Chan55 commited on
Commit
061266d
Β·
1 Parent(s): e95d526

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +27 -5
app.py CHANGED
@@ -5,7 +5,7 @@ import time
5
 
6
  # from elevenlabs import set_api_key
7
 
8
-
9
  import whisper
10
 
11
 
@@ -91,7 +91,26 @@ remember to not to include question number.
91
  """)
92
 
93
 
 
 
 
 
 
 
94
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95
 
96
  def convert_tuples_to_list(tuples_list):
97
  result = []
@@ -128,8 +147,11 @@ def run_text_prompt(message, chat_history):
128
  def run_audio_prompt(audio, chat_history):
129
  if audio is None:
130
  return None, chat_history
 
 
131
 
132
- message_transcription = model.transcribe(audio)["text"]
 
133
  _, chat_history = run_text_prompt(message_transcription, chat_history)
134
  return None, chat_history
135
 
@@ -164,9 +186,9 @@ with gr.Blocks(title="hi") as app2:
164
  # fn = run_audio_prompt,
165
  # inputs = [audio, chatbot],
166
  # outputs = [audio, chatbot]
167
-
168
- send_audio_button = gr.Button("Send Audio", interactive=True)
169
- send_audio_button.click(run_audio_prompt, [audio, chatbot], [audio, chatbot])
170
 
171
  demo = gr.TabbedInterface([app2], [ "Interview"])
172
 
 
5
 
6
  # from elevenlabs import set_api_key
7
 
8
+ import requests
9
  import whisper
10
 
11
 
 
91
  """)
92
 
93
 
94
+ from gradio_client import Client
95
+ API_URL = "https://sanchit-gandhi-whisper-jax.hf.space/"
96
+
97
+ # set up the Gradio client
98
+ client = Client(API_URL)
99
+
100
 
101
+ def transcribe_audio(audio_path, task="transcribe", return_timestamps=False):
102
+ """Function to transcribe an audio file using the Whisper JAX endpoint."""
103
+ if task not in ["transcribe", "translate"]:
104
+ raise ValueError("task should be one of 'transcribe' or 'translate'.")
105
+
106
+ text, runtime = client.predict(
107
+ audio_path,
108
+ task,
109
+ return_timestamps,
110
+ api_name="/predict_1",
111
+ )
112
+ return text
113
+
114
 
115
  def convert_tuples_to_list(tuples_list):
116
  result = []
 
147
  def run_audio_prompt(audio, chat_history):
148
  if audio is None:
149
  return None, chat_history
150
+
151
+ message_transcription = transcribe_audio(audio)
152
 
153
+
154
+ # message_transcription = model.transcribe(audio)["text"]
155
  _, chat_history = run_text_prompt(message_transcription, chat_history)
156
  return None, chat_history
157
 
 
186
  # fn = run_audio_prompt,
187
  # inputs = [audio, chatbot],
188
  # outputs = [audio, chatbot]
189
+ audio.change(run_audio_prompt, [audio, chatbot], [audio, chatbot])
190
+ # send_audio_button = gr.Button("Send Audio", interactive=True)
191
+ # send_audio_button.click(run_audio_prompt, [audio, chatbot], [audio, chatbot])
192
 
193
  demo = gr.TabbedInterface([app2], [ "Interview"])
194