Training in progress, step 1000, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scaler.pt +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +166 -2293
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1583480280
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02fd35dd30945d2a331fcc4229d682044d2333729fe2d642c02c88a4f8950290
|
| 3 |
size 1583480280
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3166958572
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2bfd5531726b88f6d30b16379bf8128025316616e6a5fdc46f82548ddef8b2f4
|
| 3 |
size 3166958572
|
last-checkpoint/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 sha256:172c6e9da1198fecb1104ae5588ac154055d22275bb62749b67f1d60379ff0a7
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14ae2a2128444abab378aa06c09a61a84665f758fcc19fc46f5789b0bc1b5665
|
| 3 |
size 1383
|
last-checkpoint/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 sha256:68aec417c91400a5fbe9c98d7447dabd74ed3b0812272a5f21d640985e919bad
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 1.
|
| 4 |
-
"best_model_checkpoint": "hieptt/vietnamese-correction-ft/checkpoint-
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -18,2839 +18,712 @@
|
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.00024186136506554442,
|
| 21 |
-
"grad_norm": 7.
|
| 22 |
"learning_rate": 4.5e-07,
|
| 23 |
-
"loss": 4.
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.00048372273013108885,
|
| 28 |
-
"grad_norm": 9.
|
| 29 |
"learning_rate": 9.5e-07,
|
| 30 |
-
"loss": 4.
|
| 31 |
"step": 20
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.0007255840951966333,
|
| 35 |
-
"grad_norm": 12.
|
| 36 |
"learning_rate": 1.45e-06,
|
| 37 |
-
"loss": 4.
|
| 38 |
"step": 30
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.0009674454602621777,
|
| 42 |
-
"grad_norm": 13.
|
| 43 |
"learning_rate": 1.95e-06,
|
| 44 |
-
"loss": 4.
|
| 45 |
"step": 40
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.0012093068253277222,
|
| 49 |
-
"grad_norm": 14.
|
| 50 |
"learning_rate": 2.4500000000000003e-06,
|
| 51 |
-
"loss": 3.
|
| 52 |
"step": 50
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.0014511681903932665,
|
| 56 |
-
"grad_norm": 4.
|
| 57 |
"learning_rate": 2.95e-06,
|
| 58 |
-
"loss": 2.
|
| 59 |
"step": 60
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.001693029555458811,
|
| 63 |
-
"grad_norm": 2.
|
| 64 |
"learning_rate": 3.4500000000000004e-06,
|
| 65 |
"loss": 2.3633,
|
| 66 |
"step": 70
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.0019348909205243554,
|
| 70 |
-
"grad_norm": 1.
|
| 71 |
"learning_rate": 3.95e-06,
|
| 72 |
"loss": 2.0775,
|
| 73 |
"step": 80
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.0021767522855899,
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
"learning_rate": 4.45e-06,
|
| 79 |
"loss": 1.9554,
|
| 80 |
"step": 90
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.0024186136506554445,
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
"learning_rate": 4.950000000000001e-06,
|
| 86 |
"loss": 1.887,
|
| 87 |
"step": 100
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.0026604750157209886,
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
"learning_rate": 5.45e-06,
|
| 93 |
"loss": 1.8405,
|
| 94 |
"step": 110
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.002902336380786533,
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
"learning_rate": 5.95e-06,
|
| 100 |
"loss": 1.8095,
|
| 101 |
"step": 120
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.0031441977458520776,
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
"learning_rate": 6.45e-06,
|
| 107 |
"loss": 1.7881,
|
| 108 |
"step": 130
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.003386059110917622,
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
"learning_rate": 6.950000000000001e-06,
|
| 114 |
"loss": 1.7647,
|
| 115 |
"step": 140
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.0036279204759831663,
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
"learning_rate": 7.45e-06,
|
| 121 |
"loss": 1.7577,
|
| 122 |
"step": 150
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.003869781841048711,
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
"learning_rate": 7.95e-06,
|
| 128 |
"loss": 1.7499,
|
| 129 |
"step": 160
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.004111643206114255,
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
"learning_rate": 8.45e-06,
|
| 135 |
"loss": 1.741,
|
| 136 |
"step": 170
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.0043535045711798,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
"learning_rate": 8.95e-06,
|
| 142 |
"loss": 1.7283,
|
| 143 |
"step": 180
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.004595365936245344,
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
"learning_rate": 9.450000000000001e-06,
|
| 149 |
"loss": 1.707,
|
| 150 |
"step": 190
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.004837227301310889,
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
"learning_rate": 9.950000000000001e-06,
|
| 156 |
"loss": 1.706,
|
| 157 |
"step": 200
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.005079088666376433,
|
| 161 |
-
"grad_norm": 0.
|
| 162 |
"learning_rate": 1.045e-05,
|
| 163 |
"loss": 1.7065,
|
| 164 |
"step": 210
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.005320950031441977,
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
"learning_rate": 1.095e-05,
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 220
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.005562811396507522,
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
"learning_rate": 1.145e-05,
|
| 177 |
"loss": 1.6822,
|
| 178 |
"step": 230
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.005804672761573066,
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
"learning_rate": 1.195e-05,
|
| 184 |
"loss": 1.6794,
|
| 185 |
"step": 240
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.006046534126638611,
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
"learning_rate": 1.2450000000000001e-05,
|
| 191 |
"loss": 1.6587,
|
| 192 |
"step": 250
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.006288395491704155,
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
"learning_rate": 1.2950000000000001e-05,
|
| 198 |
"loss": 1.6756,
|
| 199 |
"step": 260
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.006530256856769699,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
"learning_rate": 1.3450000000000002e-05,
|
| 205 |
"loss": 1.6723,
|
| 206 |
"step": 270
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.006772118221835244,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
"learning_rate": 1.3950000000000002e-05,
|
| 212 |
"loss": 1.6605,
|
| 213 |
"step": 280
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.007013979586900788,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
"learning_rate": 1.4449999999999999e-05,
|
| 219 |
-
"loss": 1.
|
| 220 |
"step": 290
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.0072558409519663325,
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
"learning_rate": 1.4950000000000001e-05,
|
| 226 |
"loss": 1.6441,
|
| 227 |
"step": 300
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.0074977023170318775,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
"learning_rate": 1.545e-05,
|
| 233 |
"loss": 1.6478,
|
| 234 |
"step": 310
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.007739563682097422,
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
"learning_rate": 1.595e-05,
|
| 240 |
"loss": 1.6405,
|
| 241 |
"step": 320
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.007981425047162967,
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
"learning_rate": 1.645e-05,
|
| 247 |
"loss": 1.6441,
|
| 248 |
"step": 330
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.00822328641222851,
|
| 252 |
-
"grad_norm": 0.
|
| 253 |
"learning_rate": 1.6950000000000002e-05,
|
| 254 |
"loss": 1.6401,
|
| 255 |
"step": 340
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.008465147777294055,
|
| 259 |
-
"grad_norm": 0.
|
| 260 |
"learning_rate": 1.745e-05,
|
| 261 |
"loss": 1.6385,
|
| 262 |
"step": 350
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.0087070091423596,
|
| 266 |
-
"grad_norm": 0.
|
| 267 |
"learning_rate": 1.795e-05,
|
| 268 |
"loss": 1.6222,
|
| 269 |
"step": 360
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.008948870507425145,
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
"learning_rate": 1.845e-05,
|
| 275 |
"loss": 1.6267,
|
| 276 |
"step": 370
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.009190731872490688,
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
"learning_rate": 1.895e-05,
|
| 282 |
"loss": 1.6265,
|
| 283 |
"step": 380
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.009432593237556233,
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
"learning_rate": 1.9450000000000002e-05,
|
| 289 |
"loss": 1.6208,
|
| 290 |
"step": 390
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.009674454602621778,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
"learning_rate": 1.995e-05,
|
| 296 |
"loss": 1.6137,
|
| 297 |
"step": 400
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.009916315967687321,
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
"learning_rate": 2.045e-05,
|
| 303 |
"loss": 1.6196,
|
| 304 |
"step": 410
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.010158177332752866,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
"learning_rate": 2.095e-05,
|
| 310 |
"loss": 1.6049,
|
| 311 |
"step": 420
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.010400038697818411,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
"learning_rate": 2.145e-05,
|
| 317 |
"loss": 1.6083,
|
| 318 |
"step": 430
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.010641900062883954,
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
"learning_rate": 2.195e-05,
|
| 324 |
"loss": 1.6074,
|
| 325 |
"step": 440
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.0108837614279495,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
"learning_rate": 2.245e-05,
|
| 331 |
"loss": 1.6033,
|
| 332 |
"step": 450
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.011125622793015044,
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
"learning_rate": 2.2950000000000002e-05,
|
| 338 |
"loss": 1.5976,
|
| 339 |
"step": 460
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.011367484158080587,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
"learning_rate": 2.345e-05,
|
| 345 |
"loss": 1.5982,
|
| 346 |
"step": 470
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.011609345523146132,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
"learning_rate": 2.395e-05,
|
| 352 |
"loss": 1.5983,
|
| 353 |
"step": 480
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.011851206888211677,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
"learning_rate": 2.445e-05,
|
| 359 |
-
"loss": 1.
|
| 360 |
"step": 490
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.012093068253277222,
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
"learning_rate": 2.495e-05,
|
| 366 |
"loss": 1.5944,
|
| 367 |
"step": 500
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.012334929618342766,
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
"learning_rate": 2.5450000000000002e-05,
|
| 373 |
-
"loss": 1.
|
| 374 |
"step": 510
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.01257679098340831,
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
"learning_rate": 2.595e-05,
|
| 380 |
"loss": 1.5945,
|
| 381 |
"step": 520
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.012818652348473855,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
"learning_rate": 2.6450000000000003e-05,
|
| 387 |
-
"loss": 1.
|
| 388 |
"step": 530
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.013060513713539399,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
"learning_rate": 2.6950000000000005e-05,
|
| 394 |
"loss": 1.586,
|
| 395 |
"step": 540
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.013302375078604944,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
"learning_rate": 2.7450000000000003e-05,
|
| 401 |
"loss": 1.591,
|
| 402 |
"step": 550
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.013544236443670489,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
"learning_rate": 2.7950000000000005e-05,
|
| 408 |
-
"loss": 1.
|
| 409 |
"step": 560
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.013786097808736032,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
"learning_rate": 2.845e-05,
|
| 415 |
-
"loss": 1.
|
| 416 |
"step": 570
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.014027959173801577,
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
"learning_rate": 2.895e-05,
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 580
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.014269820538867122,
|
| 427 |
-
"grad_norm": 0.
|
| 428 |
"learning_rate": 2.945e-05,
|
| 429 |
-
"loss": 1.
|
| 430 |
"step": 590
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.014511681903932665,
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
"learning_rate": 2.995e-05,
|
| 436 |
"loss": 1.5796,
|
| 437 |
"step": 600
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.01475354326899821,
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
"learning_rate": 3.045e-05,
|
| 443 |
-
"loss": 1.
|
| 444 |
"step": 610
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.014995404634063755,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
"learning_rate": 3.095e-05,
|
| 450 |
-
"loss": 1.
|
| 451 |
"step": 620
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.015237265999129298,
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
"learning_rate": 3.145e-05,
|
| 457 |
-
"loss": 1.
|
| 458 |
"step": 630
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.015479127364194843,
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
"learning_rate": 3.1950000000000004e-05,
|
| 464 |
-
"loss": 1.
|
| 465 |
"step": 640
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.015720988729260386,
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
"learning_rate": 3.245e-05,
|
| 471 |
-
"loss": 1.
|
| 472 |
"step": 650
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.015962850094325933,
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
"learning_rate": 3.295e-05,
|
| 478 |
-
"loss": 1.
|
| 479 |
"step": 660
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.016204711459391476,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
"learning_rate": 3.345000000000001e-05,
|
| 485 |
-
"loss": 1.
|
| 486 |
"step": 670
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.01644657282445702,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
"learning_rate": 3.3950000000000005e-05,
|
| 492 |
-
"loss": 1.
|
| 493 |
"step": 680
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.016688434189522566,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
"learning_rate": 3.445e-05,
|
| 499 |
-
"loss": 1.
|
| 500 |
"step": 690
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.01693029555458811,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
"learning_rate": 3.495e-05,
|
| 506 |
-
"loss": 1.
|
| 507 |
"step": 700
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.017172156919653656,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
"learning_rate": 3.545e-05,
|
| 513 |
-
"loss": 1.
|
| 514 |
"step": 710
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.0174140182847192,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
"learning_rate": 3.595e-05,
|
| 520 |
-
"loss": 1.
|
| 521 |
"step": 720
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.017655879649784743,
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
"learning_rate": 3.645e-05,
|
| 527 |
-
"loss": 1.
|
| 528 |
"step": 730
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.01789774101485029,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
"learning_rate": 3.6950000000000004e-05,
|
| 534 |
-
"loss": 1.
|
| 535 |
"step": 740
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.018139602379915833,
|
| 539 |
-
"grad_norm": 0.
|
| 540 |
"learning_rate": 3.745e-05,
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 750
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.018381463744981376,
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
"learning_rate": 3.795e-05,
|
| 548 |
-
"loss": 1.
|
| 549 |
"step": 760
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.018623325110046923,
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
"learning_rate": 3.845e-05,
|
| 555 |
-
"loss": 1.
|
| 556 |
"step": 770
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.018865186475112466,
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
"learning_rate": 3.8950000000000005e-05,
|
| 562 |
-
"loss": 1.
|
| 563 |
"step": 780
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.01910704784017801,
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
"learning_rate": 3.9450000000000003e-05,
|
| 569 |
-
"loss": 1.
|
| 570 |
"step": 790
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.019348909205243556,
|
| 574 |
-
"grad_norm": 0.
|
| 575 |
"learning_rate": 3.995e-05,
|
| 576 |
-
"loss": 1.
|
| 577 |
"step": 800
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.0195907705703091,
|
| 581 |
-
"grad_norm": 0.
|
| 582 |
"learning_rate": 4.045000000000001e-05,
|
| 583 |
-
"loss": 1.
|
| 584 |
"step": 810
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.019832631935374642,
|
| 588 |
-
"grad_norm": 0.
|
| 589 |
"learning_rate": 4.095e-05,
|
| 590 |
-
"loss": 1.
|
| 591 |
"step": 820
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.02007449330044019,
|
| 595 |
-
"grad_norm": 0.
|
| 596 |
"learning_rate": 4.145e-05,
|
| 597 |
-
"loss": 1.
|
| 598 |
"step": 830
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.020316354665505732,
|
| 602 |
-
"grad_norm": 0.
|
| 603 |
"learning_rate": 4.195e-05,
|
| 604 |
-
"loss": 1.
|
| 605 |
"step": 840
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.020558216030571275,
|
| 609 |
-
"grad_norm": 0.
|
| 610 |
"learning_rate": 4.245e-05,
|
| 611 |
-
"loss": 1.
|
| 612 |
"step": 850
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.020800077395636822,
|
| 616 |
-
"grad_norm": 0.
|
| 617 |
"learning_rate": 4.295e-05,
|
| 618 |
-
"loss": 1.
|
| 619 |
"step": 860
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.021041938760702365,
|
| 623 |
-
"grad_norm": 0.
|
| 624 |
"learning_rate": 4.345e-05,
|
| 625 |
-
"loss": 1.
|
| 626 |
"step": 870
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.02128380012576791,
|
| 630 |
-
"grad_norm": 0.
|
| 631 |
"learning_rate": 4.3950000000000004e-05,
|
| 632 |
-
"loss": 1.
|
| 633 |
"step": 880
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.021525661490833455,
|
| 637 |
-
"grad_norm": 0.
|
| 638 |
"learning_rate": 4.445e-05,
|
| 639 |
-
"loss": 1.
|
| 640 |
"step": 890
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.021767522855899,
|
| 644 |
-
"grad_norm": 0.
|
| 645 |
"learning_rate": 4.495e-05,
|
| 646 |
-
"loss": 1.
|
| 647 |
"step": 900
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.02200938422096454,
|
| 651 |
-
"grad_norm": 0.
|
| 652 |
"learning_rate": 4.545000000000001e-05,
|
| 653 |
-
"loss": 1.
|
| 654 |
"step": 910
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.02225124558603009,
|
| 658 |
-
"grad_norm": 0.
|
| 659 |
"learning_rate": 4.5950000000000006e-05,
|
| 660 |
-
"loss": 1.
|
| 661 |
"step": 920
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.02249310695109563,
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
"learning_rate": 4.6450000000000004e-05,
|
| 667 |
-
"loss": 1.
|
| 668 |
"step": 930
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.022734968316161175,
|
| 672 |
-
"grad_norm": 0.
|
| 673 |
"learning_rate": 4.695e-05,
|
| 674 |
-
"loss": 1.
|
| 675 |
"step": 940
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.02297682968122672,
|
| 679 |
-
"grad_norm": 0.
|
| 680 |
"learning_rate": 4.745e-05,
|
| 681 |
-
"loss": 1.
|
| 682 |
"step": 950
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.023218691046292265,
|
| 686 |
-
"grad_norm": 0.
|
| 687 |
"learning_rate": 4.795e-05,
|
| 688 |
-
"loss": 1.
|
| 689 |
"step": 960
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.023460552411357808,
|
| 693 |
-
"grad_norm": 0.
|
| 694 |
"learning_rate": 4.845e-05,
|
| 695 |
-
"loss": 1.
|
| 696 |
"step": 970
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.023702413776423355,
|
| 700 |
-
"grad_norm":
|
| 701 |
"learning_rate": 4.8950000000000004e-05,
|
| 702 |
-
"loss": 1.
|
| 703 |
"step": 980
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.023944275141488898,
|
| 707 |
-
"grad_norm":
|
| 708 |
"learning_rate": 4.945e-05,
|
| 709 |
-
"loss": 1.
|
| 710 |
"step": 990
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.024186136506554445,
|
| 714 |
-
"grad_norm": 0.
|
| 715 |
"learning_rate": 4.995e-05,
|
| 716 |
-
"loss": 1.
|
| 717 |
"step": 1000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.024186136506554445,
|
| 721 |
-
"eval_loss": 1.
|
| 722 |
-
"eval_runtime":
|
| 723 |
-
"eval_sacrebleu":
|
| 724 |
-
"eval_samples_per_second":
|
| 725 |
-
"eval_steps_per_second": 0.
|
| 726 |
"step": 1000
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 0.024427997871619988,
|
| 730 |
-
"grad_norm": 0.8520330786705017,
|
| 731 |
-
"learning_rate": 4.999999583799493e-05,
|
| 732 |
-
"loss": 1.7127,
|
| 733 |
-
"step": 1010
|
| 734 |
-
},
|
| 735 |
-
{
|
| 736 |
-
"epoch": 0.02466985923668553,
|
| 737 |
-
"grad_norm": 3.2178781032562256,
|
| 738 |
-
"learning_rate": 4.999998145081868e-05,
|
| 739 |
-
"loss": 1.6545,
|
| 740 |
-
"step": 1020
|
| 741 |
-
},
|
| 742 |
-
{
|
| 743 |
-
"epoch": 0.024911720601751078,
|
| 744 |
-
"grad_norm": 0.7130942940711975,
|
| 745 |
-
"learning_rate": 4.999995678709439e-05,
|
| 746 |
-
"loss": 1.5647,
|
| 747 |
-
"step": 1030
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 0.02515358196681662,
|
| 751 |
-
"grad_norm": 0.487717866897583,
|
| 752 |
-
"learning_rate": 4.999992184683219e-05,
|
| 753 |
-
"loss": 1.5527,
|
| 754 |
-
"step": 1040
|
| 755 |
-
},
|
| 756 |
-
{
|
| 757 |
-
"epoch": 0.025395443331882164,
|
| 758 |
-
"grad_norm": 0.4338527023792267,
|
| 759 |
-
"learning_rate": 4.999987663004646e-05,
|
| 760 |
-
"loss": 1.5538,
|
| 761 |
-
"step": 1050
|
| 762 |
-
},
|
| 763 |
-
{
|
| 764 |
-
"epoch": 0.02563730469694771,
|
| 765 |
-
"grad_norm": 0.4418632388114929,
|
| 766 |
-
"learning_rate": 4.9999821136755766e-05,
|
| 767 |
-
"loss": 1.5522,
|
| 768 |
-
"step": 1060
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 0.025879166062013254,
|
| 772 |
-
"grad_norm": 0.4217578172683716,
|
| 773 |
-
"learning_rate": 4.9999755366982925e-05,
|
| 774 |
-
"loss": 1.5358,
|
| 775 |
-
"step": 1070
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 0.026121027427078797,
|
| 779 |
-
"grad_norm": 0.6195696592330933,
|
| 780 |
-
"learning_rate": 4.999967932075499e-05,
|
| 781 |
-
"loss": 1.5551,
|
| 782 |
-
"step": 1080
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 0.026362888792144344,
|
| 786 |
-
"grad_norm": 0.48337095975875854,
|
| 787 |
-
"learning_rate": 4.99995929981032e-05,
|
| 788 |
-
"loss": 1.5554,
|
| 789 |
-
"step": 1090
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 0.026604750157209887,
|
| 793 |
-
"grad_norm": 0.3626619279384613,
|
| 794 |
-
"learning_rate": 4.999949639906304e-05,
|
| 795 |
-
"loss": 1.5524,
|
| 796 |
-
"step": 1100
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 0.02684661152227543,
|
| 800 |
-
"grad_norm": 0.4561481773853302,
|
| 801 |
-
"learning_rate": 4.999938952367422e-05,
|
| 802 |
-
"loss": 1.5543,
|
| 803 |
-
"step": 1110
|
| 804 |
-
},
|
| 805 |
-
{
|
| 806 |
-
"epoch": 0.027088472887340977,
|
| 807 |
-
"grad_norm": 0.9597444534301758,
|
| 808 |
-
"learning_rate": 4.999927237198069e-05,
|
| 809 |
-
"loss": 1.5503,
|
| 810 |
-
"step": 1120
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 0.02733033425240652,
|
| 814 |
-
"grad_norm": 0.4445475935935974,
|
| 815 |
-
"learning_rate": 4.999914494403059e-05,
|
| 816 |
-
"loss": 1.5405,
|
| 817 |
-
"step": 1130
|
| 818 |
-
},
|
| 819 |
-
{
|
| 820 |
-
"epoch": 0.027572195617472064,
|
| 821 |
-
"grad_norm": 0.617169201374054,
|
| 822 |
-
"learning_rate": 4.9999007239876294e-05,
|
| 823 |
-
"loss": 1.5472,
|
| 824 |
-
"step": 1140
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 0.02781405698253761,
|
| 828 |
-
"grad_norm": 0.43119344115257263,
|
| 829 |
-
"learning_rate": 4.999885925957443e-05,
|
| 830 |
-
"loss": 1.5397,
|
| 831 |
-
"step": 1150
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 0.028055918347603154,
|
| 835 |
-
"grad_norm": 0.36087217926979065,
|
| 836 |
-
"learning_rate": 4.99987010031858e-05,
|
| 837 |
-
"loss": 1.5461,
|
| 838 |
-
"step": 1160
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 0.028297779712668697,
|
| 842 |
-
"grad_norm": 0.38970816135406494,
|
| 843 |
-
"learning_rate": 4.9998532470775484e-05,
|
| 844 |
-
"loss": 1.5311,
|
| 845 |
-
"step": 1170
|
| 846 |
-
},
|
| 847 |
-
{
|
| 848 |
-
"epoch": 0.028539641077734244,
|
| 849 |
-
"grad_norm": 0.4160713255405426,
|
| 850 |
-
"learning_rate": 4.999835366241274e-05,
|
| 851 |
-
"loss": 1.5447,
|
| 852 |
-
"step": 1180
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 0.028781502442799787,
|
| 856 |
-
"grad_norm": 1.10483980178833,
|
| 857 |
-
"learning_rate": 4.9998164578171076e-05,
|
| 858 |
-
"loss": 1.5359,
|
| 859 |
-
"step": 1190
|
| 860 |
-
},
|
| 861 |
-
{
|
| 862 |
-
"epoch": 0.02902336380786533,
|
| 863 |
-
"grad_norm": 0.3932547867298126,
|
| 864 |
-
"learning_rate": 4.999796521812822e-05,
|
| 865 |
-
"loss": 1.5291,
|
| 866 |
-
"step": 1200
|
| 867 |
-
},
|
| 868 |
-
{
|
| 869 |
-
"epoch": 0.029265225172930877,
|
| 870 |
-
"grad_norm": 0.5235729813575745,
|
| 871 |
-
"learning_rate": 4.999775558236611e-05,
|
| 872 |
-
"loss": 1.5406,
|
| 873 |
-
"step": 1210
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 0.02950708653799642,
|
| 877 |
-
"grad_norm": 0.5899940729141235,
|
| 878 |
-
"learning_rate": 4.999753567097094e-05,
|
| 879 |
-
"loss": 1.5393,
|
| 880 |
-
"step": 1220
|
| 881 |
-
},
|
| 882 |
-
{
|
| 883 |
-
"epoch": 0.029748947903061963,
|
| 884 |
-
"grad_norm": 0.6470837593078613,
|
| 885 |
-
"learning_rate": 4.9997305484033085e-05,
|
| 886 |
-
"loss": 1.5414,
|
| 887 |
-
"step": 1230
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 0.02999080926812751,
|
| 891 |
-
"grad_norm": 0.5609555244445801,
|
| 892 |
-
"learning_rate": 4.999706502164718e-05,
|
| 893 |
-
"loss": 1.5487,
|
| 894 |
-
"step": 1240
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 0.030232670633193053,
|
| 898 |
-
"grad_norm": 0.34050312638282776,
|
| 899 |
-
"learning_rate": 4.999681428391207e-05,
|
| 900 |
-
"loss": 1.5438,
|
| 901 |
-
"step": 1250
|
| 902 |
-
},
|
| 903 |
-
{
|
| 904 |
-
"epoch": 0.030474531998258596,
|
| 905 |
-
"grad_norm": 0.49677538871765137,
|
| 906 |
-
"learning_rate": 4.999655327093081e-05,
|
| 907 |
-
"loss": 1.543,
|
| 908 |
-
"step": 1260
|
| 909 |
-
},
|
| 910 |
-
{
|
| 911 |
-
"epoch": 0.030716393363324143,
|
| 912 |
-
"grad_norm": 0.5405974388122559,
|
| 913 |
-
"learning_rate": 4.999628198281072e-05,
|
| 914 |
-
"loss": 1.5373,
|
| 915 |
-
"step": 1270
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 0.030958254728389686,
|
| 919 |
-
"grad_norm": 0.5481676459312439,
|
| 920 |
-
"learning_rate": 4.999600041966328e-05,
|
| 921 |
-
"loss": 1.5377,
|
| 922 |
-
"step": 1280
|
| 923 |
-
},
|
| 924 |
-
{
|
| 925 |
-
"epoch": 0.031200116093455233,
|
| 926 |
-
"grad_norm": 0.39385202527046204,
|
| 927 |
-
"learning_rate": 4.999570858160426e-05,
|
| 928 |
-
"loss": 1.5287,
|
| 929 |
-
"step": 1290
|
| 930 |
-
},
|
| 931 |
-
{
|
| 932 |
-
"epoch": 0.03144197745852077,
|
| 933 |
-
"grad_norm": 0.4990335702896118,
|
| 934 |
-
"learning_rate": 4.999540646875361e-05,
|
| 935 |
-
"loss": 1.5345,
|
| 936 |
-
"step": 1300
|
| 937 |
-
},
|
| 938 |
-
{
|
| 939 |
-
"epoch": 0.03168383882358632,
|
| 940 |
-
"grad_norm": 0.9431979060173035,
|
| 941 |
-
"learning_rate": 4.9995094081235524e-05,
|
| 942 |
-
"loss": 1.5393,
|
| 943 |
-
"step": 1310
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 0.031925700188651866,
|
| 947 |
-
"grad_norm": 0.3421996235847473,
|
| 948 |
-
"learning_rate": 4.9994771419178396e-05,
|
| 949 |
-
"loss": 1.5367,
|
| 950 |
-
"step": 1320
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 0.03216756155371741,
|
| 954 |
-
"grad_norm": 0.38486698269844055,
|
| 955 |
-
"learning_rate": 4.999443848271489e-05,
|
| 956 |
-
"loss": 1.5335,
|
| 957 |
-
"step": 1330
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 0.03240942291878295,
|
| 961 |
-
"grad_norm": 0.4813336431980133,
|
| 962 |
-
"learning_rate": 4.9994095271981835e-05,
|
| 963 |
-
"loss": 1.5402,
|
| 964 |
-
"step": 1340
|
| 965 |
-
},
|
| 966 |
-
{
|
| 967 |
-
"epoch": 0.032651284283848496,
|
| 968 |
-
"grad_norm": 0.5125822424888611,
|
| 969 |
-
"learning_rate": 4.999374178712032e-05,
|
| 970 |
-
"loss": 1.5313,
|
| 971 |
-
"step": 1350
|
| 972 |
-
},
|
| 973 |
-
{
|
| 974 |
-
"epoch": 0.03289314564891404,
|
| 975 |
-
"grad_norm": 0.4205283224582672,
|
| 976 |
-
"learning_rate": 4.999337802827566e-05,
|
| 977 |
-
"loss": 1.5293,
|
| 978 |
-
"step": 1360
|
| 979 |
-
},
|
| 980 |
-
{
|
| 981 |
-
"epoch": 0.03313500701397959,
|
| 982 |
-
"grad_norm": 0.4249746799468994,
|
| 983 |
-
"learning_rate": 4.999300399559738e-05,
|
| 984 |
-
"loss": 1.5246,
|
| 985 |
-
"step": 1370
|
| 986 |
-
},
|
| 987 |
-
{
|
| 988 |
-
"epoch": 0.03337686837904513,
|
| 989 |
-
"grad_norm": 0.49504849314689636,
|
| 990 |
-
"learning_rate": 4.999261968923922e-05,
|
| 991 |
-
"loss": 1.5298,
|
| 992 |
-
"step": 1380
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 0.033618729744110676,
|
| 996 |
-
"grad_norm": 0.7440537214279175,
|
| 997 |
-
"learning_rate": 4.999222510935915e-05,
|
| 998 |
-
"loss": 1.5309,
|
| 999 |
-
"step": 1390
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 0.03386059110917622,
|
| 1003 |
-
"grad_norm": 0.4096105992794037,
|
| 1004 |
-
"learning_rate": 4.9991820256119385e-05,
|
| 1005 |
-
"loss": 1.5357,
|
| 1006 |
-
"step": 1400
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 0.03410245247424176,
|
| 1010 |
-
"grad_norm": 0.35250145196914673,
|
| 1011 |
-
"learning_rate": 4.999140512968634e-05,
|
| 1012 |
-
"loss": 1.531,
|
| 1013 |
-
"step": 1410
|
| 1014 |
-
},
|
| 1015 |
-
{
|
| 1016 |
-
"epoch": 0.03434431383930731,
|
| 1017 |
-
"grad_norm": 0.4399813711643219,
|
| 1018 |
-
"learning_rate": 4.999097973023065e-05,
|
| 1019 |
-
"loss": 1.5252,
|
| 1020 |
-
"step": 1420
|
| 1021 |
-
},
|
| 1022 |
-
{
|
| 1023 |
-
"epoch": 0.034586175204372856,
|
| 1024 |
-
"grad_norm": 0.41037410497665405,
|
| 1025 |
-
"learning_rate": 4.999054405792718e-05,
|
| 1026 |
-
"loss": 1.5276,
|
| 1027 |
-
"step": 1430
|
| 1028 |
-
},
|
| 1029 |
-
{
|
| 1030 |
-
"epoch": 0.0348280365694384,
|
| 1031 |
-
"grad_norm": 0.41494110226631165,
|
| 1032 |
-
"learning_rate": 4.999009811295503e-05,
|
| 1033 |
-
"loss": 1.5291,
|
| 1034 |
-
"step": 1440
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 0.03506989793450394,
|
| 1038 |
-
"grad_norm": 0.4767596423625946,
|
| 1039 |
-
"learning_rate": 4.998964189549751e-05,
|
| 1040 |
-
"loss": 1.5352,
|
| 1041 |
-
"step": 1450
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 0.035311759299569485,
|
| 1045 |
-
"grad_norm": 0.4275621771812439,
|
| 1046 |
-
"learning_rate": 4.9989175405742135e-05,
|
| 1047 |
-
"loss": 1.5303,
|
| 1048 |
-
"step": 1460
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 0.03555362066463503,
|
| 1052 |
-
"grad_norm": 0.44721493124961853,
|
| 1053 |
-
"learning_rate": 4.998869864388068e-05,
|
| 1054 |
-
"loss": 1.5371,
|
| 1055 |
-
"step": 1470
|
| 1056 |
-
},
|
| 1057 |
-
{
|
| 1058 |
-
"epoch": 0.03579548202970058,
|
| 1059 |
-
"grad_norm": 0.5296744108200073,
|
| 1060 |
-
"learning_rate": 4.998821161010912e-05,
|
| 1061 |
-
"loss": 1.5342,
|
| 1062 |
-
"step": 1480
|
| 1063 |
-
},
|
| 1064 |
-
{
|
| 1065 |
-
"epoch": 0.03603734339476612,
|
| 1066 |
-
"grad_norm": 0.39906907081604004,
|
| 1067 |
-
"learning_rate": 4.9987714304627655e-05,
|
| 1068 |
-
"loss": 1.5258,
|
| 1069 |
-
"step": 1490
|
| 1070 |
-
},
|
| 1071 |
-
{
|
| 1072 |
-
"epoch": 0.036279204759831665,
|
| 1073 |
-
"grad_norm": 0.5269441604614258,
|
| 1074 |
-
"learning_rate": 4.9987206727640703e-05,
|
| 1075 |
-
"loss": 1.5317,
|
| 1076 |
-
"step": 1500
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 0.03652106612489721,
|
| 1080 |
-
"grad_norm": 0.3892313838005066,
|
| 1081 |
-
"learning_rate": 4.998668887935691e-05,
|
| 1082 |
-
"loss": 1.5462,
|
| 1083 |
-
"step": 1510
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 0.03676292748996275,
|
| 1087 |
-
"grad_norm": 0.5325940847396851,
|
| 1088 |
-
"learning_rate": 4.998616075998916e-05,
|
| 1089 |
-
"loss": 1.5189,
|
| 1090 |
-
"step": 1520
|
| 1091 |
-
},
|
| 1092 |
-
{
|
| 1093 |
-
"epoch": 0.037004788855028295,
|
| 1094 |
-
"grad_norm": 0.3709082305431366,
|
| 1095 |
-
"learning_rate": 4.9985622369754525e-05,
|
| 1096 |
-
"loss": 1.5219,
|
| 1097 |
-
"step": 1530
|
| 1098 |
-
},
|
| 1099 |
-
{
|
| 1100 |
-
"epoch": 0.037246650220093845,
|
| 1101 |
-
"grad_norm": 0.3708791732788086,
|
| 1102 |
-
"learning_rate": 4.998507370887433e-05,
|
| 1103 |
-
"loss": 1.5202,
|
| 1104 |
-
"step": 1540
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 0.03748851158515939,
|
| 1108 |
-
"grad_norm": 0.6688795685768127,
|
| 1109 |
-
"learning_rate": 4.9984514777574085e-05,
|
| 1110 |
-
"loss": 1.5273,
|
| 1111 |
-
"step": 1550
|
| 1112 |
-
},
|
| 1113 |
-
{
|
| 1114 |
-
"epoch": 0.03773037295022493,
|
| 1115 |
-
"grad_norm": 0.4270065128803253,
|
| 1116 |
-
"learning_rate": 4.998394557608358e-05,
|
| 1117 |
-
"loss": 1.5227,
|
| 1118 |
-
"step": 1560
|
| 1119 |
-
},
|
| 1120 |
-
{
|
| 1121 |
-
"epoch": 0.037972234315290475,
|
| 1122 |
-
"grad_norm": 1.07551109790802,
|
| 1123 |
-
"learning_rate": 4.998336610463677e-05,
|
| 1124 |
-
"loss": 1.5212,
|
| 1125 |
-
"step": 1570
|
| 1126 |
-
},
|
| 1127 |
-
{
|
| 1128 |
-
"epoch": 0.03821409568035602,
|
| 1129 |
-
"grad_norm": 0.5083074569702148,
|
| 1130 |
-
"learning_rate": 4.998277636347186e-05,
|
| 1131 |
-
"loss": 1.5259,
|
| 1132 |
-
"step": 1580
|
| 1133 |
-
},
|
| 1134 |
-
{
|
| 1135 |
-
"epoch": 0.03845595704542156,
|
| 1136 |
-
"grad_norm": 0.3792552351951599,
|
| 1137 |
-
"learning_rate": 4.998217635283127e-05,
|
| 1138 |
-
"loss": 1.5191,
|
| 1139 |
-
"step": 1590
|
| 1140 |
-
},
|
| 1141 |
-
{
|
| 1142 |
-
"epoch": 0.03869781841048711,
|
| 1143 |
-
"grad_norm": 0.46655386686325073,
|
| 1144 |
-
"learning_rate": 4.998156607296163e-05,
|
| 1145 |
-
"loss": 1.5218,
|
| 1146 |
-
"step": 1600
|
| 1147 |
-
},
|
| 1148 |
-
{
|
| 1149 |
-
"epoch": 0.038939679775552655,
|
| 1150 |
-
"grad_norm": 0.37935832142829895,
|
| 1151 |
-
"learning_rate": 4.998094552411382e-05,
|
| 1152 |
-
"loss": 1.5278,
|
| 1153 |
-
"step": 1610
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 0.0391815411406182,
|
| 1157 |
-
"grad_norm": 1.3031045198440552,
|
| 1158 |
-
"learning_rate": 4.9980314706542916e-05,
|
| 1159 |
-
"loss": 1.5278,
|
| 1160 |
-
"step": 1620
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 0.03942340250568374,
|
| 1164 |
-
"grad_norm": 0.38517051935195923,
|
| 1165 |
-
"learning_rate": 4.997967362050824e-05,
|
| 1166 |
-
"loss": 1.5157,
|
| 1167 |
-
"step": 1630
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 0.039665263870749284,
|
| 1171 |
-
"grad_norm": 0.3939484655857086,
|
| 1172 |
-
"learning_rate": 4.997902226627329e-05,
|
| 1173 |
-
"loss": 1.5242,
|
| 1174 |
-
"step": 1640
|
| 1175 |
-
},
|
| 1176 |
-
{
|
| 1177 |
-
"epoch": 0.03990712523581483,
|
| 1178 |
-
"grad_norm": 0.45832833647727966,
|
| 1179 |
-
"learning_rate": 4.997836064410583e-05,
|
| 1180 |
-
"loss": 1.522,
|
| 1181 |
-
"step": 1650
|
| 1182 |
-
},
|
| 1183 |
-
{
|
| 1184 |
-
"epoch": 0.04014898660088038,
|
| 1185 |
-
"grad_norm": 0.4101678729057312,
|
| 1186 |
-
"learning_rate": 4.997768875427782e-05,
|
| 1187 |
-
"loss": 1.5233,
|
| 1188 |
-
"step": 1660
|
| 1189 |
-
},
|
| 1190 |
-
{
|
| 1191 |
-
"epoch": 0.04039084796594592,
|
| 1192 |
-
"grad_norm": 0.4063393771648407,
|
| 1193 |
-
"learning_rate": 4.997700659706545e-05,
|
| 1194 |
-
"loss": 1.5182,
|
| 1195 |
-
"step": 1670
|
| 1196 |
-
},
|
| 1197 |
-
{
|
| 1198 |
-
"epoch": 0.040632709331011464,
|
| 1199 |
-
"grad_norm": 0.433902770280838,
|
| 1200 |
-
"learning_rate": 4.997631417274914e-05,
|
| 1201 |
-
"loss": 1.5267,
|
| 1202 |
-
"step": 1680
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 0.04087457069607701,
|
| 1206 |
-
"grad_norm": 0.43226370215415955,
|
| 1207 |
-
"learning_rate": 4.997561148161351e-05,
|
| 1208 |
-
"loss": 1.5242,
|
| 1209 |
-
"step": 1690
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 0.04111643206114255,
|
| 1213 |
-
"grad_norm": 0.5391157269477844,
|
| 1214 |
-
"learning_rate": 4.997489852394741e-05,
|
| 1215 |
-
"loss": 1.5183,
|
| 1216 |
-
"step": 1700
|
| 1217 |
-
},
|
| 1218 |
-
{
|
| 1219 |
-
"epoch": 0.0413582934262081,
|
| 1220 |
-
"grad_norm": 0.3768923878669739,
|
| 1221 |
-
"learning_rate": 4.997417530004391e-05,
|
| 1222 |
-
"loss": 1.5232,
|
| 1223 |
-
"step": 1710
|
| 1224 |
-
},
|
| 1225 |
-
{
|
| 1226 |
-
"epoch": 0.041600154791273644,
|
| 1227 |
-
"grad_norm": 0.5486100912094116,
|
| 1228 |
-
"learning_rate": 4.9973441810200306e-05,
|
| 1229 |
-
"loss": 1.5149,
|
| 1230 |
-
"step": 1720
|
| 1231 |
-
},
|
| 1232 |
-
{
|
| 1233 |
-
"epoch": 0.04184201615633919,
|
| 1234 |
-
"grad_norm": 0.4516407549381256,
|
| 1235 |
-
"learning_rate": 4.997269805471809e-05,
|
| 1236 |
-
"loss": 1.518,
|
| 1237 |
-
"step": 1730
|
| 1238 |
-
},
|
| 1239 |
-
{
|
| 1240 |
-
"epoch": 0.04208387752140473,
|
| 1241 |
-
"grad_norm": 0.36857402324676514,
|
| 1242 |
-
"learning_rate": 4.997194403390302e-05,
|
| 1243 |
-
"loss": 1.532,
|
| 1244 |
-
"step": 1740
|
| 1245 |
-
},
|
| 1246 |
-
{
|
| 1247 |
-
"epoch": 0.042325738886470274,
|
| 1248 |
-
"grad_norm": 0.3924587368965149,
|
| 1249 |
-
"learning_rate": 4.9971179748065024e-05,
|
| 1250 |
-
"loss": 1.5213,
|
| 1251 |
-
"step": 1750
|
| 1252 |
-
},
|
| 1253 |
-
{
|
| 1254 |
-
"epoch": 0.04256760025153582,
|
| 1255 |
-
"grad_norm": 0.635565996170044,
|
| 1256 |
-
"learning_rate": 4.997040519751828e-05,
|
| 1257 |
-
"loss": 1.5167,
|
| 1258 |
-
"step": 1760
|
| 1259 |
-
},
|
| 1260 |
-
{
|
| 1261 |
-
"epoch": 0.04280946161660137,
|
| 1262 |
-
"grad_norm": 0.6264420747756958,
|
| 1263 |
-
"learning_rate": 4.996962038258117e-05,
|
| 1264 |
-
"loss": 1.5209,
|
| 1265 |
-
"step": 1770
|
| 1266 |
-
},
|
| 1267 |
-
{
|
| 1268 |
-
"epoch": 0.04305132298166691,
|
| 1269 |
-
"grad_norm": 0.46929728984832764,
|
| 1270 |
-
"learning_rate": 4.9968825303576314e-05,
|
| 1271 |
-
"loss": 1.5236,
|
| 1272 |
-
"step": 1780
|
| 1273 |
-
},
|
| 1274 |
-
{
|
| 1275 |
-
"epoch": 0.043293184346732454,
|
| 1276 |
-
"grad_norm": 0.3539489209651947,
|
| 1277 |
-
"learning_rate": 4.996801996083052e-05,
|
| 1278 |
-
"loss": 1.5202,
|
| 1279 |
-
"step": 1790
|
| 1280 |
-
},
|
| 1281 |
-
{
|
| 1282 |
-
"epoch": 0.043535045711798,
|
| 1283 |
-
"grad_norm": 0.5325206518173218,
|
| 1284 |
-
"learning_rate": 4.996720435467485e-05,
|
| 1285 |
-
"loss": 1.5163,
|
| 1286 |
-
"step": 1800
|
| 1287 |
-
},
|
| 1288 |
-
{
|
| 1289 |
-
"epoch": 0.04377690707686354,
|
| 1290 |
-
"grad_norm": 0.37687698006629944,
|
| 1291 |
-
"learning_rate": 4.9966378485444567e-05,
|
| 1292 |
-
"loss": 1.5181,
|
| 1293 |
-
"step": 1810
|
| 1294 |
-
},
|
| 1295 |
-
{
|
| 1296 |
-
"epoch": 0.04401876844192908,
|
| 1297 |
-
"grad_norm": 0.39997562766075134,
|
| 1298 |
-
"learning_rate": 4.9965542353479144e-05,
|
| 1299 |
-
"loss": 1.518,
|
| 1300 |
-
"step": 1820
|
| 1301 |
-
},
|
| 1302 |
-
{
|
| 1303 |
-
"epoch": 0.044260629806994634,
|
| 1304 |
-
"grad_norm": 0.42250344157218933,
|
| 1305 |
-
"learning_rate": 4.9964695959122294e-05,
|
| 1306 |
-
"loss": 1.5123,
|
| 1307 |
-
"step": 1830
|
| 1308 |
-
},
|
| 1309 |
-
{
|
| 1310 |
-
"epoch": 0.04450249117206018,
|
| 1311 |
-
"grad_norm": 0.4876804053783417,
|
| 1312 |
-
"learning_rate": 4.9963839302721936e-05,
|
| 1313 |
-
"loss": 1.5224,
|
| 1314 |
-
"step": 1840
|
| 1315 |
-
},
|
| 1316 |
-
{
|
| 1317 |
-
"epoch": 0.04474435253712572,
|
| 1318 |
-
"grad_norm": 0.4264649748802185,
|
| 1319 |
-
"learning_rate": 4.99629723846302e-05,
|
| 1320 |
-
"loss": 1.5226,
|
| 1321 |
-
"step": 1850
|
| 1322 |
-
},
|
| 1323 |
-
{
|
| 1324 |
-
"epoch": 0.04498621390219126,
|
| 1325 |
-
"grad_norm": 0.3468644320964813,
|
| 1326 |
-
"learning_rate": 4.996209520520346e-05,
|
| 1327 |
-
"loss": 1.5091,
|
| 1328 |
-
"step": 1860
|
| 1329 |
-
},
|
| 1330 |
-
{
|
| 1331 |
-
"epoch": 0.045228075267256806,
|
| 1332 |
-
"grad_norm": 0.7444997429847717,
|
| 1333 |
-
"learning_rate": 4.9961207764802275e-05,
|
| 1334 |
-
"loss": 1.5164,
|
| 1335 |
-
"step": 1870
|
| 1336 |
-
},
|
| 1337 |
-
{
|
| 1338 |
-
"epoch": 0.04546993663232235,
|
| 1339 |
-
"grad_norm": 0.3771494925022125,
|
| 1340 |
-
"learning_rate": 4.996031006379145e-05,
|
| 1341 |
-
"loss": 1.5119,
|
| 1342 |
-
"step": 1880
|
| 1343 |
-
},
|
| 1344 |
-
{
|
| 1345 |
-
"epoch": 0.0457117979973879,
|
| 1346 |
-
"grad_norm": 0.34328994154930115,
|
| 1347 |
-
"learning_rate": 4.9959402102539986e-05,
|
| 1348 |
-
"loss": 1.5087,
|
| 1349 |
-
"step": 1890
|
| 1350 |
-
},
|
| 1351 |
-
{
|
| 1352 |
-
"epoch": 0.04595365936245344,
|
| 1353 |
-
"grad_norm": 0.4548104405403137,
|
| 1354 |
-
"learning_rate": 4.995848388142112e-05,
|
| 1355 |
-
"loss": 1.5268,
|
| 1356 |
-
"step": 1900
|
| 1357 |
-
},
|
| 1358 |
-
{
|
| 1359 |
-
"epoch": 0.046195520727518986,
|
| 1360 |
-
"grad_norm": 0.5128782987594604,
|
| 1361 |
-
"learning_rate": 4.995755540081229e-05,
|
| 1362 |
-
"loss": 1.5127,
|
| 1363 |
-
"step": 1910
|
| 1364 |
-
},
|
| 1365 |
-
{
|
| 1366 |
-
"epoch": 0.04643738209258453,
|
| 1367 |
-
"grad_norm": 0.5447189807891846,
|
| 1368 |
-
"learning_rate": 4.995661666109518e-05,
|
| 1369 |
-
"loss": 1.5148,
|
| 1370 |
-
"step": 1920
|
| 1371 |
-
},
|
| 1372 |
-
{
|
| 1373 |
-
"epoch": 0.04667924345765007,
|
| 1374 |
-
"grad_norm": 0.42812883853912354,
|
| 1375 |
-
"learning_rate": 4.9955667662655636e-05,
|
| 1376 |
-
"loss": 1.515,
|
| 1377 |
-
"step": 1930
|
| 1378 |
-
},
|
| 1379 |
-
{
|
| 1380 |
-
"epoch": 0.046921104822715616,
|
| 1381 |
-
"grad_norm": 0.46662941575050354,
|
| 1382 |
-
"learning_rate": 4.995470840588379e-05,
|
| 1383 |
-
"loss": 1.5354,
|
| 1384 |
-
"step": 1940
|
| 1385 |
-
},
|
| 1386 |
-
{
|
| 1387 |
-
"epoch": 0.047162966187781166,
|
| 1388 |
-
"grad_norm": 0.38454869389533997,
|
| 1389 |
-
"learning_rate": 4.995373889117393e-05,
|
| 1390 |
-
"loss": 1.5254,
|
| 1391 |
-
"step": 1950
|
| 1392 |
-
},
|
| 1393 |
-
{
|
| 1394 |
-
"epoch": 0.04740482755284671,
|
| 1395 |
-
"grad_norm": 0.4641868472099304,
|
| 1396 |
-
"learning_rate": 4.99527591189246e-05,
|
| 1397 |
-
"loss": 1.5115,
|
| 1398 |
-
"step": 1960
|
| 1399 |
-
},
|
| 1400 |
-
{
|
| 1401 |
-
"epoch": 0.04764668891791225,
|
| 1402 |
-
"grad_norm": 0.4224710464477539,
|
| 1403 |
-
"learning_rate": 4.995176908953854e-05,
|
| 1404 |
-
"loss": 1.5091,
|
| 1405 |
-
"step": 1970
|
| 1406 |
-
},
|
| 1407 |
-
{
|
| 1408 |
-
"epoch": 0.047888550282977796,
|
| 1409 |
-
"grad_norm": 0.5580614805221558,
|
| 1410 |
-
"learning_rate": 4.995076880342271e-05,
|
| 1411 |
-
"loss": 1.507,
|
| 1412 |
-
"step": 1980
|
| 1413 |
-
},
|
| 1414 |
-
{
|
| 1415 |
-
"epoch": 0.04813041164804334,
|
| 1416 |
-
"grad_norm": 0.42475947737693787,
|
| 1417 |
-
"learning_rate": 4.994975826098831e-05,
|
| 1418 |
-
"loss": 1.5142,
|
| 1419 |
-
"step": 1990
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 0.04837227301310889,
|
| 1423 |
-
"grad_norm": 0.48980775475502014,
|
| 1424 |
-
"learning_rate": 4.994873746265073e-05,
|
| 1425 |
-
"loss": 1.52,
|
| 1426 |
-
"step": 2000
|
| 1427 |
-
},
|
| 1428 |
-
{
|
| 1429 |
-
"epoch": 0.04837227301310889,
|
| 1430 |
-
"eval_loss": 1.509605050086975,
|
| 1431 |
-
"eval_runtime": 1199.6508,
|
| 1432 |
-
"eval_sacrebleu": 96.64726244926639,
|
| 1433 |
-
"eval_samples_per_second": 84.23,
|
| 1434 |
-
"eval_steps_per_second": 0.659,
|
| 1435 |
-
"step": 2000
|
| 1436 |
-
},
|
| 1437 |
-
{
|
| 1438 |
-
"epoch": 0.04861413437817443,
|
| 1439 |
-
"grad_norm": 0.3497089147567749,
|
| 1440 |
-
"learning_rate": 4.9947706408829565e-05,
|
| 1441 |
-
"loss": 1.5106,
|
| 1442 |
-
"step": 2010
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 0.048855995743239976,
|
| 1446 |
-
"grad_norm": 0.37651902437210083,
|
| 1447 |
-
"learning_rate": 4.994666509994865e-05,
|
| 1448 |
-
"loss": 1.5115,
|
| 1449 |
-
"step": 2020
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 0.04909785710830552,
|
| 1453 |
-
"grad_norm": 0.5997866988182068,
|
| 1454 |
-
"learning_rate": 4.994561353643604e-05,
|
| 1455 |
-
"loss": 1.5086,
|
| 1456 |
-
"step": 2030
|
| 1457 |
-
},
|
| 1458 |
-
{
|
| 1459 |
-
"epoch": 0.04933971847337106,
|
| 1460 |
-
"grad_norm": 0.4603705108165741,
|
| 1461 |
-
"learning_rate": 4.994455171872398e-05,
|
| 1462 |
-
"loss": 1.529,
|
| 1463 |
-
"step": 2040
|
| 1464 |
-
},
|
| 1465 |
-
{
|
| 1466 |
-
"epoch": 0.049581579838436605,
|
| 1467 |
-
"grad_norm": 0.35874149203300476,
|
| 1468 |
-
"learning_rate": 4.994347964724894e-05,
|
| 1469 |
-
"loss": 1.5076,
|
| 1470 |
-
"step": 2050
|
| 1471 |
-
},
|
| 1472 |
-
{
|
| 1473 |
-
"epoch": 0.049823441203502156,
|
| 1474 |
-
"grad_norm": 0.9553395509719849,
|
| 1475 |
-
"learning_rate": 4.994239732245162e-05,
|
| 1476 |
-
"loss": 1.5138,
|
| 1477 |
-
"step": 2060
|
| 1478 |
-
},
|
| 1479 |
-
{
|
| 1480 |
-
"epoch": 0.0500653025685677,
|
| 1481 |
-
"grad_norm": 0.42904314398765564,
|
| 1482 |
-
"learning_rate": 4.994130474477692e-05,
|
| 1483 |
-
"loss": 1.5135,
|
| 1484 |
-
"step": 2070
|
| 1485 |
-
},
|
| 1486 |
-
{
|
| 1487 |
-
"epoch": 0.05030716393363324,
|
| 1488 |
-
"grad_norm": 0.44861099123954773,
|
| 1489 |
-
"learning_rate": 4.994020191467395e-05,
|
| 1490 |
-
"loss": 1.5137,
|
| 1491 |
-
"step": 2080
|
| 1492 |
-
},
|
| 1493 |
-
{
|
| 1494 |
-
"epoch": 0.050549025298698785,
|
| 1495 |
-
"grad_norm": 0.46189823746681213,
|
| 1496 |
-
"learning_rate": 4.993908883259605e-05,
|
| 1497 |
-
"loss": 1.5088,
|
| 1498 |
-
"step": 2090
|
| 1499 |
-
},
|
| 1500 |
-
{
|
| 1501 |
-
"epoch": 0.05079088666376433,
|
| 1502 |
-
"grad_norm": 0.31554293632507324,
|
| 1503 |
-
"learning_rate": 4.993796549900076e-05,
|
| 1504 |
-
"loss": 1.5161,
|
| 1505 |
-
"step": 2100
|
| 1506 |
-
},
|
| 1507 |
-
{
|
| 1508 |
-
"epoch": 0.05103274802882987,
|
| 1509 |
-
"grad_norm": 0.48815393447875977,
|
| 1510 |
-
"learning_rate": 4.9936831914349845e-05,
|
| 1511 |
-
"loss": 1.5148,
|
| 1512 |
-
"step": 2110
|
| 1513 |
-
},
|
| 1514 |
-
{
|
| 1515 |
-
"epoch": 0.05127460939389542,
|
| 1516 |
-
"grad_norm": 0.5335953235626221,
|
| 1517 |
-
"learning_rate": 4.9935688079109277e-05,
|
| 1518 |
-
"loss": 1.5144,
|
| 1519 |
-
"step": 2120
|
| 1520 |
-
},
|
| 1521 |
-
{
|
| 1522 |
-
"epoch": 0.051516470758960965,
|
| 1523 |
-
"grad_norm": 0.4184572398662567,
|
| 1524 |
-
"learning_rate": 4.9934533993749236e-05,
|
| 1525 |
-
"loss": 1.5134,
|
| 1526 |
-
"step": 2130
|
| 1527 |
-
},
|
| 1528 |
-
{
|
| 1529 |
-
"epoch": 0.05175833212402651,
|
| 1530 |
-
"grad_norm": 0.3346949517726898,
|
| 1531 |
-
"learning_rate": 4.993336965874414e-05,
|
| 1532 |
-
"loss": 1.5042,
|
| 1533 |
-
"step": 2140
|
| 1534 |
-
},
|
| 1535 |
-
{
|
| 1536 |
-
"epoch": 0.05200019348909205,
|
| 1537 |
-
"grad_norm": 0.34112387895584106,
|
| 1538 |
-
"learning_rate": 4.993219507457259e-05,
|
| 1539 |
-
"loss": 1.5109,
|
| 1540 |
-
"step": 2150
|
| 1541 |
-
},
|
| 1542 |
-
{
|
| 1543 |
-
"epoch": 0.052242054854157595,
|
| 1544 |
-
"grad_norm": 0.42028605937957764,
|
| 1545 |
-
"learning_rate": 4.9931010241717414e-05,
|
| 1546 |
-
"loss": 1.5159,
|
| 1547 |
-
"step": 2160
|
| 1548 |
-
},
|
| 1549 |
-
{
|
| 1550 |
-
"epoch": 0.05248391621922314,
|
| 1551 |
-
"grad_norm": 0.4802249073982239,
|
| 1552 |
-
"learning_rate": 4.992981516066565e-05,
|
| 1553 |
-
"loss": 1.5063,
|
| 1554 |
-
"step": 2170
|
| 1555 |
-
},
|
| 1556 |
-
{
|
| 1557 |
-
"epoch": 0.05272577758428869,
|
| 1558 |
-
"grad_norm": 0.3522288501262665,
|
| 1559 |
-
"learning_rate": 4.9928609831908564e-05,
|
| 1560 |
-
"loss": 1.5067,
|
| 1561 |
-
"step": 2180
|
| 1562 |
-
},
|
| 1563 |
-
{
|
| 1564 |
-
"epoch": 0.05296763894935423,
|
| 1565 |
-
"grad_norm": 0.47284331917762756,
|
| 1566 |
-
"learning_rate": 4.9927394255941606e-05,
|
| 1567 |
-
"loss": 1.5035,
|
| 1568 |
-
"step": 2190
|
| 1569 |
-
},
|
| 1570 |
-
{
|
| 1571 |
-
"epoch": 0.053209500314419775,
|
| 1572 |
-
"grad_norm": 0.4390985071659088,
|
| 1573 |
-
"learning_rate": 4.992616843326446e-05,
|
| 1574 |
-
"loss": 1.5081,
|
| 1575 |
-
"step": 2200
|
| 1576 |
-
},
|
| 1577 |
-
{
|
| 1578 |
-
"epoch": 0.05345136167948532,
|
| 1579 |
-
"grad_norm": 0.37876608967781067,
|
| 1580 |
-
"learning_rate": 4.992493236438102e-05,
|
| 1581 |
-
"loss": 1.5084,
|
| 1582 |
-
"step": 2210
|
| 1583 |
-
},
|
| 1584 |
-
{
|
| 1585 |
-
"epoch": 0.05369322304455086,
|
| 1586 |
-
"grad_norm": 0.45138484239578247,
|
| 1587 |
-
"learning_rate": 4.992368604979938e-05,
|
| 1588 |
-
"loss": 1.5125,
|
| 1589 |
-
"step": 2220
|
| 1590 |
-
},
|
| 1591 |
-
{
|
| 1592 |
-
"epoch": 0.053935084409616404,
|
| 1593 |
-
"grad_norm": 0.40889862179756165,
|
| 1594 |
-
"learning_rate": 4.9922429490031844e-05,
|
| 1595 |
-
"loss": 1.5211,
|
| 1596 |
-
"step": 2230
|
| 1597 |
-
},
|
| 1598 |
-
{
|
| 1599 |
-
"epoch": 0.054176945774681955,
|
| 1600 |
-
"grad_norm": 0.43350115418434143,
|
| 1601 |
-
"learning_rate": 4.992116268559496e-05,
|
| 1602 |
-
"loss": 1.5106,
|
| 1603 |
-
"step": 2240
|
| 1604 |
-
},
|
| 1605 |
-
{
|
| 1606 |
-
"epoch": 0.0544188071397475,
|
| 1607 |
-
"grad_norm": 0.404316246509552,
|
| 1608 |
-
"learning_rate": 4.991988563700944e-05,
|
| 1609 |
-
"loss": 1.5125,
|
| 1610 |
-
"step": 2250
|
| 1611 |
-
},
|
| 1612 |
-
{
|
| 1613 |
-
"epoch": 0.05466066850481304,
|
| 1614 |
-
"grad_norm": 0.26847347617149353,
|
| 1615 |
-
"learning_rate": 4.991859834480025e-05,
|
| 1616 |
-
"loss": 1.5067,
|
| 1617 |
-
"step": 2260
|
| 1618 |
-
},
|
| 1619 |
-
{
|
| 1620 |
-
"epoch": 0.054902529869878584,
|
| 1621 |
-
"grad_norm": 0.36783504486083984,
|
| 1622 |
-
"learning_rate": 4.991730080949654e-05,
|
| 1623 |
-
"loss": 1.5137,
|
| 1624 |
-
"step": 2270
|
| 1625 |
-
},
|
| 1626 |
-
{
|
| 1627 |
-
"epoch": 0.05514439123494413,
|
| 1628 |
-
"grad_norm": 0.3214723765850067,
|
| 1629 |
-
"learning_rate": 4.991599303163167e-05,
|
| 1630 |
-
"loss": 1.5136,
|
| 1631 |
-
"step": 2280
|
| 1632 |
-
},
|
| 1633 |
-
{
|
| 1634 |
-
"epoch": 0.05538625260000968,
|
| 1635 |
-
"grad_norm": 0.6154227256774902,
|
| 1636 |
-
"learning_rate": 4.991467501174323e-05,
|
| 1637 |
-
"loss": 1.51,
|
| 1638 |
-
"step": 2290
|
| 1639 |
-
},
|
| 1640 |
-
{
|
| 1641 |
-
"epoch": 0.05562811396507522,
|
| 1642 |
-
"grad_norm": 0.388557493686676,
|
| 1643 |
-
"learning_rate": 4.991334675037299e-05,
|
| 1644 |
-
"loss": 1.4995,
|
| 1645 |
-
"step": 2300
|
| 1646 |
-
},
|
| 1647 |
-
{
|
| 1648 |
-
"epoch": 0.055869975330140764,
|
| 1649 |
-
"grad_norm": 0.5019369125366211,
|
| 1650 |
-
"learning_rate": 4.991200824806697e-05,
|
| 1651 |
-
"loss": 1.5063,
|
| 1652 |
-
"step": 2310
|
| 1653 |
-
},
|
| 1654 |
-
{
|
| 1655 |
-
"epoch": 0.05611183669520631,
|
| 1656 |
-
"grad_norm": 0.5222612619400024,
|
| 1657 |
-
"learning_rate": 4.991065950537537e-05,
|
| 1658 |
-
"loss": 1.5134,
|
| 1659 |
-
"step": 2320
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 0.05635369806027185,
|
| 1663 |
-
"grad_norm": 0.43677157163619995,
|
| 1664 |
-
"learning_rate": 4.9909300522852595e-05,
|
| 1665 |
-
"loss": 1.5115,
|
| 1666 |
-
"step": 2330
|
| 1667 |
-
},
|
| 1668 |
-
{
|
| 1669 |
-
"epoch": 0.056595559425337394,
|
| 1670 |
-
"grad_norm": 0.4183470606803894,
|
| 1671 |
-
"learning_rate": 4.990793130105729e-05,
|
| 1672 |
-
"loss": 1.5003,
|
| 1673 |
-
"step": 2340
|
| 1674 |
-
},
|
| 1675 |
-
{
|
| 1676 |
-
"epoch": 0.056837420790402944,
|
| 1677 |
-
"grad_norm": 0.9535141587257385,
|
| 1678 |
-
"learning_rate": 4.9906551840552285e-05,
|
| 1679 |
-
"loss": 1.5132,
|
| 1680 |
-
"step": 2350
|
| 1681 |
-
},
|
| 1682 |
-
{
|
| 1683 |
-
"epoch": 0.05707928215546849,
|
| 1684 |
-
"grad_norm": 0.35690176486968994,
|
| 1685 |
-
"learning_rate": 4.9905162141904616e-05,
|
| 1686 |
-
"loss": 1.51,
|
| 1687 |
-
"step": 2360
|
| 1688 |
-
},
|
| 1689 |
-
{
|
| 1690 |
-
"epoch": 0.05732114352053403,
|
| 1691 |
-
"grad_norm": 0.3144702911376953,
|
| 1692 |
-
"learning_rate": 4.990376220568554e-05,
|
| 1693 |
-
"loss": 1.501,
|
| 1694 |
-
"step": 2370
|
| 1695 |
-
},
|
| 1696 |
-
{
|
| 1697 |
-
"epoch": 0.057563004885599574,
|
| 1698 |
-
"grad_norm": 0.3751768469810486,
|
| 1699 |
-
"learning_rate": 4.990235203247052e-05,
|
| 1700 |
-
"loss": 1.5177,
|
| 1701 |
-
"step": 2380
|
| 1702 |
-
},
|
| 1703 |
-
{
|
| 1704 |
-
"epoch": 0.05780486625066512,
|
| 1705 |
-
"grad_norm": 0.46490389108657837,
|
| 1706 |
-
"learning_rate": 4.9900931622839216e-05,
|
| 1707 |
-
"loss": 1.5075,
|
| 1708 |
-
"step": 2390
|
| 1709 |
-
},
|
| 1710 |
-
{
|
| 1711 |
-
"epoch": 0.05804672761573066,
|
| 1712 |
-
"grad_norm": 0.42116859555244446,
|
| 1713 |
-
"learning_rate": 4.989950097737552e-05,
|
| 1714 |
-
"loss": 1.5057,
|
| 1715 |
-
"step": 2400
|
| 1716 |
-
},
|
| 1717 |
-
{
|
| 1718 |
-
"epoch": 0.05828858898079621,
|
| 1719 |
-
"grad_norm": 0.4353298544883728,
|
| 1720 |
-
"learning_rate": 4.9898060096667506e-05,
|
| 1721 |
-
"loss": 1.5026,
|
| 1722 |
-
"step": 2410
|
| 1723 |
-
},
|
| 1724 |
-
{
|
| 1725 |
-
"epoch": 0.058530450345861754,
|
| 1726 |
-
"grad_norm": 0.3899039328098297,
|
| 1727 |
-
"learning_rate": 4.989660898130748e-05,
|
| 1728 |
-
"loss": 1.509,
|
| 1729 |
-
"step": 2420
|
| 1730 |
-
},
|
| 1731 |
-
{
|
| 1732 |
-
"epoch": 0.0587723117109273,
|
| 1733 |
-
"grad_norm": 0.4386129677295685,
|
| 1734 |
-
"learning_rate": 4.9895147631891916e-05,
|
| 1735 |
-
"loss": 1.5059,
|
| 1736 |
-
"step": 2430
|
| 1737 |
-
},
|
| 1738 |
-
{
|
| 1739 |
-
"epoch": 0.05901417307599284,
|
| 1740 |
-
"grad_norm": 0.3924826681613922,
|
| 1741 |
-
"learning_rate": 4.989367604902153e-05,
|
| 1742 |
-
"loss": 1.508,
|
| 1743 |
-
"step": 2440
|
| 1744 |
-
},
|
| 1745 |
-
{
|
| 1746 |
-
"epoch": 0.05925603444105838,
|
| 1747 |
-
"grad_norm": 0.36131057143211365,
|
| 1748 |
-
"learning_rate": 4.9892194233301246e-05,
|
| 1749 |
-
"loss": 1.5045,
|
| 1750 |
-
"step": 2450
|
| 1751 |
-
},
|
| 1752 |
-
{
|
| 1753 |
-
"epoch": 0.05949789580612393,
|
| 1754 |
-
"grad_norm": 0.29293808341026306,
|
| 1755 |
-
"learning_rate": 4.989070218534017e-05,
|
| 1756 |
-
"loss": 1.5065,
|
| 1757 |
-
"step": 2460
|
| 1758 |
-
},
|
| 1759 |
-
{
|
| 1760 |
-
"epoch": 0.05973975717118948,
|
| 1761 |
-
"grad_norm": 0.6674768328666687,
|
| 1762 |
-
"learning_rate": 4.988919990575163e-05,
|
| 1763 |
-
"loss": 1.5142,
|
| 1764 |
-
"step": 2470
|
| 1765 |
-
},
|
| 1766 |
-
{
|
| 1767 |
-
"epoch": 0.05998161853625502,
|
| 1768 |
-
"grad_norm": 0.3673650920391083,
|
| 1769 |
-
"learning_rate": 4.988768739515315e-05,
|
| 1770 |
-
"loss": 1.5105,
|
| 1771 |
-
"step": 2480
|
| 1772 |
-
},
|
| 1773 |
-
{
|
| 1774 |
-
"epoch": 0.06022347990132056,
|
| 1775 |
-
"grad_norm": 0.4137951135635376,
|
| 1776 |
-
"learning_rate": 4.988616465416648e-05,
|
| 1777 |
-
"loss": 1.5067,
|
| 1778 |
-
"step": 2490
|
| 1779 |
-
},
|
| 1780 |
-
{
|
| 1781 |
-
"epoch": 0.060465341266386106,
|
| 1782 |
-
"grad_norm": 0.38709238171577454,
|
| 1783 |
-
"learning_rate": 4.988463168341755e-05,
|
| 1784 |
-
"loss": 1.5043,
|
| 1785 |
-
"step": 2500
|
| 1786 |
-
},
|
| 1787 |
-
{
|
| 1788 |
-
"epoch": 0.06070720263145165,
|
| 1789 |
-
"grad_norm": 0.4081590175628662,
|
| 1790 |
-
"learning_rate": 4.9883088483536514e-05,
|
| 1791 |
-
"loss": 1.5047,
|
| 1792 |
-
"step": 2510
|
| 1793 |
-
},
|
| 1794 |
-
{
|
| 1795 |
-
"epoch": 0.06094906399651719,
|
| 1796 |
-
"grad_norm": 0.5782541036605835,
|
| 1797 |
-
"learning_rate": 4.988153505515771e-05,
|
| 1798 |
-
"loss": 1.5004,
|
| 1799 |
-
"step": 2520
|
| 1800 |
-
},
|
| 1801 |
-
{
|
| 1802 |
-
"epoch": 0.06119092536158274,
|
| 1803 |
-
"grad_norm": 0.5798467993736267,
|
| 1804 |
-
"learning_rate": 4.987997139891971e-05,
|
| 1805 |
-
"loss": 1.5108,
|
| 1806 |
-
"step": 2530
|
| 1807 |
-
},
|
| 1808 |
-
{
|
| 1809 |
-
"epoch": 0.061432786726648286,
|
| 1810 |
-
"grad_norm": 0.4626334607601166,
|
| 1811 |
-
"learning_rate": 4.987839751546526e-05,
|
| 1812 |
-
"loss": 1.5064,
|
| 1813 |
-
"step": 2540
|
| 1814 |
-
},
|
| 1815 |
-
{
|
| 1816 |
-
"epoch": 0.06167464809171383,
|
| 1817 |
-
"grad_norm": 0.38564711809158325,
|
| 1818 |
-
"learning_rate": 4.987681340544134e-05,
|
| 1819 |
-
"loss": 1.5139,
|
| 1820 |
-
"step": 2550
|
| 1821 |
-
},
|
| 1822 |
-
{
|
| 1823 |
-
"epoch": 0.06191650945677937,
|
| 1824 |
-
"grad_norm": 0.3602082431316376,
|
| 1825 |
-
"learning_rate": 4.98752190694991e-05,
|
| 1826 |
-
"loss": 1.5083,
|
| 1827 |
-
"step": 2560
|
| 1828 |
-
},
|
| 1829 |
-
{
|
| 1830 |
-
"epoch": 0.062158370821844916,
|
| 1831 |
-
"grad_norm": 0.41802382469177246,
|
| 1832 |
-
"learning_rate": 4.987361450829392e-05,
|
| 1833 |
-
"loss": 1.5173,
|
| 1834 |
-
"step": 2570
|
| 1835 |
-
},
|
| 1836 |
-
{
|
| 1837 |
-
"epoch": 0.062400232186910466,
|
| 1838 |
-
"grad_norm": 0.46094614267349243,
|
| 1839 |
-
"learning_rate": 4.9871999722485375e-05,
|
| 1840 |
-
"loss": 1.504,
|
| 1841 |
-
"step": 2580
|
| 1842 |
-
},
|
| 1843 |
-
{
|
| 1844 |
-
"epoch": 0.062642093551976,
|
| 1845 |
-
"grad_norm": 0.3688693046569824,
|
| 1846 |
-
"learning_rate": 4.987037471273725e-05,
|
| 1847 |
-
"loss": 1.5077,
|
| 1848 |
-
"step": 2590
|
| 1849 |
-
},
|
| 1850 |
-
{
|
| 1851 |
-
"epoch": 0.06288395491704155,
|
| 1852 |
-
"grad_norm": 0.42533257603645325,
|
| 1853 |
-
"learning_rate": 4.986873947971751e-05,
|
| 1854 |
-
"loss": 1.492,
|
| 1855 |
-
"step": 2600
|
| 1856 |
-
},
|
| 1857 |
-
{
|
| 1858 |
-
"epoch": 0.0631258162821071,
|
| 1859 |
-
"grad_norm": 0.43037179112434387,
|
| 1860 |
-
"learning_rate": 4.986709402409834e-05,
|
| 1861 |
-
"loss": 1.5045,
|
| 1862 |
-
"step": 2610
|
| 1863 |
-
},
|
| 1864 |
-
{
|
| 1865 |
-
"epoch": 0.06336767764717265,
|
| 1866 |
-
"grad_norm": 0.4197692573070526,
|
| 1867 |
-
"learning_rate": 4.986543834655614e-05,
|
| 1868 |
-
"loss": 1.5038,
|
| 1869 |
-
"step": 2620
|
| 1870 |
-
},
|
| 1871 |
-
{
|
| 1872 |
-
"epoch": 0.06360953901223819,
|
| 1873 |
-
"grad_norm": 0.38508251309394836,
|
| 1874 |
-
"learning_rate": 4.986377244777147e-05,
|
| 1875 |
-
"loss": 1.5128,
|
| 1876 |
-
"step": 2630
|
| 1877 |
-
},
|
| 1878 |
-
{
|
| 1879 |
-
"epoch": 0.06385140037730373,
|
| 1880 |
-
"grad_norm": 0.44845107197761536,
|
| 1881 |
-
"learning_rate": 4.9862096328429145e-05,
|
| 1882 |
-
"loss": 1.5071,
|
| 1883 |
-
"step": 2640
|
| 1884 |
-
},
|
| 1885 |
-
{
|
| 1886 |
-
"epoch": 0.06409326174236928,
|
| 1887 |
-
"grad_norm": 0.41548463702201843,
|
| 1888 |
-
"learning_rate": 4.986040998921814e-05,
|
| 1889 |
-
"loss": 1.5008,
|
| 1890 |
-
"step": 2650
|
| 1891 |
-
},
|
| 1892 |
-
{
|
| 1893 |
-
"epoch": 0.06433512310743482,
|
| 1894 |
-
"grad_norm": 0.400920569896698,
|
| 1895 |
-
"learning_rate": 4.9858713430831646e-05,
|
| 1896 |
-
"loss": 1.5008,
|
| 1897 |
-
"step": 2660
|
| 1898 |
-
},
|
| 1899 |
-
{
|
| 1900 |
-
"epoch": 0.06457698447250036,
|
| 1901 |
-
"grad_norm": 0.339394211769104,
|
| 1902 |
-
"learning_rate": 4.9857006653967064e-05,
|
| 1903 |
-
"loss": 1.5015,
|
| 1904 |
-
"step": 2670
|
| 1905 |
-
},
|
| 1906 |
-
{
|
| 1907 |
-
"epoch": 0.0648188458375659,
|
| 1908 |
-
"grad_norm": 0.4258403778076172,
|
| 1909 |
-
"learning_rate": 4.9855289659325974e-05,
|
| 1910 |
-
"loss": 1.5036,
|
| 1911 |
-
"step": 2680
|
| 1912 |
-
},
|
| 1913 |
-
{
|
| 1914 |
-
"epoch": 0.06506070720263145,
|
| 1915 |
-
"grad_norm": 0.34458810091018677,
|
| 1916 |
-
"learning_rate": 4.9853562447614164e-05,
|
| 1917 |
-
"loss": 1.5041,
|
| 1918 |
-
"step": 2690
|
| 1919 |
-
},
|
| 1920 |
-
{
|
| 1921 |
-
"epoch": 0.06530256856769699,
|
| 1922 |
-
"grad_norm": 0.3317313492298126,
|
| 1923 |
-
"learning_rate": 4.9851825019541644e-05,
|
| 1924 |
-
"loss": 1.4998,
|
| 1925 |
-
"step": 2700
|
| 1926 |
-
},
|
| 1927 |
-
{
|
| 1928 |
-
"epoch": 0.06554442993276254,
|
| 1929 |
-
"grad_norm": 0.3822150230407715,
|
| 1930 |
-
"learning_rate": 4.985007737582259e-05,
|
| 1931 |
-
"loss": 1.5205,
|
| 1932 |
-
"step": 2710
|
| 1933 |
-
},
|
| 1934 |
-
{
|
| 1935 |
-
"epoch": 0.06578629129782808,
|
| 1936 |
-
"grad_norm": 0.4381140470504761,
|
| 1937 |
-
"learning_rate": 4.984831951717539e-05,
|
| 1938 |
-
"loss": 1.5063,
|
| 1939 |
-
"step": 2720
|
| 1940 |
-
},
|
| 1941 |
-
{
|
| 1942 |
-
"epoch": 0.06602815266289364,
|
| 1943 |
-
"grad_norm": 0.36855047941207886,
|
| 1944 |
-
"learning_rate": 4.984655144432264e-05,
|
| 1945 |
-
"loss": 1.5115,
|
| 1946 |
-
"step": 2730
|
| 1947 |
-
},
|
| 1948 |
-
{
|
| 1949 |
-
"epoch": 0.06627001402795918,
|
| 1950 |
-
"grad_norm": 0.6179637908935547,
|
| 1951 |
-
"learning_rate": 4.984477315799114e-05,
|
| 1952 |
-
"loss": 1.5141,
|
| 1953 |
-
"step": 2740
|
| 1954 |
-
},
|
| 1955 |
-
{
|
| 1956 |
-
"epoch": 0.06651187539302472,
|
| 1957 |
-
"grad_norm": 0.46340298652648926,
|
| 1958 |
-
"learning_rate": 4.984298465891185e-05,
|
| 1959 |
-
"loss": 1.5102,
|
| 1960 |
-
"step": 2750
|
| 1961 |
-
},
|
| 1962 |
-
{
|
| 1963 |
-
"epoch": 0.06675373675809027,
|
| 1964 |
-
"grad_norm": 0.3800617456436157,
|
| 1965 |
-
"learning_rate": 4.984118594781998e-05,
|
| 1966 |
-
"loss": 1.4986,
|
| 1967 |
-
"step": 2760
|
| 1968 |
-
},
|
| 1969 |
-
{
|
| 1970 |
-
"epoch": 0.06699559812315581,
|
| 1971 |
-
"grad_norm": 0.38096559047698975,
|
| 1972 |
-
"learning_rate": 4.983937702545489e-05,
|
| 1973 |
-
"loss": 1.5079,
|
| 1974 |
-
"step": 2770
|
| 1975 |
-
},
|
| 1976 |
-
{
|
| 1977 |
-
"epoch": 0.06723745948822135,
|
| 1978 |
-
"grad_norm": 0.39892709255218506,
|
| 1979 |
-
"learning_rate": 4.983755789256018e-05,
|
| 1980 |
-
"loss": 1.5005,
|
| 1981 |
-
"step": 2780
|
| 1982 |
-
},
|
| 1983 |
-
{
|
| 1984 |
-
"epoch": 0.0674793208532869,
|
| 1985 |
-
"grad_norm": 0.45704516768455505,
|
| 1986 |
-
"learning_rate": 4.983572854988361e-05,
|
| 1987 |
-
"loss": 1.5014,
|
| 1988 |
-
"step": 2790
|
| 1989 |
-
},
|
| 1990 |
-
{
|
| 1991 |
-
"epoch": 0.06772118221835244,
|
| 1992 |
-
"grad_norm": 0.3516509234905243,
|
| 1993 |
-
"learning_rate": 4.9833888998177165e-05,
|
| 1994 |
-
"loss": 1.4981,
|
| 1995 |
-
"step": 2800
|
| 1996 |
-
},
|
| 1997 |
-
{
|
| 1998 |
-
"epoch": 0.06796304358341798,
|
| 1999 |
-
"grad_norm": 0.3279588520526886,
|
| 2000 |
-
"learning_rate": 4.9832039238197e-05,
|
| 2001 |
-
"loss": 1.5002,
|
| 2002 |
-
"step": 2810
|
| 2003 |
-
},
|
| 2004 |
-
{
|
| 2005 |
-
"epoch": 0.06820490494848352,
|
| 2006 |
-
"grad_norm": 0.37436121702194214,
|
| 2007 |
-
"learning_rate": 4.983017927070351e-05,
|
| 2008 |
-
"loss": 1.5019,
|
| 2009 |
-
"step": 2820
|
| 2010 |
-
},
|
| 2011 |
-
{
|
| 2012 |
-
"epoch": 0.06844676631354907,
|
| 2013 |
-
"grad_norm": 0.3239165246486664,
|
| 2014 |
-
"learning_rate": 4.982830909646123e-05,
|
| 2015 |
-
"loss": 1.4984,
|
| 2016 |
-
"step": 2830
|
| 2017 |
-
},
|
| 2018 |
-
{
|
| 2019 |
-
"epoch": 0.06868862767861462,
|
| 2020 |
-
"grad_norm": 0.3340087831020355,
|
| 2021 |
-
"learning_rate": 4.982642871623893e-05,
|
| 2022 |
-
"loss": 1.4952,
|
| 2023 |
-
"step": 2840
|
| 2024 |
-
},
|
| 2025 |
-
{
|
| 2026 |
-
"epoch": 0.06893048904368017,
|
| 2027 |
-
"grad_norm": 0.452729731798172,
|
| 2028 |
-
"learning_rate": 4.9824538130809556e-05,
|
| 2029 |
-
"loss": 1.505,
|
| 2030 |
-
"step": 2850
|
| 2031 |
-
},
|
| 2032 |
-
{
|
| 2033 |
-
"epoch": 0.06917235040874571,
|
| 2034 |
-
"grad_norm": 0.3184410035610199,
|
| 2035 |
-
"learning_rate": 4.9822637340950264e-05,
|
| 2036 |
-
"loss": 1.5044,
|
| 2037 |
-
"step": 2860
|
| 2038 |
-
},
|
| 2039 |
-
{
|
| 2040 |
-
"epoch": 0.06941421177381125,
|
| 2041 |
-
"grad_norm": 0.3954326808452606,
|
| 2042 |
-
"learning_rate": 4.982072634744239e-05,
|
| 2043 |
-
"loss": 1.5104,
|
| 2044 |
-
"step": 2870
|
| 2045 |
-
},
|
| 2046 |
-
{
|
| 2047 |
-
"epoch": 0.0696560731388768,
|
| 2048 |
-
"grad_norm": 0.4541665017604828,
|
| 2049 |
-
"learning_rate": 4.981880515107148e-05,
|
| 2050 |
-
"loss": 1.5009,
|
| 2051 |
-
"step": 2880
|
| 2052 |
-
},
|
| 2053 |
-
{
|
| 2054 |
-
"epoch": 0.06989793450394234,
|
| 2055 |
-
"grad_norm": 0.3360423743724823,
|
| 2056 |
-
"learning_rate": 4.981687375262727e-05,
|
| 2057 |
-
"loss": 1.5023,
|
| 2058 |
-
"step": 2890
|
| 2059 |
-
},
|
| 2060 |
-
{
|
| 2061 |
-
"epoch": 0.07013979586900788,
|
| 2062 |
-
"grad_norm": 0.40814629197120667,
|
| 2063 |
-
"learning_rate": 4.981493215290366e-05,
|
| 2064 |
-
"loss": 1.4952,
|
| 2065 |
-
"step": 2900
|
| 2066 |
-
},
|
| 2067 |
-
{
|
| 2068 |
-
"epoch": 0.07038165723407343,
|
| 2069 |
-
"grad_norm": 0.37559887766838074,
|
| 2070 |
-
"learning_rate": 4.981298035269879e-05,
|
| 2071 |
-
"loss": 1.501,
|
| 2072 |
-
"step": 2910
|
| 2073 |
-
},
|
| 2074 |
-
{
|
| 2075 |
-
"epoch": 0.07062351859913897,
|
| 2076 |
-
"grad_norm": 0.39514732360839844,
|
| 2077 |
-
"learning_rate": 4.981101835281496e-05,
|
| 2078 |
-
"loss": 1.4897,
|
| 2079 |
-
"step": 2920
|
| 2080 |
-
},
|
| 2081 |
-
{
|
| 2082 |
-
"epoch": 0.07086537996420451,
|
| 2083 |
-
"grad_norm": 0.38335472345352173,
|
| 2084 |
-
"learning_rate": 4.9809046154058694e-05,
|
| 2085 |
-
"loss": 1.4938,
|
| 2086 |
-
"step": 2930
|
| 2087 |
-
},
|
| 2088 |
-
{
|
| 2089 |
-
"epoch": 0.07110724132927006,
|
| 2090 |
-
"grad_norm": 0.35088109970092773,
|
| 2091 |
-
"learning_rate": 4.980706375724066e-05,
|
| 2092 |
-
"loss": 1.4904,
|
| 2093 |
-
"step": 2940
|
| 2094 |
-
},
|
| 2095 |
-
{
|
| 2096 |
-
"epoch": 0.0713491026943356,
|
| 2097 |
-
"grad_norm": 0.3556269109249115,
|
| 2098 |
-
"learning_rate": 4.9805071163175764e-05,
|
| 2099 |
-
"loss": 1.4952,
|
| 2100 |
-
"step": 2950
|
| 2101 |
-
},
|
| 2102 |
-
{
|
| 2103 |
-
"epoch": 0.07159096405940116,
|
| 2104 |
-
"grad_norm": 0.41015541553497314,
|
| 2105 |
-
"learning_rate": 4.980306837268308e-05,
|
| 2106 |
-
"loss": 1.5099,
|
| 2107 |
-
"step": 2960
|
| 2108 |
-
},
|
| 2109 |
-
{
|
| 2110 |
-
"epoch": 0.0718328254244667,
|
| 2111 |
-
"grad_norm": 0.3832804560661316,
|
| 2112 |
-
"learning_rate": 4.9801055386585895e-05,
|
| 2113 |
-
"loss": 1.4985,
|
| 2114 |
-
"step": 2970
|
| 2115 |
-
},
|
| 2116 |
-
{
|
| 2117 |
-
"epoch": 0.07207468678953224,
|
| 2118 |
-
"grad_norm": 0.33278173208236694,
|
| 2119 |
-
"learning_rate": 4.9799032205711644e-05,
|
| 2120 |
-
"loss": 1.4954,
|
| 2121 |
-
"step": 2980
|
| 2122 |
-
},
|
| 2123 |
-
{
|
| 2124 |
-
"epoch": 0.07231654815459779,
|
| 2125 |
-
"grad_norm": 0.445580393075943,
|
| 2126 |
-
"learning_rate": 4.9796998830891995e-05,
|
| 2127 |
-
"loss": 1.5024,
|
| 2128 |
-
"step": 2990
|
| 2129 |
-
},
|
| 2130 |
-
{
|
| 2131 |
-
"epoch": 0.07255840951966333,
|
| 2132 |
-
"grad_norm": 0.3755344748497009,
|
| 2133 |
-
"learning_rate": 4.979495526296279e-05,
|
| 2134 |
-
"loss": 1.5028,
|
| 2135 |
-
"step": 3000
|
| 2136 |
-
},
|
| 2137 |
-
{
|
| 2138 |
-
"epoch": 0.07255840951966333,
|
| 2139 |
-
"eval_loss": 1.499047875404358,
|
| 2140 |
-
"eval_runtime": 1239.9853,
|
| 2141 |
-
"eval_sacrebleu": 96.74962574504019,
|
| 2142 |
-
"eval_samples_per_second": 81.49,
|
| 2143 |
-
"eval_steps_per_second": 0.637,
|
| 2144 |
-
"step": 3000
|
| 2145 |
-
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 0.07280027088472887,
|
| 2148 |
-
"grad_norm": 0.33310985565185547,
|
| 2149 |
-
"learning_rate": 4.9792901502764075e-05,
|
| 2150 |
-
"loss": 1.4985,
|
| 2151 |
-
"step": 3010
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 0.07304213224979442,
|
| 2155 |
-
"grad_norm": 0.4175123870372772,
|
| 2156 |
-
"learning_rate": 4.979083755114006e-05,
|
| 2157 |
-
"loss": 1.496,
|
| 2158 |
-
"step": 3020
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 0.07328399361485996,
|
| 2162 |
-
"grad_norm": 0.3468244969844818,
|
| 2163 |
-
"learning_rate": 4.978876340893915e-05,
|
| 2164 |
-
"loss": 1.4988,
|
| 2165 |
-
"step": 3030
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 0.0735258549799255,
|
| 2169 |
-
"grad_norm": 0.353986531496048,
|
| 2170 |
-
"learning_rate": 4.978667907701396e-05,
|
| 2171 |
-
"loss": 1.5031,
|
| 2172 |
-
"step": 3040
|
| 2173 |
-
},
|
| 2174 |
-
{
|
| 2175 |
-
"epoch": 0.07376771634499105,
|
| 2176 |
-
"grad_norm": 0.46692317724227905,
|
| 2177 |
-
"learning_rate": 4.9784584556221275e-05,
|
| 2178 |
-
"loss": 1.496,
|
| 2179 |
-
"step": 3050
|
| 2180 |
-
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 0.07400957771005659,
|
| 2183 |
-
"grad_norm": 0.45075079798698425,
|
| 2184 |
-
"learning_rate": 4.978247984742208e-05,
|
| 2185 |
-
"loss": 1.5003,
|
| 2186 |
-
"step": 3060
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 0.07425143907512213,
|
| 2190 |
-
"grad_norm": 0.36158666014671326,
|
| 2191 |
-
"learning_rate": 4.9780364951481525e-05,
|
| 2192 |
-
"loss": 1.4975,
|
| 2193 |
-
"step": 3070
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 0.07449330044018769,
|
| 2197 |
-
"grad_norm": 0.38103577494621277,
|
| 2198 |
-
"learning_rate": 4.977823986926898e-05,
|
| 2199 |
-
"loss": 1.5029,
|
| 2200 |
-
"step": 3080
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 0.07473516180525323,
|
| 2204 |
-
"grad_norm": 0.6088710427284241,
|
| 2205 |
-
"learning_rate": 4.977610460165799e-05,
|
| 2206 |
-
"loss": 1.4975,
|
| 2207 |
-
"step": 3090
|
| 2208 |
-
},
|
| 2209 |
-
{
|
| 2210 |
-
"epoch": 0.07497702317031878,
|
| 2211 |
-
"grad_norm": 0.3984760642051697,
|
| 2212 |
-
"learning_rate": 4.977395914952627e-05,
|
| 2213 |
-
"loss": 1.4991,
|
| 2214 |
-
"step": 3100
|
| 2215 |
-
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 0.07521888453538432,
|
| 2218 |
-
"grad_norm": 0.581408679485321,
|
| 2219 |
-
"learning_rate": 4.977180351375573e-05,
|
| 2220 |
-
"loss": 1.4974,
|
| 2221 |
-
"step": 3110
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 0.07546074590044986,
|
| 2225 |
-
"grad_norm": 0.4531524181365967,
|
| 2226 |
-
"learning_rate": 4.9769637695232486e-05,
|
| 2227 |
-
"loss": 1.4981,
|
| 2228 |
-
"step": 3120
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 0.0757026072655154,
|
| 2232 |
-
"grad_norm": 0.3169078826904297,
|
| 2233 |
-
"learning_rate": 4.976746169484681e-05,
|
| 2234 |
-
"loss": 1.4993,
|
| 2235 |
-
"step": 3130
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 0.07594446863058095,
|
| 2239 |
-
"grad_norm": 0.3769000172615051,
|
| 2240 |
-
"learning_rate": 4.976527551349318e-05,
|
| 2241 |
-
"loss": 1.4966,
|
| 2242 |
-
"step": 3140
|
| 2243 |
-
},
|
| 2244 |
-
{
|
| 2245 |
-
"epoch": 0.07618632999564649,
|
| 2246 |
-
"grad_norm": 0.34887903928756714,
|
| 2247 |
-
"learning_rate": 4.976307915207026e-05,
|
| 2248 |
-
"loss": 1.4998,
|
| 2249 |
-
"step": 3150
|
| 2250 |
-
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 0.07642819136071204,
|
| 2253 |
-
"grad_norm": 0.3374789357185364,
|
| 2254 |
-
"learning_rate": 4.976087261148088e-05,
|
| 2255 |
-
"loss": 1.4968,
|
| 2256 |
-
"step": 3160
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 0.07667005272577758,
|
| 2260 |
-
"grad_norm": 0.3697901964187622,
|
| 2261 |
-
"learning_rate": 4.9758655892632063e-05,
|
| 2262 |
-
"loss": 1.5013,
|
| 2263 |
-
"step": 3170
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 0.07691191409084312,
|
| 2267 |
-
"grad_norm": 0.4903365671634674,
|
| 2268 |
-
"learning_rate": 4.975642899643503e-05,
|
| 2269 |
-
"loss": 1.493,
|
| 2270 |
-
"step": 3180
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 0.07715377545590868,
|
| 2274 |
-
"grad_norm": 0.3290081024169922,
|
| 2275 |
-
"learning_rate": 4.975419192380517e-05,
|
| 2276 |
-
"loss": 1.5078,
|
| 2277 |
-
"step": 3190
|
| 2278 |
-
},
|
| 2279 |
-
{
|
| 2280 |
-
"epoch": 0.07739563682097422,
|
| 2281 |
-
"grad_norm": 0.33302998542785645,
|
| 2282 |
-
"learning_rate": 4.975194467566206e-05,
|
| 2283 |
-
"loss": 1.4965,
|
| 2284 |
-
"step": 3200
|
| 2285 |
-
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 0.07763749818603977,
|
| 2288 |
-
"grad_norm": 0.29818981885910034,
|
| 2289 |
-
"learning_rate": 4.974968725292945e-05,
|
| 2290 |
-
"loss": 1.4978,
|
| 2291 |
-
"step": 3210
|
| 2292 |
-
},
|
| 2293 |
-
{
|
| 2294 |
-
"epoch": 0.07787935955110531,
|
| 2295 |
-
"grad_norm": 0.5125884413719177,
|
| 2296 |
-
"learning_rate": 4.97474196565353e-05,
|
| 2297 |
-
"loss": 1.5056,
|
| 2298 |
-
"step": 3220
|
| 2299 |
-
},
|
| 2300 |
-
{
|
| 2301 |
-
"epoch": 0.07812122091617085,
|
| 2302 |
-
"grad_norm": 0.39383506774902344,
|
| 2303 |
-
"learning_rate": 4.9745141887411704e-05,
|
| 2304 |
-
"loss": 1.5096,
|
| 2305 |
-
"step": 3230
|
| 2306 |
-
},
|
| 2307 |
-
{
|
| 2308 |
-
"epoch": 0.0783630822812364,
|
| 2309 |
-
"grad_norm": 0.3851451575756073,
|
| 2310 |
-
"learning_rate": 4.974285394649499e-05,
|
| 2311 |
-
"loss": 1.501,
|
| 2312 |
-
"step": 3240
|
| 2313 |
-
},
|
| 2314 |
-
{
|
| 2315 |
-
"epoch": 0.07860494364630194,
|
| 2316 |
-
"grad_norm": 0.37914183735847473,
|
| 2317 |
-
"learning_rate": 4.9740555834725646e-05,
|
| 2318 |
-
"loss": 1.4854,
|
| 2319 |
-
"step": 3250
|
| 2320 |
-
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 0.07884680501136748,
|
| 2323 |
-
"grad_norm": 0.3494224548339844,
|
| 2324 |
-
"learning_rate": 4.973824755304832e-05,
|
| 2325 |
-
"loss": 1.5,
|
| 2326 |
-
"step": 3260
|
| 2327 |
-
},
|
| 2328 |
-
{
|
| 2329 |
-
"epoch": 0.07908866637643303,
|
| 2330 |
-
"grad_norm": 0.41473448276519775,
|
| 2331 |
-
"learning_rate": 4.973592910241188e-05,
|
| 2332 |
-
"loss": 1.5026,
|
| 2333 |
-
"step": 3270
|
| 2334 |
-
},
|
| 2335 |
-
{
|
| 2336 |
-
"epoch": 0.07933052774149857,
|
| 2337 |
-
"grad_norm": 0.5182355046272278,
|
| 2338 |
-
"learning_rate": 4.973360048376934e-05,
|
| 2339 |
-
"loss": 1.4998,
|
| 2340 |
-
"step": 3280
|
| 2341 |
-
},
|
| 2342 |
-
{
|
| 2343 |
-
"epoch": 0.07957238910656411,
|
| 2344 |
-
"grad_norm": 0.3290833532810211,
|
| 2345 |
-
"learning_rate": 4.973126169807791e-05,
|
| 2346 |
-
"loss": 1.498,
|
| 2347 |
-
"step": 3290
|
| 2348 |
-
},
|
| 2349 |
-
{
|
| 2350 |
-
"epoch": 0.07981425047162966,
|
| 2351 |
-
"grad_norm": 0.3534269332885742,
|
| 2352 |
-
"learning_rate": 4.9728912746298975e-05,
|
| 2353 |
-
"loss": 1.4917,
|
| 2354 |
-
"step": 3300
|
| 2355 |
-
},
|
| 2356 |
-
{
|
| 2357 |
-
"epoch": 0.08005611183669521,
|
| 2358 |
-
"grad_norm": 0.32593613862991333,
|
| 2359 |
-
"learning_rate": 4.97265536293981e-05,
|
| 2360 |
-
"loss": 1.4909,
|
| 2361 |
-
"step": 3310
|
| 2362 |
-
},
|
| 2363 |
-
{
|
| 2364 |
-
"epoch": 0.08029797320176076,
|
| 2365 |
-
"grad_norm": 0.38789451122283936,
|
| 2366 |
-
"learning_rate": 4.972418434834504e-05,
|
| 2367 |
-
"loss": 1.5006,
|
| 2368 |
-
"step": 3320
|
| 2369 |
-
},
|
| 2370 |
-
{
|
| 2371 |
-
"epoch": 0.0805398345668263,
|
| 2372 |
-
"grad_norm": 0.46936559677124023,
|
| 2373 |
-
"learning_rate": 4.9721804904113705e-05,
|
| 2374 |
-
"loss": 1.5049,
|
| 2375 |
-
"step": 3330
|
| 2376 |
-
},
|
| 2377 |
-
{
|
| 2378 |
-
"epoch": 0.08078169593189184,
|
| 2379 |
-
"grad_norm": 0.3091851472854614,
|
| 2380 |
-
"learning_rate": 4.97194152976822e-05,
|
| 2381 |
-
"loss": 1.4966,
|
| 2382 |
-
"step": 3340
|
| 2383 |
-
},
|
| 2384 |
-
{
|
| 2385 |
-
"epoch": 0.08102355729695739,
|
| 2386 |
-
"grad_norm": 0.45702749490737915,
|
| 2387 |
-
"learning_rate": 4.97170155300328e-05,
|
| 2388 |
-
"loss": 1.5079,
|
| 2389 |
-
"step": 3350
|
| 2390 |
-
},
|
| 2391 |
-
{
|
| 2392 |
-
"epoch": 0.08126541866202293,
|
| 2393 |
-
"grad_norm": 0.40697166323661804,
|
| 2394 |
-
"learning_rate": 4.9714605602151955e-05,
|
| 2395 |
-
"loss": 1.5018,
|
| 2396 |
-
"step": 3360
|
| 2397 |
-
},
|
| 2398 |
-
{
|
| 2399 |
-
"epoch": 0.08150728002708847,
|
| 2400 |
-
"grad_norm": 0.532478392124176,
|
| 2401 |
-
"learning_rate": 4.971218551503029e-05,
|
| 2402 |
-
"loss": 1.4866,
|
| 2403 |
-
"step": 3370
|
| 2404 |
-
},
|
| 2405 |
-
{
|
| 2406 |
-
"epoch": 0.08174914139215401,
|
| 2407 |
-
"grad_norm": 0.4273573160171509,
|
| 2408 |
-
"learning_rate": 4.970975526966264e-05,
|
| 2409 |
-
"loss": 1.495,
|
| 2410 |
-
"step": 3380
|
| 2411 |
-
},
|
| 2412 |
-
{
|
| 2413 |
-
"epoch": 0.08199100275721956,
|
| 2414 |
-
"grad_norm": 0.4085642397403717,
|
| 2415 |
-
"learning_rate": 4.970731486704796e-05,
|
| 2416 |
-
"loss": 1.4925,
|
| 2417 |
-
"step": 3390
|
| 2418 |
-
},
|
| 2419 |
-
{
|
| 2420 |
-
"epoch": 0.0822328641222851,
|
| 2421 |
-
"grad_norm": 0.28924253582954407,
|
| 2422 |
-
"learning_rate": 4.9704864308189415e-05,
|
| 2423 |
-
"loss": 1.4942,
|
| 2424 |
-
"step": 3400
|
| 2425 |
-
},
|
| 2426 |
-
{
|
| 2427 |
-
"epoch": 0.08247472548735064,
|
| 2428 |
-
"grad_norm": 0.47220590710639954,
|
| 2429 |
-
"learning_rate": 4.9702403594094335e-05,
|
| 2430 |
-
"loss": 1.4882,
|
| 2431 |
-
"step": 3410
|
| 2432 |
-
},
|
| 2433 |
-
{
|
| 2434 |
-
"epoch": 0.0827165868524162,
|
| 2435 |
-
"grad_norm": 0.35851794481277466,
|
| 2436 |
-
"learning_rate": 4.9699932725774234e-05,
|
| 2437 |
-
"loss": 1.4992,
|
| 2438 |
-
"step": 3420
|
| 2439 |
-
},
|
| 2440 |
-
{
|
| 2441 |
-
"epoch": 0.08295844821748175,
|
| 2442 |
-
"grad_norm": 0.40433382987976074,
|
| 2443 |
-
"learning_rate": 4.969745170424478e-05,
|
| 2444 |
-
"loss": 1.5014,
|
| 2445 |
-
"step": 3430
|
| 2446 |
-
},
|
| 2447 |
-
{
|
| 2448 |
-
"epoch": 0.08320030958254729,
|
| 2449 |
-
"grad_norm": 0.40462055802345276,
|
| 2450 |
-
"learning_rate": 4.969496053052584e-05,
|
| 2451 |
-
"loss": 1.4984,
|
| 2452 |
-
"step": 3440
|
| 2453 |
-
},
|
| 2454 |
-
{
|
| 2455 |
-
"epoch": 0.08344217094761283,
|
| 2456 |
-
"grad_norm": 0.39987021684646606,
|
| 2457 |
-
"learning_rate": 4.969245920564144e-05,
|
| 2458 |
-
"loss": 1.4943,
|
| 2459 |
-
"step": 3450
|
| 2460 |
-
},
|
| 2461 |
-
{
|
| 2462 |
-
"epoch": 0.08368403231267837,
|
| 2463 |
-
"grad_norm": 0.4980989992618561,
|
| 2464 |
-
"learning_rate": 4.968994773061977e-05,
|
| 2465 |
-
"loss": 1.4958,
|
| 2466 |
-
"step": 3460
|
| 2467 |
-
},
|
| 2468 |
-
{
|
| 2469 |
-
"epoch": 0.08392589367774392,
|
| 2470 |
-
"grad_norm": 0.32841891050338745,
|
| 2471 |
-
"learning_rate": 4.968742610649321e-05,
|
| 2472 |
-
"loss": 1.4957,
|
| 2473 |
-
"step": 3470
|
| 2474 |
-
},
|
| 2475 |
-
{
|
| 2476 |
-
"epoch": 0.08416775504280946,
|
| 2477 |
-
"grad_norm": 0.36307471990585327,
|
| 2478 |
-
"learning_rate": 4.96848943342983e-05,
|
| 2479 |
-
"loss": 1.5073,
|
| 2480 |
-
"step": 3480
|
| 2481 |
-
},
|
| 2482 |
-
{
|
| 2483 |
-
"epoch": 0.084409616407875,
|
| 2484 |
-
"grad_norm": 0.5699554681777954,
|
| 2485 |
-
"learning_rate": 4.968235241507576e-05,
|
| 2486 |
-
"loss": 1.4998,
|
| 2487 |
-
"step": 3490
|
| 2488 |
-
},
|
| 2489 |
-
{
|
| 2490 |
-
"epoch": 0.08465147777294055,
|
| 2491 |
-
"grad_norm": 0.36965617537498474,
|
| 2492 |
-
"learning_rate": 4.967980034987048e-05,
|
| 2493 |
-
"loss": 1.4925,
|
| 2494 |
-
"step": 3500
|
| 2495 |
-
},
|
| 2496 |
-
{
|
| 2497 |
-
"epoch": 0.08489333913800609,
|
| 2498 |
-
"grad_norm": 0.3929795026779175,
|
| 2499 |
-
"learning_rate": 4.967723813973151e-05,
|
| 2500 |
-
"loss": 1.4951,
|
| 2501 |
-
"step": 3510
|
| 2502 |
-
},
|
| 2503 |
-
{
|
| 2504 |
-
"epoch": 0.08513520050307163,
|
| 2505 |
-
"grad_norm": 0.35785186290740967,
|
| 2506 |
-
"learning_rate": 4.967466578571209e-05,
|
| 2507 |
-
"loss": 1.4903,
|
| 2508 |
-
"step": 3520
|
| 2509 |
-
},
|
| 2510 |
-
{
|
| 2511 |
-
"epoch": 0.08537706186813718,
|
| 2512 |
-
"grad_norm": 0.3917689025402069,
|
| 2513 |
-
"learning_rate": 4.96720832888696e-05,
|
| 2514 |
-
"loss": 1.496,
|
| 2515 |
-
"step": 3530
|
| 2516 |
-
},
|
| 2517 |
-
{
|
| 2518 |
-
"epoch": 0.08561892323320273,
|
| 2519 |
-
"grad_norm": 0.37746888399124146,
|
| 2520 |
-
"learning_rate": 4.9669490650265624e-05,
|
| 2521 |
-
"loss": 1.488,
|
| 2522 |
-
"step": 3540
|
| 2523 |
-
},
|
| 2524 |
-
{
|
| 2525 |
-
"epoch": 0.08586078459826828,
|
| 2526 |
-
"grad_norm": 0.4309702217578888,
|
| 2527 |
-
"learning_rate": 4.9666887870965885e-05,
|
| 2528 |
-
"loss": 1.4991,
|
| 2529 |
-
"step": 3550
|
| 2530 |
-
},
|
| 2531 |
-
{
|
| 2532 |
-
"epoch": 0.08610264596333382,
|
| 2533 |
-
"grad_norm": 0.4067014753818512,
|
| 2534 |
-
"learning_rate": 4.966427495204029e-05,
|
| 2535 |
-
"loss": 1.495,
|
| 2536 |
-
"step": 3560
|
| 2537 |
-
},
|
| 2538 |
-
{
|
| 2539 |
-
"epoch": 0.08634450732839936,
|
| 2540 |
-
"grad_norm": 0.29803594946861267,
|
| 2541 |
-
"learning_rate": 4.966165189456291e-05,
|
| 2542 |
-
"loss": 1.4948,
|
| 2543 |
-
"step": 3570
|
| 2544 |
-
},
|
| 2545 |
-
{
|
| 2546 |
-
"epoch": 0.08658636869346491,
|
| 2547 |
-
"grad_norm": 0.32155147194862366,
|
| 2548 |
-
"learning_rate": 4.9659018699611995e-05,
|
| 2549 |
-
"loss": 1.4942,
|
| 2550 |
-
"step": 3580
|
| 2551 |
-
},
|
| 2552 |
-
{
|
| 2553 |
-
"epoch": 0.08682823005853045,
|
| 2554 |
-
"grad_norm": 0.30671581625938416,
|
| 2555 |
-
"learning_rate": 4.965637536826995e-05,
|
| 2556 |
-
"loss": 1.4941,
|
| 2557 |
-
"step": 3590
|
| 2558 |
-
},
|
| 2559 |
-
{
|
| 2560 |
-
"epoch": 0.087070091423596,
|
| 2561 |
-
"grad_norm": 0.3266153931617737,
|
| 2562 |
-
"learning_rate": 4.965372190162333e-05,
|
| 2563 |
-
"loss": 1.5009,
|
| 2564 |
-
"step": 3600
|
| 2565 |
-
},
|
| 2566 |
-
{
|
| 2567 |
-
"epoch": 0.08731195278866154,
|
| 2568 |
-
"grad_norm": 0.2917800545692444,
|
| 2569 |
-
"learning_rate": 4.9651058300762895e-05,
|
| 2570 |
-
"loss": 1.4879,
|
| 2571 |
-
"step": 3610
|
| 2572 |
-
},
|
| 2573 |
-
{
|
| 2574 |
-
"epoch": 0.08755381415372708,
|
| 2575 |
-
"grad_norm": 0.3614349067211151,
|
| 2576 |
-
"learning_rate": 4.9648384566783545e-05,
|
| 2577 |
-
"loss": 1.4868,
|
| 2578 |
-
"step": 3620
|
| 2579 |
-
},
|
| 2580 |
-
{
|
| 2581 |
-
"epoch": 0.08779567551879262,
|
| 2582 |
-
"grad_norm": 0.3470216691493988,
|
| 2583 |
-
"learning_rate": 4.964570070078435e-05,
|
| 2584 |
-
"loss": 1.4854,
|
| 2585 |
-
"step": 3630
|
| 2586 |
-
},
|
| 2587 |
-
{
|
| 2588 |
-
"epoch": 0.08803753688385817,
|
| 2589 |
-
"grad_norm": 0.5655723214149475,
|
| 2590 |
-
"learning_rate": 4.964300670386856e-05,
|
| 2591 |
-
"loss": 1.4926,
|
| 2592 |
-
"step": 3640
|
| 2593 |
-
},
|
| 2594 |
-
{
|
| 2595 |
-
"epoch": 0.08827939824892371,
|
| 2596 |
-
"grad_norm": 1.6314070224761963,
|
| 2597 |
-
"learning_rate": 4.9640302577143536e-05,
|
| 2598 |
-
"loss": 1.5252,
|
| 2599 |
-
"step": 3650
|
| 2600 |
-
},
|
| 2601 |
-
{
|
| 2602 |
-
"epoch": 0.08852125961398927,
|
| 2603 |
-
"grad_norm": 34.27970886230469,
|
| 2604 |
-
"learning_rate": 4.963758832172088e-05,
|
| 2605 |
-
"loss": 4.7597,
|
| 2606 |
-
"step": 3660
|
| 2607 |
-
},
|
| 2608 |
-
{
|
| 2609 |
-
"epoch": 0.08876312097905481,
|
| 2610 |
-
"grad_norm": 9.672608375549316,
|
| 2611 |
-
"learning_rate": 4.9634863938716315e-05,
|
| 2612 |
-
"loss": 9.2436,
|
| 2613 |
-
"step": 3670
|
| 2614 |
-
},
|
| 2615 |
-
{
|
| 2616 |
-
"epoch": 0.08900498234412035,
|
| 2617 |
-
"grad_norm": 18.468290328979492,
|
| 2618 |
-
"learning_rate": 4.9632129429249716e-05,
|
| 2619 |
-
"loss": 8.3476,
|
| 2620 |
-
"step": 3680
|
| 2621 |
-
},
|
| 2622 |
-
{
|
| 2623 |
-
"epoch": 0.0892468437091859,
|
| 2624 |
-
"grad_norm": 4.984256744384766,
|
| 2625 |
-
"learning_rate": 4.9629384794445157e-05,
|
| 2626 |
-
"loss": 8.0705,
|
| 2627 |
-
"step": 3690
|
| 2628 |
-
},
|
| 2629 |
-
{
|
| 2630 |
-
"epoch": 0.08948870507425144,
|
| 2631 |
-
"grad_norm": 3.9478564262390137,
|
| 2632 |
-
"learning_rate": 4.962663003543083e-05,
|
| 2633 |
-
"loss": 7.7813,
|
| 2634 |
-
"step": 3700
|
| 2635 |
-
},
|
| 2636 |
-
{
|
| 2637 |
-
"epoch": 0.08973056643931698,
|
| 2638 |
-
"grad_norm": 3.284527063369751,
|
| 2639 |
-
"learning_rate": 4.962386515333913e-05,
|
| 2640 |
-
"loss": 7.5749,
|
| 2641 |
-
"step": 3710
|
| 2642 |
-
},
|
| 2643 |
-
{
|
| 2644 |
-
"epoch": 0.08997242780438253,
|
| 2645 |
-
"grad_norm": 4.682032585144043,
|
| 2646 |
-
"learning_rate": 4.962109014930659e-05,
|
| 2647 |
-
"loss": 7.3946,
|
| 2648 |
-
"step": 3720
|
| 2649 |
-
},
|
| 2650 |
-
{
|
| 2651 |
-
"epoch": 0.09021428916944807,
|
| 2652 |
-
"grad_norm": 3.8620240688323975,
|
| 2653 |
-
"learning_rate": 4.9618305024473905e-05,
|
| 2654 |
-
"loss": 5.399,
|
| 2655 |
-
"step": 3730
|
| 2656 |
-
},
|
| 2657 |
-
{
|
| 2658 |
-
"epoch": 0.09045615053451361,
|
| 2659 |
-
"grad_norm": 1.1052254438400269,
|
| 2660 |
-
"learning_rate": 4.961550977998594e-05,
|
| 2661 |
-
"loss": 1.7208,
|
| 2662 |
-
"step": 3740
|
| 2663 |
-
},
|
| 2664 |
-
{
|
| 2665 |
-
"epoch": 0.09069801189957916,
|
| 2666 |
-
"grad_norm": 45.039066314697266,
|
| 2667 |
-
"learning_rate": 4.9612704416991716e-05,
|
| 2668 |
-
"loss": 1.6211,
|
| 2669 |
-
"step": 3750
|
| 2670 |
-
},
|
| 2671 |
-
{
|
| 2672 |
-
"epoch": 0.0909398732646447,
|
| 2673 |
-
"grad_norm": 0.7098454236984253,
|
| 2674 |
-
"learning_rate": 4.960988893664441e-05,
|
| 2675 |
-
"loss": 1.6957,
|
| 2676 |
-
"step": 3760
|
| 2677 |
-
},
|
| 2678 |
-
{
|
| 2679 |
-
"epoch": 0.09118173462971026,
|
| 2680 |
-
"grad_norm": 2.7060585021972656,
|
| 2681 |
-
"learning_rate": 4.9607063340101354e-05,
|
| 2682 |
-
"loss": 1.6564,
|
| 2683 |
-
"step": 3770
|
| 2684 |
-
},
|
| 2685 |
-
{
|
| 2686 |
-
"epoch": 0.0914235959947758,
|
| 2687 |
-
"grad_norm": 2.3458268642425537,
|
| 2688 |
-
"learning_rate": 4.960422762852406e-05,
|
| 2689 |
-
"loss": 2.4888,
|
| 2690 |
-
"step": 3780
|
| 2691 |
-
},
|
| 2692 |
-
{
|
| 2693 |
-
"epoch": 0.09166545735984134,
|
| 2694 |
-
"grad_norm": 12.697403907775879,
|
| 2695 |
-
"learning_rate": 4.960138180307816e-05,
|
| 2696 |
-
"loss": 2.405,
|
| 2697 |
-
"step": 3790
|
| 2698 |
-
},
|
| 2699 |
-
{
|
| 2700 |
-
"epoch": 0.09190731872490689,
|
| 2701 |
-
"grad_norm": 7.4579854011535645,
|
| 2702 |
-
"learning_rate": 4.959852586493349e-05,
|
| 2703 |
-
"loss": 3.2748,
|
| 2704 |
-
"step": 3800
|
| 2705 |
-
},
|
| 2706 |
-
{
|
| 2707 |
-
"epoch": 0.09214918008997243,
|
| 2708 |
-
"grad_norm": 0.6929895877838135,
|
| 2709 |
-
"learning_rate": 4.9595659815263985e-05,
|
| 2710 |
-
"loss": 2.112,
|
| 2711 |
-
"step": 3810
|
| 2712 |
-
},
|
| 2713 |
-
{
|
| 2714 |
-
"epoch": 0.09239104145503797,
|
| 2715 |
-
"grad_norm": 0.563450276851654,
|
| 2716 |
-
"learning_rate": 4.95927836552478e-05,
|
| 2717 |
-
"loss": 1.6898,
|
| 2718 |
-
"step": 3820
|
| 2719 |
-
},
|
| 2720 |
-
{
|
| 2721 |
-
"epoch": 0.09263290282010352,
|
| 2722 |
-
"grad_norm": 0.4547874629497528,
|
| 2723 |
-
"learning_rate": 4.9589897386067205e-05,
|
| 2724 |
-
"loss": 1.5361,
|
| 2725 |
-
"step": 3830
|
| 2726 |
-
},
|
| 2727 |
-
{
|
| 2728 |
-
"epoch": 0.09287476418516906,
|
| 2729 |
-
"grad_norm": 0.525037407875061,
|
| 2730 |
-
"learning_rate": 4.958700100890863e-05,
|
| 2731 |
-
"loss": 1.5204,
|
| 2732 |
-
"step": 3840
|
| 2733 |
-
},
|
| 2734 |
-
{
|
| 2735 |
-
"epoch": 0.0931166255502346,
|
| 2736 |
-
"grad_norm": 0.826223611831665,
|
| 2737 |
-
"learning_rate": 4.9584094524962684e-05,
|
| 2738 |
-
"loss": 1.5188,
|
| 2739 |
-
"step": 3850
|
| 2740 |
-
},
|
| 2741 |
-
{
|
| 2742 |
-
"epoch": 0.09335848691530015,
|
| 2743 |
-
"grad_norm": 0.45551028847694397,
|
| 2744 |
-
"learning_rate": 4.95811779354241e-05,
|
| 2745 |
-
"loss": 1.5779,
|
| 2746 |
-
"step": 3860
|
| 2747 |
-
},
|
| 2748 |
-
{
|
| 2749 |
-
"epoch": 0.09360034828036569,
|
| 2750 |
-
"grad_norm": 0.4146782159805298,
|
| 2751 |
-
"learning_rate": 4.957825124149177e-05,
|
| 2752 |
-
"loss": 1.5293,
|
| 2753 |
-
"step": 3870
|
| 2754 |
-
},
|
| 2755 |
-
{
|
| 2756 |
-
"epoch": 0.09384220964543123,
|
| 2757 |
-
"grad_norm": 0.40724071860313416,
|
| 2758 |
-
"learning_rate": 4.9575314444368756e-05,
|
| 2759 |
-
"loss": 1.5346,
|
| 2760 |
-
"step": 3880
|
| 2761 |
-
},
|
| 2762 |
-
{
|
| 2763 |
-
"epoch": 0.09408407101049679,
|
| 2764 |
-
"grad_norm": 0.47108572721481323,
|
| 2765 |
-
"learning_rate": 4.957236754526227e-05,
|
| 2766 |
-
"loss": 1.5205,
|
| 2767 |
-
"step": 3890
|
| 2768 |
-
},
|
| 2769 |
-
{
|
| 2770 |
-
"epoch": 0.09432593237556233,
|
| 2771 |
-
"grad_norm": 0.45449286699295044,
|
| 2772 |
-
"learning_rate": 4.9569410545383665e-05,
|
| 2773 |
-
"loss": 1.5149,
|
| 2774 |
-
"step": 3900
|
| 2775 |
-
},
|
| 2776 |
-
{
|
| 2777 |
-
"epoch": 0.09456779374062788,
|
| 2778 |
-
"grad_norm": 0.5905476808547974,
|
| 2779 |
-
"learning_rate": 4.956644344594845e-05,
|
| 2780 |
-
"loss": 1.5242,
|
| 2781 |
-
"step": 3910
|
| 2782 |
-
},
|
| 2783 |
-
{
|
| 2784 |
-
"epoch": 0.09480965510569342,
|
| 2785 |
-
"grad_norm": 0.4378545582294464,
|
| 2786 |
-
"learning_rate": 4.956346624817629e-05,
|
| 2787 |
-
"loss": 1.495,
|
| 2788 |
-
"step": 3920
|
| 2789 |
-
},
|
| 2790 |
-
{
|
| 2791 |
-
"epoch": 0.09505151647075896,
|
| 2792 |
-
"grad_norm": 0.3891792893409729,
|
| 2793 |
-
"learning_rate": 4.956047895329101e-05,
|
| 2794 |
-
"loss": 1.5011,
|
| 2795 |
-
"step": 3930
|
| 2796 |
-
},
|
| 2797 |
-
{
|
| 2798 |
-
"epoch": 0.0952933778358245,
|
| 2799 |
-
"grad_norm": 0.4357585310935974,
|
| 2800 |
-
"learning_rate": 4.9557481562520546e-05,
|
| 2801 |
-
"loss": 1.5061,
|
| 2802 |
-
"step": 3940
|
| 2803 |
-
},
|
| 2804 |
-
{
|
| 2805 |
-
"epoch": 0.09553523920089005,
|
| 2806 |
-
"grad_norm": 0.36777547001838684,
|
| 2807 |
-
"learning_rate": 4.955447407709704e-05,
|
| 2808 |
-
"loss": 1.4988,
|
| 2809 |
-
"step": 3950
|
| 2810 |
-
},
|
| 2811 |
-
{
|
| 2812 |
-
"epoch": 0.09577710056595559,
|
| 2813 |
-
"grad_norm": 0.42386820912361145,
|
| 2814 |
-
"learning_rate": 4.9551456498256744e-05,
|
| 2815 |
-
"loss": 1.5029,
|
| 2816 |
-
"step": 3960
|
| 2817 |
-
},
|
| 2818 |
-
{
|
| 2819 |
-
"epoch": 0.09601896193102114,
|
| 2820 |
-
"grad_norm": 0.5050960183143616,
|
| 2821 |
-
"learning_rate": 4.954842882724007e-05,
|
| 2822 |
-
"loss": 1.5055,
|
| 2823 |
-
"step": 3970
|
| 2824 |
-
},
|
| 2825 |
-
{
|
| 2826 |
-
"epoch": 0.09626082329608668,
|
| 2827 |
-
"grad_norm": 0.32713472843170166,
|
| 2828 |
-
"learning_rate": 4.954539106529157e-05,
|
| 2829 |
-
"loss": 1.5023,
|
| 2830 |
-
"step": 3980
|
| 2831 |
-
},
|
| 2832 |
-
{
|
| 2833 |
-
"epoch": 0.09650268466115222,
|
| 2834 |
-
"grad_norm": 0.2444874346256256,
|
| 2835 |
-
"learning_rate": 4.9542343213659974e-05,
|
| 2836 |
-
"loss": 1.4971,
|
| 2837 |
-
"step": 3990
|
| 2838 |
-
},
|
| 2839 |
-
{
|
| 2840 |
-
"epoch": 0.09674454602621778,
|
| 2841 |
-
"grad_norm": 0.4041503965854645,
|
| 2842 |
-
"learning_rate": 4.953928527359812e-05,
|
| 2843 |
-
"loss": 1.503,
|
| 2844 |
-
"step": 4000
|
| 2845 |
-
},
|
| 2846 |
-
{
|
| 2847 |
-
"epoch": 0.09674454602621778,
|
| 2848 |
-
"eval_loss": 1.4903924465179443,
|
| 2849 |
-
"eval_runtime": 1236.6443,
|
| 2850 |
-
"eval_sacrebleu": 96.6133417933895,
|
| 2851 |
-
"eval_samples_per_second": 81.71,
|
| 2852 |
-
"eval_steps_per_second": 0.639,
|
| 2853 |
-
"step": 4000
|
| 2854 |
}
|
| 2855 |
],
|
| 2856 |
"logging_steps": 10,
|
|
@@ -2879,7 +752,7 @@
|
|
| 2879 |
"attributes": {}
|
| 2880 |
}
|
| 2881 |
},
|
| 2882 |
-
"total_flos":
|
| 2883 |
"train_batch_size": 64,
|
| 2884 |
"trial_name": null,
|
| 2885 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 1000,
|
| 3 |
+
"best_metric": 1.5370113849639893,
|
| 4 |
+
"best_model_checkpoint": "hieptt/vietnamese-correction-ft/checkpoint-1000",
|
| 5 |
+
"epoch": 0.024186136506554445,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.00024186136506554442,
|
| 21 |
+
"grad_norm": 7.38713264465332,
|
| 22 |
"learning_rate": 4.5e-07,
|
| 23 |
+
"loss": 4.9047,
|
| 24 |
"step": 10
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.00048372273013108885,
|
| 28 |
+
"grad_norm": 9.883085250854492,
|
| 29 |
"learning_rate": 9.5e-07,
|
| 30 |
+
"loss": 4.7161,
|
| 31 |
"step": 20
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.0007255840951966333,
|
| 35 |
+
"grad_norm": 12.030326843261719,
|
| 36 |
"learning_rate": 1.45e-06,
|
| 37 |
+
"loss": 4.3936,
|
| 38 |
"step": 30
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.0009674454602621777,
|
| 42 |
+
"grad_norm": 13.895113945007324,
|
| 43 |
"learning_rate": 1.95e-06,
|
| 44 |
+
"loss": 4.0292,
|
| 45 |
"step": 40
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.0012093068253277222,
|
| 49 |
+
"grad_norm": 14.019319534301758,
|
| 50 |
"learning_rate": 2.4500000000000003e-06,
|
| 51 |
+
"loss": 3.5471,
|
| 52 |
"step": 50
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.0014511681903932665,
|
| 56 |
+
"grad_norm": 4.721616744995117,
|
| 57 |
"learning_rate": 2.95e-06,
|
| 58 |
+
"loss": 2.88,
|
| 59 |
"step": 60
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.001693029555458811,
|
| 63 |
+
"grad_norm": 2.526538372039795,
|
| 64 |
"learning_rate": 3.4500000000000004e-06,
|
| 65 |
"loss": 2.3633,
|
| 66 |
"step": 70
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.0019348909205243554,
|
| 70 |
+
"grad_norm": 1.627066969871521,
|
| 71 |
"learning_rate": 3.95e-06,
|
| 72 |
"loss": 2.0775,
|
| 73 |
"step": 80
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.0021767522855899,
|
| 77 |
+
"grad_norm": 0.8922048807144165,
|
| 78 |
"learning_rate": 4.45e-06,
|
| 79 |
"loss": 1.9554,
|
| 80 |
"step": 90
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.0024186136506554445,
|
| 84 |
+
"grad_norm": 0.8262912631034851,
|
| 85 |
"learning_rate": 4.950000000000001e-06,
|
| 86 |
"loss": 1.887,
|
| 87 |
"step": 100
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.0026604750157209886,
|
| 91 |
+
"grad_norm": 0.8611723184585571,
|
| 92 |
"learning_rate": 5.45e-06,
|
| 93 |
"loss": 1.8405,
|
| 94 |
"step": 110
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.002902336380786533,
|
| 98 |
+
"grad_norm": 0.7151685357093811,
|
| 99 |
"learning_rate": 5.95e-06,
|
| 100 |
"loss": 1.8095,
|
| 101 |
"step": 120
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.0031441977458520776,
|
| 105 |
+
"grad_norm": 0.6910758018493652,
|
| 106 |
"learning_rate": 6.45e-06,
|
| 107 |
"loss": 1.7881,
|
| 108 |
"step": 130
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.003386059110917622,
|
| 112 |
+
"grad_norm": 0.7488411664962769,
|
| 113 |
"learning_rate": 6.950000000000001e-06,
|
| 114 |
"loss": 1.7647,
|
| 115 |
"step": 140
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.0036279204759831663,
|
| 119 |
+
"grad_norm": 0.6839690208435059,
|
| 120 |
"learning_rate": 7.45e-06,
|
| 121 |
"loss": 1.7577,
|
| 122 |
"step": 150
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.003869781841048711,
|
| 126 |
+
"grad_norm": 0.7202065587043762,
|
| 127 |
"learning_rate": 7.95e-06,
|
| 128 |
"loss": 1.7499,
|
| 129 |
"step": 160
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.004111643206114255,
|
| 133 |
+
"grad_norm": 0.6052983403205872,
|
| 134 |
"learning_rate": 8.45e-06,
|
| 135 |
"loss": 1.741,
|
| 136 |
"step": 170
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.0043535045711798,
|
| 140 |
+
"grad_norm": 0.7364121079444885,
|
| 141 |
"learning_rate": 8.95e-06,
|
| 142 |
"loss": 1.7283,
|
| 143 |
"step": 180
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.004595365936245344,
|
| 147 |
+
"grad_norm": 0.7102776169776917,
|
| 148 |
"learning_rate": 9.450000000000001e-06,
|
| 149 |
"loss": 1.707,
|
| 150 |
"step": 190
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.004837227301310889,
|
| 154 |
+
"grad_norm": 0.6977587342262268,
|
| 155 |
"learning_rate": 9.950000000000001e-06,
|
| 156 |
"loss": 1.706,
|
| 157 |
"step": 200
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.005079088666376433,
|
| 161 |
+
"grad_norm": 0.6137486100196838,
|
| 162 |
"learning_rate": 1.045e-05,
|
| 163 |
"loss": 1.7065,
|
| 164 |
"step": 210
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.005320950031441977,
|
| 168 |
+
"grad_norm": 0.5842403769493103,
|
| 169 |
"learning_rate": 1.095e-05,
|
| 170 |
+
"loss": 1.6961,
|
| 171 |
"step": 220
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.005562811396507522,
|
| 175 |
+
"grad_norm": 0.6366673111915588,
|
| 176 |
"learning_rate": 1.145e-05,
|
| 177 |
"loss": 1.6822,
|
| 178 |
"step": 230
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.005804672761573066,
|
| 182 |
+
"grad_norm": 0.5458840131759644,
|
| 183 |
"learning_rate": 1.195e-05,
|
| 184 |
"loss": 1.6794,
|
| 185 |
"step": 240
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.006046534126638611,
|
| 189 |
+
"grad_norm": 0.5614727735519409,
|
| 190 |
"learning_rate": 1.2450000000000001e-05,
|
| 191 |
"loss": 1.6587,
|
| 192 |
"step": 250
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.006288395491704155,
|
| 196 |
+
"grad_norm": 0.6590563058853149,
|
| 197 |
"learning_rate": 1.2950000000000001e-05,
|
| 198 |
"loss": 1.6756,
|
| 199 |
"step": 260
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.006530256856769699,
|
| 203 |
+
"grad_norm": 0.6136428713798523,
|
| 204 |
"learning_rate": 1.3450000000000002e-05,
|
| 205 |
"loss": 1.6723,
|
| 206 |
"step": 270
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.006772118221835244,
|
| 210 |
+
"grad_norm": 0.6550738215446472,
|
| 211 |
"learning_rate": 1.3950000000000002e-05,
|
| 212 |
"loss": 1.6605,
|
| 213 |
"step": 280
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.007013979586900788,
|
| 217 |
+
"grad_norm": 0.5237274765968323,
|
| 218 |
"learning_rate": 1.4449999999999999e-05,
|
| 219 |
+
"loss": 1.6496,
|
| 220 |
"step": 290
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.0072558409519663325,
|
| 224 |
+
"grad_norm": 0.6175047159194946,
|
| 225 |
"learning_rate": 1.4950000000000001e-05,
|
| 226 |
"loss": 1.6441,
|
| 227 |
"step": 300
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.0074977023170318775,
|
| 231 |
+
"grad_norm": 0.5207669138908386,
|
| 232 |
"learning_rate": 1.545e-05,
|
| 233 |
"loss": 1.6478,
|
| 234 |
"step": 310
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.007739563682097422,
|
| 238 |
+
"grad_norm": 0.4118206799030304,
|
| 239 |
"learning_rate": 1.595e-05,
|
| 240 |
"loss": 1.6405,
|
| 241 |
"step": 320
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.007981425047162967,
|
| 245 |
+
"grad_norm": 0.5392739772796631,
|
| 246 |
"learning_rate": 1.645e-05,
|
| 247 |
"loss": 1.6441,
|
| 248 |
"step": 330
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.00822328641222851,
|
| 252 |
+
"grad_norm": 0.5163934826850891,
|
| 253 |
"learning_rate": 1.6950000000000002e-05,
|
| 254 |
"loss": 1.6401,
|
| 255 |
"step": 340
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.008465147777294055,
|
| 259 |
+
"grad_norm": 0.5002100467681885,
|
| 260 |
"learning_rate": 1.745e-05,
|
| 261 |
"loss": 1.6385,
|
| 262 |
"step": 350
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.0087070091423596,
|
| 266 |
+
"grad_norm": 0.6193915009498596,
|
| 267 |
"learning_rate": 1.795e-05,
|
| 268 |
"loss": 1.6222,
|
| 269 |
"step": 360
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.008948870507425145,
|
| 273 |
+
"grad_norm": 0.45196324586868286,
|
| 274 |
"learning_rate": 1.845e-05,
|
| 275 |
"loss": 1.6267,
|
| 276 |
"step": 370
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.009190731872490688,
|
| 280 |
+
"grad_norm": 0.45798471570014954,
|
| 281 |
"learning_rate": 1.895e-05,
|
| 282 |
"loss": 1.6265,
|
| 283 |
"step": 380
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.009432593237556233,
|
| 287 |
+
"grad_norm": 0.5480896234512329,
|
| 288 |
"learning_rate": 1.9450000000000002e-05,
|
| 289 |
"loss": 1.6208,
|
| 290 |
"step": 390
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.009674454602621778,
|
| 294 |
+
"grad_norm": 0.4133257567882538,
|
| 295 |
"learning_rate": 1.995e-05,
|
| 296 |
"loss": 1.6137,
|
| 297 |
"step": 400
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.009916315967687321,
|
| 301 |
+
"grad_norm": 0.45199403166770935,
|
| 302 |
"learning_rate": 2.045e-05,
|
| 303 |
"loss": 1.6196,
|
| 304 |
"step": 410
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.010158177332752866,
|
| 308 |
+
"grad_norm": 0.5875248312950134,
|
| 309 |
"learning_rate": 2.095e-05,
|
| 310 |
"loss": 1.6049,
|
| 311 |
"step": 420
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.010400038697818411,
|
| 315 |
+
"grad_norm": 0.6048296093940735,
|
| 316 |
"learning_rate": 2.145e-05,
|
| 317 |
"loss": 1.6083,
|
| 318 |
"step": 430
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.010641900062883954,
|
| 322 |
+
"grad_norm": 0.5220174193382263,
|
| 323 |
"learning_rate": 2.195e-05,
|
| 324 |
"loss": 1.6074,
|
| 325 |
"step": 440
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.0108837614279495,
|
| 329 |
+
"grad_norm": 0.5288923978805542,
|
| 330 |
"learning_rate": 2.245e-05,
|
| 331 |
"loss": 1.6033,
|
| 332 |
"step": 450
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.011125622793015044,
|
| 336 |
+
"grad_norm": 0.5983248353004456,
|
| 337 |
"learning_rate": 2.2950000000000002e-05,
|
| 338 |
"loss": 1.5976,
|
| 339 |
"step": 460
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.011367484158080587,
|
| 343 |
+
"grad_norm": 0.5582709312438965,
|
| 344 |
"learning_rate": 2.345e-05,
|
| 345 |
"loss": 1.5982,
|
| 346 |
"step": 470
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.011609345523146132,
|
| 350 |
+
"grad_norm": 0.43671730160713196,
|
| 351 |
"learning_rate": 2.395e-05,
|
| 352 |
"loss": 1.5983,
|
| 353 |
"step": 480
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.011851206888211677,
|
| 357 |
+
"grad_norm": 0.49004241824150085,
|
| 358 |
"learning_rate": 2.445e-05,
|
| 359 |
+
"loss": 1.5916,
|
| 360 |
"step": 490
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.012093068253277222,
|
| 364 |
+
"grad_norm": 0.41542279720306396,
|
| 365 |
"learning_rate": 2.495e-05,
|
| 366 |
"loss": 1.5944,
|
| 367 |
"step": 500
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.012334929618342766,
|
| 371 |
+
"grad_norm": 0.42180851101875305,
|
| 372 |
"learning_rate": 2.5450000000000002e-05,
|
| 373 |
+
"loss": 1.5958,
|
| 374 |
"step": 510
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.01257679098340831,
|
| 378 |
+
"grad_norm": 0.5838198661804199,
|
| 379 |
"learning_rate": 2.595e-05,
|
| 380 |
"loss": 1.5945,
|
| 381 |
"step": 520
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.012818652348473855,
|
| 385 |
+
"grad_norm": 0.4461694359779358,
|
| 386 |
"learning_rate": 2.6450000000000003e-05,
|
| 387 |
+
"loss": 1.5975,
|
| 388 |
"step": 530
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.013060513713539399,
|
| 392 |
+
"grad_norm": 0.445316344499588,
|
| 393 |
"learning_rate": 2.6950000000000005e-05,
|
| 394 |
"loss": 1.586,
|
| 395 |
"step": 540
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.013302375078604944,
|
| 399 |
+
"grad_norm": 0.3972742557525635,
|
| 400 |
"learning_rate": 2.7450000000000003e-05,
|
| 401 |
"loss": 1.591,
|
| 402 |
"step": 550
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.013544236443670489,
|
| 406 |
+
"grad_norm": 0.4089615046977997,
|
| 407 |
"learning_rate": 2.7950000000000005e-05,
|
| 408 |
+
"loss": 1.5944,
|
| 409 |
"step": 560
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.013786097808736032,
|
| 413 |
+
"grad_norm": 0.42213934659957886,
|
| 414 |
"learning_rate": 2.845e-05,
|
| 415 |
+
"loss": 1.5924,
|
| 416 |
"step": 570
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.014027959173801577,
|
| 420 |
+
"grad_norm": 0.5217621326446533,
|
| 421 |
"learning_rate": 2.895e-05,
|
| 422 |
+
"loss": 1.5837,
|
| 423 |
"step": 580
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.014269820538867122,
|
| 427 |
+
"grad_norm": 0.601890504360199,
|
| 428 |
"learning_rate": 2.945e-05,
|
| 429 |
+
"loss": 1.5856,
|
| 430 |
"step": 590
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.014511681903932665,
|
| 434 |
+
"grad_norm": 0.4616837501525879,
|
| 435 |
"learning_rate": 2.995e-05,
|
| 436 |
"loss": 1.5796,
|
| 437 |
"step": 600
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.01475354326899821,
|
| 441 |
+
"grad_norm": 0.5522668957710266,
|
| 442 |
"learning_rate": 3.045e-05,
|
| 443 |
+
"loss": 1.5768,
|
| 444 |
"step": 610
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.014995404634063755,
|
| 448 |
+
"grad_norm": 0.5148155093193054,
|
| 449 |
"learning_rate": 3.095e-05,
|
| 450 |
+
"loss": 1.5797,
|
| 451 |
"step": 620
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.015237265999129298,
|
| 455 |
+
"grad_norm": 0.444976270198822,
|
| 456 |
"learning_rate": 3.145e-05,
|
| 457 |
+
"loss": 1.5811,
|
| 458 |
"step": 630
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.015479127364194843,
|
| 462 |
+
"grad_norm": 0.4654428958892822,
|
| 463 |
"learning_rate": 3.1950000000000004e-05,
|
| 464 |
+
"loss": 1.5825,
|
| 465 |
"step": 640
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.015720988729260386,
|
| 469 |
+
"grad_norm": 0.4910729229450226,
|
| 470 |
"learning_rate": 3.245e-05,
|
| 471 |
+
"loss": 1.5678,
|
| 472 |
"step": 650
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.015962850094325933,
|
| 476 |
+
"grad_norm": 0.42272791266441345,
|
| 477 |
"learning_rate": 3.295e-05,
|
| 478 |
+
"loss": 1.5799,
|
| 479 |
"step": 660
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.016204711459391476,
|
| 483 |
+
"grad_norm": 0.5058956146240234,
|
| 484 |
"learning_rate": 3.345000000000001e-05,
|
| 485 |
+
"loss": 1.5821,
|
| 486 |
"step": 670
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.01644657282445702,
|
| 490 |
+
"grad_norm": 0.4395345151424408,
|
| 491 |
"learning_rate": 3.3950000000000005e-05,
|
| 492 |
+
"loss": 1.5787,
|
| 493 |
"step": 680
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.016688434189522566,
|
| 497 |
+
"grad_norm": 0.3926360607147217,
|
| 498 |
"learning_rate": 3.445e-05,
|
| 499 |
+
"loss": 1.5763,
|
| 500 |
"step": 690
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.01693029555458811,
|
| 504 |
+
"grad_norm": 0.3995387554168701,
|
| 505 |
"learning_rate": 3.495e-05,
|
| 506 |
+
"loss": 1.5719,
|
| 507 |
"step": 700
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.017172156919653656,
|
| 511 |
+
"grad_norm": 0.48855945467948914,
|
| 512 |
"learning_rate": 3.545e-05,
|
| 513 |
+
"loss": 1.5754,
|
| 514 |
"step": 710
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.0174140182847192,
|
| 518 |
+
"grad_norm": 0.7298448085784912,
|
| 519 |
"learning_rate": 3.595e-05,
|
| 520 |
+
"loss": 1.5686,
|
| 521 |
"step": 720
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.017655879649784743,
|
| 525 |
+
"grad_norm": 0.5547767281532288,
|
| 526 |
"learning_rate": 3.645e-05,
|
| 527 |
+
"loss": 1.5784,
|
| 528 |
"step": 730
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.01789774101485029,
|
| 532 |
+
"grad_norm": 0.46988898515701294,
|
| 533 |
"learning_rate": 3.6950000000000004e-05,
|
| 534 |
+
"loss": 1.5653,
|
| 535 |
"step": 740
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.018139602379915833,
|
| 539 |
+
"grad_norm": 0.48516085743904114,
|
| 540 |
"learning_rate": 3.745e-05,
|
| 541 |
+
"loss": 1.5718,
|
| 542 |
"step": 750
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.018381463744981376,
|
| 546 |
+
"grad_norm": 0.6289165019989014,
|
| 547 |
"learning_rate": 3.795e-05,
|
| 548 |
+
"loss": 1.5604,
|
| 549 |
"step": 760
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.018623325110046923,
|
| 553 |
+
"grad_norm": 0.5190830826759338,
|
| 554 |
"learning_rate": 3.845e-05,
|
| 555 |
+
"loss": 1.5656,
|
| 556 |
"step": 770
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.018865186475112466,
|
| 560 |
+
"grad_norm": 0.5255008935928345,
|
| 561 |
"learning_rate": 3.8950000000000005e-05,
|
| 562 |
+
"loss": 1.5719,
|
| 563 |
"step": 780
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.01910704784017801,
|
| 567 |
+
"grad_norm": 0.5320749878883362,
|
| 568 |
"learning_rate": 3.9450000000000003e-05,
|
| 569 |
+
"loss": 1.5645,
|
| 570 |
"step": 790
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.019348909205243556,
|
| 574 |
+
"grad_norm": 0.5073422193527222,
|
| 575 |
"learning_rate": 3.995e-05,
|
| 576 |
+
"loss": 1.5757,
|
| 577 |
"step": 800
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.0195907705703091,
|
| 581 |
+
"grad_norm": 0.5081677436828613,
|
| 582 |
"learning_rate": 4.045000000000001e-05,
|
| 583 |
+
"loss": 1.5618,
|
| 584 |
"step": 810
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.019832631935374642,
|
| 588 |
+
"grad_norm": 0.4249745309352875,
|
| 589 |
"learning_rate": 4.095e-05,
|
| 590 |
+
"loss": 1.5565,
|
| 591 |
"step": 820
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.02007449330044019,
|
| 595 |
+
"grad_norm": 0.4423615634441376,
|
| 596 |
"learning_rate": 4.145e-05,
|
| 597 |
+
"loss": 1.5645,
|
| 598 |
"step": 830
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.020316354665505732,
|
| 602 |
+
"grad_norm": 0.5799271464347839,
|
| 603 |
"learning_rate": 4.195e-05,
|
| 604 |
+
"loss": 1.5584,
|
| 605 |
"step": 840
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.020558216030571275,
|
| 609 |
+
"grad_norm": 0.4171762466430664,
|
| 610 |
"learning_rate": 4.245e-05,
|
| 611 |
+
"loss": 1.5527,
|
| 612 |
"step": 850
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.020800077395636822,
|
| 616 |
+
"grad_norm": 0.418618768453598,
|
| 617 |
"learning_rate": 4.295e-05,
|
| 618 |
+
"loss": 1.5571,
|
| 619 |
"step": 860
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.021041938760702365,
|
| 623 |
+
"grad_norm": 0.4215683937072754,
|
| 624 |
"learning_rate": 4.345e-05,
|
| 625 |
+
"loss": 1.5569,
|
| 626 |
"step": 870
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.02128380012576791,
|
| 630 |
+
"grad_norm": 0.4244352877140045,
|
| 631 |
"learning_rate": 4.3950000000000004e-05,
|
| 632 |
+
"loss": 1.5581,
|
| 633 |
"step": 880
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.021525661490833455,
|
| 637 |
+
"grad_norm": 0.5420682430267334,
|
| 638 |
"learning_rate": 4.445e-05,
|
| 639 |
+
"loss": 1.5539,
|
| 640 |
"step": 890
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.021767522855899,
|
| 644 |
+
"grad_norm": 0.37200865149497986,
|
| 645 |
"learning_rate": 4.495e-05,
|
| 646 |
+
"loss": 1.5503,
|
| 647 |
"step": 900
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.02200938422096454,
|
| 651 |
+
"grad_norm": 0.46333831548690796,
|
| 652 |
"learning_rate": 4.545000000000001e-05,
|
| 653 |
+
"loss": 1.5514,
|
| 654 |
"step": 910
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.02225124558603009,
|
| 658 |
+
"grad_norm": 0.4164145290851593,
|
| 659 |
"learning_rate": 4.5950000000000006e-05,
|
| 660 |
+
"loss": 1.5466,
|
| 661 |
"step": 920
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.02249310695109563,
|
| 665 |
+
"grad_norm": 0.3915160596370697,
|
| 666 |
"learning_rate": 4.6450000000000004e-05,
|
| 667 |
+
"loss": 1.5373,
|
| 668 |
"step": 930
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.022734968316161175,
|
| 672 |
+
"grad_norm": 0.5574844479560852,
|
| 673 |
"learning_rate": 4.695e-05,
|
| 674 |
+
"loss": 1.5484,
|
| 675 |
"step": 940
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.02297682968122672,
|
| 679 |
+
"grad_norm": 0.48569104075431824,
|
| 680 |
"learning_rate": 4.745e-05,
|
| 681 |
+
"loss": 1.5494,
|
| 682 |
"step": 950
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.023218691046292265,
|
| 686 |
+
"grad_norm": 0.5068873167037964,
|
| 687 |
"learning_rate": 4.795e-05,
|
| 688 |
+
"loss": 1.5598,
|
| 689 |
"step": 960
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.023460552411357808,
|
| 693 |
+
"grad_norm": 0.680260419845581,
|
| 694 |
"learning_rate": 4.845e-05,
|
| 695 |
+
"loss": 1.5707,
|
| 696 |
"step": 970
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.023702413776423355,
|
| 700 |
+
"grad_norm": 0.41554513573646545,
|
| 701 |
"learning_rate": 4.8950000000000004e-05,
|
| 702 |
+
"loss": 1.5526,
|
| 703 |
"step": 980
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.023944275141488898,
|
| 707 |
+
"grad_norm": 0.4825437366962433,
|
| 708 |
"learning_rate": 4.945e-05,
|
| 709 |
+
"loss": 1.5479,
|
| 710 |
"step": 990
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.024186136506554445,
|
| 714 |
+
"grad_norm": 0.42075395584106445,
|
| 715 |
"learning_rate": 4.995e-05,
|
| 716 |
+
"loss": 1.5491,
|
| 717 |
"step": 1000
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.024186136506554445,
|
| 721 |
+
"eval_loss": 1.5370113849639893,
|
| 722 |
+
"eval_runtime": 1410.8458,
|
| 723 |
+
"eval_sacrebleu": 96.55473159449107,
|
| 724 |
+
"eval_samples_per_second": 71.621,
|
| 725 |
+
"eval_steps_per_second": 0.56,
|
| 726 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 727 |
}
|
| 728 |
],
|
| 729 |
"logging_steps": 10,
|
|
|
|
| 752 |
"attributes": {}
|
| 753 |
}
|
| 754 |
},
|
| 755 |
+
"total_flos": 2.054005594049741e+16,
|
| 756 |
"train_batch_size": 64,
|
| 757 |
"trial_name": null,
|
| 758 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6033
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8597eaaf5d2c011bf1e394385b7665a1fc4a2649d8b9e1d4bf08323419d3cdc
|
| 3 |
size 6033
|