metadata
license: llama2
base_model: codellama/CodeLlama-13b-Instruct-hf
tags:
- trl
- sft
- unsloth
- generated_from_trainer
- unsloth
datasets:
- generator
model-index:
- name: ShaderLLM-CodeLlama-13b-it
results: []
ShaderLLM-CodeLlama-13b-it
This model is a fine-tuned version of codellama/CodeLlama-13b-Instruct-hf on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.7616
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: 0.0025
- train_batch_size: 4
- eval_batch_size: 1
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: polynomial
- num_epochs: 2
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.5616 | 0.24 | 1000 | 1.2367 |
| 1.0239 | 0.49 | 2000 | 1.0618 |
| 0.9445 | 0.73 | 3000 | 0.9659 |
| 1.0534 | 0.98 | 4000 | 0.8703 |
| 0.653 | 1.22 | 5000 | 0.8155 |
| 0.6389 | 1.46 | 6000 | 0.7856 |
| 0.6987 | 1.71 | 7000 | 0.7670 |
| 0.6333 | 1.95 | 8000 | 0.7616 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.2.2
- Datasets 2.18.0
- Tokenizers 0.15.2