File size: 2,865 Bytes
741de50 13f6141 84a900a 48d2f79 84a900a 48d2f79 84a900a f7e3d24 84a900a 48d2f79 84a900a 40e93a5 84a900a 40e93a5 84a900a 40e93a5 84a900a 40e93a5 84a900a 48d2f79 84a900a 48d2f79 84a900a 40e93a5 84a900a 40e93a5 84a900a 40e93a5 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 | ---
base_model: GestaltLabs/Ornstein-3.6-27B
language:
- en
license: apache-2.0
pipeline_tag: text-generation
tags:
- qwen3.5
- qwen3.6
- rys
- text-generation
- transformers
- safetensors
- canada
- sovereign-ai
---
[
# Ornstein-3.6-27B-RYS
RYS-enhanced variant of the Ornstein-3.6-27B dense model. Layer 33 is duplicated using the **Repeat Your Self (RYS)** method, improving reasoning and instruction-following performance without increasing active parameter count at inference time.
> **GGUF quantizations:** [GestaltLabs/Ornstein-3.6-27B-RYS-GGUF](https://huggingface.co/GestaltLabs/Ornstein-3.6-27B-RYS-GGUF)
## About Gestalt Lab
We are a proudly Canadian research collective working to advance **sovereign Canadian AI** — open-weight models that Canadians (and everyone else) can run locally, study, and build on without dependence on closed foreign APIs. All training, fine-tuning, and quantization is done on local and self-funded compute. By supporting this work, you help keep frontier model development accessible, transparent, and under Canadian stewardship.
## Important: requires a patched llama.cpp
RYS duplicates one of the middle layers, which breaks the hardcoded `full_attention_interval = 4` assumption in stock llama.cpp's Qwen3.5 loader. This model is converted with **per-layer `head_count_kv` baked in**, and you need a llama.cpp that reads that per-layer metadata instead of falling back to the interval formula.
**Patched fork:** [https://github.com/DJLougen/llama.cpp](https://github.com/DJLougen/llama.cpp) (default branch `rys-qwen35`, fully backward-compatible).
Stock llama.cpp, Ollama, LM Studio, and any other inference runtime built on stock llama.cpp will currently fail to load this model with a `check_tensor_dims` error — this is expected until/unless the patch is upstreamed.
## Support This Work
Our training compute is entirely self-funded. If this model is useful to you, consider supporting the lab:
**[Support on Ko-fi](https://ko-fi.com/djlougen)**
* * *
## Model Details
* **Architecture:** Qwen3.5 dense
* **Parameters:** ~27B active
* **Layers:** 65 (64 original + 1 RYS-duplicated layer 33)
* **Context length:** 131,072 tokens
* **License:** Apache-2.0
## Usage
### Build the patched llama.cpp
```bash
git clone https://github.com/DJLougen/llama.cpp.git
cd llama.cpp
git checkout rys-qwen35
cmake -B build -DGGML_CUDA=ON -DCMAKE_BUILD_TYPE=Release
cmake --build build -j
```
Drop `-DGGML_CUDA=ON` for a CPU-only build. The patch touches the GGUF loader; backend selection is independent.
### Download + run
```bash
./build/bin/llama-server \
-m ornstein-3.6-27b-rys-q4_k_m.gguf \
--host 0.0.0.0 --port 8080 \
--n-gpu-layers 99 --ctx-size 131072 \
--flash-attn on --jinja \
-ctk q4_0 -ctv q4_0
```
## License
Apache 2.0
|