FP8

4 posts · 3 sources

NVIDIABlog16 Jun

How to Optimize Transformer-Based Models for Low-Precision Training

ModalBlog11 Jun

Making FlashAttention-4 faster for inference

NVIDIABlog9 Jun

Model Quantization: Turn FP8 Checkpoints into High-Performance Inference Engines with NVIDIA TensorRT

BasetenBlog7 Mar

FP8: Efficient model inference with 8-bit floating point numbers