brnpistone/ Qwen3.5-4B-AgentCoder-q6-k:latest

351 2 weeks ago

A Quantized, Fine-Tuned Model for Enhanced Tool Calling, Code Generation, and Reasoning

tools thinking 
ollama run brnpistone/Qwen3.5-4B-AgentCoder-q6-k

Details

2 weeks ago

c73a04c05f45 · 3.5GB ·

model
qwen35
·
4.21B
·
Q6_K
template
{{ .Prompt }}
system
You are a helpful assistant. Only use tools when they are explicitly provided to you. If no tools ar
params
{ "min_p": 0, "repeat_last_n": 256, "repeat_penalty": 1.15, "stop": [ "<|im_

Readme

🧠 Qwen3.5-4B-AgentCoder-Q6-GGUF

A Quantized, Fine-Tuned Model for Enhanced Tool Calling, Code Generation, and Reasoning

Model Description

Qwen3.5-4B-AgentCoder-Q6-K is a fine-tuned version of the Qwen/Qwen3.5-4B model, optimized for: - 🧮 Complex reasoning tasks - 🧰 Tool calling - 💻 Code generation

The model was developed through sequential fine-tuning, followed by a Direct Preference Optimization (DPO) post-training stage to improve alignment, coherence, and reasoning accuracy.

Highlights

  • Post-trained with DPO using chosen/rejected pairs for better alignment
  • Excellent balance between tool use, code generation, and reasoning

🚀 Direct Use

Qwen3.5-4B-AgentCoder-Q6-K can be used directly for: - ✅ Tool calling in complex reasoning tasks - ✅ Code generation for Python, JS, and other languages - ✅ Multi-domain reasoning (math, logic, Q&A)

⚠️ Out-of-Scope Use

  • ❌ Highly sensitive or confidential data
  • ❌ Domains requiring expert-level specialization
  • ❌ Tasks where full explainability is mandatory

💻 Getting Started

ollama run brnpistone/Qwen3-4B-AgentCoder-q6-k

🧠 Training Details

Training Procedure

Phase 1 — Post-Training - Direct Preference Optimization (DPO)

After sequential fine-tuning, the model underwent a DPO phase to enhance response alignment, reasoning robustness, and factual consistency.

  • Learning rate: 3e-6
  • Batch size: 1
  • Gradient accumulation: 4
  • Epochs: 1
  • Beta: 0.1
  • Loss type: sigmoid
  • Warmup steps: 27
  • Sequence length: ~2.5K tokens
DPO Data
  • ~2.5K chosen/rejected response pairs
  • Rejected samples synthetically generated to represent poor or incoherent answers
  • Chosen samples tagged from real conversations

Objective - Encourage the model to prefer chosen completions - Improve clarity, correctness, and helpfulness - Reduce hallucinations and verbosity

🖥️ Technical Specifications

Model Architecture

  • Model type: Causal language model
  • Parameters: 4.0B
  • Context length: ~264K tokens
  • Thinking mode: Enabled

Compute Infrastructure

Hardware - GPU: NVIDIA H100 (80 GB VRAM)
- System RAM: 2 TiB
- Memory per vCPU: 10.67 GiB

Software - Python: 3.12
- Transformers: 5.3.0
- Libraries: bitsandbytes, safetensors, torch, trl, scikit-learn, tokenizers, psutil, py7zr

🧾 Citation

BibTeX

@article{qwen3.5-4b-thinking-2507-toolcode,
  title={Qwen3.5-4B-AgentCoder-Q6-K: A Fine-Tuned Model for Enhanced Tool Calling, Code Generation, and Reasoning},
  author={Bruno Pistone},
  year={2025},
  journal={Hugging Face Model Hub}
}

APA

Bruno Pistone. (2026). Qwen3.5-4B-AgentCoder-Q6-K: A Fine-Tuned Model for Enhanced Tool Calling, Code Generation, and Reasoning.

🧭 Recommendations

  • Tool use accuracy depends on task complexity
  • Code generation may occasionally produce minor syntax issues
  • Reasoning strongest in structured, logical, and mathematical contexts
  • Avoid using this model for confidential or safety-critical applications

🧠 Qwen3.5-4B-AgentCoder — created by Bruno Pistone
Enhanced reasoning, tool calling, and code generation — refined with DPO alignment