Commit History

  1. A shorter intro with cross links, and drop the status footer Running

    multimodalart HF Staff
    dbb6d13
  2. Add a 1024x1024 1:1 canvas

    multimodalart HF Staff
    9553ab5
  3. Make the first frame optional: no keyframe runs t2va with the same locked audio rows

    multimodalart HF Staff
    8453576
  4. Prompt upsampling, and one configuration: locked audio rows on larryvrh v4

    multimodalart HF Staff
    a09476b
  5. A turbo LoRA selector: Off, larryvrh v4 (default) and joyfox, each with its own step contract

    multimodalart HF Staff
    5860db9
  6. Drop the __pycache__ that slipped into the last commit

    multimodalart HF Staff
    ea0d7e8
  7. Run the joyfox 4-step turbo LoRA by default, on diffusers main

    multimodalart HF Staff
    9c9c841

  1. Credit the example assets to the sibling Space

    multimodalart HF Staff
    a39d189
  2. Encode the soundtrack under native attention: the float32 audio VAE has no cuDNN kernel

    multimodalart HF Staff
    5311d91
  3. Clone the locked soundtrack back in: the next step overwrites the rows in place

    multimodalart HF Staff
    1cae2c0
  4. MiniMax-H3 audio to video: fl2va with the audio rows given

    multimodalart HF Staff
    9d32de2
  5. initial commit

    multimodalart HF Staff
    Verified signature
    658d9a2