Pablo Vela

Would be great to add support for whole body pose estimation dataset (body+face+hands) via [COCO-WholeBody](https://github.com/jin-s13/COCO-WholeBody)

Support for [InterHand2.6](https://mks0601.github.io/InterHand2.6M/)

Would be great to see the integration of a hyperparameter tuner like [Optuna](https://github.com/optuna/optuna)

Albumentations augmetnations similar to [mmclassification](https://github.com/open-mmlab/mmclassification/blob/996d2a76412422bcfb6e2ebd486c4cff97ad49b8/mmcls/datasets/pipelines/transforms.py#L499)

[Lite-HRNet](https://github.com/HRNet/Lite-HRNet), its already built with mmpose, so including into the main repo should be super simple. Would be amazing if it could work with the pytorch2onnx tool for deployment

Motivation: Have a smaller model more well suited for realtime inference for the interhand 3d dataset Configs: mobilenetv2_interhand3d_all_256x256.py Datasets: Interhand3D Details: Would also be great to see smaller input sizes...

You'd have to convert the trained model to torchscript and do some optimizations, you'll be able to find more info here https://pytorch.org/mobile/home/ and example apps here https://github.com/pytorch/android-demo-app

Understood, so I had a chance to try and train the model using the provided config. I'm using a machine with 128gb of ram and 2 A6000 gpus. When I...

@zengwang430521 so with the current implementation it seems like there are basically two solutions if using distributed single node training 1. Reduce the number of workers (in my case I...

Good questions! Here are my proposals **Number of images** Visualization here in my opinion is mostly as a sanity check to ensure the network is actually learning what's expected. This...