Update app.py

app.py

@@ -152,6 +152,17 @@ def generate_video(text: str, video_path: str,
         time.sleep(0.01)
         yield buffer
 
+# Define examples for image and video inference
+image_examples = [
+    ["OCR the Text in the Image", "rolm/1.jpeg"],
+    ["Describe the objects in the image", "rolm/2.jpeg"]
+]
+
+video_examples = [
+    ["Explain the Ad in Detail", "examples/videoplayback.mp4"],
+    ["Identify the main actions in the video", "examples/demo_video.mp4"]
+]
+
 # Create the Gradio Interface
 with gr.Blocks() as demo:
     gr.Markdown("# **Cosmos-Reason1 by [NVIDIA](https://huggingface.co/nvidia/Cosmos-Reason1-7B)**")
@@ -161,10 +172,18 @@ with gr.Blocks() as demo:
                 with gr.TabItem("Image Inference"):
                     image_query = gr.Textbox(label="Query Input", placeholder="Enter your query here...")
                     image_upload = gr.Image(type="pil", label="Upload Image")
+                    gr.Examples(
+                        examples=image_examples,
+                        inputs=[image_query, image_upload]
+                    )
                     image_submit = gr.Button("Submit")
                 with gr.TabItem("Video Inference"):
                     video_query = gr.Textbox(label="Query Input", placeholder="Enter your query here...")
                     video_upload = gr.Video(label="Upload Video")
+                    gr.Examples(
+                        examples=video_examples,
+                        inputs=[video_query, video_upload]
+                    )
                     video_submit = gr.Button("Submit")
             with gr.Accordion("Advanced options", open=False):
                 max_new_tokens = gr.Slider(label="Max new tokens", minimum=1, maximum=MAX_MAX_NEW_TOKENS, step=1, value=DEFAULT_MAX_NEW_TOKENS)