WIP: video_unet_atten

[wr0124]

• [ ] step0 Get a similar architecture of UNet(ResBlock+AttentionBlock) in joliGEN comparable as in AnimateDiff(ResBlock+TransformerBlock+MotionModule)

• [ ] step1 Modify ResBlock to process 5D tensor image for input and output

• [ ] step1.1 test ResBlock to process 5D tensor for input and output

• [ ] step1.2 ResBlock embedding ?

• [ ] step2 Replace AttentionBlock by MotionModule

• [ ] step2.1 Use QKVAttention class for attention score in MotionModule

• [ ] step2.2 Use PositionalEncoding1D for position encoding ?

• [ ] step2.3 test AttentionBlock by MotionModule for 5D tensor for input and output

• [ ] step2.4 MotionModule embedding ?

• [ ] step3 Merge MotionModule in the Video_generator_attn file

• [ ] step3.1 Aligned attention head, input/output channels and maybe other parameters.

• [ ] step3.2 test MotionModule in the Video_generator_attn file for 5D tensor input/output

• [ ] step4 Test UNet for 5D tensor input/output ?

• [ ] step5 Create Dataloader

• [ ] step5 Test UNet with Dataloader