Everyone Coder 33b Base

🚀 Everyone-Coder-33b-Base

The EveryoneLLM series of models is created by the community for the community. This is a coding-specific model fine-tuned from deekseekcoder-33b-base.

🚀 Quick Start

This section provides a quick overview of the model and how to start using it.

Model Overview

The Everyone-Coder-33b-Base is a powerful coding model that shows excellent performance in coding challenges compared to up-to-date GPT-4. However, it has some issues with end tokens, and a custom prompt template is provided to address this problem.

Custom Prompt Template

Always end with "<|EOT|>"

{System}

<|User|>

{User}. Always end your response with "<|EOT|>"

<|Assistant|>

{Assistant}

Add <|EOT|> as a custom stop string in your LM text generating interface.

✨ Features

Community-Driven

The model is part of the EveryoneLLM series, which is developed by the community for the community, leveraging the collective wisdom and resources.

Coding-Specific

Fine-tuned from deekseekcoder-33b-base, it is specifically designed for coding tasks and has shown strong performance in coding challenges.

📦 Installation

No specific installation steps are provided in the original document, so this section is skipped.

💻 Usage Examples

No code examples are provided in the original document, so this section is skipped.

📚 Documentation

Model Merger

The models used in this merger are:

• https://huggingface.co/deepseek-ai/deepseek-coder-33b-instruct
• https://huggingface.co/codefuse-ai/CodeFuse-DeepSeek-33B
• https://huggingface.co/WizardLM/WizardCoder-33B-V1.1

You can find the write-up for merging models here.

Merger Config

models:
  - model: WizardLM_WizardCoder-33B-V1.1
    parameters:
      density: 1
      weight: .5
  - model: codefuse-ai_CodeFuse-DeepSeek-33B
    parameters:
      density: 1
      weight: .5
merge_method: ties
base_model: deepseek-ai_deepseek-coder-33b-instruct
parameters:
  normalize: true
  int8_mask: true
dtype: float16

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

🔧 Technical Details

No technical implementation details are provided in the original document, so this section is skipped.

📄 License

The model is under the DeepSeek license.

⚠️ Important Note

The author has trouble benchmarking this model due to lack of experience in running LLM benchmarks. Also, the model has trouble with end tokens, and a custom prompt template is provided to solve this problem.

💡 Usage Tip

Add (Always end with "<|EOT|>") in addition to your system prompt and (Always end your response with "<|EOT|>") at the end of the User message. Then add <|EOT|> as a custom stop string in your LM text generating interface.