palbha commited on
Commit
176259d
·
verified ·
1 Parent(s): 6a5b9ba

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +12 -11
app.py CHANGED
@@ -1,14 +1,14 @@
1
  import os
2
  import gradio as gr
3
- import google.generativeai as genai
4
  from gtts import gTTS
5
  import tempfile
6
  import time
7
 
8
-
9
  # Configure the Gemini API
10
  GOOGLE_API_KEY = os.getenv("gemini_api") # Replace with your actual API key
11
- genai.configure(api_key=GOOGLE_API_KEY)
 
12
 
13
  # Initialize the model
14
  model = genai.GenerativeModel('gemini-pro')
@@ -21,7 +21,12 @@ def transcribe_audio(audio_path):
21
  """
22
  # For demonstration, we're returning a placeholder message
23
  # In a real app, you would connect to a speech-to-text service
24
- return "This is a placeholder for speech-to-text transcription. In a real application, this would be the transcribed text from your audio."
 
 
 
 
 
25
 
26
  def text_to_speech(text):
27
  """Convert text to speech using gTTS and return the path to the audio file"""
@@ -37,13 +42,9 @@ def chat_with_gemini(user_input, history):
37
  # Initialize conversation or continue existing one
38
  if not history:
39
  history = []
40
- chat = model.start_chat(history=[])
41
- else:
42
- # Reconstruct the chat session with history
43
- chat = model.start_chat(history=[
44
- {"role": "user" if i % 2 == 0 else "model", "parts": [msg]}
45
- for i, msg in enumerate(history)
46
- ])
47
 
48
  # Generate response
49
  response = chat.send_message(user_input)
 
1
  import os
2
  import gradio as gr
3
+ from google import genai
4
  from gtts import gTTS
5
  import tempfile
6
  import time
7
 
 
8
  # Configure the Gemini API
9
  GOOGLE_API_KEY = os.getenv("gemini_api") # Replace with your actual API key
10
+
11
+ client = genai.Client(api_key=GOOGLE_API_KEY)
12
 
13
  # Initialize the model
14
  model = genai.GenerativeModel('gemini-pro')
 
21
  """
22
  # For demonstration, we're returning a placeholder message
23
  # In a real app, you would connect to a speech-to-text service
24
+ response = client.models.generate_content(
25
+ model='gemini-2.0-flash',
26
+ contents=['Describe this audio clip', audio_path]
27
+ )
28
+
29
+ return response.text
30
 
31
  def text_to_speech(text):
32
  """Convert text to speech using gTTS and return the path to the audio file"""
 
42
  # Initialize conversation or continue existing one
43
  if not history:
44
  history = []
45
+
46
+ chat = client.chats.create(model="gemini-2.0-flash")
47
+
 
 
 
 
48
 
49
  # Generate response
50
  response = chat.send_message(user_input)