Fix sqlite3 import

[pearu]

As in the title.

Addresses sqlite3 import error seen in https://github.com/pytorch/audio/actions/runs/19372860904/job/55432849539:

Using Sphinx-Gallery to convert rst text blocks to markdown for .ipynb files.
  generating gallery for tutorials... [  7%] audio_datasets_tutorial.py
  WARNING: /pytorch/audio/examples/tutorials/audio_datasets_tutorial.py failed to execute correctly: Traceback (most recent call last):
    File "/pytorch/audio/examples/tutorials/audio_datasets_tutorial.py", line 23, in <module>
      import IPython
    File "/pytorch/audio/ci_env/lib/python3.11/site-packages/IPython/__init__.py", line 57, in <module>
      from .terminal.embed import embed
    File "/pytorch/audio/ci_env/lib/python3.11/site-packages/IPython/terminal/embed.py", line 15, in <module>
      from IPython.core.interactiveshell import InteractiveShell, make_main_module_type
    File "/pytorch/audio/ci_env/lib/python3.11/site-packages/IPython/core/interactiveshell.py", line 69, in <module>
      from IPython.core.displayhook import DisplayHook
    File "/pytorch/audio/ci_env/lib/python3.11/site-packages/IPython/core/displayhook.py", line 19, in <module>
      from .history import HistoryOutput
    File "/pytorch/audio/ci_env/lib/python3.11/site-packages/IPython/core/history.py", line 48, in <module>
      from sqlite3 import DatabaseError, OperationalError
    File "/pytorch/audio/ci_env/lib/python3.11/sqlite3/__init__.py", line 57, in <module>
      from sqlite3.dbapi2 import *
    File "/pytorch/audio/ci_env/lib/python3.11/sqlite3/dbapi2.py", line 27, in <module>
      from _sqlite3 import *
  ImportError: /usr/lib64/libstdc++.so.6: version `GLIBCXX_3.4.30' not found (required by /pytorch/audio/ci_env/lib/python3.11/lib-dynload/../.././libicuuc.so.75)

Inserting $CONDA_PREFIX/lib to LD_LIBRARY_PATH fixes the issue. However, the origin of the issue remains unknown.

PLEASE NOTE THAT THE TORCHAUDIO REPOSITORY IS NO LONGER ACTIVELY MONITORED. You may not get a response. For open discussions, visit https://discuss.pytorch.org/.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/audio/4130

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 5 New Failures

As of commit 8178ed6 with merge base 07b6cbd ():

NEW FAILURES - The following jobs have failed:

• Build Windows Wheels / pytorch/audio / build-wheel-py3_10-cuda12_6 (gh)
  ninja: build stopped: subcommand failed
• Build Windows Wheels / pytorch/audio / build-wheel-py3_10-cuda13_0 (gh)
  ninja: build stopped: subcommand failed
• Build Windows Wheels / pytorch/audio / upload / upload-wheel-py3_10-cuda12_6 (gh)
  Unable to download artifact(s): Artifact not found for name: pytorch_audio__3.10_cu126_x64
• Build Windows Wheels / pytorch/audio / upload / upload-wheel-py3_10-cuda13_0 (gh)
  Unable to download artifact(s): Artifact not found for name: pytorch_audio__3.10_cu130_x64
• Unit-tests on Windows GPU / build / windows-job (gh)
  ninja: build stopped: subcommand failed

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[pearu]

The fix here is same as in meta-pytorch/torchcodec#1034 and windows CI failures are unrelated.