Upload folder using huggingface_hub
Browse files- README.md +6 -6
- all_results.json +6 -6
- checkpoint-1000/config.json +2 -2
- checkpoint-1000/optimizer.pt +1 -1
- checkpoint-1000/pytorch_model.bin +1 -1
- checkpoint-1000/rng_state.pth +1 -1
- checkpoint-1000/trainer_state.json +404 -404
- checkpoint-1000/training_args.bin +1 -1
- config.json +1 -1
- pytorch_model.bin +1 -1
- train_results.json +6 -6
- trainer_state.json +1401 -33
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
---
|
| 2 |
-
base_model: ./
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
| 5 |
model-index:
|
|
@@ -12,7 +12,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 12 |
|
| 13 |
# output5
|
| 14 |
|
| 15 |
-
This model is a fine-tuned version of [./
|
| 16 |
|
| 17 |
## Model description
|
| 18 |
|
|
@@ -31,15 +31,15 @@ More information needed
|
|
| 31 |
### Training hyperparameters
|
| 32 |
|
| 33 |
The following hyperparameters were used during training:
|
| 34 |
-
- learning_rate: 0.
|
| 35 |
- train_batch_size: 1
|
| 36 |
- eval_batch_size: 1
|
| 37 |
- seed: 42
|
| 38 |
-
- gradient_accumulation_steps:
|
| 39 |
-
- total_train_batch_size:
|
| 40 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 41 |
- lr_scheduler_type: constant
|
| 42 |
-
- num_epochs:
|
| 43 |
|
| 44 |
### Training results
|
| 45 |
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model: ./output5
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
| 5 |
model-index:
|
|
|
|
| 12 |
|
| 13 |
# output5
|
| 14 |
|
| 15 |
+
This model is a fine-tuned version of [./output5](https://huggingface.co/./output5) on an unknown dataset.
|
| 16 |
|
| 17 |
## Model description
|
| 18 |
|
|
|
|
| 31 |
### Training hyperparameters
|
| 32 |
|
| 33 |
The following hyperparameters were used during training:
|
| 34 |
+
- learning_rate: 0.0001
|
| 35 |
- train_batch_size: 1
|
| 36 |
- eval_batch_size: 1
|
| 37 |
- seed: 42
|
| 38 |
+
- gradient_accumulation_steps: 32
|
| 39 |
+
- total_train_batch_size: 32
|
| 40 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 41 |
- lr_scheduler_type: constant
|
| 42 |
+
- num_epochs: 25.0
|
| 43 |
|
| 44 |
### Training results
|
| 45 |
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 1.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 24.61,
|
| 3 |
+
"train_loss": 1.3988467488390335,
|
| 4 |
+
"train_runtime": 1205.3351,
|
| 5 |
+
"train_samples": 1528,
|
| 6 |
+
"train_samples_per_second": 31.692,
|
| 7 |
+
"train_steps_per_second": 0.975
|
| 8 |
}
|
checkpoint-1000/config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "./
|
| 3 |
"architectures": [
|
| 4 |
"LlamaForCausalLM"
|
| 5 |
],
|
|
@@ -14,7 +14,7 @@
|
|
| 14 |
"num_attention_heads": 4,
|
| 15 |
"num_hidden_layers": 2,
|
| 16 |
"num_key_value_heads": 4,
|
| 17 |
-
"pretraining_tp":
|
| 18 |
"rms_norm_eps": 1e-05,
|
| 19 |
"rope_scaling": null,
|
| 20 |
"rope_theta": 10000.0,
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "./output5",
|
| 3 |
"architectures": [
|
| 4 |
"LlamaForCausalLM"
|
| 5 |
],
|
|
|
|
| 14 |
"num_attention_heads": 4,
|
| 15 |
"num_hidden_layers": 2,
|
| 16 |
"num_key_value_heads": 4,
|
| 17 |
+
"pretraining_tp": 2,
|
| 18 |
"rms_norm_eps": 1e-05,
|
| 19 |
"rope_scaling": null,
|
| 20 |
"rope_theta": 10000.0,
|
checkpoint-1000/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 141585770
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff915a4035235786ff34aca643fbdb2a4ef583f6fa0790194e53ec78ac7eb34e
|
| 3 |
size 141585770
|
checkpoint-1000/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 70790896
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f00c657f63a7216e25e186f2d29946dac0ba91a74376974df8944e985170d76b
|
| 3 |
size 70790896
|
checkpoint-1000/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5432f961e7c82cbe7b6ec30d027cf3db4d17372f65774be8cf037a208cbc1ac
|
| 3 |
size 14575
|
checkpoint-1000/trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -9,1211 +9,1211 @@
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
"learning_rate": 0.0001,
|
| 14 |
-
"loss":
|
| 15 |
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
"learning_rate": 0.0001,
|
| 20 |
-
"loss":
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
"learning_rate": 0.0001,
|
| 26 |
-
"loss":
|
| 27 |
"step": 15
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
"learning_rate": 0.0001,
|
| 32 |
-
"loss":
|
| 33 |
"step": 20
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
"learning_rate": 0.0001,
|
| 38 |
-
"loss":
|
| 39 |
"step": 25
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
"learning_rate": 0.0001,
|
| 44 |
-
"loss":
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
"learning_rate": 0.0001,
|
| 50 |
-
"loss":
|
| 51 |
"step": 35
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
"learning_rate": 0.0001,
|
| 56 |
-
"loss":
|
| 57 |
"step": 40
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
"learning_rate": 0.0001,
|
| 62 |
-
"loss":
|
| 63 |
"step": 45
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch":
|
| 67 |
"learning_rate": 0.0001,
|
| 68 |
-
"loss":
|
| 69 |
"step": 50
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch":
|
| 73 |
"learning_rate": 0.0001,
|
| 74 |
-
"loss":
|
| 75 |
"step": 55
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
"learning_rate": 0.0001,
|
| 80 |
-
"loss":
|
| 81 |
"step": 60
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch":
|
| 85 |
"learning_rate": 0.0001,
|
| 86 |
-
"loss":
|
| 87 |
"step": 65
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch":
|
| 91 |
"learning_rate": 0.0001,
|
| 92 |
-
"loss":
|
| 93 |
"step": 70
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
"learning_rate": 0.0001,
|
| 98 |
-
"loss":
|
| 99 |
"step": 75
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
| 103 |
"learning_rate": 0.0001,
|
| 104 |
-
"loss":
|
| 105 |
"step": 80
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
"learning_rate": 0.0001,
|
| 110 |
-
"loss":
|
| 111 |
"step": 85
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
"learning_rate": 0.0001,
|
| 116 |
-
"loss":
|
| 117 |
"step": 90
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
"learning_rate": 0.0001,
|
| 122 |
-
"loss":
|
| 123 |
"step": 95
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
"learning_rate": 0.0001,
|
| 128 |
-
"loss":
|
| 129 |
"step": 100
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
"learning_rate": 0.0001,
|
| 134 |
-
"loss":
|
| 135 |
"step": 105
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
"learning_rate": 0.0001,
|
| 140 |
-
"loss":
|
| 141 |
"step": 110
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
"learning_rate": 0.0001,
|
| 146 |
-
"loss":
|
| 147 |
"step": 115
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch":
|
| 151 |
"learning_rate": 0.0001,
|
| 152 |
-
"loss":
|
| 153 |
"step": 120
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch":
|
| 157 |
"learning_rate": 0.0001,
|
| 158 |
-
"loss":
|
| 159 |
"step": 125
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
"learning_rate": 0.0001,
|
| 164 |
-
"loss":
|
| 165 |
"step": 130
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
"learning_rate": 0.0001,
|
| 170 |
-
"loss":
|
| 171 |
"step": 135
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
"learning_rate": 0.0001,
|
| 176 |
-
"loss":
|
| 177 |
"step": 140
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch":
|
| 181 |
"learning_rate": 0.0001,
|
| 182 |
-
"loss":
|
| 183 |
"step": 145
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch":
|
| 187 |
"learning_rate": 0.0001,
|
| 188 |
-
"loss":
|
| 189 |
"step": 150
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch":
|
| 193 |
"learning_rate": 0.0001,
|
| 194 |
-
"loss":
|
| 195 |
"step": 155
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch":
|
| 199 |
"learning_rate": 0.0001,
|
| 200 |
-
"loss":
|
| 201 |
"step": 160
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
"learning_rate": 0.0001,
|
| 206 |
-
"loss":
|
| 207 |
"step": 165
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
"learning_rate": 0.0001,
|
| 212 |
-
"loss":
|
| 213 |
"step": 170
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
"learning_rate": 0.0001,
|
| 218 |
-
"loss":
|
| 219 |
"step": 175
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch":
|
| 223 |
"learning_rate": 0.0001,
|
| 224 |
-
"loss":
|
| 225 |
"step": 180
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch":
|
| 229 |
"learning_rate": 0.0001,
|
| 230 |
-
"loss":
|
| 231 |
"step": 185
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch":
|
| 235 |
"learning_rate": 0.0001,
|
| 236 |
-
"loss":
|
| 237 |
"step": 190
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch":
|
| 241 |
"learning_rate": 0.0001,
|
| 242 |
-
"loss":
|
| 243 |
"step": 195
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch":
|
| 247 |
"learning_rate": 0.0001,
|
| 248 |
-
"loss":
|
| 249 |
"step": 200
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch":
|
| 253 |
"learning_rate": 0.0001,
|
| 254 |
-
"loss":
|
| 255 |
"step": 205
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
"learning_rate": 0.0001,
|
| 260 |
-
"loss":
|
| 261 |
"step": 210
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
"learning_rate": 0.0001,
|
| 266 |
-
"loss":
|
| 267 |
"step": 215
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch":
|
| 271 |
"learning_rate": 0.0001,
|
| 272 |
-
"loss":
|
| 273 |
"step": 220
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch":
|
| 277 |
"learning_rate": 0.0001,
|
| 278 |
-
"loss":
|
| 279 |
"step": 225
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch":
|
| 283 |
"learning_rate": 0.0001,
|
| 284 |
-
"loss":
|
| 285 |
"step": 230
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch":
|
| 289 |
"learning_rate": 0.0001,
|
| 290 |
-
"loss":
|
| 291 |
"step": 235
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
"learning_rate": 0.0001,
|
| 296 |
-
"loss":
|
| 297 |
"step": 240
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch":
|
| 301 |
"learning_rate": 0.0001,
|
| 302 |
-
"loss":
|
| 303 |
"step": 245
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
"learning_rate": 0.0001,
|
| 308 |
-
"loss":
|
| 309 |
"step": 250
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
"learning_rate": 0.0001,
|
| 314 |
-
"loss":
|
| 315 |
"step": 255
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch":
|
| 319 |
"learning_rate": 0.0001,
|
| 320 |
-
"loss":
|
| 321 |
"step": 260
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch":
|
| 325 |
"learning_rate": 0.0001,
|
| 326 |
-
"loss":
|
| 327 |
"step": 265
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch":
|
| 331 |
"learning_rate": 0.0001,
|
| 332 |
-
"loss":
|
| 333 |
"step": 270
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch":
|
| 337 |
"learning_rate": 0.0001,
|
| 338 |
-
"loss":
|
| 339 |
"step": 275
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch":
|
| 343 |
"learning_rate": 0.0001,
|
| 344 |
-
"loss":
|
| 345 |
"step": 280
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch":
|
| 349 |
"learning_rate": 0.0001,
|
| 350 |
-
"loss":
|
| 351 |
"step": 285
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch":
|
| 355 |
"learning_rate": 0.0001,
|
| 356 |
-
"loss":
|
| 357 |
"step": 290
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch":
|
| 361 |
"learning_rate": 0.0001,
|
| 362 |
-
"loss":
|
| 363 |
"step": 295
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch":
|
| 367 |
"learning_rate": 0.0001,
|
| 368 |
-
"loss":
|
| 369 |
"step": 300
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch":
|
| 373 |
"learning_rate": 0.0001,
|
| 374 |
-
"loss":
|
| 375 |
"step": 305
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch":
|
| 379 |
"learning_rate": 0.0001,
|
| 380 |
-
"loss":
|
| 381 |
"step": 310
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch":
|
| 385 |
"learning_rate": 0.0001,
|
| 386 |
-
"loss":
|
| 387 |
"step": 315
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch":
|
| 391 |
"learning_rate": 0.0001,
|
| 392 |
-
"loss":
|
| 393 |
"step": 320
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch":
|
| 397 |
"learning_rate": 0.0001,
|
| 398 |
-
"loss":
|
| 399 |
"step": 325
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch":
|
| 403 |
"learning_rate": 0.0001,
|
| 404 |
-
"loss":
|
| 405 |
"step": 330
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
"learning_rate": 0.0001,
|
| 410 |
-
"loss":
|
| 411 |
"step": 335
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch":
|
| 415 |
"learning_rate": 0.0001,
|
| 416 |
-
"loss":
|
| 417 |
"step": 340
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch":
|
| 421 |
"learning_rate": 0.0001,
|
| 422 |
-
"loss":
|
| 423 |
"step": 345
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
"learning_rate": 0.0001,
|
| 428 |
-
"loss":
|
| 429 |
"step": 350
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
"learning_rate": 0.0001,
|
| 434 |
-
"loss":
|
| 435 |
"step": 355
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch":
|
| 439 |
"learning_rate": 0.0001,
|
| 440 |
-
"loss":
|
| 441 |
"step": 360
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch":
|
| 445 |
"learning_rate": 0.0001,
|
| 446 |
-
"loss":
|
| 447 |
"step": 365
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch":
|
| 451 |
"learning_rate": 0.0001,
|
| 452 |
-
"loss":
|
| 453 |
"step": 370
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
"learning_rate": 0.0001,
|
| 458 |
-
"loss":
|
| 459 |
"step": 375
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
"learning_rate": 0.0001,
|
| 464 |
-
"loss":
|
| 465 |
"step": 380
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
"learning_rate": 0.0001,
|
| 470 |
-
"loss":
|
| 471 |
"step": 385
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
"learning_rate": 0.0001,
|
| 476 |
-
"loss":
|
| 477 |
"step": 390
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
"learning_rate": 0.0001,
|
| 482 |
-
"loss":
|
| 483 |
"step": 395
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch":
|
| 487 |
"learning_rate": 0.0001,
|
| 488 |
-
"loss":
|
| 489 |
"step": 400
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch":
|
| 493 |
"learning_rate": 0.0001,
|
| 494 |
-
"loss":
|
| 495 |
"step": 405
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
"learning_rate": 0.0001,
|
| 500 |
-
"loss":
|
| 501 |
"step": 410
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch":
|
| 505 |
"learning_rate": 0.0001,
|
| 506 |
-
"loss":
|
| 507 |
"step": 415
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
"learning_rate": 0.0001,
|
| 512 |
-
"loss":
|
| 513 |
"step": 420
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch":
|
| 517 |
"learning_rate": 0.0001,
|
| 518 |
-
"loss":
|
| 519 |
"step": 425
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
"learning_rate": 0.0001,
|
| 524 |
-
"loss":
|
| 525 |
"step": 430
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch":
|
| 529 |
"learning_rate": 0.0001,
|
| 530 |
-
"loss":
|
| 531 |
"step": 435
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch":
|
| 535 |
"learning_rate": 0.0001,
|
| 536 |
-
"loss":
|
| 537 |
"step": 440
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch":
|
| 541 |
"learning_rate": 0.0001,
|
| 542 |
-
"loss":
|
| 543 |
"step": 445
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch":
|
| 547 |
"learning_rate": 0.0001,
|
| 548 |
-
"loss":
|
| 549 |
"step": 450
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch":
|
| 553 |
"learning_rate": 0.0001,
|
| 554 |
-
"loss":
|
| 555 |
"step": 455
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch":
|
| 559 |
"learning_rate": 0.0001,
|
| 560 |
-
"loss":
|
| 561 |
"step": 460
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch":
|
| 565 |
"learning_rate": 0.0001,
|
| 566 |
-
"loss":
|
| 567 |
"step": 465
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch":
|
| 571 |
"learning_rate": 0.0001,
|
| 572 |
-
"loss":
|
| 573 |
"step": 470
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch":
|
| 577 |
"learning_rate": 0.0001,
|
| 578 |
-
"loss":
|
| 579 |
"step": 475
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch":
|
| 583 |
"learning_rate": 0.0001,
|
| 584 |
-
"loss":
|
| 585 |
"step": 480
|
| 586 |
},
|
| 587 |
{
|
| 588 |
-
"epoch":
|
| 589 |
"learning_rate": 0.0001,
|
| 590 |
-
"loss":
|
| 591 |
"step": 485
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
"learning_rate": 0.0001,
|
| 596 |
-
"loss":
|
| 597 |
"step": 490
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
"learning_rate": 0.0001,
|
| 602 |
-
"loss":
|
| 603 |
"step": 495
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch":
|
| 607 |
"learning_rate": 0.0001,
|
| 608 |
-
"loss":
|
| 609 |
"step": 500
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch":
|
| 613 |
"learning_rate": 0.0001,
|
| 614 |
-
"loss":
|
| 615 |
"step": 505
|
| 616 |
},
|
| 617 |
{
|
| 618 |
-
"epoch":
|
| 619 |
"learning_rate": 0.0001,
|
| 620 |
-
"loss":
|
| 621 |
"step": 510
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch":
|
| 625 |
"learning_rate": 0.0001,
|
| 626 |
-
"loss":
|
| 627 |
"step": 515
|
| 628 |
},
|
| 629 |
{
|
| 630 |
-
"epoch":
|
| 631 |
"learning_rate": 0.0001,
|
| 632 |
-
"loss":
|
| 633 |
"step": 520
|
| 634 |
},
|
| 635 |
{
|
| 636 |
-
"epoch":
|
| 637 |
"learning_rate": 0.0001,
|
| 638 |
-
"loss":
|
| 639 |
"step": 525
|
| 640 |
},
|
| 641 |
{
|
| 642 |
-
"epoch": 1
|
| 643 |
"learning_rate": 0.0001,
|
| 644 |
-
"loss":
|
| 645 |
"step": 530
|
| 646 |
},
|
| 647 |
{
|
| 648 |
-
"epoch":
|
| 649 |
"learning_rate": 0.0001,
|
| 650 |
-
"loss":
|
| 651 |
"step": 535
|
| 652 |
},
|
| 653 |
{
|
| 654 |
-
"epoch":
|
| 655 |
"learning_rate": 0.0001,
|
| 656 |
-
"loss":
|
| 657 |
"step": 540
|
| 658 |
},
|
| 659 |
{
|
| 660 |
-
"epoch":
|
| 661 |
"learning_rate": 0.0001,
|
| 662 |
-
"loss":
|
| 663 |
"step": 545
|
| 664 |
},
|
| 665 |
{
|
| 666 |
-
"epoch":
|
| 667 |
"learning_rate": 0.0001,
|
| 668 |
-
"loss":
|
| 669 |
"step": 550
|
| 670 |
},
|
| 671 |
{
|
| 672 |
-
"epoch":
|
| 673 |
"learning_rate": 0.0001,
|
| 674 |
-
"loss":
|
| 675 |
"step": 555
|
| 676 |
},
|
| 677 |
{
|
| 678 |
-
"epoch":
|
| 679 |
"learning_rate": 0.0001,
|
| 680 |
-
"loss":
|
| 681 |
"step": 560
|
| 682 |
},
|
| 683 |
{
|
| 684 |
-
"epoch":
|
| 685 |
"learning_rate": 0.0001,
|
| 686 |
-
"loss":
|
| 687 |
"step": 565
|
| 688 |
},
|
| 689 |
{
|
| 690 |
-
"epoch":
|
| 691 |
"learning_rate": 0.0001,
|
| 692 |
-
"loss":
|
| 693 |
"step": 570
|
| 694 |
},
|
| 695 |
{
|
| 696 |
-
"epoch":
|
| 697 |
"learning_rate": 0.0001,
|
| 698 |
-
"loss":
|
| 699 |
"step": 575
|
| 700 |
},
|
| 701 |
{
|
| 702 |
-
"epoch":
|
| 703 |
"learning_rate": 0.0001,
|
| 704 |
-
"loss":
|
| 705 |
"step": 580
|
| 706 |
},
|
| 707 |
{
|
| 708 |
-
"epoch":
|
| 709 |
"learning_rate": 0.0001,
|
| 710 |
-
"loss":
|
| 711 |
"step": 585
|
| 712 |
},
|
| 713 |
{
|
| 714 |
-
"epoch":
|
| 715 |
"learning_rate": 0.0001,
|
| 716 |
-
"loss":
|
| 717 |
"step": 590
|
| 718 |
},
|
| 719 |
{
|
| 720 |
-
"epoch":
|
| 721 |
"learning_rate": 0.0001,
|
| 722 |
-
"loss":
|
| 723 |
"step": 595
|
| 724 |
},
|
| 725 |
{
|
| 726 |
-
"epoch":
|
| 727 |
"learning_rate": 0.0001,
|
| 728 |
-
"loss":
|
| 729 |
"step": 600
|
| 730 |
},
|
| 731 |
{
|
| 732 |
-
"epoch":
|
| 733 |
"learning_rate": 0.0001,
|
| 734 |
-
"loss":
|
| 735 |
"step": 605
|
| 736 |
},
|
| 737 |
{
|
| 738 |
-
"epoch":
|
| 739 |
"learning_rate": 0.0001,
|
| 740 |
-
"loss":
|
| 741 |
"step": 610
|
| 742 |
},
|
| 743 |
{
|
| 744 |
-
"epoch":
|
| 745 |
"learning_rate": 0.0001,
|
| 746 |
-
"loss":
|
| 747 |
"step": 615
|
| 748 |
},
|
| 749 |
{
|
| 750 |
-
"epoch":
|
| 751 |
"learning_rate": 0.0001,
|
| 752 |
-
"loss":
|
| 753 |
"step": 620
|
| 754 |
},
|
| 755 |
{
|
| 756 |
-
"epoch":
|
| 757 |
"learning_rate": 0.0001,
|
| 758 |
-
"loss":
|
| 759 |
"step": 625
|
| 760 |
},
|
| 761 |
{
|
| 762 |
-
"epoch":
|
| 763 |
"learning_rate": 0.0001,
|
| 764 |
-
"loss":
|
| 765 |
"step": 630
|
| 766 |
},
|
| 767 |
{
|
| 768 |
-
"epoch":
|
| 769 |
"learning_rate": 0.0001,
|
| 770 |
-
"loss":
|
| 771 |
"step": 635
|
| 772 |
},
|
| 773 |
{
|
| 774 |
-
"epoch":
|
| 775 |
"learning_rate": 0.0001,
|
| 776 |
-
"loss":
|
| 777 |
"step": 640
|
| 778 |
},
|
| 779 |
{
|
| 780 |
-
"epoch":
|
| 781 |
"learning_rate": 0.0001,
|
| 782 |
-
"loss":
|
| 783 |
"step": 645
|
| 784 |
},
|
| 785 |
{
|
| 786 |
-
"epoch":
|
| 787 |
"learning_rate": 0.0001,
|
| 788 |
-
"loss":
|
| 789 |
"step": 650
|
| 790 |
},
|
| 791 |
{
|
| 792 |
-
"epoch":
|
| 793 |
"learning_rate": 0.0001,
|
| 794 |
-
"loss":
|
| 795 |
"step": 655
|
| 796 |
},
|
| 797 |
{
|
| 798 |
-
"epoch":
|
| 799 |
"learning_rate": 0.0001,
|
| 800 |
-
"loss":
|
| 801 |
"step": 660
|
| 802 |
},
|
| 803 |
{
|
| 804 |
-
"epoch":
|
| 805 |
"learning_rate": 0.0001,
|
| 806 |
-
"loss":
|
| 807 |
"step": 665
|
| 808 |
},
|
| 809 |
{
|
| 810 |
-
"epoch":
|
| 811 |
"learning_rate": 0.0001,
|
| 812 |
-
"loss":
|
| 813 |
"step": 670
|
| 814 |
},
|
| 815 |
{
|
| 816 |
-
"epoch":
|
| 817 |
"learning_rate": 0.0001,
|
| 818 |
-
"loss":
|
| 819 |
"step": 675
|
| 820 |
},
|
| 821 |
{
|
| 822 |
-
"epoch":
|
| 823 |
"learning_rate": 0.0001,
|
| 824 |
-
"loss":
|
| 825 |
"step": 680
|
| 826 |
},
|
| 827 |
{
|
| 828 |
-
"epoch":
|
| 829 |
"learning_rate": 0.0001,
|
| 830 |
-
"loss":
|
| 831 |
"step": 685
|
| 832 |
},
|
| 833 |
{
|
| 834 |
-
"epoch":
|
| 835 |
"learning_rate": 0.0001,
|
| 836 |
-
"loss":
|
| 837 |
"step": 690
|
| 838 |
},
|
| 839 |
{
|
| 840 |
-
"epoch":
|
| 841 |
"learning_rate": 0.0001,
|
| 842 |
-
"loss":
|
| 843 |
"step": 695
|
| 844 |
},
|
| 845 |
{
|
| 846 |
-
"epoch":
|
| 847 |
"learning_rate": 0.0001,
|
| 848 |
-
"loss":
|
| 849 |
"step": 700
|
| 850 |
},
|
| 851 |
{
|
| 852 |
-
"epoch":
|
| 853 |
"learning_rate": 0.0001,
|
| 854 |
-
"loss":
|
| 855 |
"step": 705
|
| 856 |
},
|
| 857 |
{
|
| 858 |
-
"epoch":
|
| 859 |
"learning_rate": 0.0001,
|
| 860 |
-
"loss":
|
| 861 |
"step": 710
|
| 862 |
},
|
| 863 |
{
|
| 864 |
-
"epoch":
|
| 865 |
"learning_rate": 0.0001,
|
| 866 |
-
"loss":
|
| 867 |
"step": 715
|
| 868 |
},
|
| 869 |
{
|
| 870 |
-
"epoch":
|
| 871 |
"learning_rate": 0.0001,
|
| 872 |
-
"loss":
|
| 873 |
"step": 720
|
| 874 |
},
|
| 875 |
{
|
| 876 |
-
"epoch":
|
| 877 |
"learning_rate": 0.0001,
|
| 878 |
-
"loss":
|
| 879 |
"step": 725
|
| 880 |
},
|
| 881 |
{
|
| 882 |
-
"epoch":
|
| 883 |
"learning_rate": 0.0001,
|
| 884 |
-
"loss":
|
| 885 |
"step": 730
|
| 886 |
},
|
| 887 |
{
|
| 888 |
-
"epoch":
|
| 889 |
"learning_rate": 0.0001,
|
| 890 |
-
"loss":
|
| 891 |
"step": 735
|
| 892 |
},
|
| 893 |
{
|
| 894 |
-
"epoch":
|
| 895 |
"learning_rate": 0.0001,
|
| 896 |
-
"loss":
|
| 897 |
"step": 740
|
| 898 |
},
|
| 899 |
{
|
| 900 |
-
"epoch":
|
| 901 |
"learning_rate": 0.0001,
|
| 902 |
-
"loss":
|
| 903 |
"step": 745
|
| 904 |
},
|
| 905 |
{
|
| 906 |
-
"epoch":
|
| 907 |
"learning_rate": 0.0001,
|
| 908 |
-
"loss":
|
| 909 |
"step": 750
|
| 910 |
},
|
| 911 |
{
|
| 912 |
-
"epoch":
|
| 913 |
"learning_rate": 0.0001,
|
| 914 |
-
"loss":
|
| 915 |
"step": 755
|
| 916 |
},
|
| 917 |
{
|
| 918 |
-
"epoch":
|
| 919 |
"learning_rate": 0.0001,
|
| 920 |
-
"loss":
|
| 921 |
"step": 760
|
| 922 |
},
|
| 923 |
{
|
| 924 |
-
"epoch":
|
| 925 |
"learning_rate": 0.0001,
|
| 926 |
-
"loss":
|
| 927 |
"step": 765
|
| 928 |
},
|
| 929 |
{
|
| 930 |
-
"epoch":
|
| 931 |
"learning_rate": 0.0001,
|
| 932 |
-
"loss":
|
| 933 |
"step": 770
|
| 934 |
},
|
| 935 |
{
|
| 936 |
-
"epoch":
|
| 937 |
"learning_rate": 0.0001,
|
| 938 |
-
"loss":
|
| 939 |
"step": 775
|
| 940 |
},
|
| 941 |
{
|
| 942 |
-
"epoch":
|
| 943 |
"learning_rate": 0.0001,
|
| 944 |
-
"loss":
|
| 945 |
"step": 780
|
| 946 |
},
|
| 947 |
{
|
| 948 |
-
"epoch":
|
| 949 |
"learning_rate": 0.0001,
|
| 950 |
-
"loss":
|
| 951 |
"step": 785
|
| 952 |
},
|
| 953 |
{
|
| 954 |
-
"epoch":
|
| 955 |
"learning_rate": 0.0001,
|
| 956 |
-
"loss":
|
| 957 |
"step": 790
|
| 958 |
},
|
| 959 |
{
|
| 960 |
-
"epoch":
|
| 961 |
"learning_rate": 0.0001,
|
| 962 |
-
"loss":
|
| 963 |
"step": 795
|
| 964 |
},
|
| 965 |
{
|
| 966 |
-
"epoch":
|
| 967 |
"learning_rate": 0.0001,
|
| 968 |
-
"loss":
|
| 969 |
"step": 800
|
| 970 |
},
|
| 971 |
{
|
| 972 |
-
"epoch":
|
| 973 |
"learning_rate": 0.0001,
|
| 974 |
-
"loss":
|
| 975 |
"step": 805
|
| 976 |
},
|
| 977 |
{
|
| 978 |
-
"epoch":
|
| 979 |
"learning_rate": 0.0001,
|
| 980 |
-
"loss":
|
| 981 |
"step": 810
|
| 982 |
},
|
| 983 |
{
|
| 984 |
-
"epoch":
|
| 985 |
"learning_rate": 0.0001,
|
| 986 |
-
"loss":
|
| 987 |
"step": 815
|
| 988 |
},
|
| 989 |
{
|
| 990 |
-
"epoch":
|
| 991 |
"learning_rate": 0.0001,
|
| 992 |
-
"loss":
|
| 993 |
"step": 820
|
| 994 |
},
|
| 995 |
{
|
| 996 |
-
"epoch":
|
| 997 |
"learning_rate": 0.0001,
|
| 998 |
-
"loss":
|
| 999 |
"step": 825
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
-
"epoch":
|
| 1003 |
"learning_rate": 0.0001,
|
| 1004 |
-
"loss":
|
| 1005 |
"step": 830
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
-
"epoch":
|
| 1009 |
"learning_rate": 0.0001,
|
| 1010 |
-
"loss":
|
| 1011 |
"step": 835
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
-
"epoch":
|
| 1015 |
"learning_rate": 0.0001,
|
| 1016 |
-
"loss":
|
| 1017 |
"step": 840
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
-
"epoch":
|
| 1021 |
"learning_rate": 0.0001,
|
| 1022 |
-
"loss":
|
| 1023 |
"step": 845
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
-
"epoch":
|
| 1027 |
"learning_rate": 0.0001,
|
| 1028 |
-
"loss":
|
| 1029 |
"step": 850
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
-
"epoch":
|
| 1033 |
"learning_rate": 0.0001,
|
| 1034 |
-
"loss":
|
| 1035 |
"step": 855
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
-
"epoch":
|
| 1039 |
"learning_rate": 0.0001,
|
| 1040 |
-
"loss":
|
| 1041 |
"step": 860
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
-
"epoch":
|
| 1045 |
"learning_rate": 0.0001,
|
| 1046 |
-
"loss":
|
| 1047 |
"step": 865
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
-
"epoch":
|
| 1051 |
"learning_rate": 0.0001,
|
| 1052 |
-
"loss":
|
| 1053 |
"step": 870
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
-
"epoch":
|
| 1057 |
"learning_rate": 0.0001,
|
| 1058 |
-
"loss":
|
| 1059 |
"step": 875
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch":
|
| 1063 |
"learning_rate": 0.0001,
|
| 1064 |
-
"loss":
|
| 1065 |
"step": 880
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
-
"epoch":
|
| 1069 |
"learning_rate": 0.0001,
|
| 1070 |
-
"loss":
|
| 1071 |
"step": 885
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
-
"epoch":
|
| 1075 |
"learning_rate": 0.0001,
|
| 1076 |
-
"loss":
|
| 1077 |
"step": 890
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
-
"epoch":
|
| 1081 |
"learning_rate": 0.0001,
|
| 1082 |
-
"loss":
|
| 1083 |
"step": 895
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
-
"epoch":
|
| 1087 |
"learning_rate": 0.0001,
|
| 1088 |
-
"loss":
|
| 1089 |
"step": 900
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
-
"epoch":
|
| 1093 |
"learning_rate": 0.0001,
|
| 1094 |
-
"loss":
|
| 1095 |
"step": 905
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
-
"epoch":
|
| 1099 |
"learning_rate": 0.0001,
|
| 1100 |
-
"loss":
|
| 1101 |
"step": 910
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
-
"epoch":
|
| 1105 |
"learning_rate": 0.0001,
|
| 1106 |
-
"loss":
|
| 1107 |
"step": 915
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
-
"epoch":
|
| 1111 |
"learning_rate": 0.0001,
|
| 1112 |
-
"loss":
|
| 1113 |
"step": 920
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
-
"epoch":
|
| 1117 |
"learning_rate": 0.0001,
|
| 1118 |
-
"loss":
|
| 1119 |
"step": 925
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
-
"epoch":
|
| 1123 |
"learning_rate": 0.0001,
|
| 1124 |
-
"loss":
|
| 1125 |
"step": 930
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
-
"epoch":
|
| 1129 |
"learning_rate": 0.0001,
|
| 1130 |
-
"loss":
|
| 1131 |
"step": 935
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
-
"epoch":
|
| 1135 |
"learning_rate": 0.0001,
|
| 1136 |
-
"loss":
|
| 1137 |
"step": 940
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
-
"epoch":
|
| 1141 |
"learning_rate": 0.0001,
|
| 1142 |
-
"loss":
|
| 1143 |
"step": 945
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
-
"epoch":
|
| 1147 |
"learning_rate": 0.0001,
|
| 1148 |
-
"loss":
|
| 1149 |
"step": 950
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
-
"epoch":
|
| 1153 |
"learning_rate": 0.0001,
|
| 1154 |
-
"loss":
|
| 1155 |
"step": 955
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
-
"epoch":
|
| 1159 |
"learning_rate": 0.0001,
|
| 1160 |
-
"loss":
|
| 1161 |
"step": 960
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
-
"epoch":
|
| 1165 |
"learning_rate": 0.0001,
|
| 1166 |
-
"loss":
|
| 1167 |
"step": 965
|
| 1168 |
},
|
| 1169 |
{
|
| 1170 |
-
"epoch":
|
| 1171 |
"learning_rate": 0.0001,
|
| 1172 |
-
"loss":
|
| 1173 |
"step": 970
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
-
"epoch":
|
| 1177 |
"learning_rate": 0.0001,
|
| 1178 |
-
"loss":
|
| 1179 |
"step": 975
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
-
"epoch":
|
| 1183 |
"learning_rate": 0.0001,
|
| 1184 |
-
"loss":
|
| 1185 |
"step": 980
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
-
"epoch":
|
| 1189 |
"learning_rate": 0.0001,
|
| 1190 |
-
"loss":
|
| 1191 |
"step": 985
|
| 1192 |
},
|
| 1193 |
{
|
| 1194 |
-
"epoch":
|
| 1195 |
"learning_rate": 0.0001,
|
| 1196 |
-
"loss":
|
| 1197 |
"step": 990
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
-
"epoch":
|
| 1201 |
"learning_rate": 0.0001,
|
| 1202 |
-
"loss":
|
| 1203 |
"step": 995
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
-
"epoch":
|
| 1207 |
"learning_rate": 0.0001,
|
| 1208 |
-
"loss":
|
| 1209 |
"step": 1000
|
| 1210 |
}
|
| 1211 |
],
|
| 1212 |
"logging_steps": 5,
|
| 1213 |
-
"max_steps":
|
| 1214 |
-
"num_train_epochs":
|
| 1215 |
"save_steps": 1000,
|
| 1216 |
-
"total_flos":
|
| 1217 |
"trial_name": null,
|
| 1218 |
"trial_params": null
|
| 1219 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.94240837696335,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
"learning_rate": 0.0001,
|
| 14 |
+
"loss": 1.7979,
|
| 15 |
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.21,
|
| 19 |
"learning_rate": 0.0001,
|
| 20 |
+
"loss": 1.7693,
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.31,
|
| 25 |
"learning_rate": 0.0001,
|
| 26 |
+
"loss": 1.7635,
|
| 27 |
"step": 15
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.42,
|
| 31 |
"learning_rate": 0.0001,
|
| 32 |
+
"loss": 1.7587,
|
| 33 |
"step": 20
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.52,
|
| 37 |
"learning_rate": 0.0001,
|
| 38 |
+
"loss": 1.7384,
|
| 39 |
"step": 25
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.63,
|
| 43 |
"learning_rate": 0.0001,
|
| 44 |
+
"loss": 1.7832,
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.73,
|
| 49 |
"learning_rate": 0.0001,
|
| 50 |
+
"loss": 1.7652,
|
| 51 |
"step": 35
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.84,
|
| 55 |
"learning_rate": 0.0001,
|
| 56 |
+
"loss": 1.7417,
|
| 57 |
"step": 40
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.94,
|
| 61 |
"learning_rate": 0.0001,
|
| 62 |
+
"loss": 1.7321,
|
| 63 |
"step": 45
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 1.05,
|
| 67 |
"learning_rate": 0.0001,
|
| 68 |
+
"loss": 1.6915,
|
| 69 |
"step": 50
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 1.15,
|
| 73 |
"learning_rate": 0.0001,
|
| 74 |
+
"loss": 1.6773,
|
| 75 |
"step": 55
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 1.26,
|
| 79 |
"learning_rate": 0.0001,
|
| 80 |
+
"loss": 1.6713,
|
| 81 |
"step": 60
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.36,
|
| 85 |
"learning_rate": 0.0001,
|
| 86 |
+
"loss": 1.6701,
|
| 87 |
"step": 65
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 1.47,
|
| 91 |
"learning_rate": 0.0001,
|
| 92 |
+
"loss": 1.6941,
|
| 93 |
"step": 70
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 1.57,
|
| 97 |
"learning_rate": 0.0001,
|
| 98 |
+
"loss": 1.6116,
|
| 99 |
"step": 75
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 1.68,
|
| 103 |
"learning_rate": 0.0001,
|
| 104 |
+
"loss": 1.7088,
|
| 105 |
"step": 80
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 1.78,
|
| 109 |
"learning_rate": 0.0001,
|
| 110 |
+
"loss": 1.661,
|
| 111 |
"step": 85
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 1.88,
|
| 115 |
"learning_rate": 0.0001,
|
| 116 |
+
"loss": 1.7321,
|
| 117 |
"step": 90
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 1.99,
|
| 121 |
"learning_rate": 0.0001,
|
| 122 |
+
"loss": 1.6995,
|
| 123 |
"step": 95
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 2.09,
|
| 127 |
"learning_rate": 0.0001,
|
| 128 |
+
"loss": 1.6298,
|
| 129 |
"step": 100
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 2.2,
|
| 133 |
"learning_rate": 0.0001,
|
| 134 |
+
"loss": 1.6231,
|
| 135 |
"step": 105
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 2.3,
|
| 139 |
"learning_rate": 0.0001,
|
| 140 |
+
"loss": 1.6427,
|
| 141 |
"step": 110
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 2.41,
|
| 145 |
"learning_rate": 0.0001,
|
| 146 |
+
"loss": 1.5985,
|
| 147 |
"step": 115
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 2.51,
|
| 151 |
"learning_rate": 0.0001,
|
| 152 |
+
"loss": 1.6267,
|
| 153 |
"step": 120
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 2.62,
|
| 157 |
"learning_rate": 0.0001,
|
| 158 |
+
"loss": 1.5373,
|
| 159 |
"step": 125
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 2.72,
|
| 163 |
"learning_rate": 0.0001,
|
| 164 |
+
"loss": 1.6516,
|
| 165 |
"step": 130
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 2.83,
|
| 169 |
"learning_rate": 0.0001,
|
| 170 |
+
"loss": 1.6861,
|
| 171 |
"step": 135
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 2.93,
|
| 175 |
"learning_rate": 0.0001,
|
| 176 |
+
"loss": 1.6746,
|
| 177 |
"step": 140
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 3.04,
|
| 181 |
"learning_rate": 0.0001,
|
| 182 |
+
"loss": 1.6478,
|
| 183 |
"step": 145
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 3.14,
|
| 187 |
"learning_rate": 0.0001,
|
| 188 |
+
"loss": 1.6206,
|
| 189 |
"step": 150
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 3.25,
|
| 193 |
"learning_rate": 0.0001,
|
| 194 |
+
"loss": 1.5848,
|
| 195 |
"step": 155
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 3.35,
|
| 199 |
"learning_rate": 0.0001,
|
| 200 |
+
"loss": 1.5988,
|
| 201 |
"step": 160
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 3.46,
|
| 205 |
"learning_rate": 0.0001,
|
| 206 |
+
"loss": 1.6014,
|
| 207 |
"step": 165
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 3.56,
|
| 211 |
"learning_rate": 0.0001,
|
| 212 |
+
"loss": 1.5327,
|
| 213 |
"step": 170
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 3.66,
|
| 217 |
"learning_rate": 0.0001,
|
| 218 |
+
"loss": 1.5811,
|
| 219 |
"step": 175
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 3.77,
|
| 223 |
"learning_rate": 0.0001,
|
| 224 |
+
"loss": 1.6202,
|
| 225 |
"step": 180
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 3.87,
|
| 229 |
"learning_rate": 0.0001,
|
| 230 |
+
"loss": 1.5986,
|
| 231 |
"step": 185
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 3.98,
|
| 235 |
"learning_rate": 0.0001,
|
| 236 |
+
"loss": 1.6548,
|
| 237 |
"step": 190
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 4.08,
|
| 241 |
"learning_rate": 0.0001,
|
| 242 |
+
"loss": 1.5288,
|
| 243 |
"step": 195
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 4.19,
|
| 247 |
"learning_rate": 0.0001,
|
| 248 |
+
"loss": 1.5604,
|
| 249 |
"step": 200
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 4.29,
|
| 253 |
"learning_rate": 0.0001,
|
| 254 |
+
"loss": 1.5302,
|
| 255 |
"step": 205
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 4.4,
|
| 259 |
"learning_rate": 0.0001,
|
| 260 |
+
"loss": 1.5751,
|
| 261 |
"step": 210
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 4.5,
|
| 265 |
"learning_rate": 0.0001,
|
| 266 |
+
"loss": 1.5865,
|
| 267 |
"step": 215
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 4.61,
|
| 271 |
"learning_rate": 0.0001,
|
| 272 |
+
"loss": 1.5476,
|
| 273 |
"step": 220
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 4.71,
|
| 277 |
"learning_rate": 0.0001,
|
| 278 |
+
"loss": 1.5636,
|
| 279 |
"step": 225
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 4.82,
|
| 283 |
"learning_rate": 0.0001,
|
| 284 |
+
"loss": 1.5472,
|
| 285 |
"step": 230
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 4.92,
|
| 289 |
"learning_rate": 0.0001,
|
| 290 |
+
"loss": 1.598,
|
| 291 |
"step": 235
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 5.03,
|
| 295 |
"learning_rate": 0.0001,
|
| 296 |
+
"loss": 1.6288,
|
| 297 |
"step": 240
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 5.13,
|
| 301 |
"learning_rate": 0.0001,
|
| 302 |
+
"loss": 1.4983,
|
| 303 |
"step": 245
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 5.24,
|
| 307 |
"learning_rate": 0.0001,
|
| 308 |
+
"loss": 1.5407,
|
| 309 |
"step": 250
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 5.34,
|
| 313 |
"learning_rate": 0.0001,
|
| 314 |
+
"loss": 1.486,
|
| 315 |
"step": 255
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 5.45,
|
| 319 |
"learning_rate": 0.0001,
|
| 320 |
+
"loss": 1.575,
|
| 321 |
"step": 260
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 5.55,
|
| 325 |
"learning_rate": 0.0001,
|
| 326 |
+
"loss": 1.4745,
|
| 327 |
"step": 265
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 5.65,
|
| 331 |
"learning_rate": 0.0001,
|
| 332 |
+
"loss": 1.5294,
|
| 333 |
"step": 270
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 5.76,
|
| 337 |
"learning_rate": 0.0001,
|
| 338 |
+
"loss": 1.5501,
|
| 339 |
"step": 275
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 5.86,
|
| 343 |
"learning_rate": 0.0001,
|
| 344 |
+
"loss": 1.5791,
|
| 345 |
"step": 280
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 5.97,
|
| 349 |
"learning_rate": 0.0001,
|
| 350 |
+
"loss": 1.5737,
|
| 351 |
"step": 285
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 6.07,
|
| 355 |
"learning_rate": 0.0001,
|
| 356 |
+
"loss": 1.5181,
|
| 357 |
"step": 290
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 6.18,
|
| 361 |
"learning_rate": 0.0001,
|
| 362 |
+
"loss": 1.4805,
|
| 363 |
"step": 295
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 6.28,
|
| 367 |
"learning_rate": 0.0001,
|
| 368 |
+
"loss": 1.5105,
|
| 369 |
"step": 300
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 6.39,
|
| 373 |
"learning_rate": 0.0001,
|
| 374 |
+
"loss": 1.5258,
|
| 375 |
"step": 305
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 6.49,
|
| 379 |
"learning_rate": 0.0001,
|
| 380 |
+
"loss": 1.5081,
|
| 381 |
"step": 310
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 6.6,
|
| 385 |
"learning_rate": 0.0001,
|
| 386 |
+
"loss": 1.4754,
|
| 387 |
"step": 315
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 6.7,
|
| 391 |
"learning_rate": 0.0001,
|
| 392 |
+
"loss": 1.5619,
|
| 393 |
"step": 320
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 6.81,
|
| 397 |
"learning_rate": 0.0001,
|
| 398 |
+
"loss": 1.5239,
|
| 399 |
"step": 325
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 6.91,
|
| 403 |
"learning_rate": 0.0001,
|
| 404 |
+
"loss": 1.4911,
|
| 405 |
"step": 330
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 7.02,
|
| 409 |
"learning_rate": 0.0001,
|
| 410 |
+
"loss": 1.5315,
|
| 411 |
"step": 335
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 7.12,
|
| 415 |
"learning_rate": 0.0001,
|
| 416 |
+
"loss": 1.477,
|
| 417 |
"step": 340
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 7.23,
|
| 421 |
"learning_rate": 0.0001,
|
| 422 |
+
"loss": 1.4767,
|
| 423 |
"step": 345
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 7.33,
|
| 427 |
"learning_rate": 0.0001,
|
| 428 |
+
"loss": 1.4628,
|
| 429 |
"step": 350
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 7.43,
|
| 433 |
"learning_rate": 0.0001,
|
| 434 |
+
"loss": 1.4545,
|
| 435 |
"step": 355
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 7.54,
|
| 439 |
"learning_rate": 0.0001,
|
| 440 |
+
"loss": 1.4682,
|
| 441 |
"step": 360
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 7.64,
|
| 445 |
"learning_rate": 0.0001,
|
| 446 |
+
"loss": 1.4623,
|
| 447 |
"step": 365
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 7.75,
|
| 451 |
"learning_rate": 0.0001,
|
| 452 |
+
"loss": 1.5164,
|
| 453 |
"step": 370
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 7.85,
|
| 457 |
"learning_rate": 0.0001,
|
| 458 |
+
"loss": 1.5296,
|
| 459 |
"step": 375
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 7.96,
|
| 463 |
"learning_rate": 0.0001,
|
| 464 |
+
"loss": 1.4896,
|
| 465 |
"step": 380
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 8.06,
|
| 469 |
"learning_rate": 0.0001,
|
| 470 |
+
"loss": 1.4365,
|
| 471 |
"step": 385
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 8.17,
|
| 475 |
"learning_rate": 0.0001,
|
| 476 |
+
"loss": 1.4186,
|
| 477 |
"step": 390
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 8.27,
|
| 481 |
"learning_rate": 0.0001,
|
| 482 |
+
"loss": 1.4972,
|
| 483 |
"step": 395
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 8.38,
|
| 487 |
"learning_rate": 0.0001,
|
| 488 |
+
"loss": 1.4749,
|
| 489 |
"step": 400
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 8.48,
|
| 493 |
"learning_rate": 0.0001,
|
| 494 |
+
"loss": 1.4914,
|
| 495 |
"step": 405
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 8.59,
|
| 499 |
"learning_rate": 0.0001,
|
| 500 |
+
"loss": 1.4091,
|
| 501 |
"step": 410
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 8.69,
|
| 505 |
"learning_rate": 0.0001,
|
| 506 |
+
"loss": 1.4599,
|
| 507 |
"step": 415
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 8.8,
|
| 511 |
"learning_rate": 0.0001,
|
| 512 |
+
"loss": 1.4714,
|
| 513 |
"step": 420
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 8.9,
|
| 517 |
"learning_rate": 0.0001,
|
| 518 |
+
"loss": 1.4636,
|
| 519 |
"step": 425
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 9.01,
|
| 523 |
"learning_rate": 0.0001,
|
| 524 |
+
"loss": 1.4848,
|
| 525 |
"step": 430
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 9.11,
|
| 529 |
"learning_rate": 0.0001,
|
| 530 |
+
"loss": 1.4412,
|
| 531 |
"step": 435
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 9.21,
|
| 535 |
"learning_rate": 0.0001,
|
| 536 |
+
"loss": 1.4057,
|
| 537 |
"step": 440
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 9.32,
|
| 541 |
"learning_rate": 0.0001,
|
| 542 |
+
"loss": 1.4464,
|
| 543 |
"step": 445
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 9.42,
|
| 547 |
"learning_rate": 0.0001,
|
| 548 |
+
"loss": 1.4228,
|
| 549 |
"step": 450
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 9.53,
|
| 553 |
"learning_rate": 0.0001,
|
| 554 |
+
"loss": 1.4315,
|
| 555 |
"step": 455
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 9.63,
|
| 559 |
"learning_rate": 0.0001,
|
| 560 |
+
"loss": 1.442,
|
| 561 |
"step": 460
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 9.74,
|
| 565 |
"learning_rate": 0.0001,
|
| 566 |
+
"loss": 1.4008,
|
| 567 |
"step": 465
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 9.84,
|
| 571 |
"learning_rate": 0.0001,
|
| 572 |
+
"loss": 1.4473,
|
| 573 |
"step": 470
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 9.95,
|
| 577 |
"learning_rate": 0.0001,
|
| 578 |
+
"loss": 1.4487,
|
| 579 |
"step": 475
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 10.05,
|
| 583 |
"learning_rate": 0.0001,
|
| 584 |
+
"loss": 1.476,
|
| 585 |
"step": 480
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 10.16,
|
| 589 |
"learning_rate": 0.0001,
|
| 590 |
+
"loss": 1.3862,
|
| 591 |
"step": 485
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 10.26,
|
| 595 |
"learning_rate": 0.0001,
|
| 596 |
+
"loss": 1.4721,
|
| 597 |
"step": 490
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 10.37,
|
| 601 |
"learning_rate": 0.0001,
|
| 602 |
+
"loss": 1.3839,
|
| 603 |
"step": 495
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 10.47,
|
| 607 |
"learning_rate": 0.0001,
|
| 608 |
+
"loss": 1.3948,
|
| 609 |
"step": 500
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 10.58,
|
| 613 |
"learning_rate": 0.0001,
|
| 614 |
+
"loss": 1.4057,
|
| 615 |
"step": 505
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 10.68,
|
| 619 |
"learning_rate": 0.0001,
|
| 620 |
+
"loss": 1.4158,
|
| 621 |
"step": 510
|
| 622 |
},
|
| 623 |
{
|
| 624 |
+
"epoch": 10.79,
|
| 625 |
"learning_rate": 0.0001,
|
| 626 |
+
"loss": 1.4162,
|
| 627 |
"step": 515
|
| 628 |
},
|
| 629 |
{
|
| 630 |
+
"epoch": 10.89,
|
| 631 |
"learning_rate": 0.0001,
|
| 632 |
+
"loss": 1.468,
|
| 633 |
"step": 520
|
| 634 |
},
|
| 635 |
{
|
| 636 |
+
"epoch": 10.99,
|
| 637 |
"learning_rate": 0.0001,
|
| 638 |
+
"loss": 1.3958,
|
| 639 |
"step": 525
|
| 640 |
},
|
| 641 |
{
|
| 642 |
+
"epoch": 11.1,
|
| 643 |
"learning_rate": 0.0001,
|
| 644 |
+
"loss": 1.4011,
|
| 645 |
"step": 530
|
| 646 |
},
|
| 647 |
{
|
| 648 |
+
"epoch": 11.2,
|
| 649 |
"learning_rate": 0.0001,
|
| 650 |
+
"loss": 1.3536,
|
| 651 |
"step": 535
|
| 652 |
},
|
| 653 |
{
|
| 654 |
+
"epoch": 11.31,
|
| 655 |
"learning_rate": 0.0001,
|
| 656 |
+
"loss": 1.4058,
|
| 657 |
"step": 540
|
| 658 |
},
|
| 659 |
{
|
| 660 |
+
"epoch": 11.41,
|
| 661 |
"learning_rate": 0.0001,
|
| 662 |
+
"loss": 1.4246,
|
| 663 |
"step": 545
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 11.52,
|
| 667 |
"learning_rate": 0.0001,
|
| 668 |
+
"loss": 1.3732,
|
| 669 |
"step": 550
|
| 670 |
},
|
| 671 |
{
|
| 672 |
+
"epoch": 11.62,
|
| 673 |
"learning_rate": 0.0001,
|
| 674 |
+
"loss": 1.4389,
|
| 675 |
"step": 555
|
| 676 |
},
|
| 677 |
{
|
| 678 |
+
"epoch": 11.73,
|
| 679 |
"learning_rate": 0.0001,
|
| 680 |
+
"loss": 1.3583,
|
| 681 |
"step": 560
|
| 682 |
},
|
| 683 |
{
|
| 684 |
+
"epoch": 11.83,
|
| 685 |
"learning_rate": 0.0001,
|
| 686 |
+
"loss": 1.383,
|
| 687 |
"step": 565
|
| 688 |
},
|
| 689 |
{
|
| 690 |
+
"epoch": 11.94,
|
| 691 |
"learning_rate": 0.0001,
|
| 692 |
+
"loss": 1.3902,
|
| 693 |
"step": 570
|
| 694 |
},
|
| 695 |
{
|
| 696 |
+
"epoch": 12.04,
|
| 697 |
"learning_rate": 0.0001,
|
| 698 |
+
"loss": 1.3814,
|
| 699 |
"step": 575
|
| 700 |
},
|
| 701 |
{
|
| 702 |
+
"epoch": 12.15,
|
| 703 |
"learning_rate": 0.0001,
|
| 704 |
+
"loss": 1.3598,
|
| 705 |
"step": 580
|
| 706 |
},
|
| 707 |
{
|
| 708 |
+
"epoch": 12.25,
|
| 709 |
"learning_rate": 0.0001,
|
| 710 |
+
"loss": 1.3826,
|
| 711 |
"step": 585
|
| 712 |
},
|
| 713 |
{
|
| 714 |
+
"epoch": 12.36,
|
| 715 |
"learning_rate": 0.0001,
|
| 716 |
+
"loss": 1.3724,
|
| 717 |
"step": 590
|
| 718 |
},
|
| 719 |
{
|
| 720 |
+
"epoch": 12.46,
|
| 721 |
"learning_rate": 0.0001,
|
| 722 |
+
"loss": 1.3413,
|
| 723 |
"step": 595
|
| 724 |
},
|
| 725 |
{
|
| 726 |
+
"epoch": 12.57,
|
| 727 |
"learning_rate": 0.0001,
|
| 728 |
+
"loss": 1.3661,
|
| 729 |
"step": 600
|
| 730 |
},
|
| 731 |
{
|
| 732 |
+
"epoch": 12.67,
|
| 733 |
"learning_rate": 0.0001,
|
| 734 |
+
"loss": 1.3633,
|
| 735 |
"step": 605
|
| 736 |
},
|
| 737 |
{
|
| 738 |
+
"epoch": 12.77,
|
| 739 |
"learning_rate": 0.0001,
|
| 740 |
+
"loss": 1.3524,
|
| 741 |
"step": 610
|
| 742 |
},
|
| 743 |
{
|
| 744 |
+
"epoch": 12.88,
|
| 745 |
"learning_rate": 0.0001,
|
| 746 |
+
"loss": 1.4059,
|
| 747 |
"step": 615
|
| 748 |
},
|
| 749 |
{
|
| 750 |
+
"epoch": 12.98,
|
| 751 |
"learning_rate": 0.0001,
|
| 752 |
+
"loss": 1.3891,
|
| 753 |
"step": 620
|
| 754 |
},
|
| 755 |
{
|
| 756 |
+
"epoch": 13.09,
|
| 757 |
"learning_rate": 0.0001,
|
| 758 |
+
"loss": 1.4142,
|
| 759 |
"step": 625
|
| 760 |
},
|
| 761 |
{
|
| 762 |
+
"epoch": 13.19,
|
| 763 |
"learning_rate": 0.0001,
|
| 764 |
+
"loss": 1.2695,
|
| 765 |
"step": 630
|
| 766 |
},
|
| 767 |
{
|
| 768 |
+
"epoch": 13.3,
|
| 769 |
"learning_rate": 0.0001,
|
| 770 |
+
"loss": 1.3521,
|
| 771 |
"step": 635
|
| 772 |
},
|
| 773 |
{
|
| 774 |
+
"epoch": 13.4,
|
| 775 |
"learning_rate": 0.0001,
|
| 776 |
+
"loss": 1.264,
|
| 777 |
"step": 640
|
| 778 |
},
|
| 779 |
{
|
| 780 |
+
"epoch": 13.51,
|
| 781 |
"learning_rate": 0.0001,
|
| 782 |
+
"loss": 1.3551,
|
| 783 |
"step": 645
|
| 784 |
},
|
| 785 |
{
|
| 786 |
+
"epoch": 13.61,
|
| 787 |
"learning_rate": 0.0001,
|
| 788 |
+
"loss": 1.4012,
|
| 789 |
"step": 650
|
| 790 |
},
|
| 791 |
{
|
| 792 |
+
"epoch": 13.72,
|
| 793 |
"learning_rate": 0.0001,
|
| 794 |
+
"loss": 1.3733,
|
| 795 |
"step": 655
|
| 796 |
},
|
| 797 |
{
|
| 798 |
+
"epoch": 13.82,
|
| 799 |
"learning_rate": 0.0001,
|
| 800 |
+
"loss": 1.3957,
|
| 801 |
"step": 660
|
| 802 |
},
|
| 803 |
{
|
| 804 |
+
"epoch": 13.93,
|
| 805 |
"learning_rate": 0.0001,
|
| 806 |
+
"loss": 1.3557,
|
| 807 |
"step": 665
|
| 808 |
},
|
| 809 |
{
|
| 810 |
+
"epoch": 14.03,
|
| 811 |
"learning_rate": 0.0001,
|
| 812 |
+
"loss": 1.3732,
|
| 813 |
"step": 670
|
| 814 |
},
|
| 815 |
{
|
| 816 |
+
"epoch": 14.14,
|
| 817 |
"learning_rate": 0.0001,
|
| 818 |
+
"loss": 1.248,
|
| 819 |
"step": 675
|
| 820 |
},
|
| 821 |
{
|
| 822 |
+
"epoch": 14.24,
|
| 823 |
"learning_rate": 0.0001,
|
| 824 |
+
"loss": 1.352,
|
| 825 |
"step": 680
|
| 826 |
},
|
| 827 |
{
|
| 828 |
+
"epoch": 14.35,
|
| 829 |
"learning_rate": 0.0001,
|
| 830 |
+
"loss": 1.3024,
|
| 831 |
"step": 685
|
| 832 |
},
|
| 833 |
{
|
| 834 |
+
"epoch": 14.45,
|
| 835 |
"learning_rate": 0.0001,
|
| 836 |
+
"loss": 1.3343,
|
| 837 |
"step": 690
|
| 838 |
},
|
| 839 |
{
|
| 840 |
+
"epoch": 14.55,
|
| 841 |
"learning_rate": 0.0001,
|
| 842 |
+
"loss": 1.3487,
|
| 843 |
"step": 695
|
| 844 |
},
|
| 845 |
{
|
| 846 |
+
"epoch": 14.66,
|
| 847 |
"learning_rate": 0.0001,
|
| 848 |
+
"loss": 1.3051,
|
| 849 |
"step": 700
|
| 850 |
},
|
| 851 |
{
|
| 852 |
+
"epoch": 14.76,
|
| 853 |
"learning_rate": 0.0001,
|
| 854 |
+
"loss": 1.3331,
|
| 855 |
"step": 705
|
| 856 |
},
|
| 857 |
{
|
| 858 |
+
"epoch": 14.87,
|
| 859 |
"learning_rate": 0.0001,
|
| 860 |
+
"loss": 1.371,
|
| 861 |
"step": 710
|
| 862 |
},
|
| 863 |
{
|
| 864 |
+
"epoch": 14.97,
|
| 865 |
"learning_rate": 0.0001,
|
| 866 |
+
"loss": 1.3779,
|
| 867 |
"step": 715
|
| 868 |
},
|
| 869 |
{
|
| 870 |
+
"epoch": 15.08,
|
| 871 |
"learning_rate": 0.0001,
|
| 872 |
+
"loss": 1.3086,
|
| 873 |
"step": 720
|
| 874 |
},
|
| 875 |
{
|
| 876 |
+
"epoch": 15.18,
|
| 877 |
"learning_rate": 0.0001,
|
| 878 |
+
"loss": 1.2871,
|
| 879 |
"step": 725
|
| 880 |
},
|
| 881 |
{
|
| 882 |
+
"epoch": 15.29,
|
| 883 |
"learning_rate": 0.0001,
|
| 884 |
+
"loss": 1.2899,
|
| 885 |
"step": 730
|
| 886 |
},
|
| 887 |
{
|
| 888 |
+
"epoch": 15.39,
|
| 889 |
"learning_rate": 0.0001,
|
| 890 |
+
"loss": 1.3205,
|
| 891 |
"step": 735
|
| 892 |
},
|
| 893 |
{
|
| 894 |
+
"epoch": 15.5,
|
| 895 |
"learning_rate": 0.0001,
|
| 896 |
+
"loss": 1.3799,
|
| 897 |
"step": 740
|
| 898 |
},
|
| 899 |
{
|
| 900 |
+
"epoch": 15.6,
|
| 901 |
"learning_rate": 0.0001,
|
| 902 |
+
"loss": 1.3251,
|
| 903 |
"step": 745
|
| 904 |
},
|
| 905 |
{
|
| 906 |
+
"epoch": 15.71,
|
| 907 |
"learning_rate": 0.0001,
|
| 908 |
+
"loss": 1.3209,
|
| 909 |
"step": 750
|
| 910 |
},
|
| 911 |
{
|
| 912 |
+
"epoch": 15.81,
|
| 913 |
"learning_rate": 0.0001,
|
| 914 |
+
"loss": 1.3015,
|
| 915 |
"step": 755
|
| 916 |
},
|
| 917 |
{
|
| 918 |
+
"epoch": 15.92,
|
| 919 |
"learning_rate": 0.0001,
|
| 920 |
+
"loss": 1.2979,
|
| 921 |
"step": 760
|
| 922 |
},
|
| 923 |
{
|
| 924 |
+
"epoch": 16.02,
|
| 925 |
"learning_rate": 0.0001,
|
| 926 |
+
"loss": 1.3355,
|
| 927 |
"step": 765
|
| 928 |
},
|
| 929 |
{
|
| 930 |
+
"epoch": 16.13,
|
| 931 |
"learning_rate": 0.0001,
|
| 932 |
+
"loss": 1.2764,
|
| 933 |
"step": 770
|
| 934 |
},
|
| 935 |
{
|
| 936 |
+
"epoch": 16.23,
|
| 937 |
"learning_rate": 0.0001,
|
| 938 |
+
"loss": 1.319,
|
| 939 |
"step": 775
|
| 940 |
},
|
| 941 |
{
|
| 942 |
+
"epoch": 16.34,
|
| 943 |
"learning_rate": 0.0001,
|
| 944 |
+
"loss": 1.2541,
|
| 945 |
"step": 780
|
| 946 |
},
|
| 947 |
{
|
| 948 |
+
"epoch": 16.44,
|
| 949 |
"learning_rate": 0.0001,
|
| 950 |
+
"loss": 1.3251,
|
| 951 |
"step": 785
|
| 952 |
},
|
| 953 |
{
|
| 954 |
+
"epoch": 16.54,
|
| 955 |
"learning_rate": 0.0001,
|
| 956 |
+
"loss": 1.2479,
|
| 957 |
"step": 790
|
| 958 |
},
|
| 959 |
{
|
| 960 |
+
"epoch": 16.65,
|
| 961 |
"learning_rate": 0.0001,
|
| 962 |
+
"loss": 1.2912,
|
| 963 |
"step": 795
|
| 964 |
},
|
| 965 |
{
|
| 966 |
+
"epoch": 16.75,
|
| 967 |
"learning_rate": 0.0001,
|
| 968 |
+
"loss": 1.3064,
|
| 969 |
"step": 800
|
| 970 |
},
|
| 971 |
{
|
| 972 |
+
"epoch": 16.86,
|
| 973 |
"learning_rate": 0.0001,
|
| 974 |
+
"loss": 1.3029,
|
| 975 |
"step": 805
|
| 976 |
},
|
| 977 |
{
|
| 978 |
+
"epoch": 16.96,
|
| 979 |
"learning_rate": 0.0001,
|
| 980 |
+
"loss": 1.3103,
|
| 981 |
"step": 810
|
| 982 |
},
|
| 983 |
{
|
| 984 |
+
"epoch": 17.07,
|
| 985 |
"learning_rate": 0.0001,
|
| 986 |
+
"loss": 1.2733,
|
| 987 |
"step": 815
|
| 988 |
},
|
| 989 |
{
|
| 990 |
+
"epoch": 17.17,
|
| 991 |
"learning_rate": 0.0001,
|
| 992 |
+
"loss": 1.3373,
|
| 993 |
"step": 820
|
| 994 |
},
|
| 995 |
{
|
| 996 |
+
"epoch": 17.28,
|
| 997 |
"learning_rate": 0.0001,
|
| 998 |
+
"loss": 1.2371,
|
| 999 |
"step": 825
|
| 1000 |
},
|
| 1001 |
{
|
| 1002 |
+
"epoch": 17.38,
|
| 1003 |
"learning_rate": 0.0001,
|
| 1004 |
+
"loss": 1.2919,
|
| 1005 |
"step": 830
|
| 1006 |
},
|
| 1007 |
{
|
| 1008 |
+
"epoch": 17.49,
|
| 1009 |
"learning_rate": 0.0001,
|
| 1010 |
+
"loss": 1.2806,
|
| 1011 |
"step": 835
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
+
"epoch": 17.59,
|
| 1015 |
"learning_rate": 0.0001,
|
| 1016 |
+
"loss": 1.2925,
|
| 1017 |
"step": 840
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
+
"epoch": 17.7,
|
| 1021 |
"learning_rate": 0.0001,
|
| 1022 |
+
"loss": 1.2465,
|
| 1023 |
"step": 845
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
+
"epoch": 17.8,
|
| 1027 |
"learning_rate": 0.0001,
|
| 1028 |
+
"loss": 1.2417,
|
| 1029 |
"step": 850
|
| 1030 |
},
|
| 1031 |
{
|
| 1032 |
+
"epoch": 17.91,
|
| 1033 |
"learning_rate": 0.0001,
|
| 1034 |
+
"loss": 1.3069,
|
| 1035 |
"step": 855
|
| 1036 |
},
|
| 1037 |
{
|
| 1038 |
+
"epoch": 18.01,
|
| 1039 |
"learning_rate": 0.0001,
|
| 1040 |
+
"loss": 1.2598,
|
| 1041 |
"step": 860
|
| 1042 |
},
|
| 1043 |
{
|
| 1044 |
+
"epoch": 18.12,
|
| 1045 |
"learning_rate": 0.0001,
|
| 1046 |
+
"loss": 1.2807,
|
| 1047 |
"step": 865
|
| 1048 |
},
|
| 1049 |
{
|
| 1050 |
+
"epoch": 18.22,
|
| 1051 |
"learning_rate": 0.0001,
|
| 1052 |
+
"loss": 1.2548,
|
| 1053 |
"step": 870
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
+
"epoch": 18.32,
|
| 1057 |
"learning_rate": 0.0001,
|
| 1058 |
+
"loss": 1.2531,
|
| 1059 |
"step": 875
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
+
"epoch": 18.43,
|
| 1063 |
"learning_rate": 0.0001,
|
| 1064 |
+
"loss": 1.1707,
|
| 1065 |
"step": 880
|
| 1066 |
},
|
| 1067 |
{
|
| 1068 |
+
"epoch": 18.53,
|
| 1069 |
"learning_rate": 0.0001,
|
| 1070 |
+
"loss": 1.2652,
|
| 1071 |
"step": 885
|
| 1072 |
},
|
| 1073 |
{
|
| 1074 |
+
"epoch": 18.64,
|
| 1075 |
"learning_rate": 0.0001,
|
| 1076 |
+
"loss": 1.2448,
|
| 1077 |
"step": 890
|
| 1078 |
},
|
| 1079 |
{
|
| 1080 |
+
"epoch": 18.74,
|
| 1081 |
"learning_rate": 0.0001,
|
| 1082 |
+
"loss": 1.2969,
|
| 1083 |
"step": 895
|
| 1084 |
},
|
| 1085 |
{
|
| 1086 |
+
"epoch": 18.85,
|
| 1087 |
"learning_rate": 0.0001,
|
| 1088 |
+
"loss": 1.279,
|
| 1089 |
"step": 900
|
| 1090 |
},
|
| 1091 |
{
|
| 1092 |
+
"epoch": 18.95,
|
| 1093 |
"learning_rate": 0.0001,
|
| 1094 |
+
"loss": 1.2714,
|
| 1095 |
"step": 905
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
+
"epoch": 19.06,
|
| 1099 |
"learning_rate": 0.0001,
|
| 1100 |
+
"loss": 1.277,
|
| 1101 |
"step": 910
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
+
"epoch": 19.16,
|
| 1105 |
"learning_rate": 0.0001,
|
| 1106 |
+
"loss": 1.2272,
|
| 1107 |
"step": 915
|
| 1108 |
},
|
| 1109 |
{
|
| 1110 |
+
"epoch": 19.27,
|
| 1111 |
"learning_rate": 0.0001,
|
| 1112 |
+
"loss": 1.2387,
|
| 1113 |
"step": 920
|
| 1114 |
},
|
| 1115 |
{
|
| 1116 |
+
"epoch": 19.37,
|
| 1117 |
"learning_rate": 0.0001,
|
| 1118 |
+
"loss": 1.2494,
|
| 1119 |
"step": 925
|
| 1120 |
},
|
| 1121 |
{
|
| 1122 |
+
"epoch": 19.48,
|
| 1123 |
"learning_rate": 0.0001,
|
| 1124 |
+
"loss": 1.3238,
|
| 1125 |
"step": 930
|
| 1126 |
},
|
| 1127 |
{
|
| 1128 |
+
"epoch": 19.58,
|
| 1129 |
"learning_rate": 0.0001,
|
| 1130 |
+
"loss": 1.258,
|
| 1131 |
"step": 935
|
| 1132 |
},
|
| 1133 |
{
|
| 1134 |
+
"epoch": 19.69,
|
| 1135 |
"learning_rate": 0.0001,
|
| 1136 |
+
"loss": 1.2089,
|
| 1137 |
"step": 940
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
+
"epoch": 19.79,
|
| 1141 |
"learning_rate": 0.0001,
|
| 1142 |
+
"loss": 1.2216,
|
| 1143 |
"step": 945
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
+
"epoch": 19.9,
|
| 1147 |
"learning_rate": 0.0001,
|
| 1148 |
+
"loss": 1.2264,
|
| 1149 |
"step": 950
|
| 1150 |
},
|
| 1151 |
{
|
| 1152 |
+
"epoch": 20.0,
|
| 1153 |
"learning_rate": 0.0001,
|
| 1154 |
+
"loss": 1.2316,
|
| 1155 |
"step": 955
|
| 1156 |
},
|
| 1157 |
{
|
| 1158 |
+
"epoch": 20.1,
|
| 1159 |
"learning_rate": 0.0001,
|
| 1160 |
+
"loss": 1.2038,
|
| 1161 |
"step": 960
|
| 1162 |
},
|
| 1163 |
{
|
| 1164 |
+
"epoch": 20.21,
|
| 1165 |
"learning_rate": 0.0001,
|
| 1166 |
+
"loss": 1.2001,
|
| 1167 |
"step": 965
|
| 1168 |
},
|
| 1169 |
{
|
| 1170 |
+
"epoch": 20.31,
|
| 1171 |
"learning_rate": 0.0001,
|
| 1172 |
+
"loss": 1.2475,
|
| 1173 |
"step": 970
|
| 1174 |
},
|
| 1175 |
{
|
| 1176 |
+
"epoch": 20.42,
|
| 1177 |
"learning_rate": 0.0001,
|
| 1178 |
+
"loss": 1.2513,
|
| 1179 |
"step": 975
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
+
"epoch": 20.52,
|
| 1183 |
"learning_rate": 0.0001,
|
| 1184 |
+
"loss": 1.2635,
|
| 1185 |
"step": 980
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
+
"epoch": 20.63,
|
| 1189 |
"learning_rate": 0.0001,
|
| 1190 |
+
"loss": 1.1817,
|
| 1191 |
"step": 985
|
| 1192 |
},
|
| 1193 |
{
|
| 1194 |
+
"epoch": 20.73,
|
| 1195 |
"learning_rate": 0.0001,
|
| 1196 |
+
"loss": 1.2474,
|
| 1197 |
"step": 990
|
| 1198 |
},
|
| 1199 |
{
|
| 1200 |
+
"epoch": 20.84,
|
| 1201 |
"learning_rate": 0.0001,
|
| 1202 |
+
"loss": 1.2309,
|
| 1203 |
"step": 995
|
| 1204 |
},
|
| 1205 |
{
|
| 1206 |
+
"epoch": 20.94,
|
| 1207 |
"learning_rate": 0.0001,
|
| 1208 |
+
"loss": 1.1871,
|
| 1209 |
"step": 1000
|
| 1210 |
}
|
| 1211 |
],
|
| 1212 |
"logging_steps": 5,
|
| 1213 |
+
"max_steps": 1175,
|
| 1214 |
+
"num_train_epochs": 25,
|
| 1215 |
"save_steps": 1000,
|
| 1216 |
+
"total_flos": 3737124864000000.0,
|
| 1217 |
"trial_name": null,
|
| 1218 |
"trial_params": null
|
| 1219 |
}
|
checkpoint-1000/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4027
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99eae4f19c67acd3f95af7fb6d5de5d000413bb16589081d407c5b736dcd91ad
|
| 3 |
size 4027
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "./
|
| 3 |
"architectures": [
|
| 4 |
"LlamaForCausalLM"
|
| 5 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "./output5",
|
| 3 |
"architectures": [
|
| 4 |
"LlamaForCausalLM"
|
| 5 |
],
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 70790896
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26a4787242d38406a2c2c06b43a99efd856e660e0f388a40204bfb62fd1ac8d0
|
| 3 |
size 70790896
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 1.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 24.61,
|
| 3 |
+
"train_loss": 1.3988467488390335,
|
| 4 |
+
"train_runtime": 1205.3351,
|
| 5 |
+
"train_samples": 1528,
|
| 6 |
+
"train_samples_per_second": 31.692,
|
| 7 |
+
"train_steps_per_second": 0.975
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,70 +1,1438 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"learning_rate": 0.
|
| 14 |
-
"loss":
|
| 15 |
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 0.
|
| 20 |
-
"loss": 1.
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 0.
|
| 26 |
-
"loss": 1.
|
| 27 |
"step": 15
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 0.
|
| 32 |
-
"loss": 1.
|
| 33 |
"step": 20
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 0.
|
| 38 |
-
"loss": 1.
|
| 39 |
"step": 25
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 0.
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 0.
|
| 50 |
-
"loss":
|
| 51 |
"step": 35
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"
|
| 56 |
-
"
|
| 57 |
-
"
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 61 |
}
|
| 62 |
],
|
| 63 |
"logging_steps": 5,
|
| 64 |
-
"max_steps":
|
| 65 |
-
"num_train_epochs":
|
| 66 |
"save_steps": 1000,
|
| 67 |
-
"total_flos":
|
| 68 |
"trial_name": null,
|
| 69 |
"trial_params": null
|
| 70 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 24.607329842931936,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1175,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.1,
|
| 13 |
+
"learning_rate": 0.0001,
|
| 14 |
+
"loss": 1.7979,
|
| 15 |
"step": 5
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.21,
|
| 19 |
+
"learning_rate": 0.0001,
|
| 20 |
+
"loss": 1.7693,
|
| 21 |
"step": 10
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.31,
|
| 25 |
+
"learning_rate": 0.0001,
|
| 26 |
+
"loss": 1.7635,
|
| 27 |
"step": 15
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.42,
|
| 31 |
+
"learning_rate": 0.0001,
|
| 32 |
+
"loss": 1.7587,
|
| 33 |
"step": 20
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.52,
|
| 37 |
+
"learning_rate": 0.0001,
|
| 38 |
+
"loss": 1.7384,
|
| 39 |
"step": 25
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.63,
|
| 43 |
+
"learning_rate": 0.0001,
|
| 44 |
+
"loss": 1.7832,
|
| 45 |
"step": 30
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.73,
|
| 49 |
+
"learning_rate": 0.0001,
|
| 50 |
+
"loss": 1.7652,
|
| 51 |
"step": 35
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.84,
|
| 55 |
+
"learning_rate": 0.0001,
|
| 56 |
+
"loss": 1.7417,
|
| 57 |
+
"step": 40
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.94,
|
| 61 |
+
"learning_rate": 0.0001,
|
| 62 |
+
"loss": 1.7321,
|
| 63 |
+
"step": 45
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.05,
|
| 67 |
+
"learning_rate": 0.0001,
|
| 68 |
+
"loss": 1.6915,
|
| 69 |
+
"step": 50
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.15,
|
| 73 |
+
"learning_rate": 0.0001,
|
| 74 |
+
"loss": 1.6773,
|
| 75 |
+
"step": 55
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 1.26,
|
| 79 |
+
"learning_rate": 0.0001,
|
| 80 |
+
"loss": 1.6713,
|
| 81 |
+
"step": 60
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.36,
|
| 85 |
+
"learning_rate": 0.0001,
|
| 86 |
+
"loss": 1.6701,
|
| 87 |
+
"step": 65
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.47,
|
| 91 |
+
"learning_rate": 0.0001,
|
| 92 |
+
"loss": 1.6941,
|
| 93 |
+
"step": 70
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.57,
|
| 97 |
+
"learning_rate": 0.0001,
|
| 98 |
+
"loss": 1.6116,
|
| 99 |
+
"step": 75
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.68,
|
| 103 |
+
"learning_rate": 0.0001,
|
| 104 |
+
"loss": 1.7088,
|
| 105 |
+
"step": 80
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.78,
|
| 109 |
+
"learning_rate": 0.0001,
|
| 110 |
+
"loss": 1.661,
|
| 111 |
+
"step": 85
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 1.88,
|
| 115 |
+
"learning_rate": 0.0001,
|
| 116 |
+
"loss": 1.7321,
|
| 117 |
+
"step": 90
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.99,
|
| 121 |
+
"learning_rate": 0.0001,
|
| 122 |
+
"loss": 1.6995,
|
| 123 |
+
"step": 95
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.09,
|
| 127 |
+
"learning_rate": 0.0001,
|
| 128 |
+
"loss": 1.6298,
|
| 129 |
+
"step": 100
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 2.2,
|
| 133 |
+
"learning_rate": 0.0001,
|
| 134 |
+
"loss": 1.6231,
|
| 135 |
+
"step": 105
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.3,
|
| 139 |
+
"learning_rate": 0.0001,
|
| 140 |
+
"loss": 1.6427,
|
| 141 |
+
"step": 110
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 2.41,
|
| 145 |
+
"learning_rate": 0.0001,
|
| 146 |
+
"loss": 1.5985,
|
| 147 |
+
"step": 115
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 2.51,
|
| 151 |
+
"learning_rate": 0.0001,
|
| 152 |
+
"loss": 1.6267,
|
| 153 |
+
"step": 120
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 2.62,
|
| 157 |
+
"learning_rate": 0.0001,
|
| 158 |
+
"loss": 1.5373,
|
| 159 |
+
"step": 125
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.72,
|
| 163 |
+
"learning_rate": 0.0001,
|
| 164 |
+
"loss": 1.6516,
|
| 165 |
+
"step": 130
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.83,
|
| 169 |
+
"learning_rate": 0.0001,
|
| 170 |
+
"loss": 1.6861,
|
| 171 |
+
"step": 135
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.93,
|
| 175 |
+
"learning_rate": 0.0001,
|
| 176 |
+
"loss": 1.6746,
|
| 177 |
+
"step": 140
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 3.04,
|
| 181 |
+
"learning_rate": 0.0001,
|
| 182 |
+
"loss": 1.6478,
|
| 183 |
+
"step": 145
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.14,
|
| 187 |
+
"learning_rate": 0.0001,
|
| 188 |
+
"loss": 1.6206,
|
| 189 |
+
"step": 150
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 3.25,
|
| 193 |
+
"learning_rate": 0.0001,
|
| 194 |
+
"loss": 1.5848,
|
| 195 |
+
"step": 155
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 3.35,
|
| 199 |
+
"learning_rate": 0.0001,
|
| 200 |
+
"loss": 1.5988,
|
| 201 |
+
"step": 160
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 3.46,
|
| 205 |
+
"learning_rate": 0.0001,
|
| 206 |
+
"loss": 1.6014,
|
| 207 |
+
"step": 165
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 3.56,
|
| 211 |
+
"learning_rate": 0.0001,
|
| 212 |
+
"loss": 1.5327,
|
| 213 |
+
"step": 170
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 3.66,
|
| 217 |
+
"learning_rate": 0.0001,
|
| 218 |
+
"loss": 1.5811,
|
| 219 |
+
"step": 175
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.77,
|
| 223 |
+
"learning_rate": 0.0001,
|
| 224 |
+
"loss": 1.6202,
|
| 225 |
+
"step": 180
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.87,
|
| 229 |
+
"learning_rate": 0.0001,
|
| 230 |
+
"loss": 1.5986,
|
| 231 |
+
"step": 185
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 3.98,
|
| 235 |
+
"learning_rate": 0.0001,
|
| 236 |
+
"loss": 1.6548,
|
| 237 |
+
"step": 190
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 4.08,
|
| 241 |
+
"learning_rate": 0.0001,
|
| 242 |
+
"loss": 1.5288,
|
| 243 |
+
"step": 195
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.19,
|
| 247 |
+
"learning_rate": 0.0001,
|
| 248 |
+
"loss": 1.5604,
|
| 249 |
+
"step": 200
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 4.29,
|
| 253 |
+
"learning_rate": 0.0001,
|
| 254 |
+
"loss": 1.5302,
|
| 255 |
+
"step": 205
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 4.4,
|
| 259 |
+
"learning_rate": 0.0001,
|
| 260 |
+
"loss": 1.5751,
|
| 261 |
+
"step": 210
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.5,
|
| 265 |
+
"learning_rate": 0.0001,
|
| 266 |
+
"loss": 1.5865,
|
| 267 |
+
"step": 215
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 4.61,
|
| 271 |
+
"learning_rate": 0.0001,
|
| 272 |
+
"loss": 1.5476,
|
| 273 |
+
"step": 220
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 4.71,
|
| 277 |
+
"learning_rate": 0.0001,
|
| 278 |
+
"loss": 1.5636,
|
| 279 |
+
"step": 225
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 4.82,
|
| 283 |
+
"learning_rate": 0.0001,
|
| 284 |
+
"loss": 1.5472,
|
| 285 |
+
"step": 230
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.92,
|
| 289 |
+
"learning_rate": 0.0001,
|
| 290 |
+
"loss": 1.598,
|
| 291 |
+
"step": 235
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 5.03,
|
| 295 |
+
"learning_rate": 0.0001,
|
| 296 |
+
"loss": 1.6288,
|
| 297 |
+
"step": 240
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 5.13,
|
| 301 |
+
"learning_rate": 0.0001,
|
| 302 |
+
"loss": 1.4983,
|
| 303 |
+
"step": 245
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.24,
|
| 307 |
+
"learning_rate": 0.0001,
|
| 308 |
+
"loss": 1.5407,
|
| 309 |
+
"step": 250
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 5.34,
|
| 313 |
+
"learning_rate": 0.0001,
|
| 314 |
+
"loss": 1.486,
|
| 315 |
+
"step": 255
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 5.45,
|
| 319 |
+
"learning_rate": 0.0001,
|
| 320 |
+
"loss": 1.575,
|
| 321 |
+
"step": 260
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 5.55,
|
| 325 |
+
"learning_rate": 0.0001,
|
| 326 |
+
"loss": 1.4745,
|
| 327 |
+
"step": 265
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 5.65,
|
| 331 |
+
"learning_rate": 0.0001,
|
| 332 |
+
"loss": 1.5294,
|
| 333 |
+
"step": 270
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 5.76,
|
| 337 |
+
"learning_rate": 0.0001,
|
| 338 |
+
"loss": 1.5501,
|
| 339 |
+
"step": 275
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 5.86,
|
| 343 |
+
"learning_rate": 0.0001,
|
| 344 |
+
"loss": 1.5791,
|
| 345 |
+
"step": 280
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.97,
|
| 349 |
+
"learning_rate": 0.0001,
|
| 350 |
+
"loss": 1.5737,
|
| 351 |
+
"step": 285
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 6.07,
|
| 355 |
+
"learning_rate": 0.0001,
|
| 356 |
+
"loss": 1.5181,
|
| 357 |
+
"step": 290
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 6.18,
|
| 361 |
+
"learning_rate": 0.0001,
|
| 362 |
+
"loss": 1.4805,
|
| 363 |
+
"step": 295
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.28,
|
| 367 |
+
"learning_rate": 0.0001,
|
| 368 |
+
"loss": 1.5105,
|
| 369 |
+
"step": 300
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 6.39,
|
| 373 |
+
"learning_rate": 0.0001,
|
| 374 |
+
"loss": 1.5258,
|
| 375 |
+
"step": 305
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 6.49,
|
| 379 |
+
"learning_rate": 0.0001,
|
| 380 |
+
"loss": 1.5081,
|
| 381 |
+
"step": 310
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 6.6,
|
| 385 |
+
"learning_rate": 0.0001,
|
| 386 |
+
"loss": 1.4754,
|
| 387 |
+
"step": 315
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.7,
|
| 391 |
+
"learning_rate": 0.0001,
|
| 392 |
+
"loss": 1.5619,
|
| 393 |
+
"step": 320
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 6.81,
|
| 397 |
+
"learning_rate": 0.0001,
|
| 398 |
+
"loss": 1.5239,
|
| 399 |
+
"step": 325
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 6.91,
|
| 403 |
+
"learning_rate": 0.0001,
|
| 404 |
+
"loss": 1.4911,
|
| 405 |
+
"step": 330
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 7.02,
|
| 409 |
+
"learning_rate": 0.0001,
|
| 410 |
+
"loss": 1.5315,
|
| 411 |
+
"step": 335
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 7.12,
|
| 415 |
+
"learning_rate": 0.0001,
|
| 416 |
+
"loss": 1.477,
|
| 417 |
+
"step": 340
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 7.23,
|
| 421 |
+
"learning_rate": 0.0001,
|
| 422 |
+
"loss": 1.4767,
|
| 423 |
+
"step": 345
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.33,
|
| 427 |
+
"learning_rate": 0.0001,
|
| 428 |
+
"loss": 1.4628,
|
| 429 |
+
"step": 350
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.43,
|
| 433 |
+
"learning_rate": 0.0001,
|
| 434 |
+
"loss": 1.4545,
|
| 435 |
+
"step": 355
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 7.54,
|
| 439 |
+
"learning_rate": 0.0001,
|
| 440 |
+
"loss": 1.4682,
|
| 441 |
+
"step": 360
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 7.64,
|
| 445 |
+
"learning_rate": 0.0001,
|
| 446 |
+
"loss": 1.4623,
|
| 447 |
+
"step": 365
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 7.75,
|
| 451 |
+
"learning_rate": 0.0001,
|
| 452 |
+
"loss": 1.5164,
|
| 453 |
+
"step": 370
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 7.85,
|
| 457 |
+
"learning_rate": 0.0001,
|
| 458 |
+
"loss": 1.5296,
|
| 459 |
+
"step": 375
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 7.96,
|
| 463 |
+
"learning_rate": 0.0001,
|
| 464 |
+
"loss": 1.4896,
|
| 465 |
+
"step": 380
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 8.06,
|
| 469 |
+
"learning_rate": 0.0001,
|
| 470 |
+
"loss": 1.4365,
|
| 471 |
+
"step": 385
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 8.17,
|
| 475 |
+
"learning_rate": 0.0001,
|
| 476 |
+
"loss": 1.4186,
|
| 477 |
+
"step": 390
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 8.27,
|
| 481 |
+
"learning_rate": 0.0001,
|
| 482 |
+
"loss": 1.4972,
|
| 483 |
+
"step": 395
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 8.38,
|
| 487 |
+
"learning_rate": 0.0001,
|
| 488 |
+
"loss": 1.4749,
|
| 489 |
+
"step": 400
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 8.48,
|
| 493 |
+
"learning_rate": 0.0001,
|
| 494 |
+
"loss": 1.4914,
|
| 495 |
+
"step": 405
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 8.59,
|
| 499 |
+
"learning_rate": 0.0001,
|
| 500 |
+
"loss": 1.4091,
|
| 501 |
+
"step": 410
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 8.69,
|
| 505 |
+
"learning_rate": 0.0001,
|
| 506 |
+
"loss": 1.4599,
|
| 507 |
+
"step": 415
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 8.8,
|
| 511 |
+
"learning_rate": 0.0001,
|
| 512 |
+
"loss": 1.4714,
|
| 513 |
+
"step": 420
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 8.9,
|
| 517 |
+
"learning_rate": 0.0001,
|
| 518 |
+
"loss": 1.4636,
|
| 519 |
+
"step": 425
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 9.01,
|
| 523 |
+
"learning_rate": 0.0001,
|
| 524 |
+
"loss": 1.4848,
|
| 525 |
+
"step": 430
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 9.11,
|
| 529 |
+
"learning_rate": 0.0001,
|
| 530 |
+
"loss": 1.4412,
|
| 531 |
+
"step": 435
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 9.21,
|
| 535 |
+
"learning_rate": 0.0001,
|
| 536 |
+
"loss": 1.4057,
|
| 537 |
+
"step": 440
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 9.32,
|
| 541 |
+
"learning_rate": 0.0001,
|
| 542 |
+
"loss": 1.4464,
|
| 543 |
+
"step": 445
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.42,
|
| 547 |
+
"learning_rate": 0.0001,
|
| 548 |
+
"loss": 1.4228,
|
| 549 |
+
"step": 450
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 9.53,
|
| 553 |
+
"learning_rate": 0.0001,
|
| 554 |
+
"loss": 1.4315,
|
| 555 |
+
"step": 455
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 9.63,
|
| 559 |
+
"learning_rate": 0.0001,
|
| 560 |
+
"loss": 1.442,
|
| 561 |
+
"step": 460
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 9.74,
|
| 565 |
+
"learning_rate": 0.0001,
|
| 566 |
+
"loss": 1.4008,
|
| 567 |
+
"step": 465
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 9.84,
|
| 571 |
+
"learning_rate": 0.0001,
|
| 572 |
+
"loss": 1.4473,
|
| 573 |
+
"step": 470
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 9.95,
|
| 577 |
+
"learning_rate": 0.0001,
|
| 578 |
+
"loss": 1.4487,
|
| 579 |
+
"step": 475
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 10.05,
|
| 583 |
+
"learning_rate": 0.0001,
|
| 584 |
+
"loss": 1.476,
|
| 585 |
+
"step": 480
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 10.16,
|
| 589 |
+
"learning_rate": 0.0001,
|
| 590 |
+
"loss": 1.3862,
|
| 591 |
+
"step": 485
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 10.26,
|
| 595 |
+
"learning_rate": 0.0001,
|
| 596 |
+
"loss": 1.4721,
|
| 597 |
+
"step": 490
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 10.37,
|
| 601 |
+
"learning_rate": 0.0001,
|
| 602 |
+
"loss": 1.3839,
|
| 603 |
+
"step": 495
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.47,
|
| 607 |
+
"learning_rate": 0.0001,
|
| 608 |
+
"loss": 1.3948,
|
| 609 |
+
"step": 500
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 10.58,
|
| 613 |
+
"learning_rate": 0.0001,
|
| 614 |
+
"loss": 1.4057,
|
| 615 |
+
"step": 505
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 10.68,
|
| 619 |
+
"learning_rate": 0.0001,
|
| 620 |
+
"loss": 1.4158,
|
| 621 |
+
"step": 510
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 10.79,
|
| 625 |
+
"learning_rate": 0.0001,
|
| 626 |
+
"loss": 1.4162,
|
| 627 |
+
"step": 515
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 10.89,
|
| 631 |
+
"learning_rate": 0.0001,
|
| 632 |
+
"loss": 1.468,
|
| 633 |
+
"step": 520
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 10.99,
|
| 637 |
+
"learning_rate": 0.0001,
|
| 638 |
+
"loss": 1.3958,
|
| 639 |
+
"step": 525
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 11.1,
|
| 643 |
+
"learning_rate": 0.0001,
|
| 644 |
+
"loss": 1.4011,
|
| 645 |
+
"step": 530
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 11.2,
|
| 649 |
+
"learning_rate": 0.0001,
|
| 650 |
+
"loss": 1.3536,
|
| 651 |
+
"step": 535
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 11.31,
|
| 655 |
+
"learning_rate": 0.0001,
|
| 656 |
+
"loss": 1.4058,
|
| 657 |
+
"step": 540
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 11.41,
|
| 661 |
+
"learning_rate": 0.0001,
|
| 662 |
+
"loss": 1.4246,
|
| 663 |
+
"step": 545
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 11.52,
|
| 667 |
+
"learning_rate": 0.0001,
|
| 668 |
+
"loss": 1.3732,
|
| 669 |
+
"step": 550
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 11.62,
|
| 673 |
+
"learning_rate": 0.0001,
|
| 674 |
+
"loss": 1.4389,
|
| 675 |
+
"step": 555
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 11.73,
|
| 679 |
+
"learning_rate": 0.0001,
|
| 680 |
+
"loss": 1.3583,
|
| 681 |
+
"step": 560
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 11.83,
|
| 685 |
+
"learning_rate": 0.0001,
|
| 686 |
+
"loss": 1.383,
|
| 687 |
+
"step": 565
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 11.94,
|
| 691 |
+
"learning_rate": 0.0001,
|
| 692 |
+
"loss": 1.3902,
|
| 693 |
+
"step": 570
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 12.04,
|
| 697 |
+
"learning_rate": 0.0001,
|
| 698 |
+
"loss": 1.3814,
|
| 699 |
+
"step": 575
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 12.15,
|
| 703 |
+
"learning_rate": 0.0001,
|
| 704 |
+
"loss": 1.3598,
|
| 705 |
+
"step": 580
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 12.25,
|
| 709 |
+
"learning_rate": 0.0001,
|
| 710 |
+
"loss": 1.3826,
|
| 711 |
+
"step": 585
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 12.36,
|
| 715 |
+
"learning_rate": 0.0001,
|
| 716 |
+
"loss": 1.3724,
|
| 717 |
+
"step": 590
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 12.46,
|
| 721 |
+
"learning_rate": 0.0001,
|
| 722 |
+
"loss": 1.3413,
|
| 723 |
+
"step": 595
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.57,
|
| 727 |
+
"learning_rate": 0.0001,
|
| 728 |
+
"loss": 1.3661,
|
| 729 |
+
"step": 600
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 12.67,
|
| 733 |
+
"learning_rate": 0.0001,
|
| 734 |
+
"loss": 1.3633,
|
| 735 |
+
"step": 605
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 12.77,
|
| 739 |
+
"learning_rate": 0.0001,
|
| 740 |
+
"loss": 1.3524,
|
| 741 |
+
"step": 610
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 12.88,
|
| 745 |
+
"learning_rate": 0.0001,
|
| 746 |
+
"loss": 1.4059,
|
| 747 |
+
"step": 615
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 12.98,
|
| 751 |
+
"learning_rate": 0.0001,
|
| 752 |
+
"loss": 1.3891,
|
| 753 |
+
"step": 620
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 13.09,
|
| 757 |
+
"learning_rate": 0.0001,
|
| 758 |
+
"loss": 1.4142,
|
| 759 |
+
"step": 625
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 13.19,
|
| 763 |
+
"learning_rate": 0.0001,
|
| 764 |
+
"loss": 1.2695,
|
| 765 |
+
"step": 630
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 13.3,
|
| 769 |
+
"learning_rate": 0.0001,
|
| 770 |
+
"loss": 1.3521,
|
| 771 |
+
"step": 635
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 13.4,
|
| 775 |
+
"learning_rate": 0.0001,
|
| 776 |
+
"loss": 1.264,
|
| 777 |
+
"step": 640
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 13.51,
|
| 781 |
+
"learning_rate": 0.0001,
|
| 782 |
+
"loss": 1.3551,
|
| 783 |
+
"step": 645
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 13.61,
|
| 787 |
+
"learning_rate": 0.0001,
|
| 788 |
+
"loss": 1.4012,
|
| 789 |
+
"step": 650
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 13.72,
|
| 793 |
+
"learning_rate": 0.0001,
|
| 794 |
+
"loss": 1.3733,
|
| 795 |
+
"step": 655
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 13.82,
|
| 799 |
+
"learning_rate": 0.0001,
|
| 800 |
+
"loss": 1.3957,
|
| 801 |
+
"step": 660
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 13.93,
|
| 805 |
+
"learning_rate": 0.0001,
|
| 806 |
+
"loss": 1.3557,
|
| 807 |
+
"step": 665
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 14.03,
|
| 811 |
+
"learning_rate": 0.0001,
|
| 812 |
+
"loss": 1.3732,
|
| 813 |
+
"step": 670
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 14.14,
|
| 817 |
+
"learning_rate": 0.0001,
|
| 818 |
+
"loss": 1.248,
|
| 819 |
+
"step": 675
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 14.24,
|
| 823 |
+
"learning_rate": 0.0001,
|
| 824 |
+
"loss": 1.352,
|
| 825 |
+
"step": 680
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 14.35,
|
| 829 |
+
"learning_rate": 0.0001,
|
| 830 |
+
"loss": 1.3024,
|
| 831 |
+
"step": 685
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 14.45,
|
| 835 |
+
"learning_rate": 0.0001,
|
| 836 |
+
"loss": 1.3343,
|
| 837 |
+
"step": 690
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 14.55,
|
| 841 |
+
"learning_rate": 0.0001,
|
| 842 |
+
"loss": 1.3487,
|
| 843 |
+
"step": 695
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 14.66,
|
| 847 |
+
"learning_rate": 0.0001,
|
| 848 |
+
"loss": 1.3051,
|
| 849 |
+
"step": 700
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 14.76,
|
| 853 |
+
"learning_rate": 0.0001,
|
| 854 |
+
"loss": 1.3331,
|
| 855 |
+
"step": 705
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 14.87,
|
| 859 |
+
"learning_rate": 0.0001,
|
| 860 |
+
"loss": 1.371,
|
| 861 |
+
"step": 710
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 14.97,
|
| 865 |
+
"learning_rate": 0.0001,
|
| 866 |
+
"loss": 1.3779,
|
| 867 |
+
"step": 715
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 15.08,
|
| 871 |
+
"learning_rate": 0.0001,
|
| 872 |
+
"loss": 1.3086,
|
| 873 |
+
"step": 720
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 15.18,
|
| 877 |
+
"learning_rate": 0.0001,
|
| 878 |
+
"loss": 1.2871,
|
| 879 |
+
"step": 725
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 15.29,
|
| 883 |
+
"learning_rate": 0.0001,
|
| 884 |
+
"loss": 1.2899,
|
| 885 |
+
"step": 730
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 15.39,
|
| 889 |
+
"learning_rate": 0.0001,
|
| 890 |
+
"loss": 1.3205,
|
| 891 |
+
"step": 735
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 15.5,
|
| 895 |
+
"learning_rate": 0.0001,
|
| 896 |
+
"loss": 1.3799,
|
| 897 |
+
"step": 740
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 15.6,
|
| 901 |
+
"learning_rate": 0.0001,
|
| 902 |
+
"loss": 1.3251,
|
| 903 |
+
"step": 745
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 15.71,
|
| 907 |
+
"learning_rate": 0.0001,
|
| 908 |
+
"loss": 1.3209,
|
| 909 |
+
"step": 750
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 15.81,
|
| 913 |
+
"learning_rate": 0.0001,
|
| 914 |
+
"loss": 1.3015,
|
| 915 |
+
"step": 755
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 15.92,
|
| 919 |
+
"learning_rate": 0.0001,
|
| 920 |
+
"loss": 1.2979,
|
| 921 |
+
"step": 760
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 16.02,
|
| 925 |
+
"learning_rate": 0.0001,
|
| 926 |
+
"loss": 1.3355,
|
| 927 |
+
"step": 765
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 16.13,
|
| 931 |
+
"learning_rate": 0.0001,
|
| 932 |
+
"loss": 1.2764,
|
| 933 |
+
"step": 770
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 16.23,
|
| 937 |
+
"learning_rate": 0.0001,
|
| 938 |
+
"loss": 1.319,
|
| 939 |
+
"step": 775
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 16.34,
|
| 943 |
+
"learning_rate": 0.0001,
|
| 944 |
+
"loss": 1.2541,
|
| 945 |
+
"step": 780
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 16.44,
|
| 949 |
+
"learning_rate": 0.0001,
|
| 950 |
+
"loss": 1.3251,
|
| 951 |
+
"step": 785
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 16.54,
|
| 955 |
+
"learning_rate": 0.0001,
|
| 956 |
+
"loss": 1.2479,
|
| 957 |
+
"step": 790
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 16.65,
|
| 961 |
+
"learning_rate": 0.0001,
|
| 962 |
+
"loss": 1.2912,
|
| 963 |
+
"step": 795
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 16.75,
|
| 967 |
+
"learning_rate": 0.0001,
|
| 968 |
+
"loss": 1.3064,
|
| 969 |
+
"step": 800
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 16.86,
|
| 973 |
+
"learning_rate": 0.0001,
|
| 974 |
+
"loss": 1.3029,
|
| 975 |
+
"step": 805
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 16.96,
|
| 979 |
+
"learning_rate": 0.0001,
|
| 980 |
+
"loss": 1.3103,
|
| 981 |
+
"step": 810
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 17.07,
|
| 985 |
+
"learning_rate": 0.0001,
|
| 986 |
+
"loss": 1.2733,
|
| 987 |
+
"step": 815
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 17.17,
|
| 991 |
+
"learning_rate": 0.0001,
|
| 992 |
+
"loss": 1.3373,
|
| 993 |
+
"step": 820
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 17.28,
|
| 997 |
+
"learning_rate": 0.0001,
|
| 998 |
+
"loss": 1.2371,
|
| 999 |
+
"step": 825
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 17.38,
|
| 1003 |
+
"learning_rate": 0.0001,
|
| 1004 |
+
"loss": 1.2919,
|
| 1005 |
+
"step": 830
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 17.49,
|
| 1009 |
+
"learning_rate": 0.0001,
|
| 1010 |
+
"loss": 1.2806,
|
| 1011 |
+
"step": 835
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 17.59,
|
| 1015 |
+
"learning_rate": 0.0001,
|
| 1016 |
+
"loss": 1.2925,
|
| 1017 |
+
"step": 840
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 17.7,
|
| 1021 |
+
"learning_rate": 0.0001,
|
| 1022 |
+
"loss": 1.2465,
|
| 1023 |
+
"step": 845
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 17.8,
|
| 1027 |
+
"learning_rate": 0.0001,
|
| 1028 |
+
"loss": 1.2417,
|
| 1029 |
+
"step": 850
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 17.91,
|
| 1033 |
+
"learning_rate": 0.0001,
|
| 1034 |
+
"loss": 1.3069,
|
| 1035 |
+
"step": 855
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 18.01,
|
| 1039 |
+
"learning_rate": 0.0001,
|
| 1040 |
+
"loss": 1.2598,
|
| 1041 |
+
"step": 860
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 18.12,
|
| 1045 |
+
"learning_rate": 0.0001,
|
| 1046 |
+
"loss": 1.2807,
|
| 1047 |
+
"step": 865
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 18.22,
|
| 1051 |
+
"learning_rate": 0.0001,
|
| 1052 |
+
"loss": 1.2548,
|
| 1053 |
+
"step": 870
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 18.32,
|
| 1057 |
+
"learning_rate": 0.0001,
|
| 1058 |
+
"loss": 1.2531,
|
| 1059 |
+
"step": 875
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 18.43,
|
| 1063 |
+
"learning_rate": 0.0001,
|
| 1064 |
+
"loss": 1.1707,
|
| 1065 |
+
"step": 880
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 18.53,
|
| 1069 |
+
"learning_rate": 0.0001,
|
| 1070 |
+
"loss": 1.2652,
|
| 1071 |
+
"step": 885
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 18.64,
|
| 1075 |
+
"learning_rate": 0.0001,
|
| 1076 |
+
"loss": 1.2448,
|
| 1077 |
+
"step": 890
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 18.74,
|
| 1081 |
+
"learning_rate": 0.0001,
|
| 1082 |
+
"loss": 1.2969,
|
| 1083 |
+
"step": 895
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 18.85,
|
| 1087 |
+
"learning_rate": 0.0001,
|
| 1088 |
+
"loss": 1.279,
|
| 1089 |
+
"step": 900
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 18.95,
|
| 1093 |
+
"learning_rate": 0.0001,
|
| 1094 |
+
"loss": 1.2714,
|
| 1095 |
+
"step": 905
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 19.06,
|
| 1099 |
+
"learning_rate": 0.0001,
|
| 1100 |
+
"loss": 1.277,
|
| 1101 |
+
"step": 910
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 19.16,
|
| 1105 |
+
"learning_rate": 0.0001,
|
| 1106 |
+
"loss": 1.2272,
|
| 1107 |
+
"step": 915
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 19.27,
|
| 1111 |
+
"learning_rate": 0.0001,
|
| 1112 |
+
"loss": 1.2387,
|
| 1113 |
+
"step": 920
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 19.37,
|
| 1117 |
+
"learning_rate": 0.0001,
|
| 1118 |
+
"loss": 1.2494,
|
| 1119 |
+
"step": 925
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 19.48,
|
| 1123 |
+
"learning_rate": 0.0001,
|
| 1124 |
+
"loss": 1.3238,
|
| 1125 |
+
"step": 930
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 19.58,
|
| 1129 |
+
"learning_rate": 0.0001,
|
| 1130 |
+
"loss": 1.258,
|
| 1131 |
+
"step": 935
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 19.69,
|
| 1135 |
+
"learning_rate": 0.0001,
|
| 1136 |
+
"loss": 1.2089,
|
| 1137 |
+
"step": 940
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 19.79,
|
| 1141 |
+
"learning_rate": 0.0001,
|
| 1142 |
+
"loss": 1.2216,
|
| 1143 |
+
"step": 945
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 19.9,
|
| 1147 |
+
"learning_rate": 0.0001,
|
| 1148 |
+
"loss": 1.2264,
|
| 1149 |
+
"step": 950
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 20.0,
|
| 1153 |
+
"learning_rate": 0.0001,
|
| 1154 |
+
"loss": 1.2316,
|
| 1155 |
+
"step": 955
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 20.1,
|
| 1159 |
+
"learning_rate": 0.0001,
|
| 1160 |
+
"loss": 1.2038,
|
| 1161 |
+
"step": 960
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 20.21,
|
| 1165 |
+
"learning_rate": 0.0001,
|
| 1166 |
+
"loss": 1.2001,
|
| 1167 |
+
"step": 965
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 20.31,
|
| 1171 |
+
"learning_rate": 0.0001,
|
| 1172 |
+
"loss": 1.2475,
|
| 1173 |
+
"step": 970
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 20.42,
|
| 1177 |
+
"learning_rate": 0.0001,
|
| 1178 |
+
"loss": 1.2513,
|
| 1179 |
+
"step": 975
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 20.52,
|
| 1183 |
+
"learning_rate": 0.0001,
|
| 1184 |
+
"loss": 1.2635,
|
| 1185 |
+
"step": 980
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 20.63,
|
| 1189 |
+
"learning_rate": 0.0001,
|
| 1190 |
+
"loss": 1.1817,
|
| 1191 |
+
"step": 985
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 20.73,
|
| 1195 |
+
"learning_rate": 0.0001,
|
| 1196 |
+
"loss": 1.2474,
|
| 1197 |
+
"step": 990
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 20.84,
|
| 1201 |
+
"learning_rate": 0.0001,
|
| 1202 |
+
"loss": 1.2309,
|
| 1203 |
+
"step": 995
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 20.94,
|
| 1207 |
+
"learning_rate": 0.0001,
|
| 1208 |
+
"loss": 1.1871,
|
| 1209 |
+
"step": 1000
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 21.05,
|
| 1213 |
+
"learning_rate": 0.0001,
|
| 1214 |
+
"loss": 1.2312,
|
| 1215 |
+
"step": 1005
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 21.15,
|
| 1219 |
+
"learning_rate": 0.0001,
|
| 1220 |
+
"loss": 1.2152,
|
| 1221 |
+
"step": 1010
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 21.26,
|
| 1225 |
+
"learning_rate": 0.0001,
|
| 1226 |
+
"loss": 1.1812,
|
| 1227 |
+
"step": 1015
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 21.36,
|
| 1231 |
+
"learning_rate": 0.0001,
|
| 1232 |
+
"loss": 1.2133,
|
| 1233 |
+
"step": 1020
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 21.47,
|
| 1237 |
+
"learning_rate": 0.0001,
|
| 1238 |
+
"loss": 1.1851,
|
| 1239 |
+
"step": 1025
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 21.57,
|
| 1243 |
+
"learning_rate": 0.0001,
|
| 1244 |
+
"loss": 1.1767,
|
| 1245 |
+
"step": 1030
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 21.68,
|
| 1249 |
+
"learning_rate": 0.0001,
|
| 1250 |
+
"loss": 1.242,
|
| 1251 |
+
"step": 1035
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 21.78,
|
| 1255 |
+
"learning_rate": 0.0001,
|
| 1256 |
+
"loss": 1.2135,
|
| 1257 |
+
"step": 1040
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 21.88,
|
| 1261 |
+
"learning_rate": 0.0001,
|
| 1262 |
+
"loss": 1.1895,
|
| 1263 |
+
"step": 1045
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 21.99,
|
| 1267 |
+
"learning_rate": 0.0001,
|
| 1268 |
+
"loss": 1.2546,
|
| 1269 |
+
"step": 1050
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 22.09,
|
| 1273 |
+
"learning_rate": 0.0001,
|
| 1274 |
+
"loss": 1.198,
|
| 1275 |
+
"step": 1055
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 22.2,
|
| 1279 |
+
"learning_rate": 0.0001,
|
| 1280 |
+
"loss": 1.1872,
|
| 1281 |
+
"step": 1060
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 22.3,
|
| 1285 |
+
"learning_rate": 0.0001,
|
| 1286 |
+
"loss": 1.1824,
|
| 1287 |
+
"step": 1065
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 22.41,
|
| 1291 |
+
"learning_rate": 0.0001,
|
| 1292 |
+
"loss": 1.1962,
|
| 1293 |
+
"step": 1070
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 22.51,
|
| 1297 |
+
"learning_rate": 0.0001,
|
| 1298 |
+
"loss": 1.1711,
|
| 1299 |
+
"step": 1075
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 22.62,
|
| 1303 |
+
"learning_rate": 0.0001,
|
| 1304 |
+
"loss": 1.1717,
|
| 1305 |
+
"step": 1080
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 22.72,
|
| 1309 |
+
"learning_rate": 0.0001,
|
| 1310 |
+
"loss": 1.1856,
|
| 1311 |
+
"step": 1085
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 22.83,
|
| 1315 |
+
"learning_rate": 0.0001,
|
| 1316 |
+
"loss": 1.2152,
|
| 1317 |
+
"step": 1090
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 22.93,
|
| 1321 |
+
"learning_rate": 0.0001,
|
| 1322 |
+
"loss": 1.2333,
|
| 1323 |
+
"step": 1095
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 23.04,
|
| 1327 |
+
"learning_rate": 0.0001,
|
| 1328 |
+
"loss": 1.1765,
|
| 1329 |
+
"step": 1100
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 23.14,
|
| 1333 |
+
"learning_rate": 0.0001,
|
| 1334 |
+
"loss": 1.1416,
|
| 1335 |
+
"step": 1105
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 23.25,
|
| 1339 |
+
"learning_rate": 0.0001,
|
| 1340 |
+
"loss": 1.1475,
|
| 1341 |
+
"step": 1110
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 23.35,
|
| 1345 |
+
"learning_rate": 0.0001,
|
| 1346 |
+
"loss": 1.1745,
|
| 1347 |
+
"step": 1115
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 23.46,
|
| 1351 |
+
"learning_rate": 0.0001,
|
| 1352 |
+
"loss": 1.1599,
|
| 1353 |
+
"step": 1120
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 23.56,
|
| 1357 |
+
"learning_rate": 0.0001,
|
| 1358 |
+
"loss": 1.1557,
|
| 1359 |
+
"step": 1125
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 23.66,
|
| 1363 |
+
"learning_rate": 0.0001,
|
| 1364 |
+
"loss": 1.21,
|
| 1365 |
+
"step": 1130
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 23.77,
|
| 1369 |
+
"learning_rate": 0.0001,
|
| 1370 |
+
"loss": 1.196,
|
| 1371 |
+
"step": 1135
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 23.87,
|
| 1375 |
+
"learning_rate": 0.0001,
|
| 1376 |
+
"loss": 1.1653,
|
| 1377 |
+
"step": 1140
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 23.98,
|
| 1381 |
+
"learning_rate": 0.0001,
|
| 1382 |
+
"loss": 1.2687,
|
| 1383 |
+
"step": 1145
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 24.08,
|
| 1387 |
+
"learning_rate": 0.0001,
|
| 1388 |
+
"loss": 1.1043,
|
| 1389 |
+
"step": 1150
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 24.19,
|
| 1393 |
+
"learning_rate": 0.0001,
|
| 1394 |
+
"loss": 1.1802,
|
| 1395 |
+
"step": 1155
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 24.29,
|
| 1399 |
+
"learning_rate": 0.0001,
|
| 1400 |
+
"loss": 1.158,
|
| 1401 |
+
"step": 1160
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 24.4,
|
| 1405 |
+
"learning_rate": 0.0001,
|
| 1406 |
+
"loss": 1.1153,
|
| 1407 |
+
"step": 1165
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 24.5,
|
| 1411 |
+
"learning_rate": 0.0001,
|
| 1412 |
+
"loss": 1.1722,
|
| 1413 |
+
"step": 1170
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 24.61,
|
| 1417 |
+
"learning_rate": 0.0001,
|
| 1418 |
+
"loss": 1.2067,
|
| 1419 |
+
"step": 1175
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 24.61,
|
| 1423 |
+
"step": 1175,
|
| 1424 |
+
"total_flos": 4391121715200000.0,
|
| 1425 |
+
"train_loss": 1.3988467488390335,
|
| 1426 |
+
"train_runtime": 1205.3351,
|
| 1427 |
+
"train_samples_per_second": 31.692,
|
| 1428 |
+
"train_steps_per_second": 0.975
|
| 1429 |
}
|
| 1430 |
],
|
| 1431 |
"logging_steps": 5,
|
| 1432 |
+
"max_steps": 1175,
|
| 1433 |
+
"num_train_epochs": 25,
|
| 1434 |
"save_steps": 1000,
|
| 1435 |
+
"total_flos": 4391121715200000.0,
|
| 1436 |
"trial_name": null,
|
| 1437 |
"trial_params": null
|
| 1438 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4027
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99eae4f19c67acd3f95af7fb6d5de5d000413bb16589081d407c5b736dcd91ad
|
| 3 |
size 4027
|