Unified-Robots comments

Repositories
Issues
Comments

Results 2 comments of


                                            Unified-Robots

Why the parameters of the generator(Stage_C) model need to calculate the gradient when training the control net？

In my opinion, the training of the controlnet dependents on the parameters, as the controlnet injects the information into the stageC modules to control the contents of the generated images.

How to set "use_fsdp=True" with "SLURM_LOCALID" and "SLURM_PROCID" for multi-gpus training?

> > 1. Run training script with torchrun instead of srun > > (e.g., torchrun --standalone --nnodes 1 --nproc_per_node 2 train/~~~~) > > 2. Exchange SLURM_LOCALID, SLURM_PROCID, SLURM_NNODES to LOCAL_RANK,...