Bhchen
Bhchen
In paper, it claims that Maxnorm regularization is applied per iteration, but in the code, it seems to be applied per epoch? Am i right?
I find with this configuration, 80G A100 is not enough for even batch_size=1. Am I right?
Will the code and model be released?
When I replace the attn processor with this project, I find the used GPU memory will be increased. Who knows the reason?
https://github.com/AIGCDesignGroup/ReplaceAnything/issues/20 congratulations to this work first. But I found the gradio demo on HF is highly similar as our ReplaceAnything. If it borrows our work, please add some acknowledgements to...
Hi, I met the following problem: ```python Handling 3D object files: 36%|████████████████████████████████ | 4/11 [00:02
OSError: Lightricks/LTX-Video does not appear to have a file named ltx-video-2b-v0.9.5.safetensors.
when using ` transformer = LTXVideoTransformer3DModel.from_single_file("https://huggingface.co/Lightricks/LTX-Video/ltx-video-2b-v0.9.5.safetensors", torch_dtype=torch.bfloat16) ` error shows out like ` OSError: Lightricks/LTX-Video does not appear to have a file named ltx-video-2b-v0.9.5.safetensors. `