<h3 id="ltx-2.3">LTX 2.3 </h3>Ltx is quite a diva to train... This is my internal 12th revision of the lora. But this one i like. I have decided to take a revision that tends to do more slow motion but with sharp faces and good bounce.Enjoy and let me know in comments how you like it, for more speed you can add the word fastlyRecommended Strength 0.75Example Workflow: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1598638/ltx-wan-i2v-22-21-ai-build-art-workflows?modelVersionId=3002323">https://civarchive.com/models/1598638/ltx-wan-i2v-22-21-ai-build-art-workflows?modelVersionId=3002323</a> Wan 2.2V1: Initial Version fast movement less bounceV2: More bounce, a little less speed, more WAN 2.1 like Works good with my sample workflow: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1598638?modelVersionId=2212578">https://civarchive.com/models/1598638?modelVersionId=2212578</a><hr />This lora is for bounce walk videos. It's quite good in the continuation of the original image.Used I2V-14B-720P as base.Since ppl were asking about my training Workflow: I have used 7 videos with about 4 second length, with faces cut off. I have described what happens e.g. Here some captions:<ul><li>A brunette girl, is sitting on the bed of her bedroom and standing up, she wears white bikini, her breasts are bouncing, her face is cut off, she is walking into the direction of the moving camera, noisy image quality</li><li>A brunette girl, is walking through the floor of her office, she wears a white top and and a black skirt, her breasts are bouncing, her face is cut off, she is walking into the direction of the turning camera</li></ul> Then I used Musubi-Tuner with default parameters and trained about 32 epochs and then 16 epochs with larger training set, with my 4090 it was done in about 1h 35 min. That's all :)

Bouncy Walk - LTX / Wan I2V 14B

WalkV1_0_LTX23_I2V-000040.comfy.safetensors

LTX 2.3 v1.0

Files with SHA256: f67ad8e0151b73736d76444a482432c9458e291174b9c2a26c5ea53bad73a48f