Skip to content

Latest commit

 

History

History
70 lines (48 loc) · 1.47 KB

README.md

File metadata and controls

70 lines (48 loc) · 1.47 KB

Llama3-JP-using-NoteRED

Make a web service using Llama3 -8B for Japanese conversation. Model supported by Llama-3-ELYZA-JP-8B-GGUF

This repo just a sample for using docker to create the composed continers.

The flow like as follows:

  1. create a llama api
  2. test your api (can skip)
  3. create a requirements list
  4. make dockerfile for llama continer.
  5. make docker-compose.yaml
  6. open the NoteRED webpage
  7. edit your service

How to use this repo:

1. Download the Llama model by yourself.

run this for Linux:

sudo wget https://huggingface.co/elyza/Llama-3-ELYZA-JP-8B-GGUF/resolve/main/Llama-3-ELYZA-JP-8B-q4_k_m.gguf

or download by yourself:

https://huggingface.co/elyza/Llama-3-ELYZA-JP-8B-GGUF/blob/main/Llama-3-ELYZA-JP-8B-q4_k_m.gguf

2. Download this repo

using:

git clone https://github.com/WUBIN10086/Llama3-JP-using-NoteRED.git

and add your downloaded model file (in step 2) to the repo floder, the structure should like:

web_llama
├── API-python
│   ├── llama_api.py
│   └── test_llama_fastapi.py
├── Llama-3-ELYZA-JP-8B-q4_k_m.gguf
├── requirements.txt
└── docker-compose.yml

!!! Dont forget add the model (Llama-3-ELYZA-JP-8B-q4_k_m.gguf).

3. create the docker continers

run this command:

docker-compose up --build

4. Open NoteRED in browser

127.0.0.1:1880

Import the content in the NoteRED.txt