End of training
Browse files- README.md +4 -2
- all_results.json +10 -10
- eval_results.json +5 -5
- runs/Dec05_11-28-22_ip-10-90-1-182/events.out.tfevents.1733398455.ip-10-90-1-182.93862.1 +3 -0
- train_results.json +5 -5
- trainer_state.json +822 -822
README.md
CHANGED
|
@@ -3,6 +3,8 @@ library_name: transformers
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: timm/resnet101.a1_in1k
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
@@ -16,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# vit-base-beans
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [timm/resnet101.a1_in1k](https://huggingface.co/timm/resnet101.a1_in1k) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
- Accuracy: 0.8571
|
| 23 |
|
| 24 |
## Model description
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: timm/resnet101.a1_in1k
|
| 5 |
tags:
|
| 6 |
+
- image-classification
|
| 7 |
+
- vision
|
| 8 |
- generated_from_trainer
|
| 9 |
metrics:
|
| 10 |
- accuracy
|
|
|
|
| 18 |
|
| 19 |
# vit-base-beans
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [timm/resnet101.a1_in1k](https://huggingface.co/timm/resnet101.a1_in1k) on the beans dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.4913
|
| 24 |
- Accuracy: 0.8571
|
| 25 |
|
| 26 |
## Model description
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 20.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 0.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second":
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.8571428571428571,
|
| 4 |
+
"eval_loss": 0.49125248193740845,
|
| 5 |
+
"eval_runtime": 0.945,
|
| 6 |
+
"eval_samples_per_second": 140.745,
|
| 7 |
+
"eval_steps_per_second": 17.99,
|
| 8 |
+
"total_flos": 7.939121542823117e+17,
|
| 9 |
+
"train_loss": 0.8216404274793772,
|
| 10 |
+
"train_runtime": 338.1702,
|
| 11 |
+
"train_samples_per_second": 61.153,
|
| 12 |
+
"train_steps_per_second": 7.688
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 20.0,
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 0.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 20.0,
|
| 3 |
+
"eval_accuracy": 0.8571428571428571,
|
| 4 |
+
"eval_loss": 0.49125248193740845,
|
| 5 |
+
"eval_runtime": 0.945,
|
| 6 |
+
"eval_samples_per_second": 140.745,
|
| 7 |
+
"eval_steps_per_second": 17.99
|
| 8 |
}
|
runs/Dec05_11-28-22_ip-10-90-1-182/events.out.tfevents.1733398455.ip-10-90-1-182.93862.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:868519f0e84b5fcf3194b3a9f32e6a75ef4fea3aa53c5005eaa76b23482456a0
|
| 3 |
+
size 411
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 20.0,
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 20.0,
|
| 3 |
+
"total_flos": 7.939121542823117e+17,
|
| 4 |
+
"train_loss": 0.8216404274793772,
|
| 5 |
+
"train_runtime": 338.1702,
|
| 6 |
+
"train_samples_per_second": 61.153,
|
| 7 |
+
"train_steps_per_second": 7.688
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./beans_outputs/checkpoint-
|
| 4 |
"epoch": 20.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 2600,
|
|
@@ -10,2012 +10,2012 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.07692307692307693,
|
| 13 |
-
"grad_norm": 2.
|
| 14 |
-
"learning_rate": 1.
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.15384615384615385,
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 1.
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.23076923076923078,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate": 1.
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.3076923076923077,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate": 1.
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.38461538461538464,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate": 1.
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.46153846153846156,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate": 1.
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.5384615384615384,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate": 1.
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.6153846153846154,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate": 1.
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.6923076923076923,
|
| 69 |
-
"grad_norm": 2.
|
| 70 |
-
"learning_rate": 1.
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.7692307692307693,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate": 1.
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.8461538461538461,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate": 1.
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.9230769230769231,
|
| 90 |
-
"grad_norm": 2.
|
| 91 |
-
"learning_rate": 1.
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 1.0,
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate": 1.
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 1.0,
|
| 104 |
-
"eval_accuracy": 0.
|
| 105 |
-
"eval_loss": 1.
|
| 106 |
-
"eval_runtime":
|
| 107 |
-
"eval_samples_per_second":
|
| 108 |
-
"eval_steps_per_second":
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 1.0769230769230769,
|
| 113 |
-
"grad_norm": 2.
|
| 114 |
-
"learning_rate": 1.
|
| 115 |
-
"loss": 1.
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 1.1538461538461537,
|
| 120 |
-
"grad_norm":
|
| 121 |
-
"learning_rate": 1.
|
| 122 |
-
"loss": 1.
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.2307692307692308,
|
| 127 |
-
"grad_norm":
|
| 128 |
-
"learning_rate": 1.
|
| 129 |
-
"loss": 1.
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 1.3076923076923077,
|
| 134 |
-
"grad_norm":
|
| 135 |
-
"learning_rate": 1.
|
| 136 |
-
"loss": 1.
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 1.3846153846153846,
|
| 141 |
-
"grad_norm":
|
| 142 |
-
"learning_rate": 1.
|
| 143 |
-
"loss": 1.
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.4615384615384617,
|
| 148 |
-
"grad_norm":
|
| 149 |
-
"learning_rate": 1.
|
| 150 |
-
"loss": 1.
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.5384615384615383,
|
| 155 |
-
"grad_norm":
|
| 156 |
-
"learning_rate": 1.
|
| 157 |
-
"loss": 1.
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 1.6153846153846154,
|
| 162 |
-
"grad_norm":
|
| 163 |
-
"learning_rate": 1.
|
| 164 |
-
"loss": 1.
|
| 165 |
"step": 210
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.6923076923076923,
|
| 169 |
-
"grad_norm":
|
| 170 |
-
"learning_rate": 1.
|
| 171 |
-
"loss": 1.
|
| 172 |
"step": 220
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.7692307692307692,
|
| 176 |
-
"grad_norm":
|
| 177 |
-
"learning_rate": 1.
|
| 178 |
-
"loss": 1.
|
| 179 |
"step": 230
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 1.8461538461538463,
|
| 183 |
-
"grad_norm":
|
| 184 |
-
"learning_rate": 1.
|
| 185 |
-
"loss": 1.
|
| 186 |
"step": 240
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 1.9230769230769231,
|
| 190 |
-
"grad_norm":
|
| 191 |
-
"learning_rate": 1.
|
| 192 |
-
"loss": 1.
|
| 193 |
"step": 250
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 2.0,
|
| 197 |
-
"grad_norm":
|
| 198 |
-
"learning_rate": 1.
|
| 199 |
-
"loss": 1.
|
| 200 |
"step": 260
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 2.0,
|
| 204 |
-
"eval_accuracy": 0.
|
| 205 |
-
"eval_loss": 1.
|
| 206 |
-
"eval_runtime": 0.
|
| 207 |
-
"eval_samples_per_second":
|
| 208 |
-
"eval_steps_per_second":
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 2.076923076923077,
|
| 213 |
-
"grad_norm":
|
| 214 |
-
"learning_rate": 1.
|
| 215 |
-
"loss": 1.
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
"epoch": 2.1538461538461537,
|
| 220 |
-
"grad_norm":
|
| 221 |
-
"learning_rate": 1.
|
| 222 |
-
"loss": 1.
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
"epoch": 2.230769230769231,
|
| 227 |
-
"grad_norm":
|
| 228 |
-
"learning_rate": 1.
|
| 229 |
-
"loss": 1.
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 2.3076923076923075,
|
| 234 |
-
"grad_norm": 3.
|
| 235 |
-
"learning_rate": 1.
|
| 236 |
-
"loss": 1.
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 2.3846153846153846,
|
| 241 |
-
"grad_norm":
|
| 242 |
-
"learning_rate": 1.
|
| 243 |
-
"loss": 1.
|
| 244 |
"step": 310
|
| 245 |
},
|
| 246 |
{
|
| 247 |
"epoch": 2.4615384615384617,
|
| 248 |
-
"grad_norm":
|
| 249 |
-
"learning_rate": 1.
|
| 250 |
-
"loss": 1.
|
| 251 |
"step": 320
|
| 252 |
},
|
| 253 |
{
|
| 254 |
"epoch": 2.5384615384615383,
|
| 255 |
-
"grad_norm":
|
| 256 |
-
"learning_rate": 1.
|
| 257 |
-
"loss": 1.
|
| 258 |
"step": 330
|
| 259 |
},
|
| 260 |
{
|
| 261 |
"epoch": 2.6153846153846154,
|
| 262 |
-
"grad_norm": 2.
|
| 263 |
-
"learning_rate": 1.
|
| 264 |
-
"loss": 1.
|
| 265 |
"step": 340
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"epoch": 2.6923076923076925,
|
| 269 |
-
"grad_norm":
|
| 270 |
-
"learning_rate": 1.
|
| 271 |
-
"loss": 1.
|
| 272 |
"step": 350
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 2.769230769230769,
|
| 276 |
-
"grad_norm":
|
| 277 |
-
"learning_rate": 1.
|
| 278 |
-
"loss": 1.
|
| 279 |
"step": 360
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 2.8461538461538463,
|
| 283 |
-
"grad_norm": 2.
|
| 284 |
-
"learning_rate": 1.
|
| 285 |
-
"loss": 1.
|
| 286 |
"step": 370
|
| 287 |
},
|
| 288 |
{
|
| 289 |
"epoch": 2.9230769230769234,
|
| 290 |
-
"grad_norm":
|
| 291 |
-
"learning_rate": 1.
|
| 292 |
-
"loss": 1.
|
| 293 |
"step": 380
|
| 294 |
},
|
| 295 |
{
|
| 296 |
"epoch": 3.0,
|
| 297 |
-
"grad_norm":
|
| 298 |
-
"learning_rate": 1.
|
| 299 |
-
"loss": 1.
|
| 300 |
"step": 390
|
| 301 |
},
|
| 302 |
{
|
| 303 |
"epoch": 3.0,
|
| 304 |
-
"eval_accuracy": 0.
|
| 305 |
-
"eval_loss": 1.
|
| 306 |
-
"eval_runtime": 0.
|
| 307 |
-
"eval_samples_per_second":
|
| 308 |
-
"eval_steps_per_second":
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 3.076923076923077,
|
| 313 |
-
"grad_norm":
|
| 314 |
-
"learning_rate": 1.
|
| 315 |
-
"loss": 1.
|
| 316 |
"step": 400
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 3.1538461538461537,
|
| 320 |
-
"grad_norm":
|
| 321 |
-
"learning_rate": 1.
|
| 322 |
-
"loss": 1.
|
| 323 |
"step": 410
|
| 324 |
},
|
| 325 |
{
|
| 326 |
"epoch": 3.230769230769231,
|
| 327 |
-
"grad_norm":
|
| 328 |
-
"learning_rate": 1.
|
| 329 |
-
"loss": 1.
|
| 330 |
"step": 420
|
| 331 |
},
|
| 332 |
{
|
| 333 |
"epoch": 3.3076923076923075,
|
| 334 |
-
"grad_norm":
|
| 335 |
-
"learning_rate": 1.
|
| 336 |
-
"loss": 1.
|
| 337 |
"step": 430
|
| 338 |
},
|
| 339 |
{
|
| 340 |
"epoch": 3.3846153846153846,
|
| 341 |
-
"grad_norm":
|
| 342 |
-
"learning_rate": 1.
|
| 343 |
-
"loss": 1.
|
| 344 |
"step": 440
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 3.4615384615384617,
|
| 348 |
-
"grad_norm":
|
| 349 |
-
"learning_rate": 1.
|
| 350 |
-
"loss": 1.
|
| 351 |
"step": 450
|
| 352 |
},
|
| 353 |
{
|
| 354 |
"epoch": 3.5384615384615383,
|
| 355 |
-
"grad_norm":
|
| 356 |
-
"learning_rate": 1.
|
| 357 |
-
"loss": 1.
|
| 358 |
"step": 460
|
| 359 |
},
|
| 360 |
{
|
| 361 |
"epoch": 3.6153846153846154,
|
| 362 |
-
"grad_norm":
|
| 363 |
-
"learning_rate": 1.
|
| 364 |
-
"loss": 1.
|
| 365 |
"step": 470
|
| 366 |
},
|
| 367 |
{
|
| 368 |
"epoch": 3.6923076923076925,
|
| 369 |
-
"grad_norm":
|
| 370 |
-
"learning_rate": 1.
|
| 371 |
-
"loss": 1.
|
| 372 |
"step": 480
|
| 373 |
},
|
| 374 |
{
|
| 375 |
"epoch": 3.769230769230769,
|
| 376 |
-
"grad_norm":
|
| 377 |
-
"learning_rate": 1.
|
| 378 |
-
"loss": 0.
|
| 379 |
"step": 490
|
| 380 |
},
|
| 381 |
{
|
| 382 |
"epoch": 3.8461538461538463,
|
| 383 |
-
"grad_norm":
|
| 384 |
-
"learning_rate": 1.
|
| 385 |
-
"loss": 1.
|
| 386 |
"step": 500
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 3.9230769230769234,
|
| 390 |
-
"grad_norm":
|
| 391 |
-
"learning_rate": 1.
|
| 392 |
-
"loss": 0.
|
| 393 |
"step": 510
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 4.0,
|
| 397 |
-
"grad_norm":
|
| 398 |
-
"learning_rate": 1.
|
| 399 |
-
"loss": 1.
|
| 400 |
"step": 520
|
| 401 |
},
|
| 402 |
{
|
| 403 |
"epoch": 4.0,
|
| 404 |
-
"eval_accuracy": 0.
|
| 405 |
-
"eval_loss": 0.
|
| 406 |
-
"eval_runtime": 0.
|
| 407 |
-
"eval_samples_per_second":
|
| 408 |
-
"eval_steps_per_second":
|
| 409 |
"step": 520
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 4.076923076923077,
|
| 413 |
-
"grad_norm":
|
| 414 |
-
"learning_rate": 1.
|
| 415 |
-
"loss":
|
| 416 |
"step": 530
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 4.153846153846154,
|
| 420 |
-
"grad_norm":
|
| 421 |
-
"learning_rate": 1.
|
| 422 |
-
"loss": 1.
|
| 423 |
"step": 540
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 4.230769230769231,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate": 1.
|
| 429 |
-
"loss": 1.
|
| 430 |
"step": 550
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 4.3076923076923075,
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate": 1.
|
| 436 |
-
"loss": 1.
|
| 437 |
"step": 560
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 4.384615384615385,
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate": 1.
|
| 443 |
-
"loss": 0.
|
| 444 |
"step": 570
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 4.461538461538462,
|
| 448 |
-
"grad_norm":
|
| 449 |
-
"learning_rate": 1.
|
| 450 |
-
"loss": 0.
|
| 451 |
"step": 580
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 4.538461538461538,
|
| 455 |
-
"grad_norm":
|
| 456 |
-
"learning_rate": 1.
|
| 457 |
-
"loss": 0.
|
| 458 |
"step": 590
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 4.615384615384615,
|
| 462 |
-
"grad_norm":
|
| 463 |
-
"learning_rate": 1.
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 600
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 4.6923076923076925,
|
| 469 |
-
"grad_norm":
|
| 470 |
-
"learning_rate": 1.
|
| 471 |
-
"loss":
|
| 472 |
"step": 610
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 4.769230769230769,
|
| 476 |
-
"grad_norm": 2.
|
| 477 |
-
"learning_rate": 1.
|
| 478 |
-
"loss": 1.
|
| 479 |
"step": 620
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 4.846153846153846,
|
| 483 |
-
"grad_norm":
|
| 484 |
-
"learning_rate": 1.
|
| 485 |
-
"loss": 0.
|
| 486 |
"step": 630
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 4.923076923076923,
|
| 490 |
-
"grad_norm":
|
| 491 |
-
"learning_rate": 1.
|
| 492 |
-
"loss": 0.
|
| 493 |
"step": 640
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 5.0,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate": 1.
|
| 499 |
-
"loss":
|
| 500 |
"step": 650
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 5.0,
|
| 504 |
-
"eval_accuracy": 0.
|
| 505 |
-
"eval_loss": 0.
|
| 506 |
-
"eval_runtime": 0.
|
| 507 |
-
"eval_samples_per_second":
|
| 508 |
-
"eval_steps_per_second":
|
| 509 |
"step": 650
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 5.076923076923077,
|
| 513 |
-
"grad_norm":
|
| 514 |
-
"learning_rate": 1.
|
| 515 |
-
"loss": 0.
|
| 516 |
"step": 660
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 5.153846153846154,
|
| 520 |
-
"grad_norm":
|
| 521 |
-
"learning_rate": 1.
|
| 522 |
-
"loss":
|
| 523 |
"step": 670
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 5.230769230769231,
|
| 527 |
-
"grad_norm":
|
| 528 |
-
"learning_rate": 1.
|
| 529 |
-
"loss": 0.
|
| 530 |
"step": 680
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 5.3076923076923075,
|
| 534 |
-
"grad_norm":
|
| 535 |
-
"learning_rate": 1.
|
| 536 |
-
"loss": 0.
|
| 537 |
"step": 690
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 5.384615384615385,
|
| 541 |
-
"grad_norm":
|
| 542 |
-
"learning_rate": 1.
|
| 543 |
-
"loss": 0.
|
| 544 |
"step": 700
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 5.461538461538462,
|
| 548 |
-
"grad_norm":
|
| 549 |
-
"learning_rate": 1.
|
| 550 |
-
"loss": 0.
|
| 551 |
"step": 710
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 5.538461538461538,
|
| 555 |
-
"grad_norm":
|
| 556 |
-
"learning_rate": 1.
|
| 557 |
-
"loss": 0.
|
| 558 |
"step": 720
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 5.615384615384615,
|
| 562 |
-
"grad_norm": 2.
|
| 563 |
-
"learning_rate": 1.
|
| 564 |
-
"loss": 0.
|
| 565 |
"step": 730
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 5.6923076923076925,
|
| 569 |
-
"grad_norm":
|
| 570 |
-
"learning_rate": 1.
|
| 571 |
-
"loss": 0.
|
| 572 |
"step": 740
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 5.769230769230769,
|
| 576 |
-
"grad_norm":
|
| 577 |
-
"learning_rate": 1.
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 750
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 5.846153846153846,
|
| 583 |
-
"grad_norm":
|
| 584 |
-
"learning_rate": 1.
|
| 585 |
-
"loss": 0.
|
| 586 |
"step": 760
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 5.923076923076923,
|
| 590 |
-
"grad_norm":
|
| 591 |
-
"learning_rate": 1.
|
| 592 |
-
"loss": 0.
|
| 593 |
"step": 770
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 6.0,
|
| 597 |
-
"grad_norm":
|
| 598 |
-
"learning_rate": 1.
|
| 599 |
-
"loss": 0.
|
| 600 |
"step": 780
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 6.0,
|
| 604 |
"eval_accuracy": 0.7819548872180451,
|
| 605 |
-
"eval_loss": 0.
|
| 606 |
-
"eval_runtime": 0.
|
| 607 |
-
"eval_samples_per_second":
|
| 608 |
-
"eval_steps_per_second":
|
| 609 |
"step": 780
|
| 610 |
},
|
| 611 |
{
|
| 612 |
"epoch": 6.076923076923077,
|
| 613 |
-
"grad_norm":
|
| 614 |
-
"learning_rate": 1.
|
| 615 |
-
"loss": 0.
|
| 616 |
"step": 790
|
| 617 |
},
|
| 618 |
{
|
| 619 |
"epoch": 6.153846153846154,
|
| 620 |
-
"grad_norm":
|
| 621 |
-
"learning_rate": 1.
|
| 622 |
-
"loss": 0.
|
| 623 |
"step": 800
|
| 624 |
},
|
| 625 |
{
|
| 626 |
"epoch": 6.230769230769231,
|
| 627 |
-
"grad_norm":
|
| 628 |
-
"learning_rate": 1.
|
| 629 |
-
"loss": 0.
|
| 630 |
"step": 810
|
| 631 |
},
|
| 632 |
{
|
| 633 |
"epoch": 6.3076923076923075,
|
| 634 |
-
"grad_norm":
|
| 635 |
-
"learning_rate": 1.
|
| 636 |
-
"loss": 0.
|
| 637 |
"step": 820
|
| 638 |
},
|
| 639 |
{
|
| 640 |
"epoch": 6.384615384615385,
|
| 641 |
-
"grad_norm":
|
| 642 |
-
"learning_rate": 1.
|
| 643 |
-
"loss": 0.
|
| 644 |
"step": 830
|
| 645 |
},
|
| 646 |
{
|
| 647 |
"epoch": 6.461538461538462,
|
| 648 |
-
"grad_norm":
|
| 649 |
-
"learning_rate": 1.
|
| 650 |
-
"loss": 0.
|
| 651 |
"step": 840
|
| 652 |
},
|
| 653 |
{
|
| 654 |
"epoch": 6.538461538461538,
|
| 655 |
-
"grad_norm":
|
| 656 |
-
"learning_rate": 1.
|
| 657 |
-
"loss": 0.
|
| 658 |
"step": 850
|
| 659 |
},
|
| 660 |
{
|
| 661 |
"epoch": 6.615384615384615,
|
| 662 |
-
"grad_norm":
|
| 663 |
-
"learning_rate": 1.
|
| 664 |
-
"loss": 0.
|
| 665 |
"step": 860
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"epoch": 6.6923076923076925,
|
| 669 |
-
"grad_norm":
|
| 670 |
-
"learning_rate": 1.
|
| 671 |
-
"loss": 0.
|
| 672 |
"step": 870
|
| 673 |
},
|
| 674 |
{
|
| 675 |
"epoch": 6.769230769230769,
|
| 676 |
-
"grad_norm":
|
| 677 |
-
"learning_rate": 1.
|
| 678 |
-
"loss": 0.
|
| 679 |
"step": 880
|
| 680 |
},
|
| 681 |
{
|
| 682 |
"epoch": 6.846153846153846,
|
| 683 |
-
"grad_norm":
|
| 684 |
-
"learning_rate": 1.
|
| 685 |
-
"loss": 0.
|
| 686 |
"step": 890
|
| 687 |
},
|
| 688 |
{
|
| 689 |
"epoch": 6.923076923076923,
|
| 690 |
-
"grad_norm": 3.
|
| 691 |
-
"learning_rate": 1.
|
| 692 |
-
"loss": 0.
|
| 693 |
"step": 900
|
| 694 |
},
|
| 695 |
{
|
| 696 |
"epoch": 7.0,
|
| 697 |
-
"grad_norm":
|
| 698 |
-
"learning_rate": 1.
|
| 699 |
-
"loss": 0.
|
| 700 |
"step": 910
|
| 701 |
},
|
| 702 |
{
|
| 703 |
"epoch": 7.0,
|
| 704 |
-
"eval_accuracy": 0.
|
| 705 |
-
"eval_loss": 0.
|
| 706 |
-
"eval_runtime": 0.
|
| 707 |
-
"eval_samples_per_second":
|
| 708 |
-
"eval_steps_per_second":
|
| 709 |
"step": 910
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 7.076923076923077,
|
| 713 |
-
"grad_norm":
|
| 714 |
-
"learning_rate": 1.
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 920
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 7.153846153846154,
|
| 720 |
-
"grad_norm": 3.
|
| 721 |
-
"learning_rate": 1.
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 930
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 7.230769230769231,
|
| 727 |
-
"grad_norm":
|
| 728 |
-
"learning_rate": 1.
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 940
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 7.3076923076923075,
|
| 734 |
-
"grad_norm":
|
| 735 |
-
"learning_rate": 1.
|
| 736 |
-
"loss": 0.
|
| 737 |
"step": 950
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 7.384615384615385,
|
| 741 |
-
"grad_norm":
|
| 742 |
-
"learning_rate": 1.
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 960
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 7.461538461538462,
|
| 748 |
-
"grad_norm":
|
| 749 |
-
"learning_rate": 1.
|
| 750 |
-
"loss": 0.
|
| 751 |
"step": 970
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 7.538461538461538,
|
| 755 |
-
"grad_norm":
|
| 756 |
-
"learning_rate":
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 980
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 7.615384615384615,
|
| 762 |
-
"grad_norm":
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 990
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 7.6923076923076925,
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 1000
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 7.769230769230769,
|
| 776 |
-
"grad_norm":
|
| 777 |
-
"learning_rate":
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 1010
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 7.846153846153846,
|
| 783 |
-
"grad_norm":
|
| 784 |
-
"learning_rate":
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 1020
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 7.923076923076923,
|
| 790 |
-
"grad_norm":
|
| 791 |
-
"learning_rate":
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 1030
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 8.0,
|
| 797 |
-
"grad_norm":
|
| 798 |
-
"learning_rate":
|
| 799 |
-
"loss": 0.
|
| 800 |
"step": 1040
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 8.0,
|
| 804 |
-
"eval_accuracy": 0.
|
| 805 |
-
"eval_loss": 0.
|
| 806 |
-
"eval_runtime": 0.
|
| 807 |
-
"eval_samples_per_second":
|
| 808 |
-
"eval_steps_per_second":
|
| 809 |
"step": 1040
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 8.076923076923077,
|
| 813 |
-
"grad_norm":
|
| 814 |
-
"learning_rate":
|
| 815 |
-
"loss": 0.
|
| 816 |
"step": 1050
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 8.153846153846153,
|
| 820 |
-
"grad_norm":
|
| 821 |
-
"learning_rate":
|
| 822 |
-
"loss": 0.
|
| 823 |
"step": 1060
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 8.23076923076923,
|
| 827 |
-
"grad_norm":
|
| 828 |
-
"learning_rate":
|
| 829 |
-
"loss": 0.
|
| 830 |
"step": 1070
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 8.307692307692308,
|
| 834 |
-
"grad_norm":
|
| 835 |
-
"learning_rate":
|
| 836 |
-
"loss": 0.
|
| 837 |
"step": 1080
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 8.384615384615385,
|
| 841 |
-
"grad_norm":
|
| 842 |
-
"learning_rate":
|
| 843 |
-
"loss": 0.
|
| 844 |
"step": 1090
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 8.461538461538462,
|
| 848 |
-
"grad_norm":
|
| 849 |
-
"learning_rate":
|
| 850 |
-
"loss": 0.
|
| 851 |
"step": 1100
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 8.538461538461538,
|
| 855 |
-
"grad_norm":
|
| 856 |
-
"learning_rate":
|
| 857 |
-
"loss": 0.
|
| 858 |
"step": 1110
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 8.615384615384615,
|
| 862 |
-
"grad_norm":
|
| 863 |
-
"learning_rate":
|
| 864 |
-
"loss": 0.
|
| 865 |
"step": 1120
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 8.692307692307692,
|
| 869 |
-
"grad_norm":
|
| 870 |
-
"learning_rate":
|
| 871 |
-
"loss": 0.
|
| 872 |
"step": 1130
|
| 873 |
},
|
| 874 |
{
|
| 875 |
"epoch": 8.76923076923077,
|
| 876 |
-
"grad_norm":
|
| 877 |
-
"learning_rate":
|
| 878 |
-
"loss": 0.
|
| 879 |
"step": 1140
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 8.846153846153847,
|
| 883 |
-
"grad_norm":
|
| 884 |
-
"learning_rate":
|
| 885 |
-
"loss": 0.
|
| 886 |
"step": 1150
|
| 887 |
},
|
| 888 |
{
|
| 889 |
"epoch": 8.923076923076923,
|
| 890 |
-
"grad_norm":
|
| 891 |
-
"learning_rate":
|
| 892 |
-
"loss": 0.
|
| 893 |
"step": 1160
|
| 894 |
},
|
| 895 |
{
|
| 896 |
"epoch": 9.0,
|
| 897 |
-
"grad_norm":
|
| 898 |
-
"learning_rate":
|
| 899 |
-
"loss": 0.
|
| 900 |
"step": 1170
|
| 901 |
},
|
| 902 |
{
|
| 903 |
"epoch": 9.0,
|
| 904 |
-
"eval_accuracy": 0.
|
| 905 |
-
"eval_loss": 0.
|
| 906 |
-
"eval_runtime": 0.
|
| 907 |
-
"eval_samples_per_second":
|
| 908 |
-
"eval_steps_per_second":
|
| 909 |
"step": 1170
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 9.076923076923077,
|
| 913 |
-
"grad_norm":
|
| 914 |
-
"learning_rate":
|
| 915 |
-
"loss": 0.
|
| 916 |
"step": 1180
|
| 917 |
},
|
| 918 |
{
|
| 919 |
"epoch": 9.153846153846153,
|
| 920 |
-
"grad_norm":
|
| 921 |
-
"learning_rate":
|
| 922 |
-
"loss": 0.
|
| 923 |
"step": 1190
|
| 924 |
},
|
| 925 |
{
|
| 926 |
"epoch": 9.23076923076923,
|
| 927 |
-
"grad_norm":
|
| 928 |
-
"learning_rate":
|
| 929 |
-
"loss": 0.
|
| 930 |
"step": 1200
|
| 931 |
},
|
| 932 |
{
|
| 933 |
"epoch": 9.307692307692308,
|
| 934 |
-
"grad_norm":
|
| 935 |
-
"learning_rate":
|
| 936 |
-
"loss": 0.
|
| 937 |
"step": 1210
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 9.384615384615385,
|
| 941 |
-
"grad_norm":
|
| 942 |
-
"learning_rate":
|
| 943 |
-
"loss": 0.
|
| 944 |
"step": 1220
|
| 945 |
},
|
| 946 |
{
|
| 947 |
"epoch": 9.461538461538462,
|
| 948 |
-
"grad_norm":
|
| 949 |
-
"learning_rate":
|
| 950 |
-
"loss": 0.
|
| 951 |
"step": 1230
|
| 952 |
},
|
| 953 |
{
|
| 954 |
"epoch": 9.538461538461538,
|
| 955 |
-
"grad_norm":
|
| 956 |
-
"learning_rate":
|
| 957 |
-
"loss": 0.
|
| 958 |
"step": 1240
|
| 959 |
},
|
| 960 |
{
|
| 961 |
"epoch": 9.615384615384615,
|
| 962 |
-
"grad_norm":
|
| 963 |
-
"learning_rate":
|
| 964 |
-
"loss": 0.
|
| 965 |
"step": 1250
|
| 966 |
},
|
| 967 |
{
|
| 968 |
"epoch": 9.692307692307692,
|
| 969 |
-
"grad_norm":
|
| 970 |
-
"learning_rate":
|
| 971 |
-
"loss": 0.
|
| 972 |
"step": 1260
|
| 973 |
},
|
| 974 |
{
|
| 975 |
"epoch": 9.76923076923077,
|
| 976 |
-
"grad_norm":
|
| 977 |
-
"learning_rate":
|
| 978 |
-
"loss": 0.
|
| 979 |
"step": 1270
|
| 980 |
},
|
| 981 |
{
|
| 982 |
"epoch": 9.846153846153847,
|
| 983 |
-
"grad_norm":
|
| 984 |
-
"learning_rate":
|
| 985 |
-
"loss": 0.
|
| 986 |
"step": 1280
|
| 987 |
},
|
| 988 |
{
|
| 989 |
"epoch": 9.923076923076923,
|
| 990 |
-
"grad_norm":
|
| 991 |
-
"learning_rate":
|
| 992 |
-
"loss": 0.
|
| 993 |
"step": 1290
|
| 994 |
},
|
| 995 |
{
|
| 996 |
"epoch": 10.0,
|
| 997 |
-
"grad_norm":
|
| 998 |
-
"learning_rate":
|
| 999 |
-
"loss": 0.
|
| 1000 |
"step": 1300
|
| 1001 |
},
|
| 1002 |
{
|
| 1003 |
"epoch": 10.0,
|
| 1004 |
-
"eval_accuracy": 0.
|
| 1005 |
-
"eval_loss": 0.
|
| 1006 |
-
"eval_runtime": 0.
|
| 1007 |
-
"eval_samples_per_second":
|
| 1008 |
-
"eval_steps_per_second":
|
| 1009 |
"step": 1300
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 10.076923076923077,
|
| 1013 |
-
"grad_norm":
|
| 1014 |
-
"learning_rate":
|
| 1015 |
-
"loss": 0.
|
| 1016 |
"step": 1310
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 10.153846153846153,
|
| 1020 |
-
"grad_norm":
|
| 1021 |
-
"learning_rate":
|
| 1022 |
-
"loss": 0.
|
| 1023 |
"step": 1320
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 10.23076923076923,
|
| 1027 |
-
"grad_norm":
|
| 1028 |
-
"learning_rate":
|
| 1029 |
-
"loss": 0.
|
| 1030 |
"step": 1330
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 10.307692307692308,
|
| 1034 |
-
"grad_norm":
|
| 1035 |
-
"learning_rate":
|
| 1036 |
-
"loss": 0.
|
| 1037 |
"step": 1340
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 10.384615384615385,
|
| 1041 |
-
"grad_norm":
|
| 1042 |
-
"learning_rate":
|
| 1043 |
-
"loss": 0.
|
| 1044 |
"step": 1350
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 10.461538461538462,
|
| 1048 |
-
"grad_norm":
|
| 1049 |
-
"learning_rate":
|
| 1050 |
-
"loss": 0.
|
| 1051 |
"step": 1360
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
"epoch": 10.538461538461538,
|
| 1055 |
-
"grad_norm":
|
| 1056 |
-
"learning_rate":
|
| 1057 |
-
"loss": 0.
|
| 1058 |
"step": 1370
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
"epoch": 10.615384615384615,
|
| 1062 |
-
"grad_norm":
|
| 1063 |
-
"learning_rate":
|
| 1064 |
-
"loss": 0.
|
| 1065 |
"step": 1380
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 10.692307692307692,
|
| 1069 |
-
"grad_norm":
|
| 1070 |
-
"learning_rate":
|
| 1071 |
-
"loss": 0.
|
| 1072 |
"step": 1390
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
"epoch": 10.76923076923077,
|
| 1076 |
-
"grad_norm":
|
| 1077 |
-
"learning_rate":
|
| 1078 |
-
"loss": 0.
|
| 1079 |
"step": 1400
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
"epoch": 10.846153846153847,
|
| 1083 |
-
"grad_norm":
|
| 1084 |
-
"learning_rate":
|
| 1085 |
-
"loss": 0.
|
| 1086 |
"step": 1410
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 10.923076923076923,
|
| 1090 |
-
"grad_norm":
|
| 1091 |
-
"learning_rate":
|
| 1092 |
-
"loss": 0.
|
| 1093 |
"step": 1420
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
"epoch": 11.0,
|
| 1097 |
-
"grad_norm":
|
| 1098 |
-
"learning_rate":
|
| 1099 |
-
"loss": 0.
|
| 1100 |
"step": 1430
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
"epoch": 11.0,
|
| 1104 |
-
"eval_accuracy": 0.
|
| 1105 |
-
"eval_loss": 0.
|
| 1106 |
-
"eval_runtime": 0.
|
| 1107 |
-
"eval_samples_per_second":
|
| 1108 |
-
"eval_steps_per_second":
|
| 1109 |
"step": 1430
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 11.076923076923077,
|
| 1113 |
-
"grad_norm":
|
| 1114 |
-
"learning_rate":
|
| 1115 |
-
"loss": 0.
|
| 1116 |
"step": 1440
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 11.153846153846153,
|
| 1120 |
-
"grad_norm":
|
| 1121 |
-
"learning_rate":
|
| 1122 |
-
"loss": 0.
|
| 1123 |
"step": 1450
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 11.23076923076923,
|
| 1127 |
-
"grad_norm":
|
| 1128 |
-
"learning_rate":
|
| 1129 |
-
"loss": 0.
|
| 1130 |
"step": 1460
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 11.307692307692308,
|
| 1134 |
-
"grad_norm":
|
| 1135 |
-
"learning_rate":
|
| 1136 |
-
"loss": 0.
|
| 1137 |
"step": 1470
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 11.384615384615385,
|
| 1141 |
-
"grad_norm":
|
| 1142 |
-
"learning_rate":
|
| 1143 |
-
"loss": 0.
|
| 1144 |
"step": 1480
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 11.461538461538462,
|
| 1148 |
-
"grad_norm":
|
| 1149 |
-
"learning_rate":
|
| 1150 |
-
"loss": 0.
|
| 1151 |
"step": 1490
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 11.538461538461538,
|
| 1155 |
-
"grad_norm":
|
| 1156 |
-
"learning_rate":
|
| 1157 |
-
"loss": 0.
|
| 1158 |
"step": 1500
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 11.615384615384615,
|
| 1162 |
-
"grad_norm": 3.
|
| 1163 |
-
"learning_rate":
|
| 1164 |
-
"loss": 0.
|
| 1165 |
"step": 1510
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 11.692307692307692,
|
| 1169 |
-
"grad_norm":
|
| 1170 |
-
"learning_rate":
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 1520
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 11.76923076923077,
|
| 1176 |
-
"grad_norm":
|
| 1177 |
-
"learning_rate":
|
| 1178 |
-
"loss": 0.
|
| 1179 |
"step": 1530
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 11.846153846153847,
|
| 1183 |
-
"grad_norm":
|
| 1184 |
-
"learning_rate":
|
| 1185 |
-
"loss": 0.
|
| 1186 |
"step": 1540
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 11.923076923076923,
|
| 1190 |
-
"grad_norm":
|
| 1191 |
-
"learning_rate":
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 1550
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 12.0,
|
| 1197 |
-
"grad_norm":
|
| 1198 |
-
"learning_rate":
|
| 1199 |
-
"loss": 0.
|
| 1200 |
"step": 1560
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 12.0,
|
| 1204 |
-
"eval_accuracy": 0.
|
| 1205 |
-
"eval_loss": 0.
|
| 1206 |
-
"eval_runtime": 0.
|
| 1207 |
-
"eval_samples_per_second":
|
| 1208 |
-
"eval_steps_per_second":
|
| 1209 |
"step": 1560
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 12.076923076923077,
|
| 1213 |
-
"grad_norm":
|
| 1214 |
-
"learning_rate":
|
| 1215 |
-
"loss": 0.
|
| 1216 |
"step": 1570
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
"epoch": 12.153846153846153,
|
| 1220 |
-
"grad_norm":
|
| 1221 |
-
"learning_rate":
|
| 1222 |
-
"loss": 0.
|
| 1223 |
"step": 1580
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
"epoch": 12.23076923076923,
|
| 1227 |
-
"grad_norm":
|
| 1228 |
-
"learning_rate":
|
| 1229 |
-
"loss": 0.
|
| 1230 |
"step": 1590
|
| 1231 |
},
|
| 1232 |
{
|
| 1233 |
"epoch": 12.307692307692308,
|
| 1234 |
-
"grad_norm":
|
| 1235 |
-
"learning_rate":
|
| 1236 |
-
"loss": 0.
|
| 1237 |
"step": 1600
|
| 1238 |
},
|
| 1239 |
{
|
| 1240 |
"epoch": 12.384615384615385,
|
| 1241 |
-
"grad_norm":
|
| 1242 |
-
"learning_rate":
|
| 1243 |
-
"loss": 0.
|
| 1244 |
"step": 1610
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
"epoch": 12.461538461538462,
|
| 1248 |
-
"grad_norm": 3.
|
| 1249 |
-
"learning_rate":
|
| 1250 |
-
"loss": 0.
|
| 1251 |
"step": 1620
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
"epoch": 12.538461538461538,
|
| 1255 |
-
"grad_norm":
|
| 1256 |
-
"learning_rate":
|
| 1257 |
-
"loss": 0.
|
| 1258 |
"step": 1630
|
| 1259 |
},
|
| 1260 |
{
|
| 1261 |
"epoch": 12.615384615384615,
|
| 1262 |
-
"grad_norm":
|
| 1263 |
-
"learning_rate":
|
| 1264 |
-
"loss": 0.
|
| 1265 |
"step": 1640
|
| 1266 |
},
|
| 1267 |
{
|
| 1268 |
"epoch": 12.692307692307692,
|
| 1269 |
-
"grad_norm":
|
| 1270 |
-
"learning_rate":
|
| 1271 |
-
"loss": 0.
|
| 1272 |
"step": 1650
|
| 1273 |
},
|
| 1274 |
{
|
| 1275 |
"epoch": 12.76923076923077,
|
| 1276 |
-
"grad_norm":
|
| 1277 |
-
"learning_rate":
|
| 1278 |
-
"loss": 0.
|
| 1279 |
"step": 1660
|
| 1280 |
},
|
| 1281 |
{
|
| 1282 |
"epoch": 12.846153846153847,
|
| 1283 |
-
"grad_norm":
|
| 1284 |
-
"learning_rate":
|
| 1285 |
-
"loss": 0.
|
| 1286 |
"step": 1670
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
"epoch": 12.923076923076923,
|
| 1290 |
-
"grad_norm":
|
| 1291 |
-
"learning_rate":
|
| 1292 |
-
"loss": 0.
|
| 1293 |
"step": 1680
|
| 1294 |
},
|
| 1295 |
{
|
| 1296 |
"epoch": 13.0,
|
| 1297 |
-
"grad_norm":
|
| 1298 |
-
"learning_rate":
|
| 1299 |
-
"loss": 0.
|
| 1300 |
"step": 1690
|
| 1301 |
},
|
| 1302 |
{
|
| 1303 |
"epoch": 13.0,
|
| 1304 |
-
"eval_accuracy": 0.
|
| 1305 |
-
"eval_loss": 0.
|
| 1306 |
-
"eval_runtime": 0.
|
| 1307 |
-
"eval_samples_per_second":
|
| 1308 |
-
"eval_steps_per_second":
|
| 1309 |
"step": 1690
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
"epoch": 13.076923076923077,
|
| 1313 |
-
"grad_norm":
|
| 1314 |
-
"learning_rate":
|
| 1315 |
-
"loss": 0.
|
| 1316 |
"step": 1700
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
"epoch": 13.153846153846153,
|
| 1320 |
-
"grad_norm":
|
| 1321 |
-
"learning_rate":
|
| 1322 |
-
"loss": 0.
|
| 1323 |
"step": 1710
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
"epoch": 13.23076923076923,
|
| 1327 |
-
"grad_norm":
|
| 1328 |
-
"learning_rate":
|
| 1329 |
-
"loss": 0.
|
| 1330 |
"step": 1720
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 13.307692307692308,
|
| 1334 |
-
"grad_norm":
|
| 1335 |
-
"learning_rate":
|
| 1336 |
-
"loss": 0.
|
| 1337 |
"step": 1730
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"epoch": 13.384615384615385,
|
| 1341 |
-
"grad_norm":
|
| 1342 |
-
"learning_rate":
|
| 1343 |
-
"loss": 0.
|
| 1344 |
"step": 1740
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
"epoch": 13.461538461538462,
|
| 1348 |
-
"grad_norm":
|
| 1349 |
-
"learning_rate":
|
| 1350 |
-
"loss": 0.
|
| 1351 |
"step": 1750
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
"epoch": 13.538461538461538,
|
| 1355 |
-
"grad_norm":
|
| 1356 |
-
"learning_rate":
|
| 1357 |
-
"loss": 0.
|
| 1358 |
"step": 1760
|
| 1359 |
},
|
| 1360 |
{
|
| 1361 |
"epoch": 13.615384615384615,
|
| 1362 |
-
"grad_norm":
|
| 1363 |
-
"learning_rate":
|
| 1364 |
-
"loss": 0.
|
| 1365 |
"step": 1770
|
| 1366 |
},
|
| 1367 |
{
|
| 1368 |
"epoch": 13.692307692307692,
|
| 1369 |
-
"grad_norm":
|
| 1370 |
-
"learning_rate":
|
| 1371 |
-
"loss": 0.
|
| 1372 |
"step": 1780
|
| 1373 |
},
|
| 1374 |
{
|
| 1375 |
"epoch": 13.76923076923077,
|
| 1376 |
-
"grad_norm":
|
| 1377 |
-
"learning_rate":
|
| 1378 |
-
"loss": 0.
|
| 1379 |
"step": 1790
|
| 1380 |
},
|
| 1381 |
{
|
| 1382 |
"epoch": 13.846153846153847,
|
| 1383 |
-
"grad_norm":
|
| 1384 |
-
"learning_rate":
|
| 1385 |
-
"loss": 0.
|
| 1386 |
"step": 1800
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
"epoch": 13.923076923076923,
|
| 1390 |
-
"grad_norm":
|
| 1391 |
-
"learning_rate":
|
| 1392 |
-
"loss": 0.
|
| 1393 |
"step": 1810
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
"epoch": 14.0,
|
| 1397 |
-
"grad_norm":
|
| 1398 |
-
"learning_rate":
|
| 1399 |
-
"loss": 0.
|
| 1400 |
"step": 1820
|
| 1401 |
},
|
| 1402 |
{
|
| 1403 |
"epoch": 14.0,
|
| 1404 |
-
"eval_accuracy": 0.
|
| 1405 |
-
"eval_loss": 0.
|
| 1406 |
-
"eval_runtime": 0.
|
| 1407 |
-
"eval_samples_per_second":
|
| 1408 |
-
"eval_steps_per_second":
|
| 1409 |
"step": 1820
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
"epoch": 14.076923076923077,
|
| 1413 |
-
"grad_norm":
|
| 1414 |
-
"learning_rate":
|
| 1415 |
-
"loss": 0.
|
| 1416 |
"step": 1830
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
"epoch": 14.153846153846153,
|
| 1420 |
-
"grad_norm":
|
| 1421 |
-
"learning_rate":
|
| 1422 |
-
"loss": 0.
|
| 1423 |
"step": 1840
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
"epoch": 14.23076923076923,
|
| 1427 |
-
"grad_norm":
|
| 1428 |
-
"learning_rate":
|
| 1429 |
-
"loss": 0.
|
| 1430 |
"step": 1850
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 14.307692307692308,
|
| 1434 |
-
"grad_norm":
|
| 1435 |
-
"learning_rate":
|
| 1436 |
-
"loss": 0.
|
| 1437 |
"step": 1860
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
"epoch": 14.384615384615385,
|
| 1441 |
-
"grad_norm":
|
| 1442 |
-
"learning_rate":
|
| 1443 |
-
"loss": 0.
|
| 1444 |
"step": 1870
|
| 1445 |
},
|
| 1446 |
{
|
| 1447 |
"epoch": 14.461538461538462,
|
| 1448 |
-
"grad_norm":
|
| 1449 |
-
"learning_rate":
|
| 1450 |
-
"loss": 0.
|
| 1451 |
"step": 1880
|
| 1452 |
},
|
| 1453 |
{
|
| 1454 |
"epoch": 14.538461538461538,
|
| 1455 |
-
"grad_norm":
|
| 1456 |
-
"learning_rate":
|
| 1457 |
-
"loss": 0.
|
| 1458 |
"step": 1890
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 14.615384615384615,
|
| 1462 |
-
"grad_norm":
|
| 1463 |
-
"learning_rate": 5.
|
| 1464 |
-
"loss": 0.
|
| 1465 |
"step": 1900
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 14.692307692307692,
|
| 1469 |
-
"grad_norm":
|
| 1470 |
-
"learning_rate":
|
| 1471 |
-
"loss": 0.
|
| 1472 |
"step": 1910
|
| 1473 |
},
|
| 1474 |
{
|
| 1475 |
"epoch": 14.76923076923077,
|
| 1476 |
-
"grad_norm":
|
| 1477 |
-
"learning_rate":
|
| 1478 |
-
"loss": 0.
|
| 1479 |
"step": 1920
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 14.846153846153847,
|
| 1483 |
-
"grad_norm": 3.
|
| 1484 |
-
"learning_rate":
|
| 1485 |
-
"loss": 0.
|
| 1486 |
"step": 1930
|
| 1487 |
},
|
| 1488 |
{
|
| 1489 |
"epoch": 14.923076923076923,
|
| 1490 |
-
"grad_norm":
|
| 1491 |
-
"learning_rate":
|
| 1492 |
-
"loss": 0.
|
| 1493 |
"step": 1940
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
"epoch": 15.0,
|
| 1497 |
-
"grad_norm":
|
| 1498 |
-
"learning_rate":
|
| 1499 |
-
"loss": 0.
|
| 1500 |
"step": 1950
|
| 1501 |
},
|
| 1502 |
{
|
| 1503 |
"epoch": 15.0,
|
| 1504 |
-
"eval_accuracy": 0.
|
| 1505 |
-
"eval_loss": 0.
|
| 1506 |
-
"eval_runtime": 0.
|
| 1507 |
-
"eval_samples_per_second":
|
| 1508 |
-
"eval_steps_per_second":
|
| 1509 |
"step": 1950
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
"epoch": 15.076923076923077,
|
| 1513 |
-
"grad_norm":
|
| 1514 |
"learning_rate": 4.923076923076924e-06,
|
| 1515 |
-
"loss": 0.
|
| 1516 |
"step": 1960
|
| 1517 |
},
|
| 1518 |
{
|
| 1519 |
"epoch": 15.153846153846153,
|
| 1520 |
-
"grad_norm":
|
| 1521 |
"learning_rate": 4.8461538461538465e-06,
|
| 1522 |
-
"loss": 0.
|
| 1523 |
"step": 1970
|
| 1524 |
},
|
| 1525 |
{
|
| 1526 |
"epoch": 15.23076923076923,
|
| 1527 |
-
"grad_norm":
|
| 1528 |
"learning_rate": 4.76923076923077e-06,
|
| 1529 |
-
"loss": 0.
|
| 1530 |
"step": 1980
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 15.307692307692308,
|
| 1534 |
-
"grad_norm":
|
| 1535 |
"learning_rate": 4.692307692307693e-06,
|
| 1536 |
-
"loss": 0.
|
| 1537 |
"step": 1990
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
"epoch": 15.384615384615385,
|
| 1541 |
-
"grad_norm":
|
| 1542 |
"learning_rate": 4.615384615384616e-06,
|
| 1543 |
-
"loss": 0.
|
| 1544 |
"step": 2000
|
| 1545 |
},
|
| 1546 |
{
|
| 1547 |
"epoch": 15.461538461538462,
|
| 1548 |
-
"grad_norm":
|
| 1549 |
"learning_rate": 4.538461538461539e-06,
|
| 1550 |
-
"loss": 0.
|
| 1551 |
"step": 2010
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
"epoch": 15.538461538461538,
|
| 1555 |
-
"grad_norm":
|
| 1556 |
"learning_rate": 4.461538461538462e-06,
|
| 1557 |
-
"loss": 0.
|
| 1558 |
"step": 2020
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
"epoch": 15.615384615384615,
|
| 1562 |
-
"grad_norm":
|
| 1563 |
"learning_rate": 4.384615384615385e-06,
|
| 1564 |
-
"loss": 0.
|
| 1565 |
"step": 2030
|
| 1566 |
},
|
| 1567 |
{
|
| 1568 |
"epoch": 15.692307692307692,
|
| 1569 |
-
"grad_norm":
|
| 1570 |
"learning_rate": 4.307692307692308e-06,
|
| 1571 |
-
"loss": 0.
|
| 1572 |
"step": 2040
|
| 1573 |
},
|
| 1574 |
{
|
| 1575 |
"epoch": 15.76923076923077,
|
| 1576 |
-
"grad_norm":
|
| 1577 |
"learning_rate": 4.230769230769231e-06,
|
| 1578 |
-
"loss": 0.
|
| 1579 |
"step": 2050
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
"epoch": 15.846153846153847,
|
| 1583 |
-
"grad_norm":
|
| 1584 |
"learning_rate": 4.1538461538461545e-06,
|
| 1585 |
-
"loss": 0.
|
| 1586 |
"step": 2060
|
| 1587 |
},
|
| 1588 |
{
|
| 1589 |
"epoch": 15.923076923076923,
|
| 1590 |
-
"grad_norm":
|
| 1591 |
"learning_rate": 4.076923076923077e-06,
|
| 1592 |
-
"loss": 0.
|
| 1593 |
"step": 2070
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 16.0,
|
| 1597 |
-
"grad_norm":
|
| 1598 |
"learning_rate": 4.000000000000001e-06,
|
| 1599 |
-
"loss": 0.
|
| 1600 |
"step": 2080
|
| 1601 |
},
|
| 1602 |
{
|
| 1603 |
"epoch": 16.0,
|
| 1604 |
-
"eval_accuracy": 0.
|
| 1605 |
-
"eval_loss": 0.
|
| 1606 |
-
"eval_runtime": 0.
|
| 1607 |
-
"eval_samples_per_second":
|
| 1608 |
-
"eval_steps_per_second":
|
| 1609 |
"step": 2080
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
"epoch": 16.076923076923077,
|
| 1613 |
-
"grad_norm":
|
| 1614 |
"learning_rate": 3.923076923076923e-06,
|
| 1615 |
-
"loss": 0.
|
| 1616 |
"step": 2090
|
| 1617 |
},
|
| 1618 |
{
|
| 1619 |
"epoch": 16.153846153846153,
|
| 1620 |
-
"grad_norm":
|
| 1621 |
"learning_rate": 3.846153846153847e-06,
|
| 1622 |
-
"loss": 0.
|
| 1623 |
"step": 2100
|
| 1624 |
},
|
| 1625 |
{
|
| 1626 |
"epoch": 16.23076923076923,
|
| 1627 |
-
"grad_norm":
|
| 1628 |
"learning_rate": 3.7692307692307694e-06,
|
| 1629 |
-
"loss": 0.
|
| 1630 |
"step": 2110
|
| 1631 |
},
|
| 1632 |
{
|
| 1633 |
"epoch": 16.307692307692307,
|
| 1634 |
-
"grad_norm":
|
| 1635 |
"learning_rate": 3.692307692307693e-06,
|
| 1636 |
-
"loss": 0.
|
| 1637 |
"step": 2120
|
| 1638 |
},
|
| 1639 |
{
|
| 1640 |
"epoch": 16.384615384615383,
|
| 1641 |
-
"grad_norm":
|
| 1642 |
"learning_rate": 3.6153846153846156e-06,
|
| 1643 |
-
"loss": 0.
|
| 1644 |
"step": 2130
|
| 1645 |
},
|
| 1646 |
{
|
| 1647 |
"epoch": 16.46153846153846,
|
| 1648 |
-
"grad_norm":
|
| 1649 |
"learning_rate": 3.538461538461539e-06,
|
| 1650 |
-
"loss": 0.
|
| 1651 |
"step": 2140
|
| 1652 |
},
|
| 1653 |
{
|
| 1654 |
"epoch": 16.53846153846154,
|
| 1655 |
-
"grad_norm":
|
| 1656 |
"learning_rate": 3.4615384615384617e-06,
|
| 1657 |
-
"loss": 0.
|
| 1658 |
"step": 2150
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
"epoch": 16.615384615384617,
|
| 1662 |
-
"grad_norm":
|
| 1663 |
"learning_rate": 3.384615384615385e-06,
|
| 1664 |
-
"loss": 0.
|
| 1665 |
"step": 2160
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 16.692307692307693,
|
| 1669 |
-
"grad_norm":
|
| 1670 |
"learning_rate": 3.307692307692308e-06,
|
| 1671 |
-
"loss": 0.
|
| 1672 |
"step": 2170
|
| 1673 |
},
|
| 1674 |
{
|
| 1675 |
"epoch": 16.76923076923077,
|
| 1676 |
-
"grad_norm":
|
| 1677 |
"learning_rate": 3.2307692307692313e-06,
|
| 1678 |
-
"loss": 0.
|
| 1679 |
"step": 2180
|
| 1680 |
},
|
| 1681 |
{
|
| 1682 |
"epoch": 16.846153846153847,
|
| 1683 |
-
"grad_norm":
|
| 1684 |
"learning_rate": 3.153846153846154e-06,
|
| 1685 |
-
"loss": 0.
|
| 1686 |
"step": 2190
|
| 1687 |
},
|
| 1688 |
{
|
| 1689 |
"epoch": 16.923076923076923,
|
| 1690 |
-
"grad_norm": 2.
|
| 1691 |
"learning_rate": 3.0769230769230774e-06,
|
| 1692 |
-
"loss": 0.
|
| 1693 |
"step": 2200
|
| 1694 |
},
|
| 1695 |
{
|
| 1696 |
"epoch": 17.0,
|
| 1697 |
-
"grad_norm":
|
| 1698 |
"learning_rate": 3e-06,
|
| 1699 |
-
"loss": 0.
|
| 1700 |
"step": 2210
|
| 1701 |
},
|
| 1702 |
{
|
| 1703 |
"epoch": 17.0,
|
| 1704 |
-
"eval_accuracy": 0.
|
| 1705 |
-
"eval_loss": 0.
|
| 1706 |
-
"eval_runtime": 0.
|
| 1707 |
-
"eval_samples_per_second":
|
| 1708 |
-
"eval_steps_per_second":
|
| 1709 |
"step": 2210
|
| 1710 |
},
|
| 1711 |
{
|
| 1712 |
"epoch": 17.076923076923077,
|
| 1713 |
-
"grad_norm":
|
| 1714 |
"learning_rate": 2.9230769230769236e-06,
|
| 1715 |
-
"loss": 0.
|
| 1716 |
"step": 2220
|
| 1717 |
},
|
| 1718 |
{
|
| 1719 |
"epoch": 17.153846153846153,
|
| 1720 |
-
"grad_norm":
|
| 1721 |
"learning_rate": 2.846153846153846e-06,
|
| 1722 |
-
"loss": 0.
|
| 1723 |
"step": 2230
|
| 1724 |
},
|
| 1725 |
{
|
| 1726 |
"epoch": 17.23076923076923,
|
| 1727 |
-
"grad_norm":
|
| 1728 |
"learning_rate": 2.7692307692307697e-06,
|
| 1729 |
-
"loss": 0.
|
| 1730 |
"step": 2240
|
| 1731 |
},
|
| 1732 |
{
|
| 1733 |
"epoch": 17.307692307692307,
|
| 1734 |
-
"grad_norm":
|
| 1735 |
"learning_rate": 2.6923076923076923e-06,
|
| 1736 |
-
"loss": 0.
|
| 1737 |
"step": 2250
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 17.384615384615383,
|
| 1741 |
-
"grad_norm": 3.
|
| 1742 |
"learning_rate": 2.615384615384616e-06,
|
| 1743 |
-
"loss": 0.
|
| 1744 |
"step": 2260
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 17.46153846153846,
|
| 1748 |
-
"grad_norm":
|
| 1749 |
"learning_rate": 2.5384615384615385e-06,
|
| 1750 |
-
"loss": 0.
|
| 1751 |
"step": 2270
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 17.53846153846154,
|
| 1755 |
-
"grad_norm":
|
| 1756 |
"learning_rate": 2.461538461538462e-06,
|
| 1757 |
-
"loss": 0.
|
| 1758 |
"step": 2280
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 17.615384615384617,
|
| 1762 |
-
"grad_norm":
|
| 1763 |
"learning_rate": 2.384615384615385e-06,
|
| 1764 |
-
"loss": 0.
|
| 1765 |
"step": 2290
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 17.692307692307693,
|
| 1769 |
-
"grad_norm": 4.
|
| 1770 |
"learning_rate": 2.307692307692308e-06,
|
| 1771 |
-
"loss": 0.
|
| 1772 |
"step": 2300
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 17.76923076923077,
|
| 1776 |
-
"grad_norm":
|
| 1777 |
"learning_rate": 2.230769230769231e-06,
|
| 1778 |
-
"loss": 0.
|
| 1779 |
"step": 2310
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 17.846153846153847,
|
| 1783 |
-
"grad_norm":
|
| 1784 |
"learning_rate": 2.153846153846154e-06,
|
| 1785 |
-
"loss": 0.
|
| 1786 |
"step": 2320
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 17.923076923076923,
|
| 1790 |
-
"grad_norm":
|
| 1791 |
"learning_rate": 2.0769230769230773e-06,
|
| 1792 |
-
"loss": 0.
|
| 1793 |
"step": 2330
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 18.0,
|
| 1797 |
-
"grad_norm":
|
| 1798 |
"learning_rate": 2.0000000000000003e-06,
|
| 1799 |
-
"loss": 0.
|
| 1800 |
"step": 2340
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 18.0,
|
| 1804 |
-
"eval_accuracy": 0.
|
| 1805 |
-
"eval_loss": 0.
|
| 1806 |
-
"eval_runtime": 0.
|
| 1807 |
-
"eval_samples_per_second":
|
| 1808 |
-
"eval_steps_per_second":
|
| 1809 |
"step": 2340
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 18.076923076923077,
|
| 1813 |
-
"grad_norm":
|
| 1814 |
"learning_rate": 1.9230769230769234e-06,
|
| 1815 |
-
"loss": 0.
|
| 1816 |
"step": 2350
|
| 1817 |
},
|
| 1818 |
{
|
| 1819 |
"epoch": 18.153846153846153,
|
| 1820 |
-
"grad_norm":
|
| 1821 |
"learning_rate": 1.8461538461538465e-06,
|
| 1822 |
-
"loss": 0.
|
| 1823 |
"step": 2360
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
"epoch": 18.23076923076923,
|
| 1827 |
-
"grad_norm":
|
| 1828 |
"learning_rate": 1.7692307692307695e-06,
|
| 1829 |
-
"loss": 0.
|
| 1830 |
"step": 2370
|
| 1831 |
},
|
| 1832 |
{
|
| 1833 |
"epoch": 18.307692307692307,
|
| 1834 |
-
"grad_norm":
|
| 1835 |
"learning_rate": 1.6923076923076926e-06,
|
| 1836 |
-
"loss": 0.
|
| 1837 |
"step": 2380
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
"epoch": 18.384615384615383,
|
| 1841 |
-
"grad_norm":
|
| 1842 |
"learning_rate": 1.6153846153846157e-06,
|
| 1843 |
-
"loss": 0.
|
| 1844 |
"step": 2390
|
| 1845 |
},
|
| 1846 |
{
|
| 1847 |
"epoch": 18.46153846153846,
|
| 1848 |
-
"grad_norm":
|
| 1849 |
"learning_rate": 1.5384615384615387e-06,
|
| 1850 |
-
"loss": 0.
|
| 1851 |
"step": 2400
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
"epoch": 18.53846153846154,
|
| 1855 |
-
"grad_norm":
|
| 1856 |
"learning_rate": 1.4615384615384618e-06,
|
| 1857 |
-
"loss": 0.
|
| 1858 |
"step": 2410
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
"epoch": 18.615384615384617,
|
| 1862 |
-
"grad_norm":
|
| 1863 |
"learning_rate": 1.3846153846153848e-06,
|
| 1864 |
-
"loss": 0.
|
| 1865 |
"step": 2420
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
"epoch": 18.692307692307693,
|
| 1869 |
-
"grad_norm":
|
| 1870 |
"learning_rate": 1.307692307692308e-06,
|
| 1871 |
-
"loss": 0.
|
| 1872 |
"step": 2430
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 18.76923076923077,
|
| 1876 |
-
"grad_norm":
|
| 1877 |
"learning_rate": 1.230769230769231e-06,
|
| 1878 |
-
"loss": 0.
|
| 1879 |
"step": 2440
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
"epoch": 18.846153846153847,
|
| 1883 |
-
"grad_norm":
|
| 1884 |
"learning_rate": 1.153846153846154e-06,
|
| 1885 |
-
"loss": 0.
|
| 1886 |
"step": 2450
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
"epoch": 18.923076923076923,
|
| 1890 |
-
"grad_norm":
|
| 1891 |
"learning_rate": 1.076923076923077e-06,
|
| 1892 |
-
"loss": 0.
|
| 1893 |
"step": 2460
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
"epoch": 19.0,
|
| 1897 |
-
"grad_norm":
|
| 1898 |
"learning_rate": 1.0000000000000002e-06,
|
| 1899 |
-
"loss": 0.
|
| 1900 |
"step": 2470
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
"epoch": 19.0,
|
| 1904 |
-
"eval_accuracy": 0.
|
| 1905 |
-
"eval_loss": 0.
|
| 1906 |
-
"eval_runtime": 0.
|
| 1907 |
-
"eval_samples_per_second":
|
| 1908 |
-
"eval_steps_per_second":
|
| 1909 |
"step": 2470
|
| 1910 |
},
|
| 1911 |
{
|
| 1912 |
"epoch": 19.076923076923077,
|
| 1913 |
-
"grad_norm":
|
| 1914 |
"learning_rate": 9.230769230769232e-07,
|
| 1915 |
-
"loss": 0.
|
| 1916 |
"step": 2480
|
| 1917 |
},
|
| 1918 |
{
|
| 1919 |
"epoch": 19.153846153846153,
|
| 1920 |
-
"grad_norm":
|
| 1921 |
"learning_rate": 8.461538461538463e-07,
|
| 1922 |
-
"loss": 0.
|
| 1923 |
"step": 2490
|
| 1924 |
},
|
| 1925 |
{
|
| 1926 |
"epoch": 19.23076923076923,
|
| 1927 |
-
"grad_norm":
|
| 1928 |
"learning_rate": 7.692307692307694e-07,
|
| 1929 |
-
"loss": 0.
|
| 1930 |
"step": 2500
|
| 1931 |
},
|
| 1932 |
{
|
| 1933 |
"epoch": 19.307692307692307,
|
| 1934 |
-
"grad_norm":
|
| 1935 |
"learning_rate": 6.923076923076924e-07,
|
| 1936 |
-
"loss": 0.
|
| 1937 |
"step": 2510
|
| 1938 |
},
|
| 1939 |
{
|
| 1940 |
"epoch": 19.384615384615383,
|
| 1941 |
-
"grad_norm":
|
| 1942 |
"learning_rate": 6.153846153846155e-07,
|
| 1943 |
-
"loss": 0.
|
| 1944 |
"step": 2520
|
| 1945 |
},
|
| 1946 |
{
|
| 1947 |
"epoch": 19.46153846153846,
|
| 1948 |
-
"grad_norm":
|
| 1949 |
"learning_rate": 5.384615384615386e-07,
|
| 1950 |
-
"loss": 0.
|
| 1951 |
"step": 2530
|
| 1952 |
},
|
| 1953 |
{
|
| 1954 |
"epoch": 19.53846153846154,
|
| 1955 |
-
"grad_norm":
|
| 1956 |
"learning_rate": 4.615384615384616e-07,
|
| 1957 |
-
"loss": 0.
|
| 1958 |
"step": 2540
|
| 1959 |
},
|
| 1960 |
{
|
| 1961 |
"epoch": 19.615384615384617,
|
| 1962 |
-
"grad_norm":
|
| 1963 |
"learning_rate": 3.846153846153847e-07,
|
| 1964 |
-
"loss": 0.
|
| 1965 |
"step": 2550
|
| 1966 |
},
|
| 1967 |
{
|
| 1968 |
"epoch": 19.692307692307693,
|
| 1969 |
-
"grad_norm":
|
| 1970 |
"learning_rate": 3.0769230769230774e-07,
|
| 1971 |
-
"loss": 0.
|
| 1972 |
"step": 2560
|
| 1973 |
},
|
| 1974 |
{
|
| 1975 |
"epoch": 19.76923076923077,
|
| 1976 |
-
"grad_norm":
|
| 1977 |
"learning_rate": 2.307692307692308e-07,
|
| 1978 |
-
"loss": 0.
|
| 1979 |
"step": 2570
|
| 1980 |
},
|
| 1981 |
{
|
| 1982 |
"epoch": 19.846153846153847,
|
| 1983 |
-
"grad_norm":
|
| 1984 |
"learning_rate": 1.5384615384615387e-07,
|
| 1985 |
-
"loss": 0.
|
| 1986 |
"step": 2580
|
| 1987 |
},
|
| 1988 |
{
|
| 1989 |
"epoch": 19.923076923076923,
|
| 1990 |
-
"grad_norm":
|
| 1991 |
"learning_rate": 7.692307692307694e-08,
|
| 1992 |
-
"loss": 0.
|
| 1993 |
"step": 2590
|
| 1994 |
},
|
| 1995 |
{
|
| 1996 |
"epoch": 20.0,
|
| 1997 |
-
"grad_norm":
|
| 1998 |
"learning_rate": 0.0,
|
| 1999 |
-
"loss": 0.
|
| 2000 |
"step": 2600
|
| 2001 |
},
|
| 2002 |
{
|
| 2003 |
"epoch": 20.0,
|
| 2004 |
-
"eval_accuracy": 0.
|
| 2005 |
-
"eval_loss": 0.
|
| 2006 |
-
"eval_runtime":
|
| 2007 |
-
"eval_samples_per_second":
|
| 2008 |
-
"eval_steps_per_second":
|
| 2009 |
"step": 2600
|
| 2010 |
},
|
| 2011 |
{
|
| 2012 |
"epoch": 20.0,
|
| 2013 |
"step": 2600,
|
| 2014 |
-
"total_flos":
|
| 2015 |
-
"train_loss": 0.
|
| 2016 |
-
"train_runtime":
|
| 2017 |
-
"train_samples_per_second":
|
| 2018 |
-
"train_steps_per_second":
|
| 2019 |
}
|
| 2020 |
],
|
| 2021 |
"logging_steps": 10,
|
|
@@ -2035,7 +2035,7 @@
|
|
| 2035 |
"attributes": {}
|
| 2036 |
}
|
| 2037 |
},
|
| 2038 |
-
"total_flos":
|
| 2039 |
"train_batch_size": 8,
|
| 2040 |
"trial_name": null,
|
| 2041 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.49125248193740845,
|
| 3 |
+
"best_model_checkpoint": "./beans_outputs/checkpoint-2340",
|
| 4 |
"epoch": 20.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 2600,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.07692307692307693,
|
| 13 |
+
"grad_norm": 2.631645917892456,
|
| 14 |
+
"learning_rate": 1.9923076923076926e-05,
|
| 15 |
+
"loss": 1.0806,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.15384615384615385,
|
| 20 |
+
"grad_norm": 2.7555034160614014,
|
| 21 |
+
"learning_rate": 1.9846153846153847e-05,
|
| 22 |
+
"loss": 1.0886,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.23076923076923078,
|
| 27 |
+
"grad_norm": 3.3691883087158203,
|
| 28 |
+
"learning_rate": 1.976923076923077e-05,
|
| 29 |
+
"loss": 1.0765,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.3076923076923077,
|
| 34 |
+
"grad_norm": 3.109200954437256,
|
| 35 |
+
"learning_rate": 1.9692307692307696e-05,
|
| 36 |
+
"loss": 1.0931,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.38461538461538464,
|
| 41 |
+
"grad_norm": 2.8181138038635254,
|
| 42 |
+
"learning_rate": 1.9615384615384617e-05,
|
| 43 |
+
"loss": 1.0809,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.46153846153846156,
|
| 48 |
+
"grad_norm": 2.4915010929107666,
|
| 49 |
+
"learning_rate": 1.953846153846154e-05,
|
| 50 |
+
"loss": 1.0718,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.5384615384615384,
|
| 55 |
+
"grad_norm": 3.8907675743103027,
|
| 56 |
+
"learning_rate": 1.9461538461538462e-05,
|
| 57 |
+
"loss": 1.0905,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.6153846153846154,
|
| 62 |
+
"grad_norm": 3.9334473609924316,
|
| 63 |
+
"learning_rate": 1.9384615384615386e-05,
|
| 64 |
+
"loss": 1.09,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.6923076923076923,
|
| 69 |
+
"grad_norm": 2.9074010848999023,
|
| 70 |
+
"learning_rate": 1.930769230769231e-05,
|
| 71 |
+
"loss": 1.0835,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.7692307692307693,
|
| 76 |
+
"grad_norm": 3.810696840286255,
|
| 77 |
+
"learning_rate": 1.923076923076923e-05,
|
| 78 |
+
"loss": 1.0851,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.8461538461538461,
|
| 83 |
+
"grad_norm": 2.3626232147216797,
|
| 84 |
+
"learning_rate": 1.9153846153846156e-05,
|
| 85 |
+
"loss": 1.0751,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.9230769230769231,
|
| 90 |
+
"grad_norm": 2.9881410598754883,
|
| 91 |
+
"learning_rate": 1.907692307692308e-05,
|
| 92 |
+
"loss": 1.0728,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 1.0,
|
| 97 |
+
"grad_norm": 8.272817611694336,
|
| 98 |
+
"learning_rate": 1.9e-05,
|
| 99 |
+
"loss": 1.07,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 1.0,
|
| 104 |
+
"eval_accuracy": 0.41353383458646614,
|
| 105 |
+
"eval_loss": 1.0682790279388428,
|
| 106 |
+
"eval_runtime": 1.1629,
|
| 107 |
+
"eval_samples_per_second": 114.366,
|
| 108 |
+
"eval_steps_per_second": 14.618,
|
| 109 |
"step": 130
|
| 110 |
},
|
| 111 |
{
|
| 112 |
"epoch": 1.0769230769230769,
|
| 113 |
+
"grad_norm": 2.923727035522461,
|
| 114 |
+
"learning_rate": 1.8923076923076925e-05,
|
| 115 |
+
"loss": 1.0742,
|
| 116 |
"step": 140
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 1.1538461538461537,
|
| 120 |
+
"grad_norm": 4.396390914916992,
|
| 121 |
+
"learning_rate": 1.8846153846153846e-05,
|
| 122 |
+
"loss": 1.0647,
|
| 123 |
"step": 150
|
| 124 |
},
|
| 125 |
{
|
| 126 |
"epoch": 1.2307692307692308,
|
| 127 |
+
"grad_norm": 2.367124080657959,
|
| 128 |
+
"learning_rate": 1.876923076923077e-05,
|
| 129 |
+
"loss": 1.0739,
|
| 130 |
"step": 160
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 1.3076923076923077,
|
| 134 |
+
"grad_norm": 3.049018621444702,
|
| 135 |
+
"learning_rate": 1.8692307692307695e-05,
|
| 136 |
+
"loss": 1.0518,
|
| 137 |
"step": 170
|
| 138 |
},
|
| 139 |
{
|
| 140 |
"epoch": 1.3846153846153846,
|
| 141 |
+
"grad_norm": 3.515366554260254,
|
| 142 |
+
"learning_rate": 1.8615384615384616e-05,
|
| 143 |
+
"loss": 1.0581,
|
| 144 |
"step": 180
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.4615384615384617,
|
| 148 |
+
"grad_norm": 3.3465802669525146,
|
| 149 |
+
"learning_rate": 1.853846153846154e-05,
|
| 150 |
+
"loss": 1.0578,
|
| 151 |
"step": 190
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.5384615384615383,
|
| 155 |
+
"grad_norm": 3.190157175064087,
|
| 156 |
+
"learning_rate": 1.8461538461538465e-05,
|
| 157 |
+
"loss": 1.0674,
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 1.6153846153846154,
|
| 162 |
+
"grad_norm": 4.284377574920654,
|
| 163 |
+
"learning_rate": 1.8384615384615386e-05,
|
| 164 |
+
"loss": 1.0656,
|
| 165 |
"step": 210
|
| 166 |
},
|
| 167 |
{
|
| 168 |
"epoch": 1.6923076923076923,
|
| 169 |
+
"grad_norm": 4.362184524536133,
|
| 170 |
+
"learning_rate": 1.830769230769231e-05,
|
| 171 |
+
"loss": 1.0597,
|
| 172 |
"step": 220
|
| 173 |
},
|
| 174 |
{
|
| 175 |
"epoch": 1.7692307692307692,
|
| 176 |
+
"grad_norm": 3.011751651763916,
|
| 177 |
+
"learning_rate": 1.823076923076923e-05,
|
| 178 |
+
"loss": 1.05,
|
| 179 |
"step": 230
|
| 180 |
},
|
| 181 |
{
|
| 182 |
"epoch": 1.8461538461538463,
|
| 183 |
+
"grad_norm": 2.408613920211792,
|
| 184 |
+
"learning_rate": 1.8153846153846155e-05,
|
| 185 |
+
"loss": 1.0617,
|
| 186 |
"step": 240
|
| 187 |
},
|
| 188 |
{
|
| 189 |
"epoch": 1.9230769230769231,
|
| 190 |
+
"grad_norm": 16.68673324584961,
|
| 191 |
+
"learning_rate": 1.807692307692308e-05,
|
| 192 |
+
"loss": 1.0474,
|
| 193 |
"step": 250
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 2.0,
|
| 197 |
+
"grad_norm": 6.727813243865967,
|
| 198 |
+
"learning_rate": 1.8e-05,
|
| 199 |
+
"loss": 1.0523,
|
| 200 |
"step": 260
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 2.0,
|
| 204 |
+
"eval_accuracy": 0.6240601503759399,
|
| 205 |
+
"eval_loss": 1.035576581954956,
|
| 206 |
+
"eval_runtime": 0.9256,
|
| 207 |
+
"eval_samples_per_second": 143.683,
|
| 208 |
+
"eval_steps_per_second": 18.365,
|
| 209 |
"step": 260
|
| 210 |
},
|
| 211 |
{
|
| 212 |
"epoch": 2.076923076923077,
|
| 213 |
+
"grad_norm": 2.9416935443878174,
|
| 214 |
+
"learning_rate": 1.7923076923076925e-05,
|
| 215 |
+
"loss": 1.0505,
|
| 216 |
"step": 270
|
| 217 |
},
|
| 218 |
{
|
| 219 |
"epoch": 2.1538461538461537,
|
| 220 |
+
"grad_norm": 3.15810489654541,
|
| 221 |
+
"learning_rate": 1.784615384615385e-05,
|
| 222 |
+
"loss": 1.0492,
|
| 223 |
"step": 280
|
| 224 |
},
|
| 225 |
{
|
| 226 |
"epoch": 2.230769230769231,
|
| 227 |
+
"grad_norm": 3.2049734592437744,
|
| 228 |
+
"learning_rate": 1.776923076923077e-05,
|
| 229 |
+
"loss": 1.0467,
|
| 230 |
"step": 290
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 2.3076923076923075,
|
| 234 |
+
"grad_norm": 3.3393309116363525,
|
| 235 |
+
"learning_rate": 1.7692307692307694e-05,
|
| 236 |
+
"loss": 1.0297,
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 2.3846153846153846,
|
| 241 |
+
"grad_norm": 2.938563346862793,
|
| 242 |
+
"learning_rate": 1.7615384615384615e-05,
|
| 243 |
+
"loss": 1.0442,
|
| 244 |
"step": 310
|
| 245 |
},
|
| 246 |
{
|
| 247 |
"epoch": 2.4615384615384617,
|
| 248 |
+
"grad_norm": 2.6086559295654297,
|
| 249 |
+
"learning_rate": 1.753846153846154e-05,
|
| 250 |
+
"loss": 1.0506,
|
| 251 |
"step": 320
|
| 252 |
},
|
| 253 |
{
|
| 254 |
"epoch": 2.5384615384615383,
|
| 255 |
+
"grad_norm": 3.588426113128662,
|
| 256 |
+
"learning_rate": 1.7461538461538464e-05,
|
| 257 |
+
"loss": 1.0351,
|
| 258 |
"step": 330
|
| 259 |
},
|
| 260 |
{
|
| 261 |
"epoch": 2.6153846153846154,
|
| 262 |
+
"grad_norm": 2.821547269821167,
|
| 263 |
+
"learning_rate": 1.7384615384615385e-05,
|
| 264 |
+
"loss": 1.0213,
|
| 265 |
"step": 340
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"epoch": 2.6923076923076925,
|
| 269 |
+
"grad_norm": 3.074611186981201,
|
| 270 |
+
"learning_rate": 1.730769230769231e-05,
|
| 271 |
+
"loss": 1.0406,
|
| 272 |
"step": 350
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 2.769230769230769,
|
| 276 |
+
"grad_norm": 2.6791045665740967,
|
| 277 |
+
"learning_rate": 1.7230769230769234e-05,
|
| 278 |
+
"loss": 1.0237,
|
| 279 |
"step": 360
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 2.8461538461538463,
|
| 283 |
+
"grad_norm": 2.8163557052612305,
|
| 284 |
+
"learning_rate": 1.7153846153846155e-05,
|
| 285 |
+
"loss": 1.0337,
|
| 286 |
"step": 370
|
| 287 |
},
|
| 288 |
{
|
| 289 |
"epoch": 2.9230769230769234,
|
| 290 |
+
"grad_norm": 3.1033220291137695,
|
| 291 |
+
"learning_rate": 1.707692307692308e-05,
|
| 292 |
+
"loss": 1.0288,
|
| 293 |
"step": 380
|
| 294 |
},
|
| 295 |
{
|
| 296 |
"epoch": 3.0,
|
| 297 |
+
"grad_norm": 6.048186302185059,
|
| 298 |
+
"learning_rate": 1.7e-05,
|
| 299 |
+
"loss": 1.0439,
|
| 300 |
"step": 390
|
| 301 |
},
|
| 302 |
{
|
| 303 |
"epoch": 3.0,
|
| 304 |
+
"eval_accuracy": 0.6616541353383458,
|
| 305 |
+
"eval_loss": 1.0045464038848877,
|
| 306 |
+
"eval_runtime": 0.9488,
|
| 307 |
+
"eval_samples_per_second": 140.174,
|
| 308 |
+
"eval_steps_per_second": 17.917,
|
| 309 |
"step": 390
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 3.076923076923077,
|
| 313 |
+
"grad_norm": 3.378181219100952,
|
| 314 |
+
"learning_rate": 1.6923076923076924e-05,
|
| 315 |
+
"loss": 1.0363,
|
| 316 |
"step": 400
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 3.1538461538461537,
|
| 320 |
+
"grad_norm": 3.51971697807312,
|
| 321 |
+
"learning_rate": 1.684615384615385e-05,
|
| 322 |
+
"loss": 1.0182,
|
| 323 |
"step": 410
|
| 324 |
},
|
| 325 |
{
|
| 326 |
"epoch": 3.230769230769231,
|
| 327 |
+
"grad_norm": 3.365443706512451,
|
| 328 |
+
"learning_rate": 1.676923076923077e-05,
|
| 329 |
+
"loss": 1.0095,
|
| 330 |
"step": 420
|
| 331 |
},
|
| 332 |
{
|
| 333 |
"epoch": 3.3076923076923075,
|
| 334 |
+
"grad_norm": 3.7726383209228516,
|
| 335 |
+
"learning_rate": 1.6692307692307694e-05,
|
| 336 |
+
"loss": 1.005,
|
| 337 |
"step": 430
|
| 338 |
},
|
| 339 |
{
|
| 340 |
"epoch": 3.3846153846153846,
|
| 341 |
+
"grad_norm": 4.8340864181518555,
|
| 342 |
+
"learning_rate": 1.6615384615384618e-05,
|
| 343 |
+
"loss": 1.0283,
|
| 344 |
"step": 440
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 3.4615384615384617,
|
| 348 |
+
"grad_norm": 3.5189855098724365,
|
| 349 |
+
"learning_rate": 1.653846153846154e-05,
|
| 350 |
+
"loss": 1.0159,
|
| 351 |
"step": 450
|
| 352 |
},
|
| 353 |
{
|
| 354 |
"epoch": 3.5384615384615383,
|
| 355 |
+
"grad_norm": 3.1230361461639404,
|
| 356 |
+
"learning_rate": 1.6461538461538463e-05,
|
| 357 |
+
"loss": 1.0098,
|
| 358 |
"step": 460
|
| 359 |
},
|
| 360 |
{
|
| 361 |
"epoch": 3.6153846153846154,
|
| 362 |
+
"grad_norm": 2.575528860092163,
|
| 363 |
+
"learning_rate": 1.6384615384615384e-05,
|
| 364 |
+
"loss": 1.0065,
|
| 365 |
"step": 470
|
| 366 |
},
|
| 367 |
{
|
| 368 |
"epoch": 3.6923076923076925,
|
| 369 |
+
"grad_norm": 3.465209722518921,
|
| 370 |
+
"learning_rate": 1.630769230769231e-05,
|
| 371 |
+
"loss": 1.0066,
|
| 372 |
"step": 480
|
| 373 |
},
|
| 374 |
{
|
| 375 |
"epoch": 3.769230769230769,
|
| 376 |
+
"grad_norm": 2.744354486465454,
|
| 377 |
+
"learning_rate": 1.6230769230769233e-05,
|
| 378 |
+
"loss": 0.9909,
|
| 379 |
"step": 490
|
| 380 |
},
|
| 381 |
{
|
| 382 |
"epoch": 3.8461538461538463,
|
| 383 |
+
"grad_norm": 2.649618625640869,
|
| 384 |
+
"learning_rate": 1.6153846153846154e-05,
|
| 385 |
+
"loss": 1.0106,
|
| 386 |
"step": 500
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 3.9230769230769234,
|
| 390 |
+
"grad_norm": 3.7920305728912354,
|
| 391 |
+
"learning_rate": 1.607692307692308e-05,
|
| 392 |
+
"loss": 0.9962,
|
| 393 |
"step": 510
|
| 394 |
},
|
| 395 |
{
|
| 396 |
"epoch": 4.0,
|
| 397 |
+
"grad_norm": 6.38316535949707,
|
| 398 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 399 |
+
"loss": 1.0056,
|
| 400 |
"step": 520
|
| 401 |
},
|
| 402 |
{
|
| 403 |
"epoch": 4.0,
|
| 404 |
+
"eval_accuracy": 0.7293233082706767,
|
| 405 |
+
"eval_loss": 0.9671366810798645,
|
| 406 |
+
"eval_runtime": 0.9722,
|
| 407 |
+
"eval_samples_per_second": 136.81,
|
| 408 |
+
"eval_steps_per_second": 17.487,
|
| 409 |
"step": 520
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 4.076923076923077,
|
| 413 |
+
"grad_norm": 4.276334285736084,
|
| 414 |
+
"learning_rate": 1.5923076923076924e-05,
|
| 415 |
+
"loss": 0.9717,
|
| 416 |
"step": 530
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 4.153846153846154,
|
| 420 |
+
"grad_norm": 3.4239840507507324,
|
| 421 |
+
"learning_rate": 1.5846153846153848e-05,
|
| 422 |
+
"loss": 1.0061,
|
| 423 |
"step": 540
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 4.230769230769231,
|
| 427 |
+
"grad_norm": 3.4140846729278564,
|
| 428 |
+
"learning_rate": 1.576923076923077e-05,
|
| 429 |
+
"loss": 1.0057,
|
| 430 |
"step": 550
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 4.3076923076923075,
|
| 434 |
+
"grad_norm": 3.539780855178833,
|
| 435 |
+
"learning_rate": 1.5692307692307693e-05,
|
| 436 |
+
"loss": 1.0197,
|
| 437 |
"step": 560
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 4.384615384615385,
|
| 441 |
+
"grad_norm": 4.03983211517334,
|
| 442 |
+
"learning_rate": 1.5615384615384618e-05,
|
| 443 |
+
"loss": 0.9655,
|
| 444 |
"step": 570
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 4.461538461538462,
|
| 448 |
+
"grad_norm": 3.4873383045196533,
|
| 449 |
+
"learning_rate": 1.553846153846154e-05,
|
| 450 |
+
"loss": 0.9859,
|
| 451 |
"step": 580
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 4.538461538461538,
|
| 455 |
+
"grad_norm": 3.594426393508911,
|
| 456 |
+
"learning_rate": 1.5461538461538463e-05,
|
| 457 |
+
"loss": 0.9938,
|
| 458 |
"step": 590
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 4.615384615384615,
|
| 462 |
+
"grad_norm": 3.7454657554626465,
|
| 463 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 464 |
+
"loss": 0.97,
|
| 465 |
"step": 600
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 4.6923076923076925,
|
| 469 |
+
"grad_norm": 4.500467300415039,
|
| 470 |
+
"learning_rate": 1.5307692307692308e-05,
|
| 471 |
+
"loss": 0.976,
|
| 472 |
"step": 610
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 4.769230769230769,
|
| 476 |
+
"grad_norm": 2.860618829727173,
|
| 477 |
+
"learning_rate": 1.523076923076923e-05,
|
| 478 |
+
"loss": 1.0125,
|
| 479 |
"step": 620
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 4.846153846153846,
|
| 483 |
+
"grad_norm": 4.1896796226501465,
|
| 484 |
+
"learning_rate": 1.5153846153846155e-05,
|
| 485 |
+
"loss": 0.9496,
|
| 486 |
"step": 630
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 4.923076923076923,
|
| 490 |
+
"grad_norm": 3.3220174312591553,
|
| 491 |
+
"learning_rate": 1.5076923076923078e-05,
|
| 492 |
+
"loss": 0.9541,
|
| 493 |
"step": 640
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 5.0,
|
| 497 |
+
"grad_norm": 7.564222812652588,
|
| 498 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 499 |
+
"loss": 0.9853,
|
| 500 |
"step": 650
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 5.0,
|
| 504 |
+
"eval_accuracy": 0.7894736842105263,
|
| 505 |
+
"eval_loss": 0.9245139956474304,
|
| 506 |
+
"eval_runtime": 0.9408,
|
| 507 |
+
"eval_samples_per_second": 141.37,
|
| 508 |
+
"eval_steps_per_second": 18.07,
|
| 509 |
"step": 650
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 5.076923076923077,
|
| 513 |
+
"grad_norm": 3.3493552207946777,
|
| 514 |
+
"learning_rate": 1.4923076923076925e-05,
|
| 515 |
+
"loss": 0.9562,
|
| 516 |
"step": 660
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 5.153846153846154,
|
| 520 |
+
"grad_norm": 4.18095588684082,
|
| 521 |
+
"learning_rate": 1.4846153846153847e-05,
|
| 522 |
+
"loss": 0.9853,
|
| 523 |
"step": 670
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 5.230769230769231,
|
| 527 |
+
"grad_norm": 3.0175259113311768,
|
| 528 |
+
"learning_rate": 1.4769230769230772e-05,
|
| 529 |
+
"loss": 0.9418,
|
| 530 |
"step": 680
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 5.3076923076923075,
|
| 534 |
+
"grad_norm": 4.064915180206299,
|
| 535 |
+
"learning_rate": 1.4692307692307694e-05,
|
| 536 |
+
"loss": 0.9644,
|
| 537 |
"step": 690
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 5.384615384615385,
|
| 541 |
+
"grad_norm": 4.862594127655029,
|
| 542 |
+
"learning_rate": 1.4615384615384615e-05,
|
| 543 |
+
"loss": 0.9521,
|
| 544 |
"step": 700
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 5.461538461538462,
|
| 548 |
+
"grad_norm": 3.188516855239868,
|
| 549 |
+
"learning_rate": 1.453846153846154e-05,
|
| 550 |
+
"loss": 0.9612,
|
| 551 |
"step": 710
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 5.538461538461538,
|
| 555 |
+
"grad_norm": 3.2765607833862305,
|
| 556 |
+
"learning_rate": 1.4461538461538462e-05,
|
| 557 |
+
"loss": 0.9351,
|
| 558 |
"step": 720
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 5.615384615384615,
|
| 562 |
+
"grad_norm": 2.763471841812134,
|
| 563 |
+
"learning_rate": 1.4384615384615387e-05,
|
| 564 |
+
"loss": 0.956,
|
| 565 |
"step": 730
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 5.6923076923076925,
|
| 569 |
+
"grad_norm": 4.148519515991211,
|
| 570 |
+
"learning_rate": 1.430769230769231e-05,
|
| 571 |
+
"loss": 0.9196,
|
| 572 |
"step": 740
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 5.769230769230769,
|
| 576 |
+
"grad_norm": 3.0915586948394775,
|
| 577 |
+
"learning_rate": 1.4230769230769232e-05,
|
| 578 |
+
"loss": 0.9266,
|
| 579 |
"step": 750
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 5.846153846153846,
|
| 583 |
+
"grad_norm": 3.6650705337524414,
|
| 584 |
+
"learning_rate": 1.4153846153846156e-05,
|
| 585 |
+
"loss": 0.9385,
|
| 586 |
"step": 760
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 5.923076923076923,
|
| 590 |
+
"grad_norm": 4.015691757202148,
|
| 591 |
+
"learning_rate": 1.4076923076923079e-05,
|
| 592 |
+
"loss": 0.9356,
|
| 593 |
"step": 770
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 6.0,
|
| 597 |
+
"grad_norm": 11.575014114379883,
|
| 598 |
+
"learning_rate": 1.4e-05,
|
| 599 |
+
"loss": 0.9581,
|
| 600 |
"step": 780
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 6.0,
|
| 604 |
"eval_accuracy": 0.7819548872180451,
|
| 605 |
+
"eval_loss": 0.8743670582771301,
|
| 606 |
+
"eval_runtime": 0.9201,
|
| 607 |
+
"eval_samples_per_second": 144.551,
|
| 608 |
+
"eval_steps_per_second": 18.477,
|
| 609 |
"step": 780
|
| 610 |
},
|
| 611 |
{
|
| 612 |
"epoch": 6.076923076923077,
|
| 613 |
+
"grad_norm": 3.6309585571289062,
|
| 614 |
+
"learning_rate": 1.3923076923076924e-05,
|
| 615 |
+
"loss": 0.9073,
|
| 616 |
"step": 790
|
| 617 |
},
|
| 618 |
{
|
| 619 |
"epoch": 6.153846153846154,
|
| 620 |
+
"grad_norm": 4.175745010375977,
|
| 621 |
+
"learning_rate": 1.3846153846153847e-05,
|
| 622 |
+
"loss": 0.9238,
|
| 623 |
"step": 800
|
| 624 |
},
|
| 625 |
{
|
| 626 |
"epoch": 6.230769230769231,
|
| 627 |
+
"grad_norm": 3.9674503803253174,
|
| 628 |
+
"learning_rate": 1.3769230769230771e-05,
|
| 629 |
+
"loss": 0.9014,
|
| 630 |
"step": 810
|
| 631 |
},
|
| 632 |
{
|
| 633 |
"epoch": 6.3076923076923075,
|
| 634 |
+
"grad_norm": 3.084416627883911,
|
| 635 |
+
"learning_rate": 1.3692307692307694e-05,
|
| 636 |
+
"loss": 0.9023,
|
| 637 |
"step": 820
|
| 638 |
},
|
| 639 |
{
|
| 640 |
"epoch": 6.384615384615385,
|
| 641 |
+
"grad_norm": 4.2650580406188965,
|
| 642 |
+
"learning_rate": 1.3615384615384616e-05,
|
| 643 |
+
"loss": 0.8993,
|
| 644 |
"step": 830
|
| 645 |
},
|
| 646 |
{
|
| 647 |
"epoch": 6.461538461538462,
|
| 648 |
+
"grad_norm": 4.604613304138184,
|
| 649 |
+
"learning_rate": 1.353846153846154e-05,
|
| 650 |
+
"loss": 0.8902,
|
| 651 |
"step": 840
|
| 652 |
},
|
| 653 |
{
|
| 654 |
"epoch": 6.538461538461538,
|
| 655 |
+
"grad_norm": 4.572556018829346,
|
| 656 |
+
"learning_rate": 1.3461538461538463e-05,
|
| 657 |
+
"loss": 0.9039,
|
| 658 |
"step": 850
|
| 659 |
},
|
| 660 |
{
|
| 661 |
"epoch": 6.615384615384615,
|
| 662 |
+
"grad_norm": 4.689060211181641,
|
| 663 |
+
"learning_rate": 1.3384615384615384e-05,
|
| 664 |
+
"loss": 0.8743,
|
| 665 |
"step": 860
|
| 666 |
},
|
| 667 |
{
|
| 668 |
"epoch": 6.6923076923076925,
|
| 669 |
+
"grad_norm": 4.0536394119262695,
|
| 670 |
+
"learning_rate": 1.3307692307692309e-05,
|
| 671 |
+
"loss": 0.9188,
|
| 672 |
"step": 870
|
| 673 |
},
|
| 674 |
{
|
| 675 |
"epoch": 6.769230769230769,
|
| 676 |
+
"grad_norm": 3.3945388793945312,
|
| 677 |
+
"learning_rate": 1.3230769230769231e-05,
|
| 678 |
+
"loss": 0.9175,
|
| 679 |
"step": 880
|
| 680 |
},
|
| 681 |
{
|
| 682 |
"epoch": 6.846153846153846,
|
| 683 |
+
"grad_norm": 4.679390907287598,
|
| 684 |
+
"learning_rate": 1.3153846153846156e-05,
|
| 685 |
+
"loss": 0.8946,
|
| 686 |
"step": 890
|
| 687 |
},
|
| 688 |
{
|
| 689 |
"epoch": 6.923076923076923,
|
| 690 |
+
"grad_norm": 3.4296305179595947,
|
| 691 |
+
"learning_rate": 1.3076923076923078e-05,
|
| 692 |
+
"loss": 0.8895,
|
| 693 |
"step": 900
|
| 694 |
},
|
| 695 |
{
|
| 696 |
"epoch": 7.0,
|
| 697 |
+
"grad_norm": 6.18251895904541,
|
| 698 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 699 |
+
"loss": 0.9044,
|
| 700 |
"step": 910
|
| 701 |
},
|
| 702 |
{
|
| 703 |
"epoch": 7.0,
|
| 704 |
+
"eval_accuracy": 0.7819548872180451,
|
| 705 |
+
"eval_loss": 0.8171929717063904,
|
| 706 |
+
"eval_runtime": 0.9448,
|
| 707 |
+
"eval_samples_per_second": 140.765,
|
| 708 |
+
"eval_steps_per_second": 17.993,
|
| 709 |
"step": 910
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 7.076923076923077,
|
| 713 |
+
"grad_norm": 4.493593692779541,
|
| 714 |
+
"learning_rate": 1.2923076923076925e-05,
|
| 715 |
+
"loss": 0.9023,
|
| 716 |
"step": 920
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 7.153846153846154,
|
| 720 |
+
"grad_norm": 3.8425426483154297,
|
| 721 |
+
"learning_rate": 1.2846153846153848e-05,
|
| 722 |
+
"loss": 0.8691,
|
| 723 |
"step": 930
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 7.230769230769231,
|
| 727 |
+
"grad_norm": 4.417716026306152,
|
| 728 |
+
"learning_rate": 1.2769230769230769e-05,
|
| 729 |
+
"loss": 0.8817,
|
| 730 |
"step": 940
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 7.3076923076923075,
|
| 734 |
+
"grad_norm": 4.091881275177002,
|
| 735 |
+
"learning_rate": 1.2692307692307693e-05,
|
| 736 |
+
"loss": 0.8901,
|
| 737 |
"step": 950
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 7.384615384615385,
|
| 741 |
+
"grad_norm": 4.237718105316162,
|
| 742 |
+
"learning_rate": 1.2615384615384616e-05,
|
| 743 |
+
"loss": 0.863,
|
| 744 |
"step": 960
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 7.461538461538462,
|
| 748 |
+
"grad_norm": 6.6138410568237305,
|
| 749 |
+
"learning_rate": 1.253846153846154e-05,
|
| 750 |
+
"loss": 0.8556,
|
| 751 |
"step": 970
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 7.538461538461538,
|
| 755 |
+
"grad_norm": 4.715096473693848,
|
| 756 |
+
"learning_rate": 1.2461538461538463e-05,
|
| 757 |
+
"loss": 0.8851,
|
| 758 |
"step": 980
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 7.615384615384615,
|
| 762 |
+
"grad_norm": 3.7215723991394043,
|
| 763 |
+
"learning_rate": 1.2384615384615385e-05,
|
| 764 |
+
"loss": 0.8334,
|
| 765 |
"step": 990
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 7.6923076923076925,
|
| 769 |
+
"grad_norm": 3.3076252937316895,
|
| 770 |
+
"learning_rate": 1.230769230769231e-05,
|
| 771 |
+
"loss": 0.8542,
|
| 772 |
"step": 1000
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 7.769230769230769,
|
| 776 |
+
"grad_norm": 4.20552921295166,
|
| 777 |
+
"learning_rate": 1.2230769230769232e-05,
|
| 778 |
+
"loss": 0.8271,
|
| 779 |
"step": 1010
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 7.846153846153846,
|
| 783 |
+
"grad_norm": 4.651803016662598,
|
| 784 |
+
"learning_rate": 1.2153846153846153e-05,
|
| 785 |
+
"loss": 0.8346,
|
| 786 |
"step": 1020
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 7.923076923076923,
|
| 790 |
+
"grad_norm": 4.573400974273682,
|
| 791 |
+
"learning_rate": 1.2076923076923078e-05,
|
| 792 |
+
"loss": 0.8315,
|
| 793 |
"step": 1030
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 8.0,
|
| 797 |
+
"grad_norm": 6.5735182762146,
|
| 798 |
+
"learning_rate": 1.2e-05,
|
| 799 |
+
"loss": 0.869,
|
| 800 |
"step": 1040
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 8.0,
|
| 804 |
+
"eval_accuracy": 0.8270676691729323,
|
| 805 |
+
"eval_loss": 0.773723304271698,
|
| 806 |
+
"eval_runtime": 0.9363,
|
| 807 |
+
"eval_samples_per_second": 142.041,
|
| 808 |
+
"eval_steps_per_second": 18.156,
|
| 809 |
"step": 1040
|
| 810 |
},
|
| 811 |
{
|
| 812 |
"epoch": 8.076923076923077,
|
| 813 |
+
"grad_norm": 4.215645790100098,
|
| 814 |
+
"learning_rate": 1.1923076923076925e-05,
|
| 815 |
+
"loss": 0.8822,
|
| 816 |
"step": 1050
|
| 817 |
},
|
| 818 |
{
|
| 819 |
"epoch": 8.153846153846153,
|
| 820 |
+
"grad_norm": 3.5975663661956787,
|
| 821 |
+
"learning_rate": 1.1846153846153847e-05,
|
| 822 |
+
"loss": 0.8305,
|
| 823 |
"step": 1060
|
| 824 |
},
|
| 825 |
{
|
| 826 |
"epoch": 8.23076923076923,
|
| 827 |
+
"grad_norm": 4.703883647918701,
|
| 828 |
+
"learning_rate": 1.176923076923077e-05,
|
| 829 |
+
"loss": 0.8559,
|
| 830 |
"step": 1070
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 8.307692307692308,
|
| 834 |
+
"grad_norm": 3.2407641410827637,
|
| 835 |
+
"learning_rate": 1.1692307692307694e-05,
|
| 836 |
+
"loss": 0.8117,
|
| 837 |
"step": 1080
|
| 838 |
},
|
| 839 |
{
|
| 840 |
"epoch": 8.384615384615385,
|
| 841 |
+
"grad_norm": 4.487489700317383,
|
| 842 |
+
"learning_rate": 1.1615384615384617e-05,
|
| 843 |
+
"loss": 0.8241,
|
| 844 |
"step": 1090
|
| 845 |
},
|
| 846 |
{
|
| 847 |
"epoch": 8.461538461538462,
|
| 848 |
+
"grad_norm": 3.8978068828582764,
|
| 849 |
+
"learning_rate": 1.1538461538461538e-05,
|
| 850 |
+
"loss": 0.8537,
|
| 851 |
"step": 1100
|
| 852 |
},
|
| 853 |
{
|
| 854 |
"epoch": 8.538461538461538,
|
| 855 |
+
"grad_norm": 2.5178143978118896,
|
| 856 |
+
"learning_rate": 1.1461538461538462e-05,
|
| 857 |
+
"loss": 0.8062,
|
| 858 |
"step": 1110
|
| 859 |
},
|
| 860 |
{
|
| 861 |
"epoch": 8.615384615384615,
|
| 862 |
+
"grad_norm": 4.473972320556641,
|
| 863 |
+
"learning_rate": 1.1384615384615385e-05,
|
| 864 |
+
"loss": 0.8532,
|
| 865 |
"step": 1120
|
| 866 |
},
|
| 867 |
{
|
| 868 |
"epoch": 8.692307692307692,
|
| 869 |
+
"grad_norm": 5.469933032989502,
|
| 870 |
+
"learning_rate": 1.1307692307692309e-05,
|
| 871 |
+
"loss": 0.8532,
|
| 872 |
"step": 1130
|
| 873 |
},
|
| 874 |
{
|
| 875 |
"epoch": 8.76923076923077,
|
| 876 |
+
"grad_norm": 2.7200920581817627,
|
| 877 |
+
"learning_rate": 1.1230769230769232e-05,
|
| 878 |
+
"loss": 0.8203,
|
| 879 |
"step": 1140
|
| 880 |
},
|
| 881 |
{
|
| 882 |
"epoch": 8.846153846153847,
|
| 883 |
+
"grad_norm": 3.509950876235962,
|
| 884 |
+
"learning_rate": 1.1153846153846154e-05,
|
| 885 |
+
"loss": 0.82,
|
| 886 |
"step": 1150
|
| 887 |
},
|
| 888 |
{
|
| 889 |
"epoch": 8.923076923076923,
|
| 890 |
+
"grad_norm": 3.90140962600708,
|
| 891 |
+
"learning_rate": 1.1076923076923079e-05,
|
| 892 |
+
"loss": 0.8182,
|
| 893 |
"step": 1160
|
| 894 |
},
|
| 895 |
{
|
| 896 |
"epoch": 9.0,
|
| 897 |
+
"grad_norm": 16.341764450073242,
|
| 898 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 899 |
+
"loss": 0.8804,
|
| 900 |
"step": 1170
|
| 901 |
},
|
| 902 |
{
|
| 903 |
"epoch": 9.0,
|
| 904 |
+
"eval_accuracy": 0.8270676691729323,
|
| 905 |
+
"eval_loss": 0.7098143100738525,
|
| 906 |
+
"eval_runtime": 0.9276,
|
| 907 |
+
"eval_samples_per_second": 143.385,
|
| 908 |
+
"eval_steps_per_second": 18.327,
|
| 909 |
"step": 1170
|
| 910 |
},
|
| 911 |
{
|
| 912 |
"epoch": 9.076923076923077,
|
| 913 |
+
"grad_norm": 5.392989635467529,
|
| 914 |
+
"learning_rate": 1.0923076923076922e-05,
|
| 915 |
+
"loss": 0.8143,
|
| 916 |
"step": 1180
|
| 917 |
},
|
| 918 |
{
|
| 919 |
"epoch": 9.153846153846153,
|
| 920 |
+
"grad_norm": 3.0304982662200928,
|
| 921 |
+
"learning_rate": 1.0846153846153847e-05,
|
| 922 |
+
"loss": 0.7586,
|
| 923 |
"step": 1190
|
| 924 |
},
|
| 925 |
{
|
| 926 |
"epoch": 9.23076923076923,
|
| 927 |
+
"grad_norm": 3.1372382640838623,
|
| 928 |
+
"learning_rate": 1.076923076923077e-05,
|
| 929 |
+
"loss": 0.7662,
|
| 930 |
"step": 1200
|
| 931 |
},
|
| 932 |
{
|
| 933 |
"epoch": 9.307692307692308,
|
| 934 |
+
"grad_norm": 3.1745128631591797,
|
| 935 |
+
"learning_rate": 1.0692307692307694e-05,
|
| 936 |
+
"loss": 0.8316,
|
| 937 |
"step": 1210
|
| 938 |
},
|
| 939 |
{
|
| 940 |
"epoch": 9.384615384615385,
|
| 941 |
+
"grad_norm": 6.142508029937744,
|
| 942 |
+
"learning_rate": 1.0615384615384616e-05,
|
| 943 |
+
"loss": 0.799,
|
| 944 |
"step": 1220
|
| 945 |
},
|
| 946 |
{
|
| 947 |
"epoch": 9.461538461538462,
|
| 948 |
+
"grad_norm": 4.660125255584717,
|
| 949 |
+
"learning_rate": 1.0538461538461539e-05,
|
| 950 |
+
"loss": 0.7981,
|
| 951 |
"step": 1230
|
| 952 |
},
|
| 953 |
{
|
| 954 |
"epoch": 9.538461538461538,
|
| 955 |
+
"grad_norm": 3.9741530418395996,
|
| 956 |
+
"learning_rate": 1.0461538461538463e-05,
|
| 957 |
+
"loss": 0.7887,
|
| 958 |
"step": 1240
|
| 959 |
},
|
| 960 |
{
|
| 961 |
"epoch": 9.615384615384615,
|
| 962 |
+
"grad_norm": 4.778607368469238,
|
| 963 |
+
"learning_rate": 1.0384615384615386e-05,
|
| 964 |
+
"loss": 0.7304,
|
| 965 |
"step": 1250
|
| 966 |
},
|
| 967 |
{
|
| 968 |
"epoch": 9.692307692307692,
|
| 969 |
+
"grad_norm": 4.286049842834473,
|
| 970 |
+
"learning_rate": 1.0307692307692307e-05,
|
| 971 |
+
"loss": 0.8002,
|
| 972 |
"step": 1260
|
| 973 |
},
|
| 974 |
{
|
| 975 |
"epoch": 9.76923076923077,
|
| 976 |
+
"grad_norm": 3.7198238372802734,
|
| 977 |
+
"learning_rate": 1.0230769230769231e-05,
|
| 978 |
+
"loss": 0.7861,
|
| 979 |
"step": 1270
|
| 980 |
},
|
| 981 |
{
|
| 982 |
"epoch": 9.846153846153847,
|
| 983 |
+
"grad_norm": 3.3311705589294434,
|
| 984 |
+
"learning_rate": 1.0153846153846154e-05,
|
| 985 |
+
"loss": 0.7667,
|
| 986 |
"step": 1280
|
| 987 |
},
|
| 988 |
{
|
| 989 |
"epoch": 9.923076923076923,
|
| 990 |
+
"grad_norm": 4.488588333129883,
|
| 991 |
+
"learning_rate": 1.0076923076923078e-05,
|
| 992 |
+
"loss": 0.8,
|
| 993 |
"step": 1290
|
| 994 |
},
|
| 995 |
{
|
| 996 |
"epoch": 10.0,
|
| 997 |
+
"grad_norm": 8.365084648132324,
|
| 998 |
+
"learning_rate": 1e-05,
|
| 999 |
+
"loss": 0.7757,
|
| 1000 |
"step": 1300
|
| 1001 |
},
|
| 1002 |
{
|
| 1003 |
"epoch": 10.0,
|
| 1004 |
+
"eval_accuracy": 0.8120300751879699,
|
| 1005 |
+
"eval_loss": 0.6705044507980347,
|
| 1006 |
+
"eval_runtime": 0.9309,
|
| 1007 |
+
"eval_samples_per_second": 142.866,
|
| 1008 |
+
"eval_steps_per_second": 18.261,
|
| 1009 |
"step": 1300
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 10.076923076923077,
|
| 1013 |
+
"grad_norm": 4.813228130340576,
|
| 1014 |
+
"learning_rate": 9.923076923076923e-06,
|
| 1015 |
+
"loss": 0.7808,
|
| 1016 |
"step": 1310
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 10.153846153846153,
|
| 1020 |
+
"grad_norm": 3.86871337890625,
|
| 1021 |
+
"learning_rate": 9.846153846153848e-06,
|
| 1022 |
+
"loss": 0.7749,
|
| 1023 |
"step": 1320
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 10.23076923076923,
|
| 1027 |
+
"grad_norm": 4.498291492462158,
|
| 1028 |
+
"learning_rate": 9.76923076923077e-06,
|
| 1029 |
+
"loss": 0.7842,
|
| 1030 |
"step": 1330
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 10.307692307692308,
|
| 1034 |
+
"grad_norm": 7.416966438293457,
|
| 1035 |
+
"learning_rate": 9.692307692307693e-06,
|
| 1036 |
+
"loss": 0.7397,
|
| 1037 |
"step": 1340
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 10.384615384615385,
|
| 1041 |
+
"grad_norm": 3.8194034099578857,
|
| 1042 |
+
"learning_rate": 9.615384615384616e-06,
|
| 1043 |
+
"loss": 0.7101,
|
| 1044 |
"step": 1350
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 10.461538461538462,
|
| 1048 |
+
"grad_norm": 4.42163610458374,
|
| 1049 |
+
"learning_rate": 9.53846153846154e-06,
|
| 1050 |
+
"loss": 0.7694,
|
| 1051 |
"step": 1360
|
| 1052 |
},
|
| 1053 |
{
|
| 1054 |
"epoch": 10.538461538461538,
|
| 1055 |
+
"grad_norm": 4.710392951965332,
|
| 1056 |
+
"learning_rate": 9.461538461538463e-06,
|
| 1057 |
+
"loss": 0.7536,
|
| 1058 |
"step": 1370
|
| 1059 |
},
|
| 1060 |
{
|
| 1061 |
"epoch": 10.615384615384615,
|
| 1062 |
+
"grad_norm": 4.0583906173706055,
|
| 1063 |
+
"learning_rate": 9.384615384615385e-06,
|
| 1064 |
+
"loss": 0.7591,
|
| 1065 |
"step": 1380
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
"epoch": 10.692307692307692,
|
| 1069 |
+
"grad_norm": 4.696585655212402,
|
| 1070 |
+
"learning_rate": 9.307692307692308e-06,
|
| 1071 |
+
"loss": 0.7469,
|
| 1072 |
"step": 1390
|
| 1073 |
},
|
| 1074 |
{
|
| 1075 |
"epoch": 10.76923076923077,
|
| 1076 |
+
"grad_norm": 4.995838642120361,
|
| 1077 |
+
"learning_rate": 9.230769230769232e-06,
|
| 1078 |
+
"loss": 0.776,
|
| 1079 |
"step": 1400
|
| 1080 |
},
|
| 1081 |
{
|
| 1082 |
"epoch": 10.846153846153847,
|
| 1083 |
+
"grad_norm": 3.7382941246032715,
|
| 1084 |
+
"learning_rate": 9.153846153846155e-06,
|
| 1085 |
+
"loss": 0.7281,
|
| 1086 |
"step": 1410
|
| 1087 |
},
|
| 1088 |
{
|
| 1089 |
"epoch": 10.923076923076923,
|
| 1090 |
+
"grad_norm": 4.595252513885498,
|
| 1091 |
+
"learning_rate": 9.076923076923078e-06,
|
| 1092 |
+
"loss": 0.7781,
|
| 1093 |
"step": 1420
|
| 1094 |
},
|
| 1095 |
{
|
| 1096 |
"epoch": 11.0,
|
| 1097 |
+
"grad_norm": 10.765867233276367,
|
| 1098 |
+
"learning_rate": 9e-06,
|
| 1099 |
+
"loss": 0.7694,
|
| 1100 |
"step": 1430
|
| 1101 |
},
|
| 1102 |
{
|
| 1103 |
"epoch": 11.0,
|
| 1104 |
+
"eval_accuracy": 0.8571428571428571,
|
| 1105 |
+
"eval_loss": 0.638173520565033,
|
| 1106 |
+
"eval_runtime": 0.9524,
|
| 1107 |
+
"eval_samples_per_second": 139.647,
|
| 1108 |
+
"eval_steps_per_second": 17.85,
|
| 1109 |
"step": 1430
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 11.076923076923077,
|
| 1113 |
+
"grad_norm": 4.3198041915893555,
|
| 1114 |
+
"learning_rate": 8.923076923076925e-06,
|
| 1115 |
+
"loss": 0.7603,
|
| 1116 |
"step": 1440
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 11.153846153846153,
|
| 1120 |
+
"grad_norm": 6.03615140914917,
|
| 1121 |
+
"learning_rate": 8.846153846153847e-06,
|
| 1122 |
+
"loss": 0.7594,
|
| 1123 |
"step": 1450
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 11.23076923076923,
|
| 1127 |
+
"grad_norm": 4.512632846832275,
|
| 1128 |
+
"learning_rate": 8.76923076923077e-06,
|
| 1129 |
+
"loss": 0.7111,
|
| 1130 |
"step": 1460
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 11.307692307692308,
|
| 1134 |
+
"grad_norm": 3.5640311241149902,
|
| 1135 |
+
"learning_rate": 8.692307692307692e-06,
|
| 1136 |
+
"loss": 0.7783,
|
| 1137 |
"step": 1470
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 11.384615384615385,
|
| 1141 |
+
"grad_norm": 3.612410306930542,
|
| 1142 |
+
"learning_rate": 8.615384615384617e-06,
|
| 1143 |
+
"loss": 0.7285,
|
| 1144 |
"step": 1480
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 11.461538461538462,
|
| 1148 |
+
"grad_norm": 5.0590996742248535,
|
| 1149 |
+
"learning_rate": 8.53846153846154e-06,
|
| 1150 |
+
"loss": 0.7128,
|
| 1151 |
"step": 1490
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 11.538461538461538,
|
| 1155 |
+
"grad_norm": 4.820272445678711,
|
| 1156 |
+
"learning_rate": 8.461538461538462e-06,
|
| 1157 |
+
"loss": 0.798,
|
| 1158 |
"step": 1500
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 11.615384615384615,
|
| 1162 |
+
"grad_norm": 3.6416468620300293,
|
| 1163 |
+
"learning_rate": 8.384615384615385e-06,
|
| 1164 |
+
"loss": 0.7408,
|
| 1165 |
"step": 1510
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 11.692307692307692,
|
| 1169 |
+
"grad_norm": 4.562481880187988,
|
| 1170 |
+
"learning_rate": 8.307692307692309e-06,
|
| 1171 |
+
"loss": 0.7682,
|
| 1172 |
"step": 1520
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 11.76923076923077,
|
| 1176 |
+
"grad_norm": 3.1171770095825195,
|
| 1177 |
+
"learning_rate": 8.230769230769232e-06,
|
| 1178 |
+
"loss": 0.7337,
|
| 1179 |
"step": 1530
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 11.846153846153847,
|
| 1183 |
+
"grad_norm": 3.2231085300445557,
|
| 1184 |
+
"learning_rate": 8.153846153846154e-06,
|
| 1185 |
+
"loss": 0.7348,
|
| 1186 |
"step": 1540
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 11.923076923076923,
|
| 1190 |
+
"grad_norm": 3.5213522911071777,
|
| 1191 |
+
"learning_rate": 8.076923076923077e-06,
|
| 1192 |
+
"loss": 0.7064,
|
| 1193 |
"step": 1550
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 12.0,
|
| 1197 |
+
"grad_norm": 7.278842449188232,
|
| 1198 |
+
"learning_rate": 8.000000000000001e-06,
|
| 1199 |
+
"loss": 0.7966,
|
| 1200 |
"step": 1560
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 12.0,
|
| 1204 |
+
"eval_accuracy": 0.7894736842105263,
|
| 1205 |
+
"eval_loss": 0.6087508797645569,
|
| 1206 |
+
"eval_runtime": 0.9347,
|
| 1207 |
+
"eval_samples_per_second": 142.285,
|
| 1208 |
+
"eval_steps_per_second": 18.187,
|
| 1209 |
"step": 1560
|
| 1210 |
},
|
| 1211 |
{
|
| 1212 |
"epoch": 12.076923076923077,
|
| 1213 |
+
"grad_norm": 8.028270721435547,
|
| 1214 |
+
"learning_rate": 7.923076923076924e-06,
|
| 1215 |
+
"loss": 0.7035,
|
| 1216 |
"step": 1570
|
| 1217 |
},
|
| 1218 |
{
|
| 1219 |
"epoch": 12.153846153846153,
|
| 1220 |
+
"grad_norm": 3.4982223510742188,
|
| 1221 |
+
"learning_rate": 7.846153846153847e-06,
|
| 1222 |
+
"loss": 0.6953,
|
| 1223 |
"step": 1580
|
| 1224 |
},
|
| 1225 |
{
|
| 1226 |
"epoch": 12.23076923076923,
|
| 1227 |
+
"grad_norm": 3.474436044692993,
|
| 1228 |
+
"learning_rate": 7.76923076923077e-06,
|
| 1229 |
+
"loss": 0.7272,
|
| 1230 |
"step": 1590
|
| 1231 |
},
|
| 1232 |
{
|
| 1233 |
"epoch": 12.307692307692308,
|
| 1234 |
+
"grad_norm": 5.147262096405029,
|
| 1235 |
+
"learning_rate": 7.692307692307694e-06,
|
| 1236 |
+
"loss": 0.7639,
|
| 1237 |
"step": 1600
|
| 1238 |
},
|
| 1239 |
{
|
| 1240 |
"epoch": 12.384615384615385,
|
| 1241 |
+
"grad_norm": 3.383554458618164,
|
| 1242 |
+
"learning_rate": 7.615384615384615e-06,
|
| 1243 |
+
"loss": 0.7094,
|
| 1244 |
"step": 1610
|
| 1245 |
},
|
| 1246 |
{
|
| 1247 |
"epoch": 12.461538461538462,
|
| 1248 |
+
"grad_norm": 3.3019802570343018,
|
| 1249 |
+
"learning_rate": 7.538461538461539e-06,
|
| 1250 |
+
"loss": 0.685,
|
| 1251 |
"step": 1620
|
| 1252 |
},
|
| 1253 |
{
|
| 1254 |
"epoch": 12.538461538461538,
|
| 1255 |
+
"grad_norm": 3.602853775024414,
|
| 1256 |
+
"learning_rate": 7.461538461538462e-06,
|
| 1257 |
+
"loss": 0.7068,
|
| 1258 |
"step": 1630
|
| 1259 |
},
|
| 1260 |
{
|
| 1261 |
"epoch": 12.615384615384615,
|
| 1262 |
+
"grad_norm": 3.3312697410583496,
|
| 1263 |
+
"learning_rate": 7.384615384615386e-06,
|
| 1264 |
+
"loss": 0.7045,
|
| 1265 |
"step": 1640
|
| 1266 |
},
|
| 1267 |
{
|
| 1268 |
"epoch": 12.692307692307692,
|
| 1269 |
+
"grad_norm": 6.843470573425293,
|
| 1270 |
+
"learning_rate": 7.307692307692308e-06,
|
| 1271 |
+
"loss": 0.6643,
|
| 1272 |
"step": 1650
|
| 1273 |
},
|
| 1274 |
{
|
| 1275 |
"epoch": 12.76923076923077,
|
| 1276 |
+
"grad_norm": 9.701897621154785,
|
| 1277 |
+
"learning_rate": 7.230769230769231e-06,
|
| 1278 |
+
"loss": 0.8052,
|
| 1279 |
"step": 1660
|
| 1280 |
},
|
| 1281 |
{
|
| 1282 |
"epoch": 12.846153846153847,
|
| 1283 |
+
"grad_norm": 5.923687934875488,
|
| 1284 |
+
"learning_rate": 7.153846153846155e-06,
|
| 1285 |
+
"loss": 0.7118,
|
| 1286 |
"step": 1670
|
| 1287 |
},
|
| 1288 |
{
|
| 1289 |
"epoch": 12.923076923076923,
|
| 1290 |
+
"grad_norm": 3.330617666244507,
|
| 1291 |
+
"learning_rate": 7.076923076923078e-06,
|
| 1292 |
+
"loss": 0.701,
|
| 1293 |
"step": 1680
|
| 1294 |
},
|
| 1295 |
{
|
| 1296 |
"epoch": 13.0,
|
| 1297 |
+
"grad_norm": 10.613458633422852,
|
| 1298 |
+
"learning_rate": 7e-06,
|
| 1299 |
+
"loss": 0.7425,
|
| 1300 |
"step": 1690
|
| 1301 |
},
|
| 1302 |
{
|
| 1303 |
"epoch": 13.0,
|
| 1304 |
+
"eval_accuracy": 0.849624060150376,
|
| 1305 |
+
"eval_loss": 0.572424054145813,
|
| 1306 |
+
"eval_runtime": 0.9506,
|
| 1307 |
+
"eval_samples_per_second": 139.913,
|
| 1308 |
+
"eval_steps_per_second": 17.884,
|
| 1309 |
"step": 1690
|
| 1310 |
},
|
| 1311 |
{
|
| 1312 |
"epoch": 13.076923076923077,
|
| 1313 |
+
"grad_norm": 4.795916557312012,
|
| 1314 |
+
"learning_rate": 6.923076923076923e-06,
|
| 1315 |
+
"loss": 0.7275,
|
| 1316 |
"step": 1700
|
| 1317 |
},
|
| 1318 |
{
|
| 1319 |
"epoch": 13.153846153846153,
|
| 1320 |
+
"grad_norm": 3.3990397453308105,
|
| 1321 |
+
"learning_rate": 6.846153846153847e-06,
|
| 1322 |
+
"loss": 0.737,
|
| 1323 |
"step": 1710
|
| 1324 |
},
|
| 1325 |
{
|
| 1326 |
"epoch": 13.23076923076923,
|
| 1327 |
+
"grad_norm": 5.3640851974487305,
|
| 1328 |
+
"learning_rate": 6.76923076923077e-06,
|
| 1329 |
+
"loss": 0.7653,
|
| 1330 |
"step": 1720
|
| 1331 |
},
|
| 1332 |
{
|
| 1333 |
"epoch": 13.307692307692308,
|
| 1334 |
+
"grad_norm": 2.9321601390838623,
|
| 1335 |
+
"learning_rate": 6.692307692307692e-06,
|
| 1336 |
+
"loss": 0.741,
|
| 1337 |
"step": 1730
|
| 1338 |
},
|
| 1339 |
{
|
| 1340 |
"epoch": 13.384615384615385,
|
| 1341 |
+
"grad_norm": 5.113746166229248,
|
| 1342 |
+
"learning_rate": 6.615384615384616e-06,
|
| 1343 |
+
"loss": 0.6832,
|
| 1344 |
"step": 1740
|
| 1345 |
},
|
| 1346 |
{
|
| 1347 |
"epoch": 13.461538461538462,
|
| 1348 |
+
"grad_norm": 4.589268207550049,
|
| 1349 |
+
"learning_rate": 6.538461538461539e-06,
|
| 1350 |
+
"loss": 0.7463,
|
| 1351 |
"step": 1750
|
| 1352 |
},
|
| 1353 |
{
|
| 1354 |
"epoch": 13.538461538461538,
|
| 1355 |
+
"grad_norm": 3.7653889656066895,
|
| 1356 |
+
"learning_rate": 6.461538461538463e-06,
|
| 1357 |
+
"loss": 0.7569,
|
| 1358 |
"step": 1760
|
| 1359 |
},
|
| 1360 |
{
|
| 1361 |
"epoch": 13.615384615384615,
|
| 1362 |
+
"grad_norm": 4.248018264770508,
|
| 1363 |
+
"learning_rate": 6.384615384615384e-06,
|
| 1364 |
+
"loss": 0.6853,
|
| 1365 |
"step": 1770
|
| 1366 |
},
|
| 1367 |
{
|
| 1368 |
"epoch": 13.692307692307692,
|
| 1369 |
+
"grad_norm": 4.481900691986084,
|
| 1370 |
+
"learning_rate": 6.307692307692308e-06,
|
| 1371 |
+
"loss": 0.6679,
|
| 1372 |
"step": 1780
|
| 1373 |
},
|
| 1374 |
{
|
| 1375 |
"epoch": 13.76923076923077,
|
| 1376 |
+
"grad_norm": 3.7759058475494385,
|
| 1377 |
+
"learning_rate": 6.230769230769231e-06,
|
| 1378 |
+
"loss": 0.7159,
|
| 1379 |
"step": 1790
|
| 1380 |
},
|
| 1381 |
{
|
| 1382 |
"epoch": 13.846153846153847,
|
| 1383 |
+
"grad_norm": 7.013620853424072,
|
| 1384 |
+
"learning_rate": 6.153846153846155e-06,
|
| 1385 |
+
"loss": 0.6663,
|
| 1386 |
"step": 1800
|
| 1387 |
},
|
| 1388 |
{
|
| 1389 |
"epoch": 13.923076923076923,
|
| 1390 |
+
"grad_norm": 3.7434396743774414,
|
| 1391 |
+
"learning_rate": 6.076923076923077e-06,
|
| 1392 |
+
"loss": 0.6043,
|
| 1393 |
"step": 1810
|
| 1394 |
},
|
| 1395 |
{
|
| 1396 |
"epoch": 14.0,
|
| 1397 |
+
"grad_norm": 8.63290023803711,
|
| 1398 |
+
"learning_rate": 6e-06,
|
| 1399 |
+
"loss": 0.7698,
|
| 1400 |
"step": 1820
|
| 1401 |
},
|
| 1402 |
{
|
| 1403 |
"epoch": 14.0,
|
| 1404 |
+
"eval_accuracy": 0.8195488721804511,
|
| 1405 |
+
"eval_loss": 0.5665194392204285,
|
| 1406 |
+
"eval_runtime": 0.9506,
|
| 1407 |
+
"eval_samples_per_second": 139.918,
|
| 1408 |
+
"eval_steps_per_second": 17.884,
|
| 1409 |
"step": 1820
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
"epoch": 14.076923076923077,
|
| 1413 |
+
"grad_norm": 6.073098659515381,
|
| 1414 |
+
"learning_rate": 5.923076923076924e-06,
|
| 1415 |
+
"loss": 0.701,
|
| 1416 |
"step": 1830
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
"epoch": 14.153846153846153,
|
| 1420 |
+
"grad_norm": 4.466890811920166,
|
| 1421 |
+
"learning_rate": 5.846153846153847e-06,
|
| 1422 |
+
"loss": 0.712,
|
| 1423 |
"step": 1840
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
"epoch": 14.23076923076923,
|
| 1427 |
+
"grad_norm": 4.844086170196533,
|
| 1428 |
+
"learning_rate": 5.769230769230769e-06,
|
| 1429 |
+
"loss": 0.6637,
|
| 1430 |
"step": 1850
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 14.307692307692308,
|
| 1434 |
+
"grad_norm": 5.5982184410095215,
|
| 1435 |
+
"learning_rate": 5.692307692307692e-06,
|
| 1436 |
+
"loss": 0.6735,
|
| 1437 |
"step": 1860
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
"epoch": 14.384615384615385,
|
| 1441 |
+
"grad_norm": 6.027699947357178,
|
| 1442 |
+
"learning_rate": 5.615384615384616e-06,
|
| 1443 |
+
"loss": 0.672,
|
| 1444 |
"step": 1870
|
| 1445 |
},
|
| 1446 |
{
|
| 1447 |
"epoch": 14.461538461538462,
|
| 1448 |
+
"grad_norm": 3.128363847732544,
|
| 1449 |
+
"learning_rate": 5.538461538461539e-06,
|
| 1450 |
+
"loss": 0.6864,
|
| 1451 |
"step": 1880
|
| 1452 |
},
|
| 1453 |
{
|
| 1454 |
"epoch": 14.538461538461538,
|
| 1455 |
+
"grad_norm": 5.279370307922363,
|
| 1456 |
+
"learning_rate": 5.461538461538461e-06,
|
| 1457 |
+
"loss": 0.6765,
|
| 1458 |
"step": 1890
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 14.615384615384615,
|
| 1462 |
+
"grad_norm": 3.779651165008545,
|
| 1463 |
+
"learning_rate": 5.384615384615385e-06,
|
| 1464 |
+
"loss": 0.6902,
|
| 1465 |
"step": 1900
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 14.692307692307692,
|
| 1469 |
+
"grad_norm": 7.5796003341674805,
|
| 1470 |
+
"learning_rate": 5.307692307692308e-06,
|
| 1471 |
+
"loss": 0.6971,
|
| 1472 |
"step": 1910
|
| 1473 |
},
|
| 1474 |
{
|
| 1475 |
"epoch": 14.76923076923077,
|
| 1476 |
+
"grad_norm": 5.17158842086792,
|
| 1477 |
+
"learning_rate": 5.230769230769232e-06,
|
| 1478 |
+
"loss": 0.6997,
|
| 1479 |
"step": 1920
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 14.846153846153847,
|
| 1483 |
+
"grad_norm": 3.0330209732055664,
|
| 1484 |
+
"learning_rate": 5.1538461538461534e-06,
|
| 1485 |
+
"loss": 0.778,
|
| 1486 |
"step": 1930
|
| 1487 |
},
|
| 1488 |
{
|
| 1489 |
"epoch": 14.923076923076923,
|
| 1490 |
+
"grad_norm": 7.733983993530273,
|
| 1491 |
+
"learning_rate": 5.076923076923077e-06,
|
| 1492 |
+
"loss": 0.6478,
|
| 1493 |
"step": 1940
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
"epoch": 15.0,
|
| 1497 |
+
"grad_norm": 7.919536590576172,
|
| 1498 |
+
"learning_rate": 5e-06,
|
| 1499 |
+
"loss": 0.6632,
|
| 1500 |
"step": 1950
|
| 1501 |
},
|
| 1502 |
{
|
| 1503 |
"epoch": 15.0,
|
| 1504 |
+
"eval_accuracy": 0.8571428571428571,
|
| 1505 |
+
"eval_loss": 0.5307806730270386,
|
| 1506 |
+
"eval_runtime": 0.9229,
|
| 1507 |
+
"eval_samples_per_second": 144.113,
|
| 1508 |
+
"eval_steps_per_second": 18.421,
|
| 1509 |
"step": 1950
|
| 1510 |
},
|
| 1511 |
{
|
| 1512 |
"epoch": 15.076923076923077,
|
| 1513 |
+
"grad_norm": 4.707977771759033,
|
| 1514 |
"learning_rate": 4.923076923076924e-06,
|
| 1515 |
+
"loss": 0.6156,
|
| 1516 |
"step": 1960
|
| 1517 |
},
|
| 1518 |
{
|
| 1519 |
"epoch": 15.153846153846153,
|
| 1520 |
+
"grad_norm": 5.254363536834717,
|
| 1521 |
"learning_rate": 4.8461538461538465e-06,
|
| 1522 |
+
"loss": 0.6478,
|
| 1523 |
"step": 1970
|
| 1524 |
},
|
| 1525 |
{
|
| 1526 |
"epoch": 15.23076923076923,
|
| 1527 |
+
"grad_norm": 4.477289199829102,
|
| 1528 |
"learning_rate": 4.76923076923077e-06,
|
| 1529 |
+
"loss": 0.6786,
|
| 1530 |
"step": 1980
|
| 1531 |
},
|
| 1532 |
{
|
| 1533 |
"epoch": 15.307692307692308,
|
| 1534 |
+
"grad_norm": 6.381091117858887,
|
| 1535 |
"learning_rate": 4.692307692307693e-06,
|
| 1536 |
+
"loss": 0.7183,
|
| 1537 |
"step": 1990
|
| 1538 |
},
|
| 1539 |
{
|
| 1540 |
"epoch": 15.384615384615385,
|
| 1541 |
+
"grad_norm": 3.8760223388671875,
|
| 1542 |
"learning_rate": 4.615384615384616e-06,
|
| 1543 |
+
"loss": 0.6391,
|
| 1544 |
"step": 2000
|
| 1545 |
},
|
| 1546 |
{
|
| 1547 |
"epoch": 15.461538461538462,
|
| 1548 |
+
"grad_norm": 7.011289119720459,
|
| 1549 |
"learning_rate": 4.538461538461539e-06,
|
| 1550 |
+
"loss": 0.6135,
|
| 1551 |
"step": 2010
|
| 1552 |
},
|
| 1553 |
{
|
| 1554 |
"epoch": 15.538461538461538,
|
| 1555 |
+
"grad_norm": 5.04637336730957,
|
| 1556 |
"learning_rate": 4.461538461538462e-06,
|
| 1557 |
+
"loss": 0.628,
|
| 1558 |
"step": 2020
|
| 1559 |
},
|
| 1560 |
{
|
| 1561 |
"epoch": 15.615384615384615,
|
| 1562 |
+
"grad_norm": 7.818150520324707,
|
| 1563 |
"learning_rate": 4.384615384615385e-06,
|
| 1564 |
+
"loss": 0.6537,
|
| 1565 |
"step": 2030
|
| 1566 |
},
|
| 1567 |
{
|
| 1568 |
"epoch": 15.692307692307692,
|
| 1569 |
+
"grad_norm": 3.9413349628448486,
|
| 1570 |
"learning_rate": 4.307692307692308e-06,
|
| 1571 |
+
"loss": 0.6662,
|
| 1572 |
"step": 2040
|
| 1573 |
},
|
| 1574 |
{
|
| 1575 |
"epoch": 15.76923076923077,
|
| 1576 |
+
"grad_norm": 5.911675930023193,
|
| 1577 |
"learning_rate": 4.230769230769231e-06,
|
| 1578 |
+
"loss": 0.6284,
|
| 1579 |
"step": 2050
|
| 1580 |
},
|
| 1581 |
{
|
| 1582 |
"epoch": 15.846153846153847,
|
| 1583 |
+
"grad_norm": 4.451909065246582,
|
| 1584 |
"learning_rate": 4.1538461538461545e-06,
|
| 1585 |
+
"loss": 0.6906,
|
| 1586 |
"step": 2060
|
| 1587 |
},
|
| 1588 |
{
|
| 1589 |
"epoch": 15.923076923076923,
|
| 1590 |
+
"grad_norm": 5.271285057067871,
|
| 1591 |
"learning_rate": 4.076923076923077e-06,
|
| 1592 |
+
"loss": 0.6453,
|
| 1593 |
"step": 2070
|
| 1594 |
},
|
| 1595 |
{
|
| 1596 |
"epoch": 16.0,
|
| 1597 |
+
"grad_norm": 8.229835510253906,
|
| 1598 |
"learning_rate": 4.000000000000001e-06,
|
| 1599 |
+
"loss": 0.6162,
|
| 1600 |
"step": 2080
|
| 1601 |
},
|
| 1602 |
{
|
| 1603 |
"epoch": 16.0,
|
| 1604 |
+
"eval_accuracy": 0.8345864661654135,
|
| 1605 |
+
"eval_loss": 0.5261984467506409,
|
| 1606 |
+
"eval_runtime": 0.9205,
|
| 1607 |
+
"eval_samples_per_second": 144.487,
|
| 1608 |
+
"eval_steps_per_second": 18.468,
|
| 1609 |
"step": 2080
|
| 1610 |
},
|
| 1611 |
{
|
| 1612 |
"epoch": 16.076923076923077,
|
| 1613 |
+
"grad_norm": 12.219758033752441,
|
| 1614 |
"learning_rate": 3.923076923076923e-06,
|
| 1615 |
+
"loss": 0.7381,
|
| 1616 |
"step": 2090
|
| 1617 |
},
|
| 1618 |
{
|
| 1619 |
"epoch": 16.153846153846153,
|
| 1620 |
+
"grad_norm": 4.024055480957031,
|
| 1621 |
"learning_rate": 3.846153846153847e-06,
|
| 1622 |
+
"loss": 0.708,
|
| 1623 |
"step": 2100
|
| 1624 |
},
|
| 1625 |
{
|
| 1626 |
"epoch": 16.23076923076923,
|
| 1627 |
+
"grad_norm": 5.345643043518066,
|
| 1628 |
"learning_rate": 3.7692307692307694e-06,
|
| 1629 |
+
"loss": 0.7214,
|
| 1630 |
"step": 2110
|
| 1631 |
},
|
| 1632 |
{
|
| 1633 |
"epoch": 16.307692307692307,
|
| 1634 |
+
"grad_norm": 7.9586992263793945,
|
| 1635 |
"learning_rate": 3.692307692307693e-06,
|
| 1636 |
+
"loss": 0.7163,
|
| 1637 |
"step": 2120
|
| 1638 |
},
|
| 1639 |
{
|
| 1640 |
"epoch": 16.384615384615383,
|
| 1641 |
+
"grad_norm": 7.007926940917969,
|
| 1642 |
"learning_rate": 3.6153846153846156e-06,
|
| 1643 |
+
"loss": 0.6692,
|
| 1644 |
"step": 2130
|
| 1645 |
},
|
| 1646 |
{
|
| 1647 |
"epoch": 16.46153846153846,
|
| 1648 |
+
"grad_norm": 8.54269027709961,
|
| 1649 |
"learning_rate": 3.538461538461539e-06,
|
| 1650 |
+
"loss": 0.661,
|
| 1651 |
"step": 2140
|
| 1652 |
},
|
| 1653 |
{
|
| 1654 |
"epoch": 16.53846153846154,
|
| 1655 |
+
"grad_norm": 6.417664051055908,
|
| 1656 |
"learning_rate": 3.4615384615384617e-06,
|
| 1657 |
+
"loss": 0.6186,
|
| 1658 |
"step": 2150
|
| 1659 |
},
|
| 1660 |
{
|
| 1661 |
"epoch": 16.615384615384617,
|
| 1662 |
+
"grad_norm": 3.766784429550171,
|
| 1663 |
"learning_rate": 3.384615384615385e-06,
|
| 1664 |
+
"loss": 0.6254,
|
| 1665 |
"step": 2160
|
| 1666 |
},
|
| 1667 |
{
|
| 1668 |
"epoch": 16.692307692307693,
|
| 1669 |
+
"grad_norm": 4.219268798828125,
|
| 1670 |
"learning_rate": 3.307692307692308e-06,
|
| 1671 |
+
"loss": 0.6308,
|
| 1672 |
"step": 2170
|
| 1673 |
},
|
| 1674 |
{
|
| 1675 |
"epoch": 16.76923076923077,
|
| 1676 |
+
"grad_norm": 6.51754903793335,
|
| 1677 |
"learning_rate": 3.2307692307692313e-06,
|
| 1678 |
+
"loss": 0.7395,
|
| 1679 |
"step": 2180
|
| 1680 |
},
|
| 1681 |
{
|
| 1682 |
"epoch": 16.846153846153847,
|
| 1683 |
+
"grad_norm": 4.084061622619629,
|
| 1684 |
"learning_rate": 3.153846153846154e-06,
|
| 1685 |
+
"loss": 0.7093,
|
| 1686 |
"step": 2190
|
| 1687 |
},
|
| 1688 |
{
|
| 1689 |
"epoch": 16.923076923076923,
|
| 1690 |
+
"grad_norm": 2.852893590927124,
|
| 1691 |
"learning_rate": 3.0769230769230774e-06,
|
| 1692 |
+
"loss": 0.6213,
|
| 1693 |
"step": 2200
|
| 1694 |
},
|
| 1695 |
{
|
| 1696 |
"epoch": 17.0,
|
| 1697 |
+
"grad_norm": 9.559004783630371,
|
| 1698 |
"learning_rate": 3e-06,
|
| 1699 |
+
"loss": 0.6128,
|
| 1700 |
"step": 2210
|
| 1701 |
},
|
| 1702 |
{
|
| 1703 |
"epoch": 17.0,
|
| 1704 |
+
"eval_accuracy": 0.8421052631578947,
|
| 1705 |
+
"eval_loss": 0.5081294178962708,
|
| 1706 |
+
"eval_runtime": 0.9208,
|
| 1707 |
+
"eval_samples_per_second": 144.443,
|
| 1708 |
+
"eval_steps_per_second": 18.463,
|
| 1709 |
"step": 2210
|
| 1710 |
},
|
| 1711 |
{
|
| 1712 |
"epoch": 17.076923076923077,
|
| 1713 |
+
"grad_norm": 5.92500114440918,
|
| 1714 |
"learning_rate": 2.9230769230769236e-06,
|
| 1715 |
+
"loss": 0.6286,
|
| 1716 |
"step": 2220
|
| 1717 |
},
|
| 1718 |
{
|
| 1719 |
"epoch": 17.153846153846153,
|
| 1720 |
+
"grad_norm": 5.656437397003174,
|
| 1721 |
"learning_rate": 2.846153846153846e-06,
|
| 1722 |
+
"loss": 0.6561,
|
| 1723 |
"step": 2230
|
| 1724 |
},
|
| 1725 |
{
|
| 1726 |
"epoch": 17.23076923076923,
|
| 1727 |
+
"grad_norm": 5.492016792297363,
|
| 1728 |
"learning_rate": 2.7692307692307697e-06,
|
| 1729 |
+
"loss": 0.6089,
|
| 1730 |
"step": 2240
|
| 1731 |
},
|
| 1732 |
{
|
| 1733 |
"epoch": 17.307692307692307,
|
| 1734 |
+
"grad_norm": 5.833240985870361,
|
| 1735 |
"learning_rate": 2.6923076923076923e-06,
|
| 1736 |
+
"loss": 0.6818,
|
| 1737 |
"step": 2250
|
| 1738 |
},
|
| 1739 |
{
|
| 1740 |
"epoch": 17.384615384615383,
|
| 1741 |
+
"grad_norm": 3.1228320598602295,
|
| 1742 |
"learning_rate": 2.615384615384616e-06,
|
| 1743 |
+
"loss": 0.6544,
|
| 1744 |
"step": 2260
|
| 1745 |
},
|
| 1746 |
{
|
| 1747 |
"epoch": 17.46153846153846,
|
| 1748 |
+
"grad_norm": 5.364021301269531,
|
| 1749 |
"learning_rate": 2.5384615384615385e-06,
|
| 1750 |
+
"loss": 0.5678,
|
| 1751 |
"step": 2270
|
| 1752 |
},
|
| 1753 |
{
|
| 1754 |
"epoch": 17.53846153846154,
|
| 1755 |
+
"grad_norm": 5.820409297943115,
|
| 1756 |
"learning_rate": 2.461538461538462e-06,
|
| 1757 |
+
"loss": 0.7722,
|
| 1758 |
"step": 2280
|
| 1759 |
},
|
| 1760 |
{
|
| 1761 |
"epoch": 17.615384615384617,
|
| 1762 |
+
"grad_norm": 4.316230297088623,
|
| 1763 |
"learning_rate": 2.384615384615385e-06,
|
| 1764 |
+
"loss": 0.6169,
|
| 1765 |
"step": 2290
|
| 1766 |
},
|
| 1767 |
{
|
| 1768 |
"epoch": 17.692307692307693,
|
| 1769 |
+
"grad_norm": 4.034291744232178,
|
| 1770 |
"learning_rate": 2.307692307692308e-06,
|
| 1771 |
+
"loss": 0.616,
|
| 1772 |
"step": 2300
|
| 1773 |
},
|
| 1774 |
{
|
| 1775 |
"epoch": 17.76923076923077,
|
| 1776 |
+
"grad_norm": 9.319421768188477,
|
| 1777 |
"learning_rate": 2.230769230769231e-06,
|
| 1778 |
+
"loss": 0.6772,
|
| 1779 |
"step": 2310
|
| 1780 |
},
|
| 1781 |
{
|
| 1782 |
"epoch": 17.846153846153847,
|
| 1783 |
+
"grad_norm": 4.886960029602051,
|
| 1784 |
"learning_rate": 2.153846153846154e-06,
|
| 1785 |
+
"loss": 0.6425,
|
| 1786 |
"step": 2320
|
| 1787 |
},
|
| 1788 |
{
|
| 1789 |
"epoch": 17.923076923076923,
|
| 1790 |
+
"grad_norm": 3.6336989402770996,
|
| 1791 |
"learning_rate": 2.0769230769230773e-06,
|
| 1792 |
+
"loss": 0.6684,
|
| 1793 |
"step": 2330
|
| 1794 |
},
|
| 1795 |
{
|
| 1796 |
"epoch": 18.0,
|
| 1797 |
+
"grad_norm": 6.341444492340088,
|
| 1798 |
"learning_rate": 2.0000000000000003e-06,
|
| 1799 |
+
"loss": 0.685,
|
| 1800 |
"step": 2340
|
| 1801 |
},
|
| 1802 |
{
|
| 1803 |
"epoch": 18.0,
|
| 1804 |
+
"eval_accuracy": 0.8571428571428571,
|
| 1805 |
+
"eval_loss": 0.49125248193740845,
|
| 1806 |
+
"eval_runtime": 0.9184,
|
| 1807 |
+
"eval_samples_per_second": 144.82,
|
| 1808 |
+
"eval_steps_per_second": 18.511,
|
| 1809 |
"step": 2340
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 18.076923076923077,
|
| 1813 |
+
"grad_norm": 4.5043158531188965,
|
| 1814 |
"learning_rate": 1.9230769230769234e-06,
|
| 1815 |
+
"loss": 0.5884,
|
| 1816 |
"step": 2350
|
| 1817 |
},
|
| 1818 |
{
|
| 1819 |
"epoch": 18.153846153846153,
|
| 1820 |
+
"grad_norm": 5.0124592781066895,
|
| 1821 |
"learning_rate": 1.8461538461538465e-06,
|
| 1822 |
+
"loss": 0.7235,
|
| 1823 |
"step": 2360
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
"epoch": 18.23076923076923,
|
| 1827 |
+
"grad_norm": 3.5005030632019043,
|
| 1828 |
"learning_rate": 1.7692307692307695e-06,
|
| 1829 |
+
"loss": 0.6181,
|
| 1830 |
"step": 2370
|
| 1831 |
},
|
| 1832 |
{
|
| 1833 |
"epoch": 18.307692307692307,
|
| 1834 |
+
"grad_norm": 5.270286560058594,
|
| 1835 |
"learning_rate": 1.6923076923076926e-06,
|
| 1836 |
+
"loss": 0.6754,
|
| 1837 |
"step": 2380
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
"epoch": 18.384615384615383,
|
| 1841 |
+
"grad_norm": 3.9892470836639404,
|
| 1842 |
"learning_rate": 1.6153846153846157e-06,
|
| 1843 |
+
"loss": 0.6778,
|
| 1844 |
"step": 2390
|
| 1845 |
},
|
| 1846 |
{
|
| 1847 |
"epoch": 18.46153846153846,
|
| 1848 |
+
"grad_norm": 4.945601940155029,
|
| 1849 |
"learning_rate": 1.5384615384615387e-06,
|
| 1850 |
+
"loss": 0.6981,
|
| 1851 |
"step": 2400
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
"epoch": 18.53846153846154,
|
| 1855 |
+
"grad_norm": 6.190303802490234,
|
| 1856 |
"learning_rate": 1.4615384615384618e-06,
|
| 1857 |
+
"loss": 0.6147,
|
| 1858 |
"step": 2410
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
"epoch": 18.615384615384617,
|
| 1862 |
+
"grad_norm": 11.984601974487305,
|
| 1863 |
"learning_rate": 1.3846153846153848e-06,
|
| 1864 |
+
"loss": 0.695,
|
| 1865 |
"step": 2420
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
"epoch": 18.692307692307693,
|
| 1869 |
+
"grad_norm": 5.748422622680664,
|
| 1870 |
"learning_rate": 1.307692307692308e-06,
|
| 1871 |
+
"loss": 0.6945,
|
| 1872 |
"step": 2430
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 18.76923076923077,
|
| 1876 |
+
"grad_norm": 7.194711685180664,
|
| 1877 |
"learning_rate": 1.230769230769231e-06,
|
| 1878 |
+
"loss": 0.7725,
|
| 1879 |
"step": 2440
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
"epoch": 18.846153846153847,
|
| 1883 |
+
"grad_norm": 4.310044288635254,
|
| 1884 |
"learning_rate": 1.153846153846154e-06,
|
| 1885 |
+
"loss": 0.6847,
|
| 1886 |
"step": 2450
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
"epoch": 18.923076923076923,
|
| 1890 |
+
"grad_norm": 4.310553073883057,
|
| 1891 |
"learning_rate": 1.076923076923077e-06,
|
| 1892 |
+
"loss": 0.7078,
|
| 1893 |
"step": 2460
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
"epoch": 19.0,
|
| 1897 |
+
"grad_norm": 7.30472993850708,
|
| 1898 |
"learning_rate": 1.0000000000000002e-06,
|
| 1899 |
+
"loss": 0.6614,
|
| 1900 |
"step": 2470
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
"epoch": 19.0,
|
| 1904 |
+
"eval_accuracy": 0.849624060150376,
|
| 1905 |
+
"eval_loss": 0.49367815256118774,
|
| 1906 |
+
"eval_runtime": 0.9215,
|
| 1907 |
+
"eval_samples_per_second": 144.328,
|
| 1908 |
+
"eval_steps_per_second": 18.448,
|
| 1909 |
"step": 2470
|
| 1910 |
},
|
| 1911 |
{
|
| 1912 |
"epoch": 19.076923076923077,
|
| 1913 |
+
"grad_norm": 3.58013653755188,
|
| 1914 |
"learning_rate": 9.230769230769232e-07,
|
| 1915 |
+
"loss": 0.621,
|
| 1916 |
"step": 2480
|
| 1917 |
},
|
| 1918 |
{
|
| 1919 |
"epoch": 19.153846153846153,
|
| 1920 |
+
"grad_norm": 5.2532453536987305,
|
| 1921 |
"learning_rate": 8.461538461538463e-07,
|
| 1922 |
+
"loss": 0.6367,
|
| 1923 |
"step": 2490
|
| 1924 |
},
|
| 1925 |
{
|
| 1926 |
"epoch": 19.23076923076923,
|
| 1927 |
+
"grad_norm": 11.078712463378906,
|
| 1928 |
"learning_rate": 7.692307692307694e-07,
|
| 1929 |
+
"loss": 0.6835,
|
| 1930 |
"step": 2500
|
| 1931 |
},
|
| 1932 |
{
|
| 1933 |
"epoch": 19.307692307692307,
|
| 1934 |
+
"grad_norm": 4.118062973022461,
|
| 1935 |
"learning_rate": 6.923076923076924e-07,
|
| 1936 |
+
"loss": 0.6719,
|
| 1937 |
"step": 2510
|
| 1938 |
},
|
| 1939 |
{
|
| 1940 |
"epoch": 19.384615384615383,
|
| 1941 |
+
"grad_norm": 4.630733966827393,
|
| 1942 |
"learning_rate": 6.153846153846155e-07,
|
| 1943 |
+
"loss": 0.6546,
|
| 1944 |
"step": 2520
|
| 1945 |
},
|
| 1946 |
{
|
| 1947 |
"epoch": 19.46153846153846,
|
| 1948 |
+
"grad_norm": 4.452598571777344,
|
| 1949 |
"learning_rate": 5.384615384615386e-07,
|
| 1950 |
+
"loss": 0.6052,
|
| 1951 |
"step": 2530
|
| 1952 |
},
|
| 1953 |
{
|
| 1954 |
"epoch": 19.53846153846154,
|
| 1955 |
+
"grad_norm": 5.847692966461182,
|
| 1956 |
"learning_rate": 4.615384615384616e-07,
|
| 1957 |
+
"loss": 0.6786,
|
| 1958 |
"step": 2540
|
| 1959 |
},
|
| 1960 |
{
|
| 1961 |
"epoch": 19.615384615384617,
|
| 1962 |
+
"grad_norm": 5.957764625549316,
|
| 1963 |
"learning_rate": 3.846153846153847e-07,
|
| 1964 |
+
"loss": 0.677,
|
| 1965 |
"step": 2550
|
| 1966 |
},
|
| 1967 |
{
|
| 1968 |
"epoch": 19.692307692307693,
|
| 1969 |
+
"grad_norm": 6.010245323181152,
|
| 1970 |
"learning_rate": 3.0769230769230774e-07,
|
| 1971 |
+
"loss": 0.6106,
|
| 1972 |
"step": 2560
|
| 1973 |
},
|
| 1974 |
{
|
| 1975 |
"epoch": 19.76923076923077,
|
| 1976 |
+
"grad_norm": 5.3892412185668945,
|
| 1977 |
"learning_rate": 2.307692307692308e-07,
|
| 1978 |
+
"loss": 0.5836,
|
| 1979 |
"step": 2570
|
| 1980 |
},
|
| 1981 |
{
|
| 1982 |
"epoch": 19.846153846153847,
|
| 1983 |
+
"grad_norm": 4.214204788208008,
|
| 1984 |
"learning_rate": 1.5384615384615387e-07,
|
| 1985 |
+
"loss": 0.6448,
|
| 1986 |
"step": 2580
|
| 1987 |
},
|
| 1988 |
{
|
| 1989 |
"epoch": 19.923076923076923,
|
| 1990 |
+
"grad_norm": 4.230797290802002,
|
| 1991 |
"learning_rate": 7.692307692307694e-08,
|
| 1992 |
+
"loss": 0.5617,
|
| 1993 |
"step": 2590
|
| 1994 |
},
|
| 1995 |
{
|
| 1996 |
"epoch": 20.0,
|
| 1997 |
+
"grad_norm": 10.618528366088867,
|
| 1998 |
"learning_rate": 0.0,
|
| 1999 |
+
"loss": 0.6934,
|
| 2000 |
"step": 2600
|
| 2001 |
},
|
| 2002 |
{
|
| 2003 |
"epoch": 20.0,
|
| 2004 |
+
"eval_accuracy": 0.8571428571428571,
|
| 2005 |
+
"eval_loss": 0.5026500821113586,
|
| 2006 |
+
"eval_runtime": 1.0302,
|
| 2007 |
+
"eval_samples_per_second": 129.097,
|
| 2008 |
+
"eval_steps_per_second": 16.501,
|
| 2009 |
"step": 2600
|
| 2010 |
},
|
| 2011 |
{
|
| 2012 |
"epoch": 20.0,
|
| 2013 |
"step": 2600,
|
| 2014 |
+
"total_flos": 7.939121542823117e+17,
|
| 2015 |
+
"train_loss": 0.8216404274793772,
|
| 2016 |
+
"train_runtime": 338.1702,
|
| 2017 |
+
"train_samples_per_second": 61.153,
|
| 2018 |
+
"train_steps_per_second": 7.688
|
| 2019 |
}
|
| 2020 |
],
|
| 2021 |
"logging_steps": 10,
|
|
|
|
| 2035 |
"attributes": {}
|
| 2036 |
}
|
| 2037 |
},
|
| 2038 |
+
"total_flos": 7.939121542823117e+17,
|
| 2039 |
"train_batch_size": 8,
|
| 2040 |
"trial_name": null,
|
| 2041 |
"trial_params": null
|