change

app.py

@@ -1,10 +1,7 @@
-from cProfile import label
-from email.policy import default
-
-from altair import value
 import gradio as gr
 import os
 from lang_list import TEXT_SOURCE_LANGUAGE_NAMES
+from gradio_client import Client
 
 HF_API = os.getenv("HF_API")
 API_URL = os.getenv("API_URL")  # path to Seamlessm4t API endpoint
@@ -14,7 +11,6 @@ DEFAULT_TARGET_LANGUAGE = "Western Persian"
 DESCRIPTION = """
 # Seamlessm4t + Speaker Diarization + Voice Activity Detection
 Here we use seamlessm4t to generate captions for full audios. Audio can be of arbitrary length. 
-
 """
 
 DUPLICATE = """
@@ -28,6 +24,42 @@ To duplicate this repo, you have to give permission from three reopsitories and
 
 """
 
+pipeline = Pipeline.from_pretrained("pyannote/speaker-diarization",use_auth_token=HF_API)
+def predict(target_language, number_of_speakers, final_audio):
+    if number_of_speakers == 0:
+        diarization = pipeline(final_audio)
+    else:
+        diarization = pipeline(final_audio, num_speakers=number_of_speakers)
+
+    for turn, value, speaker in diarization.itertracks(yield_label=True):
+        print(f"start={turn.start}s stop={turn.end}s speaker_{speaker}")
+
+    song = AudioSegment.from_wav(sample_file)
+
+    client = Client(API_URL)
+    output_text = ""
+    for turn, value, speaker in diarization.itertracks(yield_label=True):
+        print(turn)
+        try:
+            clipped = song[turn.start * 1000: turn.end * 1000]
+            clipped.export(f'my.wav', format='wav', bitrate=16000)
+
+            _, result = client.predict(
+                "ASR (Automatic Speech Recognition)",
+                "file",  # str in 'Audio source' Radio component
+                f"my.wav",
+                f"my.wav",
+                "text",
+                target_language,
+                target_language,
+                api_name="/run"
+            )
+
+            output_text = output_text + "\n" + (f'start: {turn.start:.1f} end: {turn.end:.1f} text: {result} speaker: {speaker}')
+
+
+        except Exception as e:
+            print(e)
 
 def update_audio_ui(audio_source: str) -> tuple[dict, dict]:
     mic = audio_source == "microphone"
@@ -40,13 +72,14 @@ def update_audio_ui(audio_source: str) -> tuple[dict, dict]:
 with gr.Blocks(css="style.css") as demo:
     gr.Markdown(DESCRIPTION)
     with gr.Group():
-        target_language = gr.Dropdown(
-            choices=TEXT_SOURCE_LANGUAGE_NAMES,
-            label="Output Language",
-            value=DEFAULT_TARGET_LANGUAGE,
-            interactive=True,
-        )
-        target_language.update(value=DEFAULT_TARGET_LANGUAGE)
+        with gr.Row():
+            target_language = gr.Dropdown(
+                choices=TEXT_SOURCE_LANGUAGE_NAMES,
+                label="Output Language",
+                value=DEFAULT_TARGET_LANGUAGE,
+                interactive=True,
+            )
+            number_of_speakers=gr.Number(label="Number of Speakers",info="Keep it zero, if you want the model to automatically detect the number of speakers")
         with gr.Row() as audio_box:
             audio_source = gr.Radio(
                 choices=["file", "microphone"], value="file", interactive=True
@@ -63,7 +96,7 @@ with gr.Blocks(css="style.css") as demo:
                 source="upload",
                 visible=True,
             )
-            output = gr.Audio(label="Output", visible=False)
+            final_audio = gr.Audio(label="Output", visible=False)
         audio_source.change(
             fn=update_audio_ui,
             inputs=audio_source,
@@ -71,12 +104,13 @@ with gr.Blocks(css="style.css") as demo:
             queue=False,
             api_name=False,
         )
-        input_audio_mic.change(lambda x: x, input_audio_mic, output)
-        input_audio_file.change(lambda x: x, input_audio_file, output)
+        input_audio_mic.change(lambda x: x, input_audio_mic, final_audio)
+        input_audio_file.change(lambda x: x, input_audio_file, final_audio)
         submit = gr.Button("Submit")
         text_output = gr.Textbox(label="Transcribed Text", value="", interactive=False)
 
-    gr.Markdown(DUPLICATE)
+        submit.click(fn=predict, inputs=[target_language,number_of_speakers, final_audio], outputs=[text_output],api_name="predict")
 
+    gr.Markdown(DUPLICATE)
 
 demo.queue(max_size=50).launch()