Completed
INT8 and FP8 have the same number of values but different distributions.
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Advances in Quantization for Efficient On-Device Neural Network Inference
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Low-precision numerical formats
- 3 INT8 and FP8 have the same number of values but different distributions.
- 4 INT8 and FP8 accuracy
- 5 Challenges in using integer quantization
- 6 Introduction to Quantization-Aware Training (QAT)
- 7 Oscillating weights in QAT
- 8 MobileNetV2 - comparison to literature
- 9 Why do outliers occur?
- 10 Outliers in Transformers
