ModelVerificadoExterno

Qwen3.5 9B SFT Claude Opus Reasoning Unsloth GGUF

GGUF quantizations of a LoRA fine-tune of unsloth/Qwen3.5-9B, supervised fine-tuned on ermiaazarkhalili/claude-reasoning-distillation (private) (config sft).

Fontehuggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF/tree/b96c426fa295cef7e7f13b71ebbe14a223bb9b53

TipoOtherModelo baseunsloth/Qwen3.5-9BVersãoRevision b96c426Publicador@ermiaazarkhaliliCnota

Fonte do modelo

Tipo: Other
Tamanho do arquivo: 3,56–8,87 GB
Modelo base: unsloth/Qwen3.5-9B
Versão: vRevision b96c426
Atualizado: 2 de out. de 2026
Publicador: @ermiaazarkhalili
Downloads: 226
Parâmetros: 9B
Tarefas: Geração de texto
Fonte: Hugging Face

Descrição da fonte

Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF

GGUF quantizations of a LoRA fine-tune of unsloth/Qwen3.5-9B, supervised fine-tuned on ermiaazarkhalili/claude-reasoning-distillation (private) (config sft).

Quantized from ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth. See that repository for the full-precision weights.


Base model	`unsloth/Qwen3.5-9B`
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
`qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q2_k.gguf`	3.83 GB
`qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q3_k_m.gguf`	4.62 GB
`qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf`	5.63 GB
`qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q5_k_m.gguf`	6.47 GB
`qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q6_k.gguf`	7.36 GB
`qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q8_0.gguf`	9.53 GB

Usage

llama.cpp

huggingface-cli download ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf --local-dir .
llama-cli -m qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf -p "Explain gradient checkpointing in two sentences." -n 256

Ollama

echo 'FROM ./qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf' > Modelfile
ollama create qwen3.5-9b-sft-claude-opus-reasoning-unsloth-gguf -f Modelfile
ollama run qwen3.5-9b-sft-claude-opus-reasoning-unsloth-gguf

Training configuration

Setting	Value
LoRA rank (r)	64
LoRA alpha	64
Learning rate	0.0002
Epochs	1
Effective batch size	8 (1 x 8 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`, `out_proj`

Observed training loss

Measured from our SLURM logs for this configuration. These are training-loss observations only — no downstream benchmark evaluation has been run on this model, so they should not be read as a quality claim.

SLURM job	Steps	First loss	Final loss
`37216727`	1,310	0.9383	0.7393

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_qwen3.5-9b_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.

Fontes

1 fonte

Hugging FacePrimárioVersão b96c426fa295cef7e7f13b71ebbe14a223bb9b53

huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF/tree/b96c426fa295cef7e7f13b71ebbe14a223bb9b53

Verificado 2 de out.

Artefatos de modelo

6 artefatos

qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q2_k.gguf

gguf · 3,56 GB · SHA-256 fefe9baa116b…476a · Hugging Face

Baixar

qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q3_k_m.gguf

gguf · 4,31 GB · SHA-256 e37e57413654…d0bb · Hugging Face

Baixar

qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf

gguf · 5,24 GB · SHA-256 e80746898cfc…b016 · Hugging Face

Baixar

qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q5_k_m.gguf

gguf · 6,02 GB · SHA-256 c631efe4ffbe…2927 · Hugging Face

Baixar

qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q6_k.gguf

gguf · 6,85 GB · SHA-256 4761e1ca51e7…414b · Hugging Face

Baixar

qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q8_0.gguf

gguf · 8,87 GB · SHA-256 1f0281c5d8a5…adc2 · Hugging Face

Baixar

Trechos de fonte

3 trechos

Readme2 de out.huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF

--- license: apache-2.0 base_model: - unsloth/Qwen3.5-9B library_name: gguf pipeline_tag: text-generation tags: - gguf - llama.cpp - quantized - unsloth - lora - trl - sft --- # Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF GGUF quantizations of a LoRA fine-tune of [`unsloth/Qwen3.5-9B`](https://huggingface.co/unsloth/Qwen3.5-9B), supervised fine-tuned on `ermiaazarkhalili/claude-reasoning-distillation` (private) (config `sft`). Quantized from [`ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth`](https://huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth). See that repository for the full-precision weights. | | | | --- | --- | | **Base model** | [`unsloth/Qwen3.5-9B`](https://huggingface.co/unsloth/Qwen3.5-9B) | | **Training data** | `ermiaazarkhalili/claude-reasoning-distillation` (private) (config `sft`) | | **Method** | LoRA supervised fine-tuning via [Unsloth](https://github.com/unslothai/unsloth) + [TRL](https://github.com/huggingface/trl) | | **License** | `apache-2.0` (inherited from the base model) | ## Available quantizations | File | Size | | --- | --- | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q2_k.gguf` | 3.83 GB | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q3_k_m.gguf` | 4.62 GB | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf` | 5.63 GB | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q5_k_m.gguf` | 6.47 GB | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q6_k.gguf` | 7.36 GB | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q8_0.gguf` | 9.53 GB | ## Usage ### llama.cpp ```bash huggingface-cli download ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf --local-dir . llama-cli -m qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf -p "Explain gradient checkpointing in two sentences." -n 256 ``` ### Ollama ```bash echo 'FROM ./qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf' > Modelfile ollama create qwen3.5-9b-sft-claude-opus-reasoning-unsloth-gguf -f Modelfile ollama run qwen3.5-9b-sft-claude-opus-reasoning-unsloth-gguf ``` ## Training configuration | Setting | Value | | --- | --- | | LoRA rank (r) | 64 | | LoRA alpha | 64 | | Learning rate | 0.0002 | | Epochs | 1 | | Effective batch size | 8 (1 x 8 grad accum) | | Max sequence length | 2048 | | Base precision | 4-bit...

Readme11 de jul.huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF

--- base_model: ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth tags: - gguf - llama.cpp - unsloth - qwen3.5 - sft - distillation - reasoning - quantized license: apache-2.0 language: - en datasets: - ermiaazarkhalili/claude-reasoning-distillation pipeline_tag: text-generation --- # Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth — GGUF quantized GGUF quantizations of [`ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth`](https://huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth), produced via [Unsloth](https://github.com/unslothai/unsloth) + llama.cpp's conversion scripts. | Field | Value | |---|---| | **Source checkpoint** | [`ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth`](https://huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth) | | **Base model** | [`unsloth/Qwen3.5-9B`](https://huggingface.co/unsloth/Qwen3.5-9B) | | **Dataset** | [`claude-reasoning-distillation`](https://huggingface.co/datasets/ermiaazarkhalili/claude-reasoning-distillation) (10,477 samples) | | **Architecture** | Qwen3.5 hybrid transformer + state-space (SSM) blocks | | **Training** | 1 full epoch (effective batch=8 via per_device=1 × grad_accum=8) | | **LoRA** | r=64, α=64 (Jackrong recipe), dropout=0 | | **Final training loss** | 0.795 (job 37216727, runtime 2h 32m on H100 MIG 3g.40gb, peak VRAM 20.25 GB) | | **Conversion** | Unsloth `save_pretrained` → llama.cpp `convert_hf_to_gguf.py` → `llama-quantize` | | **Quantization tool** | llama.cpp `llama-quantize` (cached toolchain) | ## Available quantizations | File | Bits | Size | Notes | |---|---|---|---| | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q2_k.gguf` | 2-bit | 3.83 GB | Smallest; aggressive quality loss | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q3_k_m.gguf` | 3-bit | 4.62 GB | Small; noticeable quality loss | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q4_k_m.gguf` | 4-bit | 5.63 GB | **Recommended** — best size/quality balance | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q5_k_m.gguf` | 5-bit | 6.47 GB | Near-full quality | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q6_k.gguf` | 6-bit | 7.36 GB | Very close to Q8_0 at smaller size | | `qwen3.5-9b-sft-claude-opus-reasoning-unsloth.q8_0.gguf` | 8-bit | 9.53 GB | Largest; closest to bf16 source | **Recommended default:** `Q4_K_M` (4-bit, K-quant medium...

Note2 de out.huggingface.co/ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF

Source context: 226 downloads · 0 likes · Pipeline text-generation · Library gguf · Repo ermiaazarkhalili/Qwen3.5-9B-SFT-Claude-Opus-Reasoning-Unsloth-GGUF