DeepSpeed
Optimize the fp-dequantizer to get high memory-BW utilization
#5373

Merged

Optimize the fp-dequantizer to get high memory-BW utilization #5373

loadams merged 3 commits into deepspeedai:master from RezaYazdaniAminabadi:master

Optimize the fp-dequantizer to get high memory-BW utilization

10bad7de

RezaYazdaniAminabadi requested a review from

awan-10 1 year ago

RezaYazdaniAminabadi requested a review from

mrwyattii 1 year ago

RezaYazdaniAminabadi requested a review from

arashb 1 year ago

fix formating

c5ba68e3

JamesTheZ approved these changes on 2024-04-10

Merge branch 'master' into master

9975f753

loadams enabled auto-merge 1 year ago

loadams merged a8b82153 into master 1 year ago

Reviewers

JamesTheZ

awan-10

mrwyattii

arashb

Assignees

No one assigned

Labels

None yet

Milestone

No milestone