Update app.py

app.py

@@ -4,12 +4,22 @@ from transformers import AutoTokenizer
 import torch
 import os
 from transformers import pipeline
+
+token  = AutoTokenizer.from_pretrained('distilroberta-base')
+
+inf_session = onnxruntime.InferenceSession('classifier1-quantized.onnx')
+input_name = inf_session.get_inputs()[0].name
+output_name = inf_session.get_outputs()[0].name
+
+classes = ['Art', 'Astrology', 'Biology', 'Chemistry', 'Economics', 'History', 'Literature', 'Philosophy', 'Physics', 'Politics', 'Psychology', 'Sociology']
+
 ### --- Audio/Video to txt ---###
 device = "cuda:0" if torch.cuda.is_available() else "cpu"
 pipe = pipeline("automatic-speech-recognition", 
                     model="openai/whisper-base.en",
                     chunk_length_s=30, device=device)
 
+### --- Text Summary --- ###
 summarizer = pipeline("summarization", model="facebook/bart-large-cnn", device=device)
      
 
@@ -40,18 +50,6 @@ def summary(text):
     summ = summarizer(chunks,max_length = 100)
 
     return summ
-    
-    
-
-
-
-token  = AutoTokenizer.from_pretrained('distilroberta-base')
-
-inf_session = onnxruntime.InferenceSession('classifier1-quantized.onnx')
-input_name = inf_session.get_inputs()[0].name
-output_name = inf_session.get_outputs()[0].name
-
-classes = ['Art', 'Astrology', 'Biology', 'Chemistry', 'Economics', 'History', 'Literature', 'Philosophy', 'Physics', 'Politics', 'Psychology', 'Sociology']
 
 def classify(vid):
     full_text = video_identity(vid)
@@ -64,9 +62,9 @@ def classify(vid):
     probs = torch.sigmoid(logits)[0]
     return full_text, sum, dict(zip(classes,map(float,probs)))
 
-# label = gr.outputs.Label(num_top_classes=5)
+    
 iface = gr.Interface(fn=classify,
-                     inputs=gr.inputs.Audio(source="upload", type="filepath"),
+                     inputs=gr.inputs.Audio(source="upload", type="filepath",label="Model Card"),
                      outputs = ['text','text',gr.outputs.Label(num_top_classes=3)])
 iface.launch(inline=False)