Instructions to use NaamFung/ZatFung 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 NaamFung/ZatFung 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 NaamFung/ZatFung:IQ3_XXS # Run inference directly in the terminal: llama cli -hf NaamFung/ZatFung:IQ3_XXS
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf NaamFung/ZatFung:IQ3_XXS # Run inference directly in the terminal: llama cli -hf NaamFung/ZatFung:IQ3_XXS
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 NaamFung/ZatFung:IQ3_XXS # Run inference directly in the terminal: ./build/bin/llama cli -hf NaamFung/ZatFung:IQ3_XXS
Use Docker
docker model run hf.co/NaamFung/ZatFung:IQ3_XXS
- Llama new
- LM Studio
- Jan
- Ollama
How to use NaamFung/ZatFung with Ollama:
ollama run hf.co/NaamFung/ZatFung:IQ3_XXS
- Unsloth Desktop
- Pi
How to use NaamFung/ZatFung with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf NaamFung/ZatFung:IQ3_XXS
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": "NaamFung/ZatFung:IQ3_XXS" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use NaamFung/ZatFung with Docker Model Runner:
docker model run hf.co/NaamFung/ZatFung:IQ3_XXS
- Lemonade
How to use NaamFung/ZatFung with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull NaamFung/ZatFung:IQ3_XXS
Run and chat with the model
lemonade run user.ZatFung-IQ3_XXS
List all available models
lemonade list
- Hermes Agent
How to use NaamFung/ZatFung with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf NaamFung/ZatFung:IQ3_XXS
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 NaamFung/ZatFung:IQ3_XXS
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use NaamFung/ZatFung with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf NaamFung/ZatFung:IQ3_XXS
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 "NaamFung/ZatFung:IQ3_XXS" \ --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"
ZatFung ยท Qwen3.8-Flash-Next๏ผไธญๆ + ไปฃ็ ๏ผ๏ผSwift-1.5 ไบๆกฃ + Unsloth Dynamic 3.1 320E
ๆฌไปๅบๆไธคไธชๆฅๆบใไธค็งๅ่๏ผZatFung-โฆ ๅ็ผ็ๆฏไธชๅญ็ฎๅฝ้ฝๆฏไธไธช็ฌ็ซๅฏ็จ็ๆจกๅ๏ผ
| ็ณปๅ | ไธๆธธ | ไธๅฎถๆกฃ | ๅญ็ฎๅฝ |
|---|---|---|---|
| Swift-1.5๏ผIQ3_S๏ผ | SC117/Swift-1.5-Qwen3.8-Flash-Next-GSQ-RCO-abliterated-GGUF | 192 / 256 / 288 / 320 / 384 | 5 |
| Unsloth Dynamic 3.1๏ผUD-IQ3_XXS๏ผ | SC117/Qwen3.8-Flash-Next-abliterated-UD-GGUF | 320 | 1๏ผๆจ่๏ผ |
Swift ็ณปๅๆฏ IQ3_S ๅบ็ก้ๅ๏ผๅจๅๅง 512 ไธชไธๅฎถไธญไฟ็ 192 / 256 / 288 / 320 / 384 ไธช๏ผๅ ฑไบๆกฃ๏ผ ไฟ็้จๅ็ๆ้ไฟๆๅๅง้ๅ็ฒพๅบฆ๏ผๆช้ๆฐ้ๅใUD ็ณปๅๆฏ้ๅฑๅจๆ้ๅ๏ผๅไฝ็งฏไธ่ดจ้ๆดๅฅฝใ ้ๅบฆ็จๆ ขโโ่งไธ้ข็ไธ่ใ
ๆฌไปๅบๆจกๅ็ไผๅฟๆฏใไธญๆ + ไปฃ็ ใ๏ผๆฌ็ๆฌ้ๅฏนไธญๆ้ฟๆใ็ๅฎไปฃ็ ๏ผZig / Go / Rust / C++๏ผไธ ็ฎๆณ้ข่ฟไธ็ฑป่ด่ฝฝๅไบๅ่๏ผๅ ๆญคๅจ่ฟไธค็ฑปๅบๆฏไธ๏ผๅไฝ็งฏ็่กจ็ฐๆๆพๅฅฝไบๆ้็จ่ฏญๆๅ่็็ๆฌใ ้ๅไปฃ็ ็ผ็จ๏ผไปๅบ็บงไปปๅกใๅค่ฏญ่จใ็ฎๆณ้ข๏ผไธๅฐ่ฏดๅไฝ๏ผ้ฟไธญๆใ้่ฆ็จณๅฎๆ้ฃไธ่ฟ่ดฏไธไธๆ๏ผใ
ไบๆกฃ่งๆ ผ
| ็ฎๅฝ | ไธๅฎถๆฐ | ็ๅบ้ๅๅทฎ โ | ไธ็ๅ่ฎก | ๅปบ่ฎฎ |
|---|---|---|---|---|
ZatFung-Swift-384E-v5/ |
384 | 0.88% | 66.05 GiB | ่ดจ้ไผๅ |
ZatFung-Swift-320E-v5/ |
320 | 2.27% | 60.16 GiB | ๆจ่๏ผ่ดจ้/ไฝ็งฏๅนณ่กก |
ZatFung-Swift-288E-v5/ |
288 | 3.18% | 57.22 GiB | ๅฐๆพๅญ / ๅฐๅ ๅญ |
ZatFung-Swift-256E-v5/ |
256 | 4.55% | 54.27 GiB | ๆ้ๅ็ผฉ |
ZatFung-Swift-192E-v5/ |
192 | 8.59% | 48.39 GiB | ไป ๅฎ้ช๏ผไธๅปบ่ฎฎ็ไบง |
ๅปบ่ฎฎไฝฟ็จ 320 ๆกฃๅไปฅไธใ ใ็ๅบ้ๅๅทฎใๆฏๅจๆชๅไธๅ่็็ๅบ้ไธ๏ผไธๅๅง 512 ไธๅฎถๆจกๅ็ธๆฏ ่พๅบ่ทฏๅพๅ็ๆนๅ็ๆฏไพโโ่ถไฝ่ถๆฅ่ฟๅๆจกๅใ192 ๆกฃๅทฒๆฅ่ฟ 9%๏ผๆฅๅธธไฝฟ็จไผๆๆพๆๅฐๆ็นใ
Unsloth Dynamic 3.1 ็ ยท ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5/๏ผๆจ่๏ผ
**ๅบไบ SC117/Qwen3.8-Flash-Next-abliterated-UD-GGUF
็ UD-IQ3_XXS**๏ผๅๆ ทๆ 512 ไธชไธๅฎถ็ ๅฐ 320 ไธช๏ผๆฉๆ็ๆฏ่ทฏ็ฑๅฎฝๅบฆไธไธๅฎถๅ๏ผ็ไธๆฅ็ๅ
้ๅญ่ๆฌ่ฟใไธ้ๆฐ้ๅ๏ผไธ Swift ็ณปๅๅไธๅฅๅๆณ๏ผๅชๆฏๆบๅคดๆขๆไบ Unsloth ็้ๅฑๅจๆ้ๅ็ๆฌ๏ผใ
ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5/
ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5-00001-of-00002.gguf 26.83 GiB per-layer token embedding ่กจ๏ผPLE๏ผ
ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5-00002-of-00002.gguf 33.40 GiB ๅ
ถไฝ 1223 ไธชๅผ ้
ๅ่ฎก 60.23 GiB
- ไธค็ๅฟ
้กปไธ้ฝไธ่ฝฝ๏ผๅ ่ฝฝๆถๅช็ป็ฌฌไธ็็่ทฏๅพ๏ผ็ฌฌไบ็ๆ
-0000N-of-0000M่ชๅจๆพ้ฝใ - ็ฌฌไธ็้ๅชๆ PLE ่กจไธๅผ ๅผ ้๏ผ
split.no = 0๏ผ160 ร 320001536๏ผ๏ผๆไปฅๅฎ 26.83 GiB ๅดๅชๆไธๅผ ่กจ๏ผๆฏๆญฃๅธธ็ใ - ่ฟไธๆกฃ็ไธๅฎถๆฏ้ๅฑๅจๆ้ๅ๏ผไธๅๅฑ็ gate/up ๅๅซๆฏ IQ2_S / IQ3_XXS / IQ3_S / IQ4_XS๏ผdown ๆฏ IQ4_NL / Q2_0๏ผๅ ๆญคๅไฝ็งฏไธ้ๅ่ฏฏๅทฎๆดๅฐ๏ผไปฃไปทๆฏๅ ๆ ธ็ปๅๆด็ข๏ผ้ขๅกซไธ่งฃ็ ้ฝไผๆฏ Swift ็ณปๅๆ ขไธไบใ
ๅฆไฝ้๏ผๆไปฌ่ชๅทฑๅจไธญๆ้ฟๆ๏ผๅฐ่ฏด๏ผไธ็ๅฎไปฃ็ ๏ผZig / Go / Rust / C++๏ผไธๅฎๆต๏ผUD3.1 320E ็่ดจ้ๆๆพๅฅฝไบ Swift 320E๏ผไฝ้ๅบฆๆดๆ ขใๆ ขๅ ้ธ ๅคๅฎๅ จๅๅณไบไฝ ็ๆจ็็ซฏใไธไธๆ้ฟๅบฆใKV ่ฎพ็ฝฎไธๆฏๅฆๆๆบ่งฃ็ โโๅปบ่ฎฎ่ชๅทฑๅฎๆตไธๆฌก๏ผ่ฆ่ดจ้้ UD๏ผ่ฆ้ๅบฆ้ Swiftใ
ๆไปถ็ปๆ
Swift ไบๆกฃๆฏๆกฃไธไธชๅญ็ฎๅฝ๏ผๅ
้จๆฏ 3 ไธชๅ็๏ผllama.cpp gguf-split ๅฝๅ๏ผ๏ผUD ้ฃไธๆกฃๆฏ 2 ไธชๅ็๏ผ่งไธไธ่๏ผ
ZatFung-Swift-320E-v5/
ZatFung-Swift-320E-v5-00001-of-00003.gguf 10.01 GiB ๅ
ๆฐๆฎ + embedding / attention / SSM / router
ZatFung-Swift-320E-v5-00002-of-00003.gguf 23.33 GiB ็ฌฌ 14โ47 ๅฑ็ไธๅฎถไธๅ
ถไฝๆ้
ZatFung-Swift-320E-v5-00003-of-00003.gguf 26.82 GiB per-layer token embedding ่กจ๏ผPLE / n-gram ่กจ๏ผ
- ็ฌฌ 3 ็๏ผ
per_layer_token_embd.weight๏ผไบๆกฃ้ฝไธๆ ทๅคง๏ผ26.82 GiB๏ผ๏ผไธ้ไธๅฎถๆฐๅๅโโ ๆไปฅๆๅฐๆกฃไนๆ 48 GiBใ - ๆฏ็ๅธฆ
split.count/split.no/split.tensors.count๏ผๅไธๆกฃ็ไธ็ๅฟ ้กปไธ้ฝไธ่ฝฝใ - ๅ ่ฝฝๆถๅช็ป็ฌฌไธ็็่ทฏๅพ๏ผๅ ถไฝๅ็ๆๆไปถๅ่ชๅจๆพ้ฝ๏ผ็ผบ็ไผๆ็กฎๆฅๅบ็ผบๅชไธ็ใ
้ ๅฅๆไปถ๏ผไปๅบๆ น๏ผ
| ๆไปถ | ๅคงๅฐ | ็จ้ |
|---|---|---|
mmproj-Qwen3.8-Flash-Next-BF16.gguf |
0.85 GiB | ่ง่ง็ผ็ ๅจ๏ผBF16๏ผใ่ฆๅๅพๅ่พๅ ฅๆถไธไธปๆจกๅ้ ๅฏนไฝฟ็จ |
tmpls/qwen3.8-froggeric-v22.5.jinja |
28 KB | ่ๅคฉๆจกๆฟ๏ผๅ็ |
tmpls/qwen3.8-froggeric-v22.5-hont.jinja |
28 KB | ๅไธ๏ผๆ่ๅ็ผ็จ็นไฝใๆ่้็จ๏ผใ |
tmpls/qwen3.8-froggeric-v22.5-hons.jinja |
28 KB | ๅไธ๏ผๆ่ๅ็ผ็จ็ฎไฝใๆ่่ฟ็จ๏ผใ |
mtp/*.gguf |
0.4โ3.9 GiB | ๅฏ้็ MTP draft ๅคด๏ผ5 ไธช๏ผ้ๅไธไธๅฎถๆฐไธๅ๏ผใไธไธปๆจกๅๅฎๅ จ็ฌ็ซ๏ผๅชๆๅๆๆบ่งฃ็ ๅ ้ๆถๆ้่ฆ |
mtp/rt/ |
787 MB | ๆไธช่็จฟๅคด็่ฟ่กๆถ็ฎๅฝๅฝขๆ๏ผdense.bin + experts.bin + draft_vocab.bin + dense.txt๏ผ๏ผไพ STRATA ๅผๆ็ดๆฅๅ ่ฝฝ๏ผไธๅฟ
ๅไป GGUF ่ฝฌๆข |
ๆๆๆกฃไฝ๏ผSwift ไบๆกฃไธ UD ไธๆกฃ๏ผ้ฝไธๅ ๅซ่ฟไบๆไปถ๏ผๆ้ๅ็ฌไธ่ฝฝใ
ๅฆไฝ่ท
่ฟๆฏๆ ๅ GGUF ๅ็๏ผ้่ฆๆฏๆ qwen4exp๏ผQwen3.8-Flash-Next๏ผๆถๆ็ๆจ็็ซฏๅ ่ฝฝ๏ผ
# Swift ๆกฃ๏ผ3 ็๏ผ
<ไฝ ็ๆจ็็ซฏ> --model ZatFung-Swift-320E-v5/ZatFung-Swift-320E-v5-00001-of-00003.gguf --ctx 131072
# UD ๆกฃ๏ผ2 ็๏ผๆจ่๏ผ
<ไฝ ็ๆจ็็ซฏ> --model ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5/ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5-00001-of-00002.gguf --ctx 131072
ไธไธๆๅ็ๆฏๆ 262144๏ผๆพๅญ/ๅ ๅญไธ่ถณๆถไผๅ ้ไธไธๆๆ้ๆดๅฐ็ๆกฃ๏ผ่ไธๆฏๆๆ้ๅ้ๅไธๆฌกใ
่ฎธๅฏ
่ฎธๅฏ็ปงๆฟไธๆธธ๏ผSwift ๆกฃ็ๅบ็กๆจกๅๆฅ่ช SC117 ็ Swift-1.5-Qwen3.8-Flash-Next๏ผGSQ-RCO abliterated๏ผ๏ผ UD ๆกฃๆฅ่ช SC117 ็ Qwen3.8-Flash-Next-abliterated-UD-GGUF๏ผUnsloth Dynamic ้ๅ๏ผใไธค่ ็ๅบๅฑ้ฝๆฏๅ้ฎ ๏ผQwen๏ผๅบ็กๆจกๅ๏ผๅ ๆญคๆฌไปๅบๆฒฟ็จๅฎไธๆธธ็่ฎธๅฏไธไฝฟ็จๆกๆฌพ๏ผabliterated ไธๅจๆ้ๅ้จๅ็่ฏดๆไปฅไธๆธธไปๅบไธบๅใ ๅ็จๅ่ฏท่ช่กๆ ธๅฏนไธๆธธ่ฎธๅฏๅๆใ
English (short)
Expert-reduced variants of SC117/Swift-1.5-Qwen3.8-Flash-Next-GSQ-RCO-abliterated-GGUF
(IQ3_S), keeping 192 / 256 / 288 / 320 / 384 of the original 512 experts. Weights keep their original
quantisation (nothing is re-quantised). This release is tuned for Chinese long-form text, real code
(Zig / Go / Rust / C++) and algorithm tasks โ hence the "Chinese + code" strength, aimed at coding and
long-form Chinese fiction. Each tier is 3 gguf-split shards; the third shard is the 26.82 GiB
per-layer token embedding table and is identical across tiers. Use 320E or 384E; 192E/256E deviate
8.6% / 4.6% from the original model and are not recommended for production. Licence follows the upstream
Qwen-based model.
The second family, ZatFung-Unsloth-Dynamic-3-1-IQ3_XXS-320E-v5/, is the same 320-expert reduction of
SC117/Qwen3.8-Flash-Next-abliterated-UD-GGUF
(Unsloth Dynamic UD-IQ3_XXS, per-layer mixed quantisation) in 2 shards (shard 1 is the PLE table alone).
We measured it to be better quality than the Swift 320E and somewhat slower on Chinese long-form and real
code; pick it for quality, the Swift tiers for speed.
- Downloads last month
- 1,761
Model tree for NaamFung/ZatFung
Base model
Qwen/Qwen3.8-Flash-Next