_rate": 4.1804353527406244e-05,
|
| 171 |
-
"loss": 2.6965,
|
| 172 |
-
"step": 12500
|
| 173 |
-
},
|
| 174 |
-
{
|
| 175 |
-
"epoch": 1.7,
|
| 176 |
-
"learning_rate": 4.1476527668502494e-05,
|
| 177 |
-
"loss": 2.6693,
|
| 178 |
-
"step": 13000
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 1.77,
|
| 182 |
-
"learning_rate": 4.1148701809598744e-05,
|
| 183 |
-
"loss": 2.6322,
|
| 184 |
-
"step": 13500
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 1.84,
|
| 188 |
-
"learning_rate": 4.082087595069499e-05,
|
| 189 |
-
"loss": 2.6495,
|
| 190 |
-
"step": 14000
|
| 191 |
-
},
|
| 192 |
-
{
|
| 193 |
-
"epoch": 1.9,
|
| 194 |
-
"learning_rate": 4.049305009179124e-05,
|
| 195 |
-
"loss": 2.6525,
|
| 196 |
-
"step": 14500
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 1.97,
|
| 200 |
-
"learning_rate": 4.016522423288749e-05,
|
| 201 |
-
"loss": 2.6366,
|
| 202 |
-
"step": 15000
|
| 203 |
-
},
|
| 204 |
-
{
|
| 205 |
-
"epoch": 2.0,
|
| 206 |
-
"eval_gen_len": 10.9786072587784,
|
| 207 |
-
"eval_loss": 2.5554187297821045,
|
| 208 |
-
"eval_rouge1": 2.6765,
|
| 209 |
-
"eval_rouge2": 0.8682,
|
| 210 |
-
"eval_rougeL": 2.6696,
|
| 211 |
-
"eval_rougeLsum": 2.6713,
|
| 212 |
-
"eval_runtime": 495.1658,
|
| 213 |
-
"eval_samples_per_second": 13.688,
|
| 214 |
-
"eval_steps_per_second": 1.713,
|
| 215 |
-
"step": 15252
|
| 216 |
-
},
|
| 217 |
-
{
|
| 218 |
-
"epoch": 2.03,
|
| 219 |
-
"learning_rate": 3.983739837398374e-05,
|
| 220 |
-
"loss": 2.6367,
|
| 221 |
-
"step": 15500
|
| 222 |
-
},
|
| 223 |
-
{
|
| 224 |
-
"epoch": 2.1,
|
| 225 |
-
"learning_rate": 3.950957251507999e-05,
|
| 226 |
-
"loss": 2.5734,
|
| 227 |
-
"step": 16000
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 2.16,
|
| 231 |
-
"learning_rate": 3.918174665617624e-05,
|
| 232 |
-
"loss": 2.5957,
|
| 233 |
-
"step": 16500
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 2.23,
|
| 237 |
-
"learning_rate": 3.885392079727249e-05,
|
| 238 |
-
"loss": 2.5152,
|
| 239 |
-
"step": 17000
|
| 240 |
-
},
|
| 241 |
-
{
|
| 242 |
-
"epoch": 2.29,
|
| 243 |
-
"learning_rate": 3.852609493836874e-05,
|
| 244 |
-
"loss": 2.5383,
|
| 245 |
-
"step": 17500
|
| 246 |
-
},
|
| 247 |
-
{
|
| 248 |
-
"epoch": 2.36,
|
| 249 |
-
"learning_rate": 3.819826907946499e-05,
|
| 250 |
-
"loss": 2.6066,
|
| 251 |
-
"step": 18000
|
| 252 |
-
},
|
| 253 |
-
{
|
| 254 |
-
"epoch": 2.43,
|
| 255 |
-
"learning_rate": 3.787044322056124e-05,
|
| 256 |
-
"loss": 2.5666,
|
| 257 |
-
"step": 18500
|
| 258 |
-
},
|
| 259 |
-
{
|
| 260 |
-
"epoch": 2.49,
|
| 261 |
-
"learning_rate": 3.754261736165749e-05,
|
| 262 |
-
"loss": 2.5145,
|
| 263 |
-
"step": 19000
|
| 264 |
-
},
|
| 265 |
-
{
|
| 266 |
-
"epoch": 2.56,
|
| 267 |
-
"learning_rate": 3.721479150275374e-05,
|
| 268 |
-
"loss": 2.5282,
|
| 269 |
-
"step": 19500
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 2.62,
|
| 273 |
-
"learning_rate": 3.688696564384999e-05,
|
| 274 |
-
"loss": 2.5698,
|
| 275 |
-
"step": 20000
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 2.69,
|
| 279 |
-
"learning_rate": 3.655913978494624e-05,
|
| 280 |
-
"loss": 2.5273,
|
| 281 |
-
"step": 20500
|
| 282 |
-
},
|
| 283 |
-
{
|
| 284 |
-
"epoch": 2.75,
|
| 285 |
-
"learning_rate": 3.623131392604249e-05,
|
| 286 |
-
"loss": 2.5781,
|
| 287 |
-
"step": 21000
|
| 288 |
-
},
|
| 289 |
-
{
|
| 290 |
-
"epoch": 2.82,
|
| 291 |
-
"learning_rate": 3.590348806713874e-05,
|
| 292 |
-
"loss": 2.5602,
|
| 293 |
-
"step": 21500
|
| 294 |
-
},
|
| 295 |
-
{
|
| 296 |
-
"epoch": 2.88,
|
| 297 |
-
"learning_rate": 3.557566220823499e-05,
|
| 298 |
-
"loss": 2.5263,
|
| 299 |
-
"step": 22000
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 2.95,
|
| 303 |
-
"learning_rate": 3.524783634933124e-05,
|
| 304 |
-
"loss": 2.5724,
|
| 305 |
-
"step": 22500
|
| 306 |
-
},
|
| 307 |
-
{
|
| 308 |
-
"epoch": 3.0,
|
| 309 |
-
"eval_gen_len": 9.474918855119505,
|
| 310 |
-
"eval_loss": 2.504915952682495,
|
| 311 |
-
"eval_rouge1": 3.2194,
|
| 312 |
-
"eval_rouge2": 0.916,
|
| 313 |
-
"eval_rougeL": 3.2175,
|
| 314 |
-
"eval_rougeLsum": 3.2137,
|
| 315 |
-
"eval_runtime": 484.055,
|
| 316 |
-
"eval_samples_per_second": 14.003,
|
| 317 |
-
"eval_steps_per_second": 1.752,
|
| 318 |
-
"step": 22878
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 3.02,
|
| 322 |
-
"learning_rate": 3.492001049042749e-05,
|
| 323 |
-
"loss": 2.488,
|
| 324 |
-
"step": 23000
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 3.08,
|
| 328 |
-
"learning_rate": 3.459218463152374e-05,
|
| 329 |
-
"loss": 2.3958,
|
| 330 |
-
"step": 23500
|
| 331 |
-
},
|
| 332 |
-
{
|
| 333 |
-
"epoch": 3.15,
|
| 334 |
-
"learning_rate": 3.426435877261999e-05,
|
| 335 |
-
"loss": 2.4125,
|
| 336 |
-
"step": 24000
|
| 337 |
-
},
|
| 338 |
-
{
|
| 339 |
-
"epoch": 3.21,
|
| 340 |
-
"learning_rate": 3.3936532913716236e-05,
|
| 341 |
-
"loss": 2.4189,
|
| 342 |
-
"step": 24500
|
| 343 |
-
},
|
| 344 |
-
{
|
| 345 |
-
"epoch": 3.28,
|
| 346 |
-
"learning_rate": 3.3608707054812486e-05,
|
| 347 |
-
"loss": 2.4619,
|
| 348 |
-
"step": 25000
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 3.34,
|
| 352 |
-
"learning_rate": 3.3280881195908736e-05,
|
| 353 |
-
"loss": 2.4608,
|
| 354 |
-
"step": 25500
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 3.41,
|
| 358 |
-
"learning_rate": 3.2953055337004985e-05,
|
| 359 |
-
"loss": 2.4737,
|
| 360 |
-
"step": 26000
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 3.47,
|
| 364 |
-
"learning_rate": 3.2625229478101235e-05,
|
| 365 |
-
"loss": 2.4373,
|
| 366 |
-
"step": 26500
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 3.54,
|
| 370 |
-
"learning_rate": 3.2297403619197485e-05,
|
| 371 |
-
"loss": 2.4892,
|
| 372 |
-
"step": 27000
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"epoch": 3.61,
|
| 376 |
-
"learning_rate": 3.1969577760293734e-05,
|
| 377 |
-
"loss": 2.5006,
|
| 378 |
-
"step": 27500
|
| 379 |
-
},
|
| 380 |
-
{
|
| 381 |
-
"epoch": 3.67,
|
| 382 |
-
"learning_rate": 3.1641751901389984e-05,
|
| 383 |
-
"loss": 2.4974,
|
| 384 |
-
"step": 28000
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 3.74,
|
| 388 |
-
"learning_rate": 3.131392604248623e-05,
|
| 389 |
-
"loss": 2.4321,
|
| 390 |
-
"step": 28500
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 3.8,
|
| 394 |
-
"learning_rate": 3.0986100183582483e-05,
|
| 395 |
-
"loss": 2.4587,
|
| 396 |
-
"step": 29000
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"epoch": 3.87,
|
| 400 |
-
"learning_rate": 3.065827432467873e-05,
|
| 401 |
-
"loss": 2.4401,
|
| 402 |
-
"step": 29500
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 3.93,
|
| 406 |
-
"learning_rate": 3.033044846577498e-05,
|
| 407 |
-
"loss": 2.4593,
|
| 408 |
-
"step": 30000
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 4.0,
|
| 412 |
-
"learning_rate": 3.0002622606871232e-05,
|
| 413 |
-
"loss": 2.4562,
|
| 414 |
-
"step": 30500
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 4.0,
|
| 418 |
-
"eval_gen_len": 8.58940690469165,
|
| 419 |
-
"eval_loss": 2.4669947624206543,
|
| 420 |
-
"eval_rouge1": 4.0735,
|
| 421 |
-
"eval_rouge2": 1.2636,
|
| 422 |
-
"eval_rougeL": 4.0604,
|
| 423 |
-
"eval_rougeLsum": 4.0621,
|
| 424 |
-
"eval_runtime": 475.9174,
|
| 425 |
-
"eval_samples_per_second": 14.242,
|
| 426 |
-
"eval_steps_per_second": 1.782,
|
| 427 |
-
"step": 30504
|
| 428 |
-
},
|
| 429 |
-
{
|
| 430 |
-
"epoch": 4.07,
|
| 431 |
-
"learning_rate": 2.9674796747967482e-05,
|
| 432 |
-
"loss": 2.4016,
|
| 433 |
-
"step": 31000
|
| 434 |
-
},
|
| 435 |
-
{
|
| 436 |
-
"epoch": 4.13,
|
| 437 |
-
"learning_rate": 2.934697088906373e-05,
|
| 438 |
-
"loss": 2.3893,
|
| 439 |
-
"step": 31500
|
| 440 |
-
},
|
| 441 |
-
{
|
| 442 |
-
"epoch": 4.2,
|
| 443 |
-
"learning_rate": 2.9019145030159978e-05,
|
| 444 |
-
"loss": 2.3954,
|
| 445 |
-
"step": 32000
|
| 446 |
-
},
|
| 447 |
-
{
|
| 448 |
-
"epoch": 4.26,
|
| 449 |
-
"learning_rate": 2.869131917125623e-05,
|
| 450 |
-
"loss": 2.3312,
|
| 451 |
-
"step": 32500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 4.33,
|
| 455 |
-
"learning_rate": 2.836349331235248e-05,
|
| 456 |
-
"loss": 2.3718,
|
| 457 |
-
"step": 33000
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 4.39,
|
| 461 |
-
"learning_rate": 2.8035667453448727e-05,
|
| 462 |
-
"loss": 2.374,
|
| 463 |
-
"step": 33500
|
| 464 |
-
},
|
| 465 |
-
{
|
| 466 |
-
"epoch": 4.46,
|
| 467 |
-
"learning_rate": 2.770784159454498e-05,
|
| 468 |
-
"loss": 2.3837,
|
| 469 |
-
"step": 34000
|
| 470 |
-
},
|
| 471 |
-
{
|
| 472 |
-
"epoch": 4.52,
|
| 473 |
-
"learning_rate": 2.738001573564123e-05,
|
| 474 |
-
"loss": 2.3719,
|
| 475 |
-
"step": 34500
|
| 476 |
-
},
|
| 477 |
-
{
|
| 478 |
-
"epoch": 4.59,
|
| 479 |
-
"learning_rate": 2.7052189876737476e-05,
|
| 480 |
-
"loss": 2.3966,
|
| 481 |
-
"step": 35000
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 4.66,
|
| 485 |
-
"learning_rate": 2.672436401783373e-05,
|
| 486 |
-
"loss": 2.37,
|
| 487 |
-
"step": 35500
|
| 488 |
-
},
|
| 489 |
-
{
|
| 490 |
-
"epoch": 4.72,
|
| 491 |
-
"learning_rate": 2.639653815892998e-05,
|
| 492 |
-
"loss": 2.4051,
|
| 493 |
-
"step": 36000
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 4.79,
|
| 497 |
-
"learning_rate": 2.6068712300026225e-05,
|
| 498 |
-
"loss": 2.4123,
|
| 499 |
-
"step": 36500
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 4.85,
|
| 503 |
-
"learning_rate": 2.5740886441122475e-05,
|
| 504 |
-
"loss": 2.3227,
|
| 505 |
-
"step": 37000
|
| 506 |
-
},
|
| 507 |
-
{
|
| 508 |
-
"epoch": 4.92,
|
| 509 |
-
"learning_rate": 2.5413060582218728e-05,
|
| 510 |
-
"loss": 2.3622,
|
| 511 |
-
"step": 37500
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 4.98,
|
| 515 |
-
"learning_rate": 2.5085234723314978e-05,
|
| 516 |
-
"loss": 2.3699,
|
| 517 |
-
"step": 38000
|
| 518 |
-
},
|
| 519 |
-
{
|
| 520 |
-
"epoch": 5.0,
|
| 521 |
-
"eval_gen_len": 8.69371496016524,
|
| 522 |
-
"eval_loss": 2.4463706016540527,
|
| 523 |
-
"eval_rouge1": 4.2784,
|
| 524 |
-
"eval_rouge2": 1.3764,
|
| 525 |
-
"eval_rougeL": 4.2679,
|
| 526 |
-
"eval_rougeLsum": 4.2621,
|
| 527 |
-
"eval_runtime": 478.3802,
|
| 528 |
-
"eval_samples_per_second": 14.169,
|
| 529 |
-
"eval_steps_per_second": 1.773,
|
| 530 |
-
"step": 38130
|
| 531 |
-
},
|
| 532 |
-
{
|
| 533 |
-
"epoch": 5.05,
|
| 534 |
-
"learning_rate": 2.4757408864411227e-05,
|
| 535 |
-
"loss": 2.3002,
|
| 536 |
-
"step": 38500
|
| 537 |
-
},
|
| 538 |
-
{
|
| 539 |
-
"epoch": 5.11,
|
| 540 |
-
"learning_rate": 2.4429583005507477e-05,
|
| 541 |
-
"loss": 2.307,
|
| 542 |
-
"step": 39000
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 5.18,
|
| 546 |
-
"learning_rate": 2.4101757146603727e-05,
|
| 547 |
-
"loss": 2.3295,
|
| 548 |
-
"step": 39500
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 5.25,
|
| 552 |
-
"learning_rate": 2.3773931287699973e-05,
|
| 553 |
-
"loss": 2.3213,
|
| 554 |
-
"step": 40000
|
| 555 |
-
},
|
| 556 |
-
{
|
| 557 |
-
"epoch": 5.31,
|
| 558 |
-
"learning_rate": 2.3446105428796226e-05,
|
| 559 |
-
"loss": 2.3208,
|
| 560 |
-
"step": 40500
|
| 561 |
-
},
|
| 562 |
-
{
|
| 563 |
-
"epoch": 5.38,
|
| 564 |
-
"learning_rate": 2.3118279569892472e-05,
|
| 565 |
-
"loss": 2.335,
|
| 566 |
-
"step": 41000
|
| 567 |
-
},
|
| 568 |
-
{
|
| 569 |
-
"epoch": 5.44,
|
| 570 |
-
"learning_rate": 2.2790453710988725e-05,
|
| 571 |
-
"loss": 2.3047,
|
| 572 |
-
"step": 41500
|
| 573 |
-
},
|
| 574 |
-
{
|
| 575 |
-
"epoch": 5.51,
|
| 576 |
-
"learning_rate": 2.2462627852084975e-05,
|
| 577 |
-
"loss": 2.3212,
|
| 578 |
-
"step": 42000
|
| 579 |
-
},
|
| 580 |
-
{
|
| 581 |
-
"epoch": 5.57,
|
| 582 |
-
"learning_rate": 2.213480199318122e-05,
|
| 583 |
-
"loss": 2.2925,
|
| 584 |
-
"step": 42500
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 5.64,
|
| 588 |
-
"learning_rate": 2.1806976134277474e-05,
|
| 589 |
-
"loss": 2.2996,
|
| 590 |
-
"step": 43000
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 5.7,
|
| 594 |
-
"learning_rate": 2.147915027537372e-05,
|
| 595 |
-
"loss": 2.3415,
|
| 596 |
-
"step": 43500
|
| 597 |
-
},
|
| 598 |
-
{
|
| 599 |
-
"epoch": 5.77,
|
| 600 |
-
"learning_rate": 2.1151324416469974e-05,
|
| 601 |
-
"loss": 2.2935,
|
| 602 |
-
"step": 44000
|
| 603 |
-
},
|
| 604 |
-
{
|
| 605 |
-
"epoch": 5.84,
|
| 606 |
-
"learning_rate": 2.082349855756622e-05,
|
| 607 |
-
"loss": 2.3273,
|
| 608 |
-
"step": 44500
|
| 609 |
-
},
|
| 610 |
-
{
|
| 611 |
-
"epoch": 5.9,
|
| 612 |
-
"learning_rate": 2.0495672698662473e-05,
|
| 613 |
-
"loss": 2.3358,
|
| 614 |
-
"step": 45000
|
| 615 |
-
},
|
| 616 |
-
{
|
| 617 |
-
"epoch": 5.97,
|
| 618 |
-
"learning_rate": 2.0167846839758723e-05,
|
| 619 |
-
"loss": 2.3031,
|
| 620 |
-
"step": 45500
|
| 621 |
-
},
|
| 622 |
-
{
|
| 623 |
-
"epoch": 6.0,
|
| 624 |
-
"eval_gen_len": 10.01962230746533,
|
| 625 |
-
"eval_loss": 2.4326746463775635,
|
| 626 |
-
"eval_rouge1": 4.6968,
|
| 627 |
-
"eval_rouge2": 1.6015,
|
| 628 |
-
"eval_rougeL": 4.6813,
|
| 629 |
-
"eval_rougeLsum": 4.6911,
|
| 630 |
-
"eval_runtime": 492.1017,
|
| 631 |
-
"eval_samples_per_second": 13.774,
|
| 632 |
-
"eval_steps_per_second": 1.723,
|
| 633 |
-
"step": 45756
|
| 634 |
}
|
| 635 |
],
|
| 636 |
"logging_steps": 500,
|
| 637 |
-
"max_steps":
|
| 638 |
-
"num_train_epochs":
|
| 639 |
"save_steps": 500,
|
| 640 |
-
"total_flos": 2.
|
| 641 |
"trial_name": null,
|
| 642 |
"trial_params": null
|
| 643 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 2.766650676727295,
|
| 3 |
+
"best_model_checkpoint": "dq158/morbius/checkpoint-3042",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3042,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.16,
|
| 13 |
+
"learning_rate": 4.835634451019067e-05,
|
| 14 |
+
"loss": 3.6556,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.33,
|
| 19 |
+
"learning_rate": 4.671268902038133e-05,
|
| 20 |
+
"loss": 3.223,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.49,
|
| 25 |
+
"learning_rate": 4.5069033530571994e-05,
|
| 26 |
+
"loss": 3.1336,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.66,
|
| 31 |
+
"learning_rate": 4.342537804076266e-05,
|
| 32 |
+
"loss": 3.066,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.82,
|
| 37 |
+
"learning_rate": 4.178172255095332e-05,
|
| 38 |
+
"loss": 3.0564,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.99,
|
| 43 |
+
"learning_rate": 4.0138067061143986e-05,
|
| 44 |
+
"loss": 2.9224,
|
| 45 |
"step": 3000
|
| 46 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
{
|
| 48 |
"epoch": 1.0,
|
| 49 |
+
"eval_bleu": 0.023938349196740243,
|
| 50 |
+
"eval_brevity_penalty": 0.2829444266697748,
|
| 51 |
+
"eval_length_ratio": 0.4419880178887858,
|
| 52 |
+
"eval_loss": 2.766650676727295,
|
| 53 |
+
"eval_precisions": [
|
| 54 |
+
0.20208094692630776,
|
| 55 |
+
0.11387310606060606,
|
| 56 |
+
0.04758566978193146,
|
| 57 |
+
0.0467896174863388
|
| 58 |
+
],
|
| 59 |
+
"eval_reference_length": 47404,
|
| 60 |
+
"eval_runtime": 168.3486,
|
| 61 |
+
"eval_samples_per_second": 24.093,
|
| 62 |
+
"eval_steps_per_second": 2.008,
|
| 63 |
+
"eval_translation_length": 20952,
|
| 64 |
+
"step": 3042
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
}
|
| 66 |
],
|
| 67 |
"logging_steps": 500,
|
| 68 |
+
"max_steps": 15210,
|
| 69 |
+
"num_train_epochs": 5,
|
| 70 |
"save_steps": 500,
|
| 71 |
+
"total_flos": 2.499021957758976e+16,
|
| 72 |
"trial_name": null,
|
| 73 |
"trial_params": null
|
| 74 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4219
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cdb1e797f594b30976363588c0c374fa82672278e85e67cfbe87db449cc3a94c
|
| 3 |
size 4219
|