pre model card
Browse files
README.md
CHANGED
|
@@ -15,4 +15,83 @@ language:
|
|
| 15 |
- en
|
| 16 |
base_model: meta-math/MetaMath-Mistral-7B
|
| 17 |
---
|
|
|
|
| 18 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
- en
|
| 16 |
base_model: meta-math/MetaMath-Mistral-7B
|
| 17 |
---
|
| 18 |
+
# 🔢 Einstein-v6-7B
|
| 19 |
|
| 20 |
+
This model is a full fine-tuned version of [meta-math/MetaMath-Mistral-7B](meta-math/MetaMath-Mistral-7B) on the following datasets:
|
| 21 |
+
|
| 22 |
+
- 🧮 [TIGER-Lab/MathInstruct](https://huggingface.co/datasets/TIGER-Lab/MathInstruct)
|
| 23 |
+
- 📐 [microsoft/orca-math-word-problems-200k](https://huggingface.co/datasets/microsoft/orca-math-word-problems-200k)
|
| 24 |
+
|
| 25 |
+
This model is finetuned using `8xRTX3090` + `1xRTXA6000` using [axolotl](https://github.com/OpenAccess-AI-Collective/axolotl).
|
| 26 |
+
|
| 27 |
+
This model's training was sponsored by [sablo.ai](https://sablo.ai).
|
| 28 |
+
|
| 29 |
+
<details><summary>See axolotl config</summary>
|
| 30 |
+
|
| 31 |
+
axolotl version: `0.4.0`
|
| 32 |
+
```yaml
|
| 33 |
+
|
| 34 |
+
```
|
| 35 |
+
|
| 36 |
+
</details><br>
|
| 37 |
+
|
| 38 |
+
# 💬 Prompt Template
|
| 39 |
+
|
| 40 |
+
You can use this prompt template while using the model:
|
| 41 |
+
|
| 42 |
+
### Alpaca
|
| 43 |
+
|
| 44 |
+
```
|
| 45 |
+
Below is an instruction that describes a task. Write a response that appropriately completes the request.
|
| 46 |
+
|
| 47 |
+
### Instruction:
|
| 48 |
+
{instruction}
|
| 49 |
+
|
| 50 |
+
### Response:
|
| 51 |
+
|
| 52 |
+
```
|
| 53 |
+
|
| 54 |
+
This prompt template is available as a [chat template](https://huggingface.co/docs/transformers/main/chat_templating), which means you can format messages using the
|
| 55 |
+
`tokenizer.apply_chat_template()` method:
|
| 56 |
+
|
| 57 |
+
```python
|
| 58 |
+
messages = [
|
| 59 |
+
{"role": "system", "content": "You are helpful AI asistant."},
|
| 60 |
+
{"role": "user", "content": "Hello!"}
|
| 61 |
+
]
|
| 62 |
+
gen_input = tokenizer.apply_chat_template(message, return_tensors="pt")
|
| 63 |
+
model.generate(**gen_input)
|
| 64 |
+
```
|
| 65 |
+
|
| 66 |
+
# 🔄 Quantizationed versions
|
| 67 |
+
|
| 68 |
+
Quantizationed versions of this model is currently not available. It will be available soon :)
|
| 69 |
+
|
| 70 |
+
# 🎯 [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)
|
| 71 |
+
|
| 72 |
+
# 🤖 Additional information about training
|
| 73 |
+
|
| 74 |
+
This model is full fine-tuned for 2 epoch.
|
| 75 |
+
|
| 76 |
+
Total number of steps was x.
|
| 77 |
+
|
| 78 |
+
<details><summary>Loss graph</summary>
|
| 79 |
+
|
| 80 |
+
|
| 81 |
+
</details><br>
|
| 82 |
+
|
| 83 |
+
# 🤝 Acknowledgments
|
| 84 |
+
|
| 85 |
+
Thanks to [sablo.ai](https://sablo.ai) for sponsoring this model.
|
| 86 |
+
|
| 87 |
+
Thanks to all the dataset authors mentioned in the datasets section.
|
| 88 |
+
|
| 89 |
+
Thanks to [axolotl](https://github.com/OpenAccess-AI-Collective/axolotl) for making the repository I used to make this model.
|
| 90 |
+
|
| 91 |
+
Thanks to all open source AI community.
|
| 92 |
+
|
| 93 |
+
[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
|
| 94 |
+
|
| 95 |
+
If you would like to support me:
|
| 96 |
+
|
| 97 |
+
[☕ Buy Me a Coffee](https://www.buymeacoffee.com/weyaxi)
|