mirror of
https://github.com/zebrajr/pytorch.git
synced 2025-12-08 07:39:33 +01:00
torch.profiler.record_function and torch.profiler.profile are ignored by dynamo. In the common case, users have `record_function` in the middle of their program in order to annotate a section of the profile. The previous error message was `Profiler will be ignored`. Users would think that profiling would be completely ignored. Now the message will look like `Profiler function <class 'torch.autograd.profiler.record_function'> will be ignored`. Pull Request resolved: https://github.com/pytorch/pytorch/pull/105362 Approved by: https://github.com/yanboliang, https://github.com/aaronenyeshi |
||
|---|---|---|
| .. | ||
| __init__.py | ||
| base.py | ||
| builder.py | ||
| builtin.py | ||
| constant.py | ||
| ctx_manager.py | ||
| dicts.py | ||
| functions.py | ||
| higher_order_ops.py | ||
| lists.py | ||
| misc.py | ||
| nn_module.py | ||
| optimizer.py | ||
| tensor.py | ||
| torch.py | ||
| user_defined.py | ||