Update app.py

app.py

@@ -274,7 +274,7 @@ def generate_30(
     guidance_scale: float = 4,
     num_inference_steps: int = 125,
     randomize_seed: bool = False,
-    latent_file = None,  # Add latents file input
+    latent_file = gr.File(),  # Add latents file input
     progress=gr.Progress(track_tqdm=True)  # Add progress as a keyword argument
 ):
     #torch.backends.cudnn.benchmark = False
@@ -286,12 +286,7 @@ def generate_30(
     seed = int(randomize_seed_fn(seed, randomize_seed))
     generator = torch.Generator(device='cuda').manual_seed(seed)
     if latent_file is not None:  # Check if a latent file is provided
-        if not isinstance(latent_file, Image.Image):
-            image = Image.fromarray(latent_file)
-        image = image.resize((width, height))  # Example resize, adjust as needed
-        image = image.convert("RGB")  # Ensure it's a 3-channel image
-        image = np.array(image)
-        sd_image_a = torch.from_numpy(image).permute(2, 0, 1) / 127.5 - 1.0  # Normalize
+        sd_image_a = Image.open(latent_file.name)
         timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
         filename= f'rv_IP_{timestamp}.txt'
         print("-- using image file --")
@@ -332,7 +327,7 @@ def generate_60(
     guidance_scale: float = 4,
     num_inference_steps: int = 250,
     randomize_seed: bool = False,
-    latent_file = None,  # Add latents file input
+    latent_file = gr.File(),  # Add latents file input
     progress=gr.Progress(track_tqdm=True)  # Add progress as a keyword argument
 ):
     #torch.backends.cudnn.benchmark = True
@@ -344,12 +339,7 @@ def generate_60(
     seed = int(randomize_seed_fn(seed, randomize_seed))
     generator = torch.Generator(device='cuda').manual_seed(seed)
     if latent_file is not None:  # Check if a latent file is provided
-        if not isinstance(latent_file, Image.Image):
-            image = Image.fromarray(latent_file)
-        image = image.resize((width, height))  # Example resize, adjust as needed
-        image = image.convert("RGB")  # Ensure it's a 3-channel image
-        image = np.array(image)
-        sd_image_a = torch.from_numpy(image).permute(2, 0, 1) / 127.5 - 1.0  # Normalize
+        sd_image_a = Image.open(latent_file.name)
         timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
         filename= f'rv_IP_{timestamp}.txt'
         print("-- using image file --")
@@ -390,7 +380,7 @@ def generate_90(
     guidance_scale: float = 4,
     num_inference_steps: int = 250,
     randomize_seed: bool = False,
-    latent_file = None,  # Add latents file input
+    latent_file = gr.File(),  # Add latents file input
     progress=gr.Progress(track_tqdm=True)  # Add progress as a keyword argument
 ):
     #torch.backends.cudnn.benchmark = True
@@ -402,12 +392,7 @@ def generate_90(
     seed = int(randomize_seed_fn(seed, randomize_seed))
     generator = torch.Generator(device='cuda').manual_seed(seed)
     if latent_file is not None:  # Check if a latent file is provided
-        if not isinstance(latent_file, Image.Image):
-            image = Image.fromarray(latent_file)
-        image = image.resize((width, height))  # Example resize, adjust as needed
-        image = image.convert("RGB")  # Ensure it's a 3-channel image
-        image = np.array(image)
-        sd_image_a = torch.from_numpy(image).permute(2, 0, 1) / 127.5 - 1.0  # Normalize
+        sd_image_a = Image.open(latent_file.name)
         timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
         filename= f'rv_IP_{timestamp}.txt'
         print("-- using image file --")
@@ -480,7 +465,7 @@ with gr.Blocks(theme=gr.themes.Origin(),css=css) as demo:
     result = gr.Gallery(label="Result", columns=1, show_label=False) 
 
     with gr.Row():
-        latent_file = gr.Image(label="Image Prompt (Required)", interactive=True)
+        latent_file = gr.File(label="Image Prompt (Required)")
         style_selection = gr.Radio(
             show_label=True,
             container=True,