Glenn Jocher

https://github.com/karpathy/minGPT/blob/5433de6158cc8bf9f1bf6c79b4e408292d6bc394/mingpt/trainer.py#L53-L58

@aletote it's used in the training loop for saving checkpoints. You can see this usage in trainer.py: https://github.com/karpathy/minGPT/blob/8909e1b646d6fd5235ec33259fb22fdc2c91037c/mingpt/trainer.py#L123-L129

@marxav @1910mauri ah, you're absolutely right, wrong repo (!). I've deleted my post, please ignore it.

@nebw I saw you also `with autocast():` inference when not training in sampler(). That's interesting, I haven't seen that use-case before. Do you know if autocast would correctly handle running...

@testproducts20 this seems unrelated to YOLOv5. If I run this Colab I see the same error: ```python !pip install -U albumentations import cv2 --------------------------------------------------------------------------- ImportError Traceback (most recent call last)...

@BloodAxe following up on this. Albumentation seems to have a serious Colab problem that users are encountering i.e. https://github.com/ultralytics/yolov5/issues/7014, https://github.com/albumentations-team/albumentations/issues/1140. You can reproduce in Colab with: ```bash !pip install -U...

@brownrc that's correct, albumentations are additive to existing augmentations. Though I think your conclusion is inconsistent with statistical additive noise, which always results in more noise and never cancels itself...

@dapiaoGe I'd refer your questions to the authors of the papers you mentioned if you have questions about them

@triple-Mu thanks for the PR! The easiest argument structure is to simply use an `--nms` arg that would be handled accordingly for formats that are nms capable.

@savchenkoyana PR is merged. Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐