-
Notifications
You must be signed in to change notification settings - Fork 24.9k
Bump transformers pin #159291
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Bump transformers pin #159291
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/159291
Note: Links to docs will display an error until the docs builds have been completed. ❌ 4 New FailuresAs of commit 61fc201 with merge base 334ecbd ( NEW FAILURES - The following jobs have failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
fb506c0
to
1013de5
Compare
@@ -14,6 +14,7 @@ | |||
"detectron2_maskrcnn_r_101_c4", | |||
"timm_efficientnet", # see https://github.com/pytorch/pytorch/issues/148699 | |||
"XGLMForCausalLM", # discovered in https://github.com/pytorch/pytorch/pull/128148 | |||
"moondream", # discovered in https://github.com/pytorch/pytorch/pull/159291 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
should we file an issue?
# Real transformers benchmarks will be added soon using a different | ||
# infra. | ||
if model_name.startswith("hf") and hasattr(model.config, "use_cache"): | ||
model.config.use_cache = False |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
what would be the issue if we don't turn off kv cache? We probably want to track a real-world setting for regression..
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We should just rely on Angela's transformers models here, instead of trying to work with these old models. There are a few different types of caches in transfomers DynamicCache, StaticCache etc, and its better to rely on the generate
API (which Angela uses in the new model benchmarks).
My other extreme suggestion is to just remove these models from the CI, but thats a discussion for another day.
Thanks for bumping the transfomers pin! Should we add screenshot of dashboard before/after this pr to the description? So we know the dashboard change from pin update. |
f79f1fe
to
07dfd59
Compare
Converting to draft because the dashboard is red even though CI passes. |
97a64f2
to
3302756
Compare
07a9eae
to
4ff8560
Compare
(accidentally delete the branch, but I'm re-opening the PR now) |
Signed-off-by: Huy Do <huydhn@gmail.com>
The dashboard run is almost done. I will update the rest of the accuracy values tomorrow to get a green PR. Among them, https://github.com/pytorch/pytorch/actions/runs/16821307931/job/47652611212 looks like a real regression but it comes from my change, so I will also push a fix for this in one go. |
@huydhn from the failing tuns, it seems that the new version was not picked up |
Signed-off-by: Huy Do <huydhn@gmail.com>
Signed-off-by: Huy Do <huydhn@gmail.com>
This reverts commit 505c7b3.
Signed-off-by: Huy Do <huydhn@gmail.com>
Signed-off-by: Huy Do <huydhn@gmail.com>
Signed-off-by: Huy Do <huydhn@gmail.com>
Signed-off-by: Huy Do <huydhn@gmail.com>
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Merge failedReason: 1 mandatory check(s) failed. The first few are: Dig deeper by viewing the failures on hud |
@pytorchbot merge -i |
Merge startedYour change will be merged while ignoring the following 4 checks: pull / linux-jammy-py3.9-gcc11 / test (default, 5, 5, linux.2xlarge), pull / linux-jammy-py3.9-clang12 / test (default, 2, 5, linux.4xlarge), inductor / unit-test / linux-jammy-cpu-py3.9-gcc11-inductor / test (inductor_amx, 1, 2, linux.8xlarge.amx), inductor / unit-test / linux-jammy-cpu-py3.9-gcc11-inductor / test (inductor_avx2, 1, 2, linux.10xlarge.avx2) Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Trying to update hf pin.
Benchmarking run to figure out issues
Retrying - #156118
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @chenyang78 @kadeng @muchulee8 @amjames @chauhang @aakhundov @coconutruben @Lucaskabela