David Berard
David Berard
@davidberard98 has imported this pull request. If you are a Facebook employee, you can view this diff [on Phabricator](https://www.internalfb.com/diff/D38560212).
@pytorchbot merge
@DavidSlayback does the same error occur with LayerNorm+CUDA but without JIT?
@jjsjann123 so does the issue go away when we turn off nvfuser?
ConstantChunk is inserted as part of CanonicalizeOps, which I think will happen either way: https://github.com/pytorch/pytorch/blob/master/torch/csrc/jit/runtime/profiling_graph_executor_impl.cpp#L340 Is that the optimization you're talking about or is there another detail ConstantChunk that you're...
@xuzhao9 @zzpmiracle so, a few comments: 1. https://pytorch.org/docs/stable/jit.html#attributes is actually outdated, typically you will not need to annotate to preserve parameters. See below for an example. 2. In this case,...
@davidberard98 has imported this pull request. If you are a Meta employee, you can view this diff [on Phabricator](https://www.internalfb.com/diff/D53950998).
@pytorchbot merge
@davidberard98 has imported this pull request. If you are a Meta employee, you can view this diff [on Phabricator](https://www.internalfb.com/diff/D54269082).
@davidberard98 has imported this pull request. If you are a Meta employee, you can view this diff [on Phabricator](https://www.internalfb.com/diff/D54269082).