PR #20237 [js/webgpu] Support GroupQueryAttention

[js/webgpu] Support GroupQueryAttention #20237

guschmue merged 11 commits into microsoft:main from axinging:groupqueryattn

axinging force pushed from 11556cb3 to 46f14e6f 1 year ago

axinging force pushed from 46f14e6f to cc13e089 1 year ago

axinging changed the title ~~[Nijs/webgpu] Support GroupQueryAttention~~ [js/webgpu] Support GroupQueryAttention 1 year ago

axinging force pushed from cc13e089 to 3cd27537 1 year ago

axinging force pushed from 3cd27537 to df80fe7a 1 year ago

axinging force pushed from 9c94d809 to 9a70a6d2 1 year ago

axinging force pushed from 9a70a6d2 to b745c24b 1 year ago

axinging force pushed from b745c24b to fef358f9 1 year ago

axinging force pushed from fef358f9 to 8cfe9333 1 year ago

axinging force pushed from 8cfe9333 to e6e2a93a 1 year ago

axinging force pushed from e6e2a93a to 94aa95e8 1 year ago

axinging force pushed from 9924d509 to c60f0eed 1 year ago

axinging force pushed from c60f0eed to 45478c97 1 year ago

axinging force pushed from 45478c97 to feccc9f3 1 year ago

axinging force pushed from feccc9f3 to 2c1c0625 1 year ago

axinging force pushed from 2c1c0625 to 54ff1c46 1 year ago

axinging marked this pull request as ready for review 1 year ago

axinging marked this pull request as draft 1 year ago

axinging marked this pull request as ready for review 1 year ago

axinging marked this pull request as draft 1 year ago

guschmue added ep:WebGPU

axinging changed the title ~~[js/webgpu] Support GroupQueryAttention~~ [js/webgpu] Support GroupQueryAttention(NotForReview) 1 year ago

axinging force pushed from 5464d0d2 to 32dcaf4e 1 year ago

axinging force pushed from 32dcaf4e to 275fa2bb 1 year ago

axinging force pushed from 275fa2bb to 7d044503 1 year ago

axinging force pushed from 7d044503 to dffd2728 1 year ago

axinging force pushed from 3a4f38cd to eaf89180 1 year ago

axinging force pushed from eaf89180 to 790376fe 1 year ago

axinging force pushed from 790376fe to 20fe303b 1 year ago

axinging changed the title ~~[js/webgpu] Support GroupQueryAttention(NotForReview)~~ [js/webgpu] Support GroupQueryAttention 1 year ago

axinging force pushed from 20fe303b to 8b0456ee 1 year ago

axinging marked this pull request as ready for review 1 year ago

axinging force pushed from 8b0456ee to 1ffa3c42 1 year ago

satyajandhyala commented on 2024-04-27

axinging force pushed from 1ffa3c42 to 0aa45db6 1 year ago

axinging marked this pull request as draft 1 year ago

axinging marked this pull request as ready for review 1 year ago

satyajandhyala commented on 2024-04-28

axinging marked this pull request as draft 1 year ago

axinging marked this pull request as ready for review 1 year ago

guschmue dismissed these changes on 2024-05-06

axinging dismissed their stale review via 9495a7c6 1 year ago

[js/webgpu] Support GroupQueryAttention

cac2caad

Format jsonc and fix commensts&typo

982c9944

Remove space

595d7998

Turn past format to BNSH

c3637f51

Support present key value as output

7d583b52

Nit

735b77fb

Add ceil when calc H

b49b1733

Update doc

2bd89587

Use BSNH as past

f660df7e

Merge branch 'main' into groupqueryattn

57578645

axinging force pushed from 1fa8a467 to 57578645 1 year ago

Run npm run prebuild

4a64ff98

guschmue approved these changes on 2024-05-13

guschmue merged 8c59cd4f into main 1 year ago

Reviewers

guschmue

satyajandhyala

Assignees

No one assigned

Labels

ep:WebGPU

Milestone

No milestone

onnxruntime [js/webgpu] Support GroupQueryAttention #20237 Merged

[js/webgpu] Support GroupQueryAttention #20237

onnxruntime
[js/webgpu] Support GroupQueryAttention
#20237

Merged