Skip to content

Latest commit

 

History

History
111 lines (74 loc) · 4.98 KB

codellama.md

File metadata and controls

111 lines (74 loc) · 4.98 KB
Raw

codellama

Introduction

codellama features enhanced coding capabilities. It can generate code and natural language about code, from both code and natural language prompts (e.g., “Write me a function that outputs the fibonacci sequence”). It can also be used for code completion and debugging. It supports many of the most popular programming languages used today, including Python, C++, Java, PHP, Typescript (Javascript), C#, Bash and more.

There are three sizes (7b, 13b, 34b) as well as three flavours (base model, Python fine-tuned, and instruction tuned) released on HuggingFace.

Base Model Python Instruct
codellama/CodeLlama-7b-hf codellama/CodeLlama-7b-Python-hf codellama/CodeLlama-7b-Instruct-hf
codellama/CodeLlama-13b-hf codellama/CodeLlama-13b-Python-hf codellama/CodeLlama-13b-Instruct-hf
codellama/CodeLlama-34b-hf codellama/CodeLlama-34b-Python-hf codellama/CodeLlama-34b-Instruct-hf

The correspondence between the model and capabilities is:

models code completion infilling instructions / chat python specialist
Base Model Y Y(7B,13B), N(34B) N N
Python Y N N Y
Instruct Y Y(7B,13B), N(34B) Y N

Inference

Based on the above table, download the model that meets your requirements. Execute the following command to interact with the model in the console:

# install lmdeploy
python3 -m pip install lmdeploy[all]

# convert weight layout
lmdeploy convert codellama /the/path/of/codellama/model

Then, you can communicate with codellama in consolo by following instructions in next sections

Note:

  • minimum requirement of transformers is v4.33.0
  • lmdeploy supports copying code blocks to the console. But you have to press enter, input "!!" and press enter again to end the prompt. The way to input prompt for other supported models keeps unchanged, i.e., double pressing enter.

Completion

lmdeploy chat ./workspace --cap completion

Infilling

lmdeploy chat ./workspace --cap infilling

The input code is supposed to have a special placeholder <FILL>. For example,

def remove_non_ascii(s: str) -> str:
    """ <FILL>
    return result

And the generated code piece by turbomind.chat is the one to be filled in <FILL>

Chat

lmdeploy chat ./workspace --cap chat --meta-instruct "Provide answers in Python"

--meta-instruct instruction can be changed to other coding languages as long as codellama supports it

Python specialist

lmdeploy chat ./workspace --cap python

Python fine-tuned model is highly recommended when 'python specialist' capability is required.

Quantization

TBD

Serving

LMDeploy server only supports chat capabllity. The res ones are going to be supported soon.

Launch inference server by:

# --tp: the number of GPUs used in tensor parallelism
lmdeploy serve api_server ./workspace --server-name ${server_ip} --server-port ${server_port} --tp 1

Then, you can communicate with it by command line,

# restful_api_url is what printed in api_server.py, e.g. http://localhost:23333
lmdeploy serve api_client api_server_url

or through webui after launching gradio,

# api_server_url is what printed in api_server.py, e.g. http://localhost:23333
# server_ip and server_port here are for gradio ui
# example: lmdeploy serve gradio http://localhost:23333 --server-name localhost --server-port 6006
lmdeploy serve gradio api_server_url --server-name ${gradio_ui_ip} --server-port ${gradio_ui_port}

Regarding the detailed information of RESTful API, you can refer to the guide.