Quantification and Optimization

8-bit Floating Point Representation (FP8)

Very low-precision numerical data format using 8 bits to represent floating-point numbers, enabling significant accelerations on modern GPUs while maintaining the training stability of large models.

← Quay lại