AI_for_biomedical_students

Designing a chatbot in Sarvam

• Create an account in Sarvam

• get API key

• open Google Colab

• save your API key in secrets tab in Google Colab

securely in Google Colab, you should use Colab’s Secrets feature. This allows you to store sensitive information like API keys or tokens without embedding them directly in your code, which is important for security and when sharing notebooks.

1. Open the Secrets Panel: In the left sidebar of your Colab notebook, click on the “🔑 Secrets” icon (it looks like a key).
2. Add a new secret: Click the ”+ Add new secret” button.
3. Name the secret: For your Sarvam API key, you could name it YOUR_AUTHTOKEN (or any other descriptive name you prefer, but remember it for later).
4. Enter the secret value: Paste your Sarvam API key into the “Value” field.
5. Save: Click “Save secret”.

Once saved, you can access this secret in your Python code using from google.colab import userdata and then token = userdata.get('YOUR_AUTHTOKEN') (replacing 'YOUR_AUTHTOKEN' with the name you gave your secret).

• type the following code in Colab

'''
Sarvam voice model starter kit

Documentation:
- https://docs.sarvam.ai/api-reference-docs/getting-started/quickstart

Installation:

    python -m venv sarvam_venv

    source sarvam_venv/bin/activate

    pip install -r requirements_sarvam.txt

Usage:
    python sarvam_voice_model.py
    
Author: Soumya Banerjee

'''

from sarvamai import SarvamAI
import dotenv
import os
import logging
import wave
import sounddevice as sd
import numpy as np

# Load environment variables from .env file
# CAUTION: add SARVAM_API_KEY to .env file before running this code
# Also please add .env to .gitignore to avoid pushing your API key to github
dotenv.load_dotenv()
client = SarvamAI(
    api_subscription_key=os.getenv("SARVAM_API_KEY")
)

# create wav file from mic 
def record_wav(filename="hello.wav", seconds=5, samplerate=16000):
    '''
    record audio from mic and save file
    '''
    print(f"Recording for {seconds} seconds... speak now")
    audio = sd.rec(
        int(seconds * samplerate),
        samplerate=samplerate,
        channels=1,
        dtype="int16",
    )
    sd.wait()

    with wave.open(filename, "wb") as wf:
        wf.setnchannels(1)
        wf.setsampwidth(2)  # int16 = 2 bytes
        wf.setframerate(samplerate)
        wf.writeframes(audio.tobytes())

    print(f"Saved {filename}")

record_wav(filename="audio.wav",
           seconds=5,
           samplerate=16000)    

str_mode = "translate" # or "transcribe", "verbatim", "translit", "codemix"

response = client.speech_to_text.transcribe(
    file = open("audio.wav", "rb"),
    model = "saaras:v3",
    mode = str_mode
    # mode="transcribe" # or "translate", "verbatim", "translit", "codemix"
)

print("Response from SarvAM API: \n")
print(response)

# TODO: appify using streamlit or gradio or huggingface spaces

This site is open source. Improve this page.