AI/diffusers
1
0
Fork 0
mirror of https://github.com/huggingface/diffusers.git synced 2026-01-27 17:22:53 +03:00
Code Activity

Fix HunyuanVideo produces NaN on PyTorch<2.5 (#10482)

Browse Source 
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
This commit is contained in:
hlky
2025-01-07 23:13:55 +00:00
committed by DN6
parent 2b432ac5a8
commit 13ea83f0fa
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/diffusers/models/transformers/transformer_hunyuan_video.py
View File

@@ -713,15 +713,15 @@ class HunyuanVideoTransformer3DModel(ModelMixin, ConfigMixin, PeftAdapterMixin,
condition_sequence_length = encoder_hidden_states.shape[1]
sequence_length = latent_sequence_length + condition_sequence_length
attention_mask = torch.zeros(
batch_size, sequence_length, sequence_length, device=hidden_states.device, dtype=torch.bool
) # [B, N, N]
batch_size, sequence_length, device=hidden_states.device, dtype=torch.bool
) # [B, N]
effective_condition_sequence_length = encoder_attention_mask.sum(dim=1, dtype=torch.int) # [B,]
effective_sequence_length = latent_sequence_length + effective_condition_sequence_length
for i in range(batch_size):
attention_mask[i, : effective_sequence_length[i], : effective_sequence_length[i]] = True
attention_mask = attention_mask.unsqueeze(1) # [B, 1, N, N], for broadcasting across attention heads
attention_mask[i, : effective_sequence_length[i]] = True
attention_mask = attention_mask.unsqueeze(1) # [B, 1, N], for broadcasting across attention heads
# 4. Transformer blocks
if torch.is_grad_enabled() and self.gradient_checkpointing: