Saikrishna2511/java2py-qwen

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 5, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

Saikrishna2511/java2py-qwen is a 0.5 billion parameter Qwen2.5-Coder-0.5B-Instruct model, fine-tuned by Saikrishna2511, specifically for translating Java code into Python code. This model leverages a LoRA fine-tuning approach on Java-to-Python pairs from the AVATAR-TC dataset. It is optimized for direct code translation tasks, providing a specialized solution for migrating Java snippets to Python.

Loading preview...

Model Overview

Saikrishna2511/java2py-qwen is a specialized 0.5 billion parameter language model, fine-tuned from the Qwen2.5-Coder-0.5B-Instruct base model. Its primary function is to translate Java code into Python code, making it a focused tool for code migration tasks.

Key Capabilities

  • Java to Python Translation: The model is specifically trained to convert Java code snippets into their Python equivalents. It uses a prompt format that clearly delineates the Java input and expects Python output.
  • LoRA Fine-tuning: This checkpoint was developed using LoRA (Low-Rank Adaptation) with specific parameters (r=16, alpha=32) on the AVATAR-TC dataset, focusing on Java→Python pairs. The weights have been merged for direct inference.

Training Details

The model was fine-tuned on the Qwen/Qwen2.5-Coder-0.5B-Instruct base model using Java→Python pairs from the AVATAR-TC dataset. This targeted training ensures its proficiency in the specific task of cross-language code translation.

Limitations

  • As a relatively small 0.5B parameter model, its translation quality can vary significantly with the complexity of the input Java code.
  • For more versatile code-related tasks, such as natural language to Python or code to documentation, the multi-task checkpoint Saikrishna2511/qwen-multitask is recommended.
  • This model is not intended for production use without thorough additional evaluation and testing.