added readme

app.py

@@ -162,7 +162,19 @@ def analyze_accent(url_or_file):
 
 
 with gr.Blocks() as demo:
-    gr.Markdown("# English Accent Classifier")
+
+    gr.Markdown("""
+        # English Accent Classifier!
+
+        ### How it works?
+        - Takes video URL or video file  
+        - Converts it into audio  
+        - Uses `Whisper-tiny` to detect which language is being spoken  
+        - If the detected language is English, it uses SpeechBrain's Accent ID classifier to show the speaker's accent along with a confidence score.
+
+        **Q: What if my transformers version doesn't expose `return_language` for `whisper-tiny`?**  
+        A: Then it will approximate the language by counting which language's tokens it is using the most.
+        """)
 
     with gr.Tab("From URL"):
         url_input = gr.Textbox(label="Video URL (MP4)")
@@ -194,9 +206,10 @@ with gr.Blocks() as demo:
         )
 
 
+
 demo.css = """
 .output-box {
-    min-height: 100px;
+    min-height: 70px;
     overflow-y: auto;
     padding: 10px;
 }