lunahr / Qwen3-0.6B-Code-Expert-abliterated
Text Generation596M2290Updated May 16, 2025
Downloads
229
Likes
0
Last updated
May 16, 2025
License
apache-2.0
This project performs full fine-tuning on the Qwen3-0.6B language model to enhance its code reasoning and generation capabilities. Training was conducted exclusively on the nvidia/OpenCodeReasoning dataset, and the model was optimized using the bfloat16 (bf16) data type. Addit…
Tasks & library
- text generation
- transformers
Tags
- transformers
- qwen3
- text-generation
- unsloth
- trl
- sft
Technical specifications
Parameters596M
Librarytransformers
Languagesen
Base modelsuayptalha/Qwen3-0.6B-Code-Expert
Licenseapache-2.0