Instructions to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-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 nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-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 nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0 # Run inference directly in the terminal: llama cli -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0 # Run inference directly in the terminal: llama cli -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0 # Run inference directly in the terminal: ./build/bin/llama cli -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
Use Docker
docker model run hf.co/nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
- Llama new
- LM Studio
- Jan
- vLLM
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-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": "nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
- Ollama
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF with Ollama:
ollama run hf.co/nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
- Unsloth Desktop
- Pi
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
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:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF with Docker Model Runner:
docker model run hf.co/nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
- Lemonade
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
Run and chat with the model
lemonade run user.Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF-Q4_0
List all available models
lemonade list
- Hermes Agent
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-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 nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
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://127.0.0.1:9931/v1 hermes config set model.default nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0
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://127.0.0.1:9931/v1 \ --custom-model-id "nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF:Q4_0" \ --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"
Swift-1.5-Qwen3.8-Flash-Next Q4_0-Q8_0-out v3 (GGUF)
โ ๏ธ Runtime Compatibility Notice
This is a
qwen4exparchitecture GGUF (Gated-DeltaNet hybrid + 512-expert routed MoE + MTP). Stockllama.cppdoes not support this architecture. Use aqwen4exp-enabled llama.cpp build (e.g. theunslothai/llama.cppfork with qwen4exp support) with--mtp-draftfor speculative decoding.
Overview
Swift 1.5 Flash Miroslav2.0 โ the Swift-optimized (UkisAI "Swift" reasoning-token reduction, ~58% fewer reasoning tokens, hesitation-loop elimination) variant of Qwen3.8 Flash Next, quantized for local serving on Apple Silicon.
- Build: Q4_0 base re-quantized with Q8_0 donor tensors spliced into the output/embedding/residual
tensors (
-Q8out-v3pipeline: 575 tensors upgraded to Q8_0 from the full-precision donor shards), imatrix-calibrated (imatrix-bartv6, calibration datasetQwen3.8-Flash-Next-calibration-v6) - Architecture:
qwen4expโ 48 layers (GDN recurrent + full attention every 4th), 512 routed experts (10 active), shared expert, hyper-connections, ngram-PLE embedding, built-in MTP heads - Context: 262,144 tokens native
- Shards: 3 (
-00001-of-00003.ggufโฆ-00003-of-00003.gguf), 95 GiB total - Vision: vision projector tensors included (image-text-to-text capable)
Files
| File | Size |
|---|---|
Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-00001-of-00003.gguf |
42.4 GiB |
Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-00002-of-00003.gguf |
42.8 GiB |
Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-00003-of-00003.gguf |
10.3 GiB |
Download all three shards into the same directory, then load shard 1.
Usage (qwen4exp-enabled llama.cpp)
llama-server -m Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-00001-of-00003.gguf \
--mmproj mmproj-f16.gguf -ngl 99 -c 262144 --port 8080
Built and benchmarked locally on Apple M5 Pro (64 GB unified memory) with the Slipstream streaming-MoE llama.cpp fork.
- Downloads last month
- 5,859
4-bit