Training in progress, step 150, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 527048968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90093abcf7bd2cc38fbd8dd4a992d4a8a84765909d6c543ac763634242b4f4e8
|
| 3 |
size 527048968
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1054135994
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c9519724ca8d0709fa71e288b6a83a4f4e4d6df223c2a8f35f742642724da765
|
| 3 |
size 1054135994
|
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 sha256:5feb56512e955691dc9bb9a1e37b9dd590e06a961d7d94560b679e2730b03194
|
| 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 sha256:8cf3f988e8fed2daa2e801eb1f19b681872781cf57f0fb7b896e859a12cfe2bb
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1027 +1,157 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./output/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 150,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.43478260869565216,
|
| 13 |
-
"grad_norm": 1.
|
| 14 |
"learning_rate": 3e-06,
|
| 15 |
"loss": 0.906,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.8695652173913043,
|
| 20 |
-
"grad_norm": 1.
|
| 21 |
"learning_rate": 6e-06,
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 1.3043478260869565,
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
"learning_rate": 9e-06,
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 1.7391304347826086,
|
| 34 |
-
"grad_norm": 1.
|
| 35 |
"learning_rate": 1.2e-05,
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 2.1739130434782608,
|
| 41 |
-
"grad_norm": 1.
|
| 42 |
"learning_rate": 1.5e-05,
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 2.608695652173913,
|
| 48 |
-
"grad_norm": 2.
|
| 49 |
"learning_rate": 1.8e-05,
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 3.0434782608695654,
|
| 55 |
-
"grad_norm": 1.
|
| 56 |
"learning_rate": 2.1e-05,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 3.4782608695652173,
|
| 62 |
-
"grad_norm": 1.
|
| 63 |
"learning_rate": 2.4e-05,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 3.9130434782608696,
|
| 69 |
-
"grad_norm": 1.
|
| 70 |
"learning_rate": 2.7000000000000002e-05,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 4.3478260869565215,
|
| 76 |
-
"grad_norm": 1.
|
| 77 |
"learning_rate": 3e-05,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 4.782608695652174,
|
| 83 |
-
"grad_norm": 2.
|
| 84 |
"learning_rate": 2.999999702723963e-05,
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 5.217391304347826,
|
| 90 |
-
"grad_norm": 1.
|
| 91 |
"learning_rate": 2.9999988108959687e-05,
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 5.6521739130434785,
|
| 97 |
-
"grad_norm": 1.
|
| 98 |
"learning_rate": 2.9999973245163716e-05,
|
| 99 |
"loss": 0.7415,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 6.086956521739131,
|
| 104 |
-
"grad_norm": 1.
|
| 105 |
"learning_rate": 2.99999524358576e-05,
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 6.521739130434782,
|
| 111 |
-
"grad_norm": 1.
|
| 112 |
"learning_rate": 2.9999925681049593e-05,
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 6.521739130434782,
|
| 118 |
-
"eval_loss": 0.
|
| 119 |
"eval_runtime": 0.4775,
|
| 120 |
-
"eval_samples_per_second": 20.
|
| 121 |
-
"eval_steps_per_second": 20.
|
| 122 |
"step": 150
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 6.521739130434782,
|
| 126 |
-
"eval_loss": 0.
|
| 127 |
-
"eval_runtime": 0.
|
| 128 |
-
"eval_samples_per_second":
|
| 129 |
-
"eval_steps_per_second":
|
| 130 |
"step": 150
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 6.521739130434782,
|
| 134 |
-
"eval_loss": 0.
|
| 135 |
-
"eval_runtime": 0.
|
| 136 |
-
"eval_samples_per_second":
|
| 137 |
-
"eval_steps_per_second":
|
| 138 |
"step": 150
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"epoch": 6.521739130434782,
|
| 142 |
-
"eval_loss": 0.
|
| 143 |
-
"eval_runtime": 0.
|
| 144 |
-
"eval_samples_per_second":
|
| 145 |
-
"eval_steps_per_second":
|
| 146 |
"step": 150
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 6.521739130434782,
|
| 150 |
-
"eval_loss": 0.
|
| 151 |
-
"eval_runtime": 0.
|
| 152 |
-
"eval_samples_per_second": 24.
|
| 153 |
-
"eval_steps_per_second": 24.
|
| 154 |
"step": 150
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 6.956521739130435,
|
| 158 |
-
"grad_norm": 1.533442497253418,
|
| 159 |
-
"learning_rate": 2.9999892980750297e-05,
|
| 160 |
-
"loss": 0.6585,
|
| 161 |
-
"step": 160
|
| 162 |
-
},
|
| 163 |
-
{
|
| 164 |
-
"epoch": 7.391304347826087,
|
| 165 |
-
"grad_norm": 1.3458119630813599,
|
| 166 |
-
"learning_rate": 2.9999854334972675e-05,
|
| 167 |
-
"loss": 0.7387,
|
| 168 |
-
"step": 170
|
| 169 |
-
},
|
| 170 |
-
{
|
| 171 |
-
"epoch": 7.826086956521739,
|
| 172 |
-
"grad_norm": 1.726345419883728,
|
| 173 |
-
"learning_rate": 2.999980974373204e-05,
|
| 174 |
-
"loss": 0.729,
|
| 175 |
-
"step": 180
|
| 176 |
-
},
|
| 177 |
-
{
|
| 178 |
-
"epoch": 8.26086956521739,
|
| 179 |
-
"grad_norm": 1.540044903755188,
|
| 180 |
-
"learning_rate": 2.9999759207046075e-05,
|
| 181 |
-
"loss": 0.6246,
|
| 182 |
-
"step": 190
|
| 183 |
-
},
|
| 184 |
-
{
|
| 185 |
-
"epoch": 8.695652173913043,
|
| 186 |
-
"grad_norm": 1.7418792247772217,
|
| 187 |
-
"learning_rate": 2.9999702724934804e-05,
|
| 188 |
-
"loss": 0.6761,
|
| 189 |
-
"step": 200
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 9.130434782608695,
|
| 193 |
-
"grad_norm": 1.038259506225586,
|
| 194 |
-
"learning_rate": 2.999964029742062e-05,
|
| 195 |
-
"loss": 0.6525,
|
| 196 |
-
"step": 210
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 9.565217391304348,
|
| 200 |
-
"grad_norm": 1.2199773788452148,
|
| 201 |
-
"learning_rate": 2.9999571924528263e-05,
|
| 202 |
-
"loss": 0.5592,
|
| 203 |
-
"step": 220
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 10.0,
|
| 207 |
-
"grad_norm": 1.5277819633483887,
|
| 208 |
-
"learning_rate": 2.9999497606284837e-05,
|
| 209 |
-
"loss": 0.7559,
|
| 210 |
-
"step": 230
|
| 211 |
-
},
|
| 212 |
-
{
|
| 213 |
-
"epoch": 10.434782608695652,
|
| 214 |
-
"grad_norm": 1.4215247631072998,
|
| 215 |
-
"learning_rate": 2.9999417342719796e-05,
|
| 216 |
-
"loss": 0.7117,
|
| 217 |
-
"step": 240
|
| 218 |
-
},
|
| 219 |
-
{
|
| 220 |
-
"epoch": 10.869565217391305,
|
| 221 |
-
"grad_norm": 0.974700927734375,
|
| 222 |
-
"learning_rate": 2.9999331133864956e-05,
|
| 223 |
-
"loss": 0.5895,
|
| 224 |
-
"step": 250
|
| 225 |
-
},
|
| 226 |
-
{
|
| 227 |
-
"epoch": 11.304347826086957,
|
| 228 |
-
"grad_norm": 1.1944469213485718,
|
| 229 |
-
"learning_rate": 2.9999238979754485e-05,
|
| 230 |
-
"loss": 0.6546,
|
| 231 |
-
"step": 260
|
| 232 |
-
},
|
| 233 |
-
{
|
| 234 |
-
"epoch": 11.73913043478261,
|
| 235 |
-
"grad_norm": 1.049609661102295,
|
| 236 |
-
"learning_rate": 2.999914088042492e-05,
|
| 237 |
-
"loss": 0.6475,
|
| 238 |
-
"step": 270
|
| 239 |
-
},
|
| 240 |
-
{
|
| 241 |
-
"epoch": 12.173913043478262,
|
| 242 |
-
"grad_norm": 1.3114821910858154,
|
| 243 |
-
"learning_rate": 2.9999036835915132e-05,
|
| 244 |
-
"loss": 0.5939,
|
| 245 |
-
"step": 280
|
| 246 |
-
},
|
| 247 |
-
{
|
| 248 |
-
"epoch": 12.608695652173914,
|
| 249 |
-
"grad_norm": 1.0786523818969727,
|
| 250 |
-
"learning_rate": 2.9998926846266365e-05,
|
| 251 |
-
"loss": 0.633,
|
| 252 |
-
"step": 290
|
| 253 |
-
},
|
| 254 |
-
{
|
| 255 |
-
"epoch": 13.043478260869565,
|
| 256 |
-
"grad_norm": 1.3863022327423096,
|
| 257 |
-
"learning_rate": 2.9998810911522213e-05,
|
| 258 |
-
"loss": 0.5808,
|
| 259 |
-
"step": 300
|
| 260 |
-
},
|
| 261 |
-
{
|
| 262 |
-
"epoch": 13.043478260869565,
|
| 263 |
-
"eval_loss": 0.7308415770530701,
|
| 264 |
-
"eval_runtime": 0.5396,
|
| 265 |
-
"eval_samples_per_second": 18.534,
|
| 266 |
-
"eval_steps_per_second": 18.534,
|
| 267 |
-
"step": 300
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 13.043478260869565,
|
| 271 |
-
"eval_loss": 0.8609212040901184,
|
| 272 |
-
"eval_runtime": 0.5562,
|
| 273 |
-
"eval_samples_per_second": 17.98,
|
| 274 |
-
"eval_steps_per_second": 17.98,
|
| 275 |
-
"step": 300
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 13.043478260869565,
|
| 279 |
-
"eval_loss": 0.7308415770530701,
|
| 280 |
-
"eval_runtime": 0.4963,
|
| 281 |
-
"eval_samples_per_second": 20.15,
|
| 282 |
-
"eval_steps_per_second": 20.15,
|
| 283 |
-
"step": 300
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 13.043478260869565,
|
| 287 |
-
"eval_loss": 0.7747370004653931,
|
| 288 |
-
"eval_runtime": 0.5127,
|
| 289 |
-
"eval_samples_per_second": 19.506,
|
| 290 |
-
"eval_steps_per_second": 19.506,
|
| 291 |
-
"step": 300
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 13.043478260869565,
|
| 295 |
-
"eval_loss": 0.8607426881790161,
|
| 296 |
-
"eval_runtime": 0.5332,
|
| 297 |
-
"eval_samples_per_second": 18.756,
|
| 298 |
-
"eval_steps_per_second": 18.756,
|
| 299 |
-
"step": 300
|
| 300 |
-
},
|
| 301 |
-
{
|
| 302 |
-
"epoch": 13.478260869565217,
|
| 303 |
-
"grad_norm": 1.779449224472046,
|
| 304 |
-
"learning_rate": 2.9998689031728636e-05,
|
| 305 |
-
"loss": 0.5144,
|
| 306 |
-
"step": 310
|
| 307 |
-
},
|
| 308 |
-
{
|
| 309 |
-
"epoch": 13.91304347826087,
|
| 310 |
-
"grad_norm": 1.5321470499038696,
|
| 311 |
-
"learning_rate": 2.9998561206933938e-05,
|
| 312 |
-
"loss": 0.6497,
|
| 313 |
-
"step": 320
|
| 314 |
-
},
|
| 315 |
-
{
|
| 316 |
-
"epoch": 14.347826086956522,
|
| 317 |
-
"grad_norm": 1.4785079956054688,
|
| 318 |
-
"learning_rate": 2.9998427437188786e-05,
|
| 319 |
-
"loss": 0.5745,
|
| 320 |
-
"step": 330
|
| 321 |
-
},
|
| 322 |
-
{
|
| 323 |
-
"epoch": 14.782608695652174,
|
| 324 |
-
"grad_norm": 1.3100569248199463,
|
| 325 |
-
"learning_rate": 2.99982877225462e-05,
|
| 326 |
-
"loss": 0.6013,
|
| 327 |
-
"step": 340
|
| 328 |
-
},
|
| 329 |
-
{
|
| 330 |
-
"epoch": 15.217391304347826,
|
| 331 |
-
"grad_norm": 0.9780473709106445,
|
| 332 |
-
"learning_rate": 2.9998142063061564e-05,
|
| 333 |
-
"loss": 0.4987,
|
| 334 |
-
"step": 350
|
| 335 |
-
},
|
| 336 |
-
{
|
| 337 |
-
"epoch": 15.652173913043478,
|
| 338 |
-
"grad_norm": 1.6418719291687012,
|
| 339 |
-
"learning_rate": 2.9997990458792603e-05,
|
| 340 |
-
"loss": 0.5628,
|
| 341 |
-
"step": 360
|
| 342 |
-
},
|
| 343 |
-
{
|
| 344 |
-
"epoch": 16.08695652173913,
|
| 345 |
-
"grad_norm": 1.6354929208755493,
|
| 346 |
-
"learning_rate": 2.9997832909799417e-05,
|
| 347 |
-
"loss": 0.6671,
|
| 348 |
-
"step": 370
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 16.52173913043478,
|
| 352 |
-
"grad_norm": 0.9526194930076599,
|
| 353 |
-
"learning_rate": 2.9997669416144452e-05,
|
| 354 |
-
"loss": 0.513,
|
| 355 |
-
"step": 380
|
| 356 |
-
},
|
| 357 |
-
{
|
| 358 |
-
"epoch": 16.956521739130434,
|
| 359 |
-
"grad_norm": 0.9340882897377014,
|
| 360 |
-
"learning_rate": 2.999749997789251e-05,
|
| 361 |
-
"loss": 0.5795,
|
| 362 |
-
"step": 390
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 17.391304347826086,
|
| 366 |
-
"grad_norm": 1.1163101196289062,
|
| 367 |
-
"learning_rate": 2.9997324595110743e-05,
|
| 368 |
-
"loss": 0.518,
|
| 369 |
-
"step": 400
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 17.82608695652174,
|
| 373 |
-
"grad_norm": 1.2847086191177368,
|
| 374 |
-
"learning_rate": 2.9997143267868683e-05,
|
| 375 |
-
"loss": 0.5882,
|
| 376 |
-
"step": 410
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 18.26086956521739,
|
| 380 |
-
"grad_norm": 1.1625791788101196,
|
| 381 |
-
"learning_rate": 2.9996955996238192e-05,
|
| 382 |
-
"loss": 0.5061,
|
| 383 |
-
"step": 420
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 18.695652173913043,
|
| 387 |
-
"grad_norm": 1.2001575231552124,
|
| 388 |
-
"learning_rate": 2.9996762780293503e-05,
|
| 389 |
-
"loss": 0.5314,
|
| 390 |
-
"step": 430
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 19.130434782608695,
|
| 394 |
-
"grad_norm": 1.2133065462112427,
|
| 395 |
-
"learning_rate": 2.9996563620111197e-05,
|
| 396 |
-
"loss": 0.5337,
|
| 397 |
-
"step": 440
|
| 398 |
-
},
|
| 399 |
-
{
|
| 400 |
-
"epoch": 19.565217391304348,
|
| 401 |
-
"grad_norm": 1.4226895570755005,
|
| 402 |
-
"learning_rate": 2.9996358515770218e-05,
|
| 403 |
-
"loss": 0.5677,
|
| 404 |
-
"step": 450
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 19.565217391304348,
|
| 408 |
-
"eval_loss": 0.7167752981185913,
|
| 409 |
-
"eval_runtime": 0.5068,
|
| 410 |
-
"eval_samples_per_second": 19.73,
|
| 411 |
-
"eval_steps_per_second": 19.73,
|
| 412 |
-
"step": 450
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 19.565217391304348,
|
| 416 |
-
"eval_loss": 0.8609212040901184,
|
| 417 |
-
"eval_runtime": 0.4062,
|
| 418 |
-
"eval_samples_per_second": 24.616,
|
| 419 |
-
"eval_steps_per_second": 24.616,
|
| 420 |
-
"step": 450
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 19.565217391304348,
|
| 424 |
-
"eval_loss": 0.7167752981185913,
|
| 425 |
-
"eval_runtime": 0.4329,
|
| 426 |
-
"eval_samples_per_second": 23.098,
|
| 427 |
-
"eval_steps_per_second": 23.098,
|
| 428 |
-
"step": 450
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 19.565217391304348,
|
| 432 |
-
"eval_loss": 0.7563869953155518,
|
| 433 |
-
"eval_runtime": 0.3953,
|
| 434 |
-
"eval_samples_per_second": 25.297,
|
| 435 |
-
"eval_steps_per_second": 25.297,
|
| 436 |
-
"step": 450
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 19.565217391304348,
|
| 440 |
-
"eval_loss": 0.8611674308776855,
|
| 441 |
-
"eval_runtime": 0.3948,
|
| 442 |
-
"eval_samples_per_second": 25.331,
|
| 443 |
-
"eval_steps_per_second": 25.331,
|
| 444 |
-
"step": 450
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 20.0,
|
| 448 |
-
"grad_norm": 2.1191132068634033,
|
| 449 |
-
"learning_rate": 2.9996147467351856e-05,
|
| 450 |
-
"loss": 0.515,
|
| 451 |
-
"step": 460
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 20.434782608695652,
|
| 455 |
-
"grad_norm": 1.2782835960388184,
|
| 456 |
-
"learning_rate": 2.9995930474939773e-05,
|
| 457 |
-
"loss": 0.4784,
|
| 458 |
-
"step": 470
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 20.869565217391305,
|
| 462 |
-
"grad_norm": 1.4754245281219482,
|
| 463 |
-
"learning_rate": 2.9995707538619975e-05,
|
| 464 |
-
"loss": 0.5705,
|
| 465 |
-
"step": 480
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 21.304347826086957,
|
| 469 |
-
"grad_norm": 1.322965383529663,
|
| 470 |
-
"learning_rate": 2.9995478658480822e-05,
|
| 471 |
-
"loss": 0.5162,
|
| 472 |
-
"step": 490
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 21.73913043478261,
|
| 476 |
-
"grad_norm": 1.2421406507492065,
|
| 477 |
-
"learning_rate": 2.9995243834613043e-05,
|
| 478 |
-
"loss": 0.5209,
|
| 479 |
-
"step": 500
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 22.17391304347826,
|
| 483 |
-
"grad_norm": 1.7905986309051514,
|
| 484 |
-
"learning_rate": 2.9995003067109707e-05,
|
| 485 |
-
"loss": 0.4834,
|
| 486 |
-
"step": 510
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 22.608695652173914,
|
| 490 |
-
"grad_norm": 1.5309374332427979,
|
| 491 |
-
"learning_rate": 2.9994756356066246e-05,
|
| 492 |
-
"loss": 0.5617,
|
| 493 |
-
"step": 520
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 23.043478260869566,
|
| 497 |
-
"grad_norm": 1.7404286861419678,
|
| 498 |
-
"learning_rate": 2.999450370158046e-05,
|
| 499 |
-
"loss": 0.4927,
|
| 500 |
-
"step": 530
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 23.47826086956522,
|
| 504 |
-
"grad_norm": 1.3119419813156128,
|
| 505 |
-
"learning_rate": 2.9994245103752478e-05,
|
| 506 |
-
"loss": 0.4384,
|
| 507 |
-
"step": 540
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 23.91304347826087,
|
| 511 |
-
"grad_norm": 1.2318671941757202,
|
| 512 |
-
"learning_rate": 2.999398056268481e-05,
|
| 513 |
-
"loss": 0.5268,
|
| 514 |
-
"step": 550
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 24.347826086956523,
|
| 518 |
-
"grad_norm": 1.4077017307281494,
|
| 519 |
-
"learning_rate": 2.9993710078482306e-05,
|
| 520 |
-
"loss": 0.5206,
|
| 521 |
-
"step": 560
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 24.782608695652176,
|
| 525 |
-
"grad_norm": 0.9560300707817078,
|
| 526 |
-
"learning_rate": 2.9993433651252185e-05,
|
| 527 |
-
"loss": 0.4429,
|
| 528 |
-
"step": 570
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 25.217391304347824,
|
| 532 |
-
"grad_norm": 1.7175792455673218,
|
| 533 |
-
"learning_rate": 2.9993151281104006e-05,
|
| 534 |
-
"loss": 0.5326,
|
| 535 |
-
"step": 580
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 25.652173913043477,
|
| 539 |
-
"grad_norm": 1.1363499164581299,
|
| 540 |
-
"learning_rate": 2.9992862968149695e-05,
|
| 541 |
-
"loss": 0.4734,
|
| 542 |
-
"step": 590
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 26.08695652173913,
|
| 546 |
-
"grad_norm": 1.1709671020507812,
|
| 547 |
-
"learning_rate": 2.9992568712503533e-05,
|
| 548 |
-
"loss": 0.4608,
|
| 549 |
-
"step": 600
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 26.08695652173913,
|
| 553 |
-
"eval_loss": 0.7204815149307251,
|
| 554 |
-
"eval_runtime": 0.4403,
|
| 555 |
-
"eval_samples_per_second": 22.71,
|
| 556 |
-
"eval_steps_per_second": 22.71,
|
| 557 |
-
"step": 600
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 26.08695652173913,
|
| 561 |
-
"eval_loss": 0.8609212040901184,
|
| 562 |
-
"eval_runtime": 0.5019,
|
| 563 |
-
"eval_samples_per_second": 19.926,
|
| 564 |
-
"eval_steps_per_second": 19.926,
|
| 565 |
-
"step": 600
|
| 566 |
-
},
|
| 567 |
-
{
|
| 568 |
-
"epoch": 26.08695652173913,
|
| 569 |
-
"eval_loss": 0.7204815149307251,
|
| 570 |
-
"eval_runtime": 0.5015,
|
| 571 |
-
"eval_samples_per_second": 19.942,
|
| 572 |
-
"eval_steps_per_second": 19.942,
|
| 573 |
-
"step": 600
|
| 574 |
-
},
|
| 575 |
-
{
|
| 576 |
-
"epoch": 26.08695652173913,
|
| 577 |
-
"eval_loss": 0.7344802618026733,
|
| 578 |
-
"eval_runtime": 0.4734,
|
| 579 |
-
"eval_samples_per_second": 21.126,
|
| 580 |
-
"eval_steps_per_second": 21.126,
|
| 581 |
-
"step": 600
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 26.08695652173913,
|
| 585 |
-
"eval_loss": 0.8617879748344421,
|
| 586 |
-
"eval_runtime": 0.4638,
|
| 587 |
-
"eval_samples_per_second": 21.559,
|
| 588 |
-
"eval_steps_per_second": 21.559,
|
| 589 |
-
"step": 600
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"epoch": 26.52173913043478,
|
| 593 |
-
"grad_norm": 1.0739339590072632,
|
| 594 |
-
"learning_rate": 2.171901642542767e-06,
|
| 595 |
-
"loss": 0.5121,
|
| 596 |
-
"step": 610
|
| 597 |
-
},
|
| 598 |
-
{
|
| 599 |
-
"epoch": 26.956521739130434,
|
| 600 |
-
"grad_norm": 1.3133119344711304,
|
| 601 |
-
"learning_rate": 4.343803285085534e-06,
|
| 602 |
-
"loss": 0.4308,
|
| 603 |
-
"step": 620
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"epoch": 27.391304347826086,
|
| 607 |
-
"grad_norm": 1.1951584815979004,
|
| 608 |
-
"learning_rate": 6.5157049276283e-06,
|
| 609 |
-
"loss": 0.4829,
|
| 610 |
-
"step": 630
|
| 611 |
-
},
|
| 612 |
-
{
|
| 613 |
-
"epoch": 27.82608695652174,
|
| 614 |
-
"grad_norm": 1.3615652322769165,
|
| 615 |
-
"learning_rate": 8.687606570171068e-06,
|
| 616 |
-
"loss": 0.5071,
|
| 617 |
-
"step": 640
|
| 618 |
-
},
|
| 619 |
-
{
|
| 620 |
-
"epoch": 28.26086956521739,
|
| 621 |
-
"grad_norm": 1.3934813737869263,
|
| 622 |
-
"learning_rate": 1.0859508212713834e-05,
|
| 623 |
-
"loss": 0.4195,
|
| 624 |
-
"step": 650
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 28.695652173913043,
|
| 628 |
-
"grad_norm": 1.5364168882369995,
|
| 629 |
-
"learning_rate": 1.30314098552566e-05,
|
| 630 |
-
"loss": 0.5198,
|
| 631 |
-
"step": 660
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 29.130434782608695,
|
| 635 |
-
"grad_norm": 1.291927695274353,
|
| 636 |
-
"learning_rate": 1.5203311497799366e-05,
|
| 637 |
-
"loss": 0.4474,
|
| 638 |
-
"step": 670
|
| 639 |
-
},
|
| 640 |
-
{
|
| 641 |
-
"epoch": 29.565217391304348,
|
| 642 |
-
"grad_norm": 1.0491178035736084,
|
| 643 |
-
"learning_rate": 1.7375213140342136e-05,
|
| 644 |
-
"loss": 0.489,
|
| 645 |
-
"step": 680
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"epoch": 30.0,
|
| 649 |
-
"grad_norm": 2.413468360900879,
|
| 650 |
-
"learning_rate": 1.9547114782884902e-05,
|
| 651 |
-
"loss": 0.4297,
|
| 652 |
-
"step": 690
|
| 653 |
-
},
|
| 654 |
-
{
|
| 655 |
-
"epoch": 30.434782608695652,
|
| 656 |
-
"grad_norm": 1.3009700775146484,
|
| 657 |
-
"learning_rate": 2.1719016425427668e-05,
|
| 658 |
-
"loss": 0.498,
|
| 659 |
-
"step": 700
|
| 660 |
-
},
|
| 661 |
-
{
|
| 662 |
-
"epoch": 30.869565217391305,
|
| 663 |
-
"grad_norm": 1.5959917306900024,
|
| 664 |
-
"learning_rate": 2.1719014273246623e-05,
|
| 665 |
-
"loss": 0.4323,
|
| 666 |
-
"step": 710
|
| 667 |
-
},
|
| 668 |
-
{
|
| 669 |
-
"epoch": 31.304347826086957,
|
| 670 |
-
"grad_norm": 0.9005913138389587,
|
| 671 |
-
"learning_rate": 2.171900781670434e-05,
|
| 672 |
-
"loss": 0.4452,
|
| 673 |
-
"step": 720
|
| 674 |
-
},
|
| 675 |
-
{
|
| 676 |
-
"epoch": 31.73913043478261,
|
| 677 |
-
"grad_norm": 1.5817768573760986,
|
| 678 |
-
"learning_rate": 2.1718997055803376e-05,
|
| 679 |
-
"loss": 0.5078,
|
| 680 |
-
"step": 730
|
| 681 |
-
},
|
| 682 |
-
{
|
| 683 |
-
"epoch": 32.17391304347826,
|
| 684 |
-
"grad_norm": 1.16550874710083,
|
| 685 |
-
"learning_rate": 2.1718981990548e-05,
|
| 686 |
-
"loss": 0.4618,
|
| 687 |
-
"step": 740
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 32.608695652173914,
|
| 691 |
-
"grad_norm": 1.7140876054763794,
|
| 692 |
-
"learning_rate": 2.1718962620944183e-05,
|
| 693 |
-
"loss": 0.5004,
|
| 694 |
-
"step": 750
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"epoch": 32.608695652173914,
|
| 698 |
-
"eval_loss": 0.7238079309463501,
|
| 699 |
-
"eval_runtime": 0.3947,
|
| 700 |
-
"eval_samples_per_second": 25.334,
|
| 701 |
-
"eval_steps_per_second": 25.334,
|
| 702 |
-
"step": 750
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 32.608695652173914,
|
| 706 |
-
"eval_loss": 0.8609212040901184,
|
| 707 |
-
"eval_runtime": 0.3993,
|
| 708 |
-
"eval_samples_per_second": 25.042,
|
| 709 |
-
"eval_steps_per_second": 25.042,
|
| 710 |
-
"step": 750
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 32.608695652173914,
|
| 714 |
-
"eval_loss": 0.7238079309463501,
|
| 715 |
-
"eval_runtime": 0.389,
|
| 716 |
-
"eval_samples_per_second": 25.705,
|
| 717 |
-
"eval_steps_per_second": 25.705,
|
| 718 |
-
"step": 750
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 32.608695652173914,
|
| 722 |
-
"eval_loss": 0.729554295539856,
|
| 723 |
-
"eval_runtime": 0.3906,
|
| 724 |
-
"eval_samples_per_second": 25.605,
|
| 725 |
-
"eval_steps_per_second": 25.605,
|
| 726 |
-
"step": 750
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 32.608695652173914,
|
| 730 |
-
"eval_loss": 0.8608381152153015,
|
| 731 |
-
"eval_runtime": 0.391,
|
| 732 |
-
"eval_samples_per_second": 25.574,
|
| 733 |
-
"eval_steps_per_second": 25.574,
|
| 734 |
-
"step": 750
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 33.04347826086956,
|
| 738 |
-
"grad_norm": 1.426628589630127,
|
| 739 |
-
"learning_rate": 1.5787846818616295e-06,
|
| 740 |
-
"loss": 0.4544,
|
| 741 |
-
"step": 760
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 33.47826086956522,
|
| 745 |
-
"grad_norm": 1.549428105354309,
|
| 746 |
-
"learning_rate": 3.157569363723259e-06,
|
| 747 |
-
"loss": 0.461,
|
| 748 |
-
"step": 770
|
| 749 |
-
},
|
| 750 |
-
{
|
| 751 |
-
"epoch": 33.91304347826087,
|
| 752 |
-
"grad_norm": 1.2880816459655762,
|
| 753 |
-
"learning_rate": 4.736354045584888e-06,
|
| 754 |
-
"loss": 0.4361,
|
| 755 |
-
"step": 780
|
| 756 |
-
},
|
| 757 |
-
{
|
| 758 |
-
"epoch": 34.34782608695652,
|
| 759 |
-
"grad_norm": 1.3904523849487305,
|
| 760 |
-
"learning_rate": 6.315138727446518e-06,
|
| 761 |
-
"loss": 0.4367,
|
| 762 |
-
"step": 790
|
| 763 |
-
},
|
| 764 |
-
{
|
| 765 |
-
"epoch": 34.78260869565217,
|
| 766 |
-
"grad_norm": 1.4110171794891357,
|
| 767 |
-
"learning_rate": 7.893923409308147e-06,
|
| 768 |
-
"loss": 0.4146,
|
| 769 |
-
"step": 800
|
| 770 |
-
},
|
| 771 |
-
{
|
| 772 |
-
"epoch": 35.21739130434783,
|
| 773 |
-
"grad_norm": 1.2572520971298218,
|
| 774 |
-
"learning_rate": 9.472708091169777e-06,
|
| 775 |
-
"loss": 0.515,
|
| 776 |
-
"step": 810
|
| 777 |
-
},
|
| 778 |
-
{
|
| 779 |
-
"epoch": 35.65217391304348,
|
| 780 |
-
"grad_norm": 1.2433037757873535,
|
| 781 |
-
"learning_rate": 1.1051492773031405e-05,
|
| 782 |
-
"loss": 0.4429,
|
| 783 |
-
"step": 820
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"epoch": 36.08695652173913,
|
| 787 |
-
"grad_norm": 1.068438172340393,
|
| 788 |
-
"learning_rate": 1.2630277454893036e-05,
|
| 789 |
-
"loss": 0.4585,
|
| 790 |
-
"step": 830
|
| 791 |
-
},
|
| 792 |
-
{
|
| 793 |
-
"epoch": 36.52173913043478,
|
| 794 |
-
"grad_norm": 1.1094129085540771,
|
| 795 |
-
"learning_rate": 1.4209062136754664e-05,
|
| 796 |
-
"loss": 0.4247,
|
| 797 |
-
"step": 840
|
| 798 |
-
},
|
| 799 |
-
{
|
| 800 |
-
"epoch": 36.95652173913044,
|
| 801 |
-
"grad_norm": 1.3411928415298462,
|
| 802 |
-
"learning_rate": 1.5787846818616294e-05,
|
| 803 |
-
"loss": 0.4956,
|
| 804 |
-
"step": 850
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 37.391304347826086,
|
| 808 |
-
"grad_norm": 1.8016853332519531,
|
| 809 |
-
"learning_rate": 1.5787845254166782e-05,
|
| 810 |
-
"loss": 0.4679,
|
| 811 |
-
"step": 860
|
| 812 |
-
},
|
| 813 |
-
{
|
| 814 |
-
"epoch": 37.82608695652174,
|
| 815 |
-
"grad_norm": 1.2095096111297607,
|
| 816 |
-
"learning_rate": 1.5787840560818862e-05,
|
| 817 |
-
"loss": 0.4155,
|
| 818 |
-
"step": 870
|
| 819 |
-
},
|
| 820 |
-
{
|
| 821 |
-
"epoch": 38.26086956521739,
|
| 822 |
-
"grad_norm": 1.1868607997894287,
|
| 823 |
-
"learning_rate": 1.5787832738574396e-05,
|
| 824 |
-
"loss": 0.4514,
|
| 825 |
-
"step": 880
|
| 826 |
-
},
|
| 827 |
-
{
|
| 828 |
-
"epoch": 38.69565217391305,
|
| 829 |
-
"grad_norm": 1.377031922340393,
|
| 830 |
-
"learning_rate": 1.5787821787436484e-05,
|
| 831 |
-
"loss": 0.4023,
|
| 832 |
-
"step": 890
|
| 833 |
-
},
|
| 834 |
-
{
|
| 835 |
-
"epoch": 39.130434782608695,
|
| 836 |
-
"grad_norm": 1.88169527053833,
|
| 837 |
-
"learning_rate": 1.578780770740947e-05,
|
| 838 |
-
"loss": 0.5221,
|
| 839 |
-
"step": 900
|
| 840 |
-
},
|
| 841 |
-
{
|
| 842 |
-
"epoch": 39.130434782608695,
|
| 843 |
-
"eval_loss": 0.7288335561752319,
|
| 844 |
-
"eval_runtime": 0.4014,
|
| 845 |
-
"eval_samples_per_second": 24.912,
|
| 846 |
-
"eval_steps_per_second": 24.912,
|
| 847 |
-
"step": 900
|
| 848 |
-
},
|
| 849 |
-
{
|
| 850 |
-
"epoch": 39.130434782608695,
|
| 851 |
-
"eval_loss": 0.8609212040901184,
|
| 852 |
-
"eval_runtime": 0.5203,
|
| 853 |
-
"eval_samples_per_second": 19.219,
|
| 854 |
-
"eval_steps_per_second": 19.219,
|
| 855 |
-
"step": 900
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 39.130434782608695,
|
| 859 |
-
"eval_loss": 0.7288335561752319,
|
| 860 |
-
"eval_runtime": 0.5688,
|
| 861 |
-
"eval_samples_per_second": 17.58,
|
| 862 |
-
"eval_steps_per_second": 17.58,
|
| 863 |
-
"step": 900
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 39.130434782608695,
|
| 867 |
-
"eval_loss": 0.7238188982009888,
|
| 868 |
-
"eval_runtime": 0.4689,
|
| 869 |
-
"eval_samples_per_second": 21.324,
|
| 870 |
-
"eval_steps_per_second": 21.324,
|
| 871 |
-
"step": 900
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 39.130434782608695,
|
| 875 |
-
"eval_loss": 0.861186146736145,
|
| 876 |
-
"eval_runtime": 0.3946,
|
| 877 |
-
"eval_samples_per_second": 25.341,
|
| 878 |
-
"eval_steps_per_second": 25.341,
|
| 879 |
-
"step": 900
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 39.56521739130435,
|
| 883 |
-
"grad_norm": 1.403578758239746,
|
| 884 |
-
"learning_rate": 1.129275730490657e-06,
|
| 885 |
-
"loss": 0.4353,
|
| 886 |
-
"step": 910
|
| 887 |
-
},
|
| 888 |
-
{
|
| 889 |
-
"epoch": 40.0,
|
| 890 |
-
"grad_norm": 2.7664737701416016,
|
| 891 |
-
"learning_rate": 2.258551460981314e-06,
|
| 892 |
-
"loss": 0.4364,
|
| 893 |
-
"step": 920
|
| 894 |
-
},
|
| 895 |
-
{
|
| 896 |
-
"epoch": 40.43478260869565,
|
| 897 |
-
"grad_norm": 1.4606038331985474,
|
| 898 |
-
"learning_rate": 3.3878271914719706e-06,
|
| 899 |
-
"loss": 0.5149,
|
| 900 |
-
"step": 930
|
| 901 |
-
},
|
| 902 |
-
{
|
| 903 |
-
"epoch": 40.869565217391305,
|
| 904 |
-
"grad_norm": 1.4345122575759888,
|
| 905 |
-
"learning_rate": 4.517102921962628e-06,
|
| 906 |
-
"loss": 0.3915,
|
| 907 |
-
"step": 940
|
| 908 |
-
},
|
| 909 |
-
{
|
| 910 |
-
"epoch": 41.30434782608695,
|
| 911 |
-
"grad_norm": 1.9893536567687988,
|
| 912 |
-
"learning_rate": 5.6463786524532845e-06,
|
| 913 |
-
"loss": 0.4036,
|
| 914 |
-
"step": 950
|
| 915 |
-
},
|
| 916 |
-
{
|
| 917 |
-
"epoch": 41.73913043478261,
|
| 918 |
-
"grad_norm": 1.3536087274551392,
|
| 919 |
-
"learning_rate": 6.775654382943941e-06,
|
| 920 |
-
"loss": 0.4227,
|
| 921 |
-
"step": 960
|
| 922 |
-
},
|
| 923 |
-
{
|
| 924 |
-
"epoch": 42.17391304347826,
|
| 925 |
-
"grad_norm": 1.3940178155899048,
|
| 926 |
-
"learning_rate": 7.904930113434598e-06,
|
| 927 |
-
"loss": 0.5038,
|
| 928 |
-
"step": 970
|
| 929 |
-
},
|
| 930 |
-
{
|
| 931 |
-
"epoch": 42.608695652173914,
|
| 932 |
-
"grad_norm": 1.3348439931869507,
|
| 933 |
-
"learning_rate": 9.034205843925256e-06,
|
| 934 |
-
"loss": 0.4424,
|
| 935 |
-
"step": 980
|
| 936 |
-
},
|
| 937 |
-
{
|
| 938 |
-
"epoch": 43.04347826086956,
|
| 939 |
-
"grad_norm": 1.4961334466934204,
|
| 940 |
-
"learning_rate": 1.0163481574415913e-05,
|
| 941 |
-
"loss": 0.4336,
|
| 942 |
-
"step": 990
|
| 943 |
-
},
|
| 944 |
-
{
|
| 945 |
-
"epoch": 43.47826086956522,
|
| 946 |
-
"grad_norm": 1.2589102983474731,
|
| 947 |
-
"learning_rate": 1.1292757304906569e-05,
|
| 948 |
-
"loss": 0.4641,
|
| 949 |
-
"step": 1000
|
| 950 |
-
},
|
| 951 |
-
{
|
| 952 |
-
"epoch": 43.91304347826087,
|
| 953 |
-
"grad_norm": 1.1131060123443604,
|
| 954 |
-
"learning_rate": 1.1292756185884522e-05,
|
| 955 |
-
"loss": 0.4143,
|
| 956 |
-
"step": 1010
|
| 957 |
-
},
|
| 958 |
-
{
|
| 959 |
-
"epoch": 44.34782608695652,
|
| 960 |
-
"grad_norm": 1.8933826684951782,
|
| 961 |
-
"learning_rate": 1.1292752828818823e-05,
|
| 962 |
-
"loss": 0.4697,
|
| 963 |
-
"step": 1020
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 44.78260869565217,
|
| 967 |
-
"grad_norm": 1.4995206594467163,
|
| 968 |
-
"learning_rate": 1.1292747233710805e-05,
|
| 969 |
-
"loss": 0.4582,
|
| 970 |
-
"step": 1030
|
| 971 |
-
},
|
| 972 |
-
{
|
| 973 |
-
"epoch": 45.21739130434783,
|
| 974 |
-
"grad_norm": 1.1720516681671143,
|
| 975 |
-
"learning_rate": 1.1292739400562683e-05,
|
| 976 |
-
"loss": 0.3298,
|
| 977 |
-
"step": 1040
|
| 978 |
-
},
|
| 979 |
-
{
|
| 980 |
-
"epoch": 45.65217391304348,
|
| 981 |
-
"grad_norm": 1.5783675909042358,
|
| 982 |
-
"learning_rate": 1.1292729329377566e-05,
|
| 983 |
-
"loss": 0.4415,
|
| 984 |
-
"step": 1050
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 45.65217391304348,
|
| 988 |
-
"eval_loss": 0.7320327758789062,
|
| 989 |
-
"eval_runtime": 0.5007,
|
| 990 |
-
"eval_samples_per_second": 19.973,
|
| 991 |
-
"eval_steps_per_second": 19.973,
|
| 992 |
-
"step": 1050
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 45.65217391304348,
|
| 996 |
-
"eval_loss": 0.8609212040901184,
|
| 997 |
-
"eval_runtime": 0.425,
|
| 998 |
-
"eval_samples_per_second": 23.532,
|
| 999 |
-
"eval_steps_per_second": 23.532,
|
| 1000 |
-
"step": 1050
|
| 1001 |
-
},
|
| 1002 |
-
{
|
| 1003 |
-
"epoch": 45.65217391304348,
|
| 1004 |
-
"eval_loss": 0.7320327758789062,
|
| 1005 |
-
"eval_runtime": 0.4019,
|
| 1006 |
-
"eval_samples_per_second": 24.883,
|
| 1007 |
-
"eval_steps_per_second": 24.883,
|
| 1008 |
-
"step": 1050
|
| 1009 |
-
},
|
| 1010 |
-
{
|
| 1011 |
-
"epoch": 45.65217391304348,
|
| 1012 |
-
"eval_loss": 0.7223752737045288,
|
| 1013 |
-
"eval_runtime": 0.4083,
|
| 1014 |
-
"eval_samples_per_second": 24.492,
|
| 1015 |
-
"eval_steps_per_second": 24.492,
|
| 1016 |
-
"step": 1050
|
| 1017 |
-
},
|
| 1018 |
-
{
|
| 1019 |
-
"epoch": 45.65217391304348,
|
| 1020 |
-
"eval_loss": 0.8609784245491028,
|
| 1021 |
-
"eval_runtime": 0.4101,
|
| 1022 |
-
"eval_samples_per_second": 24.385,
|
| 1023 |
-
"eval_steps_per_second": 24.385,
|
| 1024 |
-
"step": 1050
|
| 1025 |
}
|
| 1026 |
],
|
| 1027 |
"logging_steps": 10,
|
|
@@ -1041,7 +171,7 @@
|
|
| 1041 |
"attributes": {}
|
| 1042 |
}
|
| 1043 |
},
|
| 1044 |
-
"total_flos":
|
| 1045 |
"train_batch_size": 4,
|
| 1046 |
"trial_name": null,
|
| 1047 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.7960879802703857,
|
| 3 |
+
"best_model_checkpoint": "./output/checkpoint-150",
|
| 4 |
+
"epoch": 6.521739130434782,
|
| 5 |
"eval_steps": 150,
|
| 6 |
+
"global_step": 150,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.43478260869565216,
|
| 13 |
+
"grad_norm": 1.502160668373108,
|
| 14 |
"learning_rate": 3e-06,
|
| 15 |
"loss": 0.906,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.8695652173913043,
|
| 20 |
+
"grad_norm": 1.6870523691177368,
|
| 21 |
"learning_rate": 6e-06,
|
| 22 |
+
"loss": 0.9023,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 1.3043478260869565,
|
| 27 |
+
"grad_norm": 1.7296977043151855,
|
| 28 |
"learning_rate": 9e-06,
|
| 29 |
+
"loss": 0.9002,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 1.7391304347826086,
|
| 34 |
+
"grad_norm": 1.4458173513412476,
|
| 35 |
"learning_rate": 1.2e-05,
|
| 36 |
+
"loss": 0.9095,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 2.1739130434782608,
|
| 41 |
+
"grad_norm": 1.351696252822876,
|
| 42 |
"learning_rate": 1.5e-05,
|
| 43 |
+
"loss": 0.8362,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 2.608695652173913,
|
| 48 |
+
"grad_norm": 2.046602964401245,
|
| 49 |
"learning_rate": 1.8e-05,
|
| 50 |
+
"loss": 0.8897,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 3.0434782608695654,
|
| 55 |
+
"grad_norm": 1.4842596054077148,
|
| 56 |
"learning_rate": 2.1e-05,
|
| 57 |
+
"loss": 0.8916,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 3.4782608695652173,
|
| 62 |
+
"grad_norm": 1.7293957471847534,
|
| 63 |
"learning_rate": 2.4e-05,
|
| 64 |
+
"loss": 0.8233,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 3.9130434782608696,
|
| 69 |
+
"grad_norm": 1.4240052700042725,
|
| 70 |
"learning_rate": 2.7000000000000002e-05,
|
| 71 |
+
"loss": 0.8529,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 4.3478260869565215,
|
| 76 |
+
"grad_norm": 1.3658534288406372,
|
| 77 |
"learning_rate": 3e-05,
|
| 78 |
+
"loss": 0.8646,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 4.782608695652174,
|
| 83 |
+
"grad_norm": 2.2143099308013916,
|
| 84 |
"learning_rate": 2.999999702723963e-05,
|
| 85 |
+
"loss": 0.8225,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 5.217391304347826,
|
| 90 |
+
"grad_norm": 1.0725128650665283,
|
| 91 |
"learning_rate": 2.9999988108959687e-05,
|
| 92 |
+
"loss": 0.7653,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 5.6521739130434785,
|
| 97 |
+
"grad_norm": 1.5600417852401733,
|
| 98 |
"learning_rate": 2.9999973245163716e-05,
|
| 99 |
"loss": 0.7415,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 6.086956521739131,
|
| 104 |
+
"grad_norm": 1.907906413078308,
|
| 105 |
"learning_rate": 2.99999524358576e-05,
|
| 106 |
+
"loss": 0.7656,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 6.521739130434782,
|
| 111 |
+
"grad_norm": 1.121804118156433,
|
| 112 |
"learning_rate": 2.9999925681049593e-05,
|
| 113 |
+
"loss": 0.7858,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 6.521739130434782,
|
| 118 |
+
"eval_loss": 0.7960879802703857,
|
| 119 |
"eval_runtime": 0.4775,
|
| 120 |
+
"eval_samples_per_second": 20.942,
|
| 121 |
+
"eval_steps_per_second": 20.942,
|
| 122 |
"step": 150
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 6.521739130434782,
|
| 126 |
+
"eval_loss": 0.8616224527359009,
|
| 127 |
+
"eval_runtime": 0.3994,
|
| 128 |
+
"eval_samples_per_second": 25.038,
|
| 129 |
+
"eval_steps_per_second": 25.038,
|
| 130 |
"step": 150
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 6.521739130434782,
|
| 134 |
+
"eval_loss": 0.7960879802703857,
|
| 135 |
+
"eval_runtime": 0.3973,
|
| 136 |
+
"eval_samples_per_second": 25.173,
|
| 137 |
+
"eval_steps_per_second": 25.173,
|
| 138 |
"step": 150
|
| 139 |
},
|
| 140 |
{
|
| 141 |
"epoch": 6.521739130434782,
|
| 142 |
+
"eval_loss": 0.8616224527359009,
|
| 143 |
+
"eval_runtime": 0.4031,
|
| 144 |
+
"eval_samples_per_second": 24.805,
|
| 145 |
+
"eval_steps_per_second": 24.805,
|
| 146 |
"step": 150
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 6.521739130434782,
|
| 150 |
+
"eval_loss": 0.8619683384895325,
|
| 151 |
+
"eval_runtime": 0.4079,
|
| 152 |
+
"eval_samples_per_second": 24.517,
|
| 153 |
+
"eval_steps_per_second": 24.517,
|
| 154 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 155 |
}
|
| 156 |
],
|
| 157 |
"logging_steps": 10,
|
|
|
|
| 171 |
"attributes": {}
|
| 172 |
}
|
| 173 |
},
|
| 174 |
+
"total_flos": 3894839614291968.0,
|
| 175 |
"train_batch_size": 4,
|
| 176 |
"trial_name": null,
|
| 177 |
"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 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c27b0feffa55e9783adeb1945da4877c05b7d99c3eb25293fa4481c312fbc7a4
|
| 3 |
size 5368
|