ValueError: Cannot load file containing pickled data when allow_pickle=False

[CGMikeG]

(easymocap) M:\GITHUB\EASYMOCAP>python apps/demo/1v1p_mirror.py zju-m-demo/ --out zju-m-demo-output/ --vis_smpl --video

Demo code for single view and one person with mirror:

- Input : zju-m-demo/: [backflip, dance, kaluli]
- Output: zju-m-demo-output/
- Body  : smpl => neutral, body25

-> [Loading smpl, neutral]: 4.8s -> [Loading SPIN ]: 399.0ms Traceback (most recent call last): File "apps/demo/1v1p_mirror.py", line 153, in device=body_model.device) File "C:\ProgramData\Miniconda3\envs\easymocap\lib\site-packages\easymocap\estimator\SPIN\spin_api.py", line 178, in init model = hmr(SMPL_MEAN_PARAMS).to(device) File "C:\ProgramData\Miniconda3\envs\easymocap\lib\site-packages\easymocap\estimator\SPIN\models.py", line 175, in hmr model = HMR(Bottleneck, [3, 4, 6, 3], smpl_mean_params, **kwargs) File "C:\ProgramData\Miniconda3\envs\easymocap\lib\site-packages\easymocap\estimator\SPIN\models.py", line 102, in init mean_params = np.load(smpl_mean_params) File "C:\ProgramData\Miniconda3\envs\easymocap\lib\site-packages\numpy\lib\npyio.py", line 445, in load raise ValueError("Cannot load file containing pickled data " ValueError: Cannot load file containing pickled data when allow_pickle=False

[chingswy]

We can't reproduce this error.

[CGMikeG]

I have change false to true and some other suggestion I got from google. Now I am getting out of memory error. I will try to reduce batch_size, epochs, and load and see if that will help.