maicomputer/alpaca-native

1# Stanford Alpaca2 3This is a replica of Alpaca by Stanford' tatsu4 5Trained using the original instructions with a minor modification in FSDP mode6 7# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)8 9| Metric                | Value                     |10|-----------------------|---------------------------|11| Avg.                  | 41.96   |12| ARC (25-shot)         | 52.3          |13| HellaSwag (10-shot)   | 77.09    |14| MMLU (5-shot)         | 41.6         |15| TruthfulQA (0-shot)   | 37.58   |16| Winogrande (5-shot)   | 69.46   |17| GSM8K (5-shot)        | 1.44        |18| DROP (3-shot)         | 14.23         |19