ver1.0

app.py

@@ -0,0 +1,39 @@
+import gradio as gr
+from pytube import YouTube
+import whisper
+
+#define transcription function
+def whisper_transcript(model_size, url, audio_file):
+    if url:
+        link = YouTube(url)
+        source = link.streams.filter(only_audio=True)[0].download(filename="audio.mp4")
+
+    else:
+        source = audio_file
+
+    options = whisper.DecodingOptions(without_timestamps=True)
+
+    loaded_model = whisper.load_model(model_size)
+    transcript = loaded_model.transcribe(source)
+
+    return transcript["text"]
+
+#DEFINE GRADIO INTERFACE
+gradio_ui = gr.Interface(
+    fn=whisper_transcript,
+    title="Transcribe multi-lingual audio clips with Whisper",
+    description= "**How to use**: Select a model, paste in a Youtube link or upload an audio clip, then click submit.",
+    article="**Note**: The larger the model size selected or the longer the audio clip, the more time it would take to process the transcript.",
+    inputs=[
+        gr.Dropdown(
+            label="Select Model",
+            choices=["base", "small", "medium", "large"],
+            value="base",
+        ),
+        gr.Textbox(label="Paste YouTube link here"),
+        gr.Audio(label="Upload Audio File", source="upload", type="filepath"),
+    ],
+    outputs=gr.outputs.Textbox(label="Whisper Transcript"),
+)
+
+gradio_ui.queue().launch()