Claude Opus 4.5 Distill LoRA on Qwen 3.6 27B
A LoRA for Qwen 3.6 27B distilled on a Claude 4.5 Opus reasoning dataset, trained entirely on AMD Strix Halo.
Overview
This model is Qwen 3.6 27B fine-tuned to reason and respond in the style of Claude 4.5 Opus, making it more enjoyable for interactive use and coding.
Training Details
| Parameter | Value |
|---|---|
| Base Model | Qwen 3.6 27B |
| Dataset | claude-4.5-opus-high-reasoning-250x |
| Sequence Length | 4096 |
| Training Time | ~3 hours |
| Hardware | AMD Strix Halo |
Hardware Setup
The model was trained using Unloth Studio on Strix halo. Setup guide: APPENDIX_STRIX_HALO.md
License
- Downloads last month
- 21
Hardware compatibility
Log In to add your hardware
We're not able to determine the quantization variants.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support