47015772_0
This model is a fine-tuned version of openai-community/gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6222
- Accuracy: 0.0000
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1.41e-05
- train_batch_size: 32
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- total_eval_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2000
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
1.0878 | 0.12 | 25 | 0.9843 | 0.0001 |
0.8657 | 0.25 | 50 | 0.8198 | 0.0001 |
0.808 | 0.38 | 75 | 0.7747 | 0.0001 |
0.7861 | 0.5 | 100 | 0.7512 | 0.0000 |
0.7619 | 0.62 | 125 | 0.7365 | 0.0000 |
0.7596 | 0.75 | 150 | 0.7256 | 0.0001 |
0.7449 | 0.88 | 175 | 0.7182 | 0.0001 |
0.7351 | 1.0 | 200 | 0.7104 | 0.0001 |
0.7199 | 1.12 | 225 | 0.7052 | 0.0000 |
0.7204 | 1.25 | 250 | 0.7010 | 0.0000 |
0.7142 | 1.38 | 275 | 0.6965 | 0.0000 |
0.7181 | 1.5 | 300 | 0.6929 | 0.0000 |
0.69 | 1.62 | 325 | 0.6894 | 0.0000 |
0.6944 | 1.75 | 350 | 0.6863 | 0.0000 |
0.6991 | 1.88 | 375 | 0.6823 | 0.0000 |
0.7005 | 2.0 | 400 | 0.6797 | 0.0000 |
0.6933 | 2.12 | 425 | 0.6770 | 0.0000 |
0.6875 | 2.25 | 450 | 0.6740 | 0.0000 |
0.6905 | 2.38 | 475 | 0.6721 | 0.0000 |
0.6881 | 2.5 | 500 | 0.6695 | 0.0000 |
0.6809 | 2.62 | 525 | 0.6671 | 0.0000 |
0.6763 | 2.75 | 550 | 0.6653 | 0.0000 |
0.6791 | 2.88 | 575 | 0.6635 | 0.0000 |
0.6815 | 3.0 | 600 | 0.6611 | 0.0000 |
0.6751 | 3.12 | 625 | 0.6598 | 0.0000 |
0.6631 | 3.25 | 650 | 0.6579 | 0.0000 |
0.6716 | 3.38 | 675 | 0.6569 | 0.0000 |
0.6685 | 3.5 | 700 | 0.6555 | 0.0000 |
0.6635 | 3.62 | 725 | 0.6535 | 0.0000 |
0.6763 | 3.75 | 750 | 0.6525 | 0.0000 |
0.6588 | 3.88 | 775 | 0.6513 | 0.0000 |
0.6621 | 4.0 | 800 | 0.6504 | 0.0000 |
0.6676 | 4.12 | 825 | 0.6487 | 0.0000 |
0.656 | 4.25 | 850 | 0.6475 | 0.0000 |
0.667 | 4.38 | 875 | 0.6465 | 0.0000 |
0.6589 | 4.5 | 900 | 0.6452 | 0.0000 |
0.6599 | 4.62 | 925 | 0.6446 | 0.0000 |
0.6492 | 4.75 | 950 | 0.6435 | 0.0000 |
0.6493 | 4.88 | 975 | 0.6425 | 0.0000 |
0.6544 | 5.0 | 1000 | 0.6412 | 0.0000 |
0.6577 | 5.12 | 1025 | 0.6401 | 0.0000 |
0.6464 | 5.25 | 1050 | 0.6393 | 0.0001 |
0.6466 | 5.38 | 1075 | 0.6385 | 0.0000 |
0.6608 | 5.5 | 1100 | 0.6372 | 0.0000 |
0.644 | 5.62 | 1125 | 0.6364 | 0.0001 |
0.6476 | 5.75 | 1150 | 0.6355 | 0.0000 |
0.6504 | 5.88 | 1175 | 0.6348 | 0.0001 |
0.6478 | 6.0 | 1200 | 0.6336 | 0.0000 |
0.6516 | 6.12 | 1225 | 0.6328 | 0.0000 |
0.6326 | 6.25 | 1250 | 0.6323 | 0.0000 |
0.6397 | 6.38 | 1275 | 0.6317 | 0.0001 |
0.6532 | 6.5 | 1300 | 0.6305 | 0.0001 |
0.6509 | 6.62 | 1325 | 0.6303 | 0.0000 |
0.6397 | 6.75 | 1350 | 0.6296 | 0.0000 |
0.6385 | 6.88 | 1375 | 0.6291 | 0.0000 |
0.6391 | 7.0 | 1400 | 0.6285 | 0.0000 |
0.6289 | 7.12 | 1425 | 0.6280 | 0.0000 |
0.6442 | 7.25 | 1450 | 0.6276 | 0.0000 |
0.642 | 7.38 | 1475 | 0.6271 | 0.0000 |
0.6496 | 7.5 | 1500 | 0.6266 | 0.0000 |
0.6389 | 7.62 | 1525 | 0.6263 | 0.0000 |
0.6309 | 7.75 | 1550 | 0.6259 | 0.0000 |
0.6438 | 7.88 | 1575 | 0.6255 | 0.0000 |
0.6281 | 8.0 | 1600 | 0.6251 | 0.0000 |
0.6342 | 8.12 | 1625 | 0.6248 | 0.0000 |
0.6224 | 8.25 | 1650 | 0.6245 | 0.0000 |
0.6391 | 8.38 | 1675 | 0.6241 | 0.0000 |
0.6448 | 8.5 | 1700 | 0.6240 | 0.0000 |
0.6259 | 8.62 | 1725 | 0.6236 | 0.0000 |
0.6231 | 8.75 | 1750 | 0.6234 | 0.0000 |
0.6383 | 8.88 | 1775 | 0.6231 | 0.0000 |
0.6205 | 9.0 | 1800 | 0.6229 | 0.0000 |
0.6204 | 9.12 | 1825 | 0.6228 | 0.0000 |
0.633 | 9.25 | 1850 | 0.6226 | 0.0000 |
0.6451 | 9.38 | 1875 | 0.6226 | 0.0000 |
0.639 | 9.5 | 1900 | 0.6224 | 0.0000 |
0.6292 | 9.62 | 1925 | 0.6223 | 0.0000 |
0.6295 | 9.75 | 1950 | 0.6222 | 0.0000 |
0.6425 | 9.88 | 1975 | 0.6222 | 0.0000 |
0.6323 | 10.0 | 2000 | 0.6222 | 0.0000 |
Framework versions
- Transformers 4.37.0
- Pytorch 2.0.0+cu118
- Datasets 2.16.1
- Tokenizers 0.15.1
- Downloads last month
- 0