Hello,
Firstly, I would like to thank you for the excellent work. It has been very useful. I am currently using the script (Visualization Part 2) to create SMPL meshes by running the following command:
python -m fit --dir YOUR_NPY_FOLDER --save_folder TEMP_PLY_FOLDER --cuda
However, I am encountering an issue when I attempt to increase the batchSize to enhance the conversion process. Initially, the default batchSize is set to 1. When I modified the batchSize to 16 to process multiple batches simultaneously, I started experiencing problems.
Traceback (most recent call last):
File "/root/miniconda3/envs/mgpt/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/root/miniconda3/envs/mgpt/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/root/autodl-tmp/MotionGPT/fit.py", line 204, in
pred_betas[0, :] = init_mean_shape
RuntimeError: expand(torch.cuda.FloatTensor{[16, 10]}, size=[10]): the number of sizes provided (1) must be greater or equal to the number of dimensions in the tensor (2)
Could you please provide some advice on how to modify the batchSize or any other settings that could help improve the efficiency of the script(using batchSize=1 is too slow)? Any suggestions or guidance would be greatly appreciated.
Thank you for your support.
Hello, Firstly, I would like to thank you for the excellent work. It has been very useful. I am currently using the script (Visualization Part 2) to create SMPL meshes by running the following command:
However, I am encountering an issue when I attempt to increase the
batchSize
to enhance the conversion process. Initially, the defaultbatchSize
is set to 1. When I modified thebatchSize
to 16 to process multiple batches simultaneously, I started experiencing problems.Traceback (most recent call last): File "/root/miniconda3/envs/mgpt/lib/python3.10/runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "/root/miniconda3/envs/mgpt/lib/python3.10/runpy.py", line 86, in _run_code exec(code, run_globals) File "/root/autodl-tmp/MotionGPT/fit.py", line 204, in
pred_betas[0, :] = init_mean_shape
RuntimeError: expand(torch.cuda.FloatTensor{[16, 10]}, size=[10]): the number of sizes provided (1) must be greater or equal to the number of dimensions in the tensor (2)
Could you please provide some advice on how to modify the batchSize or any other settings that could help improve the efficiency of the script(using batchSize=1 is too slow)? Any suggestions or guidance would be greatly appreciated. Thank you for your support.