updating requirements and imports

requirements.txt

@@ -1,7 +1,7 @@
 transformers
 joblib
 accelerate
-diffusers==0.12.1
+diffusers
 salesforce-lavis
 openai
-#git+https://github.com/pix2pixzero/pix2pix-zero.git
+#git+https://github.com/pix2pixzero/pix2pix-zero.git

submodules/pix2pix-zero/src/utils/base_pipeline.py

@@ -8,7 +8,8 @@ from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
 from diffusers import DiffusionPipeline
 from diffusers.models import AutoencoderKL, UNet2DConditionModel
 from diffusers.schedulers import KarrasDiffusionSchedulers
-from diffusers.utils import deprecate, is_accelerate_available, logging, randn_tensor, replace_example_docstring
+from diffusers.utils import deprecate, is_accelerate_available, logging, replace_example_docstring
+from diffusers.utils.torch_utils import randn_tensor
 from diffusers import StableDiffusionPipeline
 from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
 

submodules/pix2pix-zero/src/utils/cross_attention.py

@@ -1,5 +1,5 @@
 import torch
-from diffusers.models.attention import CrossAttention
+from diffusers.models.attention import Attention as CrossAttention
 
 class MyCrossAttnProcessor:
     def __call__(self, attn: CrossAttention, hidden_states, encoder_hidden_states=None, attention_mask=None):

submodules/pix2pix-zero/src/utils/scheduler.py

@@ -23,7 +23,8 @@ import numpy as np
 import torch
 
 from diffusers.configuration_utils import ConfigMixin, register_to_config
-from diffusers.utils import BaseOutput, randn_tensor
+from diffusers.utils import BaseOutput
+from diffusers.utils.torch_utils import randn_tensor
 from diffusers.schedulers.scheduling_utils import KarrasDiffusionSchedulers, SchedulerMixin
 
 

utils/direction_utils.py

@@ -1,7 +1,7 @@
 import os, sys, pdb
 
 import torch, torchvision
-from huggingface_hub import hf_hub_url, cached_download, hf_hub_download, HfApi
+from huggingface_hub import hf_hub_url, hf_hub_download, HfApi
 import joblib
 from pathlib import Path
 

utils/generate_synthetic.py

@@ -313,4 +313,4 @@ def launch_main(img_in_real, img_in_synth, src, src_custom, dest, dest_custom, n
 
 
 if __name__=="__main__":
-    print(flant5xl_compute_word2sentences("cat wearing sunglasses", num=100))
+    print(flant5xl_compute_word2sentences("cat wearing sunglasses", num=100))