Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,25 +1,20 @@
|
|
1 |
import gradio as gr
|
2 |
-
|
3 |
-
import
|
4 |
from huggingface_hub import InferenceClient
|
5 |
|
6 |
-
# Initialize
|
7 |
-
asr_model = pipeline("automatic-speech-recognition", model="facebook/mms-1b-all")
|
8 |
-
|
9 |
-
# Initialize Facebook MMS TTS model
|
10 |
-
tts_model = pipeline("text-to-speech", model="facebook/mms-tts")
|
11 |
-
|
12 |
-
# Initialize the Chat Model (Gemma-2-9B or Futuresony.gguf)
|
13 |
chat_client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf") # Change if needed
|
14 |
|
15 |
def asr_chat_tts(audio):
|
16 |
"""
|
17 |
-
1. Convert Speech to Text
|
18 |
-
2. Process text through Chat Model (
|
19 |
-
3. Convert response to Speech
|
20 |
"""
|
21 |
-
# Step 1:
|
22 |
-
|
|
|
23 |
|
24 |
# Step 2: Process text through the chat model
|
25 |
messages = [{"role": "system", "content": "You are a helpful AI assistant."}]
|
@@ -30,12 +25,11 @@ def asr_chat_tts(audio):
|
|
30 |
token = msg.choices[0].delta.content
|
31 |
response += token
|
32 |
|
33 |
-
# Step 3:
|
34 |
-
|
35 |
-
|
36 |
-
sf.write(output_file, speech["audio"], samplerate=speech["sampling_rate"])
|
37 |
|
38 |
-
return transcription, response,
|
39 |
|
40 |
# Gradio Interface
|
41 |
with gr.Blocks() as demo:
|
@@ -54,4 +48,3 @@ with gr.Blocks() as demo:
|
|
54 |
# Run the App
|
55 |
if __name__ == "__main__":
|
56 |
demo.launch()
|
57 |
-
|
|
|
1 |
import gradio as gr
|
2 |
+
import subprocess
|
3 |
+
import os
|
4 |
from huggingface_hub import InferenceClient
|
5 |
|
6 |
+
# Initialize Chatbot Model (Futuresony.gguf)
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
chat_client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf") # Change if needed
|
8 |
|
9 |
def asr_chat_tts(audio):
|
10 |
"""
|
11 |
+
1. Convert Speech to Text using asr.py
|
12 |
+
2. Process text through Chat Model (Futuresony.gguf)
|
13 |
+
3. Convert response to Speech using tts.py
|
14 |
"""
|
15 |
+
# Step 1: Run ASR (Speech-to-Text)
|
16 |
+
asr_output = subprocess.run(["python3", "asr.py", audio], capture_output=True, text=True)
|
17 |
+
transcription = asr_output.stdout.strip()
|
18 |
|
19 |
# Step 2: Process text through the chat model
|
20 |
messages = [{"role": "system", "content": "You are a helpful AI assistant."}]
|
|
|
25 |
token = msg.choices[0].delta.content
|
26 |
response += token
|
27 |
|
28 |
+
# Step 3: Run TTS (Text-to-Speech)
|
29 |
+
tts_output_file = "generated_speech.wav"
|
30 |
+
subprocess.run(["python3", "tts.py", response, tts_output_file])
|
|
|
31 |
|
32 |
+
return transcription, response, tts_output_file
|
33 |
|
34 |
# Gradio Interface
|
35 |
with gr.Blocks() as demo:
|
|
|
48 |
# Run the App
|
49 |
if __name__ == "__main__":
|
50 |
demo.launch()
|
|