Upload app.py with huggingface_hub

app.py

@@ -3,7 +3,7 @@ import base64
 from queue import Queue
 import threading
 import torch
-from PIL import Image
+from PIL import Image, ImageOps
 from transformers import AutoProcessor, MiniCPMV4_6ForConditionalGeneration, TextIteratorStreamer
 import gradio as gr
 
@@ -21,6 +21,13 @@ GPT_MAX_COMPLETION_TOKENS = 4096
 GPT_REASONING_EFFORT = "none"
 NOTES_PROMPT = "Transcribe the musical notes in this image. Return only the transcription."
 
+
+def env_flag(name: str, default: bool = False) -> bool:
+    value = os.environ.get(name)
+    if value is None:
+        return default
+    return value.strip().lower() in {"1", "true", "yes", "on"}
+
 print("Loading processor...")
 processor = AutoProcessor.from_pretrained(ORIGINAL_MODEL_ID, trust_remote_code=True)
 
@@ -223,7 +230,7 @@ def predict_all(image_path):
         yield message, message, message
         return
 
-    image = Image.open(image_path).convert("RGB")
+    image = ImageOps.exif_transpose(Image.open(image_path)).convert("RGB")
     updates = Queue()
     outputs = ["", "", ""]
 
@@ -268,11 +275,20 @@ with gr.Blocks(title="Noteworthy — Sheet Music Transcription") as demo:
         """
         # Noteworthy
         Sheet Music Transcription
-        Upload an image of sheet music and click **Transcribe Music** to compare models.
+        Take a photo or upload sheet music, then click **Transcribe Music** to compare models.
         """
     )
 
-    image_input = gr.Image(type="filepath", label="Upload Sheet Music Image")
+    gr.Markdown(
+        "Camera: allow camera access, frame the whole page, then tap the round capture button in the image box."
+    )
+    image_input = gr.Image(
+        type="filepath",
+        label="Sheet Music Image - camera captures are mirrored for correct orientation",
+        sources=["webcam", "upload", "clipboard"],
+        webcam_options=gr.WebcamOptions(mirror=True),
+        placeholder="Use Camera to frame the sheet music, then tap the capture button.",
+    )
     gr.Examples(
         examples=[
             ["examples/000100005-1_1_1.png"],
@@ -303,4 +319,9 @@ with gr.Blocks(title="Noteworthy — Sheet Music Transcription") as demo:
         outputs=[finetuned_output, original_output, gpt_output],
     )
 
-demo.launch(theme=gr.themes.Soft())
+demo.launch(
+    theme=gr.themes.Soft(),
+    server_name=os.environ.get("GRADIO_SERVER_NAME", "0.0.0.0"),
+    server_port=int(os.environ.get("GRADIO_SERVER_PORT", "7860")),
+    share=env_flag("GRADIO_SHARE"),
+)