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 1980860410
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ab5328b63a8ec3b53c2f26a99111049d2792ac322fccebc0739829ebfab0879
|
| 3 |
size 1980860410
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 990409330
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0afe6c8cd50d6bcf5ad6cb45258efb15f89d523168c877fac499ab3891b636e
|
| 3 |
size 990409330
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52a28bc3541f02e8de86363b8d2f634108f83fb8b7a33f774c760cce0869599d
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5afd8d09736ea35ab4f783803d2aca249bf98e9d0591be216198fe2ebe96a3c3
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,684 +1,50 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 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":
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
-
{
|
| 18 |
-
"epoch": 0.1,
|
| 19 |
-
"learning_rate": 4.895484949832776e-05,
|
| 20 |
-
"loss": 2.08,
|
| 21 |
-
"step": 1000
|
| 22 |
-
},
|
| 23 |
-
{
|
| 24 |
-
"epoch": 0.16,
|
| 25 |
-
"learning_rate": 4.8432274247491646e-05,
|
| 26 |
-
"loss": 2.0242,
|
| 27 |
-
"step": 1500
|
| 28 |
-
},
|
| 29 |
-
{
|
| 30 |
-
"epoch": 0.21,
|
| 31 |
-
"learning_rate": 4.7909698996655525e-05,
|
| 32 |
-
"loss": 2.0865,
|
| 33 |
-
"step": 2000
|
| 34 |
-
},
|
| 35 |
-
{
|
| 36 |
-
"epoch": 0.26,
|
| 37 |
-
"learning_rate": 4.7387123745819403e-05,
|
| 38 |
-
"loss": 2.0723,
|
| 39 |
-
"step": 2500
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"epoch": 0.31,
|
| 43 |
-
"learning_rate": 4.686454849498328e-05,
|
| 44 |
-
"loss": 2.0951,
|
| 45 |
-
"step": 3000
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.37,
|
| 49 |
-
"learning_rate": 4.6341973244147154e-05,
|
| 50 |
-
"loss": 2.1049,
|
| 51 |
-
"step": 3500
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.42,
|
| 55 |
-
"learning_rate": 4.581939799331103e-05,
|
| 56 |
-
"loss": 2.1124,
|
| 57 |
-
"step": 4000
|
| 58 |
-
},
|
| 59 |
-
{
|
| 60 |
-
"epoch": 0.47,
|
| 61 |
-
"learning_rate": 4.529682274247492e-05,
|
| 62 |
-
"loss": 2.0989,
|
| 63 |
-
"step": 4500
|
| 64 |
-
},
|
| 65 |
-
{
|
| 66 |
-
"epoch": 0.52,
|
| 67 |
-
"learning_rate": 4.47742474916388e-05,
|
| 68 |
-
"loss": 2.1086,
|
| 69 |
-
"step": 5000
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.57,
|
| 73 |
-
"learning_rate": 4.425167224080268e-05,
|
| 74 |
-
"loss": 2.0899,
|
| 75 |
-
"step": 5500
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.63,
|
| 79 |
-
"learning_rate": 4.3729096989966556e-05,
|
| 80 |
-
"loss": 2.0607,
|
| 81 |
-
"step": 6000
|
| 82 |
-
},
|
| 83 |
{
|
| 84 |
"epoch": 0.68,
|
| 85 |
-
"learning_rate": 4.
|
| 86 |
-
"loss":
|
| 87 |
-
"step":
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.73,
|
| 91 |
-
"learning_rate": 4.268394648829432e-05,
|
| 92 |
-
"loss": 2.0447,
|
| 93 |
-
"step": 7000
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.78,
|
| 97 |
-
"learning_rate": 4.21613712374582e-05,
|
| 98 |
-
"loss": 2.0825,
|
| 99 |
-
"step": 7500
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.84,
|
| 103 |
-
"learning_rate": 4.163879598662208e-05,
|
| 104 |
-
"loss": 2.077,
|
| 105 |
-
"step": 8000
|
| 106 |
-
},
|
| 107 |
-
{
|
| 108 |
-
"epoch": 0.89,
|
| 109 |
-
"learning_rate": 4.111622073578596e-05,
|
| 110 |
-
"loss": 2.0956,
|
| 111 |
-
"step": 8500
|
| 112 |
-
},
|
| 113 |
-
{
|
| 114 |
-
"epoch": 0.94,
|
| 115 |
-
"learning_rate": 4.0593645484949835e-05,
|
| 116 |
-
"loss": 2.0863,
|
| 117 |
-
"step": 9000
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"epoch": 0.99,
|
| 121 |
-
"learning_rate": 4.0071070234113714e-05,
|
| 122 |
-
"loss": 2.0549,
|
| 123 |
-
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.0,
|
| 127 |
-
"eval_bleu": 0
|
| 128 |
-
"eval_brevity_penalty": 0
|
| 129 |
-
"eval_length_ratio": 0
|
| 130 |
-
"eval_loss": 1.
|
| 131 |
-
"eval_precisions": [
|
| 132 |
-
0.1860885275519422,
|
| 133 |
-
0.11627011709351441,
|
| 134 |
-
0.08451872485298668,
|
| 135 |
-
0.08095923402255639
|
| 136 |
-
],
|
| 137 |
-
"eval_reference_length": 134020,
|
| 138 |
-
"eval_runtime": 811.6945,
|
| 139 |
-
"eval_samples_per_second": 15.717,
|
| 140 |
-
"eval_steps_per_second": 1.311,
|
| 141 |
-
"eval_translation_length": 106272,
|
| 142 |
-
"step": 9568
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 1.05,
|
| 146 |
-
"learning_rate": 3.954849498327759e-05,
|
| 147 |
-
"loss": 2.0367,
|
| 148 |
-
"step": 10000
|
| 149 |
-
},
|
| 150 |
-
{
|
| 151 |
-
"epoch": 1.1,
|
| 152 |
-
"learning_rate": 3.902591973244147e-05,
|
| 153 |
-
"loss": 2.0379,
|
| 154 |
-
"step": 10500
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 1.15,
|
| 158 |
-
"learning_rate": 3.850334448160535e-05,
|
| 159 |
-
"loss": 1.9553,
|
| 160 |
-
"step": 11000
|
| 161 |
-
},
|
| 162 |
-
{
|
| 163 |
-
"epoch": 1.2,
|
| 164 |
-
"learning_rate": 3.798076923076923e-05,
|
| 165 |
-
"loss": 1.9742,
|
| 166 |
-
"step": 11500
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 1.25,
|
| 170 |
-
"learning_rate": 3.745819397993311e-05,
|
| 171 |
-
"loss": 1.9989,
|
| 172 |
-
"step": 12000
|
| 173 |
-
},
|
| 174 |
-
{
|
| 175 |
-
"epoch": 1.31,
|
| 176 |
-
"learning_rate": 3.6935618729096994e-05,
|
| 177 |
-
"loss": 2.0074,
|
| 178 |
-
"step": 12500
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 1.36,
|
| 182 |
-
"learning_rate": 3.641304347826087e-05,
|
| 183 |
-
"loss": 2.04,
|
| 184 |
-
"step": 13000
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 1.41,
|
| 188 |
-
"learning_rate": 3.589046822742475e-05,
|
| 189 |
-
"loss": 1.9847,
|
| 190 |
-
"step": 13500
|
| 191 |
-
},
|
| 192 |
-
{
|
| 193 |
-
"epoch": 1.46,
|
| 194 |
-
"learning_rate": 3.536789297658863e-05,
|
| 195 |
-
"loss": 1.9823,
|
| 196 |
-
"step": 14000
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 1.52,
|
| 200 |
-
"learning_rate": 3.484531772575251e-05,
|
| 201 |
-
"loss": 2.034,
|
| 202 |
-
"step": 14500
|
| 203 |
-
},
|
| 204 |
-
{
|
| 205 |
-
"epoch": 1.57,
|
| 206 |
-
"learning_rate": 3.432274247491639e-05,
|
| 207 |
-
"loss": 2.0177,
|
| 208 |
-
"step": 15000
|
| 209 |
-
},
|
| 210 |
-
{
|
| 211 |
-
"epoch": 1.62,
|
| 212 |
-
"learning_rate": 3.380016722408027e-05,
|
| 213 |
-
"loss": 2.0167,
|
| 214 |
-
"step": 15500
|
| 215 |
-
},
|
| 216 |
-
{
|
| 217 |
-
"epoch": 1.67,
|
| 218 |
-
"learning_rate": 3.3277591973244146e-05,
|
| 219 |
-
"loss": 2.0235,
|
| 220 |
-
"step": 16000
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 1.72,
|
| 224 |
-
"learning_rate": 3.2755016722408025e-05,
|
| 225 |
-
"loss": 2.0242,
|
| 226 |
-
"step": 16500
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 1.78,
|
| 230 |
-
"learning_rate": 3.2232441471571904e-05,
|
| 231 |
-
"loss": 2.0119,
|
| 232 |
-
"step": 17000
|
| 233 |
-
},
|
| 234 |
-
{
|
| 235 |
-
"epoch": 1.83,
|
| 236 |
-
"learning_rate": 3.170986622073579e-05,
|
| 237 |
-
"loss": 1.9696,
|
| 238 |
-
"step": 17500
|
| 239 |
-
},
|
| 240 |
-
{
|
| 241 |
-
"epoch": 1.88,
|
| 242 |
-
"learning_rate": 3.118729096989967e-05,
|
| 243 |
-
"loss": 2.0214,
|
| 244 |
-
"step": 18000
|
| 245 |
-
},
|
| 246 |
-
{
|
| 247 |
-
"epoch": 1.93,
|
| 248 |
-
"learning_rate": 3.066471571906355e-05,
|
| 249 |
-
"loss": 2.0348,
|
| 250 |
-
"step": 18500
|
| 251 |
-
},
|
| 252 |
-
{
|
| 253 |
-
"epoch": 1.99,
|
| 254 |
-
"learning_rate": 3.0142140468227426e-05,
|
| 255 |
-
"loss": 1.9923,
|
| 256 |
-
"step": 19000
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 2.0,
|
| 260 |
-
"eval_bleu": 0.08658597339938905,
|
| 261 |
-
"eval_brevity_penalty": 0.7517944633429557,
|
| 262 |
-
"eval_length_ratio": 0.7780331293836741,
|
| 263 |
-
"eval_loss": 1.83921480178833,
|
| 264 |
-
"eval_precisions": [
|
| 265 |
-
0.1897824919441461,
|
| 266 |
-
0.11953231710648528,
|
| 267 |
-
0.0894031328543502,
|
| 268 |
-
0.08675598838616018
|
| 269 |
-
],
|
| 270 |
-
"eval_reference_length": 134020,
|
| 271 |
-
"eval_runtime": 809.9384,
|
| 272 |
-
"eval_samples_per_second": 15.751,
|
| 273 |
-
"eval_steps_per_second": 1.314,
|
| 274 |
-
"eval_translation_length": 104272,
|
| 275 |
-
"step": 19136
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 2.04,
|
| 279 |
-
"learning_rate": 2.9619565217391305e-05,
|
| 280 |
-
"loss": 1.9566,
|
| 281 |
-
"step": 19500
|
| 282 |
-
},
|
| 283 |
-
{
|
| 284 |
-
"epoch": 2.09,
|
| 285 |
-
"learning_rate": 2.9096989966555184e-05,
|
| 286 |
-
"loss": 1.9582,
|
| 287 |
-
"step": 20000
|
| 288 |
-
},
|
| 289 |
-
{
|
| 290 |
-
"epoch": 2.14,
|
| 291 |
-
"learning_rate": 2.8574414715719066e-05,
|
| 292 |
-
"loss": 1.9681,
|
| 293 |
-
"step": 20500
|
| 294 |
-
},
|
| 295 |
-
{
|
| 296 |
-
"epoch": 2.19,
|
| 297 |
-
"learning_rate": 2.8051839464882945e-05,
|
| 298 |
-
"loss": 1.9374,
|
| 299 |
-
"step": 21000
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 2.25,
|
| 303 |
-
"learning_rate": 2.7529264214046824e-05,
|
| 304 |
-
"loss": 1.9412,
|
| 305 |
-
"step": 21500
|
| 306 |
-
},
|
| 307 |
-
{
|
| 308 |
-
"epoch": 2.3,
|
| 309 |
-
"learning_rate": 2.7006688963210703e-05,
|
| 310 |
-
"loss": 1.9557,
|
| 311 |
-
"step": 22000
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 2.35,
|
| 315 |
-
"learning_rate": 2.6484113712374582e-05,
|
| 316 |
-
"loss": 1.9684,
|
| 317 |
-
"step": 22500
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 2.4,
|
| 321 |
-
"learning_rate": 2.5961538461538464e-05,
|
| 322 |
-
"loss": 1.9307,
|
| 323 |
-
"step": 23000
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"epoch": 2.46,
|
| 327 |
-
"learning_rate": 2.5438963210702343e-05,
|
| 328 |
-
"loss": 1.9237,
|
| 329 |
-
"step": 23500
|
| 330 |
-
},
|
| 331 |
-
{
|
| 332 |
-
"epoch": 2.51,
|
| 333 |
-
"learning_rate": 2.491638795986622e-05,
|
| 334 |
-
"loss": 1.9608,
|
| 335 |
-
"step": 24000
|
| 336 |
-
},
|
| 337 |
-
{
|
| 338 |
-
"epoch": 2.56,
|
| 339 |
-
"learning_rate": 2.43938127090301e-05,
|
| 340 |
-
"loss": 1.9549,
|
| 341 |
-
"step": 24500
|
| 342 |
-
},
|
| 343 |
-
{
|
| 344 |
-
"epoch": 2.61,
|
| 345 |
-
"learning_rate": 2.3871237458193983e-05,
|
| 346 |
-
"loss": 1.9211,
|
| 347 |
-
"step": 25000
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"epoch": 2.67,
|
| 351 |
-
"learning_rate": 2.334866220735786e-05,
|
| 352 |
-
"loss": 1.9698,
|
| 353 |
-
"step": 25500
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 2.72,
|
| 357 |
-
"learning_rate": 2.282608695652174e-05,
|
| 358 |
-
"loss": 1.9413,
|
| 359 |
-
"step": 26000
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 2.77,
|
| 363 |
-
"learning_rate": 2.230351170568562e-05,
|
| 364 |
-
"loss": 1.9943,
|
| 365 |
-
"step": 26500
|
| 366 |
-
},
|
| 367 |
-
{
|
| 368 |
-
"epoch": 2.82,
|
| 369 |
-
"learning_rate": 2.1780936454849498e-05,
|
| 370 |
-
"loss": 1.938,
|
| 371 |
-
"step": 27000
|
| 372 |
-
},
|
| 373 |
-
{
|
| 374 |
-
"epoch": 2.87,
|
| 375 |
-
"learning_rate": 2.125836120401338e-05,
|
| 376 |
-
"loss": 1.987,
|
| 377 |
-
"step": 27500
|
| 378 |
-
},
|
| 379 |
-
{
|
| 380 |
-
"epoch": 2.93,
|
| 381 |
-
"learning_rate": 2.073578595317726e-05,
|
| 382 |
-
"loss": 1.9455,
|
| 383 |
-
"step": 28000
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 2.98,
|
| 387 |
-
"learning_rate": 2.0213210702341138e-05,
|
| 388 |
-
"loss": 1.9788,
|
| 389 |
-
"step": 28500
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 3.0,
|
| 393 |
-
"eval_bleu": 0.08712036412034174,
|
| 394 |
-
"eval_brevity_penalty": 0.7810596870491452,
|
| 395 |
-
"eval_length_ratio": 0.8018579316519923,
|
| 396 |
-
"eval_loss": 1.8364616632461548,
|
| 397 |
-
"eval_precisions": [
|
| 398 |
-
0.1853626762201647,
|
| 399 |
-
0.1156396502935338,
|
| 400 |
-
0.08657517535834096,
|
| 401 |
-
0.08341024457775727
|
| 402 |
-
],
|
| 403 |
-
"eval_reference_length": 134020,
|
| 404 |
-
"eval_runtime": 809.3386,
|
| 405 |
-
"eval_samples_per_second": 15.762,
|
| 406 |
-
"eval_steps_per_second": 1.315,
|
| 407 |
-
"eval_translation_length": 107465,
|
| 408 |
-
"step": 28704
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 3.03,
|
| 412 |
-
"learning_rate": 1.9690635451505017e-05,
|
| 413 |
-
"loss": 1.9465,
|
| 414 |
-
"step": 29000
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 3.08,
|
| 418 |
-
"learning_rate": 1.9168060200668896e-05,
|
| 419 |
-
"loss": 1.9071,
|
| 420 |
-
"step": 29500
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 3.14,
|
| 424 |
-
"learning_rate": 1.8645484949832775e-05,
|
| 425 |
-
"loss": 1.893,
|
| 426 |
-
"step": 30000
|
| 427 |
-
},
|
| 428 |
-
{
|
| 429 |
-
"epoch": 3.19,
|
| 430 |
-
"learning_rate": 1.8122909698996657e-05,
|
| 431 |
-
"loss": 1.8895,
|
| 432 |
-
"step": 30500
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"epoch": 3.24,
|
| 436 |
-
"learning_rate": 1.7600334448160536e-05,
|
| 437 |
-
"loss": 1.8914,
|
| 438 |
-
"step": 31000
|
| 439 |
-
},
|
| 440 |
-
{
|
| 441 |
-
"epoch": 3.29,
|
| 442 |
-
"learning_rate": 1.7077759197324418e-05,
|
| 443 |
-
"loss": 1.9214,
|
| 444 |
-
"step": 31500
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 3.34,
|
| 448 |
-
"learning_rate": 1.6555183946488294e-05,
|
| 449 |
-
"loss": 1.8911,
|
| 450 |
-
"step": 32000
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 3.4,
|
| 454 |
-
"learning_rate": 1.6032608695652173e-05,
|
| 455 |
-
"loss": 1.9149,
|
| 456 |
-
"step": 32500
|
| 457 |
-
},
|
| 458 |
-
{
|
| 459 |
-
"epoch": 3.45,
|
| 460 |
-
"learning_rate": 1.5510033444816055e-05,
|
| 461 |
-
"loss": 1.902,
|
| 462 |
-
"step": 33000
|
| 463 |
-
},
|
| 464 |
-
{
|
| 465 |
-
"epoch": 3.5,
|
| 466 |
-
"learning_rate": 1.4987458193979934e-05,
|
| 467 |
-
"loss": 1.9066,
|
| 468 |
-
"step": 33500
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 3.55,
|
| 472 |
-
"learning_rate": 1.4464882943143812e-05,
|
| 473 |
-
"loss": 1.9027,
|
| 474 |
-
"step": 34000
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 3.61,
|
| 478 |
-
"learning_rate": 1.3942307692307693e-05,
|
| 479 |
-
"loss": 1.9232,
|
| 480 |
-
"step": 34500
|
| 481 |
-
},
|
| 482 |
-
{
|
| 483 |
-
"epoch": 3.66,
|
| 484 |
-
"learning_rate": 1.3419732441471572e-05,
|
| 485 |
-
"loss": 1.926,
|
| 486 |
-
"step": 35000
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 3.71,
|
| 490 |
-
"learning_rate": 1.2897157190635452e-05,
|
| 491 |
-
"loss": 1.87,
|
| 492 |
-
"step": 35500
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 3.76,
|
| 496 |
-
"learning_rate": 1.2374581939799331e-05,
|
| 497 |
-
"loss": 1.9055,
|
| 498 |
-
"step": 36000
|
| 499 |
-
},
|
| 500 |
-
{
|
| 501 |
-
"epoch": 3.81,
|
| 502 |
-
"learning_rate": 1.1852006688963212e-05,
|
| 503 |
-
"loss": 1.952,
|
| 504 |
-
"step": 36500
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 3.87,
|
| 508 |
-
"learning_rate": 1.132943143812709e-05,
|
| 509 |
-
"loss": 1.8896,
|
| 510 |
-
"step": 37000
|
| 511 |
-
},
|
| 512 |
-
{
|
| 513 |
-
"epoch": 3.92,
|
| 514 |
-
"learning_rate": 1.080685618729097e-05,
|
| 515 |
-
"loss": 1.9144,
|
| 516 |
-
"step": 37500
|
| 517 |
-
},
|
| 518 |
-
{
|
| 519 |
-
"epoch": 3.97,
|
| 520 |
-
"learning_rate": 1.028428093645485e-05,
|
| 521 |
-
"loss": 1.9447,
|
| 522 |
-
"step": 38000
|
| 523 |
-
},
|
| 524 |
-
{
|
| 525 |
-
"epoch": 4.0,
|
| 526 |
-
"eval_bleu": 0.08669705678202416,
|
| 527 |
-
"eval_brevity_penalty": 0.7634478532624474,
|
| 528 |
-
"eval_length_ratio": 0.7874570959558275,
|
| 529 |
-
"eval_loss": 1.8330533504486084,
|
| 530 |
-
"eval_precisions": [
|
| 531 |
-
0.1886293646657507,
|
| 532 |
-
0.11698894134385307,
|
| 533 |
-
0.08828452928243054,
|
| 534 |
-
0.08536133232489508
|
| 535 |
-
],
|
| 536 |
-
"eval_reference_length": 134020,
|
| 537 |
-
"eval_runtime": 811.3896,
|
| 538 |
-
"eval_samples_per_second": 15.722,
|
| 539 |
-
"eval_steps_per_second": 1.311,
|
| 540 |
-
"eval_translation_length": 105535,
|
| 541 |
-
"step": 38272
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 4.02,
|
| 545 |
-
"learning_rate": 9.76170568561873e-06,
|
| 546 |
-
"loss": 1.9148,
|
| 547 |
-
"step": 38500
|
| 548 |
-
},
|
| 549 |
-
{
|
| 550 |
-
"epoch": 4.08,
|
| 551 |
-
"learning_rate": 9.239130434782608e-06,
|
| 552 |
-
"loss": 1.9031,
|
| 553 |
-
"step": 39000
|
| 554 |
-
},
|
| 555 |
-
{
|
| 556 |
-
"epoch": 4.13,
|
| 557 |
-
"learning_rate": 8.716555183946488e-06,
|
| 558 |
-
"loss": 1.8884,
|
| 559 |
-
"step": 39500
|
| 560 |
-
},
|
| 561 |
-
{
|
| 562 |
-
"epoch": 4.18,
|
| 563 |
-
"learning_rate": 8.193979933110369e-06,
|
| 564 |
-
"loss": 1.8267,
|
| 565 |
-
"step": 40000
|
| 566 |
-
},
|
| 567 |
-
{
|
| 568 |
-
"epoch": 4.23,
|
| 569 |
-
"learning_rate": 7.671404682274248e-06,
|
| 570 |
-
"loss": 1.8556,
|
| 571 |
-
"step": 40500
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 4.29,
|
| 575 |
-
"learning_rate": 7.148829431438127e-06,
|
| 576 |
-
"loss": 1.9098,
|
| 577 |
-
"step": 41000
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 4.34,
|
| 581 |
-
"learning_rate": 6.6262541806020064e-06,
|
| 582 |
-
"loss": 1.8693,
|
| 583 |
-
"step": 41500
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 4.39,
|
| 587 |
-
"learning_rate": 6.103678929765887e-06,
|
| 588 |
-
"loss": 1.8622,
|
| 589 |
-
"step": 42000
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 4.44,
|
| 593 |
-
"learning_rate": 5.581103678929766e-06,
|
| 594 |
-
"loss": 1.8848,
|
| 595 |
-
"step": 42500
|
| 596 |
-
},
|
| 597 |
-
{
|
| 598 |
-
"epoch": 4.49,
|
| 599 |
-
"learning_rate": 5.0585284280936456e-06,
|
| 600 |
-
"loss": 1.8998,
|
| 601 |
-
"step": 43000
|
| 602 |
-
},
|
| 603 |
-
{
|
| 604 |
-
"epoch": 4.55,
|
| 605 |
-
"learning_rate": 4.535953177257525e-06,
|
| 606 |
-
"loss": 1.9059,
|
| 607 |
-
"step": 43500
|
| 608 |
-
},
|
| 609 |
-
{
|
| 610 |
-
"epoch": 4.6,
|
| 611 |
-
"learning_rate": 4.013377926421405e-06,
|
| 612 |
-
"loss": 1.8852,
|
| 613 |
-
"step": 44000
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 4.65,
|
| 617 |
-
"learning_rate": 3.4908026755852843e-06,
|
| 618 |
-
"loss": 1.8984,
|
| 619 |
-
"step": 44500
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 4.7,
|
| 623 |
-
"learning_rate": 2.968227424749164e-06,
|
| 624 |
-
"loss": 1.8517,
|
| 625 |
-
"step": 45000
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 4.76,
|
| 629 |
-
"learning_rate": 2.4456521739130437e-06,
|
| 630 |
-
"loss": 1.917,
|
| 631 |
-
"step": 45500
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 4.81,
|
| 635 |
-
"learning_rate": 1.9230769230769234e-06,
|
| 636 |
-
"loss": 1.9134,
|
| 637 |
-
"step": 46000
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 4.86,
|
| 641 |
-
"learning_rate": 1.4005016722408027e-06,
|
| 642 |
-
"loss": 1.8772,
|
| 643 |
-
"step": 46500
|
| 644 |
-
},
|
| 645 |
-
{
|
| 646 |
-
"epoch": 4.91,
|
| 647 |
-
"learning_rate": 8.779264214046823e-07,
|
| 648 |
-
"loss": 1.8938,
|
| 649 |
-
"step": 47000
|
| 650 |
-
},
|
| 651 |
-
{
|
| 652 |
-
"epoch": 4.96,
|
| 653 |
-
"learning_rate": 3.553511705685619e-07,
|
| 654 |
-
"loss": 1.8652,
|
| 655 |
-
"step": 47500
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 5.0,
|
| 659 |
-
"eval_bleu": 0.08741495854193439,
|
| 660 |
-
"eval_brevity_penalty": 0.7850055548072304,
|
| 661 |
-
"eval_length_ratio": 0.8051111774362035,
|
| 662 |
-
"eval_loss": 1.8320603370666504,
|
| 663 |
"eval_precisions": [
|
| 664 |
-
0
|
| 665 |
-
0
|
| 666 |
-
0
|
| 667 |
-
0
|
| 668 |
],
|
| 669 |
-
"eval_reference_length":
|
| 670 |
-
"eval_runtime":
|
| 671 |
-
"eval_samples_per_second":
|
| 672 |
-
"eval_steps_per_second":
|
| 673 |
-
"eval_translation_length":
|
| 674 |
-
"step":
|
| 675 |
}
|
| 676 |
],
|
| 677 |
"logging_steps": 500,
|
| 678 |
-
"max_steps":
|
| 679 |
"num_train_epochs": 5,
|
| 680 |
"save_steps": 500,
|
| 681 |
-
"total_flos":
|
| 682 |
"trial_name": null,
|
| 683 |
"trial_params": null
|
| 684 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.4803038835525513,
|
| 3 |
+
"best_model_checkpoint": "dq158/morbius/checkpoint-1475",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1475,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.34,
|
| 13 |
+
"learning_rate": 4.6610169491525425e-05,
|
| 14 |
+
"loss": 1.8922,
|
| 15 |
"step": 500
|
| 16 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
{
|
| 18 |
"epoch": 0.68,
|
| 19 |
+
"learning_rate": 4.3220338983050854e-05,
|
| 20 |
+
"loss": 1.7522,
|
| 21 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
},
|
| 23 |
{
|
| 24 |
"epoch": 1.0,
|
| 25 |
+
"eval_bleu": 1.0,
|
| 26 |
+
"eval_brevity_penalty": 1.0,
|
| 27 |
+
"eval_length_ratio": 1.0,
|
| 28 |
+
"eval_loss": 1.4803038835525513,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
"eval_precisions": [
|
| 30 |
+
1.0,
|
| 31 |
+
1.0,
|
| 32 |
+
1.0,
|
| 33 |
+
1.0
|
| 34 |
],
|
| 35 |
+
"eval_reference_length": 35996,
|
| 36 |
+
"eval_runtime": 295.5668,
|
| 37 |
+
"eval_samples_per_second": 6.652,
|
| 38 |
+
"eval_steps_per_second": 0.555,
|
| 39 |
+
"eval_translation_length": 35996,
|
| 40 |
+
"step": 1475
|
| 41 |
}
|
| 42 |
],
|
| 43 |
"logging_steps": 500,
|
| 44 |
+
"max_steps": 7375,
|
| 45 |
"num_train_epochs": 5,
|
| 46 |
"save_steps": 500,
|
| 47 |
+
"total_flos": 1.2116096594214912e+16,
|
| 48 |
"trial_name": null,
|
| 49 |
"trial_params": null
|
| 50 |
}
|