parryppp
parryppp
The image of bool_matrix1024, bool_matrix4096 are shown as belowed.
sorry, i am fresh to image/ video generation, what's the encoder in the paper
https://github.com/deepinsight/insightface/blob/786c4a8327398aecb4cad0cb83ebcefc12b9d3cb/recognition/arcface_torch/backbones/iresnet.py#L160 why not use fp16 these two lines?
 
https://github.com/modelscope/DiffSynth-Studio/blob/main/diffsynth/pipelines/wan_video_new.py#L1073
i do not find this file, torchrun --nproc-per-node 8 nemo/collections/diffusion/train.py --yes --factory pretrain_xl