essl
essl
@JiankunShi I suspect the issue comes from not setting the line auto_scale_lr = dict(enable=False, base_batch_size=32) to enable=True and the fact that the batch size is different. If i am not...
The mini version is called v1.0-mini and the full dataset v1.0-trainval. That's why it's not working.
@gorkemguzeler the repo is working as expected for me. Haven't trained lidar-only but I got 65 mAP after 3 epochs of training the bevfusion model with the lidar-only base. Oh...
@curiosity654 ohh sry it was a typo. I meant 3090 (24G RAM), so bs 2 per GPU. Do you think the issue could have to do with the batchnorm layers?...