metadata
license: llama2
base_model: codellama/CodeLlama-13b-hf
tags:
- generated_from_trainer
model-index:
- name: code-llama-text-to-sparql
results: []
code-llama-text-to-sparql
This model is a fine-tuned version of codellama/CodeLlama-13b-hf on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1870
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- training_steps: 400
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.0976 | 0.0710 | 20 | 1.0941 |
| 0.5959 | 0.1421 | 40 | 0.3399 |
| 0.3037 | 0.2131 | 60 | 0.2744 |
| 0.2385 | 0.2842 | 80 | 0.2441 |
| 0.2178 | 0.3552 | 100 | 0.2767 |
| 0.2271 | 0.4263 | 120 | 0.2303 |
| 0.2302 | 0.4973 | 140 | 0.2219 |
| 0.2433 | 0.5684 | 160 | 0.2162 |
| 0.2115 | 0.6394 | 180 | 0.2108 |
| 0.199 | 0.7105 | 200 | 0.2151 |
| 0.2117 | 0.7815 | 220 | 0.2071 |
| 0.2107 | 0.8526 | 240 | 0.2004 |
| 0.2008 | 0.9236 | 260 | 0.1990 |
| 0.2083 | 0.9947 | 280 | 0.1953 |
| 0.1905 | 1.0657 | 300 | 0.1933 |
| 0.2001 | 1.1368 | 320 | 0.1905 |
| 0.1887 | 1.2078 | 340 | 0.1905 |
| 0.1921 | 1.2789 | 360 | 0.1880 |
| 0.1837 | 1.3499 | 380 | 0.1870 |
| 0.1951 | 1.4210 | 400 | 0.1870 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.10.1
- Tokenizers 0.19.1