Model save
Browse files- README.md +14 -14
- all_results.json +4 -4
- train_results.json +4 -4
- trainer_state.json +695 -695
README.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
license: apache-2.0
|
| 4 |
-
base_model: mistralai/Mistral-7B-v0.
|
| 5 |
tags:
|
| 6 |
- trl
|
| 7 |
- sft
|
|
@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# zephyr-7b-sft-full
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
@@ -39,7 +39,7 @@ More information needed
|
|
| 39 |
### Training hyperparameters
|
| 40 |
|
| 41 |
The following hyperparameters were used during training:
|
| 42 |
-
- learning_rate:
|
| 43 |
- train_batch_size: 8
|
| 44 |
- eval_batch_size: 16
|
| 45 |
- seed: 42
|
|
@@ -57,16 +57,16 @@ The following hyperparameters were used during training:
|
|
| 57 |
|
| 58 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
|:-------------:|:------:|:----:|:---------------:|
|
| 60 |
-
|
|
| 61 |
-
|
|
| 62 |
-
|
|
| 63 |
-
|
|
| 64 |
-
|
|
| 65 |
-
| 0.
|
| 66 |
-
| 0.
|
| 67 |
-
| 0.
|
| 68 |
-
| 0.
|
| 69 |
-
| 0.
|
| 70 |
|
| 71 |
|
| 72 |
### Framework versions
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
license: apache-2.0
|
| 4 |
+
base_model: mistralai/Mistral-7B-v0.1
|
| 5 |
tags:
|
| 6 |
- trl
|
| 7 |
- sft
|
|
|
|
| 18 |
|
| 19 |
# zephyr-7b-sft-full
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the generator dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.9293
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
|
|
| 39 |
### Training hyperparameters
|
| 40 |
|
| 41 |
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 1e-05
|
| 43 |
- train_batch_size: 8
|
| 44 |
- eval_batch_size: 16
|
| 45 |
- seed: 42
|
|
|
|
| 57 |
|
| 58 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
|:-------------:|:------:|:----:|:---------------:|
|
| 60 |
+
| 0.9896 | 0.0923 | 100 | 0.9893 |
|
| 61 |
+
| 0.9838 | 0.1846 | 200 | 0.9935 |
|
| 62 |
+
| 0.9853 | 0.2769 | 300 | 0.9881 |
|
| 63 |
+
| 0.9638 | 0.3692 | 400 | 0.9781 |
|
| 64 |
+
| 0.9745 | 0.4615 | 500 | 0.9680 |
|
| 65 |
+
| 0.9396 | 0.5538 | 600 | 0.9568 |
|
| 66 |
+
| 0.9176 | 0.6461 | 700 | 0.9465 |
|
| 67 |
+
| 0.9067 | 0.7383 | 800 | 0.9379 |
|
| 68 |
+
| 0.9221 | 0.8306 | 900 | 0.9320 |
|
| 69 |
+
| 0.9087 | 0.9229 | 1000 | 0.9293 |
|
| 70 |
|
| 71 |
|
| 72 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 207864,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
+
"train_loss": 0.9547446678880179,
|
| 5 |
+
"train_runtime": 38927.5133,
|
| 6 |
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 3.563,
|
| 8 |
+
"train_steps_per_second": 0.028
|
| 9 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 207864,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
+
"train_loss": 0.9547446678880179,
|
| 5 |
+
"train_runtime": 38927.5133,
|
| 6 |
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 3.563,
|
| 8 |
+
"train_steps_per_second": 0.028
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -10,1611 +10,1611 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0009229349330872173,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.0046146746654360865,
|
| 20 |
-
"grad_norm": 8.
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.009229349330872173,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01384402399630826,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.018458698661744346,
|
| 41 |
-
"grad_norm": 2.
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.023073373327180433,
|
| 48 |
-
"grad_norm": 2.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.02768804799261652,
|
| 55 |
-
"grad_norm": 2.
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.032302722658052604,
|
| 62 |
-
"grad_norm": 2.
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss":
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03691739732348869,
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.04153207198892478,
|
| 76 |
-
"grad_norm": 2.
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.046146746654360866,
|
| 83 |
-
"grad_norm": 2.
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.050761421319796954,
|
| 90 |
-
"grad_norm": 2.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05537609598523304,
|
| 97 |
-
"grad_norm": 2.
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss":
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05999077065066913,
|
| 104 |
-
"grad_norm": 2.
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss":
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06460544531610521,
|
| 111 |
-
"grad_norm": 2.
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0692201199815413,
|
| 118 |
-
"grad_norm": 2.
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07383479464697738,
|
| 125 |
-
"grad_norm": 2.
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07844946931241348,
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss":
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.08306414397784956,
|
| 139 |
-
"grad_norm":
|
| 140 |
-
"learning_rate":
|
| 141 |
-
"loss":
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08767881864328565,
|
| 146 |
-
"grad_norm": 2.
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss":
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.09229349330872173,
|
| 153 |
-
"grad_norm": 1.
|
| 154 |
-
"learning_rate":
|
| 155 |
-
"loss":
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.09229349330872173,
|
| 160 |
-
"eval_loss":
|
| 161 |
-
"eval_runtime":
|
| 162 |
-
"eval_samples_per_second":
|
| 163 |
-
"eval_steps_per_second": 0.
|
| 164 |
"step": 100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.09690816797415783,
|
| 168 |
-
"grad_norm": 2.
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 105
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.10152284263959391,
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss":
|
| 178 |
"step": 110
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.10613751730503,
|
| 182 |
-
"grad_norm": 2.
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss": 1.
|
| 185 |
"step": 115
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.11075219197046608,
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss":
|
| 192 |
"step": 120
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.11536686663590216,
|
| 196 |
-
"grad_norm":
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 1.
|
| 199 |
"step": 125
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.11998154130133826,
|
| 203 |
-
"grad_norm": 1.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss":
|
| 206 |
"step": 130
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.12459621596677434,
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 135
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.12921089063221042,
|
| 217 |
-
"grad_norm": 1.
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 1.
|
| 220 |
"step": 140
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.13382556529764653,
|
| 224 |
-
"grad_norm": 1.
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss":
|
| 227 |
"step": 145
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.1384402399630826,
|
| 231 |
-
"grad_norm":
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss":
|
| 234 |
"step": 150
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.1430549146285187,
|
| 238 |
-
"grad_norm": 2.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 1.
|
| 241 |
"step": 155
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.14766958929395477,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss":
|
| 248 |
"step": 160
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.15228426395939088,
|
| 252 |
-
"grad_norm": 1.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss":
|
| 255 |
"step": 165
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.15689893862482696,
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss":
|
| 262 |
"step": 170
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.16151361329026304,
|
| 266 |
-
"grad_norm": 1.
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss":
|
| 269 |
"step": 175
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.16612828795569912,
|
| 273 |
-
"grad_norm": 1.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss":
|
| 276 |
"step": 180
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.1707429626211352,
|
| 280 |
-
"grad_norm": 1.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss":
|
| 283 |
"step": 185
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.1753576372865713,
|
| 287 |
-
"grad_norm": 1.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss":
|
| 290 |
"step": 190
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.17997231195200739,
|
| 294 |
-
"grad_norm": 1.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss":
|
| 297 |
"step": 195
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.18458698661744347,
|
| 301 |
-
"grad_norm": 1.
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss":
|
| 304 |
"step": 200
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.18458698661744347,
|
| 308 |
-
"eval_loss":
|
| 309 |
-
"eval_runtime":
|
| 310 |
-
"eval_samples_per_second":
|
| 311 |
-
"eval_steps_per_second": 0.
|
| 312 |
"step": 200
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 0.18920166128287955,
|
| 316 |
-
"grad_norm": 1.
|
| 317 |
-
"learning_rate":
|
| 318 |
-
"loss":
|
| 319 |
"step": 205
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 0.19381633594831565,
|
| 323 |
-
"grad_norm": 1.
|
| 324 |
-
"learning_rate":
|
| 325 |
-
"loss":
|
| 326 |
"step": 210
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.19843101061375173,
|
| 330 |
-
"grad_norm":
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss":
|
| 333 |
"step": 215
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 0.20304568527918782,
|
| 337 |
-
"grad_norm": 1.
|
| 338 |
-
"learning_rate":
|
| 339 |
-
"loss":
|
| 340 |
"step": 220
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 0.2076603599446239,
|
| 344 |
-
"grad_norm": 1.
|
| 345 |
-
"learning_rate":
|
| 346 |
-
"loss":
|
| 347 |
"step": 225
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 0.21227503461006,
|
| 351 |
-
"grad_norm": 1.
|
| 352 |
-
"learning_rate":
|
| 353 |
-
"loss":
|
| 354 |
"step": 230
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.21688970927549608,
|
| 358 |
-
"grad_norm": 1.
|
| 359 |
-
"learning_rate":
|
| 360 |
-
"loss":
|
| 361 |
"step": 235
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.22150438394093216,
|
| 365 |
-
"grad_norm": 1.
|
| 366 |
-
"learning_rate":
|
| 367 |
-
"loss":
|
| 368 |
"step": 240
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.22611905860636825,
|
| 372 |
-
"grad_norm": 1.
|
| 373 |
-
"learning_rate":
|
| 374 |
-
"loss":
|
| 375 |
"step": 245
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 0.23073373327180433,
|
| 379 |
-
"grad_norm": 1.
|
| 380 |
-
"learning_rate":
|
| 381 |
-
"loss":
|
| 382 |
"step": 250
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 0.23534840793724043,
|
| 386 |
-
"grad_norm": 1.
|
| 387 |
-
"learning_rate":
|
| 388 |
-
"loss":
|
| 389 |
"step": 255
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 0.23996308260267651,
|
| 393 |
-
"grad_norm": 1.
|
| 394 |
-
"learning_rate":
|
| 395 |
-
"loss":
|
| 396 |
"step": 260
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.2445777572681126,
|
| 400 |
-
"grad_norm": 1.
|
| 401 |
-
"learning_rate":
|
| 402 |
-
"loss":
|
| 403 |
"step": 265
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 0.24919243193354867,
|
| 407 |
-
"grad_norm": 1.
|
| 408 |
-
"learning_rate":
|
| 409 |
-
"loss":
|
| 410 |
"step": 270
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.25380710659898476,
|
| 414 |
-
"grad_norm": 1.
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss":
|
| 417 |
"step": 275
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.25842178126442084,
|
| 421 |
-
"grad_norm":
|
| 422 |
-
"learning_rate":
|
| 423 |
-
"loss":
|
| 424 |
"step": 280
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 0.26303645592985697,
|
| 428 |
-
"grad_norm": 1.
|
| 429 |
-
"learning_rate":
|
| 430 |
-
"loss":
|
| 431 |
"step": 285
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 0.26765113059529305,
|
| 435 |
-
"grad_norm": 1.
|
| 436 |
-
"learning_rate":
|
| 437 |
-
"loss":
|
| 438 |
"step": 290
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 0.27226580526072913,
|
| 442 |
-
"grad_norm": 1.
|
| 443 |
-
"learning_rate":
|
| 444 |
-
"loss":
|
| 445 |
"step": 295
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 0.2768804799261652,
|
| 449 |
-
"grad_norm": 1.
|
| 450 |
-
"learning_rate":
|
| 451 |
-
"loss":
|
| 452 |
"step": 300
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.2768804799261652,
|
| 456 |
-
"eval_loss":
|
| 457 |
-
"eval_runtime":
|
| 458 |
-
"eval_samples_per_second": 23.
|
| 459 |
-
"eval_steps_per_second": 0.
|
| 460 |
"step": 300
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 0.2814951545916013,
|
| 464 |
-
"grad_norm": 1.
|
| 465 |
-
"learning_rate":
|
| 466 |
-
"loss":
|
| 467 |
"step": 305
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 0.2861098292570374,
|
| 471 |
-
"grad_norm": 1.
|
| 472 |
-
"learning_rate":
|
| 473 |
-
"loss":
|
| 474 |
"step": 310
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.29072450392247345,
|
| 478 |
-
"grad_norm": 1.
|
| 479 |
-
"learning_rate":
|
| 480 |
-
"loss":
|
| 481 |
"step": 315
|
| 482 |
},
|
| 483 |
{
|
| 484 |
"epoch": 0.29533917858790953,
|
| 485 |
-
"grad_norm": 1.
|
| 486 |
-
"learning_rate":
|
| 487 |
-
"loss":
|
| 488 |
"step": 320
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.2999538532533456,
|
| 492 |
-
"grad_norm": 1.
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss":
|
| 495 |
"step": 325
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 0.30456852791878175,
|
| 499 |
-
"grad_norm": 1.
|
| 500 |
-
"learning_rate":
|
| 501 |
-
"loss":
|
| 502 |
"step": 330
|
| 503 |
},
|
| 504 |
{
|
| 505 |
"epoch": 0.30918320258421783,
|
| 506 |
-
"grad_norm": 1.
|
| 507 |
-
"learning_rate":
|
| 508 |
-
"loss":
|
| 509 |
"step": 335
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 0.3137978772496539,
|
| 513 |
-
"grad_norm": 1.
|
| 514 |
-
"learning_rate":
|
| 515 |
-
"loss":
|
| 516 |
"step": 340
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 0.31841255191509,
|
| 520 |
-
"grad_norm": 1.
|
| 521 |
-
"learning_rate":
|
| 522 |
-
"loss":
|
| 523 |
"step": 345
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 0.3230272265805261,
|
| 527 |
-
"grad_norm": 1.
|
| 528 |
-
"learning_rate":
|
| 529 |
-
"loss":
|
| 530 |
"step": 350
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.32764190124596215,
|
| 534 |
-
"grad_norm": 1.
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss":
|
| 537 |
"step": 355
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.33225657591139823,
|
| 541 |
-
"grad_norm":
|
| 542 |
-
"learning_rate":
|
| 543 |
-
"loss":
|
| 544 |
"step": 360
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 0.3368712505768343,
|
| 548 |
-
"grad_norm": 1.
|
| 549 |
-
"learning_rate":
|
| 550 |
-
"loss":
|
| 551 |
"step": 365
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 0.3414859252422704,
|
| 555 |
-
"grad_norm":
|
| 556 |
-
"learning_rate":
|
| 557 |
-
"loss":
|
| 558 |
"step": 370
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 0.34610059990770653,
|
| 562 |
-
"grad_norm": 1.
|
| 563 |
-
"learning_rate":
|
| 564 |
-
"loss":
|
| 565 |
"step": 375
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 0.3507152745731426,
|
| 569 |
-
"grad_norm": 1.
|
| 570 |
-
"learning_rate":
|
| 571 |
-
"loss":
|
| 572 |
"step": 380
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.3553299492385787,
|
| 576 |
-
"grad_norm": 1.
|
| 577 |
-
"learning_rate":
|
| 578 |
-
"loss":
|
| 579 |
"step": 385
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.35994462390401477,
|
| 583 |
-
"grad_norm": 1.
|
| 584 |
-
"learning_rate":
|
| 585 |
-
"loss":
|
| 586 |
"step": 390
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 0.36455929856945085,
|
| 590 |
-
"grad_norm": 1.
|
| 591 |
-
"learning_rate":
|
| 592 |
-
"loss":
|
| 593 |
"step": 395
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 0.36917397323488693,
|
| 597 |
-
"grad_norm": 1.
|
| 598 |
-
"learning_rate":
|
| 599 |
-
"loss":
|
| 600 |
"step": 400
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 0.36917397323488693,
|
| 604 |
-
"eval_loss":
|
| 605 |
-
"eval_runtime":
|
| 606 |
-
"eval_samples_per_second":
|
| 607 |
-
"eval_steps_per_second": 0.
|
| 608 |
"step": 400
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 0.373788647900323,
|
| 612 |
-
"grad_norm": 1.
|
| 613 |
-
"learning_rate":
|
| 614 |
-
"loss":
|
| 615 |
"step": 405
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 0.3784033225657591,
|
| 619 |
-
"grad_norm": 1.
|
| 620 |
-
"learning_rate":
|
| 621 |
-
"loss":
|
| 622 |
"step": 410
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 0.3830179972311952,
|
| 626 |
-
"grad_norm": 1.
|
| 627 |
-
"learning_rate":
|
| 628 |
-
"loss": 0.
|
| 629 |
"step": 415
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 0.3876326718966313,
|
| 633 |
-
"grad_norm": 1.
|
| 634 |
-
"learning_rate":
|
| 635 |
-
"loss":
|
| 636 |
"step": 420
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 0.3922473465620674,
|
| 640 |
-
"grad_norm": 1.
|
| 641 |
-
"learning_rate":
|
| 642 |
-
"loss":
|
| 643 |
"step": 425
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 0.39686202122750347,
|
| 647 |
-
"grad_norm": 1.
|
| 648 |
-
"learning_rate":
|
| 649 |
-
"loss":
|
| 650 |
"step": 430
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.40147669589293955,
|
| 654 |
-
"grad_norm": 1.
|
| 655 |
-
"learning_rate":
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 435
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 0.40609137055837563,
|
| 661 |
-
"grad_norm": 1.
|
| 662 |
-
"learning_rate":
|
| 663 |
-
"loss":
|
| 664 |
"step": 440
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 0.4107060452238117,
|
| 668 |
-
"grad_norm": 1.
|
| 669 |
-
"learning_rate":
|
| 670 |
-
"loss":
|
| 671 |
"step": 445
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 0.4153207198892478,
|
| 675 |
-
"grad_norm": 1.
|
| 676 |
-
"learning_rate":
|
| 677 |
-
"loss":
|
| 678 |
"step": 450
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 0.41993539455468387,
|
| 682 |
-
"grad_norm": 1.
|
| 683 |
-
"learning_rate":
|
| 684 |
-
"loss": 0.
|
| 685 |
"step": 455
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 0.42455006922012,
|
| 689 |
-
"grad_norm": 1.
|
| 690 |
-
"learning_rate":
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 460
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 0.4291647438855561,
|
| 696 |
-
"grad_norm": 1.
|
| 697 |
-
"learning_rate":
|
| 698 |
-
"loss":
|
| 699 |
"step": 465
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.43377941855099217,
|
| 703 |
-
"grad_norm": 1.
|
| 704 |
-
"learning_rate":
|
| 705 |
-
"loss":
|
| 706 |
"step": 470
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 0.43839409321642825,
|
| 710 |
-
"grad_norm": 1.
|
| 711 |
-
"learning_rate":
|
| 712 |
-
"loss":
|
| 713 |
"step": 475
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.44300876788186433,
|
| 717 |
-
"grad_norm": 1.
|
| 718 |
-
"learning_rate":
|
| 719 |
-
"loss":
|
| 720 |
"step": 480
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 0.4476234425473004,
|
| 724 |
-
"grad_norm": 1.
|
| 725 |
-
"learning_rate":
|
| 726 |
-
"loss":
|
| 727 |
"step": 485
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 0.4522381172127365,
|
| 731 |
-
"grad_norm": 1.
|
| 732 |
-
"learning_rate":
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 490
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.45685279187817257,
|
| 738 |
-
"grad_norm": 1.
|
| 739 |
-
"learning_rate":
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 495
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 0.46146746654360865,
|
| 745 |
-
"grad_norm": 1.
|
| 746 |
-
"learning_rate":
|
| 747 |
-
"loss":
|
| 748 |
"step": 500
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 0.46146746654360865,
|
| 752 |
-
"eval_loss":
|
| 753 |
-
"eval_runtime":
|
| 754 |
-
"eval_samples_per_second": 24.
|
| 755 |
-
"eval_steps_per_second": 0.
|
| 756 |
"step": 500
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"epoch": 0.4660821412090448,
|
| 760 |
-
"grad_norm": 1.
|
| 761 |
-
"learning_rate":
|
| 762 |
-
"loss": 0.
|
| 763 |
"step": 505
|
| 764 |
},
|
| 765 |
{
|
| 766 |
"epoch": 0.47069681587448087,
|
| 767 |
-
"grad_norm": 1.
|
| 768 |
-
"learning_rate":
|
| 769 |
-
"loss":
|
| 770 |
"step": 510
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.47531149053991695,
|
| 774 |
-
"grad_norm": 1.
|
| 775 |
-
"learning_rate":
|
| 776 |
-
"loss":
|
| 777 |
"step": 515
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.47992616520535303,
|
| 781 |
-
"grad_norm": 1.
|
| 782 |
-
"learning_rate":
|
| 783 |
-
"loss":
|
| 784 |
"step": 520
|
| 785 |
},
|
| 786 |
{
|
| 787 |
"epoch": 0.4845408398707891,
|
| 788 |
-
"grad_norm": 1.
|
| 789 |
-
"learning_rate":
|
| 790 |
-
"loss": 0.
|
| 791 |
"step": 525
|
| 792 |
},
|
| 793 |
{
|
| 794 |
"epoch": 0.4891555145362252,
|
| 795 |
-
"grad_norm": 1.
|
| 796 |
-
"learning_rate":
|
| 797 |
-
"loss": 0.
|
| 798 |
"step": 530
|
| 799 |
},
|
| 800 |
{
|
| 801 |
"epoch": 0.49377018920166127,
|
| 802 |
-
"grad_norm": 1.
|
| 803 |
-
"learning_rate":
|
| 804 |
-
"loss": 0.
|
| 805 |
"step": 535
|
| 806 |
},
|
| 807 |
{
|
| 808 |
"epoch": 0.49838486386709735,
|
| 809 |
-
"grad_norm": 1.
|
| 810 |
-
"learning_rate":
|
| 811 |
-
"loss": 0.
|
| 812 |
"step": 540
|
| 813 |
},
|
| 814 |
{
|
| 815 |
"epoch": 0.5029995385325334,
|
| 816 |
-
"grad_norm": 1.
|
| 817 |
-
"learning_rate":
|
| 818 |
-
"loss": 0.
|
| 819 |
"step": 545
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.5076142131979695,
|
| 823 |
-
"grad_norm": 1.
|
| 824 |
-
"learning_rate":
|
| 825 |
-
"loss": 0.
|
| 826 |
"step": 550
|
| 827 |
},
|
| 828 |
{
|
| 829 |
"epoch": 0.5122288878634056,
|
| 830 |
-
"grad_norm": 1.
|
| 831 |
-
"learning_rate":
|
| 832 |
-
"loss": 0.
|
| 833 |
"step": 555
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 0.5168435625288417,
|
| 837 |
-
"grad_norm": 1.
|
| 838 |
-
"learning_rate":
|
| 839 |
-
"loss": 0.
|
| 840 |
"step": 560
|
| 841 |
},
|
| 842 |
{
|
| 843 |
"epoch": 0.5214582371942778,
|
| 844 |
-
"grad_norm": 1.
|
| 845 |
-
"learning_rate":
|
| 846 |
-
"loss": 0.
|
| 847 |
"step": 565
|
| 848 |
},
|
| 849 |
{
|
| 850 |
"epoch": 0.5260729118597139,
|
| 851 |
-
"grad_norm": 1.
|
| 852 |
-
"learning_rate":
|
| 853 |
-
"loss": 0.
|
| 854 |
"step": 570
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.53068758652515,
|
| 858 |
-
"grad_norm": 1.
|
| 859 |
-
"learning_rate":
|
| 860 |
-
"loss": 0.
|
| 861 |
"step": 575
|
| 862 |
},
|
| 863 |
{
|
| 864 |
"epoch": 0.5353022611905861,
|
| 865 |
-
"grad_norm": 1.
|
| 866 |
-
"learning_rate":
|
| 867 |
-
"loss": 0.
|
| 868 |
"step": 580
|
| 869 |
},
|
| 870 |
{
|
| 871 |
"epoch": 0.5399169358560222,
|
| 872 |
-
"grad_norm": 1.
|
| 873 |
-
"learning_rate":
|
| 874 |
-
"loss": 0.
|
| 875 |
"step": 585
|
| 876 |
},
|
| 877 |
{
|
| 878 |
"epoch": 0.5445316105214583,
|
| 879 |
-
"grad_norm": 1.
|
| 880 |
-
"learning_rate":
|
| 881 |
-
"loss": 0.
|
| 882 |
"step": 590
|
| 883 |
},
|
| 884 |
{
|
| 885 |
"epoch": 0.5491462851868943,
|
| 886 |
-
"grad_norm": 1.
|
| 887 |
-
"learning_rate":
|
| 888 |
-
"loss": 0.
|
| 889 |
"step": 595
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.5537609598523304,
|
| 893 |
-
"grad_norm": 1.
|
| 894 |
-
"learning_rate":
|
| 895 |
-
"loss": 0.
|
| 896 |
"step": 600
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.5537609598523304,
|
| 900 |
-
"eval_loss": 0.
|
| 901 |
-
"eval_runtime":
|
| 902 |
-
"eval_samples_per_second":
|
| 903 |
-
"eval_steps_per_second": 0.
|
| 904 |
"step": 600
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 0.5583756345177665,
|
| 908 |
-
"grad_norm": 1.
|
| 909 |
-
"learning_rate":
|
| 910 |
-
"loss": 0.
|
| 911 |
"step": 605
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 0.5629903091832026,
|
| 915 |
-
"grad_norm": 1.
|
| 916 |
-
"learning_rate":
|
| 917 |
-
"loss": 0.
|
| 918 |
"step": 610
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 0.5676049838486387,
|
| 922 |
-
"grad_norm": 1.
|
| 923 |
-
"learning_rate":
|
| 924 |
-
"loss": 0.
|
| 925 |
"step": 615
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 0.5722196585140747,
|
| 929 |
-
"grad_norm": 1.
|
| 930 |
-
"learning_rate":
|
| 931 |
-
"loss": 0.
|
| 932 |
"step": 620
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 0.5768343331795108,
|
| 936 |
-
"grad_norm": 1.
|
| 937 |
-
"learning_rate":
|
| 938 |
-
"loss": 0.
|
| 939 |
"step": 625
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 0.5814490078449469,
|
| 943 |
-
"grad_norm": 1.
|
| 944 |
-
"learning_rate":
|
| 945 |
-
"loss": 0.
|
| 946 |
"step": 630
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 0.586063682510383,
|
| 950 |
-
"grad_norm": 1.
|
| 951 |
-
"learning_rate":
|
| 952 |
-
"loss": 0.
|
| 953 |
"step": 635
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.5906783571758191,
|
| 957 |
-
"grad_norm": 1.
|
| 958 |
-
"learning_rate":
|
| 959 |
-
"loss": 0.
|
| 960 |
"step": 640
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 0.5952930318412551,
|
| 964 |
-
"grad_norm": 1.
|
| 965 |
-
"learning_rate":
|
| 966 |
-
"loss": 0.
|
| 967 |
"step": 645
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 0.5999077065066912,
|
| 971 |
-
"grad_norm": 1.
|
| 972 |
-
"learning_rate":
|
| 973 |
-
"loss": 0.
|
| 974 |
"step": 650
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 0.6045223811721273,
|
| 978 |
-
"grad_norm": 1.
|
| 979 |
-
"learning_rate":
|
| 980 |
-
"loss": 0.
|
| 981 |
"step": 655
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 0.6091370558375635,
|
| 985 |
-
"grad_norm": 1.
|
| 986 |
-
"learning_rate":
|
| 987 |
-
"loss": 0.
|
| 988 |
"step": 660
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 0.6137517305029996,
|
| 992 |
-
"grad_norm": 1.
|
| 993 |
-
"learning_rate":
|
| 994 |
-
"loss": 0.
|
| 995 |
"step": 665
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 0.6183664051684357,
|
| 999 |
-
"grad_norm": 1.
|
| 1000 |
-
"learning_rate":
|
| 1001 |
-
"loss": 0.
|
| 1002 |
"step": 670
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 0.6229810798338717,
|
| 1006 |
-
"grad_norm": 1.
|
| 1007 |
-
"learning_rate":
|
| 1008 |
-
"loss": 0.
|
| 1009 |
"step": 675
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 0.6275957544993078,
|
| 1013 |
-
"grad_norm": 1.
|
| 1014 |
-
"learning_rate":
|
| 1015 |
-
"loss": 0.
|
| 1016 |
"step": 680
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 0.6322104291647439,
|
| 1020 |
-
"grad_norm": 1.
|
| 1021 |
-
"learning_rate":
|
| 1022 |
-
"loss": 0.
|
| 1023 |
"step": 685
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 0.63682510383018,
|
| 1027 |
-
"grad_norm": 1.
|
| 1028 |
-
"learning_rate":
|
| 1029 |
-
"loss": 0.
|
| 1030 |
"step": 690
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 0.6414397784956161,
|
| 1034 |
-
"grad_norm": 1.
|
| 1035 |
-
"learning_rate":
|
| 1036 |
-
"loss": 0.
|
| 1037 |
"step": 695
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 0.6460544531610521,
|
| 1041 |
-
"grad_norm": 1.
|
| 1042 |
-
"learning_rate":
|
| 1043 |
-
"loss": 0.
|
| 1044 |
"step": 700
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.6460544531610521,
|
| 1048 |
-
"eval_loss": 0.
|
| 1049 |
-
"eval_runtime":
|
| 1050 |
-
"eval_samples_per_second": 24.
|
| 1051 |
-
"eval_steps_per_second": 0.
|
| 1052 |
"step": 700
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.6506691278264882,
|
| 1056 |
-
"grad_norm": 1.
|
| 1057 |
-
"learning_rate":
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 705
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.6552838024919243,
|
| 1063 |
-
"grad_norm": 1.
|
| 1064 |
-
"learning_rate":
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 710
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.6598984771573604,
|
| 1070 |
-
"grad_norm": 1.
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 715
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.6645131518227965,
|
| 1077 |
-
"grad_norm": 1.
|
| 1078 |
-
"learning_rate":
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 720
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.6691278264882325,
|
| 1084 |
-
"grad_norm": 1.
|
| 1085 |
-
"learning_rate":
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 725
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.6737425011536686,
|
| 1091 |
-
"grad_norm": 1.
|
| 1092 |
-
"learning_rate":
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 730
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.6783571758191047,
|
| 1098 |
-
"grad_norm": 1.
|
| 1099 |
-
"learning_rate":
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 735
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.6829718504845408,
|
| 1105 |
-
"grad_norm": 1.
|
| 1106 |
-
"learning_rate":
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 740
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.687586525149977,
|
| 1112 |
-
"grad_norm": 1.
|
| 1113 |
-
"learning_rate":
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 745
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.6922011998154131,
|
| 1119 |
-
"grad_norm": 1.
|
| 1120 |
-
"learning_rate":
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 750
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.6968158744808491,
|
| 1126 |
-
"grad_norm": 1.
|
| 1127 |
-
"learning_rate":
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 755
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.7014305491462852,
|
| 1133 |
-
"grad_norm": 1.
|
| 1134 |
-
"learning_rate":
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 760
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.7060452238117213,
|
| 1140 |
-
"grad_norm": 1.
|
| 1141 |
-
"learning_rate":
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 765
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.7106598984771574,
|
| 1147 |
-
"grad_norm": 1.
|
| 1148 |
-
"learning_rate":
|
| 1149 |
-
"loss": 0.
|
| 1150 |
"step": 770
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.7152745731425935,
|
| 1154 |
-
"grad_norm": 1.
|
| 1155 |
-
"learning_rate":
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 775
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.7198892478080295,
|
| 1161 |
-
"grad_norm": 1.
|
| 1162 |
-
"learning_rate":
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 780
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.7245039224734656,
|
| 1168 |
-
"grad_norm": 1.
|
| 1169 |
-
"learning_rate":
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 785
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.7291185971389017,
|
| 1175 |
-
"grad_norm": 1.
|
| 1176 |
-
"learning_rate":
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 790
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.7337332718043378,
|
| 1182 |
-
"grad_norm": 1.
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 795
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.7383479464697739,
|
| 1189 |
-
"grad_norm": 1.
|
| 1190 |
-
"learning_rate":
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 800
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.7383479464697739,
|
| 1196 |
-
"eval_loss": 0.
|
| 1197 |
-
"eval_runtime":
|
| 1198 |
-
"eval_samples_per_second": 23.
|
| 1199 |
-
"eval_steps_per_second": 0.
|
| 1200 |
"step": 800
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.7429626211352099,
|
| 1204 |
-
"grad_norm": 1.
|
| 1205 |
-
"learning_rate":
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 805
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.747577295800646,
|
| 1211 |
-
"grad_norm": 1.
|
| 1212 |
-
"learning_rate":
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 810
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.7521919704660821,
|
| 1218 |
-
"grad_norm": 1.
|
| 1219 |
-
"learning_rate":
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 815
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.7568066451315182,
|
| 1225 |
-
"grad_norm": 1.
|
| 1226 |
-
"learning_rate":
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 820
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.7614213197969543,
|
| 1232 |
-
"grad_norm": 1.
|
| 1233 |
-
"learning_rate":
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 825
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.7660359944623903,
|
| 1239 |
-
"grad_norm": 1.
|
| 1240 |
-
"learning_rate":
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 830
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.7706506691278265,
|
| 1246 |
-
"grad_norm": 1.
|
| 1247 |
-
"learning_rate":
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 835
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.7752653437932626,
|
| 1253 |
-
"grad_norm": 1.
|
| 1254 |
-
"learning_rate":
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 840
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.7798800184586987,
|
| 1260 |
-
"grad_norm": 1.
|
| 1261 |
-
"learning_rate":
|
| 1262 |
-
"loss": 0.
|
| 1263 |
"step": 845
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.7844946931241348,
|
| 1267 |
-
"grad_norm": 1.
|
| 1268 |
-
"learning_rate":
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 850
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.7891093677895709,
|
| 1274 |
-
"grad_norm": 1.
|
| 1275 |
-
"learning_rate":
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 855
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.7937240424550069,
|
| 1281 |
-
"grad_norm": 1.
|
| 1282 |
-
"learning_rate":
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 860
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.798338717120443,
|
| 1288 |
-
"grad_norm": 1.
|
| 1289 |
-
"learning_rate":
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 865
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.8029533917858791,
|
| 1295 |
-
"grad_norm": 1.
|
| 1296 |
-
"learning_rate":
|
| 1297 |
-
"loss": 0.
|
| 1298 |
"step": 870
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.8075680664513152,
|
| 1302 |
-
"grad_norm": 1.
|
| 1303 |
-
"learning_rate":
|
| 1304 |
-
"loss": 0.
|
| 1305 |
"step": 875
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.8121827411167513,
|
| 1309 |
-
"grad_norm": 1.
|
| 1310 |
-
"learning_rate":
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 880
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.8167974157821873,
|
| 1316 |
-
"grad_norm": 1.
|
| 1317 |
-
"learning_rate":
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 885
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.8214120904476234,
|
| 1323 |
-
"grad_norm": 1.
|
| 1324 |
-
"learning_rate":
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 890
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.8260267651130595,
|
| 1330 |
-
"grad_norm": 1.
|
| 1331 |
-
"learning_rate":
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 895
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.8306414397784956,
|
| 1337 |
-
"grad_norm": 1.
|
| 1338 |
-
"learning_rate":
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 900
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.8306414397784956,
|
| 1344 |
-
"eval_loss": 0.
|
| 1345 |
-
"eval_runtime":
|
| 1346 |
-
"eval_samples_per_second": 24.
|
| 1347 |
-
"eval_steps_per_second": 0.
|
| 1348 |
"step": 900
|
| 1349 |
},
|
| 1350 |
{
|
| 1351 |
"epoch": 0.8352561144439317,
|
| 1352 |
-
"grad_norm": 1.
|
| 1353 |
-
"learning_rate":
|
| 1354 |
-
"loss": 0.
|
| 1355 |
"step": 905
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 0.8398707891093677,
|
| 1359 |
-
"grad_norm": 1.
|
| 1360 |
-
"learning_rate":
|
| 1361 |
-
"loss": 0.
|
| 1362 |
"step": 910
|
| 1363 |
},
|
| 1364 |
{
|
| 1365 |
"epoch": 0.8444854637748038,
|
| 1366 |
-
"grad_norm": 1.
|
| 1367 |
-
"learning_rate":
|
| 1368 |
-
"loss": 0.
|
| 1369 |
"step": 915
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.84910013844024,
|
| 1373 |
-
"grad_norm": 1.
|
| 1374 |
-
"learning_rate":
|
| 1375 |
-
"loss": 0.
|
| 1376 |
"step": 920
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
"epoch": 0.8537148131056761,
|
| 1380 |
-
"grad_norm": 1.
|
| 1381 |
-
"learning_rate":
|
| 1382 |
-
"loss": 0.
|
| 1383 |
"step": 925
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 0.8583294877711122,
|
| 1387 |
-
"grad_norm": 1.
|
| 1388 |
-
"learning_rate":
|
| 1389 |
-
"loss": 0.
|
| 1390 |
"step": 930
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 0.8629441624365483,
|
| 1394 |
-
"grad_norm": 1.
|
| 1395 |
-
"learning_rate":
|
| 1396 |
-
"loss": 0.
|
| 1397 |
"step": 935
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 0.8675588371019843,
|
| 1401 |
-
"grad_norm": 1.
|
| 1402 |
-
"learning_rate":
|
| 1403 |
-
"loss": 0.
|
| 1404 |
"step": 940
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 0.8721735117674204,
|
| 1408 |
-
"grad_norm": 1.
|
| 1409 |
-
"learning_rate":
|
| 1410 |
-
"loss": 0.
|
| 1411 |
"step": 945
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 0.8767881864328565,
|
| 1415 |
-
"grad_norm": 1.
|
| 1416 |
-
"learning_rate":
|
| 1417 |
-
"loss": 0.
|
| 1418 |
"step": 950
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
"epoch": 0.8814028610982926,
|
| 1422 |
-
"grad_norm": 1.
|
| 1423 |
-
"learning_rate":
|
| 1424 |
-
"loss": 0.
|
| 1425 |
"step": 955
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 0.8860175357637287,
|
| 1429 |
-
"grad_norm": 1.
|
| 1430 |
-
"learning_rate":
|
| 1431 |
-
"loss": 0.
|
| 1432 |
"step": 960
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 0.8906322104291647,
|
| 1436 |
-
"grad_norm": 1.
|
| 1437 |
-
"learning_rate":
|
| 1438 |
-
"loss": 0.
|
| 1439 |
"step": 965
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 0.8952468850946008,
|
| 1443 |
-
"grad_norm": 1.
|
| 1444 |
-
"learning_rate":
|
| 1445 |
-
"loss": 0.
|
| 1446 |
"step": 970
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 0.8998615597600369,
|
| 1450 |
-
"grad_norm": 1.
|
| 1451 |
-
"learning_rate":
|
| 1452 |
-
"loss": 0.
|
| 1453 |
"step": 975
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 0.904476234425473,
|
| 1457 |
-
"grad_norm": 1.
|
| 1458 |
-
"learning_rate":
|
| 1459 |
-
"loss": 0.
|
| 1460 |
"step": 980
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 0.9090909090909091,
|
| 1464 |
-
"grad_norm": 1.
|
| 1465 |
-
"learning_rate":
|
| 1466 |
-
"loss": 0.
|
| 1467 |
"step": 985
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.9137055837563451,
|
| 1471 |
-
"grad_norm": 1.
|
| 1472 |
-
"learning_rate":
|
| 1473 |
-
"loss": 0.
|
| 1474 |
"step": 990
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 0.9183202584217812,
|
| 1478 |
-
"grad_norm": 1.
|
| 1479 |
-
"learning_rate":
|
| 1480 |
-
"loss": 0.
|
| 1481 |
"step": 995
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.9229349330872173,
|
| 1485 |
-
"grad_norm": 1.
|
| 1486 |
-
"learning_rate":
|
| 1487 |
-
"loss": 0.
|
| 1488 |
"step": 1000
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 0.9229349330872173,
|
| 1492 |
-
"eval_loss": 0.
|
| 1493 |
-
"eval_runtime":
|
| 1494 |
-
"eval_samples_per_second":
|
| 1495 |
-
"eval_steps_per_second": 0.
|
| 1496 |
"step": 1000
|
| 1497 |
},
|
| 1498 |
{
|
| 1499 |
"epoch": 0.9275496077526535,
|
| 1500 |
-
"grad_norm": 1.
|
| 1501 |
-
"learning_rate":
|
| 1502 |
-
"loss": 0.
|
| 1503 |
"step": 1005
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 0.9321642824180896,
|
| 1507 |
-
"grad_norm": 1.
|
| 1508 |
-
"learning_rate":
|
| 1509 |
-
"loss": 0.
|
| 1510 |
"step": 1010
|
| 1511 |
},
|
| 1512 |
{
|
| 1513 |
"epoch": 0.9367789570835257,
|
| 1514 |
-
"grad_norm": 1.
|
| 1515 |
-
"learning_rate":
|
| 1516 |
-
"loss": 0.
|
| 1517 |
"step": 1015
|
| 1518 |
},
|
| 1519 |
{
|
| 1520 |
"epoch": 0.9413936317489617,
|
| 1521 |
-
"grad_norm": 1.
|
| 1522 |
-
"learning_rate":
|
| 1523 |
-
"loss": 0.
|
| 1524 |
"step": 1020
|
| 1525 |
},
|
| 1526 |
{
|
| 1527 |
"epoch": 0.9460083064143978,
|
| 1528 |
-
"grad_norm": 1.
|
| 1529 |
-
"learning_rate":
|
| 1530 |
-
"loss": 0.
|
| 1531 |
"step": 1025
|
| 1532 |
},
|
| 1533 |
{
|
| 1534 |
"epoch": 0.9506229810798339,
|
| 1535 |
-
"grad_norm": 1.
|
| 1536 |
-
"learning_rate":
|
| 1537 |
-
"loss": 0.
|
| 1538 |
"step": 1030
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
"epoch": 0.95523765574527,
|
| 1542 |
-
"grad_norm": 1.
|
| 1543 |
-
"learning_rate":
|
| 1544 |
-
"loss": 0.
|
| 1545 |
"step": 1035
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 0.9598523304107061,
|
| 1549 |
-
"grad_norm": 1.
|
| 1550 |
-
"learning_rate":
|
| 1551 |
-
"loss": 0.
|
| 1552 |
"step": 1040
|
| 1553 |
},
|
| 1554 |
{
|
| 1555 |
"epoch": 0.9644670050761421,
|
| 1556 |
-
"grad_norm": 1.
|
| 1557 |
-
"learning_rate":
|
| 1558 |
-
"loss": 0.
|
| 1559 |
"step": 1045
|
| 1560 |
},
|
| 1561 |
{
|
| 1562 |
"epoch": 0.9690816797415782,
|
| 1563 |
-
"grad_norm": 1.
|
| 1564 |
-
"learning_rate":
|
| 1565 |
-
"loss": 0.
|
| 1566 |
"step": 1050
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 0.9736963544070143,
|
| 1570 |
-
"grad_norm": 1.
|
| 1571 |
-
"learning_rate":
|
| 1572 |
-
"loss": 0.
|
| 1573 |
"step": 1055
|
| 1574 |
},
|
| 1575 |
{
|
| 1576 |
"epoch": 0.9783110290724504,
|
| 1577 |
-
"grad_norm": 1.
|
| 1578 |
-
"learning_rate":
|
| 1579 |
-
"loss": 0.
|
| 1580 |
"step": 1060
|
| 1581 |
},
|
| 1582 |
{
|
| 1583 |
"epoch": 0.9829257037378865,
|
| 1584 |
-
"grad_norm": 1.
|
| 1585 |
-
"learning_rate":
|
| 1586 |
-
"loss": 0.
|
| 1587 |
"step": 1065
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
"epoch": 0.9875403784033225,
|
| 1591 |
-
"grad_norm": 1.
|
| 1592 |
-
"learning_rate":
|
| 1593 |
-
"loss": 0.
|
| 1594 |
"step": 1070
|
| 1595 |
},
|
| 1596 |
{
|
| 1597 |
"epoch": 0.9921550530687586,
|
| 1598 |
-
"grad_norm": 1.
|
| 1599 |
-
"learning_rate":
|
| 1600 |
-
"loss": 0.
|
| 1601 |
"step": 1075
|
| 1602 |
},
|
| 1603 |
{
|
| 1604 |
"epoch": 0.9967697277341947,
|
| 1605 |
-
"grad_norm": 1.
|
| 1606 |
-
"learning_rate":
|
| 1607 |
-
"loss": 0.
|
| 1608 |
"step": 1080
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
"epoch": 0.9995385325334564,
|
| 1612 |
"step": 1083,
|
| 1613 |
"total_flos": 453306954547200.0,
|
| 1614 |
-
"train_loss": 0.
|
| 1615 |
-
"train_runtime":
|
| 1616 |
-
"train_samples_per_second":
|
| 1617 |
-
"train_steps_per_second": 0.
|
| 1618 |
}
|
| 1619 |
],
|
| 1620 |
"logging_steps": 5,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0009229349330872173,
|
| 13 |
+
"grad_norm": 9.222650171740101,
|
| 14 |
+
"learning_rate": 9.174311926605506e-08,
|
| 15 |
+
"loss": 1.1391,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.0046146746654360865,
|
| 20 |
+
"grad_norm": 8.813354760736841,
|
| 21 |
+
"learning_rate": 4.587155963302753e-07,
|
| 22 |
+
"loss": 1.1346,
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.009229349330872173,
|
| 27 |
+
"grad_norm": 5.082254385960609,
|
| 28 |
+
"learning_rate": 9.174311926605506e-07,
|
| 29 |
+
"loss": 1.0934,
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01384402399630826,
|
| 34 |
+
"grad_norm": 3.211536911547781,
|
| 35 |
+
"learning_rate": 1.3761467889908258e-06,
|
| 36 |
+
"loss": 1.0264,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.018458698661744346,
|
| 41 |
+
"grad_norm": 2.7181445543779494,
|
| 42 |
+
"learning_rate": 1.8348623853211011e-06,
|
| 43 |
+
"loss": 1.0199,
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.023073373327180433,
|
| 48 |
+
"grad_norm": 2.3056635728248183,
|
| 49 |
+
"learning_rate": 2.2935779816513764e-06,
|
| 50 |
+
"loss": 1.0075,
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.02768804799261652,
|
| 55 |
+
"grad_norm": 2.6315302800596365,
|
| 56 |
+
"learning_rate": 2.7522935779816517e-06,
|
| 57 |
+
"loss": 0.9732,
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.032302722658052604,
|
| 62 |
+
"grad_norm": 2.0678619803720886,
|
| 63 |
+
"learning_rate": 3.211009174311927e-06,
|
| 64 |
+
"loss": 0.9964,
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03691739732348869,
|
| 69 |
+
"grad_norm": 3.005634145043816,
|
| 70 |
+
"learning_rate": 3.6697247706422022e-06,
|
| 71 |
+
"loss": 1.0015,
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.04153207198892478,
|
| 76 |
+
"grad_norm": 2.495857277480277,
|
| 77 |
+
"learning_rate": 4.128440366972478e-06,
|
| 78 |
+
"loss": 0.979,
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.046146746654360866,
|
| 83 |
+
"grad_norm": 2.442092473284365,
|
| 84 |
+
"learning_rate": 4.587155963302753e-06,
|
| 85 |
+
"loss": 0.9723,
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.050761421319796954,
|
| 90 |
+
"grad_norm": 2.1284759016567136,
|
| 91 |
+
"learning_rate": 5.045871559633028e-06,
|
| 92 |
+
"loss": 0.9722,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05537609598523304,
|
| 97 |
+
"grad_norm": 2.83564108049901,
|
| 98 |
+
"learning_rate": 5.504587155963303e-06,
|
| 99 |
+
"loss": 0.9852,
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05999077065066913,
|
| 104 |
+
"grad_norm": 2.0748240090940975,
|
| 105 |
+
"learning_rate": 5.963302752293578e-06,
|
| 106 |
+
"loss": 0.9845,
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06460544531610521,
|
| 111 |
+
"grad_norm": 2.9709334092183863,
|
| 112 |
+
"learning_rate": 6.422018348623854e-06,
|
| 113 |
+
"loss": 0.9615,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0692201199815413,
|
| 118 |
+
"grad_norm": 2.184920669369361,
|
| 119 |
+
"learning_rate": 6.880733944954129e-06,
|
| 120 |
+
"loss": 0.9675,
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07383479464697738,
|
| 125 |
+
"grad_norm": 2.1738804546566817,
|
| 126 |
+
"learning_rate": 7.3394495412844045e-06,
|
| 127 |
+
"loss": 0.9589,
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07844946931241348,
|
| 132 |
+
"grad_norm": 2.223214536084122,
|
| 133 |
+
"learning_rate": 7.79816513761468e-06,
|
| 134 |
+
"loss": 0.977,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.08306414397784956,
|
| 139 |
+
"grad_norm": 1.990814551883062,
|
| 140 |
+
"learning_rate": 8.256880733944956e-06,
|
| 141 |
+
"loss": 0.9767,
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08767881864328565,
|
| 146 |
+
"grad_norm": 2.0700255239398375,
|
| 147 |
+
"learning_rate": 8.71559633027523e-06,
|
| 148 |
+
"loss": 0.9828,
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.09229349330872173,
|
| 153 |
+
"grad_norm": 1.821385794436734,
|
| 154 |
+
"learning_rate": 9.174311926605506e-06,
|
| 155 |
+
"loss": 0.9896,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.09229349330872173,
|
| 160 |
+
"eval_loss": 0.9892959594726562,
|
| 161 |
+
"eval_runtime": 901.1864,
|
| 162 |
+
"eval_samples_per_second": 17.033,
|
| 163 |
+
"eval_steps_per_second": 0.133,
|
| 164 |
"step": 100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.09690816797415783,
|
| 168 |
+
"grad_norm": 2.1363964627484155,
|
| 169 |
+
"learning_rate": 9.633027522935781e-06,
|
| 170 |
+
"loss": 1.015,
|
| 171 |
"step": 105
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.10152284263959391,
|
| 175 |
+
"grad_norm": 1.9477999761197564,
|
| 176 |
+
"learning_rate": 9.999973991131207e-06,
|
| 177 |
+
"loss": 0.9983,
|
| 178 |
"step": 110
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.10613751730503,
|
| 182 |
+
"grad_norm": 2.0399110459564516,
|
| 183 |
+
"learning_rate": 9.999063709134502e-06,
|
| 184 |
+
"loss": 1.0062,
|
| 185 |
"step": 115
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.11075219197046608,
|
| 189 |
+
"grad_norm": 1.9098211009145907,
|
| 190 |
+
"learning_rate": 9.996853254269984e-06,
|
| 191 |
+
"loss": 0.9855,
|
| 192 |
"step": 120
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.11536686663590216,
|
| 196 |
+
"grad_norm": 1.7833107957442789,
|
| 197 |
+
"learning_rate": 9.99334320144e-06,
|
| 198 |
+
"loss": 1.0093,
|
| 199 |
"step": 125
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.11998154130133826,
|
| 203 |
+
"grad_norm": 1.7815777364460414,
|
| 204 |
+
"learning_rate": 9.988534463550585e-06,
|
| 205 |
+
"loss": 0.981,
|
| 206 |
"step": 130
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.12459621596677434,
|
| 210 |
+
"grad_norm": 1.8694974323828528,
|
| 211 |
+
"learning_rate": 9.982428291274047e-06,
|
| 212 |
+
"loss": 1.0102,
|
| 213 |
"step": 135
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.12921089063221042,
|
| 217 |
+
"grad_norm": 1.8237445983099982,
|
| 218 |
+
"learning_rate": 9.975026272723677e-06,
|
| 219 |
+
"loss": 1.006,
|
| 220 |
"step": 140
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.13382556529764653,
|
| 224 |
+
"grad_norm": 1.8625032130850405,
|
| 225 |
+
"learning_rate": 9.96633033304071e-06,
|
| 226 |
+
"loss": 0.984,
|
| 227 |
"step": 145
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.1384402399630826,
|
| 231 |
+
"grad_norm": 2.072572452539365,
|
| 232 |
+
"learning_rate": 9.95634273389363e-06,
|
| 233 |
+
"loss": 0.9712,
|
| 234 |
"step": 150
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.1430549146285187,
|
| 238 |
+
"grad_norm": 2.0593175620343547,
|
| 239 |
+
"learning_rate": 9.945066072889942e-06,
|
| 240 |
+
"loss": 1.0211,
|
| 241 |
"step": 155
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.14766958929395477,
|
| 245 |
+
"grad_norm": 2.0200152592170943,
|
| 246 |
+
"learning_rate": 9.93250328290059e-06,
|
| 247 |
+
"loss": 0.9775,
|
| 248 |
"step": 160
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.15228426395939088,
|
| 252 |
+
"grad_norm": 1.875921122063454,
|
| 253 |
+
"learning_rate": 9.918657631297153e-06,
|
| 254 |
+
"loss": 0.9847,
|
| 255 |
"step": 165
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.15689893862482696,
|
| 259 |
+
"grad_norm": 1.8527340279543534,
|
| 260 |
+
"learning_rate": 9.90353271910206e-06,
|
| 261 |
+
"loss": 0.9872,
|
| 262 |
"step": 170
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.16151361329026304,
|
| 266 |
+
"grad_norm": 1.8908823561795056,
|
| 267 |
+
"learning_rate": 9.887132480052028e-06,
|
| 268 |
+
"loss": 0.9712,
|
| 269 |
"step": 175
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.16612828795569912,
|
| 273 |
+
"grad_norm": 1.8296519425504616,
|
| 274 |
+
"learning_rate": 9.869461179574963e-06,
|
| 275 |
+
"loss": 0.9859,
|
| 276 |
"step": 180
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.1707429626211352,
|
| 280 |
+
"grad_norm": 1.7574451917239733,
|
| 281 |
+
"learning_rate": 9.850523413680588e-06,
|
| 282 |
+
"loss": 0.9911,
|
| 283 |
"step": 185
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.1753576372865713,
|
| 287 |
+
"grad_norm": 1.7389467033514425,
|
| 288 |
+
"learning_rate": 9.830324107765104e-06,
|
| 289 |
+
"loss": 0.9739,
|
| 290 |
"step": 190
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.17997231195200739,
|
| 294 |
+
"grad_norm": 1.7050530099463075,
|
| 295 |
+
"learning_rate": 9.808868515330169e-06,
|
| 296 |
+
"loss": 0.9803,
|
| 297 |
"step": 195
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.18458698661744347,
|
| 301 |
+
"grad_norm": 1.756999826792102,
|
| 302 |
+
"learning_rate": 9.786162216616561e-06,
|
| 303 |
+
"loss": 0.9838,
|
| 304 |
"step": 200
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.18458698661744347,
|
| 308 |
+
"eval_loss": 0.9935115575790405,
|
| 309 |
+
"eval_runtime": 751.2551,
|
| 310 |
+
"eval_samples_per_second": 20.432,
|
| 311 |
+
"eval_steps_per_second": 0.16,
|
| 312 |
"step": 200
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 0.18920166128287955,
|
| 316 |
+
"grad_norm": 1.778225787399477,
|
| 317 |
+
"learning_rate": 9.762211117152839e-06,
|
| 318 |
+
"loss": 0.9754,
|
| 319 |
"step": 205
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 0.19381633594831565,
|
| 323 |
+
"grad_norm": 1.6039072749046104,
|
| 324 |
+
"learning_rate": 9.737021446219424e-06,
|
| 325 |
+
"loss": 0.9577,
|
| 326 |
"step": 210
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.19843101061375173,
|
| 330 |
+
"grad_norm": 1.8562210127250676,
|
| 331 |
+
"learning_rate": 9.71059975522846e-06,
|
| 332 |
+
"loss": 0.9708,
|
| 333 |
"step": 215
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 0.20304568527918782,
|
| 337 |
+
"grad_norm": 1.602175292503488,
|
| 338 |
+
"learning_rate": 9.682952916019907e-06,
|
| 339 |
+
"loss": 0.9922,
|
| 340 |
"step": 220
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 0.2076603599446239,
|
| 344 |
+
"grad_norm": 1.5958137821643004,
|
| 345 |
+
"learning_rate": 9.654088119074282e-06,
|
| 346 |
+
"loss": 0.9965,
|
| 347 |
"step": 225
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 0.21227503461006,
|
| 351 |
+
"grad_norm": 1.8573846533843932,
|
| 352 |
+
"learning_rate": 9.624012871642545e-06,
|
| 353 |
+
"loss": 0.9753,
|
| 354 |
"step": 230
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.21688970927549608,
|
| 358 |
+
"grad_norm": 1.640370350766078,
|
| 359 |
+
"learning_rate": 9.592734995793583e-06,
|
| 360 |
+
"loss": 0.9925,
|
| 361 |
"step": 235
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.22150438394093216,
|
| 365 |
+
"grad_norm": 1.6631975332565576,
|
| 366 |
+
"learning_rate": 9.560262626379824e-06,
|
| 367 |
+
"loss": 0.9829,
|
| 368 |
"step": 240
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.22611905860636825,
|
| 372 |
+
"grad_norm": 1.6393108832006937,
|
| 373 |
+
"learning_rate": 9.52660420892149e-06,
|
| 374 |
+
"loss": 0.9837,
|
| 375 |
"step": 245
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 0.23073373327180433,
|
| 379 |
+
"grad_norm": 1.5999393221352023,
|
| 380 |
+
"learning_rate": 9.49176849741007e-06,
|
| 381 |
+
"loss": 0.9822,
|
| 382 |
"step": 250
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 0.23534840793724043,
|
| 386 |
+
"grad_norm": 1.6562799165061401,
|
| 387 |
+
"learning_rate": 9.455764552031546e-06,
|
| 388 |
+
"loss": 0.9876,
|
| 389 |
"step": 255
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 0.23996308260267651,
|
| 393 |
+
"grad_norm": 1.597298992345653,
|
| 394 |
+
"learning_rate": 9.418601736809989e-06,
|
| 395 |
+
"loss": 0.99,
|
| 396 |
"step": 260
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.2445777572681126,
|
| 400 |
+
"grad_norm": 1.615615429575249,
|
| 401 |
+
"learning_rate": 9.380289717172141e-06,
|
| 402 |
+
"loss": 0.9944,
|
| 403 |
"step": 265
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 0.24919243193354867,
|
| 407 |
+
"grad_norm": 1.6391764879253228,
|
| 408 |
+
"learning_rate": 9.340838457433588e-06,
|
| 409 |
+
"loss": 0.9941,
|
| 410 |
"step": 270
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.25380710659898476,
|
| 414 |
+
"grad_norm": 1.7100035873468014,
|
| 415 |
+
"learning_rate": 9.300258218207215e-06,
|
| 416 |
+
"loss": 0.9785,
|
| 417 |
"step": 275
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.25842178126442084,
|
| 421 |
+
"grad_norm": 2.095593373075969,
|
| 422 |
+
"learning_rate": 9.258559553734597e-06,
|
| 423 |
+
"loss": 0.9883,
|
| 424 |
"step": 280
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 0.26303645592985697,
|
| 428 |
+
"grad_norm": 1.6273765059661172,
|
| 429 |
+
"learning_rate": 9.215753309141e-06,
|
| 430 |
+
"loss": 0.9772,
|
| 431 |
"step": 285
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 0.26765113059529305,
|
| 435 |
+
"grad_norm": 1.6695304390680636,
|
| 436 |
+
"learning_rate": 9.17185061761477e-06,
|
| 437 |
+
"loss": 0.9881,
|
| 438 |
"step": 290
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 0.27226580526072913,
|
| 442 |
+
"grad_norm": 1.6630409654600378,
|
| 443 |
+
"learning_rate": 9.126862897511752e-06,
|
| 444 |
+
"loss": 0.9785,
|
| 445 |
"step": 295
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 0.2768804799261652,
|
| 449 |
+
"grad_norm": 1.6120537857852493,
|
| 450 |
+
"learning_rate": 9.080801849385585e-06,
|
| 451 |
+
"loss": 0.9853,
|
| 452 |
"step": 300
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.2768804799261652,
|
| 456 |
+
"eval_loss": 0.9881044626235962,
|
| 457 |
+
"eval_runtime": 648.9048,
|
| 458 |
+
"eval_samples_per_second": 23.655,
|
| 459 |
+
"eval_steps_per_second": 0.185,
|
| 460 |
"step": 300
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 0.2814951545916013,
|
| 464 |
+
"grad_norm": 1.5682941966581576,
|
| 465 |
+
"learning_rate": 9.033679452944574e-06,
|
| 466 |
+
"loss": 0.9593,
|
| 467 |
"step": 305
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 0.2861098292570374,
|
| 471 |
+
"grad_norm": 1.7509871813668145,
|
| 472 |
+
"learning_rate": 8.98550796393597e-06,
|
| 473 |
+
"loss": 0.9833,
|
| 474 |
"step": 310
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.29072450392247345,
|
| 478 |
+
"grad_norm": 1.5617036787004468,
|
| 479 |
+
"learning_rate": 8.936299910958461e-06,
|
| 480 |
+
"loss": 0.9732,
|
| 481 |
"step": 315
|
| 482 |
},
|
| 483 |
{
|
| 484 |
"epoch": 0.29533917858790953,
|
| 485 |
+
"grad_norm": 1.6332967875738513,
|
| 486 |
+
"learning_rate": 8.886068092203684e-06,
|
| 487 |
+
"loss": 0.9744,
|
| 488 |
"step": 320
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.2999538532533456,
|
| 492 |
+
"grad_norm": 1.7354150495189573,
|
| 493 |
+
"learning_rate": 8.834825572127632e-06,
|
| 494 |
+
"loss": 0.9827,
|
| 495 |
"step": 325
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 0.30456852791878175,
|
| 499 |
+
"grad_norm": 1.588233836037739,
|
| 500 |
+
"learning_rate": 8.782585678052814e-06,
|
| 501 |
+
"loss": 0.9962,
|
| 502 |
"step": 330
|
| 503 |
},
|
| 504 |
{
|
| 505 |
"epoch": 0.30918320258421783,
|
| 506 |
+
"grad_norm": 1.5809425639553816,
|
| 507 |
+
"learning_rate": 8.729361996702032e-06,
|
| 508 |
+
"loss": 0.9903,
|
| 509 |
"step": 335
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 0.3137978772496539,
|
| 513 |
+
"grad_norm": 1.603602687498353,
|
| 514 |
+
"learning_rate": 8.675168370664706e-06,
|
| 515 |
+
"loss": 0.981,
|
| 516 |
"step": 340
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 0.31841255191509,
|
| 520 |
+
"grad_norm": 1.5782147372024207,
|
| 521 |
+
"learning_rate": 8.620018894796654e-06,
|
| 522 |
+
"loss": 0.9888,
|
| 523 |
"step": 345
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 0.3230272265805261,
|
| 527 |
+
"grad_norm": 1.5566880880848435,
|
| 528 |
+
"learning_rate": 8.56392791255424e-06,
|
| 529 |
+
"loss": 0.9648,
|
| 530 |
"step": 350
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.32764190124596215,
|
| 534 |
+
"grad_norm": 1.655119543708647,
|
| 535 |
+
"learning_rate": 8.5069100122639e-06,
|
| 536 |
+
"loss": 0.9761,
|
| 537 |
"step": 355
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.33225657591139823,
|
| 541 |
+
"grad_norm": 1.9303593174940483,
|
| 542 |
+
"learning_rate": 8.448980023327943e-06,
|
| 543 |
+
"loss": 0.978,
|
| 544 |
"step": 360
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 0.3368712505768343,
|
| 548 |
+
"grad_norm": 1.6875645085180389,
|
| 549 |
+
"learning_rate": 8.390153012367692e-06,
|
| 550 |
+
"loss": 0.9704,
|
| 551 |
"step": 365
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 0.3414859252422704,
|
| 555 |
+
"grad_norm": 1.625080279365752,
|
| 556 |
+
"learning_rate": 8.330444279304887e-06,
|
| 557 |
+
"loss": 0.9933,
|
| 558 |
"step": 370
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 0.34610059990770653,
|
| 562 |
+
"grad_norm": 1.5399521803126077,
|
| 563 |
+
"learning_rate": 8.269869353382448e-06,
|
| 564 |
+
"loss": 0.9768,
|
| 565 |
"step": 375
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 0.3507152745731426,
|
| 569 |
+
"grad_norm": 1.5199082736016265,
|
| 570 |
+
"learning_rate": 8.208443989125567e-06,
|
| 571 |
+
"loss": 0.9702,
|
| 572 |
"step": 380
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.3553299492385787,
|
| 576 |
+
"grad_norm": 1.6898798315423416,
|
| 577 |
+
"learning_rate": 8.146184162244231e-06,
|
| 578 |
+
"loss": 0.9508,
|
| 579 |
"step": 385
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.35994462390401477,
|
| 583 |
+
"grad_norm": 1.5767639905295907,
|
| 584 |
+
"learning_rate": 8.083106065478192e-06,
|
| 585 |
+
"loss": 0.9617,
|
| 586 |
"step": 390
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 0.36455929856945085,
|
| 590 |
+
"grad_norm": 1.6940722444661227,
|
| 591 |
+
"learning_rate": 8.019226104385519e-06,
|
| 592 |
+
"loss": 0.9827,
|
| 593 |
"step": 395
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 0.36917397323488693,
|
| 597 |
+
"grad_norm": 1.542549043886959,
|
| 598 |
+
"learning_rate": 7.954560893075785e-06,
|
| 599 |
+
"loss": 0.9638,
|
| 600 |
"step": 400
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 0.36917397323488693,
|
| 604 |
+
"eval_loss": 0.978050708770752,
|
| 605 |
+
"eval_runtime": 631.4932,
|
| 606 |
+
"eval_samples_per_second": 24.307,
|
| 607 |
+
"eval_steps_per_second": 0.19,
|
| 608 |
"step": 400
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 0.373788647900323,
|
| 612 |
+
"grad_norm": 1.5935524851120768,
|
| 613 |
+
"learning_rate": 7.889127249889003e-06,
|
| 614 |
+
"loss": 0.9673,
|
| 615 |
"step": 405
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 0.3784033225657591,
|
| 619 |
+
"grad_norm": 1.497119604051003,
|
| 620 |
+
"learning_rate": 7.82294219302148e-06,
|
| 621 |
+
"loss": 0.9574,
|
| 622 |
"step": 410
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 0.3830179972311952,
|
| 626 |
+
"grad_norm": 1.709877148318467,
|
| 627 |
+
"learning_rate": 7.75602293609964e-06,
|
| 628 |
+
"loss": 0.9483,
|
| 629 |
"step": 415
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 0.3876326718966313,
|
| 633 |
+
"grad_norm": 1.5527701210281757,
|
| 634 |
+
"learning_rate": 7.688386883703071e-06,
|
| 635 |
+
"loss": 0.9904,
|
| 636 |
"step": 420
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 0.3922473465620674,
|
| 640 |
+
"grad_norm": 1.651581387061671,
|
| 641 |
+
"learning_rate": 7.620051626837878e-06,
|
| 642 |
+
"loss": 0.9619,
|
| 643 |
"step": 425
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 0.39686202122750347,
|
| 647 |
+
"grad_norm": 1.5992647254822225,
|
| 648 |
+
"learning_rate": 7.55103493836155e-06,
|
| 649 |
+
"loss": 0.9688,
|
| 650 |
"step": 430
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.40147669589293955,
|
| 654 |
+
"grad_norm": 1.4883188672166754,
|
| 655 |
+
"learning_rate": 7.481354768360543e-06,
|
| 656 |
+
"loss": 0.9493,
|
| 657 |
"step": 435
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 0.40609137055837563,
|
| 661 |
+
"grad_norm": 1.540282089268457,
|
| 662 |
+
"learning_rate": 7.411029239481766e-06,
|
| 663 |
+
"loss": 0.9656,
|
| 664 |
"step": 440
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 0.4107060452238117,
|
| 668 |
+
"grad_norm": 1.57006447700181,
|
| 669 |
+
"learning_rate": 7.340076642219172e-06,
|
| 670 |
+
"loss": 0.963,
|
| 671 |
"step": 445
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 0.4153207198892478,
|
| 675 |
+
"grad_norm": 1.6069506996123204,
|
| 676 |
+
"learning_rate": 7.268515430156722e-06,
|
| 677 |
+
"loss": 0.9729,
|
| 678 |
"step": 450
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 0.41993539455468387,
|
| 682 |
+
"grad_norm": 1.5002386553771234,
|
| 683 |
+
"learning_rate": 7.196364215168901e-06,
|
| 684 |
+
"loss": 0.9521,
|
| 685 |
"step": 455
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 0.42455006922012,
|
| 689 |
+
"grad_norm": 1.4646720705420488,
|
| 690 |
+
"learning_rate": 7.123641762580089e-06,
|
| 691 |
+
"loss": 0.9507,
|
| 692 |
"step": 460
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 0.4291647438855561,
|
| 696 |
+
"grad_norm": 1.6149847953084777,
|
| 697 |
+
"learning_rate": 7.050366986284019e-06,
|
| 698 |
+
"loss": 0.9635,
|
| 699 |
"step": 465
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.43377941855099217,
|
| 703 |
+
"grad_norm": 1.554087118833947,
|
| 704 |
+
"learning_rate": 6.9765589438245765e-06,
|
| 705 |
+
"loss": 0.9744,
|
| 706 |
"step": 470
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 0.43839409321642825,
|
| 710 |
+
"grad_norm": 1.4890425150210163,
|
| 711 |
+
"learning_rate": 6.9022368314392595e-06,
|
| 712 |
+
"loss": 0.9694,
|
| 713 |
"step": 475
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.44300876788186433,
|
| 717 |
+
"grad_norm": 1.5064709526763582,
|
| 718 |
+
"learning_rate": 6.827419979066559e-06,
|
| 719 |
+
"loss": 0.9577,
|
| 720 |
"step": 480
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 0.4476234425473004,
|
| 724 |
+
"grad_norm": 1.5765372233113244,
|
| 725 |
+
"learning_rate": 6.752127845318561e-06,
|
| 726 |
+
"loss": 0.9777,
|
| 727 |
"step": 485
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 0.4522381172127365,
|
| 731 |
+
"grad_norm": 1.502803008633363,
|
| 732 |
+
"learning_rate": 6.676380012420087e-06,
|
| 733 |
+
"loss": 0.9543,
|
| 734 |
"step": 490
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.45685279187817257,
|
| 738 |
+
"grad_norm": 1.480113884170601,
|
| 739 |
+
"learning_rate": 6.600196181115692e-06,
|
| 740 |
+
"loss": 0.9413,
|
| 741 |
"step": 495
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 0.46146746654360865,
|
| 745 |
+
"grad_norm": 1.518180225480369,
|
| 746 |
+
"learning_rate": 6.523596165545818e-06,
|
| 747 |
+
"loss": 0.9745,
|
| 748 |
"step": 500
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 0.46146746654360865,
|
| 752 |
+
"eval_loss": 0.9680244326591492,
|
| 753 |
+
"eval_runtime": 631.2665,
|
| 754 |
+
"eval_samples_per_second": 24.316,
|
| 755 |
+
"eval_steps_per_second": 0.19,
|
| 756 |
"step": 500
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"epoch": 0.4660821412090448,
|
| 760 |
+
"grad_norm": 1.5182752103495631,
|
| 761 |
+
"learning_rate": 6.446599888093478e-06,
|
| 762 |
+
"loss": 0.9493,
|
| 763 |
"step": 505
|
| 764 |
},
|
| 765 |
{
|
| 766 |
"epoch": 0.47069681587448087,
|
| 767 |
+
"grad_norm": 1.6381019403661983,
|
| 768 |
+
"learning_rate": 6.369227374202776e-06,
|
| 769 |
+
"loss": 0.9655,
|
| 770 |
"step": 510
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.47531149053991695,
|
| 774 |
+
"grad_norm": 1.480944986996014,
|
| 775 |
+
"learning_rate": 6.291498747170615e-06,
|
| 776 |
+
"loss": 0.973,
|
| 777 |
"step": 515
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.47992616520535303,
|
| 781 |
+
"grad_norm": 1.515420838901527,
|
| 782 |
+
"learning_rate": 6.213434222912977e-06,
|
| 783 |
+
"loss": 0.9618,
|
| 784 |
"step": 520
|
| 785 |
},
|
| 786 |
{
|
| 787 |
"epoch": 0.4845408398707891,
|
| 788 |
+
"grad_norm": 1.5303480066641486,
|
| 789 |
+
"learning_rate": 6.135054104707093e-06,
|
| 790 |
+
"loss": 0.9439,
|
| 791 |
"step": 525
|
| 792 |
},
|
| 793 |
{
|
| 794 |
"epoch": 0.4891555145362252,
|
| 795 |
+
"grad_norm": 1.487348077399422,
|
| 796 |
+
"learning_rate": 6.056378777910898e-06,
|
| 797 |
+
"loss": 0.9565,
|
| 798 |
"step": 530
|
| 799 |
},
|
| 800 |
{
|
| 801 |
"epoch": 0.49377018920166127,
|
| 802 |
+
"grad_norm": 1.5775060237163985,
|
| 803 |
+
"learning_rate": 5.977428704661151e-06,
|
| 804 |
+
"loss": 0.9407,
|
| 805 |
"step": 535
|
| 806 |
},
|
| 807 |
{
|
| 808 |
"epoch": 0.49838486386709735,
|
| 809 |
+
"grad_norm": 1.6339716299433642,
|
| 810 |
+
"learning_rate": 5.898224418551565e-06,
|
| 811 |
+
"loss": 0.9532,
|
| 812 |
"step": 540
|
| 813 |
},
|
| 814 |
{
|
| 815 |
"epoch": 0.5029995385325334,
|
| 816 |
+
"grad_norm": 1.5688231463762927,
|
| 817 |
+
"learning_rate": 5.8187865192923644e-06,
|
| 818 |
+
"loss": 0.9433,
|
| 819 |
"step": 545
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.5076142131979695,
|
| 823 |
+
"grad_norm": 1.465375400099311,
|
| 824 |
+
"learning_rate": 5.739135667352651e-06,
|
| 825 |
+
"loss": 0.9494,
|
| 826 |
"step": 550
|
| 827 |
},
|
| 828 |
{
|
| 829 |
"epoch": 0.5122288878634056,
|
| 830 |
+
"grad_norm": 1.590916658194805,
|
| 831 |
+
"learning_rate": 5.659292578586957e-06,
|
| 832 |
+
"loss": 0.9574,
|
| 833 |
"step": 555
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 0.5168435625288417,
|
| 837 |
+
"grad_norm": 1.4545280778553549,
|
| 838 |
+
"learning_rate": 5.579278018847395e-06,
|
| 839 |
+
"loss": 0.9471,
|
| 840 |
"step": 560
|
| 841 |
},
|
| 842 |
{
|
| 843 |
"epoch": 0.5214582371942778,
|
| 844 |
+
"grad_norm": 1.4258346370497963,
|
| 845 |
+
"learning_rate": 5.499112798582814e-06,
|
| 846 |
+
"loss": 0.9456,
|
| 847 |
"step": 565
|
| 848 |
},
|
| 849 |
{
|
| 850 |
"epoch": 0.5260729118597139,
|
| 851 |
+
"grad_norm": 1.4412831745954977,
|
| 852 |
+
"learning_rate": 5.418817767426343e-06,
|
| 853 |
+
"loss": 0.9419,
|
| 854 |
"step": 570
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.53068758652515,
|
| 858 |
+
"grad_norm": 1.5161519877588197,
|
| 859 |
+
"learning_rate": 5.3384138087727555e-06,
|
| 860 |
+
"loss": 0.9429,
|
| 861 |
"step": 575
|
| 862 |
},
|
| 863 |
{
|
| 864 |
"epoch": 0.5353022611905861,
|
| 865 |
+
"grad_norm": 1.5365045658327852,
|
| 866 |
+
"learning_rate": 5.257921834347043e-06,
|
| 867 |
+
"loss": 0.9421,
|
| 868 |
"step": 580
|
| 869 |
},
|
| 870 |
{
|
| 871 |
"epoch": 0.5399169358560222,
|
| 872 |
+
"grad_norm": 1.5096905798025197,
|
| 873 |
+
"learning_rate": 5.177362778765629e-06,
|
| 874 |
+
"loss": 0.9418,
|
| 875 |
"step": 585
|
| 876 |
},
|
| 877 |
{
|
| 878 |
"epoch": 0.5445316105214583,
|
| 879 |
+
"grad_norm": 1.4931388458814197,
|
| 880 |
+
"learning_rate": 5.096757594091623e-06,
|
| 881 |
+
"loss": 0.9336,
|
| 882 |
"step": 590
|
| 883 |
},
|
| 884 |
{
|
| 885 |
"epoch": 0.5491462851868943,
|
| 886 |
+
"grad_norm": 1.54787574418171,
|
| 887 |
+
"learning_rate": 5.01612724438554e-06,
|
| 888 |
+
"loss": 0.9594,
|
| 889 |
"step": 595
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.5537609598523304,
|
| 893 |
+
"grad_norm": 1.4047855369892224,
|
| 894 |
+
"learning_rate": 4.935492700252903e-06,
|
| 895 |
+
"loss": 0.9396,
|
| 896 |
"step": 600
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.5537609598523304,
|
| 900 |
+
"eval_loss": 0.9567832350730896,
|
| 901 |
+
"eval_runtime": 1029.8073,
|
| 902 |
+
"eval_samples_per_second": 14.906,
|
| 903 |
+
"eval_steps_per_second": 0.117,
|
| 904 |
"step": 600
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 0.5583756345177665,
|
| 908 |
+
"grad_norm": 1.532026589616403,
|
| 909 |
+
"learning_rate": 4.854874933390124e-06,
|
| 910 |
+
"loss": 0.9464,
|
| 911 |
"step": 605
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 0.5629903091832026,
|
| 915 |
+
"grad_norm": 1.4445505648251018,
|
| 916 |
+
"learning_rate": 4.774294911130141e-06,
|
| 917 |
+
"loss": 0.9564,
|
| 918 |
"step": 610
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 0.5676049838486387,
|
| 922 |
+
"grad_norm": 1.3806609000939527,
|
| 923 |
+
"learning_rate": 4.6937735909891456e-06,
|
| 924 |
+
"loss": 0.9401,
|
| 925 |
"step": 615
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 0.5722196585140747,
|
| 929 |
+
"grad_norm": 1.4758989461572256,
|
| 930 |
+
"learning_rate": 4.6133319152158886e-06,
|
| 931 |
+
"loss": 0.9504,
|
| 932 |
"step": 620
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 0.5768343331795108,
|
| 936 |
+
"grad_norm": 1.411552408489387,
|
| 937 |
+
"learning_rate": 4.532990805344958e-06,
|
| 938 |
+
"loss": 0.9382,
|
| 939 |
"step": 625
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 0.5814490078449469,
|
| 943 |
+
"grad_norm": 1.4215358213534153,
|
| 944 |
+
"learning_rate": 4.452771156755423e-06,
|
| 945 |
+
"loss": 0.9457,
|
| 946 |
"step": 630
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 0.586063682510383,
|
| 950 |
+
"grad_norm": 1.5194235066664028,
|
| 951 |
+
"learning_rate": 4.372693833236319e-06,
|
| 952 |
+
"loss": 0.9538,
|
| 953 |
"step": 635
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.5906783571758191,
|
| 957 |
+
"grad_norm": 1.504251672226047,
|
| 958 |
+
"learning_rate": 4.292779661560295e-06,
|
| 959 |
+
"loss": 0.9541,
|
| 960 |
"step": 640
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 0.5952930318412551,
|
| 964 |
+
"grad_norm": 1.477570181176633,
|
| 965 |
+
"learning_rate": 4.213049426066946e-06,
|
| 966 |
+
"loss": 0.932,
|
| 967 |
"step": 645
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 0.5999077065066912,
|
| 971 |
+
"grad_norm": 1.4830574491662214,
|
| 972 |
+
"learning_rate": 4.133523863257139e-06,
|
| 973 |
+
"loss": 0.9499,
|
| 974 |
"step": 650
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 0.6045223811721273,
|
| 978 |
+
"grad_norm": 1.4506011398131606,
|
| 979 |
+
"learning_rate": 4.054223656399794e-06,
|
| 980 |
+
"loss": 0.9432,
|
| 981 |
"step": 655
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 0.6091370558375635,
|
| 985 |
+
"grad_norm": 1.5259712782334296,
|
| 986 |
+
"learning_rate": 3.975169430152524e-06,
|
| 987 |
+
"loss": 0.9336,
|
| 988 |
"step": 660
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 0.6137517305029996,
|
| 992 |
+
"grad_norm": 1.4462086241646492,
|
| 993 |
+
"learning_rate": 3.8963817451974915e-06,
|
| 994 |
+
"loss": 0.9434,
|
| 995 |
"step": 665
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 0.6183664051684357,
|
| 999 |
+
"grad_norm": 1.4279004522752485,
|
| 1000 |
+
"learning_rate": 3.817881092893934e-06,
|
| 1001 |
+
"loss": 0.9468,
|
| 1002 |
"step": 670
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 0.6229810798338717,
|
| 1006 |
+
"grad_norm": 1.4169615015471333,
|
| 1007 |
+
"learning_rate": 3.7396878899486896e-06,
|
| 1008 |
+
"loss": 0.9416,
|
| 1009 |
"step": 675
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 0.6275957544993078,
|
| 1013 |
+
"grad_norm": 1.4479507419493,
|
| 1014 |
+
"learning_rate": 3.6618224731061658e-06,
|
| 1015 |
+
"loss": 0.9388,
|
| 1016 |
"step": 680
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 0.6322104291647439,
|
| 1020 |
+
"grad_norm": 1.4589224612269058,
|
| 1021 |
+
"learning_rate": 3.584305093859082e-06,
|
| 1022 |
+
"loss": 0.9384,
|
| 1023 |
"step": 685
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 0.63682510383018,
|
| 1027 |
+
"grad_norm": 1.4183964058904668,
|
| 1028 |
+
"learning_rate": 3.507155913181402e-06,
|
| 1029 |
+
"loss": 0.9347,
|
| 1030 |
"step": 690
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 0.6414397784956161,
|
| 1034 |
+
"grad_norm": 1.4542672740750342,
|
| 1035 |
+
"learning_rate": 3.4303949962848003e-06,
|
| 1036 |
+
"loss": 0.9494,
|
| 1037 |
"step": 695
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 0.6460544531610521,
|
| 1041 |
+
"grad_norm": 1.466448569041173,
|
| 1042 |
+
"learning_rate": 3.3540423074000323e-06,
|
| 1043 |
+
"loss": 0.9176,
|
| 1044 |
"step": 700
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.6460544531610521,
|
| 1048 |
+
"eval_loss": 0.9464961290359497,
|
| 1049 |
+
"eval_runtime": 630.6464,
|
| 1050 |
+
"eval_samples_per_second": 24.34,
|
| 1051 |
+
"eval_steps_per_second": 0.19,
|
| 1052 |
"step": 700
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.6506691278264882,
|
| 1056 |
+
"grad_norm": 1.4112284829468522,
|
| 1057 |
+
"learning_rate": 3.278117704584577e-06,
|
| 1058 |
+
"loss": 0.9164,
|
| 1059 |
"step": 705
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.6552838024919243,
|
| 1063 |
+
"grad_norm": 1.4108457010710305,
|
| 1064 |
+
"learning_rate": 3.202640934557884e-06,
|
| 1065 |
+
"loss": 0.9213,
|
| 1066 |
"step": 710
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.6598984771573604,
|
| 1070 |
+
"grad_norm": 1.4189577161809213,
|
| 1071 |
+
"learning_rate": 3.127631627565586e-06,
|
| 1072 |
+
"loss": 0.9368,
|
| 1073 |
"step": 715
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.6645131518227965,
|
| 1077 |
+
"grad_norm": 1.452002361321377,
|
| 1078 |
+
"learning_rate": 3.053109292273996e-06,
|
| 1079 |
+
"loss": 0.9372,
|
| 1080 |
"step": 720
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.6691278264882325,
|
| 1084 |
+
"grad_norm": 1.4024056929659159,
|
| 1085 |
+
"learning_rate": 2.9790933106962328e-06,
|
| 1086 |
+
"loss": 0.925,
|
| 1087 |
"step": 725
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.6737425011536686,
|
| 1091 |
+
"grad_norm": 1.4306957271780452,
|
| 1092 |
+
"learning_rate": 2.9056029331512853e-06,
|
| 1093 |
+
"loss": 0.9259,
|
| 1094 |
"step": 730
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.6783571758191047,
|
| 1098 |
+
"grad_norm": 1.3638433368150233,
|
| 1099 |
+
"learning_rate": 2.8326572732573167e-06,
|
| 1100 |
+
"loss": 0.9298,
|
| 1101 |
"step": 735
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.6829718504845408,
|
| 1105 |
+
"grad_norm": 1.438840741138399,
|
| 1106 |
+
"learning_rate": 2.7602753029605456e-06,
|
| 1107 |
+
"loss": 0.9312,
|
| 1108 |
"step": 740
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.687586525149977,
|
| 1112 |
+
"grad_norm": 1.3538477302959735,
|
| 1113 |
+
"learning_rate": 2.688475847600947e-06,
|
| 1114 |
+
"loss": 0.9328,
|
| 1115 |
"step": 745
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.6922011998154131,
|
| 1119 |
+
"grad_norm": 1.473692645733402,
|
| 1120 |
+
"learning_rate": 2.6172775810161104e-06,
|
| 1121 |
+
"loss": 0.9239,
|
| 1122 |
"step": 750
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.6968158744808491,
|
| 1126 |
+
"grad_norm": 1.5623662932746458,
|
| 1127 |
+
"learning_rate": 2.546699020684471e-06,
|
| 1128 |
+
"loss": 0.9371,
|
| 1129 |
"step": 755
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.7014305491462852,
|
| 1133 |
+
"grad_norm": 1.4016556607303827,
|
| 1134 |
+
"learning_rate": 2.4767585229092368e-06,
|
| 1135 |
+
"loss": 0.9308,
|
| 1136 |
"step": 760
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.7060452238117213,
|
| 1140 |
+
"grad_norm": 1.3899751453383695,
|
| 1141 |
+
"learning_rate": 2.407474278044215e-06,
|
| 1142 |
+
"loss": 0.9332,
|
| 1143 |
"step": 765
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.7106598984771574,
|
| 1147 |
+
"grad_norm": 1.4345839815007075,
|
| 1148 |
+
"learning_rate": 2.3388643057628025e-06,
|
| 1149 |
+
"loss": 0.9283,
|
| 1150 |
"step": 770
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.7152745731425935,
|
| 1154 |
+
"grad_norm": 1.4014510611067819,
|
| 1155 |
+
"learning_rate": 2.2709464503713785e-06,
|
| 1156 |
+
"loss": 0.9196,
|
| 1157 |
"step": 775
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.7198892478080295,
|
| 1161 |
+
"grad_norm": 1.4343098995975527,
|
| 1162 |
+
"learning_rate": 2.2037383761682877e-06,
|
| 1163 |
+
"loss": 0.9211,
|
| 1164 |
"step": 780
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.7245039224734656,
|
| 1168 |
+
"grad_norm": 1.342137285764838,
|
| 1169 |
+
"learning_rate": 2.1372575628496662e-06,
|
| 1170 |
+
"loss": 0.9206,
|
| 1171 |
"step": 785
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.7291185971389017,
|
| 1175 |
+
"grad_norm": 1.396553897850799,
|
| 1176 |
+
"learning_rate": 2.071521300963246e-06,
|
| 1177 |
+
"loss": 0.9324,
|
| 1178 |
"step": 790
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.7337332718043378,
|
| 1182 |
+
"grad_norm": 1.4382708247406493,
|
| 1183 |
+
"learning_rate": 2.0065466874113944e-06,
|
| 1184 |
+
"loss": 0.9159,
|
| 1185 |
"step": 795
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.7383479464697739,
|
| 1189 |
+
"grad_norm": 1.3548453547021793,
|
| 1190 |
+
"learning_rate": 1.9423506210044746e-06,
|
| 1191 |
+
"loss": 0.9067,
|
| 1192 |
"step": 800
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.7383479464697739,
|
| 1196 |
+
"eval_loss": 0.9378637671470642,
|
| 1197 |
+
"eval_runtime": 663.1431,
|
| 1198 |
+
"eval_samples_per_second": 23.147,
|
| 1199 |
+
"eval_steps_per_second": 0.181,
|
| 1200 |
"step": 800
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.7429626211352099,
|
| 1204 |
+
"grad_norm": 1.3924317571532645,
|
| 1205 |
+
"learning_rate": 1.8789497980657644e-06,
|
| 1206 |
+
"loss": 0.9387,
|
| 1207 |
"step": 805
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.747577295800646,
|
| 1211 |
+
"grad_norm": 1.4165077817002125,
|
| 1212 |
+
"learning_rate": 1.8163607080890143e-06,
|
| 1213 |
+
"loss": 0.9593,
|
| 1214 |
"step": 810
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.7521919704660821,
|
| 1218 |
+
"grad_norm": 1.4141204258596356,
|
| 1219 |
+
"learning_rate": 1.7545996294498013e-06,
|
| 1220 |
+
"loss": 0.9374,
|
| 1221 |
"step": 815
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.7568066451315182,
|
| 1225 |
+
"grad_norm": 1.3739608338173115,
|
| 1226 |
+
"learning_rate": 1.6936826251718075e-06,
|
| 1227 |
+
"loss": 0.9345,
|
| 1228 |
"step": 820
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.7614213197969543,
|
| 1232 |
+
"grad_norm": 1.4298339467420962,
|
| 1233 |
+
"learning_rate": 1.6336255387490846e-06,
|
| 1234 |
+
"loss": 0.9185,
|
| 1235 |
"step": 825
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.7660359944623903,
|
| 1239 |
+
"grad_norm": 1.4280106425956387,
|
| 1240 |
+
"learning_rate": 1.574443990025436e-06,
|
| 1241 |
+
"loss": 0.9251,
|
| 1242 |
"step": 830
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.7706506691278265,
|
| 1246 |
+
"grad_norm": 1.4182309514153764,
|
| 1247 |
+
"learning_rate": 1.5161533711319454e-06,
|
| 1248 |
+
"loss": 0.9337,
|
| 1249 |
"step": 835
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.7752653437932626,
|
| 1253 |
+
"grad_norm": 1.3937433613798544,
|
| 1254 |
+
"learning_rate": 1.4587688424837538e-06,
|
| 1255 |
+
"loss": 0.9448,
|
| 1256 |
"step": 840
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.7798800184586987,
|
| 1260 |
+
"grad_norm": 1.456151982504558,
|
| 1261 |
+
"learning_rate": 1.4023053288370803e-06,
|
| 1262 |
+
"loss": 0.9129,
|
| 1263 |
"step": 845
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.7844946931241348,
|
| 1267 |
+
"grad_norm": 1.3853091041139631,
|
| 1268 |
+
"learning_rate": 1.3467775154075425e-06,
|
| 1269 |
+
"loss": 0.9213,
|
| 1270 |
"step": 850
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.7891093677895709,
|
| 1274 |
+
"grad_norm": 1.4547278352297517,
|
| 1275 |
+
"learning_rate": 1.2921998440507838e-06,
|
| 1276 |
+
"loss": 0.9211,
|
| 1277 |
"step": 855
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.7937240424550069,
|
| 1281 |
+
"grad_norm": 1.4019092163127755,
|
| 1282 |
+
"learning_rate": 1.2385865095063808e-06,
|
| 1283 |
+
"loss": 0.9189,
|
| 1284 |
"step": 860
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.798338717120443,
|
| 1288 |
+
"grad_norm": 1.3640880739873995,
|
| 1289 |
+
"learning_rate": 1.1859514557060358e-06,
|
| 1290 |
+
"loss": 0.9184,
|
| 1291 |
"step": 865
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.8029533917858791,
|
| 1295 |
+
"grad_norm": 1.4032390358058673,
|
| 1296 |
+
"learning_rate": 1.1343083721469867e-06,
|
| 1297 |
+
"loss": 0.9234,
|
| 1298 |
"step": 870
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.8075680664513152,
|
| 1302 |
+
"grad_norm": 1.365688843862272,
|
| 1303 |
+
"learning_rate": 1.0836706903316052e-06,
|
| 1304 |
+
"loss": 0.9244,
|
| 1305 |
"step": 875
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.8121827411167513,
|
| 1309 |
+
"grad_norm": 1.4146840493944073,
|
| 1310 |
+
"learning_rate": 1.0340515802740781e-06,
|
| 1311 |
+
"loss": 0.9197,
|
| 1312 |
"step": 880
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.8167974157821873,
|
| 1316 |
+
"grad_norm": 1.3895617974460883,
|
| 1317 |
+
"learning_rate": 9.854639470751182e-07,
|
| 1318 |
+
"loss": 0.9194,
|
| 1319 |
"step": 885
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.8214120904476234,
|
| 1323 |
+
"grad_norm": 1.4357287631638345,
|
| 1324 |
+
"learning_rate": 9.379204275655524e-07,
|
| 1325 |
+
"loss": 0.9156,
|
| 1326 |
"step": 890
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.8260267651130595,
|
| 1330 |
+
"grad_norm": 1.3844689117152469,
|
| 1331 |
+
"learning_rate": 8.914333870197022e-07,
|
| 1332 |
+
"loss": 0.9355,
|
| 1333 |
"step": 895
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.8306414397784956,
|
| 1337 |
+
"grad_norm": 1.3689079608163703,
|
| 1338 |
+
"learning_rate": 8.460149159393766e-07,
|
| 1339 |
+
"loss": 0.9221,
|
| 1340 |
"step": 900
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.8306414397784956,
|
| 1344 |
+
"eval_loss": 0.9320199489593506,
|
| 1345 |
+
"eval_runtime": 630.5251,
|
| 1346 |
+
"eval_samples_per_second": 24.345,
|
| 1347 |
+
"eval_steps_per_second": 0.19,
|
| 1348 |
"step": 900
|
| 1349 |
},
|
| 1350 |
{
|
| 1351 |
"epoch": 0.8352561144439317,
|
| 1352 |
+
"grad_norm": 1.373835743305699,
|
| 1353 |
+
"learning_rate": 8.016768269093389e-07,
|
| 1354 |
+
"loss": 0.9225,
|
| 1355 |
"step": 905
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 0.8398707891093677,
|
| 1359 |
+
"grad_norm": 1.3862642218872392,
|
| 1360 |
+
"learning_rate": 7.584306515250461e-07,
|
| 1361 |
+
"loss": 0.9118,
|
| 1362 |
"step": 910
|
| 1363 |
},
|
| 1364 |
{
|
| 1365 |
"epoch": 0.8444854637748038,
|
| 1366 |
+
"grad_norm": 1.427561354442013,
|
| 1367 |
+
"learning_rate": 7.162876373934813e-07,
|
| 1368 |
+
"loss": 0.9257,
|
| 1369 |
"step": 915
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.84910013844024,
|
| 1373 |
+
"grad_norm": 1.3318980900547395,
|
| 1374 |
+
"learning_rate": 6.752587452078297e-07,
|
| 1375 |
+
"loss": 0.9168,
|
| 1376 |
"step": 920
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
"epoch": 0.8537148131056761,
|
| 1380 |
+
"grad_norm": 1.426821326563428,
|
| 1381 |
+
"learning_rate": 6.353546458967957e-07,
|
| 1382 |
+
"loss": 0.9269,
|
| 1383 |
"step": 925
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 0.8583294877711122,
|
| 1387 |
+
"grad_norm": 1.4130164875607825,
|
| 1388 |
+
"learning_rate": 5.965857178492629e-07,
|
| 1389 |
+
"loss": 0.9177,
|
| 1390 |
"step": 930
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 0.8629441624365483,
|
| 1394 |
+
"grad_norm": 1.37338302698056,
|
| 1395 |
+
"learning_rate": 5.589620442150579e-07,
|
| 1396 |
+
"loss": 0.908,
|
| 1397 |
"step": 935
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 0.8675588371019843,
|
| 1401 |
+
"grad_norm": 1.3675248953715173,
|
| 1402 |
+
"learning_rate": 5.224934102824824e-07,
|
| 1403 |
+
"loss": 0.9018,
|
| 1404 |
"step": 940
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 0.8721735117674204,
|
| 1408 |
+
"grad_norm": 1.3837632491876184,
|
| 1409 |
+
"learning_rate": 4.871893009333345e-07,
|
| 1410 |
+
"loss": 0.9266,
|
| 1411 |
"step": 945
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 0.8767881864328565,
|
| 1415 |
+
"grad_norm": 1.3573040010924988,
|
| 1416 |
+
"learning_rate": 4.5305889817603757e-07,
|
| 1417 |
+
"loss": 0.9303,
|
| 1418 |
"step": 950
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
"epoch": 0.8814028610982926,
|
| 1422 |
+
"grad_norm": 1.3246603553725993,
|
| 1423 |
+
"learning_rate": 4.201110787575619e-07,
|
| 1424 |
+
"loss": 0.9003,
|
| 1425 |
"step": 955
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 0.8860175357637287,
|
| 1429 |
+
"grad_norm": 1.3518273420371894,
|
| 1430 |
+
"learning_rate": 3.883544118547289e-07,
|
| 1431 |
+
"loss": 0.9223,
|
| 1432 |
"step": 960
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 0.8906322104291647,
|
| 1436 |
+
"grad_norm": 1.3931507467693585,
|
| 1437 |
+
"learning_rate": 3.5779715684550966e-07,
|
| 1438 |
+
"loss": 0.9233,
|
| 1439 |
"step": 965
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 0.8952468850946008,
|
| 1443 |
+
"grad_norm": 1.4917970308046273,
|
| 1444 |
+
"learning_rate": 3.284472611609024e-07,
|
| 1445 |
+
"loss": 0.9262,
|
| 1446 |
"step": 970
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 0.8998615597600369,
|
| 1450 |
+
"grad_norm": 1.3457919919745878,
|
| 1451 |
+
"learning_rate": 3.00312358217934e-07,
|
| 1452 |
+
"loss": 0.9061,
|
| 1453 |
"step": 975
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 0.904476234425473,
|
| 1457 |
+
"grad_norm": 1.399847886161994,
|
| 1458 |
+
"learning_rate": 2.7339976543434065e-07,
|
| 1459 |
+
"loss": 0.9303,
|
| 1460 |
"step": 980
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 0.9090909090909091,
|
| 1464 |
+
"grad_norm": 1.4141769147196013,
|
| 1465 |
+
"learning_rate": 2.4771648232542524e-07,
|
| 1466 |
+
"loss": 0.9184,
|
| 1467 |
"step": 985
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.9137055837563451,
|
| 1471 |
+
"grad_norm": 1.4105606054413926,
|
| 1472 |
+
"learning_rate": 2.2326918868360636e-07,
|
| 1473 |
+
"loss": 0.923,
|
| 1474 |
"step": 990
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 0.9183202584217812,
|
| 1478 |
+
"grad_norm": 1.3589359885492762,
|
| 1479 |
+
"learning_rate": 2.000642428411087e-07,
|
| 1480 |
+
"loss": 0.9274,
|
| 1481 |
"step": 995
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.9229349330872173,
|
| 1485 |
+
"grad_norm": 1.3455010064913675,
|
| 1486 |
+
"learning_rate": 1.7810768001627455e-07,
|
| 1487 |
+
"loss": 0.9087,
|
| 1488 |
"step": 1000
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 0.9229349330872173,
|
| 1492 |
+
"eval_loss": 0.9292727112770081,
|
| 1493 |
+
"eval_runtime": 1081.8404,
|
| 1494 |
+
"eval_samples_per_second": 14.189,
|
| 1495 |
+
"eval_steps_per_second": 0.111,
|
| 1496 |
"step": 1000
|
| 1497 |
},
|
| 1498 |
{
|
| 1499 |
"epoch": 0.9275496077526535,
|
| 1500 |
+
"grad_norm": 1.3936222044373034,
|
| 1501 |
+
"learning_rate": 1.5740521074389837e-07,
|
| 1502 |
+
"loss": 0.9294,
|
| 1503 |
"step": 1005
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 0.9321642824180896,
|
| 1507 |
+
"grad_norm": 1.6247273551888408,
|
| 1508 |
+
"learning_rate": 1.3796221939001598e-07,
|
| 1509 |
+
"loss": 0.9291,
|
| 1510 |
"step": 1010
|
| 1511 |
},
|
| 1512 |
{
|
| 1513 |
"epoch": 0.9367789570835257,
|
| 1514 |
+
"grad_norm": 1.386776941906468,
|
| 1515 |
+
"learning_rate": 1.1978376275151915e-07,
|
| 1516 |
+
"loss": 0.9116,
|
| 1517 |
"step": 1015
|
| 1518 |
},
|
| 1519 |
{
|
| 1520 |
"epoch": 0.9413936317489617,
|
| 1521 |
+
"grad_norm": 1.3387125746511046,
|
| 1522 |
+
"learning_rate": 1.0287456874096824e-07,
|
| 1523 |
+
"loss": 0.9137,
|
| 1524 |
"step": 1020
|
| 1525 |
},
|
| 1526 |
{
|
| 1527 |
"epoch": 0.9460083064143978,
|
| 1528 |
+
"grad_norm": 1.3731528656657377,
|
| 1529 |
+
"learning_rate": 8.723903515694132e-08,
|
| 1530 |
+
"loss": 0.9208,
|
| 1531 |
"step": 1025
|
| 1532 |
},
|
| 1533 |
{
|
| 1534 |
"epoch": 0.9506229810798339,
|
| 1535 |
+
"grad_norm": 1.3785247362215645,
|
| 1536 |
+
"learning_rate": 7.28812285402386e-08,
|
| 1537 |
+
"loss": 0.9062,
|
| 1538 |
"step": 1030
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
"epoch": 0.95523765574527,
|
| 1542 |
+
"grad_norm": 1.4102762199212637,
|
| 1543 |
+
"learning_rate": 5.98048831162451e-08,
|
| 1544 |
+
"loss": 0.9286,
|
| 1545 |
"step": 1035
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 0.9598523304107061,
|
| 1549 |
+
"grad_norm": 1.3925062428618178,
|
| 1550 |
+
"learning_rate": 4.801339982372144e-08,
|
| 1551 |
+
"loss": 0.9066,
|
| 1552 |
"step": 1040
|
| 1553 |
},
|
| 1554 |
{
|
| 1555 |
"epoch": 0.9644670050761421,
|
| 1556 |
+
"grad_norm": 1.3653301284738,
|
| 1557 |
+
"learning_rate": 3.750984543027358e-08,
|
| 1558 |
+
"loss": 0.9197,
|
| 1559 |
"step": 1045
|
| 1560 |
},
|
| 1561 |
{
|
| 1562 |
"epoch": 0.9690816797415782,
|
| 1563 |
+
"grad_norm": 1.3996435505145486,
|
| 1564 |
+
"learning_rate": 2.8296951734740896e-08,
|
| 1565 |
+
"loss": 0.9303,
|
| 1566 |
"step": 1050
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 0.9736963544070143,
|
| 1570 |
+
"grad_norm": 1.367940583530902,
|
| 1571 |
+
"learning_rate": 2.0377114856700575e-08,
|
| 1572 |
+
"loss": 0.9286,
|
| 1573 |
"step": 1055
|
| 1574 |
},
|
| 1575 |
{
|
| 1576 |
"epoch": 0.9783110290724504,
|
| 1577 |
+
"grad_norm": 1.4028823114794313,
|
| 1578 |
+
"learning_rate": 1.3752394613274488e-08,
|
| 1579 |
+
"loss": 0.9169,
|
| 1580 |
"step": 1060
|
| 1581 |
},
|
| 1582 |
{
|
| 1583 |
"epoch": 0.9829257037378865,
|
| 1584 |
+
"grad_norm": 1.354080160910544,
|
| 1585 |
+
"learning_rate": 8.424513983408267e-09,
|
| 1586 |
+
"loss": 0.9156,
|
| 1587 |
"step": 1065
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
"epoch": 0.9875403784033225,
|
| 1591 |
+
"grad_norm": 1.3348691278230316,
|
| 1592 |
+
"learning_rate": 4.3948586597525325e-09,
|
| 1593 |
+
"loss": 0.9243,
|
| 1594 |
"step": 1070
|
| 1595 |
},
|
| 1596 |
{
|
| 1597 |
"epoch": 0.9921550530687586,
|
| 1598 |
+
"grad_norm": 1.3552426228812648,
|
| 1599 |
+
"learning_rate": 1.664476688265082e-09,
|
| 1600 |
+
"loss": 0.9195,
|
| 1601 |
"step": 1075
|
| 1602 |
},
|
| 1603 |
{
|
| 1604 |
"epoch": 0.9967697277341947,
|
| 1605 |
+
"grad_norm": 1.3989954260059947,
|
| 1606 |
+
"learning_rate": 2.3407819563503463e-10,
|
| 1607 |
+
"loss": 0.9093,
|
| 1608 |
"step": 1080
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
"epoch": 0.9995385325334564,
|
| 1612 |
"step": 1083,
|
| 1613 |
"total_flos": 453306954547200.0,
|
| 1614 |
+
"train_loss": 0.9547446678880179,
|
| 1615 |
+
"train_runtime": 38927.5133,
|
| 1616 |
+
"train_samples_per_second": 3.563,
|
| 1617 |
+
"train_steps_per_second": 0.028
|
| 1618 |
}
|
| 1619 |
],
|
| 1620 |
"logging_steps": 5,
|