Skip to content

Actions: pytorch/executorch

Actions

Propose to merge ghstack orig PRs to main

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
ghstack_land.yml
907 workflow runs
907 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Reduce memory requirement on export_llama tests with no params Propose to merge ghstack orig PRs to main #557: Pull request #11175 closed by jackzhxng
33s gh/jackzhxng/22/head
gh/jackzhxng/22/head
33s
[ET-VK] Using push constants for buffer to image prepack nodes. Propose to merge ghstack orig PRs to main #556: Pull request #11252 closed by facebook-github-bot
37s gh/trivedivivek/108/head
gh/trivedivivek/108/head
37s
[ET-VK] Introduce generalized shaders for transfer ops and use it for select and slice Propose to merge ghstack orig PRs to main #555: Pull request #11255 closed by facebook-github-bot
37s gh/SS-JIA/231/head
gh/SS-JIA/231/head
37s
[ET-VK][ez] Updates to DynamicDispatchNode Propose to merge ghstack orig PRs to main #554: Pull request #11254 closed by facebook-github-bot
42s gh/SS-JIA/230/head
gh/SS-JIA/230/head
42s
[ET-VK][ez] Enable dynamic shape support when using push constants Propose to merge ghstack orig PRs to main #553: Pull request #11253 closed by facebook-github-bot
42s gh/SS-JIA/229/head
gh/SS-JIA/229/head
42s
Make release labels optional Propose to merge ghstack orig PRs to main #552: Pull request #11287 closed by jackzhxng
39s gh/jackzhxng/23/head
gh/jackzhxng/23/head
39s
[ET-VK] Fixing out_limits_scaled calculation for conv2d pw ops. Propose to merge ghstack orig PRs to main #551: Pull request #11251 closed by facebook-github-bot
46s gh/trivedivivek/107/head
gh/trivedivivek/107/head
46s
[ET-VK] Removed shared memory usage and simplied conv2d dw op shader to improve performance. Propose to merge ghstack orig PRs to main #550: Pull request #11178 closed by facebook-github-bot
44s gh/trivedivivek/102/head
gh/trivedivivek/102/head
44s
[ET-VK] Modifying should_squeeze function in SqueezeUnsqueezeInputs to not squeeze if significant axis are all 1 and trailing axis are all > 1. Propose to merge ghstack orig PRs to main #549: Pull request #11177 closed by facebook-github-bot
41s gh/trivedivivek/101/head
gh/trivedivivek/101/head
41s
[ET-VK][Ops] Registering tan tests Propose to merge ghstack orig PRs to main #548: Pull request #11216 closed by facebook-github-bot
38s gh/ahmtox/6/head
gh/ahmtox/6/head
38s
Dtype selective build: disable when not in xplat and add fail msg Propose to merge ghstack orig PRs to main #547: Pull request #11090 closed by facebook-github-bot
40s gh/lucylq/84/head
gh/lucylq/84/head
40s
[ET-VK] Applying bias after sum calculation in conv2d pw shader to improve performance. Propose to merge ghstack orig PRs to main #546: Pull request #11150 closed by facebook-github-bot
1m 2s gh/trivedivivek/100/head
gh/trivedivivek/100/head
1m 2s
[ET-VK] Reducing precision of some in members in conv2d pw to improved performance. Propose to merge ghstack orig PRs to main #545: Pull request #11139 closed by facebook-github-bot
51s gh/trivedivivek/99/head
gh/trivedivivek/99/head
51s
[ET-VK] Storing positions in uint16 to instead of int in conv2d pw shader. Propose to merge ghstack orig PRs to main #544: Pull request #11138 closed by facebook-github-bot
55s gh/trivedivivek/98/head
gh/trivedivivek/98/head
55s
[ET-VK] Creating specialized version of conv2d pw shader for X and Y stride = 1 and padding = 0. Propose to merge ghstack orig PRs to main #543: Pull request #11137 closed by facebook-github-bot
53s gh/trivedivivek/97/head
gh/trivedivivek/97/head
53s
[ET-VK] De vectorise all vectors in conv2d pw shader to improve perf. Propose to merge ghstack orig PRs to main #542: Pull request #11136 closed by facebook-github-bot
1m 6s gh/trivedivivek/96/head
gh/trivedivivek/96/head
1m 6s
[ET-VK] Tuning local workgroup size calculation for conv2d pw to improve performance. Propose to merge ghstack orig PRs to main #541: Pull request #11135 closed by facebook-github-bot
1m 8s gh/trivedivivek/95/head
gh/trivedivivek/95/head
1m 8s
[ET-VK] Minor unroll tuning to improve conv2d pw perf. Propose to merge ghstack orig PRs to main #540: Pull request #11134 closed by facebook-github-bot
59s gh/trivedivivek/94/head
gh/trivedivivek/94/head
59s
[ET-VK] De vectorise positions in conv2d pw shader to improve perf. Propose to merge ghstack orig PRs to main #539: Pull request #11122 closed by facebook-github-bot
56s gh/trivedivivek/93/head
gh/trivedivivek/93/head
56s
[ET-VK] Minor tuning for conv2d pw op to improve performance. Propose to merge ghstack orig PRs to main #538: Pull request #11113 closed by facebook-github-bot
50s gh/trivedivivek/92/head
gh/trivedivivek/92/head
50s
[ET-VK] Tuning conv 2d pw op tile size to improve perf. Propose to merge ghstack orig PRs to main #537: Pull request #11112 closed by facebook-github-bot
42s gh/trivedivivek/91/head
gh/trivedivivek/91/head
42s
[ET-VK] Remove the use of shared memory in conv2d pw to improve perf. Propose to merge ghstack orig PRs to main #536: Pull request #11110 closed by facebook-github-bot
57s gh/trivedivivek/90/head
gh/trivedivivek/90/head
57s
[ET-VK] De vectorise conv2d pw shader to improve perf. Propose to merge ghstack orig PRs to main #535: Pull request #11108 closed by facebook-github-bot
46s gh/trivedivivek/89/head
gh/trivedivivek/89/head
46s
Fix test_llama Propose to merge ghstack orig PRs to main #534: Pull request #11165 closed by jackzhxng
31s gh/jackzhxng/15/head
gh/jackzhxng/15/head
31s
refactor: Replace self.args with LlmConfig in model.py and export_llama_lib.py Propose to merge ghstack orig PRs to main #533: Pull request #11166 closed by jackzhxng
32s gh/jackzhxng/16/head
gh/jackzhxng/16/head
32s
ProTip! You can narrow down the results and go further in time using created:<2025-05-27 or the other filters available.