Octopus b8bb5427f9 fix: make attention_mask optional in LTXBaseModel.forward (fixes #13299) ... The LTXBaseModel.forward and _forward methods required attention_mask as a positional argument with no default value. However, LTXV.extra_conds only conditionally adds attention_mask to model_conds when it is present in kwargs. If attention_mask is not provided by the text encoder, the diffusion_model forward call fails with: TypeError: LTXBaseModel.forward() missing 1 required positional argument: 'attention_mask' The model already handles attention_mask=None correctly in both _prepare_attention_mask and _prepare_context, so making the parameter optional is the minimal safe fix. This also aligns with how LTXAVDoubleStreamBlock.forward handles the parameter.		2026-04-06 13:17:51 +08:00
..
vae	ltx: vae: Fix missing init variable (#13074)	2026-03-19 22:34:58 -04:00
vocoders	LTX audio vae novram fixes. (#12796)	2026-03-05 16:31:28 -05:00
av_model.py	feat: LTX2: Support reference audio (ID-LoRA) (#13111)	2026-03-23 18:22:24 -04:00
embeddings_connector.py	Support the LTXAV 2.3 model. (#12773)	2026-03-04 20:06:20 -05:00
latent_upsampler.py	Support the LTXV 2 model. (#11632)	2026-01-05 01:58:59 -05:00
model.py	fix: make attention_mask optional in LTXBaseModel.forward (fixes #13299)	2026-04-06 13:17:51 +08:00
symmetric_patchifier.py	Support the LTXV 2 model. (#11632)	2026-01-05 01:58:59 -05:00