download python code lib packages and run it colab, offline running;

 

first mina program run with 

#to download offline

!pip download edge-tts pydub -d /content/drive/MyDrive/Lib/

offline program.

# Mount Google Drive to access your pre-downloaded wheels

from google.colab import drive

drive.mount('/content/drive')

# Install required wheel files (replace filenames as needed)

#!pip install /content/drive/MyDrive/Lib/edge_tts-1.4.0-py3-none-any.whl --quiet

#!pip install /content/drive/MyDrive/Lib/pydub-0.25.1-py2.py3-none-any.whl --quiet

!pip install /content/drive/MyDrive/Lib/*.whl

# Install ffmpeg from local .deb if downloaded OR fallback to internet-based apt

# (Optional) Uncomment below if you have ffmpeg.deb in Drive:

# !dpkg -i /content/drive/MyDrive/Lib/ffmpeg-release-i686-static.deb

# If not available locally, this will still need internet:

!apt install ffmpeg --quiet

# Now continue your original script (unchanged logic)

import os

import asyncio

import edge_tts

from pydub import AudioSegment

# Constants

INPUT_FILE = "/content/input.txt"

VOICE = "te-IN-ShrutiNeural"

CHUNK_SIZE = 4000

TEMP_DIR = "/content/temp_chunks"

FINAL_MP3 = "/content/output.mp3"

def map_speed_to_rate(value):

    mapping = {1: "-75%", 2: "-50%", 3: "0%", 4: "+25%", 5: "+50%", 6: "+75%", 7: "+100%"}

    return mapping.get(value, "0%")

def read_text(file_path):

    try:

        with open(file_path, 'r', encoding='utf-8') as f:

            return " ".join(f.read().split())

    except Exception as e:

        print(f"❌ Error reading file: {e}")

        return ""

def chunk_text(text, chunk_size):

    chunks = []

    while text:

        if len(text) <= chunk_size:

            chunks.append(text)

            break

        split_idx = text.rfind(' ', 0, chunk_size)

        if split_idx == -1:

            split_idx = chunk_size

        chunks.append(text[:split_idx])

        text = text[split_idx:].lstrip()

    return chunks

async def generate_audio(chunks, rate):

    os.makedirs(TEMP_DIR, exist_ok=True)

    segments = []

    for idx, chunk in enumerate(chunks):

        mp3_file = os.path.join(TEMP_DIR, f"chunk_{idx}.mp3")

        print(f"🎙️ Generating Chunk {idx+1}/{len(chunks)} at Speed: {rate}")

        try:

            communicate = edge_tts.Communicate(text=chunk, voice=VOICE, rate=rate)

            await communicate.save(mp3_file)

            segments.append(AudioSegment.from_file(mp3_file))

        except Exception as e:

            print(f"⚠️ Error generating chunk {idx}: {e}")

    return segments

def combine_segments(segments, output_file):

    if not segments:

        print("❌ No audio segments to combine.")

        return

    final_audio = sum(segments[1:], segments[0])

    final_audio.export(output_file, format="mp3")

    print(f"✅ Final audio saved at: {output_file}")

def cleanup_temp():

    for f in os.listdir(TEMP_DIR):

        os.remove(os.path.join(TEMP_DIR, f))

    os.rmdir(TEMP_DIR)

    print("🧹 Temp folder cleaned.")

async def main():

    text = read_text(INPUT_FILE)

    if not text:

        print("❌ No text to process.")

        return

    chunks = chunk_text(text, CHUNK_SIZE)

    speed_input = 5  # 1=slow, 3=normal, 7=fast

    rate = map_speed_to_rate(speed_input)

    segments = await generate_audio(chunks, rate)

    combine_segments(segments, FINAL_MP3)

    cleanup_temp()

await main()

test

can i use above progrlam for hindi an other languages? if api support , any code change required.

2/2

ChatGPT said:

Yes, you can absolutely use the same program for Hindi and other languages using Microsoft Edge TTS — as long as the voice name (language-voice combination) is supported.

---

✅ Only Required Code Change:

Just change the voice name in this line:

python
VOICE = "hi-IN-SwaraNeural"  # Hindi Female voice

---

🔤 Examples of Other Supported Voices:

Language	Voice Name	Gender
Hindi	hi-IN-SwaraNeural	Female
Hindi	hi-IN-MadhurNeural	Male
Tamil	ta-IN-PallaviNeural	Female
Kannada	kn-IN-GaganNeural	Male
Bengali	bn-IN-TanishaaNeural	Female
Telugu	te-IN-ShrutiNeural	Female
Marathi	mr-IN-AarohiNeural	Female
Gujarati	gu-IN-DhwaniNeural	Female
English (India)	en-IN-NeerjaNeural	Female

---

🛠 Code line to modify:

python
VOICE = "hi-IN-SwaraNeural"  # ✅ Change this line only