Embedding layer quantization
#2
by
Junghyeok
- opened
What is the loss in accuracy when the embedding layer is quantized to 8 bits per channel?
What is the loss in accuracy when the embedding layer is quantized to 8 bits per channel?