Update diffusers version

README.md

@@ -187,7 +187,7 @@ python test_stage_2.py --config ./configs/test_stage_2.yaml -W 512 -H 512
 ```
 It will generate the video at 512 x 512 first, and then resize it back to the original size of the pose video.
 
-Currently, it takes 16GB VRAM to run on 512 x 512 x 48 and takes 28GB VRAM to run on 768 x 768 x 48. However, it should be noticed that the inference resolution would affect the final results.
+Currently, it takes 16GB VRAM to run on 512 x 512 x 48 and takes 28GB VRAM to run on 768 x 768 x 48. However, it should be noticed that the inference resolution would affect the final results (especially face region).
 
 #### Face Enhancement
 

musepose/models/transformer_2d.py

@@ -4,7 +4,6 @@ from typing import Any, Dict, Optional
 
 import torch
 from diffusers.configuration_utils import ConfigMixin, register_to_config
-from diffusers.models.embeddings import CaptionProjection
 from diffusers.models.lora import LoRACompatibleConv, LoRACompatibleLinear
 from diffusers.models.modeling_utils import ModelMixin
 from diffusers.models.normalization import AdaLayerNormSingle

musepose/models/unet_2d_condition.py

@@ -20,7 +20,6 @@ from diffusers.models.embeddings import (
     ImageHintTimeEmbedding,
     ImageProjection,
     ImageTimeEmbedding,
-    PositionNet,
     TextImageProjection,
     TextImageTimeEmbedding,
     TextTimeEmbedding,

requirements.txt

@@ -7,7 +7,7 @@ accelerate==0.29.3
 av==11.0.0
 clip @ https://github.com/openai/CLIP/archive/d50d76daa670286dd6cacf3bcd80b5e4823fc8e1.zip#sha256=b5842c25da441d6c581b53a5c60e0c2127ebafe0f746f8e15561a006c6c3be6a
 decord==0.6.0
-diffusers==0.24.0
+diffusers>=0.24.0,<=0.27.2
 einops==0.4.1
 imageio==2.33.0
 imageio-ffmpeg==0.4.9
@@ -19,7 +19,6 @@ opencv-python==4.8.1.78
 scikit-image==0.21.0
 scikit-learn==1.3.2
 transformers==4.33.1
-mlflow==2.9.2
 xformers==0.0.22
 moviepy==1.0.3
 wget==3.2