Create app.py

app.py

@@ -0,0 +1,28 @@
+from transformers import pipeline
+
+import gradio as gr
+
+asr = pipeline("automatic-speech-recognition", "facebook/wav2vec2-base-960h")
+classifier = pipeline("text-classification")
+
+def speech_to_text(speech):
+    text = asr(speech)["text"]
+    return text
+
+def text_to_sentiment(text):
+    return classifier(text)[0]["label"]
+
+demo = gr.Blocks()
+
+with demo:
+    audio_file = gr.Audio(type="filepath")
+    text = gr.Textbox()
+    label = gr.Label()
+
+    b1 = gr.Button("Recognize Speech")
+    b2 = gr.Button("Classify Sentiment")
+
+    b1.click(speech_to_text, inputs=audio_file, outputs=text)
+    b2.click(text_to_sentiment, inputs=text, outputs=label)
+
+demo.launch()