After the model runs, VAE decoding prompts an error!
#1
by Feitianxi - opened
After the model runs, VAE decoding prompts an error!“Input img and context tensors must have 3 dimensions.”
Hey Feitianxi, that VAE error sounds like a shape mismatch between your image tensor and the model’s context. Could you share what dimensions your input img and context tensors are currently set to? That’ll help pinpoint where the decoding expects 3D but gets something else.