[Quantization] fix quantization pass bug#355
Merged
Aalanli merged 2 commits intohidet-org:mainfrom Aug 29, 2023
Merged
Conversation
yaoyaoding
approved these changes
Aug 25, 2023
vadiklyutiy
added a commit
that referenced
this pull request
Jul 22, 2024
Introduce `add_hint_pass`. It adds `__builtin_assume(...)` to .cu code that helps nvcc to understand bounds if `threadIdx` and `blockIdx` and optimize code better. **Performance improvements.** Models model|latency|prev_latency|ratio| |--------|--------|--------|--------| bert-base-uncased|19.8138|20.2316|2.109 densenet121|35.1161|36.7627|4.689 efficientnet_b0|18.9451|19.278|1.757 mobilenet_v2|11.5944|11.8764|2.432 resnet50|29.4878|29.9935|1.715 vit_b_16|125.787|123.672|-1.681 Operators operator|latency|prev_latency|ratio |--------|--------|--------|--------| attn|1.50402|1.50131|-0.18 attn|0.219707|0.227568|3.578 attn_mask_add|1.5892|1.62516|2.263 attn_mask_add|0.226317|0.226507|0.084 batch_matmul|5.2399|5.11547|-2.375 batch_matmul|0.0216016|0.0223425|3.43 conv2d|0.0347093|0.0341758|-1.537 conv2d|0.310521|0.308458|-0.664 conv2d_gemm_f16|0.142542|0.146412|2.715 conv2d_gemm_f16|2.0421|2.07043|1.387 matmul_f16|2.22432|2.30458|3.608 matmul_f16|0.00888628|0.00892615|0.449 reduce|0.01375|0.0138618|0.813
vadiklyutiy
added a commit
that referenced
this pull request
Jul 23, 2024
Introduce `add_hint_pass`. It adds `__builtin_assume(...)` to .cu code that helps nvcc to understand bounds if `threadIdx` and `blockIdx` and optimize code better. **Performance improvements.** Models model|latency|prev_latency|ratio| |--------|--------|--------|--------| bert-base-uncased|19.8138|20.2316|2.109 densenet121|35.1161|36.7627|4.689 efficientnet_b0|18.9451|19.278|1.757 mobilenet_v2|11.5944|11.8764|2.432 resnet50|29.4878|29.9935|1.715 vit_b_16|125.787|123.672|-1.681 Operators operator|latency|prev_latency|ratio |--------|--------|--------|--------| attn|1.50402|1.50131|-0.18 attn|0.219707|0.227568|3.578 attn_mask_add|1.5892|1.62516|2.263 attn_mask_add|0.226317|0.226507|0.084 batch_matmul|5.2399|5.11547|-2.375 batch_matmul|0.0216016|0.0223425|3.43 conv2d|0.0347093|0.0341758|-1.537 conv2d|0.310521|0.308458|-0.664 conv2d_gemm_f16|0.142542|0.146412|2.715 conv2d_gemm_f16|2.0421|2.07043|1.387 matmul_f16|2.22432|2.30458|3.608 matmul_f16|0.00888628|0.00892615|0.449 reduce|0.01375|0.0138618|0.813
vadiklyutiy
added a commit
that referenced
this pull request
Dec 26, 2024
Introduce `add_hint_pass`. It adds `__builtin_assume(...)` to .cu code that helps nvcc to understand bounds if `threadIdx` and `blockIdx` and optimize code better. **Performance improvements.** Models model|latency|prev_latency|ratio| |--------|--------|--------|--------| bert-base-uncased|19.8138|20.2316|2.109 densenet121|35.1161|36.7627|4.689 efficientnet_b0|18.9451|19.278|1.757 mobilenet_v2|11.5944|11.8764|2.432 resnet50|29.4878|29.9935|1.715 vit_b_16|125.787|123.672|-1.681 Operators operator|latency|prev_latency|ratio |--------|--------|--------|--------| attn|1.50402|1.50131|-0.18 attn|0.219707|0.227568|3.578 attn_mask_add|1.5892|1.62516|2.263 attn_mask_add|0.226317|0.226507|0.084 batch_matmul|5.2399|5.11547|-2.375 batch_matmul|0.0216016|0.0223425|3.43 conv2d|0.0347093|0.0341758|-1.537 conv2d|0.310521|0.308458|-0.664 conv2d_gemm_f16|0.142542|0.146412|2.715 conv2d_gemm_f16|2.0421|2.07043|1.387 matmul_f16|2.22432|2.30458|3.608 matmul_f16|0.00888628|0.00892615|0.449 reduce|0.01375|0.0138618|0.813
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
For llama 7b, decoding with no prefill
fp16 - 128 tokens
org_t: 3.048099994659424s
avg_t: 0.0044784750789403915s
int8 - 128 tokens
org_t: 2.2229115962982178s
avg_t: 0.0038476772606372833s