Fanucci
commited on
Training in progress, step 50, checkpoint
Browse files
last-checkpoint/model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4995335576
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7851edd1043360267b2db8577a31064ca1621c04e7df4289b379b9f8a9793999
|
| 3 |
size 4995335576
|
last-checkpoint/model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1857639032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:89a804dd9adf23eb7826763228d6ef8cbd45f3a451d852daf22e2c4f088aba1d
|
| 3 |
size 1857639032
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13706103974
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36f0fb1c7e888518dcdfd67ad8c4327597ddef100f84eb6544a0f16da3ca1898
|
| 3 |
size 13706103974
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3cf9097d4513154245c48236b6ec5137b7ee2a21c9f58f2cba798ea275c6026f
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03bfcb5cd3899a02f7a89e2033f35f63eb1a6773ac4ce6695121020cac9264f0
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": 1,
|
| 3 |
"best_metric": 1.4945952892303467,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 50,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -12,698 +12,53 @@
|
|
| 12 |
{
|
| 13 |
"epoch": 6.246096189881324e-05,
|
| 14 |
"eval_loss": 1.4945952892303467,
|
| 15 |
-
"eval_runtime": 43.
|
| 16 |
-
"eval_samples_per_second": 19.
|
| 17 |
-
"eval_steps_per_second": 19.
|
| 18 |
"step": 1
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 0.0006246096189881324,
|
| 22 |
-
"grad_norm":
|
| 23 |
"learning_rate": 0.045000000000000005,
|
| 24 |
-
"loss":
|
| 25 |
"step": 10
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.0012492192379762648,
|
| 29 |
-
"grad_norm":
|
| 30 |
"learning_rate": 0.04998980482070473,
|
| 31 |
-
"loss":
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.0018738288569643974,
|
| 36 |
-
"grad_norm":
|
| 37 |
"learning_rate": 0.049954572901111285,
|
| 38 |
-
"loss":
|
| 39 |
"step": 30
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.0024984384759525295,
|
| 43 |
-
"grad_norm":
|
| 44 |
"learning_rate": 0.04989421384191499,
|
| 45 |
-
"loss":
|
| 46 |
"step": 40
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.003123048094940662,
|
| 50 |
-
"grad_norm":
|
| 51 |
"learning_rate": 0.04980878841957203,
|
| 52 |
-
"loss":
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.003123048094940662,
|
| 57 |
-
"eval_loss":
|
| 58 |
-
"eval_runtime": 43.
|
| 59 |
-
"eval_samples_per_second": 19.
|
| 60 |
-
"eval_steps_per_second": 19.
|
| 61 |
"step": 50
|
| 62 |
-
},
|
| 63 |
-
{
|
| 64 |
-
"epoch": 0.0037476577139287947,
|
| 65 |
-
"grad_norm": 46.25,
|
| 66 |
-
"learning_rate": 0.049698382650241506,
|
| 67 |
-
"loss": 227.9017,
|
| 68 |
-
"step": 60
|
| 69 |
-
},
|
| 70 |
-
{
|
| 71 |
-
"epoch": 0.004372267332916927,
|
| 72 |
-
"grad_norm": 129.0,
|
| 73 |
-
"learning_rate": 0.04956310770317444,
|
| 74 |
-
"loss": 91.6515,
|
| 75 |
-
"step": 70
|
| 76 |
-
},
|
| 77 |
-
{
|
| 78 |
-
"epoch": 0.004996876951905059,
|
| 79 |
-
"grad_norm": 25.375,
|
| 80 |
-
"learning_rate": 0.04940309978877575,
|
| 81 |
-
"loss": 43.7426,
|
| 82 |
-
"step": 80
|
| 83 |
-
},
|
| 84 |
-
{
|
| 85 |
-
"epoch": 0.005621486570893191,
|
| 86 |
-
"grad_norm": 26.125,
|
| 87 |
-
"learning_rate": 0.04921852002145197,
|
| 88 |
-
"loss": 35.9591,
|
| 89 |
-
"step": 90
|
| 90 |
-
},
|
| 91 |
-
{
|
| 92 |
-
"epoch": 0.006246096189881324,
|
| 93 |
-
"grad_norm": 71.5,
|
| 94 |
-
"learning_rate": 0.04900955425738262,
|
| 95 |
-
"loss": 25.3901,
|
| 96 |
-
"step": 100
|
| 97 |
-
},
|
| 98 |
-
{
|
| 99 |
-
"epoch": 0.006246096189881324,
|
| 100 |
-
"eval_loss": 21.986961364746094,
|
| 101 |
-
"eval_runtime": 53.4891,
|
| 102 |
-
"eval_samples_per_second": 15.76,
|
| 103 |
-
"eval_steps_per_second": 15.76,
|
| 104 |
-
"step": 100
|
| 105 |
-
},
|
| 106 |
-
{
|
| 107 |
-
"epoch": 0.006870705808869456,
|
| 108 |
-
"grad_norm": 29.75,
|
| 109 |
-
"learning_rate": 0.048776412907378844,
|
| 110 |
-
"loss": 16.5617,
|
| 111 |
-
"step": 110
|
| 112 |
-
},
|
| 113 |
-
{
|
| 114 |
-
"epoch": 0.007495315427857589,
|
| 115 |
-
"grad_norm": 29.75,
|
| 116 |
-
"learning_rate": 0.04851933072501756,
|
| 117 |
-
"loss": 11.0683,
|
| 118 |
-
"step": 120
|
| 119 |
-
},
|
| 120 |
-
{
|
| 121 |
-
"epoch": 0.008119925046845722,
|
| 122 |
-
"grad_norm": 41.5,
|
| 123 |
-
"learning_rate": 0.048238566570264485,
|
| 124 |
-
"loss": 9.7898,
|
| 125 |
-
"step": 130
|
| 126 |
-
},
|
| 127 |
-
{
|
| 128 |
-
"epoch": 0.008744534665833853,
|
| 129 |
-
"grad_norm": 26.75,
|
| 130 |
-
"learning_rate": 0.047934403148824085,
|
| 131 |
-
"loss": 9.437,
|
| 132 |
-
"step": 140
|
| 133 |
-
},
|
| 134 |
-
{
|
| 135 |
-
"epoch": 0.009369144284821987,
|
| 136 |
-
"grad_norm": 14.1875,
|
| 137 |
-
"learning_rate": 0.047607146727478934,
|
| 138 |
-
"loss": 8.8562,
|
| 139 |
-
"step": 150
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"epoch": 0.009369144284821987,
|
| 143 |
-
"eval_loss": 10.050978660583496,
|
| 144 |
-
"eval_runtime": 54.8139,
|
| 145 |
-
"eval_samples_per_second": 15.379,
|
| 146 |
-
"eval_steps_per_second": 15.379,
|
| 147 |
-
"step": 150
|
| 148 |
-
},
|
| 149 |
-
{
|
| 150 |
-
"epoch": 0.009993753903810118,
|
| 151 |
-
"grad_norm": 26.0,
|
| 152 |
-
"learning_rate": 0.04725712682570498,
|
| 153 |
-
"loss": 10.5338,
|
| 154 |
-
"step": 160
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 0.010618363522798251,
|
| 158 |
-
"grad_norm": 28.125,
|
| 159 |
-
"learning_rate": 0.046884695883873395,
|
| 160 |
-
"loss": 9.3137,
|
| 161 |
-
"step": 170
|
| 162 |
-
},
|
| 163 |
-
{
|
| 164 |
-
"epoch": 0.011242973141786383,
|
| 165 |
-
"grad_norm": 37.75,
|
| 166 |
-
"learning_rate": 0.04649022890837298,
|
| 167 |
-
"loss": 9.4597,
|
| 168 |
-
"step": 180
|
| 169 |
-
},
|
| 170 |
-
{
|
| 171 |
-
"epoch": 0.011867582760774516,
|
| 172 |
-
"grad_norm": 46.75,
|
| 173 |
-
"learning_rate": 0.046074123094010544,
|
| 174 |
-
"loss": 9.584,
|
| 175 |
-
"step": 190
|
| 176 |
-
},
|
| 177 |
-
{
|
| 178 |
-
"epoch": 0.012492192379762648,
|
| 179 |
-
"grad_norm": 48.25,
|
| 180 |
-
"learning_rate": 0.04563679742406935,
|
| 181 |
-
"loss": 9.6385,
|
| 182 |
-
"step": 200
|
| 183 |
-
},
|
| 184 |
-
{
|
| 185 |
-
"epoch": 0.012492192379762648,
|
| 186 |
-
"eval_loss": 9.662774085998535,
|
| 187 |
-
"eval_runtime": 53.0984,
|
| 188 |
-
"eval_samples_per_second": 15.876,
|
| 189 |
-
"eval_steps_per_second": 15.876,
|
| 190 |
-
"step": 200
|
| 191 |
-
},
|
| 192 |
-
{
|
| 193 |
-
"epoch": 0.01311680199875078,
|
| 194 |
-
"grad_norm": 100.5,
|
| 195 |
-
"learning_rate": 0.045178692248428534,
|
| 196 |
-
"loss": 10.0078,
|
| 197 |
-
"step": 210
|
| 198 |
-
},
|
| 199 |
-
{
|
| 200 |
-
"epoch": 0.013741411617738912,
|
| 201 |
-
"grad_norm": 51.5,
|
| 202 |
-
"learning_rate": 0.04470026884016805,
|
| 203 |
-
"loss": 9.4955,
|
| 204 |
-
"step": 220
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 0.014366021236727046,
|
| 208 |
-
"grad_norm": 7.71875,
|
| 209 |
-
"learning_rate": 0.0442020089311058,
|
| 210 |
-
"loss": 8.3445,
|
| 211 |
-
"step": 230
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.014990630855715179,
|
| 215 |
-
"grad_norm": 13.0,
|
| 216 |
-
"learning_rate": 0.043684414226734525,
|
| 217 |
-
"loss": 8.7676,
|
| 218 |
-
"step": 240
|
| 219 |
-
},
|
| 220 |
-
{
|
| 221 |
-
"epoch": 0.01561524047470331,
|
| 222 |
-
"grad_norm": 15.625,
|
| 223 |
-
"learning_rate": 0.04314800590104691,
|
| 224 |
-
"loss": 9.141,
|
| 225 |
-
"step": 250
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.01561524047470331,
|
| 229 |
-
"eval_loss": 8.216238975524902,
|
| 230 |
-
"eval_runtime": 55.3897,
|
| 231 |
-
"eval_samples_per_second": 15.219,
|
| 232 |
-
"eval_steps_per_second": 15.219,
|
| 233 |
-
"step": 250
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.016239850093691444,
|
| 237 |
-
"grad_norm": 19.0,
|
| 238 |
-
"learning_rate": 0.04259332407175751,
|
| 239 |
-
"loss": 8.5766,
|
| 240 |
-
"step": 260
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.016864459712679577,
|
| 244 |
-
"grad_norm": 282.0,
|
| 245 |
-
"learning_rate": 0.04202092725645009,
|
| 246 |
-
"loss": 10.2655,
|
| 247 |
-
"step": 270
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.017489069331667707,
|
| 251 |
-
"grad_norm": 7.53125,
|
| 252 |
-
"learning_rate": 0.04143139181019764,
|
| 253 |
-
"loss": 8.9695,
|
| 254 |
-
"step": 280
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.01811367895065584,
|
| 258 |
-
"grad_norm": 31.5,
|
| 259 |
-
"learning_rate": 0.040825311345221764,
|
| 260 |
-
"loss": 8.7729,
|
| 261 |
-
"step": 290
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.018738288569643973,
|
| 265 |
-
"grad_norm": 14.125,
|
| 266 |
-
"learning_rate": 0.04020329613317545,
|
| 267 |
-
"loss": 9.6525,
|
| 268 |
-
"step": 300
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.018738288569643973,
|
| 272 |
-
"eval_loss": 8.672038078308105,
|
| 273 |
-
"eval_runtime": 52.539,
|
| 274 |
-
"eval_samples_per_second": 16.045,
|
| 275 |
-
"eval_steps_per_second": 16.045,
|
| 276 |
-
"step": 300
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.019362898188632106,
|
| 280 |
-
"grad_norm": 15.8125,
|
| 281 |
-
"learning_rate": 0.03956597249065126,
|
| 282 |
-
"loss": 8.3444,
|
| 283 |
-
"step": 310
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.019987507807620236,
|
| 287 |
-
"grad_norm": 26.25,
|
| 288 |
-
"learning_rate": 0.0389139821485336,
|
| 289 |
-
"loss": 8.1302,
|
| 290 |
-
"step": 320
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.02061211742660837,
|
| 294 |
-
"grad_norm": 32.25,
|
| 295 |
-
"learning_rate": 0.03824798160583012,
|
| 296 |
-
"loss": 8.5943,
|
| 297 |
-
"step": 330
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.021236727045596503,
|
| 301 |
-
"grad_norm": 22.375,
|
| 302 |
-
"learning_rate": 0.037568641468632896,
|
| 303 |
-
"loss": 8.2872,
|
| 304 |
-
"step": 340
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.021861336664584636,
|
| 308 |
-
"grad_norm": 35.5,
|
| 309 |
-
"learning_rate": 0.03687664577487488,
|
| 310 |
-
"loss": 9.0739,
|
| 311 |
-
"step": 350
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.021861336664584636,
|
| 315 |
-
"eval_loss": 10.262110710144043,
|
| 316 |
-
"eval_runtime": 53.54,
|
| 317 |
-
"eval_samples_per_second": 15.745,
|
| 318 |
-
"eval_steps_per_second": 15.745,
|
| 319 |
-
"step": 350
|
| 320 |
-
},
|
| 321 |
-
{
|
| 322 |
-
"epoch": 0.022485946283572766,
|
| 323 |
-
"grad_norm": 356.0,
|
| 324 |
-
"learning_rate": 0.03617269130556171,
|
| 325 |
-
"loss": 10.7064,
|
| 326 |
-
"step": 360
|
| 327 |
-
},
|
| 328 |
-
{
|
| 329 |
-
"epoch": 0.0231105559025609,
|
| 330 |
-
"grad_norm": 4512.0,
|
| 331 |
-
"learning_rate": 0.035457486883172316,
|
| 332 |
-
"loss": 8.8088,
|
| 333 |
-
"step": 370
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 0.023735165521549032,
|
| 337 |
-
"grad_norm": 193536.0,
|
| 338 |
-
"learning_rate": 0.03473175265793479,
|
| 339 |
-
"loss": 10.325,
|
| 340 |
-
"step": 380
|
| 341 |
-
},
|
| 342 |
-
{
|
| 343 |
-
"epoch": 0.024359775140537165,
|
| 344 |
-
"grad_norm": 173.0,
|
| 345 |
-
"learning_rate": 0.033996219382696063,
|
| 346 |
-
"loss": 9.4952,
|
| 347 |
-
"step": 390
|
| 348 |
-
},
|
| 349 |
-
{
|
| 350 |
-
"epoch": 0.024984384759525295,
|
| 351 |
-
"grad_norm": 120.0,
|
| 352 |
-
"learning_rate": 0.033251627677115835,
|
| 353 |
-
"loss": 10.3895,
|
| 354 |
-
"step": 400
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 0.024984384759525295,
|
| 358 |
-
"eval_loss": 9.044400215148926,
|
| 359 |
-
"eval_runtime": 55.2547,
|
| 360 |
-
"eval_samples_per_second": 15.257,
|
| 361 |
-
"eval_steps_per_second": 15.257,
|
| 362 |
-
"step": 400
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 0.02560899437851343,
|
| 366 |
-
"grad_norm": 64768.0,
|
| 367 |
-
"learning_rate": 0.032498727281925266,
|
| 368 |
-
"loss": 9.7613,
|
| 369 |
-
"step": 410
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.02623360399750156,
|
| 373 |
-
"grad_norm": 2179072.0,
|
| 374 |
-
"learning_rate": 0.0317382763040017,
|
| 375 |
-
"loss": 9.7948,
|
| 376 |
-
"step": 420
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 0.026858213616489695,
|
| 380 |
-
"grad_norm": 27136.0,
|
| 381 |
-
"learning_rate": 0.030971040453019225,
|
| 382 |
-
"loss": 9.0155,
|
| 383 |
-
"step": 430
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 0.027482823235477825,
|
| 387 |
-
"grad_norm": 71168.0,
|
| 388 |
-
"learning_rate": 0.03019779227044398,
|
| 389 |
-
"loss": 11.3283,
|
| 390 |
-
"step": 440
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 0.028107432854465958,
|
| 394 |
-
"grad_norm": 8768.0,
|
| 395 |
-
"learning_rate": 0.029419310351650393,
|
| 396 |
-
"loss": 8.5037,
|
| 397 |
-
"step": 450
|
| 398 |
-
},
|
| 399 |
-
{
|
| 400 |
-
"epoch": 0.028107432854465958,
|
| 401 |
-
"eval_loss": 8.112160682678223,
|
| 402 |
-
"eval_runtime": 53.9319,
|
| 403 |
-
"eval_samples_per_second": 15.631,
|
| 404 |
-
"eval_steps_per_second": 15.631,
|
| 405 |
-
"step": 450
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 0.02873204247345409,
|
| 409 |
-
"grad_norm": 650117120.0,
|
| 410 |
-
"learning_rate": 0.02863637856194159,
|
| 411 |
-
"loss": 8.5855,
|
| 412 |
-
"step": 460
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 0.029356652092442224,
|
| 416 |
-
"grad_norm": 11008.0,
|
| 417 |
-
"learning_rate": 0.027849785247263517,
|
| 418 |
-
"loss": 8.1377,
|
| 419 |
-
"step": 470
|
| 420 |
-
},
|
| 421 |
-
{
|
| 422 |
-
"epoch": 0.029981261711430358,
|
| 423 |
-
"grad_norm": 2491081031680.0,
|
| 424 |
-
"learning_rate": 0.02706032244040741,
|
| 425 |
-
"loss": 8.1602,
|
| 426 |
-
"step": 480
|
| 427 |
-
},
|
| 428 |
-
{
|
| 429 |
-
"epoch": 0.030605871330418487,
|
| 430 |
-
"grad_norm": 15269888.0,
|
| 431 |
-
"learning_rate": 0.026268785063499858,
|
| 432 |
-
"loss": 7.909,
|
| 433 |
-
"step": 490
|
| 434 |
-
},
|
| 435 |
-
{
|
| 436 |
-
"epoch": 0.03123048094940662,
|
| 437 |
-
"grad_norm": 245366784.0,
|
| 438 |
-
"learning_rate": 0.025475970127583666,
|
| 439 |
-
"loss": 7.9959,
|
| 440 |
-
"step": 500
|
| 441 |
-
},
|
| 442 |
-
{
|
| 443 |
-
"epoch": 0.03123048094940662,
|
| 444 |
-
"eval_loss": 8.136107444763184,
|
| 445 |
-
"eval_runtime": 56.6292,
|
| 446 |
-
"eval_samples_per_second": 14.886,
|
| 447 |
-
"eval_steps_per_second": 14.886,
|
| 448 |
-
"step": 500
|
| 449 |
-
},
|
| 450 |
-
{
|
| 451 |
-
"epoch": 0.03185509056839475,
|
| 452 |
-
"grad_norm": 958464.0,
|
| 453 |
-
"learning_rate": 0.024682675930095266,
|
| 454 |
-
"loss": 8.2639,
|
| 455 |
-
"step": 510
|
| 456 |
-
},
|
| 457 |
-
{
|
| 458 |
-
"epoch": 0.03247970018738289,
|
| 459 |
-
"grad_norm": 2483027968.0,
|
| 460 |
-
"learning_rate": 0.02388970125104685,
|
| 461 |
-
"loss": 8.0817,
|
| 462 |
-
"step": 520
|
| 463 |
-
},
|
| 464 |
-
{
|
| 465 |
-
"epoch": 0.03310430980637102,
|
| 466 |
-
"grad_norm": 2.386907802506363e+17,
|
| 467 |
-
"learning_rate": 0.02309784454872262,
|
| 468 |
-
"loss": 9.0499,
|
| 469 |
-
"step": 530
|
| 470 |
-
},
|
| 471 |
-
{
|
| 472 |
-
"epoch": 0.033728919425359154,
|
| 473 |
-
"grad_norm": 225280.0,
|
| 474 |
-
"learning_rate": 0.022307903155699027,
|
| 475 |
-
"loss": 7.8657,
|
| 476 |
-
"step": 540
|
| 477 |
-
},
|
| 478 |
-
{
|
| 479 |
-
"epoch": 0.034353529044347283,
|
| 480 |
-
"grad_norm": 1171456.0,
|
| 481 |
-
"learning_rate": 0.02152067247599837,
|
| 482 |
-
"loss": 7.8539,
|
| 483 |
-
"step": 550
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 0.034353529044347283,
|
| 487 |
-
"eval_loss": 7.987682819366455,
|
| 488 |
-
"eval_runtime": 56.3279,
|
| 489 |
-
"eval_samples_per_second": 14.966,
|
| 490 |
-
"eval_steps_per_second": 14.966,
|
| 491 |
-
"step": 550
|
| 492 |
-
},
|
| 493 |
-
{
|
| 494 |
-
"epoch": 0.03497813866333541,
|
| 495 |
-
"grad_norm": 83361792.0,
|
| 496 |
-
"learning_rate": 0.020736945184184407,
|
| 497 |
-
"loss": 7.8864,
|
| 498 |
-
"step": 560
|
| 499 |
-
},
|
| 500 |
-
{
|
| 501 |
-
"epoch": 0.03560274828232355,
|
| 502 |
-
"grad_norm": 47710208.0,
|
| 503 |
-
"learning_rate": 0.019957510427206296,
|
| 504 |
-
"loss": 9.0771,
|
| 505 |
-
"step": 570
|
| 506 |
-
},
|
| 507 |
-
{
|
| 508 |
-
"epoch": 0.03622735790131168,
|
| 509 |
-
"grad_norm": 367001600.0,
|
| 510 |
-
"learning_rate": 0.01918315302979444,
|
| 511 |
-
"loss": 18.2783,
|
| 512 |
-
"step": 580
|
| 513 |
-
},
|
| 514 |
-
{
|
| 515 |
-
"epoch": 0.03685196752029981,
|
| 516 |
-
"grad_norm": 19840.0,
|
| 517 |
-
"learning_rate": 0.018414652704208584,
|
| 518 |
-
"loss": 19.6834,
|
| 519 |
-
"step": 590
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 0.037476577139287946,
|
| 523 |
-
"grad_norm": 11392.0,
|
| 524 |
-
"learning_rate": 0.017652783265133608,
|
| 525 |
-
"loss": 13.4558,
|
| 526 |
-
"step": 600
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 0.037476577139287946,
|
| 530 |
-
"eval_loss": 10.698236465454102,
|
| 531 |
-
"eval_runtime": 54.0772,
|
| 532 |
-
"eval_samples_per_second": 15.589,
|
| 533 |
-
"eval_steps_per_second": 15.589,
|
| 534 |
-
"step": 600
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.038101186758276076,
|
| 538 |
-
"grad_norm": 772.0,
|
| 539 |
-
"learning_rate": 0.01689831185051374,
|
| 540 |
-
"loss": 10.3479,
|
| 541 |
-
"step": 610
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.03872579637726421,
|
| 545 |
-
"grad_norm": 30.0,
|
| 546 |
-
"learning_rate": 0.016151998149109708,
|
| 547 |
-
"loss": 9.0419,
|
| 548 |
-
"step": 620
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.03935040599625234,
|
| 552 |
-
"grad_norm": 36.0,
|
| 553 |
-
"learning_rate": 0.015414593635556518,
|
| 554 |
-
"loss": 8.9411,
|
| 555 |
-
"step": 630
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.03997501561524047,
|
| 559 |
-
"grad_norm": 5216.0,
|
| 560 |
-
"learning_rate": 0.014686840813692224,
|
| 561 |
-
"loss": 8.2684,
|
| 562 |
-
"step": 640
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 0.04059962523422861,
|
| 566 |
-
"grad_norm": 12288.0,
|
| 567 |
-
"learning_rate": 0.013969472468919462,
|
| 568 |
-
"loss": 8.5109,
|
| 569 |
-
"step": 650
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.04059962523422861,
|
| 573 |
-
"eval_loss": 8.088459014892578,
|
| 574 |
-
"eval_runtime": 53.555,
|
| 575 |
-
"eval_samples_per_second": 15.741,
|
| 576 |
-
"eval_steps_per_second": 15.741,
|
| 577 |
-
"step": 650
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.04122423485321674,
|
| 581 |
-
"grad_norm": 2672.0,
|
| 582 |
-
"learning_rate": 0.013263210930352737,
|
| 583 |
-
"loss": 7.9423,
|
| 584 |
-
"step": 660
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.04184884447220487,
|
| 588 |
-
"grad_norm": 47185920.0,
|
| 589 |
-
"learning_rate": 0.01256876734349413,
|
| 590 |
-
"loss": 8.6554,
|
| 591 |
-
"step": 670
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.042473454091193005,
|
| 595 |
-
"grad_norm": 31.125,
|
| 596 |
-
"learning_rate": 0.011886840954170141,
|
| 597 |
-
"loss": 8.6544,
|
| 598 |
-
"step": 680
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.043098063710181135,
|
| 602 |
-
"grad_norm": 916.0,
|
| 603 |
-
"learning_rate": 0.011218118404450424,
|
| 604 |
-
"loss": 8.0047,
|
| 605 |
-
"step": 690
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.04372267332916927,
|
| 609 |
-
"grad_norm": 44.25,
|
| 610 |
-
"learning_rate": 0.010563273041257332,
|
| 611 |
-
"loss": 8.3487,
|
| 612 |
-
"step": 700
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.04372267332916927,
|
| 616 |
-
"eval_loss": 8.62705135345459,
|
| 617 |
-
"eval_runtime": 56.6085,
|
| 618 |
-
"eval_samples_per_second": 14.892,
|
| 619 |
-
"eval_steps_per_second": 14.892,
|
| 620 |
-
"step": 700
|
| 621 |
-
},
|
| 622 |
-
{
|
| 623 |
-
"epoch": 0.0443472829481574,
|
| 624 |
-
"grad_norm": 47.5,
|
| 625 |
-
"learning_rate": 0.009922964238362761,
|
| 626 |
-
"loss": 8.9014,
|
| 627 |
-
"step": 710
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 0.04497189256714553,
|
| 631 |
-
"grad_norm": 7840.0,
|
| 632 |
-
"learning_rate": 0.009297836732454564,
|
| 633 |
-
"loss": 7.7725,
|
| 634 |
-
"step": 720
|
| 635 |
-
},
|
| 636 |
-
{
|
| 637 |
-
"epoch": 0.04559650218613367,
|
| 638 |
-
"grad_norm": 4194304.0,
|
| 639 |
-
"learning_rate": 0.0086885199739414,
|
| 640 |
-
"loss": 7.9401,
|
| 641 |
-
"step": 730
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch": 0.0462211118051218,
|
| 645 |
-
"grad_norm": 16.625,
|
| 646 |
-
"learning_rate": 0.00809562749314952,
|
| 647 |
-
"loss": 7.5247,
|
| 648 |
-
"step": 740
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 0.046845721424109935,
|
| 652 |
-
"grad_norm": 151.0,
|
| 653 |
-
"learning_rate": 0.0075197562825497334,
|
| 654 |
-
"loss": 8.3288,
|
| 655 |
-
"step": 750
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 0.046845721424109935,
|
| 659 |
-
"eval_loss": 8.920095443725586,
|
| 660 |
-
"eval_runtime": 56.4634,
|
| 661 |
-
"eval_samples_per_second": 14.93,
|
| 662 |
-
"eval_steps_per_second": 14.93,
|
| 663 |
-
"step": 750
|
| 664 |
-
},
|
| 665 |
-
{
|
| 666 |
-
"epoch": 0.047470331043098064,
|
| 667 |
-
"grad_norm": 12.0625,
|
| 668 |
-
"learning_rate": 0.006961486195636613,
|
| 669 |
-
"loss": 8.1405,
|
| 670 |
-
"step": 760
|
| 671 |
-
},
|
| 672 |
-
{
|
| 673 |
-
"epoch": 0.048094940662086194,
|
| 674 |
-
"grad_norm": 258.0,
|
| 675 |
-
"learning_rate": 0.006421379363065142,
|
| 676 |
-
"loss": 8.3261,
|
| 677 |
-
"step": 770
|
| 678 |
-
},
|
| 679 |
-
{
|
| 680 |
-
"epoch": 0.04871955028107433,
|
| 681 |
-
"grad_norm": 1664.0,
|
| 682 |
-
"learning_rate": 0.005899979626632835,
|
| 683 |
-
"loss": 7.6633,
|
| 684 |
-
"step": 780
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 0.04934415990006246,
|
| 688 |
-
"grad_norm": 12.25,
|
| 689 |
-
"learning_rate": 0.005397811991677107,
|
| 690 |
-
"loss": 7.9004,
|
| 691 |
-
"step": 790
|
| 692 |
-
},
|
| 693 |
-
{
|
| 694 |
-
"epoch": 0.04996876951905059,
|
| 695 |
-
"grad_norm": 23.5,
|
| 696 |
-
"learning_rate": 0.0049153820984394365,
|
| 697 |
-
"loss": 7.9413,
|
| 698 |
-
"step": 800
|
| 699 |
-
},
|
| 700 |
-
{
|
| 701 |
-
"epoch": 0.04996876951905059,
|
| 702 |
-
"eval_loss": 7.851413249969482,
|
| 703 |
-
"eval_runtime": 56.9045,
|
| 704 |
-
"eval_samples_per_second": 14.814,
|
| 705 |
-
"eval_steps_per_second": 14.814,
|
| 706 |
-
"step": 800
|
| 707 |
}
|
| 708 |
],
|
| 709 |
"logging_steps": 10,
|
|
@@ -718,7 +73,7 @@
|
|
| 718 |
"early_stopping_threshold": 0.0
|
| 719 |
},
|
| 720 |
"attributes": {
|
| 721 |
-
"early_stopping_patience_counter":
|
| 722 |
}
|
| 723 |
},
|
| 724 |
"TrainerControl": {
|
|
@@ -732,7 +87,7 @@
|
|
| 732 |
"attributes": {}
|
| 733 |
}
|
| 734 |
},
|
| 735 |
-
"total_flos":
|
| 736 |
"train_batch_size": 1,
|
| 737 |
"trial_name": null,
|
| 738 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": 1,
|
| 3 |
"best_metric": 1.4945952892303467,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.003123048094940662,
|
| 6 |
"eval_steps": 50,
|
| 7 |
+
"global_step": 50,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 12 |
{
|
| 13 |
"epoch": 6.246096189881324e-05,
|
| 14 |
"eval_loss": 1.4945952892303467,
|
| 15 |
+
"eval_runtime": 43.2738,
|
| 16 |
+
"eval_samples_per_second": 19.481,
|
| 17 |
+
"eval_steps_per_second": 19.481,
|
| 18 |
"step": 1
|
| 19 |
},
|
| 20 |
{
|
| 21 |
"epoch": 0.0006246096189881324,
|
| 22 |
+
"grad_norm": 116.0,
|
| 23 |
"learning_rate": 0.045000000000000005,
|
| 24 |
+
"loss": 44.6564,
|
| 25 |
"step": 10
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 0.0012492192379762648,
|
| 29 |
+
"grad_norm": 81.5,
|
| 30 |
"learning_rate": 0.04998980482070473,
|
| 31 |
+
"loss": 266.9058,
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.0018738288569643974,
|
| 36 |
+
"grad_norm": 126.0,
|
| 37 |
"learning_rate": 0.049954572901111285,
|
| 38 |
+
"loss": 349.7276,
|
| 39 |
"step": 30
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.0024984384759525295,
|
| 43 |
+
"grad_norm": 47.25,
|
| 44 |
"learning_rate": 0.04989421384191499,
|
| 45 |
+
"loss": 355.1523,
|
| 46 |
"step": 40
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.003123048094940662,
|
| 50 |
+
"grad_norm": 21.625,
|
| 51 |
"learning_rate": 0.04980878841957203,
|
| 52 |
+
"loss": 271.2299,
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.003123048094940662,
|
| 57 |
+
"eval_loss": 138.12083435058594,
|
| 58 |
+
"eval_runtime": 43.2248,
|
| 59 |
+
"eval_samples_per_second": 19.503,
|
| 60 |
+
"eval_steps_per_second": 19.503,
|
| 61 |
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 62 |
}
|
| 63 |
],
|
| 64 |
"logging_steps": 10,
|
|
|
|
| 73 |
"early_stopping_threshold": 0.0
|
| 74 |
},
|
| 75 |
"attributes": {
|
| 76 |
+
"early_stopping_patience_counter": 1
|
| 77 |
}
|
| 78 |
},
|
| 79 |
"TrainerControl": {
|
|
|
|
| 87 |
"attributes": {}
|
| 88 |
}
|
| 89 |
},
|
| 90 |
+
"total_flos": 2042310819840000.0,
|
| 91 |
"train_batch_size": 1,
|
| 92 |
"trial_name": null,
|
| 93 |
"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 7160
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1e61a95b469ad587100b089121f3916eacaeca053d36c6b400373d16e858628
|
| 3 |
size 7160
|