Improved AnimateDiff for ComfyUI and Advanced Sampling Support
Apache License 2.0
2.64k
stars
196
forks
source link
Is there an option to use the context of the previous queue for the next prompt queue, so that multiple video fragments can be combined into one video and the context is coherent? #391
Is there such an option? The next prompt queue uses the context of the previous queue, so that multiple video fragments can be used to synthesize a video. The context is coherent. My video has a total of 1000 frames. I want to process 20 frames each time. Use It was generated 50 times, but the context information has been lost the next time it is generated and is not in the memory. Can the last context be used? After the operation is completed, the context can be saved to the hard disk or in memory, such as a persistent context.
Is there such an option? The next prompt queue uses the context of the previous queue, so that multiple video fragments can be used to synthesize a video. The context is coherent. My video has a total of 1000 frames. I want to process 20 frames each time. Use It was generated 50 times, but the context information has been lost the next time it is generated and is not in the memory. Can the last context be used? After the operation is completed, the context can be saved to the hard disk or in memory, such as a persistent context.