output_dir_clean_df_10-100_noX_100_50_epoch_cluster
This model is a fine-tuned version of nferruz/ProtGPT2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 5.3807
- Accuracy: 0.2682
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
all AMPs from the compass dataset with the length between 10-100 AA. Only considering the 20 standard AA (NO X).
Training procedure
50 epochs training rate: 1 e-06 block size: 100
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-06
- train_batch_size: 24
- eval_batch_size: 24
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50.0
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
No log | 1.0 | 197 | 6.2811 | 0.2041 |
No log | 2.0 | 394 | 6.1540 | 0.2116 |
6.3092 | 3.0 | 591 | 6.0786 | 0.2153 |
6.3092 | 4.0 | 788 | 6.0237 | 0.2177 |
6.3092 | 5.0 | 985 | 5.9779 | 0.2200 |
6.0762 | 6.0 | 1182 | 5.9383 | 0.2222 |
6.0762 | 7.0 | 1379 | 5.9011 | 0.2249 |
5.9715 | 8.0 | 1576 | 5.8651 | 0.2273 |
5.9715 | 9.0 | 1773 | 5.8307 | 0.2302 |
5.9715 | 10.0 | 1970 | 5.8002 | 0.2320 |
5.8814 | 11.0 | 2167 | 5.7723 | 0.2338 |
5.8814 | 12.0 | 2364 | 5.7463 | 0.2355 |
5.8123 | 13.0 | 2561 | 5.7220 | 0.2371 |
5.8123 | 14.0 | 2758 | 5.6994 | 0.2386 |
5.8123 | 15.0 | 2955 | 5.6781 | 0.2404 |
5.7544 | 16.0 | 3152 | 5.6581 | 0.2419 |
5.7544 | 17.0 | 3349 | 5.6391 | 0.2433 |
5.7009 | 18.0 | 3546 | 5.6211 | 0.2450 |
5.7009 | 19.0 | 3743 | 5.6044 | 0.2466 |
5.7009 | 20.0 | 3940 | 5.5888 | 0.2482 |
5.6629 | 21.0 | 4137 | 5.5735 | 0.2493 |
5.6629 | 22.0 | 4334 | 5.5588 | 0.2507 |
5.6235 | 23.0 | 4531 | 5.5451 | 0.2520 |
5.6235 | 24.0 | 4728 | 5.5320 | 0.2531 |
5.6235 | 25.0 | 4925 | 5.5197 | 0.2541 |
5.5865 | 26.0 | 5122 | 5.5078 | 0.2552 |
5.5865 | 27.0 | 5319 | 5.4969 | 0.2562 |
5.5649 | 28.0 | 5516 | 5.4866 | 0.2573 |
5.5649 | 29.0 | 5713 | 5.4765 | 0.2583 |
5.5649 | 30.0 | 5910 | 5.4670 | 0.2595 |
5.5322 | 31.0 | 6107 | 5.4582 | 0.2604 |
5.5322 | 32.0 | 6304 | 5.4500 | 0.2612 |
5.5168 | 33.0 | 6501 | 5.4424 | 0.2618 |
5.5168 | 34.0 | 6698 | 5.4350 | 0.2627 |
5.5168 | 35.0 | 6895 | 5.4283 | 0.2633 |
5.4984 | 36.0 | 7092 | 5.4219 | 0.2640 |
5.4984 | 37.0 | 7289 | 5.4161 | 0.2647 |
5.4984 | 38.0 | 7486 | 5.4107 | 0.2651 |
5.48 | 39.0 | 7683 | 5.4058 | 0.2656 |
5.48 | 40.0 | 7880 | 5.4014 | 0.2660 |
5.4665 | 41.0 | 8077 | 5.3974 | 0.2665 |
5.4665 | 42.0 | 8274 | 5.3936 | 0.2668 |
5.4665 | 43.0 | 8471 | 5.3905 | 0.2671 |
5.4612 | 44.0 | 8668 | 5.3878 | 0.2674 |
5.4612 | 45.0 | 8865 | 5.3855 | 0.2677 |
5.4515 | 46.0 | 9062 | 5.3836 | 0.2679 |
5.4515 | 47.0 | 9259 | 5.3822 | 0.2680 |
5.4515 | 48.0 | 9456 | 5.3812 | 0.2681 |
5.4453 | 49.0 | 9653 | 5.3808 | 0.2682 |
5.4453 | 50.0 | 9850 | 5.3807 | 0.2682 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.2.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
- Downloads last month
- 0