Run on cpu anyway, even if it takes 1m years/img

app.py

@@ -12,6 +12,8 @@ from PIL import Image
 import torch
 from diffusers import PixArtAlphaPipeline, DPMSolverMultistepScheduler
 from sa_solver_diffusers import SASolverScheduler
+from transformers import AutoTokenizer
+from transformers import pipeline as pipe
 
 
 DESCRIPTION = """![Logo](https://raw.githubusercontent.com/PixArt-alpha/PixArt-alpha.github.io/master/static/images/logo.png)
@@ -100,26 +102,11 @@ def apply_style(style_name: str, positive: str, negative: str = "") -> Tuple[str
     return p.replace("{prompt}", positive), n + negative
 
 
-if torch.cuda.is_available():
-    pipe = PixArtAlphaPipeline.from_pretrained(
+pipe = PixArtAlphaPipeline.from_pretrained(
         "PixArt-alpha/PixArt-XL-2-1024-MS",
-        torch_dtype=torch.float16,
         use_safetensors=True,
     )
 
-    if ENABLE_CPU_OFFLOAD:
-        pipe.enable_model_cpu_offload()
-    else:
-        pipe.to(device)
-        print("Loaded on Device!")
-
-    # speed-up T5
-    pipe.text_encoder.to_bettertransformer()
-
-    if USE_TORCH_COMPILE:
-        pipe.transformer = torch.compile(pipe.transformer, mode="reduce-overhead", fullgraph=True)
-        print("Model Compiled!")
-
 
 def save_image(img):
     unique_name = str(uuid.uuid4()) + ".png"