---
library_name: peft
base_model: kyujinpy/Sakura-SOLAR-Instruct
license: apache-2.0
datasets:
- meta-math/MetaMathQA
language:
- en
pipeline_tag: text-generation
tags:
- SOLAR
- MetaMathQA
---

# MOLAR 10.7B

Q-LoRA fine-tuned on **kyujinpy/Sakura-SOLAR-Instruct** with **meta-math/MetaMathQA**

Adapter model and normal model is avaible, you can download with both styles.

You must use this prompt style.

```yaml
### Query: <query>
### Response: <response>
```

# Loss Curve:
![image/png](https://cdn-uploads.huggingface.co/production/uploads/63da3d7ae697e5898cb86854/5OW8k59b_3rlqWlZfYkWx.png)

# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
Detailed results will be found [Here]()

| Metric                | Value                     |
|-----------------------|---------------------------|
| Avg.                  |                |
| ARC (25-shot)         |                |
| HellaSwag (10-shot)   |                |
| MMLU (5-shot)         |                |
| TruthfulQA (0-shot)   |                |
| Winogrande (5-shot)   |                |
| GSM8K (5-shot)        |                |