---
base_model: unsloth/Qwen3.6-27B
base_model_relation: finetune
library_name: transformers
tags:
- transformers
- safetensors
- qwen3_5
- qwen3.6
- multimodal
- image-text-to-text
- unsloth
language:
- en
license: apache-2.0
pipeline_tag: image-text-to-text
---

![Ornstein-3.6-27B](ornstein3.6-27b.png)

# Ornstein-3.6-27B

A fine-tune of [Qwen 3.6 27B](https://huggingface.co/unsloth/Qwen3.6-27B), the dense multimodal (vision + text) member of the Qwen 3.6 family with hybrid linear + full attention. Part of the Ornstein series — reasoning- and agent-oriented fine-tunes built on a custom data curation pipeline.

> **GGUF quantizations available at [GestaltLabs/Ornstein-3.6-27B-GGUF](https://huggingface.co/GestaltLabs/Ornstein-3.6-27B-GGUF)** — Q8_0 down through aggressive 3-bit I-quants.

## Support This Work

I'm a PhD student in visual neuroscience at the University of Toronto who also happens to spend way too much time fine-tuning, merging, and quantizing open-weight models on rented H100s and a local DGX Spark. All training compute is self-funded — balancing GPU costs against a student budget. If my uploads have been useful to you, consider buying a PhD student a coffee. It goes a long way toward keeping these experiments running.

**[Support on Ko-fi](https://ko-fi.com/djlougen)**

---

## Details

- **Developed by:** GestaltLabs
- **Architecture:** `Qwen3_5ForConditionalGeneration` — Qwen 3.6 dense with linear + full attention interleaved (Gated Delta Net) + vision encoder
- **Parameters:** ~27B total (dense, multimodal)
- **Hidden size / layers:** 5120 / 64
- **Attention:** 24 heads, 4 KV heads, head_dim 256, full-attention every 4 layers (linear otherwise)
- **Context length:** 262,144 tokens
- **License:** Apache 2.0
- **Base model:** [unsloth/Qwen3.6-27B](https://huggingface.co/unsloth/Qwen3.6-27B)
- **Training framework:** Unsloth

## Usage

### Transformers

```python
from transformers import AutoModelForImageTextToText, AutoProcessor

model_id = "GestaltLabs/Ornstein-3.6-27B"
processor = AutoProcessor.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForImageTextToText.from_pretrained(
    model_id, torch_dtype="auto", device_map="auto", trust_remote_code=True
)

messages = [{"role": "user", "content": [{"type": "text", "text": "Write a haiku about hybrid attention."}]}]
inputs = processor.apply_chat_template(
    messages, add_generation_prompt=True, tokenize=True, return_tensors="pt"
).to(model.device)
out = model.generate(**inputs, max_new_tokens=512)
print(processor.batch_decode(out[:, inputs["input_ids"].shape[-1]:], skip_special_tokens=True)[0])
```

### llama.cpp (via GGUF)

See the [GGUF repo](https://huggingface.co/GestaltLabs/Ornstein-3.6-27B-GGUF) — pick a quant that fits your memory (Q4_K_M is a strong default for 24 GB cards).

## License

Apache 2.0 — inherited from the Qwen 3.6 base release.

[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)