pytorch/torch/_logging/__init__.py
Angela Yi d9e87fb339 [draft-export] Include guards for constraint violation errors (#138748)
Summary:
Added where logs are being added to constrain violations in draft export.

Example output:
```
1. Constraint violation error.
    The specified input dynamic_shapes spec was found to be incorrect during tracing.
    Specifically, this guard was added: Eq(s0, 3), where {'s0': "L['args'][0][0].size()[0]"}.
    This occured at the following stacktrace:
        File /data/users/angelayi/fbsource/buck-out/v2/gen/fbcode/1beb9df83fd74b9a/scripts/angelayi/draft_export/__test_draft_export__/test_draft_export#link-tree/torch/nn/modules/module.py, lineno 1736, in _wrapped_call_impl
        File /data/users/angelayi/fbsource/buck-out/v2/gen/fbcode/1beb9df83fd74b9a/scripts/angelayi/draft_export/__test_draft_export__/test_draft_export#link-tree/torch/nn/modules/module.py, lineno 1747, in _call_impl
        File /data/users/angelayi/fbsource/buck-out/v2/gen/fbcode/1beb9df83fd74b9a/scripts/angelayi/draft_export/__test_draft_export__/test_draft_export#link-tree/scripts/angelayi/draft_export/test_draft_export.py, lineno 138, in forward.
    Because of this, we have modified the dynamic shapes structure to be the following:
    ```
    dynamic_shapes = {'a': {0: 3}}
    ```
```

The result of this diff is also that `dynamic` logs are permanently turned on during draft export. Otherwise we cannot capture the `[guard added]` logs from symbolic_shapes.py.

Test Plan: `buck2 run @//mode/dev-nosan scripts/angelayi/draft_export:test_draft_export -- -r "test_shape_failure" `

Differential Revision: D64862374

Pull Request resolved: https://github.com/pytorch/pytorch/pull/138748
Approved by: https://github.com/ezyang
2024-10-30 00:24:17 +00:00

20 lines
799 B
Python

# Top level logging module for torch logging
# Design doc: https://docs.google.com/document/d/1ZRfTWKa8eaPq1AxaiHrq4ASTPouzzlPiuquSBEJYwS8/edit#
# Simple setup for onboarding (see above doc for more detail):
# 1. register any top-level log qualified name for your module in torch._logging._registrations (see there for examples)
# 2. register any artifacts (<artifact_name> below) in torch._logging._registrations
# a. call getArtifactLogger(__name__, <artifact_name>) at your logging site instead of the standard logger to log your artifact
import torch._logging._registrations
from ._internal import (
_init_logs,
DEFAULT_LOGGING,
dtrace_structured,
get_structured_logging_overhead,
getArtifactLogger,
LazyString,
set_logs,
trace_structured,
warning_once,
)