Instructions to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-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 barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

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 barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

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 barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

Use Docker

docker model run hf.co/barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-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": "barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

Ollama
How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with Ollama:
```
ollama run hf.co/barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M
```

Unsloth Studio

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF to start chatting

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with Docker Model Runner:
```
docker model run hf.co/barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M
```

Lemonade

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-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 barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

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 barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M

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 "barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF:Q4_K_M" \
  --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"

barozp commited on 28 days ago

Commit

0d150da

verified ·

1 Parent(s): c57c5eb

Add model card with quant list and measured speculative-decoding benchmarks

Browse files

Files changed (1) hide show

README.md +59 -0

README.md ADDED Viewed

	@@ -0,0 +1,59 @@

+---
+license: apache-2.0
+base_model_relation: quantized
+pipeline_tag: text-generation
+tags:
+- gguf
+- llama.cpp
+- quantized
+- moe
+- qwen3
+- reasoning
+---
+# Qwen3.6-29B-REAP-Opus-Reasoning-Distill-GGUF
+GGUF quantizations of a REAP-pruned (205/256 experts) Qwen3.6-35B-A3B MoE, merged with an Opus-reasoning-distilled LoRA adapter. This is the **plain merge — no Multi-Token Prediction (MTP) head**.
+A version with an MTP head grafted on (for self-speculative decoding) is available at [barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-MTP-GGUF](https://huggingface.co/barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-MTP-GGUF). See the benchmarks below to decide which one fits your hardware.
+Converted with `llama.cpp`'s `convert_hf_to_gguf.py`.
+## Files
+| Quant | Size | Notes |
+|---|---:|---|
+| BF16 | 56.5 GB | full precision reference |
+| Q8_0 | 30.1 GB | near-lossless |
+| Q6_K | 23.2 GB | |
+| Q5_K_M | 20.2 GB | |
+| Q4_K_M | 17.3 GB | most popular K-quant |
+| Q4_K_S | 16.2 GB | |
+| IQ4_XS | 15.3 GB | imatrix-based, smaller & often better than Q4_K_S |
+| Q3_K_M | 13.7 GB | |
+| IQ3_M | 12.6 GB | |
+| IQ3_XXS | 11.2 GB | |
+| Q2_K | 10.6 GB | |
+| IQ2_M | 9.6 GB | |
+All quants ≤Q4_K_M were calibrated with an importance matrix (imatrix) built from 512 samples of [barozp/opus-reasoning-distill-train](https://huggingface.co/datasets/barozp/opus-reasoning-distill-train). K-quants (`Q*_K*`) favor broad compatibility and fast CPU inference; IQ-quants (`IQ*`) require the imatrix and give better quality per bit at ≤4-bit, at some CPU-inference speed cost.
+## Why choose this over the MTP version?
+The MTP-grafted version carries an extra decoder layer used for self-speculative decoding. When that feature is left enabled on hardware that's memory-bandwidth-constrained (e.g. a small/weaker GPU with heavy CPU offload), the draft+verify overhead can compete with an already-scarce resource and result in **slower** generation than this plain model. This release removes that footgun entirely — no toggle to remember, always the same speed as the MTP file with speculative decoding off.
+## Benchmarks
+Measured with `llama-cli` (`Q4_K_M`, flash attention on, greedy decoding, 5 runs per config, mean ± std) on an NVIDIA RTX PRO 6000 Blackwell Server Edition (97 GB VRAM). Full methodology and the MTP comparison are on the [MTP-GGUF model card](https://huggingface.co/barozp/Qwen3.6-29B-REAP-Opus-Reasoning-Distill-MTP-GGUF#benchmarks).
+| `-ngl` | tok/s (generation) |
+|---:|---:|
+| 99 (full offload) | 219.06 ± 0.10 |
+| 20 (partial offload) | 58.60 ± 0.52 |
+This matches the MTP-GGUF file with speculative decoding disabled (`--spec-type none`) within measurement noise — confirming the MTP head, when unused, carries no VRAM/compute penalty. If your setup benefits from speculative decoding (compute-bound, strong GPU, full offload), the MTP release may be faster; see its model card for numbers (+39–67% observed in our tests).
+## Credits
+- Base architecture: Qwen3.6-35B-A3B (MoE), pruned via [REAP](https://huggingface.co/papers) (Router-weighted Expert Activation Pruning) to 205/256 experts.
+- Reasoning distillation: LoRA fine-tune on Opus-generated reasoning traces ([barozp/opus-reasoning-distill-train](https://huggingface.co/datasets/barozp/opus-reasoning-distill-train)).