WARNING: All log messages before absl::InitializeLog() is called are written to STDERR

[matanox]

Have I written custom code (as opposed to using a stock example script provided in MediaPipe)

Yes

OS Platform and Distribution

Ubuntu 22.04

Programming Language and version

python 3.10

MediaPipe version

0.10.11

Bazel version

not built from source.

Solution

Hands

Describe the actual behavior

logging going to stderr

Describe the expected behaviour

program logs gracefully

Standalone code/steps you may have used to try to get what you need

WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1714668459.875897  381002 gl_context_egl.cc:85] Successfully initialized EGL. Major : 1 Minor: 5
I0000 00:00:1714668459.877641  381112 gl_context.cc:357] GL version: 3.2 (OpenGL ES 3.2 Mesa 23.2.1-1ubuntu3.1~22.04.2), renderer: Mesa Intel(R) UHD Graphics 630 (CFL GT2)
INFO: Created TensorFlow Lite XNNPACK delegate for CPU.

It's nicer if framework code is refactored to avoid this.

[matanox]

Also, when no GPU model is used, why does it initialize OpenGL?

[matanox]

With 10.0.13, there seem to be more warnings on stdout:

.venv/lib/python3.10/site-packages/google/protobuf/symbol_database.py:55: UserWarning: SymbolDatabase.GetPrototype() is deprecated. Please use message_factory.GetMessageClass() instead. SymbolDatabase.GetPrototype() will be removed soon.
warnings.warn('SymbolDatabase.GetPrototype() is deprecated. Please '

Maybe something to clean up in 0.10.14, by avoiding soon deprecated protobuf api, or least favorably by constraining the protobuf version.

[schmidt-sebastian]

We will put this on our list of cleanup items but we are pretty busy these days :)

[matanox]

Yeah I know, sorry for being very active here.

[Kunz-David]

Im getting this too, but for the selfie model.