Training in progress, epoch 4
Browse files- README.md +188 -371
- adapter_model.safetensors +1 -1
- all_results.json +11 -11
- eval_results.json +6 -6
- train_results.json +6 -6
README.md
CHANGED
|
@@ -5,18 +5,18 @@ base_model: gpt2
|
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
model-index:
|
| 8 |
-
- name:
|
| 9 |
results: []
|
| 10 |
---
|
| 11 |
|
| 12 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
|
| 15 |
-
#
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss: 0.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
@@ -39,8 +39,8 @@ The following hyperparameters were used during training:
|
|
| 39 |
- train_batch_size: 16
|
| 40 |
- eval_batch_size: 16
|
| 41 |
- seed: 42
|
| 42 |
-
- gradient_accumulation_steps:
|
| 43 |
-
- total_train_batch_size:
|
| 44 |
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 45 |
- lr_scheduler_type: cosine
|
| 46 |
- lr_scheduler_warmup_ratio: 0.03
|
|
@@ -50,372 +50,189 @@ The following hyperparameters were used during training:
|
|
| 50 |
|
| 51 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 52 |
|:-------------:|:------:|:----:|:---------------:|
|
| 53 |
-
|
|
| 54 |
-
|
|
| 55 |
-
|
|
| 56 |
-
|
|
| 57 |
-
|
|
| 58 |
-
| 2.
|
| 59 |
-
| 2.
|
| 60 |
-
| 2.
|
| 61 |
-
| 2.
|
| 62 |
-
| 2.
|
| 63 |
-
| 2.
|
| 64 |
-
| 2.
|
| 65 |
-
|
|
| 66 |
-
| 1.
|
| 67 |
-
| 1.
|
| 68 |
-
| 1.
|
| 69 |
-
| 1.
|
| 70 |
-
| 1.
|
| 71 |
-
| 1.
|
| 72 |
-
| 1.
|
| 73 |
-
|
|
| 74 |
-
|
|
| 75 |
-
|
|
| 76 |
-
|
|
| 77 |
-
|
|
| 78 |
-
|
|
| 79 |
-
|
|
| 80 |
-
|
|
| 81 |
-
|
|
| 82 |
-
|
|
| 83 |
-
|
|
| 84 |
-
| 0.
|
| 85 |
-
| 0.
|
| 86 |
-
| 0.
|
| 87 |
-
| 0.
|
| 88 |
-
| 0.
|
| 89 |
-
| 0.
|
| 90 |
-
| 0.
|
| 91 |
-
| 0.
|
| 92 |
-
| 0.
|
| 93 |
-
| 0.
|
| 94 |
-
| 0.
|
| 95 |
-
| 0.
|
| 96 |
-
| 0.
|
| 97 |
-
| 0.
|
| 98 |
-
| 0.
|
| 99 |
-
| 0.
|
| 100 |
-
| 0.
|
| 101 |
-
| 0.
|
| 102 |
-
| 0.
|
| 103 |
-
| 0.
|
| 104 |
-
| 0.
|
| 105 |
-
| 0.
|
| 106 |
-
| 0.
|
| 107 |
-
| 0.
|
| 108 |
-
| 0.
|
| 109 |
-
| 0.
|
| 110 |
-
| 0.
|
| 111 |
-
| 0.
|
| 112 |
-
| 0.
|
| 113 |
-
| 0.
|
| 114 |
-
| 0.
|
| 115 |
-
| 0.
|
| 116 |
-
| 0.
|
| 117 |
-
| 0.
|
| 118 |
-
| 0.
|
| 119 |
-
| 0.
|
| 120 |
-
| 0.
|
| 121 |
-
| 0.
|
| 122 |
-
| 0.
|
| 123 |
-
| 0.
|
| 124 |
-
| 0.
|
| 125 |
-
| 0.
|
| 126 |
-
| 0.
|
| 127 |
-
| 0.
|
| 128 |
-
| 0.
|
| 129 |
-
| 0.
|
| 130 |
-
| 0.
|
| 131 |
-
| 0.
|
| 132 |
-
| 0.
|
| 133 |
-
| 0.
|
| 134 |
-
| 0.
|
| 135 |
-
| 0.
|
| 136 |
-
| 0.
|
| 137 |
-
| 0.
|
| 138 |
-
| 0.
|
| 139 |
-
| 0.
|
| 140 |
-
| 0.
|
| 141 |
-
| 0.
|
| 142 |
-
| 0.
|
| 143 |
-
| 0.
|
| 144 |
-
| 0.
|
| 145 |
-
| 0.
|
| 146 |
-
| 0.
|
| 147 |
-
| 0.
|
| 148 |
-
| 0.
|
| 149 |
-
| 0.
|
| 150 |
-
| 0.
|
| 151 |
-
| 0.
|
| 152 |
-
| 0.
|
| 153 |
-
| 0.
|
| 154 |
-
| 0.
|
| 155 |
-
| 0.
|
| 156 |
-
| 0.
|
| 157 |
-
| 0.
|
| 158 |
-
| 0.
|
| 159 |
-
| 0.
|
| 160 |
-
| 0.
|
| 161 |
-
| 0.
|
| 162 |
-
| 0.
|
| 163 |
-
| 0.
|
| 164 |
-
| 0.
|
| 165 |
-
| 0.
|
| 166 |
-
| 0.
|
| 167 |
-
| 0.
|
| 168 |
-
| 0.
|
| 169 |
-
| 0.
|
| 170 |
-
| 0.
|
| 171 |
-
| 0.
|
| 172 |
-
| 0.
|
| 173 |
-
| 0.
|
| 174 |
-
| 0.
|
| 175 |
-
| 0.
|
| 176 |
-
| 0.
|
| 177 |
-
| 0.
|
| 178 |
-
| 0.
|
| 179 |
-
| 0.
|
| 180 |
-
| 0.
|
| 181 |
-
| 0.
|
| 182 |
-
| 0.
|
| 183 |
-
| 0.
|
| 184 |
-
| 0.
|
| 185 |
-
| 0.
|
| 186 |
-
| 0.
|
| 187 |
-
| 0.
|
| 188 |
-
| 0.
|
| 189 |
-
| 0.
|
| 190 |
-
| 0.
|
| 191 |
-
| 0.
|
| 192 |
-
| 0.
|
| 193 |
-
| 0.
|
| 194 |
-
| 0.
|
| 195 |
-
| 0.
|
| 196 |
-
| 0.
|
| 197 |
-
| 0.
|
| 198 |
-
| 0.
|
| 199 |
-
| 0.
|
| 200 |
-
| 0.
|
| 201 |
-
| 0.
|
| 202 |
-
| 0.
|
| 203 |
-
| 0.
|
| 204 |
-
| 0.
|
| 205 |
-
| 0.
|
| 206 |
-
| 0.
|
| 207 |
-
| 0.
|
| 208 |
-
| 0.
|
| 209 |
-
| 0.
|
| 210 |
-
| 0.
|
| 211 |
-
| 0.
|
| 212 |
-
| 0.
|
| 213 |
-
| 0.
|
| 214 |
-
| 0.
|
| 215 |
-
| 0.
|
| 216 |
-
| 0.
|
| 217 |
-
| 0.
|
| 218 |
-
| 0.
|
| 219 |
-
| 0.
|
| 220 |
-
| 0.
|
| 221 |
-
| 0.
|
| 222 |
-
| 0.
|
| 223 |
-
| 0.
|
| 224 |
-
| 0.
|
| 225 |
-
| 0.
|
| 226 |
-
| 0.
|
| 227 |
-
| 0.
|
| 228 |
-
| 0.
|
| 229 |
-
| 0.
|
| 230 |
-
| 0.
|
| 231 |
-
| 0.
|
| 232 |
-
| 0.
|
| 233 |
-
| 0.
|
| 234 |
-
| 0.
|
| 235 |
-
| 0.
|
| 236 |
-
| 0.445 | 1.5063 | 3680 | 0.3999 |
|
| 237 |
-
| 0.4389 | 1.5145 | 3700 | 0.4009 |
|
| 238 |
-
| 0.4316 | 1.5227 | 3720 | 0.3988 |
|
| 239 |
-
| 0.4363 | 1.5309 | 3740 | 0.3994 |
|
| 240 |
-
| 0.4384 | 1.5391 | 3760 | 0.3995 |
|
| 241 |
-
| 0.4355 | 1.5473 | 3780 | 0.4006 |
|
| 242 |
-
| 0.436 | 1.5555 | 3800 | 0.3983 |
|
| 243 |
-
| 0.4384 | 1.5637 | 3820 | 0.3981 |
|
| 244 |
-
| 0.4394 | 1.5718 | 3840 | 0.3985 |
|
| 245 |
-
| 0.4392 | 1.5800 | 3860 | 0.3978 |
|
| 246 |
-
| 0.4456 | 1.5882 | 3880 | 0.3991 |
|
| 247 |
-
| 0.4359 | 1.5964 | 3900 | 0.3984 |
|
| 248 |
-
| 0.4328 | 1.6046 | 3920 | 0.4004 |
|
| 249 |
-
| 0.4272 | 1.6128 | 3940 | 0.3992 |
|
| 250 |
-
| 0.4352 | 1.6210 | 3960 | 0.3993 |
|
| 251 |
-
| 0.4262 | 1.6291 | 3980 | 0.3994 |
|
| 252 |
-
| 0.4406 | 1.6373 | 4000 | 0.3979 |
|
| 253 |
-
| 0.4291 | 1.6455 | 4020 | 0.3991 |
|
| 254 |
-
| 0.4262 | 1.6537 | 4040 | 0.3975 |
|
| 255 |
-
| 0.4337 | 1.6619 | 4060 | 0.3978 |
|
| 256 |
-
| 0.4404 | 1.6701 | 4080 | 0.3964 |
|
| 257 |
-
| 0.4408 | 1.6783 | 4100 | 0.3983 |
|
| 258 |
-
| 0.4378 | 1.6865 | 4120 | 0.3977 |
|
| 259 |
-
| 0.4322 | 1.6946 | 4140 | 0.3973 |
|
| 260 |
-
| 0.4343 | 1.7028 | 4160 | 0.3970 |
|
| 261 |
-
| 0.43 | 1.7110 | 4180 | 0.3961 |
|
| 262 |
-
| 0.4343 | 1.7192 | 4200 | 0.3958 |
|
| 263 |
-
| 0.4308 | 1.7274 | 4220 | 0.3965 |
|
| 264 |
-
| 0.4355 | 1.7356 | 4240 | 0.3952 |
|
| 265 |
-
| 0.4371 | 1.7438 | 4260 | 0.3966 |
|
| 266 |
-
| 0.4342 | 1.7519 | 4280 | 0.3956 |
|
| 267 |
-
| 0.4364 | 1.7601 | 4300 | 0.3962 |
|
| 268 |
-
| 0.434 | 1.7683 | 4320 | 0.3953 |
|
| 269 |
-
| 0.4335 | 1.7765 | 4340 | 0.3965 |
|
| 270 |
-
| 0.4317 | 1.7847 | 4360 | 0.3953 |
|
| 271 |
-
| 0.4298 | 1.7929 | 4380 | 0.3954 |
|
| 272 |
-
| 0.4307 | 1.8011 | 4400 | 0.3942 |
|
| 273 |
-
| 0.4345 | 1.8093 | 4420 | 0.3952 |
|
| 274 |
-
| 0.433 | 1.8174 | 4440 | 0.3943 |
|
| 275 |
-
| 0.4261 | 1.8256 | 4460 | 0.3955 |
|
| 276 |
-
| 0.4338 | 1.8338 | 4480 | 0.3950 |
|
| 277 |
-
| 0.4263 | 1.8420 | 4500 | 0.3944 |
|
| 278 |
-
| 0.4263 | 1.8502 | 4520 | 0.3939 |
|
| 279 |
-
| 0.436 | 1.8584 | 4540 | 0.3943 |
|
| 280 |
-
| 0.432 | 1.8666 | 4560 | 0.3946 |
|
| 281 |
-
| 0.4302 | 1.8747 | 4580 | 0.3942 |
|
| 282 |
-
| 0.4333 | 1.8829 | 4600 | 0.3936 |
|
| 283 |
-
| 0.4316 | 1.8911 | 4620 | 0.3936 |
|
| 284 |
-
| 0.4294 | 1.8993 | 4640 | 0.3938 |
|
| 285 |
-
| 0.4265 | 1.9075 | 4660 | 0.3936 |
|
| 286 |
-
| 0.4294 | 1.9157 | 4680 | 0.3943 |
|
| 287 |
-
| 0.4319 | 1.9239 | 4700 | 0.3942 |
|
| 288 |
-
| 0.4391 | 1.9321 | 4720 | 0.3933 |
|
| 289 |
-
| 0.4243 | 1.9402 | 4740 | 0.3944 |
|
| 290 |
-
| 0.4325 | 1.9484 | 4760 | 0.3930 |
|
| 291 |
-
| 0.4343 | 1.9566 | 4780 | 0.3924 |
|
| 292 |
-
| 0.4287 | 1.9648 | 4800 | 0.3938 |
|
| 293 |
-
| 0.4322 | 1.9730 | 4820 | 0.3933 |
|
| 294 |
-
| 0.4283 | 1.9812 | 4840 | 0.3926 |
|
| 295 |
-
| 0.4309 | 1.9894 | 4860 | 0.3935 |
|
| 296 |
-
| 0.4238 | 1.9975 | 4880 | 0.3922 |
|
| 297 |
-
| 0.4217 | 2.0057 | 4900 | 0.3925 |
|
| 298 |
-
| 0.425 | 2.0139 | 4920 | 0.3926 |
|
| 299 |
-
| 0.4389 | 2.0221 | 4940 | 0.3925 |
|
| 300 |
-
| 0.4346 | 2.0303 | 4960 | 0.3920 |
|
| 301 |
-
| 0.4254 | 2.0385 | 4980 | 0.3931 |
|
| 302 |
-
| 0.4223 | 2.0467 | 5000 | 0.3919 |
|
| 303 |
-
| 0.4268 | 2.0549 | 5020 | 0.3930 |
|
| 304 |
-
| 0.4228 | 2.0630 | 5040 | 0.3929 |
|
| 305 |
-
| 0.4325 | 2.0712 | 5060 | 0.3928 |
|
| 306 |
-
| 0.4255 | 2.0794 | 5080 | 0.3928 |
|
| 307 |
-
| 0.4305 | 2.0876 | 5100 | 0.3922 |
|
| 308 |
-
| 0.4333 | 2.0958 | 5120 | 0.3919 |
|
| 309 |
-
| 0.4332 | 2.1040 | 5140 | 0.3927 |
|
| 310 |
-
| 0.4261 | 2.1122 | 5160 | 0.3929 |
|
| 311 |
-
| 0.429 | 2.1203 | 5180 | 0.3916 |
|
| 312 |
-
| 0.4274 | 2.1285 | 5200 | 0.3921 |
|
| 313 |
-
| 0.4277 | 2.1367 | 5220 | 0.3928 |
|
| 314 |
-
| 0.4356 | 2.1449 | 5240 | 0.3913 |
|
| 315 |
-
| 0.4268 | 2.1531 | 5260 | 0.3921 |
|
| 316 |
-
| 0.4297 | 2.1613 | 5280 | 0.3921 |
|
| 317 |
-
| 0.4272 | 2.1695 | 5300 | 0.3915 |
|
| 318 |
-
| 0.4337 | 2.1777 | 5320 | 0.3922 |
|
| 319 |
-
| 0.4312 | 2.1858 | 5340 | 0.3911 |
|
| 320 |
-
| 0.426 | 2.1940 | 5360 | 0.3917 |
|
| 321 |
-
| 0.4305 | 2.2022 | 5380 | 0.3925 |
|
| 322 |
-
| 0.4373 | 2.2104 | 5400 | 0.3919 |
|
| 323 |
-
| 0.4319 | 2.2186 | 5420 | 0.3914 |
|
| 324 |
-
| 0.43 | 2.2268 | 5440 | 0.3921 |
|
| 325 |
-
| 0.4307 | 2.2350 | 5460 | 0.3910 |
|
| 326 |
-
| 0.4352 | 2.2431 | 5480 | 0.3912 |
|
| 327 |
-
| 0.4323 | 2.2513 | 5500 | 0.3907 |
|
| 328 |
-
| 0.4255 | 2.2595 | 5520 | 0.3905 |
|
| 329 |
-
| 0.4286 | 2.2677 | 5540 | 0.3913 |
|
| 330 |
-
| 0.4271 | 2.2759 | 5560 | 0.3916 |
|
| 331 |
-
| 0.4319 | 2.2841 | 5580 | 0.3915 |
|
| 332 |
-
| 0.4175 | 2.2923 | 5600 | 0.3911 |
|
| 333 |
-
| 0.424 | 2.3005 | 5620 | 0.3914 |
|
| 334 |
-
| 0.4365 | 2.3086 | 5640 | 0.3907 |
|
| 335 |
-
| 0.4322 | 2.3168 | 5660 | 0.3906 |
|
| 336 |
-
| 0.4227 | 2.3250 | 5680 | 0.3910 |
|
| 337 |
-
| 0.4308 | 2.3332 | 5700 | 0.3909 |
|
| 338 |
-
| 0.4268 | 2.3414 | 5720 | 0.3910 |
|
| 339 |
-
| 0.4352 | 2.3496 | 5740 | 0.3911 |
|
| 340 |
-
| 0.4274 | 2.3578 | 5760 | 0.3898 |
|
| 341 |
-
| 0.4255 | 2.3659 | 5780 | 0.3901 |
|
| 342 |
-
| 0.4277 | 2.3741 | 5800 | 0.3903 |
|
| 343 |
-
| 0.4209 | 2.3823 | 5820 | 0.3905 |
|
| 344 |
-
| 0.4221 | 2.3905 | 5840 | 0.3911 |
|
| 345 |
-
| 0.4247 | 2.3987 | 5860 | 0.3911 |
|
| 346 |
-
| 0.4263 | 2.4069 | 5880 | 0.3910 |
|
| 347 |
-
| 0.4284 | 2.4151 | 5900 | 0.3912 |
|
| 348 |
-
| 0.4251 | 2.4233 | 5920 | 0.3910 |
|
| 349 |
-
| 0.4275 | 2.4314 | 5940 | 0.3908 |
|
| 350 |
-
| 0.4271 | 2.4396 | 5960 | 0.3904 |
|
| 351 |
-
| 0.4333 | 2.4478 | 5980 | 0.3904 |
|
| 352 |
-
| 0.4237 | 2.4560 | 6000 | 0.3903 |
|
| 353 |
-
| 0.4351 | 2.4642 | 6020 | 0.3903 |
|
| 354 |
-
| 0.4313 | 2.4724 | 6040 | 0.3902 |
|
| 355 |
-
| 0.4243 | 2.4806 | 6060 | 0.3910 |
|
| 356 |
-
| 0.4289 | 2.4887 | 6080 | 0.3907 |
|
| 357 |
-
| 0.4299 | 2.4969 | 6100 | 0.3909 |
|
| 358 |
-
| 0.428 | 2.5051 | 6120 | 0.3903 |
|
| 359 |
-
| 0.4202 | 2.5133 | 6140 | 0.3902 |
|
| 360 |
-
| 0.4291 | 2.5215 | 6160 | 0.3899 |
|
| 361 |
-
| 0.4344 | 2.5297 | 6180 | 0.3899 |
|
| 362 |
-
| 0.4256 | 2.5379 | 6200 | 0.3902 |
|
| 363 |
-
| 0.4227 | 2.5460 | 6220 | 0.3904 |
|
| 364 |
-
| 0.43 | 2.5542 | 6240 | 0.3907 |
|
| 365 |
-
| 0.4252 | 2.5624 | 6260 | 0.3900 |
|
| 366 |
-
| 0.4224 | 2.5706 | 6280 | 0.3909 |
|
| 367 |
-
| 0.4207 | 2.5788 | 6300 | 0.3909 |
|
| 368 |
-
| 0.4265 | 2.5870 | 6320 | 0.3906 |
|
| 369 |
-
| 0.4341 | 2.5952 | 6340 | 0.3907 |
|
| 370 |
-
| 0.4228 | 2.6034 | 6360 | 0.3903 |
|
| 371 |
-
| 0.4196 | 2.6115 | 6380 | 0.3904 |
|
| 372 |
-
| 0.4216 | 2.6197 | 6400 | 0.3897 |
|
| 373 |
-
| 0.4339 | 2.6279 | 6420 | 0.3904 |
|
| 374 |
-
| 0.4255 | 2.6361 | 6440 | 0.3903 |
|
| 375 |
-
| 0.4261 | 2.6443 | 6460 | 0.3905 |
|
| 376 |
-
| 0.43 | 2.6525 | 6480 | 0.3906 |
|
| 377 |
-
| 0.4265 | 2.6607 | 6500 | 0.3907 |
|
| 378 |
-
| 0.4279 | 2.6688 | 6520 | 0.3904 |
|
| 379 |
-
| 0.4298 | 2.6770 | 6540 | 0.3901 |
|
| 380 |
-
| 0.4312 | 2.6852 | 6560 | 0.3901 |
|
| 381 |
-
| 0.4199 | 2.6934 | 6580 | 0.3898 |
|
| 382 |
-
| 0.4288 | 2.7016 | 6600 | 0.3902 |
|
| 383 |
-
| 0.4325 | 2.7098 | 6620 | 0.3905 |
|
| 384 |
-
| 0.4246 | 2.7180 | 6640 | 0.3903 |
|
| 385 |
-
| 0.4281 | 2.7262 | 6660 | 0.3899 |
|
| 386 |
-
| 0.4296 | 2.7343 | 6680 | 0.3903 |
|
| 387 |
-
| 0.4247 | 2.7425 | 6700 | 0.3898 |
|
| 388 |
-
| 0.4252 | 2.7507 | 6720 | 0.3905 |
|
| 389 |
-
| 0.4255 | 2.7589 | 6740 | 0.3904 |
|
| 390 |
-
| 0.4282 | 2.7671 | 6760 | 0.3902 |
|
| 391 |
-
| 0.4225 | 2.7753 | 6780 | 0.3900 |
|
| 392 |
-
| 0.4251 | 2.7835 | 6800 | 0.3900 |
|
| 393 |
-
| 0.4201 | 2.7916 | 6820 | 0.3903 |
|
| 394 |
-
| 0.4252 | 2.7998 | 6840 | 0.3905 |
|
| 395 |
-
| 0.427 | 2.8080 | 6860 | 0.3907 |
|
| 396 |
-
| 0.428 | 2.8162 | 6880 | 0.3907 |
|
| 397 |
-
| 0.437 | 2.8244 | 6900 | 0.3900 |
|
| 398 |
-
| 0.4257 | 2.8326 | 6920 | 0.3901 |
|
| 399 |
-
| 0.4239 | 2.8408 | 6940 | 0.3905 |
|
| 400 |
-
| 0.4276 | 2.8490 | 6960 | 0.3902 |
|
| 401 |
-
| 0.4274 | 2.8571 | 6980 | 0.3897 |
|
| 402 |
-
| 0.4327 | 2.8653 | 7000 | 0.3902 |
|
| 403 |
-
| 0.4313 | 2.8735 | 7020 | 0.3896 |
|
| 404 |
-
| 0.4277 | 2.8817 | 7040 | 0.3904 |
|
| 405 |
-
| 0.4289 | 2.8899 | 7060 | 0.3904 |
|
| 406 |
-
| 0.4321 | 2.8981 | 7080 | 0.3900 |
|
| 407 |
-
| 0.4232 | 2.9063 | 7100 | 0.3902 |
|
| 408 |
-
| 0.4274 | 2.9144 | 7120 | 0.3901 |
|
| 409 |
-
| 0.4339 | 2.9226 | 7140 | 0.3901 |
|
| 410 |
-
| 0.4226 | 2.9308 | 7160 | 0.3904 |
|
| 411 |
-
| 0.4184 | 2.9390 | 7180 | 0.3902 |
|
| 412 |
-
| 0.4242 | 2.9472 | 7200 | 0.3901 |
|
| 413 |
-
| 0.4259 | 2.9554 | 7220 | 0.3902 |
|
| 414 |
-
| 0.4297 | 2.9636 | 7240 | 0.3897 |
|
| 415 |
-
| 0.4268 | 2.9718 | 7260 | 0.3900 |
|
| 416 |
-
| 0.4281 | 2.9799 | 7280 | 0.3900 |
|
| 417 |
-
| 0.4234 | 2.9881 | 7300 | 0.3901 |
|
| 418 |
-
| 0.4196 | 2.9963 | 7320 | 0.3900 |
|
| 419 |
|
| 420 |
|
| 421 |
### Framework versions
|
|
|
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
model-index:
|
| 8 |
+
- name: Se124M100KInfPrompt_WT_EOS
|
| 9 |
results: []
|
| 10 |
---
|
| 11 |
|
| 12 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
|
| 15 |
+
# Se124M100KInfPrompt_WT_EOS
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.7591
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 39 |
- train_batch_size: 16
|
| 40 |
- eval_batch_size: 16
|
| 41 |
- seed: 42
|
| 42 |
+
- gradient_accumulation_steps: 4
|
| 43 |
+
- total_train_batch_size: 64
|
| 44 |
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 45 |
- lr_scheduler_type: cosine
|
| 46 |
- lr_scheduler_warmup_ratio: 0.03
|
|
|
|
| 50 |
|
| 51 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 52 |
|:-------------:|:------:|:----:|:---------------:|
|
| 53 |
+
| 3.2702 | 0.0164 | 20 | 3.0061 |
|
| 54 |
+
| 3.2625 | 0.0327 | 40 | 3.0018 |
|
| 55 |
+
| 3.2351 | 0.0491 | 60 | 2.9647 |
|
| 56 |
+
| 3.1987 | 0.0655 | 80 | 2.9085 |
|
| 57 |
+
| 3.0557 | 0.0819 | 100 | 2.8161 |
|
| 58 |
+
| 2.939 | 0.0982 | 120 | 2.6976 |
|
| 59 |
+
| 2.8112 | 0.1146 | 140 | 2.5745 |
|
| 60 |
+
| 2.6056 | 0.1310 | 160 | 2.4240 |
|
| 61 |
+
| 2.4701 | 0.1474 | 180 | 2.2716 |
|
| 62 |
+
| 2.3312 | 0.1637 | 200 | 2.1298 |
|
| 63 |
+
| 2.1946 | 0.1801 | 220 | 1.9889 |
|
| 64 |
+
| 2.0407 | 0.1965 | 240 | 1.8554 |
|
| 65 |
+
| 1.9256 | 0.2129 | 260 | 1.7339 |
|
| 66 |
+
| 1.803 | 0.2292 | 280 | 1.6261 |
|
| 67 |
+
| 1.6977 | 0.2456 | 300 | 1.5231 |
|
| 68 |
+
| 1.6354 | 0.2620 | 320 | 1.4378 |
|
| 69 |
+
| 1.5545 | 0.2783 | 340 | 1.3478 |
|
| 70 |
+
| 1.4609 | 0.2947 | 360 | 1.2741 |
|
| 71 |
+
| 1.3616 | 0.3111 | 380 | 1.2038 |
|
| 72 |
+
| 1.3137 | 0.3275 | 400 | 1.1477 |
|
| 73 |
+
| 1.2682 | 0.3438 | 420 | 1.1003 |
|
| 74 |
+
| 1.2262 | 0.3602 | 440 | 1.0645 |
|
| 75 |
+
| 1.2021 | 0.3766 | 460 | 1.0357 |
|
| 76 |
+
| 1.1659 | 0.3930 | 480 | 1.0129 |
|
| 77 |
+
| 1.1355 | 0.4093 | 500 | 0.9911 |
|
| 78 |
+
| 1.1085 | 0.4257 | 520 | 0.9732 |
|
| 79 |
+
| 1.0979 | 0.4421 | 540 | 0.9571 |
|
| 80 |
+
| 1.0775 | 0.4585 | 560 | 0.9420 |
|
| 81 |
+
| 1.0659 | 0.4748 | 580 | 0.9295 |
|
| 82 |
+
| 1.0356 | 0.4912 | 600 | 0.9158 |
|
| 83 |
+
| 1.0184 | 0.5076 | 620 | 0.9056 |
|
| 84 |
+
| 0.9979 | 0.5239 | 640 | 0.8965 |
|
| 85 |
+
| 0.988 | 0.5403 | 660 | 0.8879 |
|
| 86 |
+
| 0.9913 | 0.5567 | 680 | 0.8794 |
|
| 87 |
+
| 0.9681 | 0.5731 | 700 | 0.8734 |
|
| 88 |
+
| 0.959 | 0.5894 | 720 | 0.8670 |
|
| 89 |
+
| 0.9389 | 0.6058 | 740 | 0.8591 |
|
| 90 |
+
| 0.9363 | 0.6222 | 760 | 0.8556 |
|
| 91 |
+
| 0.9274 | 0.6386 | 780 | 0.8490 |
|
| 92 |
+
| 0.9265 | 0.6549 | 800 | 0.8472 |
|
| 93 |
+
| 0.9265 | 0.6713 | 820 | 0.8432 |
|
| 94 |
+
| 0.9033 | 0.6877 | 840 | 0.8404 |
|
| 95 |
+
| 0.912 | 0.7041 | 860 | 0.8372 |
|
| 96 |
+
| 0.9067 | 0.7204 | 880 | 0.8337 |
|
| 97 |
+
| 0.896 | 0.7368 | 900 | 0.8309 |
|
| 98 |
+
| 0.9061 | 0.7532 | 920 | 0.8280 |
|
| 99 |
+
| 0.8966 | 0.7695 | 940 | 0.8247 |
|
| 100 |
+
| 0.8871 | 0.7859 | 960 | 0.8233 |
|
| 101 |
+
| 0.8844 | 0.8023 | 980 | 0.8209 |
|
| 102 |
+
| 0.8903 | 0.8187 | 1000 | 0.8188 |
|
| 103 |
+
| 0.8938 | 0.8350 | 1020 | 0.8166 |
|
| 104 |
+
| 0.884 | 0.8514 | 1040 | 0.8143 |
|
| 105 |
+
| 0.8614 | 0.8678 | 1060 | 0.8128 |
|
| 106 |
+
| 0.8716 | 0.8842 | 1080 | 0.8100 |
|
| 107 |
+
| 0.8719 | 0.9005 | 1100 | 0.8082 |
|
| 108 |
+
| 0.8685 | 0.9169 | 1120 | 0.8063 |
|
| 109 |
+
| 0.8784 | 0.9333 | 1140 | 0.8049 |
|
| 110 |
+
| 0.8534 | 0.9497 | 1160 | 0.8037 |
|
| 111 |
+
| 0.8556 | 0.9660 | 1180 | 0.8021 |
|
| 112 |
+
| 0.862 | 0.9824 | 1200 | 0.8007 |
|
| 113 |
+
| 0.852 | 0.9988 | 1220 | 0.7989 |
|
| 114 |
+
| 0.857 | 1.0147 | 1240 | 0.7969 |
|
| 115 |
+
| 0.8562 | 1.0311 | 1260 | 0.7957 |
|
| 116 |
+
| 0.8508 | 1.0475 | 1280 | 0.7965 |
|
| 117 |
+
| 0.845 | 1.0639 | 1300 | 0.7952 |
|
| 118 |
+
| 0.839 | 1.0802 | 1320 | 0.7951 |
|
| 119 |
+
| 0.8472 | 1.0966 | 1340 | 0.7919 |
|
| 120 |
+
| 0.8425 | 1.1130 | 1360 | 0.7899 |
|
| 121 |
+
| 0.8285 | 1.1293 | 1380 | 0.7908 |
|
| 122 |
+
| 0.8287 | 1.1457 | 1400 | 0.7894 |
|
| 123 |
+
| 0.8371 | 1.1621 | 1420 | 0.7880 |
|
| 124 |
+
| 0.8281 | 1.1785 | 1440 | 0.7873 |
|
| 125 |
+
| 0.8468 | 1.1948 | 1460 | 0.7863 |
|
| 126 |
+
| 0.833 | 1.2112 | 1480 | 0.7851 |
|
| 127 |
+
| 0.8298 | 1.2276 | 1500 | 0.7854 |
|
| 128 |
+
| 0.8321 | 1.2440 | 1520 | 0.7841 |
|
| 129 |
+
| 0.823 | 1.2603 | 1540 | 0.7835 |
|
| 130 |
+
| 0.828 | 1.2767 | 1560 | 0.7832 |
|
| 131 |
+
| 0.8353 | 1.2931 | 1580 | 0.7830 |
|
| 132 |
+
| 0.8277 | 1.3095 | 1600 | 0.7814 |
|
| 133 |
+
| 0.8172 | 1.3258 | 1620 | 0.7815 |
|
| 134 |
+
| 0.8192 | 1.3422 | 1640 | 0.7814 |
|
| 135 |
+
| 0.8223 | 1.3586 | 1660 | 0.7800 |
|
| 136 |
+
| 0.8272 | 1.3749 | 1680 | 0.7789 |
|
| 137 |
+
| 0.8202 | 1.3913 | 1700 | 0.7785 |
|
| 138 |
+
| 0.8185 | 1.4077 | 1720 | 0.7769 |
|
| 139 |
+
| 0.8325 | 1.4241 | 1740 | 0.7778 |
|
| 140 |
+
| 0.8128 | 1.4404 | 1760 | 0.7760 |
|
| 141 |
+
| 0.8153 | 1.4568 | 1780 | 0.7767 |
|
| 142 |
+
| 0.8212 | 1.4732 | 1800 | 0.7772 |
|
| 143 |
+
| 0.8049 | 1.4896 | 1820 | 0.7753 |
|
| 144 |
+
| 0.825 | 1.5059 | 1840 | 0.7744 |
|
| 145 |
+
| 0.8175 | 1.5223 | 1860 | 0.7746 |
|
| 146 |
+
| 0.8218 | 1.5387 | 1880 | 0.7731 |
|
| 147 |
+
| 0.814 | 1.5551 | 1900 | 0.7730 |
|
| 148 |
+
| 0.8172 | 1.5714 | 1920 | 0.7723 |
|
| 149 |
+
| 0.8305 | 1.5878 | 1940 | 0.7717 |
|
| 150 |
+
| 0.8145 | 1.6042 | 1960 | 0.7728 |
|
| 151 |
+
| 0.7996 | 1.6205 | 1980 | 0.7710 |
|
| 152 |
+
| 0.809 | 1.6369 | 2000 | 0.7705 |
|
| 153 |
+
| 0.8013 | 1.6533 | 2020 | 0.7707 |
|
| 154 |
+
| 0.8106 | 1.6697 | 2040 | 0.7680 |
|
| 155 |
+
| 0.8235 | 1.6860 | 2060 | 0.7695 |
|
| 156 |
+
| 0.8075 | 1.7024 | 2080 | 0.7692 |
|
| 157 |
+
| 0.8098 | 1.7188 | 2100 | 0.7679 |
|
| 158 |
+
| 0.8036 | 1.7352 | 2120 | 0.7690 |
|
| 159 |
+
| 0.8119 | 1.7515 | 2140 | 0.7664 |
|
| 160 |
+
| 0.8111 | 1.7679 | 2160 | 0.7672 |
|
| 161 |
+
| 0.8052 | 1.7843 | 2180 | 0.7670 |
|
| 162 |
+
| 0.8084 | 1.8007 | 2200 | 0.7673 |
|
| 163 |
+
| 0.8102 | 1.8170 | 2220 | 0.7675 |
|
| 164 |
+
| 0.8099 | 1.8334 | 2240 | 0.7665 |
|
| 165 |
+
| 0.7982 | 1.8498 | 2260 | 0.7660 |
|
| 166 |
+
| 0.8052 | 1.8661 | 2280 | 0.7652 |
|
| 167 |
+
| 0.8096 | 1.8825 | 2300 | 0.7646 |
|
| 168 |
+
| 0.8099 | 1.8989 | 2320 | 0.7659 |
|
| 169 |
+
| 0.7973 | 1.9153 | 2340 | 0.7661 |
|
| 170 |
+
| 0.817 | 1.9316 | 2360 | 0.7650 |
|
| 171 |
+
| 0.7987 | 1.9480 | 2380 | 0.7646 |
|
| 172 |
+
| 0.8112 | 1.9644 | 2400 | 0.7630 |
|
| 173 |
+
| 0.8051 | 1.9808 | 2420 | 0.7635 |
|
| 174 |
+
| 0.797 | 1.9971 | 2440 | 0.7635 |
|
| 175 |
+
| 0.7871 | 2.0131 | 2460 | 0.7629 |
|
| 176 |
+
| 0.8124 | 2.0295 | 2480 | 0.7633 |
|
| 177 |
+
| 0.7969 | 2.0458 | 2500 | 0.7642 |
|
| 178 |
+
| 0.7906 | 2.0622 | 2520 | 0.7634 |
|
| 179 |
+
| 0.8032 | 2.0786 | 2540 | 0.7632 |
|
| 180 |
+
| 0.806 | 2.0950 | 2560 | 0.7625 |
|
| 181 |
+
| 0.8019 | 2.1113 | 2580 | 0.7628 |
|
| 182 |
+
| 0.8017 | 2.1277 | 2600 | 0.7614 |
|
| 183 |
+
| 0.8106 | 2.1441 | 2620 | 0.7623 |
|
| 184 |
+
| 0.7996 | 2.1605 | 2640 | 0.7622 |
|
| 185 |
+
| 0.8038 | 2.1768 | 2660 | 0.7621 |
|
| 186 |
+
| 0.8036 | 2.1932 | 2680 | 0.7607 |
|
| 187 |
+
| 0.8032 | 2.2096 | 2700 | 0.7618 |
|
| 188 |
+
| 0.797 | 2.2260 | 2720 | 0.7608 |
|
| 189 |
+
| 0.8031 | 2.2423 | 2740 | 0.7607 |
|
| 190 |
+
| 0.8004 | 2.2587 | 2760 | 0.7606 |
|
| 191 |
+
| 0.8043 | 2.2751 | 2780 | 0.7604 |
|
| 192 |
+
| 0.7971 | 2.2914 | 2800 | 0.7614 |
|
| 193 |
+
| 0.7982 | 2.3078 | 2820 | 0.7607 |
|
| 194 |
+
| 0.7972 | 2.3242 | 2840 | 0.7601 |
|
| 195 |
+
| 0.8071 | 2.3406 | 2860 | 0.7609 |
|
| 196 |
+
| 0.8094 | 2.3569 | 2880 | 0.7605 |
|
| 197 |
+
| 0.7994 | 2.3733 | 2900 | 0.7600 |
|
| 198 |
+
| 0.7918 | 2.3897 | 2920 | 0.7596 |
|
| 199 |
+
| 0.7978 | 2.4061 | 2940 | 0.7597 |
|
| 200 |
+
| 0.8026 | 2.4224 | 2960 | 0.7599 |
|
| 201 |
+
| 0.7981 | 2.4388 | 2980 | 0.7593 |
|
| 202 |
+
| 0.7957 | 2.4552 | 3000 | 0.7600 |
|
| 203 |
+
| 0.8089 | 2.4716 | 3020 | 0.7601 |
|
| 204 |
+
| 0.801 | 2.4879 | 3040 | 0.7598 |
|
| 205 |
+
| 0.8044 | 2.5043 | 3060 | 0.7592 |
|
| 206 |
+
| 0.7967 | 2.5207 | 3080 | 0.7591 |
|
| 207 |
+
| 0.803 | 2.5370 | 3100 | 0.7589 |
|
| 208 |
+
| 0.8006 | 2.5534 | 3120 | 0.7591 |
|
| 209 |
+
| 0.7933 | 2.5698 | 3140 | 0.7599 |
|
| 210 |
+
| 0.7954 | 2.5862 | 3160 | 0.7590 |
|
| 211 |
+
| 0.7996 | 2.6025 | 3180 | 0.7592 |
|
| 212 |
+
| 0.7931 | 2.6189 | 3200 | 0.7597 |
|
| 213 |
+
| 0.8034 | 2.6353 | 3220 | 0.7595 |
|
| 214 |
+
| 0.7954 | 2.6517 | 3240 | 0.7606 |
|
| 215 |
+
| 0.801 | 2.6680 | 3260 | 0.7592 |
|
| 216 |
+
| 0.7973 | 2.6844 | 3280 | 0.7595 |
|
| 217 |
+
| 0.7976 | 2.7008 | 3300 | 0.7601 |
|
| 218 |
+
| 0.8014 | 2.7172 | 3320 | 0.7606 |
|
| 219 |
+
| 0.8052 | 2.7335 | 3340 | 0.7590 |
|
| 220 |
+
| 0.7898 | 2.7499 | 3360 | 0.7609 |
|
| 221 |
+
| 0.794 | 2.7663 | 3380 | 0.7598 |
|
| 222 |
+
| 0.7984 | 2.7826 | 3400 | 0.7609 |
|
| 223 |
+
| 0.7905 | 2.7990 | 3420 | 0.7587 |
|
| 224 |
+
| 0.7994 | 2.8154 | 3440 | 0.7595 |
|
| 225 |
+
| 0.8083 | 2.8318 | 3460 | 0.7588 |
|
| 226 |
+
| 0.7979 | 2.8481 | 3480 | 0.7585 |
|
| 227 |
+
| 0.7947 | 2.8645 | 3500 | 0.7596 |
|
| 228 |
+
| 0.8081 | 2.8809 | 3520 | 0.7592 |
|
| 229 |
+
| 0.801 | 2.8973 | 3540 | 0.7615 |
|
| 230 |
+
| 0.7959 | 2.9136 | 3560 | 0.7590 |
|
| 231 |
+
| 0.8081 | 2.9300 | 3580 | 0.7594 |
|
| 232 |
+
| 0.7927 | 2.9464 | 3600 | 0.7589 |
|
| 233 |
+
| 0.8006 | 2.9628 | 3620 | 0.7594 |
|
| 234 |
+
| 0.7933 | 2.9791 | 3640 | 0.7591 |
|
| 235 |
+
| 0.7896 | 2.9955 | 3660 | 0.7591 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 236 |
|
| 237 |
|
| 238 |
### Framework versions
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 309980480
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2574635f7a05617db02480a7e197e432f6461182a1d06a59680730c008b39ddd
|
| 3 |
size 309980480
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second":
|
| 6 |
-
"eval_steps_per_second":
|
| 7 |
-
"perplexity":
|
| 8 |
-
"total_flos": 1.
|
| 9 |
-
"train_loss":
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second": 0.
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.997953336062219,
|
| 3 |
+
"eval_loss": 0.759117603302002,
|
| 4 |
+
"eval_runtime": 31.7884,
|
| 5 |
+
"eval_samples_per_second": 526.953,
|
| 6 |
+
"eval_steps_per_second": 32.937,
|
| 7 |
+
"perplexity": 2.136390244832166,
|
| 8 |
+
"total_flos": 1.415185036075008e+16,
|
| 9 |
+
"train_loss": 1.0084684829779607,
|
| 10 |
+
"train_runtime": 6791.37,
|
| 11 |
+
"train_samples_per_second": 34.531,
|
| 12 |
+
"train_steps_per_second": 0.539
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples_per_second":
|
| 6 |
-
"eval_steps_per_second":
|
| 7 |
-
"perplexity":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.997953336062219,
|
| 3 |
+
"eval_loss": 0.759117603302002,
|
| 4 |
+
"eval_runtime": 31.7884,
|
| 5 |
+
"eval_samples_per_second": 526.953,
|
| 6 |
+
"eval_steps_per_second": 32.937,
|
| 7 |
+
"perplexity": 2.136390244832166
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 1.
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.997953336062219,
|
| 3 |
+
"total_flos": 1.415185036075008e+16,
|
| 4 |
+
"train_loss": 1.0084684829779607,
|
| 5 |
+
"train_runtime": 6791.37,
|
| 6 |
+
"train_samples_per_second": 34.531,
|
| 7 |
+
"train_steps_per_second": 0.539
|
| 8 |
}
|