Closed toyxyz closed 1 year ago
When I upload the video in Vid2VId tab it works as txt2vid not vid2vid. I also tried 'Input video path' but it doesn't work.
Go to the Vid2vid tab, upload a single video file and press the Generate button.
Vid2vid, not txt2vid, should work.
webui commit id - a9fed7c364061ae6efb37f797b6b522cb3cf7aa2 txt2vid commit id - 44a82864a0134bfc8d78456e8a63532f38a57cde
3090ti
Local PC setup (Windows)
Git commit: 44a82864 (Fri Mar 24 19:57:10 2023) Starting text2video Pipeline setup config namespace(framework='pytorch', task='text-to-video-synthesis', model={'type': 'latent-text-to-video-synthesis', 'model_args': {'ckpt_clip': 'open_clip_pytorch_model.bin', 'ckpt_unet': 'text2video_pytorch_model.pth', 'ckpt_autoencoder': 'VQGAN_autoencoder.pth', 'max_frames': 16, 'tiny_gpu': 1}, 'model_cfg': {'unet_in_dim': 4, 'unet_dim': 320, 'unet_y_dim': 768, 'unet_context_dim': 1024, 'unet_out_dim': 4, 'unet_dim_mult': [1, 2, 4, 4], 'unet_num_heads': 8, 'unet_head_dim': 64, 'unet_res_blocks': 2, 'unet_attn_scales': [1, 0.5, 0.25], 'unet_dropout': 0.1, 'temporal_attention': 'True', 'num_timesteps': 1000, 'mean_type': 'eps', 'var_type': 'fixed_small', 'loss_type': 'mse'}}, pipeline={'type': 'latent-text-to-video-synthesis'}) device cuda Working in txt2vid mode latents torch.Size([1, 4, 24, 32, 32]) tensor(-0.0031, device='cuda:0') tensor(1.0040, device='cuda:0') DDIM sampling tensor(1): 100%|█████████████████████████████████████████████████████████| 31/31 [00:22<00:00, 1.35it/s] STARTING VAE ON GPU. 24 CHUNKS TO PROCESS VAE HALVED DECODING FRAMES VAE FINISHED torch.Size([24, 3, 256, 256]) output/mp4s/20230325_072326329309.mp4 0%| | 0/1 [00:00<?, ?it/s]latents torch.Size([1, 4, 24, 32, 32]) tensor(-0.0048, device='cuda:0') tensor(0.9962, device='cuda:0') DDIM sampling tensor(1): 100%|█████████████████████████████████████████████████████████| 31/31 [00:15<00:00, 2.06it/s] STARTING VAE ON GPU. 24 CHUNKS TO PROCESS██████████████████████████████████████████████| 31/31 [00:15<00:00, 2.03it/s] VAE HALVED DECODING FRAMES VAE FINISHED torch.Size([24, 3, 256, 256]) output/mp4s/20230325_072346565309.mp4 text2video finished, saving frames to F:\\Download\\SDWEB_OUTPUT\\img2img-images\text2video-modelscope\20230325072237 Got a request to stitch frames to video using FFmpeg. Frames: F:\\Download\\SDWEB_OUTPUT\\img2img-images\text2video-modelscope\20230325072237\%06d.png To Video: F:\\Download\\SDWEB_OUTPUT\\img2img-images\text2video-modelscope\20230325072237\vid.mp4 Stitching *video*... Stitching *video*... Video stitching done in 0.46 seconds! t2v complete, result saved at F:\\Download\\SDWEB_OUTPUT\\img2img-images\text2video-modelscope\20230325072237
### Additional information _No response_
See https://github.com/deforum-art/sd-webui-modelscope-text2video/issues/62
and try lowering your denoising strength from 1 to around 0.5
Is there an existing issue for this?
Are you using the latest version of the extension?
What happened?
When I upload the video in Vid2VId tab it works as txt2vid not vid2vid. I also tried 'Input video path' but it doesn't work.
Steps to reproduce the problem
Go to the Vid2vid tab, upload a single video file and press the Generate button.
What should have happened?
Vid2vid, not txt2vid, should work.
WebUI and Deforum extension Commit IDs
webui commit id - a9fed7c364061ae6efb37f797b6b522cb3cf7aa2 txt2vid commit id - 44a82864a0134bfc8d78456e8a63532f38a57cde
What GPU were you using for launching?
3090ti
On which platform are you launching the webui backend with the extension?
Local PC setup (Windows)
Settings
Console logs