---
license: apache-2.0
---

FP8 static quantized model with compressored tensors. per-channel weight and per-tensor activation.