End of training
Browse files- README.md +1 -1
- all_results.json +9 -9
- eval_results.json +4 -4
- train_results.json +5 -5
- trainer_state.json +305 -305
- training_eval_loss.png +0 -0
- training_loss.png +0 -0
README.md
CHANGED
|
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# Qwen2.5-7B-Instruct-PsyCourse-doc-fold6
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.0172
|
| 22 |
|
|
|
|
| 16 |
|
| 17 |
# Qwen2.5-7B-Instruct-PsyCourse-doc-fold6
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the course-doc-train-fold6 dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
- Loss: 0.0172
|
| 22 |
|
all_results.json
CHANGED
|
@@ -1,12 +1,12 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.938271604938271,
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second":
|
| 6 |
-
"eval_steps_per_second":
|
| 7 |
-
"total_flos": 1.
|
| 8 |
-
"train_loss": 0.
|
| 9 |
-
"train_runtime":
|
| 10 |
-
"train_samples_per_second": 0.
|
| 11 |
-
"train_steps_per_second": 0.
|
| 12 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.938271604938271,
|
| 3 |
+
"eval_loss": 0.01722303405404091,
|
| 4 |
+
"eval_runtime": 25.4525,
|
| 5 |
+
"eval_samples_per_second": 1.768,
|
| 6 |
+
"eval_steps_per_second": 1.768,
|
| 7 |
+
"total_flos": 1.8817065711747072e+17,
|
| 8 |
+
"train_loss": 0.03361496642231941,
|
| 9 |
+
"train_runtime": 5228.875,
|
| 10 |
+
"train_samples_per_second": 0.387,
|
| 11 |
+
"train_steps_per_second": 0.024
|
| 12 |
}
|
eval_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.938271604938271,
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second":
|
| 6 |
-
"eval_steps_per_second":
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.938271604938271,
|
| 3 |
+
"eval_loss": 0.01722303405404091,
|
| 4 |
+
"eval_runtime": 25.4525,
|
| 5 |
+
"eval_samples_per_second": 1.768,
|
| 6 |
+
"eval_steps_per_second": 1.768
|
| 7 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 4.938271604938271,
|
| 3 |
-
"total_flos": 1.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second": 0.
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 4.938271604938271,
|
| 3 |
+
"total_flos": 1.8817065711747072e+17,
|
| 4 |
+
"train_loss": 0.03361496642231941,
|
| 5 |
+
"train_runtime": 5228.875,
|
| 6 |
+
"train_samples_per_second": 0.387,
|
| 7 |
+
"train_steps_per_second": 0.024
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
"best_model_checkpoint": "saves/psy-course-doc/Qwen2.5-7B-Instruct/train/fold6/checkpoint-120",
|
| 4 |
"epoch": 4.938271604938271,
|
| 5 |
"eval_steps": 10,
|
|
@@ -10,983 +10,983 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.03950617283950617,
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
"learning_rate": 7.692307692307694e-06,
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.07901234567901234,
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
"learning_rate": 1.5384615384615387e-05,
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.11851851851851852,
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
"learning_rate": 2.307692307692308e-05,
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.1580246913580247,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
"learning_rate": 3.0769230769230774e-05,
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.19753086419753085,
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
"learning_rate": 3.846153846153846e-05,
|
| 43 |
-
"loss": 0.
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.23703703703703705,
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
"learning_rate": 4.615384615384616e-05,
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.2765432098765432,
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
"learning_rate": 5.384615384615385e-05,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.3160493827160494,
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
"learning_rate": 6.153846153846155e-05,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.35555555555555557,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
"learning_rate": 6.923076923076924e-05,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.3950617283950617,
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
"learning_rate": 7.692307692307693e-05,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.3950617283950617,
|
| 83 |
-
"eval_loss": 0.
|
| 84 |
-
"eval_runtime":
|
| 85 |
-
"eval_samples_per_second":
|
| 86 |
-
"eval_steps_per_second":
|
| 87 |
"step": 10
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.4345679012345679,
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
"learning_rate": 8.461538461538461e-05,
|
| 93 |
-
"loss": 0.
|
| 94 |
"step": 11
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.4740740740740741,
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
"learning_rate": 9.230769230769232e-05,
|
| 100 |
-
"loss": 0.
|
| 101 |
"step": 12
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.5135802469135803,
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
"learning_rate": 0.0001,
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 13
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.5530864197530864,
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
"learning_rate": 9.998033131915266e-05,
|
| 114 |
-
"loss": 0.
|
| 115 |
"step": 14
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.5925925925925926,
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
"learning_rate": 9.992134075089084e-05,
|
| 121 |
-
"loss": 0.
|
| 122 |
"step": 15
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.6320987654320988,
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
"learning_rate": 9.982307470588098e-05,
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 16
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.671604938271605,
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
"learning_rate": 9.968561049466214e-05,
|
| 135 |
-
"loss": 0.
|
| 136 |
"step": 17
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.7111111111111111,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
"learning_rate": 9.950905626682228e-05,
|
| 142 |
-
"loss": 0.
|
| 143 |
"step": 18
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.7506172839506173,
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
"learning_rate": 9.92935509259118e-05,
|
| 149 |
-
"loss": 0.
|
| 150 |
"step": 19
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.7901234567901234,
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
"learning_rate": 9.903926402016153e-05,
|
| 156 |
-
"loss": 0.
|
| 157 |
"step": 20
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.7901234567901234,
|
| 161 |
-
"eval_loss": 0.
|
| 162 |
-
"eval_runtime":
|
| 163 |
-
"eval_samples_per_second":
|
| 164 |
-
"eval_steps_per_second":
|
| 165 |
"step": 20
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.8296296296296296,
|
| 169 |
-
"grad_norm": 0.
|
| 170 |
"learning_rate": 9.874639560909117e-05,
|
| 171 |
-
"loss": 0.
|
| 172 |
"step": 21
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.8691358024691358,
|
| 176 |
-
"grad_norm": 0.
|
| 177 |
"learning_rate": 9.841517610611309e-05,
|
| 178 |
-
"loss": 0.
|
| 179 |
"step": 22
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.908641975308642,
|
| 183 |
-
"grad_norm": 0.
|
| 184 |
"learning_rate": 9.804586609725499e-05,
|
| 185 |
-
"loss": 0.
|
| 186 |
"step": 23
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 0.9481481481481482,
|
| 190 |
-
"grad_norm": 0.
|
| 191 |
"learning_rate": 9.763875613614482e-05,
|
| 192 |
-
"loss": 0.
|
| 193 |
"step": 24
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.9876543209876543,
|
| 197 |
-
"grad_norm": 0.
|
| 198 |
"learning_rate": 9.719416651541839e-05,
|
| 199 |
-
"loss": 0.
|
| 200 |
"step": 25
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.0271604938271606,
|
| 204 |
-
"grad_norm": 0.
|
| 205 |
"learning_rate": 9.671244701472999e-05,
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 26
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 1.0666666666666667,
|
| 211 |
-
"grad_norm": 0.
|
| 212 |
"learning_rate": 9.619397662556435e-05,
|
| 213 |
-
"loss": 0.
|
| 214 |
"step": 27
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 1.106172839506173,
|
| 218 |
-
"grad_norm": 0.
|
| 219 |
"learning_rate": 9.563916325306594e-05,
|
| 220 |
-
"loss": 0.
|
| 221 |
"step": 28
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 1.145679012345679,
|
| 225 |
-
"grad_norm": 0.
|
| 226 |
"learning_rate": 9.504844339512095e-05,
|
| 227 |
-
"loss": 0.
|
| 228 |
"step": 29
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 1.1851851851851851,
|
| 232 |
-
"grad_norm": 0.
|
| 233 |
"learning_rate": 9.442228179894362e-05,
|
| 234 |
-
"loss": 0.
|
| 235 |
"step": 30
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 1.1851851851851851,
|
| 239 |
-
"eval_loss": 0.
|
| 240 |
-
"eval_runtime":
|
| 241 |
-
"eval_samples_per_second":
|
| 242 |
-
"eval_steps_per_second":
|
| 243 |
"step": 30
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 1.2246913580246914,
|
| 247 |
-
"grad_norm": 0.
|
| 248 |
"learning_rate": 9.376117109543769e-05,
|
| 249 |
-
"loss": 0.
|
| 250 |
"step": 31
|
| 251 |
},
|
| 252 |
{
|
| 253 |
"epoch": 1.2641975308641975,
|
| 254 |
-
"grad_norm": 0.
|
| 255 |
"learning_rate": 9.306563141162046e-05,
|
| 256 |
-
"loss": 0.
|
| 257 |
"step": 32
|
| 258 |
},
|
| 259 |
{
|
| 260 |
"epoch": 1.3037037037037038,
|
| 261 |
-
"grad_norm": 0.
|
| 262 |
"learning_rate": 9.233620996141421e-05,
|
| 263 |
-
"loss": 0.
|
| 264 |
"step": 33
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 1.34320987654321,
|
| 268 |
-
"grad_norm": 0.
|
| 269 |
"learning_rate": 9.157348061512727e-05,
|
| 270 |
-
"loss": 0.
|
| 271 |
"step": 34
|
| 272 |
},
|
| 273 |
{
|
| 274 |
"epoch": 1.382716049382716,
|
| 275 |
-
"grad_norm": 0.
|
| 276 |
"learning_rate": 9.077804344796302e-05,
|
| 277 |
-
"loss": 0.
|
| 278 |
"step": 35
|
| 279 |
},
|
| 280 |
{
|
| 281 |
"epoch": 1.4222222222222223,
|
| 282 |
-
"grad_norm": 0.
|
| 283 |
"learning_rate": 8.995052426791247e-05,
|
| 284 |
-
"loss": 0.
|
| 285 |
"step": 36
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 1.4617283950617284,
|
| 289 |
-
"grad_norm": 0.
|
| 290 |
"learning_rate": 8.90915741234015e-05,
|
| 291 |
-
"loss": 0.
|
| 292 |
"step": 37
|
| 293 |
},
|
| 294 |
{
|
| 295 |
"epoch": 1.5012345679012347,
|
| 296 |
-
"grad_norm": 0.
|
| 297 |
"learning_rate": 8.820186879108038e-05,
|
| 298 |
-
"loss": 0.
|
| 299 |
"step": 38
|
| 300 |
},
|
| 301 |
{
|
| 302 |
"epoch": 1.5407407407407407,
|
| 303 |
-
"grad_norm": 0.
|
| 304 |
"learning_rate": 8.728210824415827e-05,
|
| 305 |
-
"loss": 0.
|
| 306 |
"step": 39
|
| 307 |
},
|
| 308 |
{
|
| 309 |
"epoch": 1.5802469135802468,
|
| 310 |
-
"grad_norm": 0.
|
| 311 |
"learning_rate": 8.633301610170135e-05,
|
| 312 |
-
"loss": 0.
|
| 313 |
"step": 40
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"epoch": 1.5802469135802468,
|
| 317 |
-
"eval_loss": 0.
|
| 318 |
-
"eval_runtime":
|
| 319 |
-
"eval_samples_per_second":
|
| 320 |
-
"eval_steps_per_second":
|
| 321 |
"step": 40
|
| 322 |
},
|
| 323 |
{
|
| 324 |
"epoch": 1.6197530864197531,
|
| 325 |
-
"grad_norm": 0.
|
| 326 |
"learning_rate": 8.535533905932738e-05,
|
| 327 |
-
"loss": 0.
|
| 328 |
"step": 41
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 1.6592592592592592,
|
| 332 |
-
"grad_norm": 0.
|
| 333 |
"learning_rate": 8.434984630174509e-05,
|
| 334 |
-
"loss": 0.
|
| 335 |
"step": 42
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 1.6987654320987655,
|
| 339 |
-
"grad_norm": 0.
|
| 340 |
"learning_rate": 8.33173288976002e-05,
|
| 341 |
-
"loss": 0.
|
| 342 |
"step": 43
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 1.7382716049382716,
|
| 346 |
-
"grad_norm": 0.
|
| 347 |
"learning_rate": 8.225859917710439e-05,
|
| 348 |
-
"loss": 0.
|
| 349 |
"step": 44
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 1.7777777777777777,
|
| 353 |
-
"grad_norm": 0.
|
| 354 |
"learning_rate": 8.117449009293668e-05,
|
| 355 |
-
"loss": 0.
|
| 356 |
"step": 45
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 1.817283950617284,
|
| 360 |
-
"grad_norm": 0.
|
| 361 |
"learning_rate": 8.006585456492029e-05,
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 46
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 1.8567901234567903,
|
| 367 |
-
"grad_norm": 0.
|
| 368 |
"learning_rate": 7.89335648089903e-05,
|
| 369 |
-
"loss": 0.
|
| 370 |
"step": 47
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 1.8962962962962964,
|
| 374 |
-
"grad_norm": 0.
|
| 375 |
"learning_rate": 7.777851165098012e-05,
|
| 376 |
-
"loss": 0.
|
| 377 |
"step": 48
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 1.9358024691358025,
|
| 381 |
-
"grad_norm": 0.
|
| 382 |
"learning_rate": 7.660160382576683e-05,
|
| 383 |
-
"loss": 0.
|
| 384 |
"step": 49
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 1.9753086419753085,
|
| 388 |
-
"grad_norm": 0.
|
| 389 |
"learning_rate": 7.540376726232648e-05,
|
| 390 |
-
"loss": 0.
|
| 391 |
"step": 50
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 1.9753086419753085,
|
| 395 |
-
"eval_loss": 0.
|
| 396 |
-
"eval_runtime":
|
| 397 |
-
"eval_samples_per_second":
|
| 398 |
-
"eval_steps_per_second":
|
| 399 |
"step": 50
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 2.0148148148148146,
|
| 403 |
-
"grad_norm": 0.
|
| 404 |
"learning_rate": 7.4185944355262e-05,
|
| 405 |
-
"loss": 0.
|
| 406 |
"step": 51
|
| 407 |
},
|
| 408 |
{
|
| 409 |
"epoch": 2.054320987654321,
|
| 410 |
-
"grad_norm": 0.
|
| 411 |
"learning_rate": 7.294909322337689e-05,
|
| 412 |
-
"loss": 0.
|
| 413 |
"step": 52
|
| 414 |
},
|
| 415 |
{
|
| 416 |
"epoch": 2.093827160493827,
|
| 417 |
-
"grad_norm": 0.
|
| 418 |
"learning_rate": 7.169418695587791e-05,
|
| 419 |
-
"loss": 0.
|
| 420 |
"step": 53
|
| 421 |
},
|
| 422 |
{
|
| 423 |
"epoch": 2.1333333333333333,
|
| 424 |
-
"grad_norm": 0.
|
| 425 |
"learning_rate": 7.042221284679982e-05,
|
| 426 |
-
"loss": 0.
|
| 427 |
"step": 54
|
| 428 |
},
|
| 429 |
{
|
| 430 |
"epoch": 2.1728395061728394,
|
| 431 |
-
"grad_norm": 0.
|
| 432 |
"learning_rate": 6.91341716182545e-05,
|
| 433 |
-
"loss": 0.
|
| 434 |
"step": 55
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 2.212345679012346,
|
| 438 |
-
"grad_norm": 0.
|
| 439 |
"learning_rate": 6.783107663311565e-05,
|
| 440 |
-
"loss": 0.
|
| 441 |
"step": 56
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 2.251851851851852,
|
| 445 |
-
"grad_norm": 0.
|
| 446 |
"learning_rate": 6.651395309775837e-05,
|
| 447 |
-
"loss": 0.
|
| 448 |
"step": 57
|
| 449 |
},
|
| 450 |
{
|
| 451 |
"epoch": 2.291358024691358,
|
| 452 |
-
"grad_norm": 0.
|
| 453 |
"learning_rate": 6.518383725548074e-05,
|
| 454 |
-
"loss": 0.
|
| 455 |
"step": 58
|
| 456 |
},
|
| 457 |
{
|
| 458 |
"epoch": 2.330864197530864,
|
| 459 |
-
"grad_norm": 0.
|
| 460 |
"learning_rate": 6.384177557124247e-05,
|
| 461 |
-
"loss": 0.
|
| 462 |
"step": 59
|
| 463 |
},
|
| 464 |
{
|
| 465 |
"epoch": 2.3703703703703702,
|
| 466 |
-
"grad_norm": 0.
|
| 467 |
"learning_rate": 6.248882390836135e-05,
|
| 468 |
-
"loss": 0.
|
| 469 |
"step": 60
|
| 470 |
},
|
| 471 |
{
|
| 472 |
"epoch": 2.3703703703703702,
|
| 473 |
-
"eval_loss": 0.
|
| 474 |
-
"eval_runtime":
|
| 475 |
-
"eval_samples_per_second":
|
| 476 |
-
"eval_steps_per_second":
|
| 477 |
"step": 60
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 2.4098765432098768,
|
| 481 |
-
"grad_norm": 0.
|
| 482 |
"learning_rate": 6.112604669781572e-05,
|
| 483 |
-
"loss": 0.
|
| 484 |
"step": 61
|
| 485 |
},
|
| 486 |
{
|
| 487 |
"epoch": 2.449382716049383,
|
| 488 |
-
"grad_norm": 0.
|
| 489 |
"learning_rate": 5.9754516100806423e-05,
|
| 490 |
-
"loss": 0.
|
| 491 |
"step": 62
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 2.488888888888889,
|
| 495 |
-
"grad_norm": 0.
|
| 496 |
"learning_rate": 5.837531116523682e-05,
|
| 497 |
-
"loss": 0.
|
| 498 |
"step": 63
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 2.528395061728395,
|
| 502 |
-
"grad_norm": 0.
|
| 503 |
"learning_rate": 5.698951697677498e-05,
|
| 504 |
-
"loss": 0.
|
| 505 |
"step": 64
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 2.567901234567901,
|
| 509 |
-
"grad_norm": 0.
|
| 510 |
"learning_rate": 5.559822380516539e-05,
|
| 511 |
-
"loss": 0.
|
| 512 |
"step": 65
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 2.6074074074074076,
|
| 516 |
-
"grad_norm": 0.
|
| 517 |
"learning_rate": 5.420252624646238e-05,
|
| 518 |
-
"loss": 0.
|
| 519 |
"step": 66
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 2.6469135802469137,
|
| 523 |
-
"grad_norm": 0.
|
| 524 |
"learning_rate": 5.2803522361859594e-05,
|
| 525 |
-
"loss": 0.
|
| 526 |
"step": 67
|
| 527 |
},
|
| 528 |
{
|
| 529 |
"epoch": 2.68641975308642,
|
| 530 |
-
"grad_norm": 0.
|
| 531 |
"learning_rate": 5.140231281379345e-05,
|
| 532 |
-
"loss": 0.
|
| 533 |
"step": 68
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 2.725925925925926,
|
| 537 |
-
"grad_norm": 0.
|
| 538 |
"learning_rate": 5e-05,
|
| 539 |
-
"loss": 0.
|
| 540 |
"step": 69
|
| 541 |
},
|
| 542 |
{
|
| 543 |
"epoch": 2.765432098765432,
|
| 544 |
-
"grad_norm": 0.
|
| 545 |
"learning_rate": 4.859768718620656e-05,
|
| 546 |
-
"loss": 0.
|
| 547 |
"step": 70
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 2.765432098765432,
|
| 551 |
-
"eval_loss": 0.
|
| 552 |
-
"eval_runtime":
|
| 553 |
-
"eval_samples_per_second":
|
| 554 |
-
"eval_steps_per_second":
|
| 555 |
"step": 70
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 2.8049382716049385,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
"learning_rate": 4.7196477638140404e-05,
|
| 561 |
-
"loss": 0.
|
| 562 |
"step": 71
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 2.8444444444444446,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
"learning_rate": 4.579747375353763e-05,
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 72
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 2.8839506172839506,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
"learning_rate": 4.4401776194834613e-05,
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 73
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 2.9234567901234567,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 4.3010483023225045e-05,
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 74
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 2.962962962962963,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
"learning_rate": 4.162468883476319e-05,
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 75
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 3.0024691358024693,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
"learning_rate": 4.0245483899193595e-05,
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 76
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 3.0419753086419754,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 3.887395330218429e-05,
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 77
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 3.0814814814814815,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 3.7511176091638653e-05,
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 78
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 3.1209876543209876,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 3.6158224428757535e-05,
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 79
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 3.1604938271604937,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
"learning_rate": 3.4816162744519263e-05,
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 80
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 3.1604938271604937,
|
| 629 |
-
"eval_loss": 0.
|
| 630 |
-
"eval_runtime":
|
| 631 |
-
"eval_samples_per_second":
|
| 632 |
-
"eval_steps_per_second":
|
| 633 |
"step": 80
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 3.2,
|
| 637 |
-
"grad_norm": 0.
|
| 638 |
"learning_rate": 3.3486046902241664e-05,
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 81
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 3.2395061728395063,
|
| 644 |
-
"grad_norm": 0.
|
| 645 |
"learning_rate": 3.216892336688435e-05,
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 82
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 3.2790123456790123,
|
| 651 |
-
"grad_norm": 0.
|
| 652 |
"learning_rate": 3.086582838174551e-05,
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 83
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 3.3185185185185184,
|
| 658 |
-
"grad_norm": 0.
|
| 659 |
"learning_rate": 2.9577787153200197e-05,
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 84
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 3.3580246913580245,
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
"learning_rate": 2.8305813044122097e-05,
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 85
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 3.397530864197531,
|
| 672 |
-
"grad_norm": 0.
|
| 673 |
"learning_rate": 2.705090677662311e-05,
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 86
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 3.437037037037037,
|
| 679 |
-
"grad_norm": 0.
|
| 680 |
"learning_rate": 2.581405564473801e-05,
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 87
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 3.476543209876543,
|
| 686 |
-
"grad_norm": 0.
|
| 687 |
"learning_rate": 2.459623273767354e-05,
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 88
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 3.5160493827160493,
|
| 693 |
-
"grad_norm": 0.
|
| 694 |
"learning_rate": 2.3398396174233178e-05,
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 89
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 3.5555555555555554,
|
| 700 |
-
"grad_norm": 0.
|
| 701 |
"learning_rate": 2.2221488349019903e-05,
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 90
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 3.5555555555555554,
|
| 707 |
-
"eval_loss": 0.
|
| 708 |
-
"eval_runtime":
|
| 709 |
-
"eval_samples_per_second":
|
| 710 |
-
"eval_steps_per_second":
|
| 711 |
"step": 90
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 3.595061728395062,
|
| 715 |
-
"grad_norm": 0.
|
| 716 |
"learning_rate": 2.1066435191009715e-05,
|
| 717 |
-
"loss": 0.
|
| 718 |
"step": 91
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 3.634567901234568,
|
| 722 |
-
"grad_norm": 0.
|
| 723 |
"learning_rate": 1.9934145435079702e-05,
|
| 724 |
-
"loss": 0.
|
| 725 |
"step": 92
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 3.674074074074074,
|
| 729 |
-
"grad_norm": 0.
|
| 730 |
"learning_rate": 1.8825509907063327e-05,
|
| 731 |
-
"loss": 0.
|
| 732 |
"step": 93
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 3.71358024691358,
|
| 736 |
-
"grad_norm": 0.
|
| 737 |
"learning_rate": 1.774140082289563e-05,
|
| 738 |
-
"loss": 0.
|
| 739 |
"step": 94
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 3.753086419753086,
|
| 743 |
-
"grad_norm": 0.
|
| 744 |
"learning_rate": 1.6682671102399805e-05,
|
| 745 |
-
"loss": 0.
|
| 746 |
"step": 95
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 3.7925925925925927,
|
| 750 |
-
"grad_norm": 0.
|
| 751 |
"learning_rate": 1.5650153698254916e-05,
|
| 752 |
-
"loss": 0.
|
| 753 |
"step": 96
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 3.832098765432099,
|
| 757 |
-
"grad_norm": 0.
|
| 758 |
"learning_rate": 1.4644660940672627e-05,
|
| 759 |
-
"loss": 0.
|
| 760 |
"step": 97
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 3.871604938271605,
|
| 764 |
-
"grad_norm": 0.
|
| 765 |
"learning_rate": 1.3666983898298657e-05,
|
| 766 |
-
"loss": 0.
|
| 767 |
"step": 98
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 3.911111111111111,
|
| 771 |
-
"grad_norm": 0.
|
| 772 |
"learning_rate": 1.2717891755841722e-05,
|
| 773 |
-
"loss": 0.
|
| 774 |
"step": 99
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 3.950617283950617,
|
| 778 |
-
"grad_norm": 0.
|
| 779 |
"learning_rate": 1.1798131208919627e-05,
|
| 780 |
-
"loss": 0.
|
| 781 |
"step": 100
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 3.950617283950617,
|
| 785 |
-
"eval_loss": 0.
|
| 786 |
-
"eval_runtime":
|
| 787 |
-
"eval_samples_per_second":
|
| 788 |
-
"eval_steps_per_second":
|
| 789 |
"step": 100
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 3.9901234567901236,
|
| 793 |
-
"grad_norm": 0.
|
| 794 |
"learning_rate": 1.090842587659851e-05,
|
| 795 |
-
"loss": 0.
|
| 796 |
"step": 101
|
| 797 |
},
|
| 798 |
{
|
| 799 |
"epoch": 4.029629629629629,
|
| 800 |
-
"grad_norm": 0.
|
| 801 |
"learning_rate": 1.004947573208756e-05,
|
| 802 |
-
"loss": 0.
|
| 803 |
"step": 102
|
| 804 |
},
|
| 805 |
{
|
| 806 |
"epoch": 4.069135802469136,
|
| 807 |
-
"grad_norm": 0.
|
| 808 |
"learning_rate": 9.221956552036992e-06,
|
| 809 |
-
"loss": 0.
|
| 810 |
"step": 103
|
| 811 |
},
|
| 812 |
{
|
| 813 |
"epoch": 4.108641975308642,
|
| 814 |
-
"grad_norm": 0.
|
| 815 |
"learning_rate": 8.426519384872733e-06,
|
| 816 |
-
"loss": 0.
|
| 817 |
"step": 104
|
| 818 |
},
|
| 819 |
{
|
| 820 |
"epoch": 4.148148148148148,
|
| 821 |
-
"grad_norm": 0.
|
| 822 |
"learning_rate": 7.663790038585793e-06,
|
| 823 |
-
"loss": 0.
|
| 824 |
"step": 105
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 4.187654320987654,
|
| 828 |
-
"grad_norm": 0.
|
| 829 |
"learning_rate": 6.934368588379553e-06,
|
| 830 |
-
"loss": 0.
|
| 831 |
"step": 106
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 4.22716049382716,
|
| 835 |
-
"grad_norm": 0.
|
| 836 |
"learning_rate": 6.238828904562316e-06,
|
| 837 |
-
"loss": 0.
|
| 838 |
"step": 107
|
| 839 |
},
|
| 840 |
{
|
| 841 |
"epoch": 4.266666666666667,
|
| 842 |
-
"grad_norm": 0.
|
| 843 |
"learning_rate": 5.577718201056392e-06,
|
| 844 |
-
"loss": 0.
|
| 845 |
"step": 108
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 4.306172839506173,
|
| 849 |
-
"grad_norm": 0.
|
| 850 |
"learning_rate": 4.951556604879048e-06,
|
| 851 |
-
"loss": 0.
|
| 852 |
"step": 109
|
| 853 |
},
|
| 854 |
{
|
| 855 |
"epoch": 4.345679012345679,
|
| 856 |
-
"grad_norm": 0.
|
| 857 |
"learning_rate": 4.360836746934055e-06,
|
| 858 |
-
"loss": 0.
|
| 859 |
"step": 110
|
| 860 |
},
|
| 861 |
{
|
| 862 |
"epoch": 4.345679012345679,
|
| 863 |
-
"eval_loss": 0.
|
| 864 |
-
"eval_runtime":
|
| 865 |
-
"eval_samples_per_second":
|
| 866 |
-
"eval_steps_per_second":
|
| 867 |
"step": 110
|
| 868 |
},
|
| 869 |
{
|
| 870 |
"epoch": 4.385185185185185,
|
| 871 |
-
"grad_norm": 0.
|
| 872 |
"learning_rate": 3.8060233744356633e-06,
|
| 873 |
-
"loss": 0.
|
| 874 |
"step": 111
|
| 875 |
},
|
| 876 |
{
|
| 877 |
"epoch": 4.424691358024692,
|
| 878 |
-
"grad_norm": 0.
|
| 879 |
"learning_rate": 3.2875529852700147e-06,
|
| 880 |
-
"loss": 0.
|
| 881 |
"step": 112
|
| 882 |
},
|
| 883 |
{
|
| 884 |
"epoch": 4.4641975308641975,
|
| 885 |
-
"grad_norm": 0.
|
| 886 |
"learning_rate": 2.8058334845816213e-06,
|
| 887 |
-
"loss": 0.
|
| 888 |
"step": 113
|
| 889 |
},
|
| 890 |
{
|
| 891 |
"epoch": 4.503703703703704,
|
| 892 |
-
"grad_norm": 0.
|
| 893 |
"learning_rate": 2.361243863855184e-06,
|
| 894 |
-
"loss": 0.
|
| 895 |
"step": 114
|
| 896 |
},
|
| 897 |
{
|
| 898 |
"epoch": 4.54320987654321,
|
| 899 |
-
"grad_norm": 0.
|
| 900 |
"learning_rate": 1.9541339027450256e-06,
|
| 901 |
-
"loss": 0.
|
| 902 |
"step": 115
|
| 903 |
},
|
| 904 |
{
|
| 905 |
"epoch": 4.582716049382716,
|
| 906 |
-
"grad_norm": 0.
|
| 907 |
"learning_rate": 1.584823893886933e-06,
|
| 908 |
-
"loss": 0.
|
| 909 |
"step": 116
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 4.622222222222222,
|
| 913 |
-
"grad_norm": 0.
|
| 914 |
"learning_rate": 1.2536043909088191e-06,
|
| 915 |
-
"loss": 0.
|
| 916 |
"step": 117
|
| 917 |
},
|
| 918 |
{
|
| 919 |
"epoch": 4.661728395061728,
|
| 920 |
-
"grad_norm": 0.
|
| 921 |
"learning_rate": 9.607359798384785e-07,
|
| 922 |
-
"loss": 0.
|
| 923 |
"step": 118
|
| 924 |
},
|
| 925 |
{
|
| 926 |
"epoch": 4.701234567901235,
|
| 927 |
-
"grad_norm": 0.
|
| 928 |
"learning_rate": 7.064490740882057e-07,
|
| 929 |
-
"loss": 0.
|
| 930 |
"step": 119
|
| 931 |
},
|
| 932 |
{
|
| 933 |
"epoch": 4.7407407407407405,
|
| 934 |
-
"grad_norm": 0.
|
| 935 |
"learning_rate": 4.909437331777179e-07,
|
| 936 |
-
"loss": 0.
|
| 937 |
"step": 120
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 4.7407407407407405,
|
| 941 |
-
"eval_loss": 0.
|
| 942 |
-
"eval_runtime":
|
| 943 |
-
"eval_samples_per_second":
|
| 944 |
-
"eval_steps_per_second":
|
| 945 |
"step": 120
|
| 946 |
},
|
| 947 |
{
|
| 948 |
"epoch": 4.780246913580247,
|
| 949 |
-
"grad_norm": 0.
|
| 950 |
"learning_rate": 3.143895053378698e-07,
|
| 951 |
-
"loss": 0.
|
| 952 |
"step": 121
|
| 953 |
},
|
| 954 |
{
|
| 955 |
"epoch": 4.8197530864197535,
|
| 956 |
-
"grad_norm": 0.
|
| 957 |
"learning_rate": 1.7692529411904578e-07,
|
| 958 |
-
"loss": 0.
|
| 959 |
"step": 122
|
| 960 |
},
|
| 961 |
{
|
| 962 |
"epoch": 4.859259259259259,
|
| 963 |
-
"grad_norm": 0.
|
| 964 |
"learning_rate": 7.865924910916977e-08,
|
| 965 |
-
"loss": 0.
|
| 966 |
"step": 123
|
| 967 |
},
|
| 968 |
{
|
| 969 |
"epoch": 4.898765432098766,
|
| 970 |
-
"grad_norm": 0.
|
| 971 |
"learning_rate": 1.9668680847356735e-08,
|
| 972 |
-
"loss": 0.
|
| 973 |
"step": 124
|
| 974 |
},
|
| 975 |
{
|
| 976 |
"epoch": 4.938271604938271,
|
| 977 |
-
"grad_norm": 0.
|
| 978 |
"learning_rate": 0.0,
|
| 979 |
-
"loss": 0.
|
| 980 |
"step": 125
|
| 981 |
},
|
| 982 |
{
|
| 983 |
"epoch": 4.938271604938271,
|
| 984 |
"step": 125,
|
| 985 |
-
"total_flos": 1.
|
| 986 |
-
"train_loss": 0.
|
| 987 |
-
"train_runtime":
|
| 988 |
-
"train_samples_per_second": 0.
|
| 989 |
-
"train_steps_per_second": 0.
|
| 990 |
}
|
| 991 |
],
|
| 992 |
"logging_steps": 1,
|
|
@@ -1006,7 +1006,7 @@
|
|
| 1006 |
"attributes": {}
|
| 1007 |
}
|
| 1008 |
},
|
| 1009 |
-
"total_flos": 1.
|
| 1010 |
"train_batch_size": 1,
|
| 1011 |
"trial_name": null,
|
| 1012 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.01722303405404091,
|
| 3 |
"best_model_checkpoint": "saves/psy-course-doc/Qwen2.5-7B-Instruct/train/fold6/checkpoint-120",
|
| 4 |
"epoch": 4.938271604938271,
|
| 5 |
"eval_steps": 10,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.03950617283950617,
|
| 13 |
+
"grad_norm": 0.21943485736846924,
|
| 14 |
"learning_rate": 7.692307692307694e-06,
|
| 15 |
+
"loss": 0.1756,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.07901234567901234,
|
| 20 |
+
"grad_norm": 0.1591726690530777,
|
| 21 |
"learning_rate": 1.5384615384615387e-05,
|
| 22 |
+
"loss": 0.1363,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.11851851851851852,
|
| 27 |
+
"grad_norm": 0.22766555845737457,
|
| 28 |
"learning_rate": 2.307692307692308e-05,
|
| 29 |
+
"loss": 0.177,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.1580246913580247,
|
| 34 |
+
"grad_norm": 0.15470461547374725,
|
| 35 |
"learning_rate": 3.0769230769230774e-05,
|
| 36 |
+
"loss": 0.1372,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.19753086419753085,
|
| 41 |
+
"grad_norm": 0.14102992415428162,
|
| 42 |
"learning_rate": 3.846153846153846e-05,
|
| 43 |
+
"loss": 0.1216,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.23703703703703705,
|
| 48 |
+
"grad_norm": 0.15640774369239807,
|
| 49 |
"learning_rate": 4.615384615384616e-05,
|
| 50 |
+
"loss": 0.1278,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.2765432098765432,
|
| 55 |
+
"grad_norm": 0.14861662685871124,
|
| 56 |
"learning_rate": 5.384615384615385e-05,
|
| 57 |
+
"loss": 0.1372,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.3160493827160494,
|
| 62 |
+
"grad_norm": 0.12894566357135773,
|
| 63 |
"learning_rate": 6.153846153846155e-05,
|
| 64 |
+
"loss": 0.1184,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.35555555555555557,
|
| 69 |
+
"grad_norm": 0.12572047114372253,
|
| 70 |
"learning_rate": 6.923076923076924e-05,
|
| 71 |
+
"loss": 0.1137,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.3950617283950617,
|
| 76 |
+
"grad_norm": 0.14078351855278015,
|
| 77 |
"learning_rate": 7.692307692307693e-05,
|
| 78 |
+
"loss": 0.1126,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.3950617283950617,
|
| 83 |
+
"eval_loss": 0.12602683901786804,
|
| 84 |
+
"eval_runtime": 25.9346,
|
| 85 |
+
"eval_samples_per_second": 1.735,
|
| 86 |
+
"eval_steps_per_second": 1.735,
|
| 87 |
"step": 10
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.4345679012345679,
|
| 91 |
+
"grad_norm": 0.1199629083275795,
|
| 92 |
"learning_rate": 8.461538461538461e-05,
|
| 93 |
+
"loss": 0.1142,
|
| 94 |
"step": 11
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.4740740740740741,
|
| 98 |
+
"grad_norm": 0.10679922997951508,
|
| 99 |
"learning_rate": 9.230769230769232e-05,
|
| 100 |
+
"loss": 0.0972,
|
| 101 |
"step": 12
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.5135802469135803,
|
| 105 |
+
"grad_norm": 0.09360916912555695,
|
| 106 |
"learning_rate": 0.0001,
|
| 107 |
+
"loss": 0.0876,
|
| 108 |
"step": 13
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.5530864197530864,
|
| 112 |
+
"grad_norm": 0.08126942813396454,
|
| 113 |
"learning_rate": 9.998033131915266e-05,
|
| 114 |
+
"loss": 0.0781,
|
| 115 |
"step": 14
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.5925925925925926,
|
| 119 |
+
"grad_norm": 0.07326602935791016,
|
| 120 |
"learning_rate": 9.992134075089084e-05,
|
| 121 |
+
"loss": 0.0704,
|
| 122 |
"step": 15
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.6320987654320988,
|
| 126 |
+
"grad_norm": 0.0696115791797638,
|
| 127 |
"learning_rate": 9.982307470588098e-05,
|
| 128 |
+
"loss": 0.0635,
|
| 129 |
"step": 16
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.671604938271605,
|
| 133 |
+
"grad_norm": 0.07988032698631287,
|
| 134 |
"learning_rate": 9.968561049466214e-05,
|
| 135 |
+
"loss": 0.0617,
|
| 136 |
"step": 17
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.7111111111111111,
|
| 140 |
+
"grad_norm": 0.09191024303436279,
|
| 141 |
"learning_rate": 9.950905626682228e-05,
|
| 142 |
+
"loss": 0.0637,
|
| 143 |
"step": 18
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.7506172839506173,
|
| 147 |
+
"grad_norm": 0.0852058082818985,
|
| 148 |
"learning_rate": 9.92935509259118e-05,
|
| 149 |
+
"loss": 0.0606,
|
| 150 |
"step": 19
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.7901234567901234,
|
| 154 |
+
"grad_norm": 0.07930584251880646,
|
| 155 |
"learning_rate": 9.903926402016153e-05,
|
| 156 |
+
"loss": 0.0588,
|
| 157 |
"step": 20
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.7901234567901234,
|
| 161 |
+
"eval_loss": 0.05992984399199486,
|
| 162 |
+
"eval_runtime": 25.862,
|
| 163 |
+
"eval_samples_per_second": 1.74,
|
| 164 |
+
"eval_steps_per_second": 1.74,
|
| 165 |
"step": 20
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 0.8296296296296296,
|
| 169 |
+
"grad_norm": 0.08693177998065948,
|
| 170 |
"learning_rate": 9.874639560909117e-05,
|
| 171 |
+
"loss": 0.057,
|
| 172 |
"step": 21
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 0.8691358024691358,
|
| 176 |
+
"grad_norm": 0.06873193383216858,
|
| 177 |
"learning_rate": 9.841517610611309e-05,
|
| 178 |
+
"loss": 0.0549,
|
| 179 |
"step": 22
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 0.908641975308642,
|
| 183 |
+
"grad_norm": 0.05944234877824783,
|
| 184 |
"learning_rate": 9.804586609725499e-05,
|
| 185 |
+
"loss": 0.0391,
|
| 186 |
"step": 23
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 0.9481481481481482,
|
| 190 |
+
"grad_norm": 0.0974101722240448,
|
| 191 |
"learning_rate": 9.763875613614482e-05,
|
| 192 |
+
"loss": 0.0505,
|
| 193 |
"step": 24
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.9876543209876543,
|
| 197 |
+
"grad_norm": 0.06186546012759209,
|
| 198 |
"learning_rate": 9.719416651541839e-05,
|
| 199 |
+
"loss": 0.043,
|
| 200 |
"step": 25
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 1.0271604938271606,
|
| 204 |
+
"grad_norm": 0.09650241583585739,
|
| 205 |
"learning_rate": 9.671244701472999e-05,
|
| 206 |
+
"loss": 0.0638,
|
| 207 |
"step": 26
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 1.0666666666666667,
|
| 211 |
+
"grad_norm": 0.04897386208176613,
|
| 212 |
"learning_rate": 9.619397662556435e-05,
|
| 213 |
+
"loss": 0.0303,
|
| 214 |
"step": 27
|
| 215 |
},
|
| 216 |
{
|
| 217 |
"epoch": 1.106172839506173,
|
| 218 |
+
"grad_norm": 0.046298798173666,
|
| 219 |
"learning_rate": 9.563916325306594e-05,
|
| 220 |
+
"loss": 0.0341,
|
| 221 |
"step": 28
|
| 222 |
},
|
| 223 |
{
|
| 224 |
"epoch": 1.145679012345679,
|
| 225 |
+
"grad_norm": 0.05863014981150627,
|
| 226 |
"learning_rate": 9.504844339512095e-05,
|
| 227 |
+
"loss": 0.0334,
|
| 228 |
"step": 29
|
| 229 |
},
|
| 230 |
{
|
| 231 |
"epoch": 1.1851851851851851,
|
| 232 |
+
"grad_norm": 0.05470599606633186,
|
| 233 |
"learning_rate": 9.442228179894362e-05,
|
| 234 |
+
"loss": 0.0306,
|
| 235 |
"step": 30
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 1.1851851851851851,
|
| 239 |
+
"eval_loss": 0.033436119556427,
|
| 240 |
+
"eval_runtime": 25.8919,
|
| 241 |
+
"eval_samples_per_second": 1.738,
|
| 242 |
+
"eval_steps_per_second": 1.738,
|
| 243 |
"step": 30
|
| 244 |
},
|
| 245 |
{
|
| 246 |
"epoch": 1.2246913580246914,
|
| 247 |
+
"grad_norm": 0.059330571442842484,
|
| 248 |
"learning_rate": 9.376117109543769e-05,
|
| 249 |
+
"loss": 0.0357,
|
| 250 |
"step": 31
|
| 251 |
},
|
| 252 |
{
|
| 253 |
"epoch": 1.2641975308641975,
|
| 254 |
+
"grad_norm": 0.05388804152607918,
|
| 255 |
"learning_rate": 9.306563141162046e-05,
|
| 256 |
+
"loss": 0.028,
|
| 257 |
"step": 32
|
| 258 |
},
|
| 259 |
{
|
| 260 |
"epoch": 1.3037037037037038,
|
| 261 |
+
"grad_norm": 0.04978477954864502,
|
| 262 |
"learning_rate": 9.233620996141421e-05,
|
| 263 |
+
"loss": 0.0268,
|
| 264 |
"step": 33
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 1.34320987654321,
|
| 268 |
+
"grad_norm": 0.05924646928906441,
|
| 269 |
"learning_rate": 9.157348061512727e-05,
|
| 270 |
+
"loss": 0.0267,
|
| 271 |
"step": 34
|
| 272 |
},
|
| 273 |
{
|
| 274 |
"epoch": 1.382716049382716,
|
| 275 |
+
"grad_norm": 0.055464863777160645,
|
| 276 |
"learning_rate": 9.077804344796302e-05,
|
| 277 |
+
"loss": 0.0268,
|
| 278 |
"step": 35
|
| 279 |
},
|
| 280 |
{
|
| 281 |
"epoch": 1.4222222222222223,
|
| 282 |
+
"grad_norm": 0.06895492225885391,
|
| 283 |
"learning_rate": 8.995052426791247e-05,
|
| 284 |
+
"loss": 0.0253,
|
| 285 |
"step": 36
|
| 286 |
},
|
| 287 |
{
|
| 288 |
"epoch": 1.4617283950617284,
|
| 289 |
+
"grad_norm": 0.05984394624829292,
|
| 290 |
"learning_rate": 8.90915741234015e-05,
|
| 291 |
+
"loss": 0.0235,
|
| 292 |
"step": 37
|
| 293 |
},
|
| 294 |
{
|
| 295 |
"epoch": 1.5012345679012347,
|
| 296 |
+
"grad_norm": 0.06565533578395844,
|
| 297 |
"learning_rate": 8.820186879108038e-05,
|
| 298 |
+
"loss": 0.0257,
|
| 299 |
"step": 38
|
| 300 |
},
|
| 301 |
{
|
| 302 |
"epoch": 1.5407407407407407,
|
| 303 |
+
"grad_norm": 0.05269227176904678,
|
| 304 |
"learning_rate": 8.728210824415827e-05,
|
| 305 |
+
"loss": 0.0231,
|
| 306 |
"step": 39
|
| 307 |
},
|
| 308 |
{
|
| 309 |
"epoch": 1.5802469135802468,
|
| 310 |
+
"grad_norm": 0.05268261581659317,
|
| 311 |
"learning_rate": 8.633301610170135e-05,
|
| 312 |
+
"loss": 0.0232,
|
| 313 |
"step": 40
|
| 314 |
},
|
| 315 |
{
|
| 316 |
"epoch": 1.5802469135802468,
|
| 317 |
+
"eval_loss": 0.025375891476869583,
|
| 318 |
+
"eval_runtime": 25.9029,
|
| 319 |
+
"eval_samples_per_second": 1.737,
|
| 320 |
+
"eval_steps_per_second": 1.737,
|
| 321 |
"step": 40
|
| 322 |
},
|
| 323 |
{
|
| 324 |
"epoch": 1.6197530864197531,
|
| 325 |
+
"grad_norm": 0.056495752185583115,
|
| 326 |
"learning_rate": 8.535533905932738e-05,
|
| 327 |
+
"loss": 0.0261,
|
| 328 |
"step": 41
|
| 329 |
},
|
| 330 |
{
|
| 331 |
"epoch": 1.6592592592592592,
|
| 332 |
+
"grad_norm": 0.06019185855984688,
|
| 333 |
"learning_rate": 8.434984630174509e-05,
|
| 334 |
+
"loss": 0.0257,
|
| 335 |
"step": 42
|
| 336 |
},
|
| 337 |
{
|
| 338 |
"epoch": 1.6987654320987655,
|
| 339 |
+
"grad_norm": 0.05118177831172943,
|
| 340 |
"learning_rate": 8.33173288976002e-05,
|
| 341 |
+
"loss": 0.023,
|
| 342 |
"step": 43
|
| 343 |
},
|
| 344 |
{
|
| 345 |
"epoch": 1.7382716049382716,
|
| 346 |
+
"grad_norm": 0.05333544313907623,
|
| 347 |
"learning_rate": 8.225859917710439e-05,
|
| 348 |
+
"loss": 0.0253,
|
| 349 |
"step": 44
|
| 350 |
},
|
| 351 |
{
|
| 352 |
"epoch": 1.7777777777777777,
|
| 353 |
+
"grad_norm": 0.06095046177506447,
|
| 354 |
"learning_rate": 8.117449009293668e-05,
|
| 355 |
+
"loss": 0.0257,
|
| 356 |
"step": 45
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 1.817283950617284,
|
| 360 |
+
"grad_norm": 0.043695252388715744,
|
| 361 |
"learning_rate": 8.006585456492029e-05,
|
| 362 |
+
"loss": 0.0163,
|
| 363 |
"step": 46
|
| 364 |
},
|
| 365 |
{
|
| 366 |
"epoch": 1.8567901234567903,
|
| 367 |
+
"grad_norm": 0.05009619891643524,
|
| 368 |
"learning_rate": 7.89335648089903e-05,
|
| 369 |
+
"loss": 0.0258,
|
| 370 |
"step": 47
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 1.8962962962962964,
|
| 374 |
+
"grad_norm": 0.04014035686850548,
|
| 375 |
"learning_rate": 7.777851165098012e-05,
|
| 376 |
+
"loss": 0.0165,
|
| 377 |
"step": 48
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 1.9358024691358025,
|
| 381 |
+
"grad_norm": 0.053761377930641174,
|
| 382 |
"learning_rate": 7.660160382576683e-05,
|
| 383 |
+
"loss": 0.0254,
|
| 384 |
"step": 49
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"epoch": 1.9753086419753085,
|
| 388 |
+
"grad_norm": 0.05367910489439964,
|
| 389 |
"learning_rate": 7.540376726232648e-05,
|
| 390 |
+
"loss": 0.0189,
|
| 391 |
"step": 50
|
| 392 |
},
|
| 393 |
{
|
| 394 |
"epoch": 1.9753086419753085,
|
| 395 |
+
"eval_loss": 0.021527696400880814,
|
| 396 |
+
"eval_runtime": 25.9007,
|
| 397 |
+
"eval_samples_per_second": 1.737,
|
| 398 |
+
"eval_steps_per_second": 1.737,
|
| 399 |
"step": 50
|
| 400 |
},
|
| 401 |
{
|
| 402 |
"epoch": 2.0148148148148146,
|
| 403 |
+
"grad_norm": 0.0713057592511177,
|
| 404 |
"learning_rate": 7.4185944355262e-05,
|
| 405 |
+
"loss": 0.0288,
|
| 406 |
"step": 51
|
| 407 |
},
|
| 408 |
{
|
| 409 |
"epoch": 2.054320987654321,
|
| 410 |
+
"grad_norm": 0.03315626457333565,
|
| 411 |
"learning_rate": 7.294909322337689e-05,
|
| 412 |
+
"loss": 0.0136,
|
| 413 |
"step": 52
|
| 414 |
},
|
| 415 |
{
|
| 416 |
"epoch": 2.093827160493827,
|
| 417 |
+
"grad_norm": 0.03830026090145111,
|
| 418 |
"learning_rate": 7.169418695587791e-05,
|
| 419 |
+
"loss": 0.0169,
|
| 420 |
"step": 53
|
| 421 |
},
|
| 422 |
{
|
| 423 |
"epoch": 2.1333333333333333,
|
| 424 |
+
"grad_norm": 0.04134984314441681,
|
| 425 |
"learning_rate": 7.042221284679982e-05,
|
| 426 |
+
"loss": 0.0194,
|
| 427 |
"step": 54
|
| 428 |
},
|
| 429 |
{
|
| 430 |
"epoch": 2.1728395061728394,
|
| 431 |
+
"grad_norm": 0.04015449061989784,
|
| 432 |
"learning_rate": 6.91341716182545e-05,
|
| 433 |
+
"loss": 0.0173,
|
| 434 |
"step": 55
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 2.212345679012346,
|
| 438 |
+
"grad_norm": 0.03454513102769852,
|
| 439 |
"learning_rate": 6.783107663311565e-05,
|
| 440 |
+
"loss": 0.0129,
|
| 441 |
"step": 56
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 2.251851851851852,
|
| 445 |
+
"grad_norm": 0.04263680428266525,
|
| 446 |
"learning_rate": 6.651395309775837e-05,
|
| 447 |
+
"loss": 0.0165,
|
| 448 |
"step": 57
|
| 449 |
},
|
| 450 |
{
|
| 451 |
"epoch": 2.291358024691358,
|
| 452 |
+
"grad_norm": 0.04442084953188896,
|
| 453 |
"learning_rate": 6.518383725548074e-05,
|
| 454 |
+
"loss": 0.0221,
|
| 455 |
"step": 58
|
| 456 |
},
|
| 457 |
{
|
| 458 |
"epoch": 2.330864197530864,
|
| 459 |
+
"grad_norm": 0.03828935697674751,
|
| 460 |
"learning_rate": 6.384177557124247e-05,
|
| 461 |
+
"loss": 0.0194,
|
| 462 |
"step": 59
|
| 463 |
},
|
| 464 |
{
|
| 465 |
"epoch": 2.3703703703703702,
|
| 466 |
+
"grad_norm": 0.062412988394498825,
|
| 467 |
"learning_rate": 6.248882390836135e-05,
|
| 468 |
+
"loss": 0.0168,
|
| 469 |
"step": 60
|
| 470 |
},
|
| 471 |
{
|
| 472 |
"epoch": 2.3703703703703702,
|
| 473 |
+
"eval_loss": 0.019772132858633995,
|
| 474 |
+
"eval_runtime": 25.9082,
|
| 475 |
+
"eval_samples_per_second": 1.737,
|
| 476 |
+
"eval_steps_per_second": 1.737,
|
| 477 |
"step": 60
|
| 478 |
},
|
| 479 |
{
|
| 480 |
"epoch": 2.4098765432098768,
|
| 481 |
+
"grad_norm": 0.04680776596069336,
|
| 482 |
"learning_rate": 6.112604669781572e-05,
|
| 483 |
+
"loss": 0.0175,
|
| 484 |
"step": 61
|
| 485 |
},
|
| 486 |
{
|
| 487 |
"epoch": 2.449382716049383,
|
| 488 |
+
"grad_norm": 0.043991029262542725,
|
| 489 |
"learning_rate": 5.9754516100806423e-05,
|
| 490 |
+
"loss": 0.0178,
|
| 491 |
"step": 62
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 2.488888888888889,
|
| 495 |
+
"grad_norm": 0.04611797630786896,
|
| 496 |
"learning_rate": 5.837531116523682e-05,
|
| 497 |
+
"loss": 0.0222,
|
| 498 |
"step": 63
|
| 499 |
},
|
| 500 |
{
|
| 501 |
"epoch": 2.528395061728395,
|
| 502 |
+
"grad_norm": 0.04121099039912224,
|
| 503 |
"learning_rate": 5.698951697677498e-05,
|
| 504 |
+
"loss": 0.0146,
|
| 505 |
"step": 64
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 2.567901234567901,
|
| 509 |
+
"grad_norm": 0.04799938201904297,
|
| 510 |
"learning_rate": 5.559822380516539e-05,
|
| 511 |
+
"loss": 0.0143,
|
| 512 |
"step": 65
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 2.6074074074074076,
|
| 516 |
+
"grad_norm": 0.05085166543722153,
|
| 517 |
"learning_rate": 5.420252624646238e-05,
|
| 518 |
+
"loss": 0.0184,
|
| 519 |
"step": 66
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 2.6469135802469137,
|
| 523 |
+
"grad_norm": 0.04127015918493271,
|
| 524 |
"learning_rate": 5.2803522361859594e-05,
|
| 525 |
+
"loss": 0.0157,
|
| 526 |
"step": 67
|
| 527 |
},
|
| 528 |
{
|
| 529 |
"epoch": 2.68641975308642,
|
| 530 |
+
"grad_norm": 0.04294725880026817,
|
| 531 |
"learning_rate": 5.140231281379345e-05,
|
| 532 |
+
"loss": 0.0169,
|
| 533 |
"step": 68
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 2.725925925925926,
|
| 537 |
+
"grad_norm": 0.06253976374864578,
|
| 538 |
"learning_rate": 5e-05,
|
| 539 |
+
"loss": 0.0194,
|
| 540 |
"step": 69
|
| 541 |
},
|
| 542 |
{
|
| 543 |
"epoch": 2.765432098765432,
|
| 544 |
+
"grad_norm": 0.046594616025686264,
|
| 545 |
"learning_rate": 4.859768718620656e-05,
|
| 546 |
+
"loss": 0.0201,
|
| 547 |
"step": 70
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 2.765432098765432,
|
| 551 |
+
"eval_loss": 0.018415603786706924,
|
| 552 |
+
"eval_runtime": 25.9199,
|
| 553 |
+
"eval_samples_per_second": 1.736,
|
| 554 |
+
"eval_steps_per_second": 1.736,
|
| 555 |
"step": 70
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 2.8049382716049385,
|
| 559 |
+
"grad_norm": 0.04011939465999603,
|
| 560 |
"learning_rate": 4.7196477638140404e-05,
|
| 561 |
+
"loss": 0.0161,
|
| 562 |
"step": 71
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 2.8444444444444446,
|
| 566 |
+
"grad_norm": 0.04298004135489464,
|
| 567 |
"learning_rate": 4.579747375353763e-05,
|
| 568 |
+
"loss": 0.0127,
|
| 569 |
"step": 72
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 2.8839506172839506,
|
| 573 |
+
"grad_norm": 0.042985718697309494,
|
| 574 |
"learning_rate": 4.4401776194834613e-05,
|
| 575 |
+
"loss": 0.018,
|
| 576 |
"step": 73
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 2.9234567901234567,
|
| 580 |
+
"grad_norm": 0.03870657458901405,
|
| 581 |
"learning_rate": 4.3010483023225045e-05,
|
| 582 |
+
"loss": 0.015,
|
| 583 |
"step": 74
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 2.962962962962963,
|
| 587 |
+
"grad_norm": 0.03830299526453018,
|
| 588 |
"learning_rate": 4.162468883476319e-05,
|
| 589 |
+
"loss": 0.016,
|
| 590 |
"step": 75
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 3.0024691358024693,
|
| 594 |
+
"grad_norm": 0.08254650980234146,
|
| 595 |
"learning_rate": 4.0245483899193595e-05,
|
| 596 |
+
"loss": 0.026,
|
| 597 |
"step": 76
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 3.0419753086419754,
|
| 601 |
+
"grad_norm": 0.04306948184967041,
|
| 602 |
"learning_rate": 3.887395330218429e-05,
|
| 603 |
+
"loss": 0.014,
|
| 604 |
"step": 77
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 3.0814814814814815,
|
| 608 |
+
"grad_norm": 0.04832775145769119,
|
| 609 |
"learning_rate": 3.7511176091638653e-05,
|
| 610 |
+
"loss": 0.0152,
|
| 611 |
"step": 78
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 3.1209876543209876,
|
| 615 |
+
"grad_norm": 0.03881710395216942,
|
| 616 |
"learning_rate": 3.6158224428757535e-05,
|
| 617 |
+
"loss": 0.0157,
|
| 618 |
"step": 79
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 3.1604938271604937,
|
| 622 |
+
"grad_norm": 0.05051282048225403,
|
| 623 |
"learning_rate": 3.4816162744519263e-05,
|
| 624 |
+
"loss": 0.0171,
|
| 625 |
"step": 80
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 3.1604938271604937,
|
| 629 |
+
"eval_loss": 0.01797771453857422,
|
| 630 |
+
"eval_runtime": 25.9324,
|
| 631 |
+
"eval_samples_per_second": 1.735,
|
| 632 |
+
"eval_steps_per_second": 1.735,
|
| 633 |
"step": 80
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 3.2,
|
| 637 |
+
"grad_norm": 0.03627981245517731,
|
| 638 |
"learning_rate": 3.3486046902241664e-05,
|
| 639 |
+
"loss": 0.01,
|
| 640 |
"step": 81
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 3.2395061728395063,
|
| 644 |
+
"grad_norm": 0.030452217906713486,
|
| 645 |
"learning_rate": 3.216892336688435e-05,
|
| 646 |
+
"loss": 0.0126,
|
| 647 |
"step": 82
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 3.2790123456790123,
|
| 651 |
+
"grad_norm": 0.04779364913702011,
|
| 652 |
"learning_rate": 3.086582838174551e-05,
|
| 653 |
+
"loss": 0.0143,
|
| 654 |
"step": 83
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 3.3185185185185184,
|
| 658 |
+
"grad_norm": 0.04842763394117355,
|
| 659 |
"learning_rate": 2.9577787153200197e-05,
|
| 660 |
+
"loss": 0.0149,
|
| 661 |
"step": 84
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 3.3580246913580245,
|
| 665 |
+
"grad_norm": 0.040973685681819916,
|
| 666 |
"learning_rate": 2.8305813044122097e-05,
|
| 667 |
+
"loss": 0.0112,
|
| 668 |
"step": 85
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 3.397530864197531,
|
| 672 |
+
"grad_norm": 0.045584630221128464,
|
| 673 |
"learning_rate": 2.705090677662311e-05,
|
| 674 |
+
"loss": 0.0175,
|
| 675 |
"step": 86
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 3.437037037037037,
|
| 679 |
+
"grad_norm": 0.04598742723464966,
|
| 680 |
"learning_rate": 2.581405564473801e-05,
|
| 681 |
+
"loss": 0.0158,
|
| 682 |
"step": 87
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 3.476543209876543,
|
| 686 |
+
"grad_norm": 0.04630625247955322,
|
| 687 |
"learning_rate": 2.459623273767354e-05,
|
| 688 |
+
"loss": 0.0118,
|
| 689 |
"step": 88
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 3.5160493827160493,
|
| 693 |
+
"grad_norm": 0.04400630295276642,
|
| 694 |
"learning_rate": 2.3398396174233178e-05,
|
| 695 |
+
"loss": 0.0146,
|
| 696 |
"step": 89
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 3.5555555555555554,
|
| 700 |
+
"grad_norm": 0.05139096826314926,
|
| 701 |
"learning_rate": 2.2221488349019903e-05,
|
| 702 |
+
"loss": 0.0153,
|
| 703 |
"step": 90
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 3.5555555555555554,
|
| 707 |
+
"eval_loss": 0.01771390251815319,
|
| 708 |
+
"eval_runtime": 25.9448,
|
| 709 |
+
"eval_samples_per_second": 1.734,
|
| 710 |
+
"eval_steps_per_second": 1.734,
|
| 711 |
"step": 90
|
| 712 |
},
|
| 713 |
{
|
| 714 |
"epoch": 3.595061728395062,
|
| 715 |
+
"grad_norm": 0.05596518516540527,
|
| 716 |
"learning_rate": 2.1066435191009715e-05,
|
| 717 |
+
"loss": 0.0166,
|
| 718 |
"step": 91
|
| 719 |
},
|
| 720 |
{
|
| 721 |
"epoch": 3.634567901234568,
|
| 722 |
+
"grad_norm": 0.036661356687545776,
|
| 723 |
"learning_rate": 1.9934145435079702e-05,
|
| 724 |
+
"loss": 0.0144,
|
| 725 |
"step": 92
|
| 726 |
},
|
| 727 |
{
|
| 728 |
"epoch": 3.674074074074074,
|
| 729 |
+
"grad_norm": 0.04676743596792221,
|
| 730 |
"learning_rate": 1.8825509907063327e-05,
|
| 731 |
+
"loss": 0.0129,
|
| 732 |
"step": 93
|
| 733 |
},
|
| 734 |
{
|
| 735 |
"epoch": 3.71358024691358,
|
| 736 |
+
"grad_norm": 0.04297291487455368,
|
| 737 |
"learning_rate": 1.774140082289563e-05,
|
| 738 |
+
"loss": 0.0141,
|
| 739 |
"step": 94
|
| 740 |
},
|
| 741 |
{
|
| 742 |
"epoch": 3.753086419753086,
|
| 743 |
+
"grad_norm": 0.04279183968901634,
|
| 744 |
"learning_rate": 1.6682671102399805e-05,
|
| 745 |
+
"loss": 0.0157,
|
| 746 |
"step": 95
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 3.7925925925925927,
|
| 750 |
+
"grad_norm": 0.047029346227645874,
|
| 751 |
"learning_rate": 1.5650153698254916e-05,
|
| 752 |
+
"loss": 0.0127,
|
| 753 |
"step": 96
|
| 754 |
},
|
| 755 |
{
|
| 756 |
"epoch": 3.832098765432099,
|
| 757 |
+
"grad_norm": 0.04095562919974327,
|
| 758 |
"learning_rate": 1.4644660940672627e-05,
|
| 759 |
+
"loss": 0.0127,
|
| 760 |
"step": 97
|
| 761 |
},
|
| 762 |
{
|
| 763 |
"epoch": 3.871604938271605,
|
| 764 |
+
"grad_norm": 0.043043311685323715,
|
| 765 |
"learning_rate": 1.3666983898298657e-05,
|
| 766 |
+
"loss": 0.018,
|
| 767 |
"step": 98
|
| 768 |
},
|
| 769 |
{
|
| 770 |
"epoch": 3.911111111111111,
|
| 771 |
+
"grad_norm": 0.04111286997795105,
|
| 772 |
"learning_rate": 1.2717891755841722e-05,
|
| 773 |
+
"loss": 0.014,
|
| 774 |
"step": 99
|
| 775 |
},
|
| 776 |
{
|
| 777 |
"epoch": 3.950617283950617,
|
| 778 |
+
"grad_norm": 0.04209229722619057,
|
| 779 |
"learning_rate": 1.1798131208919627e-05,
|
| 780 |
+
"loss": 0.0131,
|
| 781 |
"step": 100
|
| 782 |
},
|
| 783 |
{
|
| 784 |
"epoch": 3.950617283950617,
|
| 785 |
+
"eval_loss": 0.017345581203699112,
|
| 786 |
+
"eval_runtime": 25.965,
|
| 787 |
+
"eval_samples_per_second": 1.733,
|
| 788 |
+
"eval_steps_per_second": 1.733,
|
| 789 |
"step": 100
|
| 790 |
},
|
| 791 |
{
|
| 792 |
"epoch": 3.9901234567901236,
|
| 793 |
+
"grad_norm": 0.05126246437430382,
|
| 794 |
"learning_rate": 1.090842587659851e-05,
|
| 795 |
+
"loss": 0.0169,
|
| 796 |
"step": 101
|
| 797 |
},
|
| 798 |
{
|
| 799 |
"epoch": 4.029629629629629,
|
| 800 |
+
"grad_norm": 0.09892018884420395,
|
| 801 |
"learning_rate": 1.004947573208756e-05,
|
| 802 |
+
"loss": 0.0245,
|
| 803 |
"step": 102
|
| 804 |
},
|
| 805 |
{
|
| 806 |
"epoch": 4.069135802469136,
|
| 807 |
+
"grad_norm": 0.0428558811545372,
|
| 808 |
"learning_rate": 9.221956552036992e-06,
|
| 809 |
+
"loss": 0.0129,
|
| 810 |
"step": 103
|
| 811 |
},
|
| 812 |
{
|
| 813 |
"epoch": 4.108641975308642,
|
| 814 |
+
"grad_norm": 0.039242979139089584,
|
| 815 |
"learning_rate": 8.426519384872733e-06,
|
| 816 |
+
"loss": 0.0125,
|
| 817 |
"step": 104
|
| 818 |
},
|
| 819 |
{
|
| 820 |
"epoch": 4.148148148148148,
|
| 821 |
+
"grad_norm": 0.033157072961330414,
|
| 822 |
"learning_rate": 7.663790038585793e-06,
|
| 823 |
+
"loss": 0.0138,
|
| 824 |
"step": 105
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 4.187654320987654,
|
| 828 |
+
"grad_norm": 0.03438961133360863,
|
| 829 |
"learning_rate": 6.934368588379553e-06,
|
| 830 |
+
"loss": 0.01,
|
| 831 |
"step": 106
|
| 832 |
},
|
| 833 |
{
|
| 834 |
"epoch": 4.22716049382716,
|
| 835 |
+
"grad_norm": 0.03627100959420204,
|
| 836 |
"learning_rate": 6.238828904562316e-06,
|
| 837 |
+
"loss": 0.0139,
|
| 838 |
"step": 107
|
| 839 |
},
|
| 840 |
{
|
| 841 |
"epoch": 4.266666666666667,
|
| 842 |
+
"grad_norm": 0.041909441351890564,
|
| 843 |
"learning_rate": 5.577718201056392e-06,
|
| 844 |
+
"loss": 0.0126,
|
| 845 |
"step": 108
|
| 846 |
},
|
| 847 |
{
|
| 848 |
"epoch": 4.306172839506173,
|
| 849 |
+
"grad_norm": 0.0378437265753746,
|
| 850 |
"learning_rate": 4.951556604879048e-06,
|
| 851 |
+
"loss": 0.0133,
|
| 852 |
"step": 109
|
| 853 |
},
|
| 854 |
{
|
| 855 |
"epoch": 4.345679012345679,
|
| 856 |
+
"grad_norm": 0.046722784638404846,
|
| 857 |
"learning_rate": 4.360836746934055e-06,
|
| 858 |
+
"loss": 0.0143,
|
| 859 |
"step": 110
|
| 860 |
},
|
| 861 |
{
|
| 862 |
"epoch": 4.345679012345679,
|
| 863 |
+
"eval_loss": 0.017259059473872185,
|
| 864 |
+
"eval_runtime": 25.9839,
|
| 865 |
+
"eval_samples_per_second": 1.732,
|
| 866 |
+
"eval_steps_per_second": 1.732,
|
| 867 |
"step": 110
|
| 868 |
},
|
| 869 |
{
|
| 870 |
"epoch": 4.385185185185185,
|
| 871 |
+
"grad_norm": 0.04061531275510788,
|
| 872 |
"learning_rate": 3.8060233744356633e-06,
|
| 873 |
+
"loss": 0.0126,
|
| 874 |
"step": 111
|
| 875 |
},
|
| 876 |
{
|
| 877 |
"epoch": 4.424691358024692,
|
| 878 |
+
"grad_norm": 0.026830753311514854,
|
| 879 |
"learning_rate": 3.2875529852700147e-06,
|
| 880 |
+
"loss": 0.0095,
|
| 881 |
"step": 112
|
| 882 |
},
|
| 883 |
{
|
| 884 |
"epoch": 4.4641975308641975,
|
| 885 |
+
"grad_norm": 0.03643717244267464,
|
| 886 |
"learning_rate": 2.8058334845816213e-06,
|
| 887 |
+
"loss": 0.0131,
|
| 888 |
"step": 113
|
| 889 |
},
|
| 890 |
{
|
| 891 |
"epoch": 4.503703703703704,
|
| 892 |
+
"grad_norm": 0.0387725755572319,
|
| 893 |
"learning_rate": 2.361243863855184e-06,
|
| 894 |
+
"loss": 0.0142,
|
| 895 |
"step": 114
|
| 896 |
},
|
| 897 |
{
|
| 898 |
"epoch": 4.54320987654321,
|
| 899 |
+
"grad_norm": 0.03656044602394104,
|
| 900 |
"learning_rate": 1.9541339027450256e-06,
|
| 901 |
+
"loss": 0.0151,
|
| 902 |
"step": 115
|
| 903 |
},
|
| 904 |
{
|
| 905 |
"epoch": 4.582716049382716,
|
| 906 |
+
"grad_norm": 0.04029763862490654,
|
| 907 |
"learning_rate": 1.584823893886933e-06,
|
| 908 |
+
"loss": 0.0149,
|
| 909 |
"step": 116
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 4.622222222222222,
|
| 913 |
+
"grad_norm": 0.04136398807168007,
|
| 914 |
"learning_rate": 1.2536043909088191e-06,
|
| 915 |
+
"loss": 0.0125,
|
| 916 |
"step": 117
|
| 917 |
},
|
| 918 |
{
|
| 919 |
"epoch": 4.661728395061728,
|
| 920 |
+
"grad_norm": 0.044925201684236526,
|
| 921 |
"learning_rate": 9.607359798384785e-07,
|
| 922 |
+
"loss": 0.0143,
|
| 923 |
"step": 118
|
| 924 |
},
|
| 925 |
{
|
| 926 |
"epoch": 4.701234567901235,
|
| 927 |
+
"grad_norm": 0.03622278571128845,
|
| 928 |
"learning_rate": 7.064490740882057e-07,
|
| 929 |
+
"loss": 0.0138,
|
| 930 |
"step": 119
|
| 931 |
},
|
| 932 |
{
|
| 933 |
"epoch": 4.7407407407407405,
|
| 934 |
+
"grad_norm": 0.04068191349506378,
|
| 935 |
"learning_rate": 4.909437331777179e-07,
|
| 936 |
+
"loss": 0.0156,
|
| 937 |
"step": 120
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 4.7407407407407405,
|
| 941 |
+
"eval_loss": 0.01722303405404091,
|
| 942 |
+
"eval_runtime": 26.004,
|
| 943 |
+
"eval_samples_per_second": 1.731,
|
| 944 |
+
"eval_steps_per_second": 1.731,
|
| 945 |
"step": 120
|
| 946 |
},
|
| 947 |
{
|
| 948 |
"epoch": 4.780246913580247,
|
| 949 |
+
"grad_norm": 0.04049716144800186,
|
| 950 |
"learning_rate": 3.143895053378698e-07,
|
| 951 |
+
"loss": 0.0132,
|
| 952 |
"step": 121
|
| 953 |
},
|
| 954 |
{
|
| 955 |
"epoch": 4.8197530864197535,
|
| 956 |
+
"grad_norm": 0.0428626723587513,
|
| 957 |
"learning_rate": 1.7692529411904578e-07,
|
| 958 |
+
"loss": 0.0144,
|
| 959 |
"step": 122
|
| 960 |
},
|
| 961 |
{
|
| 962 |
"epoch": 4.859259259259259,
|
| 963 |
+
"grad_norm": 0.042141228914260864,
|
| 964 |
"learning_rate": 7.865924910916977e-08,
|
| 965 |
+
"loss": 0.0143,
|
| 966 |
"step": 123
|
| 967 |
},
|
| 968 |
{
|
| 969 |
"epoch": 4.898765432098766,
|
| 970 |
+
"grad_norm": 0.048777248710393906,
|
| 971 |
"learning_rate": 1.9668680847356735e-08,
|
| 972 |
+
"loss": 0.0147,
|
| 973 |
"step": 124
|
| 974 |
},
|
| 975 |
{
|
| 976 |
"epoch": 4.938271604938271,
|
| 977 |
+
"grad_norm": 0.03597211092710495,
|
| 978 |
"learning_rate": 0.0,
|
| 979 |
+
"loss": 0.0127,
|
| 980 |
"step": 125
|
| 981 |
},
|
| 982 |
{
|
| 983 |
"epoch": 4.938271604938271,
|
| 984 |
"step": 125,
|
| 985 |
+
"total_flos": 1.8817065711747072e+17,
|
| 986 |
+
"train_loss": 0.03361496642231941,
|
| 987 |
+
"train_runtime": 5228.875,
|
| 988 |
+
"train_samples_per_second": 0.387,
|
| 989 |
+
"train_steps_per_second": 0.024
|
| 990 |
}
|
| 991 |
],
|
| 992 |
"logging_steps": 1,
|
|
|
|
| 1006 |
"attributes": {}
|
| 1007 |
}
|
| 1008 |
},
|
| 1009 |
+
"total_flos": 1.8817065711747072e+17,
|
| 1010 |
"train_batch_size": 1,
|
| 1011 |
"trial_name": null,
|
| 1012 |
"trial_params": null
|
training_eval_loss.png
CHANGED
|
|
training_loss.png
CHANGED
|
|