metadata
license: mit
base_model: gpt2
tags:
- generated_from_trainer
model-index:
- name: '130000'
results: []
130000
This model is a fine-tuned version of gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 5.9987
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: 0.0005
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
No log | 0.92 | 3 | 7.0396 |
No log | 1.85 | 6 | 6.5398 |
No log | 2.77 | 9 | 6.3337 |
6.6916 | 4.0 | 13 | 6.3694 |
6.6916 | 4.92 | 16 | 6.2945 |
6.6916 | 5.85 | 19 | 6.3184 |
6.1092 | 6.77 | 22 | 6.3726 |
6.1092 | 8.0 | 26 | 6.2948 |
6.1092 | 8.92 | 29 | 6.3374 |
6.5151 | 9.85 | 32 | 6.3641 |
6.5151 | 10.77 | 35 | 6.2335 |
6.5151 | 12.0 | 39 | 6.1965 |
5.998 | 12.92 | 42 | 6.0595 |
5.998 | 13.85 | 45 | 6.0374 |
5.998 | 14.77 | 48 | 6.0562 |
5.6623 | 16.0 | 52 | 6.0128 |
5.6623 | 16.92 | 55 | 5.9999 |
5.6623 | 17.85 | 58 | 6.0008 |
5.611 | 18.77 | 61 | 5.9992 |
5.611 | 20.0 | 65 | 6.0017 |
5.611 | 20.92 | 68 | 6.0005 |
5.5519 | 21.85 | 71 | 5.9962 |
5.5519 | 22.77 | 74 | 5.9964 |
5.5519 | 24.0 | 78 | 5.9975 |
5.5841 | 24.92 | 81 | 5.9974 |
5.5841 | 25.85 | 84 | 6.0000 |
5.5841 | 26.77 | 87 | 6.0019 |
5.5582 | 28.0 | 91 | 6.0014 |
5.5582 | 28.92 | 94 | 6.0016 |
5.5582 | 29.85 | 97 | 5.9987 |
5.591 | 30.77 | 100 | 5.9992 |
5.591 | 32.0 | 104 | 5.9986 |
5.591 | 32.92 | 107 | 5.9982 |
5.5638 | 33.85 | 110 | 5.9983 |
5.5638 | 34.77 | 113 | 5.9987 |
5.5638 | 36.0 | 117 | 5.9989 |
5.5683 | 36.92 | 120 | 5.9992 |
5.5683 | 37.85 | 123 | 5.9995 |
5.5683 | 38.77 | 126 | 5.9991 |
5.5628 | 40.0 | 130 | 5.9992 |
5.5628 | 40.92 | 133 | 5.9992 |
5.5628 | 41.85 | 136 | 5.9991 |
5.5628 | 42.77 | 139 | 5.9989 |
5.5683 | 44.0 | 143 | 5.9987 |
5.5683 | 44.92 | 146 | 5.9987 |
5.5683 | 45.85 | 149 | 5.9987 |
5.5534 | 46.15 | 150 | 5.9987 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2