santacoder-finetuned-the-stack-bash
This model is a fine-tuned version of bigcode/santacoder on the dbrief dataset. It achieves the following results on the evaluation set:
- Loss: 0.9406
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: 5e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 5000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.1768 | 0.1 | 500 | 1.6939 |
| 2.8451 | 0.2 | 1000 | 2.0221 |
| 3.0141 | 0.3 | 1500 | 2.1031 |
| 2.7306 | 0.4 | 2000 | 2.1053 |
| 2.3385 | 0.5 | 2500 | 2.0770 |
| 1.981 | 0.6 | 3000 | 0.9364 |
| 1.6687 | 0.7 | 3500 | 0.9637 |
| 1.2663 | 0.8 | 4000 | 0.9459 |
| 1.7485 | 0.9 | 4500 | 0.9428 |
| 1.9006 | 1.0 | 5000 | 0.9406 |
Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.0
- Tokenizers 0.13.3
