DeepSpeed
4114beac - add configurable quantization for enabling 4-bit inference

Commit

3 years ago

add configurable quantization for enabling 4-bit inference

Author

Reza Yazdani Aminabadi

Parents

Loading