Shows that you can memorize HumanEval using a LoRA adapter. Axolotl ignored the last two training samples, so achieves only 99.3% on HumanEval.
To use it, merge it with the mistralai/Mistral-7B-v0.1 base model.
lora-out
This model was trained from scratch on the None dataset.
It achieves the following results on the evaluation set:
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: 2e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 128
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|
| 0.1018 | 21.33 | 32 | 1.0007 |
| 0.017 | 42.67 | 64 | 1.7790 |
| 0.0199 | 64.0 | 96 | 1.9024 |
| 0.007 | 85.33 | 128 | 1.9155 |
Framework versions
- Transformers 4.35.0.dev0
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.14.1
