deepseek-coder-6.7b-base-APR-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: 0.5697
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 |
---|---|---|---|
0.6332 | 0.05 | 100 | 0.6259 |
0.6011 | 0.1 | 200 | 0.6139 |
0.6662 | 0.15 | 300 | 0.6086 |
0.6164 | 0.2 | 400 | 0.6028 |
0.5958 | 0.25 | 500 | 0.5991 |
0.6025 | 0.3 | 600 | 0.5950 |
0.5746 | 0.35 | 700 | 0.5907 |
0.5912 | 0.4 | 800 | 0.5882 |
0.5544 | 0.45 | 900 | 0.5839 |
0.5653 | 0.5 | 1000 | 0.5796 |
0.562 | 0.55 | 1100 | 0.5782 |
0.5724 | 0.6 | 1200 | 0.5758 |
0.5307 | 0.65 | 1300 | 0.5738 |
0.5275 | 0.7 | 1400 | 0.5731 |
0.5432 | 0.75 | 1500 | 0.5714 |
0.5463 | 0.8 | 1600 | 0.5704 |
0.5459 | 0.85 | 1700 | 0.5699 |
0.5751 | 0.9 | 1800 | 0.5696 |
0.5057 | 0.95 | 1900 | 0.5696 |
0.5485 | 1.0 | 2000 | 0.5697 |
Framework versions
- PEFT 0.13.0
- Transformers 4.44.2
- Pytorch 2.1.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1
- Downloads last month
- 6
Model tree for ardalaaan/deepseek-coder-6.7b-base-APR-finetuning
Base model
deepseek-ai/deepseek-coder-6.7b-base