Modelpublic

Fazmin/solus_v1_qwopus-coder-3b-q4

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face

1likes16downloads

Model Card

QwOpus Coder 3B — Solus v1

A compact coding model that reasons before it answers. It is Qwen2.5-Coder-3B-Instruct fine-tuned with LoRA on roughly 3,200 Claude Opus 4.6 reasoning trajectories, which teaches it to work a problem through inside <think> tags and then give a clean final answer.

The result is a 3B model with a visible chain of thought that still runs fast on consumer hardware, including 4 GB VRAM laptops. Its limits are the ones you would expect at this scale: it is not the right tool for large multi-file changes or system design.

Specifications


Parameters	3B
Quantization	Q4KM
File size	1.80 GB
Minimum RAM	5.00 GB
Minimum VRAM	not required
Context length	8,192 tokens
SHA-256	`b31bbeb1c0ff218fb3e093325cf96596ee9df5db4ca0fcd7ecb4f3140275a7db`

Single file: qwen2.5-coder-3b-instruct.Q4_K_M.gguf

Quantization

Quantization performed at the Faculty of Engineering, McMaster University.

The GGUF conversion this build is derived from was produced by ryzdfm, and the weights here are a byte-for-byte copy of that file — the SHA-256 above matches the upstream artifact.

Provenance

—Original model: Qwen/Qwen2.5-Coder-3B-Instruct
—Upstream GGUF: ryzdfm/qwen2.5-coder-3b-claude_opus_4.6-distilled
—Mirrored for Solus, a desktop app for running language models entirely on your own machine.

Usage

bash

llama-cli -m qwen2.5-coder-3b-instruct.Q4_K_M.gguf -cnv