This works on 2070ti 8gb card. a bit slow but can not complain about the quality. I could not get past the sampler on the LTX sites workflow so even though this is a far more complex flow it works even in low vram state. I did have issues with the audio node and could not add the separate audio flow Rocky published but I will be trying to work that out. I did also add a text input block for the prompting so as to not have to use strictly image to video and try some text to video. I am still tweaking it but will update if it works out. I could not edit the prompt used from Florence so I added the text input to change prompting. which makes it editable for me. I used the text to video node group LTX model has in t2v wf on site. Any way another great flow that brings video generation closer to us lower level equipment peeps LOL. Thanks for the great work! I am currently working on the mochi flow and hoping it works but waiting to see what you put out for that!
