🎩 Magicoder is a family of 7B parameter models trained on 75K synthetic instruction data using OSS-Instruct, a novel approach to enlightening LLMs with open-source code snippets.

ollama run magicoder:7b-s-cl-q3_K_L

curl http://localhost:11434/api/chat \
  -d '{
    "model": "magicoder:7b-s-cl-q3_K_L",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='magicoder:7b-s-cl-q3_K_L',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'magicoder:7b-s-cl-q3_K_L',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 2 years ago

2 years ago

2bf1232fd9c1 · 3.6GB

model

archllama

parameters6.74B

quantizationQ3_K_L

3.6GB

params

{ "num_ctx": 16384 }

18B

system

You are an exceptionally intelligent coding assistant that consistently delivers accurate and reliab

134B

template

{{ .System }} @@ Instruction {{ .Prompt }} @@ Response

57B

Readme

🎩

Magicoder is a model family empowered by OSS-Instruct, a novel approach to enlightening LLMs with open-source code snippets for generating low-bias and high-quality instruction data for code.

OSS-Instruct mitigates the inherent bias of the LLM-synthesized instruction data by empowering them with a wealth of open-source references to produce more diverse, realistic, and controllable data.

References

GitHub

HuggingFace