infer.py

import base64
import json
import os
import wave

import faster_whisper
import tempfile

import numpy as np
import torch
import time
import requests
import logging
from fastapi import FastAPI, HTTPException, WebSocket, WebSocketDisconnect
import websockets
from pydantic import BaseModel
from typing import Optional
import sys
import asyncio

from model import segment_to_dict, get_raw_words_from_segments

# Configure logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s %(levelname)s: %(message)s',
                    handlers=[logging.StreamHandler(sys.stdout)], force=True)
logger = logging.getLogger(__name__)
#logging.getLogger("asyncio").setLevel(logging.DEBUG)

logging.info(torch.__version__)
logging.info(torch.version.cuda)  # Should show the installed CUDA version
device = 'cuda' if torch.cuda.is_available() else 'cpu'
logging.info(f'Device selected: {device}')

model_name = 'ivrit-ai/faster-whisper-v2-d4'
logging.info(f'Loading model: {model_name}')
model = faster_whisper.WhisperModel(model_name, device=device)
logging.info('Model loaded successfully')

# Maximum data size: 200MB
MAX_PAYLOAD_SIZE = 200 * 1024 * 1024
logging.info(f'Max payload size set to: {MAX_PAYLOAD_SIZE} bytes')

app = FastAPI()


# Define Pydantic model for input
class TranscribeInput(BaseModel):
    audio: str  # Base64-encoded audio data
    init_prompt: str = ""



# Define WebSocket endpoint
@app.websocket("/ws_transcribe_streaming")
async def websocket_transcribe(websocket: WebSocket):
    logger.info("New WebSocket connection request received.")
    await websocket.accept()
    logger.info("WebSocket connection established successfully.")

    try:
        while True:
            try:
                # Receive JSON data
                data = await websocket.receive_json()
                # Parse input data
                input_data = TranscribeInput(**data)
                # Decode base64 audio data
                audio_bytes = base64.b64decode(input_data.audio)
                # Write audio data to a temporary file
                with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_audio_file:
                    temp_audio_file.write(audio_bytes)
                    temp_audio_file.flush()
                    audio_file_path = temp_audio_file.name

                # Call the transcribe function
                # segments, info = await asyncio.to_thread(model.transcribe,
                segments, info = model.transcribe(
                    audio_file_path,
                    language='he',
                    initial_prompt=input_data.init_prompt,
                    beam_size=5,
                    word_timestamps=True,
                    condition_on_previous_text=True
                )

                # Convert segments to list and serialize
                segments_list = list(segments)
                segments_serializable = [segment_to_dict(s) for s in segments_list]
                logger.info(get_raw_words_from_segments(segments_list))
                # Send the serialized segments back to the client
                await websocket.send_json(segments_serializable)

            except WebSocketDisconnect:
                logger.info("WebSocket connection closed by the client.")
                break
            except Exception as e:
                logger.error(f"Unexpected error during WebSocket transcription: {e}")
                await websocket.send_json({"error": str(e)})
    finally:
        logger.info("Cleaning up and closing WebSocket connection.")

# class InputData(BaseModel):
#     type: str
#     data: Optional[str] = None  # Used for blob input
#     url: Optional[str] = None  # Used for url input
#
#
# def download_file(url, max_size_bytes, output_filename, api_key=None):
#     """
#     Download a file from a given URL with size limit and optional API key.
#     """
#     logging.debug(f'Starting file download from URL: {url}')
#     try:
#         headers = {}
#         if api_key:
#             headers['Authorization'] = f'Bearer {api_key}'
#             logging.debug('API key provided, added to headers')
#
#         response = requests.get(url, stream=True, headers=headers)
#         response.raise_for_status()
#
#         file_size = int(response.headers.get('Content-Length', 0))
#         logging.info(f'File size: {file_size} bytes')
#
#         if file_size > max_size_bytes:
#             logging.error(f'File size exceeds limit: {file_size} > {max_size_bytes}')
#             return False
#
#         downloaded_size = 0
#         with open(output_filename, 'wb') as file:
#             for chunk in response.iter_content(chunk_size=8192):
#                 downloaded_size += len(chunk)
#                 logging.debug(f'Downloaded {downloaded_size} bytes')
#                 if downloaded_size > max_size_bytes:
#                     logging.error('Downloaded size exceeds maximum allowed payload size')
#                     return False
#                 file.write(chunk)
#
#         logging.info(f'File downloaded successfully: {output_filename}')
#         return True
#
#     except requests.RequestException as e:
#         logging.error(f"Error downloading file: {e}")
#         return False


@app.get("/")
async def read_root():
    return {"message": "This is the Ivrit AI Streaming service."}


# async def transcribe_core_ws(audio_file):
#     ret = {'segments': []}
#
#     try:
#
#         logging.debug(f"Initiating model transcription for file: {audio_file}")
#
#         segs, _ = await asyncio.to_thread(model.transcribe, audio_file, language='he', word_timestamps=True)
#         logging.info('Transcription completed successfully.')
#     except Exception as e:
#         logging.error(f"Error during transcription: {e}")
#         raise e
#
#     # Track the new segments and update the last transcribed time
#     for s in segs:
#         logging.info(f"Processing segment with start time: {s.start} and end time: {s.end}")
#
#         # Only process segments that start after the last transcribed time
#         logging.info(f"New segment found starting at {s.start} seconds.")
#         words = [{'start': w.start, 'end': w.end, 'word': w.word, 'probability': w.probability} for w in s.words]
#
#         seg = {
#             'id': s.id, 'seek': s.seek, 'start': s.start, 'end': s.end, 'text': s.text,
#             'avg_logprob': s.avg_logprob, 'compression_ratio': s.compression_ratio,
#             'no_speech_prob': s.no_speech_prob, 'words': words
#         }
#         logging.info(f'Adding new transcription segment: {seg}')
#         ret['segements'].append(seg)
#
#         # Update the last transcribed time to the end of the current segment
#
#
# #logging.info(f"Returning {len(ret['new_segments'])} new segments and updated last transcribed time.")
#     return ret


import tempfile


async def transcribe_core_ws(audio_file):

    ret = {'segments': []}

    try:
        # Transcribe the entire audio file
        logging.debug(f"Initiating model transcription for file: {audio_file}")
        segs, _ = await asyncio.to_thread(model.transcribe,audio_file, language='he', word_timestamps=True)
        logging.info('Transcription completed successfully.')
    except Exception as e:
        logging.error(f"Error during transcription: {e}")
        raise e

    # Track the new segments and update the last transcribed time
    for s in segs:
        logging.info(f"Processing segment with start time: {s.start} and end time: {s.end}")
        words = [{'start': w.start, 'end': w.end, 'word': w.word, 'probability': w.probability} for w in s.words]

        seg = {
            'id': s.id, 'seek': s.seek, 'start': s.start, 'end': s.end, 'text': s.text,
            'avg_logprob': s.avg_logprob, 'compression_ratio': s.compression_ratio,
            'no_speech_prob': s.no_speech_prob, 'words': words
        }
        logging.info(f'Adding new transcription segment: {seg}')
        ret['segments'].append(seg)


    #logging.info(f"Returning {len(ret['new_segments'])} new segments and updated last transcribed time.")
    return ret


import tempfile


# Function to verify if the PCM data is valid
def validate_pcm_data(pcm_audio_buffer, sample_rate, channels, sample_width):
    """Validates the PCM data buffer to ensure it conforms to the expected format."""
    logging.info(f"Validating PCM data: total size = {len(pcm_audio_buffer)} bytes.")

    # Calculate the expected sample size
    expected_sample_size = sample_rate * channels * sample_width
    actual_sample_size = len(pcm_audio_buffer)

    if actual_sample_size == 0:
        logging.error("Received PCM data is empty.")
        return False

    logging.info(f"Expected sample size per second: {expected_sample_size} bytes.")

    if actual_sample_size % expected_sample_size != 0:
        logging.warning(
            f"PCM data size {actual_sample_size} is not a multiple of the expected sample size per second ({expected_sample_size} bytes). Data may be corrupted or incomplete.")

    return True


# Function to validate if the created WAV file is valid
def validate_wav_file(wav_file_path):
    """Validates if the WAV file was created correctly and can be opened."""
    try:
        with wave.open(wav_file_path, 'rb') as wav_file:
            sample_rate = wav_file.getframerate()
            channels = wav_file.getnchannels()
            sample_width = wav_file.getsampwidth()
            logging.info(
                f"WAV file details - Sample Rate: {sample_rate}, Channels: {channels}, Sample Width: {sample_width}")
            return True
    except wave.Error as e:
        logging.error(f"Error reading WAV file: {e}")
        return False


@app.websocket("/wtranscribe")
async def websocket_transcribe(websocket: WebSocket):
    logging.info("New WebSocket connection request received.")
    await websocket.accept()
    logging.info("WebSocket connection established successfully.")

    try:
        segments = []  # Keeps track of the segments already transcribed
        accumulated_audio_time = 0  # Track the total audio duration accumulated
        last_transcribed_time = 0.0
        min_transcription_time = 5.0  # Minimum duration of audio in seconds before transcription starts

        # A buffer to store raw PCM audio data
        pcm_audio_buffer = bytearray()
        logging.info("im here, is it failing?.")

        # Metadata for the incoming PCM data (sample rate, channels, and sample width should be consistent)
        sample_rate = 16000  # 16kHz
        channels = 1  # Mono
        sample_width = 2  # 2 bytes per sample (16-bit audio)

        # Ensure the /tmp directory exists
        tmp_directory = "/tmp"
        if not os.path.exists(tmp_directory):
            logging.info(f"Creating /tmp directory: {tmp_directory}")
            os.makedirs(tmp_directory)
        logging.info("im here, is it failing?2.")
        while True:
            logging.info("in while true")
            try:
                # Receive the next chunk of PCM audio data
                logging.info("in try before recive ")
                audio_chunk = await websocket.receive_bytes()
                if isinstance(audio_chunk, bytes):
                    print("is instance ")
                else:
                    print("is instance NOT")


                logging.info("after recieve")
                sys.stdout.flush()
                if not audio_chunk:
                    logging.warning("Received empty audio chunk, skipping processing.")
                    continue

                # Accumulate the raw PCM data into the buffer
                pcm_audio_buffer.extend(audio_chunk)
                print(f"len of pcm buffer: {len(pcm_audio_buffer)}")
                logging.info("after buffer extend")

                # Validate the PCM data after each chunk
                if not validate_pcm_data(pcm_audio_buffer, sample_rate, channels, sample_width):
                    logging.error("Invalid PCM data received. Aborting transcription.")
                    await websocket.send_json({"error": "Invalid PCM data received."})
                    return

                # Estimate the duration of the chunk based on its size
                chunk_duration = len(audio_chunk) / (sample_rate * channels * sample_width)
                accumulated_audio_time += chunk_duration
                logging.info(
                    f"Received and buffered {len(audio_chunk)} bytes, total buffered: {len(pcm_audio_buffer)} bytes, total time: {accumulated_audio_time:.2f} seconds")

                # Transcribe when enough time (audio) is accumulated (e.g., at least 5 seconds of audio)
                if accumulated_audio_time >= min_transcription_time:
                    logging.info("Buffered enough audio time, starting transcription.")

                    # Create a temporary WAV file in /tmp for transcription

                    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False, dir="/tmp") as temp_wav_file:
                        logging.info(f"Temporary audio file created at {temp_wav_file.name}")

                        with wave.open(temp_wav_file.name, 'wb') as wav_file:
                            wav_file.setnchannels(channels)
                            wav_file.setsampwidth(sample_width)
                            wav_file.setframerate(sample_rate)
                            wav_file.writeframes(pcm_audio_buffer)
                            temp_wav_file.flush()

                        if not validate_wav_file(temp_wav_file.name):
                            logging.error(f"Invalid WAV file created: {temp_wav_file.name}")
                            await websocket.send_json({"error": "Invalid WAV file created."})
                            return

                    logging.info(f"Temporary WAV file created at {temp_wav_file.name} for transcription.")

                    # Log to confirm that the file exists and has the expected size
                    if os.path.exists(temp_wav_file.name):
                        file_size = os.path.getsize(temp_wav_file.name)
                        logging.info(f"Temporary WAV file size: {file_size} bytes.")
                    else:
                        logging.error(f"Temporary WAV file {temp_wav_file.name} does not exist.")
                        raise Exception(f"Temporary WAV file {temp_wav_file.name} not found.")

                    with open(temp_wav_file.name, 'rb') as audio_file:
                        audio_data = audio_file.read()
                    partial_result = await asyncio.to_thread(transcribe_core_ws,audio_data)
                    segments.extend(partial_result['segments'])

                    # Clear the buffer after transcription
                    pcm_audio_buffer.clear()
                    accumulated_audio_time = 0  # Reset accumulated time

                    # Send the transcription result back to the client with both new and all processed segments
                    response = {
                        "segments": segments
                    }
                    logging.info(f"Sending {len(partial_result['segments'])}  segments to the client.")
                    await websocket.send_json(response)

                    # Optionally delete the temporary WAV file after processing
                    if os.path.exists(temp_wav_file.name):
                        os.remove(temp_wav_file.name)
                        logging.info(f"Temporary WAV file {temp_wav_file.name} removed.")

            except WebSocketDisconnect:
                logging.info("WebSocket connection closed by the client.")
                break

    except Exception as e:
        logging.error(f"Unexpected error during WebSocket transcription: {e}")
        await websocket.send_json({"error": str(e)})

    finally:
        logging.info("Cleaning up and closing WebSocket connection.")


from fastapi.responses import FileResponse


@app.get("/download_audio/{filename}")
async def download_audio(filename: str):
    file_path = f"/tmp/{filename}"

    # Ensure the file exists before serving it
    if os.path.exists(file_path):
        return FileResponse(file_path, media_type='audio/wav', filename=filename)
    else:
        return {"error": "File not found"}

# @app.websocket("/ws")
# async def websocket_endpoint(websocket: WebSocket):
#     """WebSocket endpoint to handle client connections."""
#     await websocket.accept()
#     client_ip = websocket.client.host
#     logger.info(f"Client connected: {client_ip}")
#     sys.stdout.flush()
#     try:
#         await process_audio_stream(websocket)
#     except WebSocketDisconnect:
#         logger.info(f"Client disconnected: {client_ip}")
#     except Exception as e:
#         logger.error(f"Unexpected error: {e}")
#         await websocket.close()
#
# async def process_audio_stream(websocket: WebSocket):
#     """Continuously receive audio chunks and initiate transcription tasks."""
#     sampling_rate = 16000
#     min_chunk_size = 5  # in seconds
#
#     transcription_task = None
#     chunk_counter = 0
#     total_bytes_received = 0
#
#     while True:
#         try:
#             # Receive audio data from client
#             data = await websocket.receive_bytes()
#             if not data:
#                 logger.info("No data received, closing connection")
#                 break
#             chunk_counter += 1
#             chunk_size = len(data)
#             total_bytes_received += chunk_size
#             #logger.debug(f"Received chunk {chunk_counter}: {chunk_size} bytes")
#
#             audio_chunk = process_received_audio(data)
#             #logger.debug(f"Processed audio chunk {chunk_counter}: {len(audio_chunk)} samples")
#             # Check if enough audio has been buffered
#             # if transcription_task is None or transcription_task.done():
#             #     # Start a new transcription task
#         #     # logger.info(f"Starting transcription task for {len(audio_buffer)} samples")
#             transcription_task = asyncio.create_task(
#                 transcribe_and_send(websocket, audio_chunk)
#             )
#
#             #logger.debug(f"Audio buffer size: {len(audio_buffer)} samples")
#         except Exception as e:
#             logger.error(f"Error receiving data: {e}")
#             break
#
#
# async def transcribe_and_send(websocket: WebSocket, audio_data):
#     """Run transcription in a separate thread and send the result to the client."""
#     logger.debug(f"Transcription task started for {len(audio_data)} samples")
#     transcription_result = await asyncio.to_thread(sync_transcribe_audio, audio_data)
#     if transcription_result:
#         try:
#             # Send the result as JSON
#             await websocket.send_json(transcription_result)
#             logger.info(f"Transcription JSON sent to client {transcription_result}")
#         except Exception as e:
#             logger.error(f"Error sending transcription: {e}")
#     else:
#         logger.warning("No transcription result to send")
#
# def sync_transcribe_audio(audio_data):
#     """Synchronously transcribe audio data using the ASR model and format the result."""
#     try:
#
#         logger.info('Starting transcription...')
#         segments, info = model.transcribe(
#             audio_data, language="he",compression_ratio_threshold=2.5, word_timestamps=True
#         )
#         logger.info('Transcription completed')
#
#         # Build the transcription result as per your requirement
#         ret = {'segments': []}
#
#         for s in segments:
#             logger.debug(f"Processing segment {s.id} with start time: {s.start} and end time: {s.end}")
#
#             # Process words in the segment
#             words = [{
#                 'start': float(w.start),
#                 'end': float(w.end),
#                 'word': w.word,
#                 'probability': float(w.probability)
#             } for w in s.words]
#
#             seg = {
#                 'id': int(s.id),
#                 'seek': int(s.seek),
#                 'start': float(s.start),
#                 'end': float(s.end),
#                 'text': s.text,
#                 'avg_logprob': float(s.avg_logprob),
#                 'compression_ratio': float(s.compression_ratio),
#                 'no_speech_prob': float(s.no_speech_prob),
#                 'words': words
#             }
#             logger.debug(f'Adding new transcription segment: {seg}')
#             ret['segments'].append(seg)
#
#             logger.debug(f"Total segments in transcription result: {len(ret['segments'])}")
#             return ret
#     except Exception as e:
#         logger.error(f"Transcription error: {e}")
#         return {}
#
# def process_received_audio(data):
#     """Convert received bytes into normalized float32 NumPy array."""
#     #logger.debug(f"Processing received audio data of size {len(data)} bytes")
#     audio_int16 = np.frombuffer(data, dtype=np.int16)
#     #logger.debug(f"Converted to int16 NumPy array with {len(audio_int16)} samples")
#
#     audio_float32 = audio_int16.astype(np.float32) / 32768.0  # Normalize to [-1, 1]
#     #logger.debug(f"Normalized audio data to float32 with {len(audio_float32)} samples")
#
#     return audio_float32
#
#


# @app.websocket("/wtranscribe")
# async def websocket_transcribe(websocket: WebSocket):
#     logging.info("New WebSocket connection request received.")
#     await websocket.accept()
#     logging.info("WebSocket connection established successfully.")
#
#     try:
#         while True:
#             try:
#                 audio_chunk = await websocket.receive_bytes()
#                 if not audio_chunk:
#                     logging.warning("Received empty audio chunk, skipping processing.")
#                     continue
#                 with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_audio_file: ##new temp file for every chunk
#                     logging.info(f"Temporary audio file created at {temp_audio_file.name}")
#                     # Receive the next chunk of audio data
#
#
#
#                     partial_result = await transcribe_core_ws(temp_audio_file.name)
#                     await websocket.send_json(partial_result)
#
#             except WebSocketDisconnect:
#                 logging.info("WebSocket connection closed by the client.")
#                 break
#
#     except Exception as e:
#         logging.error(f"Unexpected error during WebSocket transcription: {e}")
#         await websocket.send_json({"error": str(e)})
#
#     finally:
#         logging.info("Cleaning up and closing WebSocket connection.")

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app)