Boyuan Deng
Boyuan Deng
UPDATE: This seems a memory-related issue. When using a small model and setting `batch_size=1` and `num_workers=1`, training runs well. But increasing `batch_size` or `num_workers` easily leads to this error ```...
``` Nuitka-Wheel:INFO: Building: 'a.b' with command '['/tmp/build-env-l8xyb6zz/bin/python', '-m', 'nuitka', '--module', '--enable-plugin=pylint-warnings', '--output-dir=/tmp/build-via-sdist-n3f9j7e3/a.b-0.0.1/build/lib/a', '--nofollow-import-to=*.tests', '--remove-output', '--include-package=a.b', '/tmp/build-via-sdist-n3f9j7e3/a.b-0.0.1/build/lib/a/b']' ``` @kayhayen If you need the entire building log, I can post it below.
Hi, is there any update? I also want to customize the `local_scheme` part, to add a PyTorch-style CUDA version suffix like `+cu117` to my package version.