metadata
base_model: deepseek-ai/deepseek-coder-6.7b-base
library_name: peft
license: other
tags:
- unsloth
- generated_from_trainer
model-index:
- name: deepseek-coder-6.7b-base-APR-FIM-finetuning
results: []
deepseek-coder-6.7b-base-APR-FIM-finetuning
This model is a fine-tuned version of deepseek-ai/deepseek-coder-6.7b-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5779
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 16
- seed: 11
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- training_steps: 2000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.6471 | 0.05 | 100 | 0.6437 |
| 0.6132 | 0.1 | 200 | 0.6208 |
| 0.6719 | 0.15 | 300 | 0.6141 |
| 0.6325 | 0.2 | 400 | 0.6089 |
| 0.6124 | 0.25 | 500 | 0.6054 |
| 0.5842 | 0.3 | 600 | 0.6023 |
| 0.5537 | 0.35 | 700 | 0.5982 |
| 0.5966 | 0.4 | 800 | 0.5951 |
| 0.5757 | 0.45 | 900 | 0.5921 |
| 0.5856 | 0.5 | 1000 | 0.5879 |
| 0.6049 | 0.55 | 1100 | 0.5864 |
| 0.5611 | 0.6 | 1200 | 0.5841 |
| 0.5753 | 0.65 | 1300 | 0.5821 |
| 0.541 | 0.7 | 1400 | 0.5810 |
| 0.5838 | 0.75 | 1500 | 0.5795 |
| 0.5326 | 0.8 | 1600 | 0.5789 |
| 0.5292 | 0.85 | 1700 | 0.5784 |
| 0.5548 | 0.9 | 1800 | 0.5780 |
| 0.552 | 0.95 | 1900 | 0.5779 |
| 0.9524 | 1.0 | 2000 | 0.5779 |
Framework versions
- PEFT 0.13.0
- Transformers 4.44.2
- Pytorch 2.1.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1