No such file or directory: 'weights.tar'

[Recialhot]

Hello, thank you for your work. I would like to check if the code I installed is correct, but there is no weights.tar, so I cannot directly evaluate it. It seems that your expression can directly test the results of your paper using weights.tar, but I cannot find the corresponding file. Do you have it?

[SlimeVRX]

Sorry, I don't have weights.tar file.

[SlimeVRX]

I just discovered that there is a newly released paper also called FIP, with a different author, with source code, you can refer to it

Name: Fast Human Motion reconstruction from sparse inertial measurement units considering the human shape (FIP)

Paper: https://www.nature.com/articles/s41467-024-46662-5

Source: https://github.com/bachongyou/FIP_inference

[Recialhot]

我刚刚发现有一篇新发表的论文也叫FIP，作者不同，有源码，可以参考

名称：考虑人体形状的稀疏惯性测量单元的快速人体运动重建 （FIP）

论文：https://www.nature.com/articles/s41467-024-46662-5

来源： https://github.com/bachongyou/FIP_inference

thanks,i will see it

[Recialhot]

对不起，我没有weights.tar文件。

hello, when i run train.py, it will cause this probelm, that's why? thank you

batch_size == 128
model == RNN
fineturning == False
cuda == True
resume ==
evaluate == False
workers == 4
half == False
lr == 0.0005
weight_decay == 0
save_freq == 30
save_dir == save_pose
start_epoch == 0
epochs == 300
visdom == False
train len 150, vail len150
0%| | 0/150 [00:01<?, ?it/s]
Traceback (most recent call last):
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/train.py", line 287, in
main()
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/train.py", line 270, in main
train_loss = train(train_loader, model, criterion, optimizers, epoch, args.fineturning)
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/train.py", line 87, in train
output = model.forward_my((imu, leaf_jtr, full_jtr), refine=refine)
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/net.py", line 114, in forward_my
global_reduced_pose = self.pose_net.forward(imu)[0]
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/net.py", line 20, in forward
x, h = self.rnn(self.dropout(relu(self.linear1(x))), h)
File "/home/HDD2/yanshiqi/.conda/envs/avatarjlm/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/home/HDD2/yanshiqi/.conda/envs/avatarjlm/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1527, in call_impl
return forward_call(*args, **kwargs)
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/sru/modules.py", line 634, in forward
h, c = rnn(prevx, c0[i], mask_pad=mask_pad)
File "/home/HDD2/yanshiqi/.conda/envs/avatarjlm/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1518, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/home/HDD2/yanshiqi/.conda/envs/avatarjlm/lib/python3.9/site-packages/torch/nn/modules/module.py", line 1527, in _call_impl
return forward_call(*args, **kwargs)
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/sru/modules.py", line 276, in forward
h, c = self.apply_recurrence(U, V, residual, c0, scale_val, mask_c, mask_pad)
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/sru/modules.py", line 298, in apply_recurrence
return elementwise_recurrence_gpu(U, residual, V, self.bias, c0,
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/sru/ops.py", line 129, in elementwise_recurrence_gpu
return ElementwiseRecurrence.apply(
File "/home/HDD2/yanshiqi/.conda/envs/avatarjlm/lib/python3.9/site-packages/torch/autograd/function.py", line 539, in apply
return super().apply(*args, **kwargs) # type: ignore[misc]
File "/home/HDD2/yanshiqi/FDIP_code/FDIP/sru/cuda_functional.py", line 176, in forward
h, last_hidden, c = elementwise_recurrence_forward(
RuntimeError: Caught an unknown exception!
ERROR conda.cli.main_run:execute(41): conda run python /home/HDD2/yanshiqi/FDIP_code/FDIP/train.py -c -b 128 --epochs 300 --save-dir save_pose --lr 5e-4 -m RNN failed. (See above for error)

[SlimeVRX]

@Recialhot Sorry! I can't check errors for you at the moment.