Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/gallen881/physics_master
Physics Master is a model fine-tuned from llama3-8B-Instruct. It can answer your physics question!
https://github.com/gallen881/physics_master
ai fine-tuning finetune finetuning llama3 llm physics
Last synced: 2 months ago
JSON representation
Physics Master is a model fine-tuned from llama3-8B-Instruct. It can answer your physics question!
- Host: GitHub
- URL: https://github.com/gallen881/physics_master
- Owner: gallen881
- License: mit
- Created: 2024-08-03T09:18:32.000Z (5 months ago)
- Default Branch: master
- Last Pushed: 2024-08-24T15:19:34.000Z (4 months ago)
- Last Synced: 2024-10-10T18:03:56.582Z (2 months ago)
- Topics: ai, fine-tuning, finetune, finetuning, llama3, llm, physics
- Language: Jupyter Notebook
- Homepage:
- Size: 21.5 KB
- Stars: 13
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Physics Master
Physics Master is a model finetuned from [unsloth/llama-3-8b-Instruct-bnb-4bit](https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit) and [ArtifactAI/arxiv-physics-instruct-tune-30k](https://huggingface.co/datasets/ArtifactAI/arxiv-physics-instruct-tune-30k). Here is the python code. The model is open source on [Hugging Face](https://huggingface.co/gallen881/Llama-3-8B-Physics_Master-GGUF).
## How to use?
### Run model on local computer
1. Clone this repo.
```
git clone https://github.com/gallen881/Physics_Master.git
```
2. Move to `Physics_Master`
```
cd Physics_Master
```
3. Install dependencies.
```
pip install -r requirements.txt
```
3. Run `chat.py`.
```
python chat.py
```
### Run model on Colab
https://colab.research.google.com/github/gallen881/Physics_Master/blob/master/chat.ipynb
## Additional Function
If want to use WolframAlpha, follow these steps:
1. If don't have a WolframAlpha account, go [here](https://developer.wolframalpha.com/) and create one.
2. Go [here](https://developer.wolframalpha.com/access) and click **Get an App ID**.
3. Enter **Name** and **Description**, then select **Full Results API**.
4. Click **Submit**, then copy the **App ID**.
5. Run `set_keys.py`, then enter your App ID.
```
python set_keys.py
```
# 物理大師
物理大師是基於 [unsloth/llama-3-8b-Instruct-bnb-4bit](https://huggingface.co/unsloth/llama-3-8b-Instruct-bnb-4bit) 和 [ArtifactAI/arxiv-physics-instruct-tune-30k](https://huggingface.co/datasets/ArtifactAI/arxiv-physics-instruct-tune-30k) 微調的模型。這裡存放 Python 腳本。模型開源於 [Hugging Face](https://huggingface.co/gallen881/Llama-3-8B-Physics_Master-GGUF)。
## 如何使用？
### 在本地電腦運行
1. clone 這個 repo
```
git clone https://github.com/gallen881/Physics_Master.git
```
2. 移動目錄到 `Physics_Master`
```
cd Physics_Master
```
3. 安裝依賴
```
pip install -r requirements.txt
```
3. 運行 `chat.py`
```
python chat.py
```
### 在 colab 上運行
https://colab.research.google.com/github/gallen881/Physics_Master/blob/master/chat.ipynb
## 額外功能
如果想使用 WolframAlpha 的功能，依照以下步驟：
1. 如果沒有 WolframAlpha 帳號，前往[這裡](https://developer.wolframalpha.com/)註冊。
2. 前往[這裡](https://developer.wolframalpha.com/access) 然後點擊 **Get an App ID**。
3. 輸入 **Name** 和 **Description** 然後選擇 **Full Results API**。
4. 點擊 **Submit** 然後複製 **App ID**。
5. 執行 `set_keys.py` 然後輸入你的 App ID。
```
python set_keys.py
```