# Bizarre InternalTorchDynamoError with locally and formerly working code

**URL:** <https://deeptalk.lambda.ai/t/bizarre-internaltorchdynamoerror-with-locally-and-formerly-working-code/4499>\
**Category:** Uncategorized\
**Created:** [February 7, 2025, 8:17pm UTC](https://deeptalk.lambda.ai/t/bizarre-internaltorchdynamoerror-with-locally-and-formerly-working-code/4499 "2025-02-07T20:17:14Z")\
**Posts on this page:** 1\
**Showing post:** 3

<div class="post-metadata">

**Author:** ![EIFY](https://avatars.discourse-cdn.com/v4/letter/e/e79b87/32.png) [@EIFY](https://deeptalk.lambda.ai/u/EIFY)\
**Post date:** [February 8, 2025, 6:29am UTC](https://deeptalk.lambda.ai/t/bizarre-internaltorchdynamoerror-with-locally-and-formerly-working-code/4499/3 "2025-02-08T06:29:03Z")

</div>

Update: It turned out that I just forgot the `--multiprocessing-distributed` flag, which is critical for multi-GPU instance. See [Bizarre InternalTorchDynamoError with locally and formerly working code - #5 by ptrblck - PyTorch Forums](https://discuss.pytorch.org/t/bizarre-internaltorchdynamoerror-with-locally-and-formerly-working-code/216382/5) for details.

---

_[View the full topic](https://deeptalk.lambda.ai/t/bizarre-internaltorchdynamoerror-with-locally-and-formerly-working-code/4499)._