license, language, base_model, pipeline_tag, library_name, tags

license	language	base_model	pipeline_tag	library_name	tags
apache-2.0	en	GetSoloTech/Qwen3-Code-Reasoning-4B	text-generation	transformers	text-generation-inference Coder Code

Qwen3-Code-Reasoning-4B-f32-GGUF

GetSoloTech/Qwen3-Code-Reasoning-4B is a 4-billion parameter language model fine-tuned from Qwen3-4B-Thinking-2507 with LoRA adapters, specifically optimized for competitive programming and advanced code reasoning tasks. Trained on the high-quality Code-Reasoning dataset, including problems from TACO, APPS, CodeContests, and Codeforces, this model is engineered to deliver detailed, step-by-step solutions with ≥50% test case pass rates, improved constraint handling, and structured output for both code generation and complex reasoning scenarios. With configurable context length up to 262,144 tokens, it inherits the reasoning strengths of its base model and is recommended for users seeking well-justified code solutions and enhanced comprehensiveness in programming problem-solving.

Model Files

File Name	Quant Type	File Size
Qwen3-Code-Reasoning-4B.BF16.gguf	BF16	8.05 GB
Qwen3-Code-Reasoning-4B.F16.gguf	F16	8.05 GB
Qwen3-Code-Reasoning-4B.F32.gguf	F32	16.1 GB
Qwen3-Code-Reasoning-4B.Q2_K.gguf	Q2_K	1.67 GB
Qwen3-Code-Reasoning-4B.Q3_K_L.gguf	Q3_K_L	2.24 GB
Qwen3-Code-Reasoning-4B.Q3_K_M.gguf	Q3_K_M	2.08 GB
Qwen3-Code-Reasoning-4B.Q3_K_S.gguf	Q3_K_S	1.89 GB
Qwen3-Code-Reasoning-4B.Q4_K_M.gguf	Q4_K_M	2.5 GB
Qwen3-Code-Reasoning-4B.Q4_K_S.gguf	Q4_K_S	2.38 GB
Qwen3-Code-Reasoning-4B.Q5_K_M.gguf	Q5_K_M	2.89 GB
Qwen3-Code-Reasoning-4B.Q5_K_S.gguf	Q5_K_S	2.82 GB
Qwen3-Code-Reasoning-4B.Q6_K.gguf	Q6_K	3.31 GB
Qwen3-Code-Reasoning-4B.Q8_0.gguf	Q8_0	4.28 GB

Quants Usage

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):