diffusers 0.33.1__py3-none-any.whl → 0.35.0__py3-none-any.whl

diffusers/pipelines/hidream_image/pipeline_output.py

@@ -0,0 +1,35 @@
+# Copyright 2025 HiDream-ai Team and The HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from dataclasses import dataclass
+from typing import List, Union
+
+import numpy as np
+import PIL.Image
+
+from ...utils import BaseOutput
+
+
+@dataclass
+class HiDreamImagePipelineOutput(BaseOutput):
+    """
+    Output class for HiDreamImage pipelines.
+
+    Args:
+        images (`List[PIL.Image.Image]` or `np.ndarray`)
+            List of denoised PIL images of length `batch_size` or numpy array of shape `(batch_size, height, width,
+            num_channels)`. PIL images or numpy array present the denoised images of the diffusion pipeline.
+    """
+
+    images: Union[List[PIL.Image.Image], np.ndarray]

diffusers/pipelines/hunyuan_video/__init__.py

@@ -24,6 +24,7 @@ except OptionalDependencyNotAvailable:
 else:
     _import_structure["pipeline_hunyuan_skyreels_image2video"] = ["HunyuanSkyreelsImageToVideoPipeline"]
     _import_structure["pipeline_hunyuan_video"] = ["HunyuanVideoPipeline"]
+    _import_structure["pipeline_hunyuan_video_framepack"] = ["HunyuanVideoFramepackPipeline"]
     _import_structure["pipeline_hunyuan_video_image2video"] = ["HunyuanVideoImageToVideoPipeline"]
 
 if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
@@ -36,6 +37,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
     else:
         from .pipeline_hunyuan_skyreels_image2video import HunyuanSkyreelsImageToVideoPipeline
         from .pipeline_hunyuan_video import HunyuanVideoPipeline
+        from .pipeline_hunyuan_video_framepack import HunyuanVideoFramepackPipeline
         from .pipeline_hunyuan_video_image2video import HunyuanVideoImageToVideoPipeline
 
 else:

diffusers/pipelines/hunyuan_video/pipeline_hunyuan_skyreels_image2video.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HunyuanVideo Team and The HuggingFace Team. All rights reserved.
+# Copyright 2025 The HunyuanVideo Team and The HuggingFace Team. All rights reserved.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -575,13 +575,13 @@ class HunyuanSkyreelsImageToVideoPipeline(DiffusionPipeline, HunyuanVideoLoraLoa
             true_cfg_scale (`float`, *optional*, defaults to 1.0):
                 When > 1.0 and a provided `negative_prompt`, enables true classifier-free guidance.
             guidance_scale (`float`, defaults to `6.0`):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality. Note that the only available HunyuanVideo model is
-                CFG-distilled, which means that traditional guidance between unconditional and conditional latent is
-                not applied.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). Guidance scale is enabled by setting
+                `guidance_scale > 1`. Higher guidance scale encourages to generate images that are closely linked to
+                the text `prompt`, usually at the expense of lower image quality. Note that the only available
+                HunyuanVideo model is CFG-distilled, which means that traditional guidance between unconditional and
+                conditional latent is not applied.
             num_videos_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):

diffusers/pipelines/hunyuan_video/pipeline_hunyuan_video.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HunyuanVideo Team and The HuggingFace Team. All rights reserved.
+# Copyright 2025 The HunyuanVideo Team and The HuggingFace Team. All rights reserved.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -529,15 +529,14 @@ class HunyuanVideoPipeline(DiffusionPipeline, HunyuanVideoLoraLoaderMixin):
                 their `set_timesteps` method. If not defined, the default behavior when `num_inference_steps` is passed
                 will be used.
             true_cfg_scale (`float`, *optional*, defaults to 1.0):
-                When > 1.0 and a provided `negative_prompt`, enables true classifier-free guidance.
+                True classifier-free guidance (guidance scale) is enabled when `true_cfg_scale` > 1 and
+                `negative_prompt` is provided.
             guidance_scale (`float`, defaults to `6.0`):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality. Note that the only available HunyuanVideo model is
-                CFG-distilled, which means that traditional guidance between unconditional and conditional latent is
-                not applied.
+                Embedded guiddance scale is enabled by setting `guidance_scale` > 1. Higher `guidance_scale` encourages
+                a model to generate images more aligned with `prompt` at the expense of lower image quality.
+
+                Guidance-distilled models approximates true classifer-free guidance for `guidance_scale` > 1. Refer to
+                the [paper](https://huggingface.co/papers/2210.03142) to learn more.
             num_videos_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
@@ -693,28 +692,30 @@ class HunyuanVideoPipeline(DiffusionPipeline, HunyuanVideoLoraLoaderMixin):
                 # broadcast to batch dimension in a way that's compatible with ONNX/Core ML
                 timestep = t.expand(latents.shape[0]).to(latents.dtype)
 
-                noise_pred = self.transformer(
-                    hidden_states=latent_model_input,
-                    timestep=timestep,
-                    encoder_hidden_states=prompt_embeds,
-                    encoder_attention_mask=prompt_attention_mask,
-                    pooled_projections=pooled_prompt_embeds,
-                    guidance=guidance,
-                    attention_kwargs=attention_kwargs,
-                    return_dict=False,
-                )[0]
-
-                if do_true_cfg:
-                    neg_noise_pred = self.transformer(
+                with self.transformer.cache_context("cond"):
+                    noise_pred = self.transformer(
                         hidden_states=latent_model_input,
                         timestep=timestep,
-                        encoder_hidden_states=negative_prompt_embeds,
-                        encoder_attention_mask=negative_prompt_attention_mask,
-                        pooled_projections=negative_pooled_prompt_embeds,
+                        encoder_hidden_states=prompt_embeds,
+                        encoder_attention_mask=prompt_attention_mask,
+                        pooled_projections=pooled_prompt_embeds,
                         guidance=guidance,
                         attention_kwargs=attention_kwargs,
                         return_dict=False,
                     )[0]
+
+                if do_true_cfg:
+                    with self.transformer.cache_context("uncond"):
+                        neg_noise_pred = self.transformer(
+                            hidden_states=latent_model_input,
+                            timestep=timestep,
+                            encoder_hidden_states=negative_prompt_embeds,
+                            encoder_attention_mask=negative_prompt_attention_mask,
+                            pooled_projections=negative_pooled_prompt_embeds,
+                            guidance=guidance,
+                            attention_kwargs=attention_kwargs,
+                            return_dict=False,
+                        )[0]
                     noise_pred = neg_noise_pred + true_cfg_scale * (noise_pred - neg_noise_pred)
 
                 # compute the previous noisy sample x_t -> x_t-1