KeyError: 'time_text_embed.timestep_embedder.linear_1.weight._data'

[ssizd]

Hello, I am training on Runpod, and although I successfully conducted training before this error appeared, the actual results were not satisfactory, so I tweaked the configuration file. However, the error started appearing after that. Even after rolling back to the original settings, the error persists. Since the training did run previously, I don't think the issue is with the PyTorch version. My multidatabackend settings are identical to the example, except for the paths. I am attaching the config file as well.

config.txt

[bghira]

duplicate of #659