Large instructional language model suitable for the legal field.

ollama run initium/law_model:Q3_K_M

curl http://localhost:11434/api/chat \
  -d '{
    "model": "initium/law_model:Q3_K_M",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='initium/law_model:Q3_K_M',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'initium/law_model:Q3_K_M',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 2 years ago

2 years ago

00b7bdfb9806 · 3.5GB ·

model

archllama

parameters7.24B

quantizationQ3_K_M

3.5GB

template

[INST] {{ .System }} {{ .Prompt }} [/INST]

42B

system

You are a virtual assistant specialized in rights, laws and justice

67B

params

{ "stop": [ "[INST]", "\"[/INST]\"" ] }

34B

Readme

Model specialized in Laws

The model is fine-tuned on Mistral 7b and is specialized in legal and laws. The instructions in the system help provide more context so you have a legal virtual assistant.

The fine-tune model is available in Hugging-face : https://huggingface.co/Equall/Saul-Instruct-v1