Update app.py

app.py

@@ -6,8 +6,9 @@ default_lang = "en"
 
 # ChatGPT
 from pyChatGPT import ChatGPT
-import whisper
-whisper_model = whisper.load_model("small")
+#import whisper
+#whisper_model = whisper.load_model("small")
+whisper = gr.Interface.load(name="spaces/sanchit-gandhi/whisper-large-v2")
 import os
 session_token = os.environ.get('SessionToken') 
 
@@ -45,6 +46,30 @@ def chat_hf(audio, custom_token, language):
     return whisper_text, gpt_response, fp.name
 
 # whisper
+#def translate(audio):
+#    print("""
+#    —
+#    Sending audio to Whisper ...
+#    —
+#    """)
+#    
+#    audio = whisper.load_audio(audio)
+#    audio = whisper.pad_or_trim(audio)
+#    
+#    mel = whisper.log_mel_spectrogram(audio).to(whisper_model.device)
+#    
+#    _, probs = whisper_model.detect_language(mel)
+#    
+#    transcript_options = whisper.DecodingOptions(task="transcribe", fp16 = False)
+#    
+#    transcription = whisper.decode(whisper_model, mel, transcript_options)
+#    
+#    print("language spoken: " + transcription.language)
+#    print("transcript: " + transcription.text)
+#    print("———————————————————————————————————————————")  
+#      
+#    return transcription.text
+
 def translate(audio):
     print("""
     —
@@ -52,22 +77,10 @@ def translate(audio):
     —
     """)
     
-    audio = whisper.load_audio(audio)
-    audio = whisper.pad_or_trim(audio)
-    
-    mel = whisper.log_mel_spectrogram(audio).to(whisper_model.device)
-    
-    _, probs = whisper_model.detect_language(mel)
-    
-    transcript_options = whisper.DecodingOptions(task="transcribe", fp16 = False)
-    
-    transcription = whisper.decode(whisper_model, mel, transcript_options)
-    
-    print("language spoken: " + transcription.language)
-    print("transcript: " + transcription.text)
-    print("———————————————————————————————————————————")  
-      
-    return transcription.text
+    text_result = whisper(audio, None, "transcribe", fn_index=0)
+    print(text_result)
+    return text_result
+
 
 with gr.Blocks() as blocks:
     gr.Markdown("<h1 style='text-align: center; margin-bottom: 1rem'>"
@@ -87,7 +100,7 @@ with gr.Blocks() as blocks:
                 submit = gr.Button("Submit", variant="primary")
         with gr.Column():
             text1 = gr.Textbox(label="Speech to Text")
-            text2 = gr.Textbox(label="chatGPT response")
+            text2 = gr.Textbox(label="ChatGPT response")
             audio = gr.Audio(label="Output", interactive=False)
     #gr.Markdown(info)
     #gr.Markdown("<center>"