Update app.py

app.py

@@ -7,6 +7,7 @@ from glob import glob
 from pathlib import Path
 from typing import Optional
 
+import base64
 import tempfile
 import numpy as np
 import cv2
@@ -188,6 +189,7 @@ model_select("AnimateLCM-SVD-xt-1.1.safetensors")
 max_64_bit_int = 2**63 - 1
 
 def sample(
+    secret_token: str,
     image: Image,
     seed: Optional[int] = 42,
     randomize_seed: bool = False,
@@ -195,12 +197,16 @@ def sample(
     fps_id: int = 8,
     max_guidance_scale: float = 1.2,
     min_guidance_scale: float = 1,
-    width: int = 1024,
-    height: int = 576,
+    width: int = 768,
+    height: int = 384,
     num_inference_steps: int = 4,
     decoding_t: int = 4,  # Number of frames decoded at a time! This eats most VRAM. Reduce if necessary.
     output_folder: str = "outputs_gradio",
 ):
+    if secret_token != SECRET_TOKEN:
+        raise gr.Error(
+            f'Invalid secret token. Please fork the original space if you want to use it for yourself.')
+
     if image.mode == "RGBA":
         image = image.convert("RGB")
 
@@ -227,15 +233,19 @@ def sample(
     export_to_video(frames, video_path, fps=fps_id)
     torch.manual_seed(seed)
 
-    return video_path, seed
+    # Read the content of the video file and encode it to base64
+    with open(video_path, "rb") as video_file:
+        video_base64 = base64.b64encode(video_file.read()).decode('utf-8')
+
+    return video_base64
 
 
 with gr.Blocks() as demo:
     with gr.Row():
-        with gr.Column():
-            image = gr.Image(label="Upload your image", type="pil")
-            generate_btn = gr.Button("Generate")
-        video = gr.Video()
+
+        image_input_base64 = gr.Textbox()
+        generate_btn = gr.Button("Generate")
+        video_output_base64 = gr.Textbox()
 
         seed = gr.Slider(
             label="Seed",
@@ -266,7 +276,7 @@ with gr.Blocks() as demo:
         width = gr.Slider(
             label="Width of input image",
             info="It should be divisible by 64",
-            value=576, # 256, 320, 384, 448
+            value=768, # 576, # 256, 320, 384, 448
             minimum=256,
             maximum=2048,
             step=64,
@@ -274,7 +284,7 @@ with gr.Blocks() as demo:
         height = gr.Slider(
             label="Height of input image",
             info="It should be divisible by 64",
-            value=320, # 256, 320, 384, 448
+            value=384, # 320, # 256, 320, 384, 448
             minimum=256,
             maximum=1152,
         )
@@ -304,7 +314,8 @@ with gr.Blocks() as demo:
     generate_btn.click(
         fn=sample,
         inputs=[
-            image,
+            secret_token,
+            image_input_base64,
             seed,
             randomize_seed,
             motion_bucket_id,
@@ -315,8 +326,7 @@ with gr.Blocks() as demo:
             height,
             num_inference_steps,
         ],
-        outputs=[video, seed],
-        api_name="video",
+        outputs=video_output_base64,
     )
 
 if __name__ == "__main__":