Views
No views yet
| File | Use |
|---|---|
| model_int8.onnx | INT8 quantized — default, ~3-5x faster |
| model.onnx | FP32 — full precision, use when accuracy matters more than latency |
{0: "neutral", 1: "toxic"}1from guardex import Guard
2guard = Guard() # pulls this model on first call, caches under ~/.cache/huggingface