27 lines
916 B
Markdown
27 lines
916 B
Markdown
---
|
|
license: llama2
|
|
---
|
|
<p><h1> ShearedPlats-1.3b </h1></p>
|
|
An experimental finetune of Sheared LLaMA 1.3b with Alpaca-QLoRA
|
|
|
|
<h2> Datasets </h2>
|
|
Trained on alpca style datasets
|
|
|
|
<p><h2> Prompt Template </h2></p>
|
|
Uses alpaca style prompt template
|
|
<br><br>
|
|
|
|
# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
|
Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_vihangd__shearedplats-1.3b-v1)
|
|
|
|
| Metric | Value |
|
|
|-----------------------|---------------------------|
|
|
| Avg. | 31.63 |
|
|
| ARC (25-shot) | 35.41 |
|
|
| HellaSwag (10-shot) | 62.75 |
|
|
| MMLU (5-shot) | 24.75 |
|
|
| TruthfulQA (0-shot) | 33.93 |
|
|
| Winogrande (5-shot) | 58.48 |
|
|
| GSM8K (5-shot) | 0.53 |
|
|
| DROP (3-shot) | 5.55 |
|