Add desriptive output

app.py

@@ -1,5 +1,6 @@
 from transformers import pipeline
 import gradio as gr
+import time
 
 pipe_fine = pipeline(model="zeihers-mart/whisper-small-swedish-basic", device_map="auto") 
 pipe_raw = pipeline(model="openai/whisper-small", device_map="auto")
@@ -19,24 +20,32 @@ pipe_raw.model.config.forced_decoder_ids = (
 )
 
 def transcribe(audio):
+    start = time.time()
     text_sv = pipe_fine(audio)["text"]
-    print(f"Audio transcribed: {text_sv}")
+    time_fine = time.time() - start
+    print(f"Fine-tuned: audio transcribed in {time_fine} seconds: {text_sv}")
+
+    start = time.time()
     text_raw= pipe_raw(audio)["text"]
-    print(f"Text translated: {text_raw}")
+    time_raw = time.time() - start
+    print(f"Raw: audio transcribed in {time_raw} seconds: {text_raw}")
     sentiment= sa(text_sv)
     print(f"Sentiment result: {sentiment}")
     sentiment= sentiment[0]["label"]
     path = "https://upload.wikimedia.org/wikipedia/commons/thumb/e/e0/SNice.svg/1200px-SNice.svg.png"
     if sentiment == "NEGATIVE":
-        path = "https://upload.wikimedia.org/wikipedia/commons/thumb/4/42/Sad_smiley_yellow_simple.svg/2048px-Sad_smiley_yellow_simple.svg.png"
-    return text_sv, text_raw, path
+        path = "https://upload.wikimedia.org/wikipedia/commons/thumb/4/42/Sad_smiley_yellow_simple.svg/240px-Sad_smiley_yellow_simple.svg.png"
+    
+    description = f"The fine-tuned model took {time_fine} seconds while the original Whisper model took {raw_time} seconds.\nThe sentiment was evaluated form the fine-tuned model transcription as {lower(sentiment)}."
+    return text_sv, text_raw, path, description
 
 iface = gr.Interface(
     fn=transcribe, 
     inputs=gr.Audio(sources=["microphone"], type="filepath"), 
     outputs=[gr.Textbox(label="Fine-tuned transcription"),
              gr.Textbox(label="Whisper transcription"),
-             gr.Image(label="Sentiment from Fine-tuned transcription", width=100, height=100)],
+             gr.Image(label="Sentiment from Fine-tuned transcription", width=250, height=250),
+             gr.Textbox(label="Description")],
     title="Finetuned Whisper Swedish Small",
     description="Realtime demo for Swedish speech recognition using a fine-tuned Whisper small model.",
 )