Token Classification
PyTorch
Safetensors
GGUF
MLX
English
customs
span-tagging
pii
pii-detection
privacy
data-minimisation
on-device
Synthetic
Instructions to use NagaYu/customs-decomposer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use NagaYu/customs-decomposer with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir customs-decomposer NagaYu/customs-decomposer
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use NagaYu/customs-decomposer 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 NagaYu/customs-decomposer # Run inference directly in the terminal: llama cli -hf NagaYu/customs-decomposer
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf NagaYu/customs-decomposer # Run inference directly in the terminal: llama cli -hf NagaYu/customs-decomposer
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 NagaYu/customs-decomposer # Run inference directly in the terminal: ./llama-cli -hf NagaYu/customs-decomposer
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 NagaYu/customs-decomposer # Run inference directly in the terminal: ./build/bin/llama-cli -hf NagaYu/customs-decomposer
Use Docker
docker model run hf.co/NagaYu/customs-decomposer
- LM Studio
- Jan
- Ollama
How to use NagaYu/customs-decomposer with Ollama:
ollama run hf.co/NagaYu/customs-decomposer
- Unsloth Studio
How to use NagaYu/customs-decomposer 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 NagaYu/customs-decomposer 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 NagaYu/customs-decomposer to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for NagaYu/customs-decomposer to start chatting
- Docker Model Runner
How to use NagaYu/customs-decomposer with Docker Model Runner:
docker model run hf.co/NagaYu/customs-decomposer
- Lemonade
How to use NagaYu/customs-decomposer with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull NagaYu/customs-decomposer
Run and chat with the model
lemonade run user.customs-decomposer-{{QUANT_TAG}}List all available models
lemonade list
- Atomic Chat
| { | |
| "seed": 0, | |
| "steps": 1500, | |
| "n_parameters": 347167, | |
| "vocab_size": 481, | |
| "n_tags": 31, | |
| "corpus": { | |
| "requests": 1440, | |
| "profiles": 60, | |
| "train": 1080, | |
| "val": 360 | |
| }, | |
| "train": { | |
| "precision": 1.0, | |
| "recall": 1.0, | |
| "f1": 1.0, | |
| "value_recall": 1.0, | |
| "n_gold": 2430, | |
| "n_pred": 2430 | |
| }, | |
| "val": { | |
| "precision": 1.0, | |
| "recall": 1.0, | |
| "f1": 1.0, | |
| "value_recall": 1.0, | |
| "n_gold": 810, | |
| "n_pred": 810 | |
| }, | |
| "ambiguity_probe": { | |
| "city": { | |
| "personal": 135, | |
| "public": 102, | |
| "tagger_correct": 237, | |
| "gazetteer_correct": 135, | |
| "tagger_accuracy": 1.0, | |
| "gazetteer_accuracy": 0.5696 | |
| }, | |
| "iso_date": { | |
| "birth": 45, | |
| "travel": 45, | |
| "tagger_correct": 90, | |
| "regex_correct": 45, | |
| "tagger_accuracy": 1.0, | |
| "regex_accuracy": 0.5 | |
| } | |
| }, | |
| "history": [ | |
| { | |
| "step": 100, | |
| "loss": 0.8456018883455545, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 200, | |
| "loss": 0.0029681686707772313, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 300, | |
| "loss": 0.0010909774966421538, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 400, | |
| "loss": 0.0012425882503157481, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 500, | |
| "loss": 0.0008028568621375598, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 600, | |
| "loss": 0.0006515971080807504, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 700, | |
| "loss": 0.00020419371416210196, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 800, | |
| "loss": 0.00022941515599086414, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 900, | |
| "loss": 0.0001741512342414353, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 1000, | |
| "loss": 0.00028020317062328105, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 1100, | |
| "loss": 0.00016108047333545983, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 1200, | |
| "loss": 0.00013588617766799871, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 1300, | |
| "loss": 0.00013093463247059844, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 1400, | |
| "loss": 9.306033061875496e-05, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| }, | |
| { | |
| "step": 1500, | |
| "loss": 9.184164744510781e-05, | |
| "val_f1": 1.0, | |
| "val_value_recall": 1.0 | |
| } | |
| ], | |
| "seconds": 33.95, | |
| "data_provenance": "fully synthetic; generated by customs.world + customs.templates", | |
| "held_out_check": { | |
| "covered": 72, | |
| "total": 72, | |
| "rate": 1.0, | |
| "misses": [] | |
| } | |
| } |