deepseek-coder-v2-16b

PyTorch

1 versions

An open-source Mixture-of-Experts code language model that achieves performance comparable to GPT4-Turbo in code-specific tasks.

Run this model

  1. Install our magic package manager:

    curl -ssL https://magic.modular.com/ | bash

    Then run the source command that's printed in your terminal.

  2. Install Max Pipelines in order to run this model.

    magic global install max-pipelines
  3. Start a local endpoint for deepseek-coder-v2/16b:

    max-serve serve --huggingface-repo-id deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct

    The endpoint is ready when you see the URI printed in your terminal:

    Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
  4. Now open another terminal to send a request using curl:

    curl -N http://0.0.0.0:8000/v1/chat/completions -H "Content-Type: application/json" -d '{
        "model": "deepseek-coder-v2/16b",
        "stream": true,
        "messages": [
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": "Who won the World Series in 2020?"}
        ]
    }' | grep -o '"content":"[^"]*"' | sed 's/"content":"//g' | sed 's/"//g' | tr -d '
    ' | sed 's/\n/
    /g'
  5. šŸŽ‰ Hooray! Youā€™re running Generative AI. Our goal is to make this as easy as possible.

About

DeepSeek-Coder-V2 is an open-source Mixture-of-Experts (MoE) language model designed for code-related tasks, achieving performance comparable to GPT4-Turbo. It is further pre-trained from its base version, DeepSeek-Coder-V2-Base, using a massive dataset consisting of 6 trillion tokens. These tokens are sourced from a diverse and high-quality corpus, enabling the model to excel in coding tasks while maintaining precision and efficiency.

DeepSeek-Coder-V2 incorporates advanced techniques, leveraging its MoE architecture to optimize computational efficiency and scalability. This architecture facilitates the model's ability to handle complex problem-solving scenarios in programming, making it suitable for developers and researchers seeking robust AI-driven code generation and analysis tools.

For additional details, visit the modelā€™s Hugging Face page.

DETAILS

MODEL CLASS
PyTorch

MODULAR GITHUB

Modular

CREATED BY

deepseek-ai

MODEL

deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct

TAGS

arxiv:2401.06066
autotrain_compatible
conversational
custom_code
deepseek_v2
endpoints_compatible
license:other
region:us
safetensors
text-generation
text-generation-inference
transformers

@ Copyright - Modular Inc - 2024