Qwen3.5-35B-A3B-Opus-Reasoning-Distilled-v2-GGUF

Model Description

This is the Q4_K_M GGUF quantized version of ponytang3/Qwen3.5-35B-A3B-Opus-Reasoning-Distilled-v2.

Original Model

• Base Model: unsloth/Qwen3.5-35B-A3B
• Fine-tuned Model: ponytang3/Qwen3.5-35B-A3B-Opus-Reasoning-Distilled-v2
• Quantization: Q4_K_M (4-bit quantization)

Training Details

• Method: bf16 LoRA + response-only (train_on_responses_only)
• LoRA Rank: 16
• Epochs: 2
• Max Sequence Length: 4096
• Framework: Unsloth + TRL

Datasets

• nohurry/Opus-4.6-Reasoning-3000x-filtered
• Jackrong/Qwen3.5-reasoning-700x
• Roman1111111/claude-opus-4.6-10000x

Usage with llama.cpp

./llama-cli -m model-q4_k_m.gguf -p "Your prompt here" -n 512

Format

The model uses <think>...</think> tags for chain-of-thought reasoning.