Update app.py

app.py

@@ -1,41 +1,30 @@
-from engine.speech2text import Speech2Text
-from engine.ner_engine import NERInfer
-from engine.audio_chunk import AudioChunk
-import glob
-import time
+from transformers import pipeline
 import gradio as gr
-import os
-
-audio_chunk = AudioChunk()
-speech2text = Speech2Text()
-ner_tag = NERInfer()
-
+import time
+p = pipeline(
+            task="automatic-speech-recognition",
+            model="arthoho66/model_005_2000",
+            token="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",
+            # device="cuda:0",
+        )
 text = ""
-def streaming_process(streaming_audio_file) -> str:
-    global text 
-    text = speech2text.process_streaming(streaming_audio_file)
-    # return text
-
-def output_streaming(text_streaming,text01)-> str:
-    text_streaming+=text01
-    return text_streaming
-
 def recorded_process(recorded_audio_file) -> str:
     """ 
     to get both input 
     and use speech2text for get text
     """
-    text = speech2text.process_microphone(recorded_audio_file)
-    result = ner_tag.infer(text)
-    return {"text": text, "entities": result}
+    text = p(recorded_audio_file)["text"]
+    return text
 
 
-def uploaded_process(uploaded_audio_file) -> str:
-    audio_chunk.chunks_audio(uploaded_audio_file)
-    text = speech2text.process_audio()
-    result = ner_tag.infer(text)      
-    return {"text": text, "entities": result}
+def streaming_process(streaming_audio_file) -> str:
+    global text 
+    text = p(streaming_audio_file)["text"]
+    return text
 
+def output_streaming(text_streaming,text01)-> str:
+    text_streaming+=text01
+    return text_streaming
 
 def clear_inputs_and_outputs() -> list:
     """
@@ -44,23 +33,10 @@ def clear_inputs_and_outputs() -> list:
     audio_chunk.remove_chunk()
     return [None, None, None, None]
 
+
 text_streaming = ""
 
 with gr.Blocks() as demo:
-    """
-    buld gradio app
-
-    """
-    gr.Markdown(
-        """
-        ![logo](/file=./image/clicknext_logo2x.png)
-
-        # Automatic Speech Recognition
-         
-        #####  Experience real-time, accurate, and multilingual speech-to-text conversion with our cutting-edge ASR technology.
-        """
-    )
-
     with gr.Tab("Record File"):
         with gr.Row():  
             with gr.Column():
@@ -69,7 +45,7 @@ with gr.Blocks() as demo:
                     clr_btn = gr.Button(value="Clear", variant="secondary")
                     sub_btn = gr.Button(value="submit")
             with gr.Column():
-                lbl_output = gr.HighlightedText(label="Result")
+                lbl_output = gr.Textbox(label="Result")
 
             clr_btn.click(
                 fn=clear_inputs_and_outputs,
@@ -100,34 +76,4 @@ with gr.Blocks() as demo:
                 text_streaming = output_streaming(text_streaming,text)
                 gr.Textbox(value=text, label="Result", autofocus=True)
 
-    # with gr.Tab("Upload File"):
-    #     with gr.Row():  
-    #         with gr.Column():
-    #             upl_input = gr.Audio( type="filepath", label="Upload a file")   
-    #             with gr.Row():
-    #                     clr_btn = gr.Button(value="Clear", variant="secondary")
-    #                     sub_btn = gr.Button(value="submit")
-    #             gr.Examples(examples=[
-    #                 os.path.join(os.path.dirname(__file__),"examples/politics.mp3"),
-    #                 os.path.join(os.path.dirname(__file__),"examples/law1.mp3"),
-    #                 os.path.join(os.path.dirname(__file__),"examples/law2.mp3"),
-    #                 os.path.join(os.path.dirname(__file__),"examples/law3.mp3"),
-    #                 os.path.join(os.path.dirname(__file__),"examples/economy.mp3"),
-    #                 os.path.join(os.path.dirname(__file__),"examples/general.mp3")
-    #                 ],
-    #                 inputs = upl_input)
-    #         with gr.Column():
-    #             lbl_output = gr.HighlightedText(label="Result")
-
-    #         clr_btn.click(
-    #             fn=clear_inputs_and_outputs,
-    #             inputs=[],
-    #             outputs=[upl_input, lbl_output]
-    #         )
-    #         sub_btn.click(
-    #             fn=uploaded_process,
-    #             inputs=[upl_input],
-    #             outputs=[lbl_output]
-    #         )
-
-demo.launch(server_name="0.0.0.0", server_port=8085)
+demo.launch(server_name="0.0.0.0", server_port=8085)