internlm/SWE-Fixer-Retriever-7B

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kPublished:Jan 9, 2025License:mitArchitecture:Transformer0.0K Open Weights Warm

The internlm/SWE-Fixer-Retriever-7B is a 7.6 billion parameter language model, fine-tuned from Qwen2.5-7B, developed by InternLM. It is specifically designed as a code file retriever component within the SWE-Fixer pipeline for addressing GitHub issues. This model excels at identifying relevant code files for issue resolution, forming a crucial part of a retrieve-then-edit strategy.

Loading preview...

SWE-Fixer-Retriever-7B Overview

The internlm/SWE-Fixer-Retriever-7B is a specialized language model, fine-tuned from the Qwen2.5-7B architecture, developed by InternLM. It serves as a core component of the SWE-Fixer framework, which aims to resolve real-world GitHub issues through an efficient retrieve-then-edit pipeline. This model, with 7.6 billion parameters and a 131,072 token context length, is specifically engineered to act as a code file retriever.

Key Capabilities

  • Code File Retrieval: Its primary function is to accurately identify and retrieve relevant code files pertinent to a given GitHub issue.
  • Integration with SWE-Fixer: Designed to work seamlessly within the SWE-Fixer's two-component system, complementing the code editor.
  • GitHub Issue Resolution: Contributes to a streamlined process for addressing and fixing software bugs and issues reported on GitHub.

Good For

  • Developers and researchers working on automated software engineering tasks.
  • Applications requiring intelligent retrieval of code context for bug fixing or code modification.
  • Enhancing existing code repair or issue resolution pipelines with a dedicated retrieval component.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p