David Stefan Kleindiek

HI! I experienced the same problem. the error seems to be in the [SimOTAAssigner ](https://github.com/open-mmlab/mmpose/models/task_modules/assigners/sim_ota_assigner.py) class in mmpose\models\task_modules\assigners\sim_ota_assigner.py Line 158: `F.one_hot(gt_labels.to(torch.int64), pred_scores.shape[-1]).float().unsqueeze(0).repeat( num_valid, 1, 1))` The error is thrown here,...

I have read the CLA Document and I sign the CLA

I have read the CLA Document and I sign the CLA recheck

Hi @dengdaixiaodi666, great to hear that you are utilizing the model. I experienced the same problem initial, you will have to tune the loss weights so that the keypoint loss...

@glenn-jocher I just merged the newest main branch into the pull request, is there any intention on merging the multitask model into the main branch? I think there is quite...

Did you try setting it at the training step via `results = model.train(pose=20)`? The suggested change in the `model.yaml` file came from the ultralytics team, I never tried it.

Hi @dengdaixiaodi666 indeed the code is not official, I changed it on my own. Please try to experiment with different pose loss weights by changing it during training by modifying...

Hi everybody, thanks for your interest in this repo. However, I am currently no longer maintaining it. A lot of changes were automatically merged into this repo. Please ensure that...