This is an official implementation of our CVPR 2020 paper "HigherHRNet: Scale-Aware Representation Learning for Bottom-Up Human Pose Estimation" (https://arxiv.org/abs/1908.10357)
MIT License
1.32k
stars
271
forks
source link
Mismatch between pre trained weights file and config files in getting started #27
Just tried to run a sample training run as per install instructions but got an error on the command:
python tools/dist_train.py --cfg experiments/coco/higher_hrnet/w32_512_adam_lr1e-3.yaml
=> init stage4.2.fuse_layers.0.3.1.num_batches_tracked from models/pytorch/imagenet/hrnet_w32-36af842e.pth
Given groups=1, weight of size 32 32 3 3, expected input[1, 256, 128, 128] to have 32 channels, but got 256 channels instead
Error occurs, No graph saved
What I am I missing? What do I need to tweak in the yaml?
The COCO validation runs fine on the same pre-trained model.
Just tried to run a sample training run as per install instructions but got an error on the command: python tools/dist_train.py --cfg experiments/coco/higher_hrnet/w32_512_adam_lr1e-3.yaml
=> init stage4.2.fuse_layers.0.3.1.num_batches_tracked from models/pytorch/imagenet/hrnet_w32-36af842e.pth Given groups=1, weight of size 32 32 3 3, expected input[1, 256, 128, 128] to have 32 channels, but got 256 channels instead Error occurs, No graph saved
What I am I missing? What do I need to tweak in the yaml?
The COCO validation runs fine on the same pre-trained model.