-
Notifications
You must be signed in to change notification settings - Fork 24.5k
fix kl_div for negative targets #69212
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
Conversation
[ghstack-poisoned]
CI Flow Status⚛️ CI FlowRuleset - Version:
You can add a comment to the PR and tag @pytorchbot with the following commands: # ciflow rerun, "ciflow/default" will always be added automatically
@pytorchbot ciflow rerun
# ciflow rerun with additional labels "-l <ciflow/label_name>", which is equivalent to adding these labels manually and trigger the rerun
@pytorchbot ciflow rerun -l ciflow/scheduled -l ciflow/slow For more information, please take a look at the CI Flow Wiki. |
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit c96a77d (more details on the Dr. CI page):
🕵️ 16 new failures recognized by patternsThe following CI failures do not appear to be due to upstream breakages
|
[ghstack-poisoned]
cc albanD mruberry jbschlosser walterddr [ghstack-poisoned]
cc albanD mruberry jbschlosser walterddr [ghstack-poisoned]
cc albanD mruberry jbschlosser walterddr [ghstack-poisoned]
cc albanD mruberry jbschlosser walterddr [ghstack-poisoned]
cc albanD mruberry jbschlosser walterddr [ghstack-poisoned]
cc albanD mruberry jbschlosser walterddr [ghstack-poisoned]
@jbschlosser This one does not make it fully composite, but we could simply remove the backward formulas so that it is fully composite (I'd vouch for that). Philip is on PTO, but I could carve these PRs out of this stack and standalone PR that does so. |
@lezcano Sure, that sounds great if you're willing to do it! |
Benchmarks: #80334 (comment) Fixes #80158 Fixes #78867 Fixes #69230 Supersedes #79007 Supersedes #69212 Supersedes #19659 Pull Request resolved: #80334 Approved by: https://github.com/ezyang
Summary: Benchmarks: #80334 (comment) Fixes #80158 Fixes #78867 Fixes #69230 Supersedes #79007 Supersedes #69212 Supersedes #19659 Pull Request resolved: #80334 Approved by: https://github.com/ezyang Test Plan: contbuild & OSS CI, see https://hud.pytorch.org/commit/pytorch/pytorch/828c787ea98da39eb786925eedcb8527aae07153 Reviewed By: mehtanirav Differential Revision: D37604775 Pulled By: mehtanirav fbshipit-source-id: b188d47df5a3a820e5c15d9ce18b1a2c3f31f287
Benchmarks: #80334 (comment) Fixes #80158 Fixes #78867 Fixes #69230 Supersedes #79007 Supersedes #69212 Supersedes #19659 Pull Request resolved: #80334 Approved by: https://github.com/ezyang
Summary: Benchmarks: #80334 (comment) Fixes #80158 Fixes #78867 Fixes #69230 Supersedes #79007 Supersedes #69212 Supersedes #19659 Pull Request resolved: #80334 Approved by: https://github.com/ezyang Test Plan: contbuild & OSS CI, see https://hud.pytorch.org/commit/pytorch/pytorch/b5b9db9f844f4f100651c6afa57124fa5851edec Reviewed By: DanilBaibak Differential Revision: D37847477 Pulled By: DanilBaibak fbshipit-source-id: a04919bbd2b746c30c654b971efcf76ef27ac5a6
Stack from ghstack:
torch.nn.functional.l1_loss
for complex inputs #65681OpInfo
s fortorch.nn.functional.triplet_margin(_with_distance)?_loss
#67079OpInfo
s fornn.functional.binary_cross_entropy(_with_logits)?
#67023OpInfo
fortorch.nn.functional.pdist
#67022OpInfo
fortorch.nn.functional.(smooth_)?l1_loss
#69211cc @albanD @mruberry @jbschlosser @walterddr