CLoSD
CLoSD copied to clipboard
GuyTevet
CLoSD: Closing the Loop between Simulation and Diffusion for multi-task character control
For DIP, why the checkpoint provided in this repo is worse than the checkpoint provided in the MDM repo? Can I know what's the difference between them during training?
Hi, thank you for sharing this excellent work. I really enjoyed reading the paper. I’m trying to reproduce the training pipeline, but my GPU has limited memory. Since the README...
I'm trying to generate custom motions from user input. I've tried the solution provided here: https://github.com/GuyTevet/CLoSD/issues/6. Here's my code: ``` def update_mdm_conditions(self, env_ids): super().update_mdm_conditions(env_ids) # updates prompts and lengths try:...
Hello! Need a bit of help in setting the environment. I want to run this project on `WSL2` (with docker). (I do have Omniverse installed in the host (Window)) Since...
In AutoRegressive Sampler, when `required_frames` is a multiple of `pred_len`, the sampling wastes an iteration for nothing due to false rounding. For example, when `self.required_frames == 120` and `self.args.pred_len ==...
Thanks for the great project. I have a question regarding to headless mode. I'm running the example using closd_multitask with headless mode. However, I only see the on-going progress of...
Hi, I try To Run CLoSD Text-to-motion it run on my machine, but the output display like only give me 1 FPS the animation is very laggy on me I...
thanks for providing a way to export to blender as separate component based animation. Is there a way to export into armature format instead? I was trying to export the...
Running: ```bash python closd/run.py learning=im_big env=closd_t2m robot=smpl_humanoid learning.params.load_checkpoint=True learning.params.load_path=output/CLoSD/my_CLoSD_no_finetune/Humanoid.pth env.dip.cfg_param=2.5 env.num_envs=3072 has_eval=False epoch=-1 exp_name=my_CLoSD_t2m_finetune ``` I get: > OSError: [Errno 24] Too many open files With `HYDRA_FULL_ERROR=1`, I get the...
Hi, thank you for your awesome projects. Recently I tried to run the script to train the policy, but I got a core dumped issue when I prepared all pkgs...
