Skip to content
/ chat Public

leverage llama3.2 and other large language models to generate responses to your questions locally with no installation

License

Notifications You must be signed in to change notification settings

load1n9/chat

Repository files navigation

chat

Simply run the following command and thats it

deno run -A jsr:@loading/chat

[Optional] create a chat-config.toml file in the active directory to configure the chat

"$schema" = 'https://jsr.io/@loading/chat/0.1.16/config-schema.json'

[config]
model = "onnx-community/Llama-3.2-1B-Instruct"
system = [
  "You are an assistant designed to help with any questions the user might have."
]
max_new_tokens = 128
max_length = 20
temperature = 1.0
top_p = 1.0
repetition_penalty = 1.2

Run the server to kinda match a similar api to the openai chat api

deno serve -A jsr:@loading/chat/server

Try it out

curl -X POST http://localhost:8000/v1/completions \  -H "Content-Type: application/json" \  -d '{    "prompt": "Once upon a time",    "max_tokens": 50,    "temperature": 0.7  }'

(New) Code Companion

With the new code companion you can generate new projects and edit them and stuff

deno run -A jsr:@loading/chat/companion

type /help to get a list of commands

License

This project is licensed under the MIT License - see the LICENSE file for details.