oodbilav1980/ athenea-4b-coding:latest

49 2 days ago

Athenea-4B-Coding is a fine-tuned version of huihui-ai/Huihui-Qwen3-4B-Thinking-2507-abliterated, specialized in code reasoning, debugging, and problem solving. 

ollama run oodbilav1980/athenea-4b-coding

Details

2 days ago

e3ca0e8e9f04 · 2.9GB ·

model
qwen3
·
4.02B
·
Q5_K_M
template
{{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}{{ range .Messages }}<|im_start|
system
You are Athenea, a coding assistant based on Qwen3. You provide concise, bug-free code solutions.
params
{ "num_ctx": 16384, "repeat_penalty": 1.1, "temperature": 0.2, "top_p": 0.9 }

Readme

This is Q5_K_M quant of mradermacher/Athenea-4B-Coding-i1-GGUF model.

Here’s the Huggingface Repo for the Quant.

Athenea-4B-Coding is a fine-tuned version of huihui-ai/Huihui-Qwen3-4B-Thinking-2507-abliterated, specialized in code reasoning, debugging, and problem solving. Trained on high-quality programming data with explicit reasoning traces using and tags, the model is designed to perform detailed step-by-step reasoning for software development, algorithm design, and code comprehension tasks.

🎯 Model Description Athenea-4B-Coding extends Huihui-Qwen3’s structured reasoning capabilities into programming-related domains, showing strong performance on logical problem-solving, code completion, and debugging scenarios.

Key features:

Step-by-step code reasoning within blocks; Specialization in algorithmic and debugging tasks; Uncensored output generation for full reasoning visibility; Improved logical consistency through focused fine-tuning; Compatible with open inference frameworks (Transformers, vLLM, etc.)