minimax-H3 FL2VA train error
Author: stihuangyuanCreated Sep 1, 2026Updated Sep 16, 2026
[rank5]: noise_pred, noise_pred_audio = pipe.model_fn( [rank5]: ^^^^^^^^^^^^^^ [rank5]: TypeError: model_fn_minimax_h3() missing 1 required positional argument: 'audio_latents'
Source: modelscope/DiffSynth-Studio