From da2fb54468e50f265d3d9a97d1e8ccf528ac1821 Mon Sep 17 00:00:00 2001 From: kijai <40791699+kijai@users.noreply.github.com> Date: Thu, 19 Dec 2024 17:03:10 +0200 Subject: [PATCH] remove print --- .../pipelines/pipeline_hunyuan_video.py | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/hyvideo/diffusion/pipelines/pipeline_hunyuan_video.py b/hyvideo/diffusion/pipelines/pipeline_hunyuan_video.py index 8d8572a..78f41a2 100644 --- a/hyvideo/diffusion/pipelines/pipeline_hunyuan_video.py +++ b/hyvideo/diffusion/pipelines/pipeline_hunyuan_video.py @@ -690,17 +690,17 @@ class HunyuanVideoPipeline(DiffusionPipeline): )) for c in context_queue: partial_latent_model_input = latent_model_input[:, :, c, :, :] - print("partial_latent_model_input", partial_latent_model_input.shape) + #print("partial_latent_model_input", partial_latent_model_input.shape) with torch.autocast( device_type="cuda", dtype=self.base_dtype, enabled=True): - noise_pred[:, :, c, :, :] += self.transformer( # For an input image (129, 192, 336) (1, 256, 256) - partial_latent_model_input, # [2, 16, 33, 24, 42] - t_expand, # [2] - text_states=input_prompt_embeds, # [2, 256, 4096] - text_mask=input_prompt_mask, # [2, 256] - text_states_2=input_prompt_embeds_2, # [2, 768] - freqs_cos=freqs_cos, # [seqlen, head_dim] - freqs_sin=freqs_sin, # [seqlen, head_dim] + noise_pred[:, :, c, :, :] += self.transformer( + partial_latent_model_input, + t_expand, + text_states=input_prompt_embeds, + text_mask=input_prompt_mask, + text_states_2=input_prompt_embeds_2, + freqs_cos=freqs_cos, + freqs_sin=freqs_sin, guidance=guidance_expand, stg_block_idx=stg_block_idx, stg_mode=stg_mode,