---

license: apache-2.0

base_model:

- unsloth/Qwen3.5-0.8B

library_name: gguf

pipeline_tag: text-generation

tags:

- gguf

- llama.cpp

- quantized

- unsloth

- lora

- trl

- sft

---

Qwen3.5-0.8B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF

GGUF quantizations of a LoRA fine-tune of unsloth/Qwen3.5-0.8B, supervised fine-tuned on ermiaazarkhalili/claude-reasoning-distillation (private) (config sft).

Quantized from ermiaazarkhalili/Qwen3.5-0.8B-SFT-Claude-Opus-Reasoning-Unsloth. See that repository for the full-precision weights.

| | |

| --- | --- |

| Base model | unsloth/Qwen3.5-0.8B |

| Training data | ermiaazarkhalili/claude-reasoning-distillation (private) (config sft) |

| Method | LoRA supervised fine-tuning via Unsloth + TRL |

| License | apache-2.0 (inherited from the base model) |

Available quantizations

| File | Size |

| --- | --- |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q2_k.gguf | 422 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q3_k_m.gguf | 466 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q4_k_m.gguf | 529 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q5_k_m.gguf | 578 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q6_k.gguf | 630 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q8_0.gguf | 812 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf | 529 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth.q5_k_m.gguf | 578 MB |

| qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth.q8_0.gguf | 812 MB |

Usage

llama.cpp

huggingface-cli download ermiaazarkhalili/Qwen3.5-0.8B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q4_k_m.gguf --local-dir .
llama-cli -m qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q4_k_m.gguf -p "Explain gradient checkpointing in two sentences." -n 256

Ollama

echo 'FROM ./qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-v1.q4_k_m.gguf' > Modelfile
ollama create qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-gguf -f Modelfile
ollama run qwen3.5-0.8b-sft-claude-opus-reasoning-unsloth-gguf

Training configuration

| Setting | Value |

| --- | --- |

| LoRA rank (r) | 16 |

| LoRA alpha | 16 |

| Learning rate | 0.0002 |

| Epochs | 1 |

| Effective batch size | 8 (2 x 4 grad accum) |

| Max sequence length | 2048 |

| Base precision | 4-bit (QLoRA) |

| Target modules | q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj |

Limitations

• No benchmark evaluation has been run on this checkpoint. The only reported

numbers are training-loss observations.

• Inherits the biases, knowledge cutoff and failure modes of the base model.
• Fine-tuned on a single instruction-following dataset; behaviour outside that

distribution is untested.

• LoRA adapters were merged into the base weights, so the merged model cannot

be detached from this fine-tune.

Reproducing

Trained by notebooks/sft_distillation_qwen3.5_unsloth.ipynb, executed non-interactively with

papermill on a SLURM H100 partition (Unsloth + TRL, LoRA).

---

Card generated from the training run's own configuration and logs by

scripts/generate_hub_model_card.py.