Skip to content

[Frontend] Fix typo#127

Merged
yaoyaoding merged 1 commit intohidet-org:mainfrom
digital-nomad-cheng:main
Mar 6, 2023
Merged

[Frontend] Fix typo#127
yaoyaoding merged 1 commit intohidet-org:mainfrom
digital-nomad-cheng:main

Conversation

@digital-nomad-cheng
Copy link
Copy Markdown
Contributor

No description provided.

@yaoyaoding yaoyaoding merged commit 58207ce into hidet-org:main Mar 6, 2023
KTong821 pushed a commit to KTong821/hidet that referenced this pull request Apr 24, 2024
…ion(`implement*`) (hidet-org#127)

This PR parallelizes: 
 - `apply_prolog_epilog` (fusion)
 - IR generation (`implement*`)

Right now implemented for host only (no offload to comp server). 

resnet50 compilation speed on g5.16xlarge 
`time python tests/benchmarks/bench_vision.py resnet50 --params 1x3x224x224 --dtype float16`
Before: 14m 45s
After: 12m 51s
Speedup: 14.8%

matmul compilation speed on g5.16xlarge 
`time python tests/benchmarks/bench_op.py batch_matmul --params 1x4096x4096,1x4096x4096 --dtype float16`
Before: 5m 54s
After: 5m 31s
Speedup: 6.9%
vadiklyutiy added a commit that referenced this pull request Jul 22, 2024
…ion(`implement*`) (#127)

This PR parallelizes: 
 - `apply_prolog_epilog` (fusion)
 - IR generation (`implement*`)

Right now implemented for host only (no offload to comp server). 

resnet50 compilation speed on g5.16xlarge 
`time python tests/benchmarks/bench_vision.py resnet50 --params 1x3x224x224 --dtype float16`
Before: 14m 45s
After: 12m 51s
Speedup: 14.8%

matmul compilation speed on g5.16xlarge 
`time python tests/benchmarks/bench_op.py batch_matmul --params 1x4096x4096,1x4096x4096 --dtype float16`
Before: 5m 54s
After: 5m 31s
Speedup: 6.9%
vadiklyutiy added a commit that referenced this pull request Jul 23, 2024
…ion(`implement*`) (#127)

This PR parallelizes: 
 - `apply_prolog_epilog` (fusion)
 - IR generation (`implement*`)

Right now implemented for host only (no offload to comp server). 

resnet50 compilation speed on g5.16xlarge 
`time python tests/benchmarks/bench_vision.py resnet50 --params 1x3x224x224 --dtype float16`
Before: 14m 45s
After: 12m 51s
Speedup: 14.8%

matmul compilation speed on g5.16xlarge 
`time python tests/benchmarks/bench_op.py batch_matmul --params 1x4096x4096,1x4096x4096 --dtype float16`
Before: 5m 54s
After: 5m 31s
Speedup: 6.9%
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants