Instructions to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16 # Run inference directly in the terminal: llama cli -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16 # Run inference directly in the terminal: llama cli -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16 # Run inference directly in the terminal: ./llama-cli -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16 # Run inference directly in the terminal: ./build/bin/llama-cli -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Use Docker
docker model run hf.co/jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
- LM Studio
- Jan
- vLLM
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
- Ollama
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with Ollama:
ollama run hf.co/jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
- Unsloth Desktop
- Pi
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:8080/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with Docker Model Runner:
docker model run hf.co/jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
- Lemonade
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Run and chat with the model
lemonade run user.Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF-BF16
List all available models
lemonade list
- Hermes Agent
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:8080/v1 hermes config set model.default jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:8080/v1 \ --custom-model-id "jackasda211233/Qwen3.5-27B-Uncensored-RYS-Reasoner-GGUF:BF16" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -299,7 +299,14 @@ Prompt: Build a more complex Python CLI tool with 8 commands, 4 question types (
|
|
| 299 |
|
| 300 |
**Severity assessment of bugs:** All bugs found in Test 2 are **surface-level generation quality issues**, not architectural or logic errors. The scoring pipeline, knowledge tracker, and data persistence all work correctly. The MCQ header parsing and T/F duplicate issues would likely be fixed in a single follow-up prompt ("fix the question text including markdown headers" / "fix the T/F for/else logic"), requiring minimal effort. No bugs required architectural changes.
|
| 301 |
|
| 302 |
-
**Note on the 80k generation:** The Quiz Engine was completed entirely autonomously
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 303 |
|
| 304 |
**Note on alternative approaches:** During development, the author also attempted a more "stable" variant using clean base Qwen3.5-27B weights in the duplicated zone (instead of dnhkng's FP8-origin layers). Ironically, the variant with standard base model layers performed worse on coding tasks — exhibiting looping behavior and variable scoping errors on the same prompts that this Splice model handled cleanly. The dnhkng FP8-origin layers in the reasoning zone appear to contribute meaningfully to coding stability.
|
| 305 |
|
|
|
|
| 299 |
|
| 300 |
**Severity assessment of bugs:** All bugs found in Test 2 are **surface-level generation quality issues**, not architectural or logic errors. The scoring pipeline, knowledge tracker, and data persistence all work correctly. The MCQ header parsing and T/F duplicate issues would likely be fixed in a single follow-up prompt ("fix the question text including markdown headers" / "fix the T/F for/else logic"), requiring minimal effort. No bugs required architectural changes.
|
| 301 |
|
| 302 |
+
**Note on the 80k generation:** The Quiz Engine was completed entirely autonomously across ~80k tokens with zero human intervention. The model:
|
| 303 |
+
1. Created the full project structure and wrote all code
|
| 304 |
+
2. Set up a Python virtual environment
|
| 305 |
+
3. Ran its own test suite, reviewed the results, and debugged failures
|
| 306 |
+
4. When asked to deliver the project as a zip file — and with no `zip` utility installed on the system — the model independently found an alternative compression method (tar.gz) and used it
|
| 307 |
+
5. Autonomously SSH'd into a separate machine and transferred the compressed project to a specific download folder
|
| 308 |
+
|
| 309 |
+
This demonstrates strong agentic capability: problem-solving around missing tools, cross-machine file operations, and sustained multi-step task completion without guidance.
|
| 310 |
|
| 311 |
**Note on alternative approaches:** During development, the author also attempted a more "stable" variant using clean base Qwen3.5-27B weights in the duplicated zone (instead of dnhkng's FP8-origin layers). Ironically, the variant with standard base model layers performed worse on coding tasks — exhibiting looping behavior and variable scoping errors on the same prompts that this Splice model handled cleanly. The dnhkng FP8-origin layers in the reasoning zone appear to contribute meaningfully to coding stability.
|
| 312 |
|