ask4help
ask4help copied to clipboard
Invalid Pretrained Checkpoints
It seems as though the pretrained checkpoints don't work with the current model. Running the given commands with the correct paths yields error
RuntimeError: Error(s) in loading state_dict for ResnetTensorObjectNavActorCritic:
size mismatch for goal_visual_encoder.resnet_compressor.0.weight: copying a param with shape torch.Size([128, 512, 1, 1]) from checkpoint, th
e shape in current model is torch.Size([128, 2048, 1, 1]).
size mismatch for actor.linear.weight: copying a param with shape torch.Size([4, 512]) from checkpoint, the shape in current model is torch.S
ize([6, 512]).
size mismatch for actor.linear.bias: copying a param with shape torch.Size([4]) from checkpoint, the shape in current model is torch.Size([6]
).
perhaps the checkpoints need to be updated? Please advise.