DeepSpeed
bd2519c1 - skip quantization for tensor smaller than 500k

Commit
2 years ago
skip quantization for tensor smaller than 500k
Author
Parents
Loading