uzi007's picture
Added GPU Models
a791cf5
raw
history blame
15 kB
import os
import json
import uvicorn
from fastapi import FastAPI, Request
from fastapi.middleware.cors import CORSMiddleware
from contextlib import asynccontextmanager
from helperfunctions import *
from models import load_models
from languages import CODE2LANG
from s3_handler import S3Handler
from media_download import YoutubeDownloader
from transcription import StableWhisper
from translation import Translation
from summarizer import Extract_Summary, AudioBookNarration
from audiobook import AudioBook
### API Configurations
# Context Manager for FastAPI Start/Shutdown
@asynccontextmanager
async def lifespan(app: FastAPI):
## FastAPI Startup Code
# Loading ML models
print('Loading ML Models..')
# For Storing Models
global MODELS
MODELS = load_models()
print('ML Models Loaded!')
yield
## FastAPI Shutdown Code
# Cleaning ML Models & Releasing the Resources
MODELS.clear()
# Initializing FastAPI App
app = FastAPI(lifespan=lifespan)
# Output Directory for Files Storage
output_folder = 'Output'
# S3 Handler
s3 = S3Handler()
# Create a context variable to store the contexts for each user
users_context = dict()
# CORS (Cross-Origin Resource Sharing)
origins = [
"http://localhost",
"http://localhost:4200",
]
app.add_middleware(
CORSMiddleware,
allow_origins=["*"], # origins,
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
### APIs
@app.get("/get_media_metadata")
async def get_media_metadata(request: Request, url: str):
# Getting User's IP & Generating UUID
user_ip = request.client.host
user_id = generate_uuid(user_ip, url)
# User Folder Path
user_folder_path = os.path.join(output_folder, user_id)
# Getting User's Youtube Downloader
youtube_downloader = YoutubeDownloader(url, user_folder_path)
# Getting Youtube Media Info
media_metadata = youtube_downloader.get_media_metadata()
# Storing User's Media Metadata to Directory
media_metadata_path = os.path.join(user_folder_path, 'media_metadata.json')
with open(media_metadata_path, "w") as outfile:
json.dump(media_metadata, outfile)
# Storing User's Media Metadata to S3
s3_path = s3.upload_file(user_id, 'media_metadata.json', media_metadata_path)
# Getting Status
status = 1 if media_metadata else 0
if status:
# Storing Info in the context for this user's session
users_context[user_id] = dict()
users_context[user_id]['downloader'] = youtube_downloader
# users_context[user_id]['media_metadata'] = media_metadata
users_context[user_id]['url'] = url
return {'status': status,
'user_id': user_id,
'media_metadata': media_metadata,
'media_metadata_path': s3_path}
@app.get("/get_media_formats")
async def get_media_formats(user_id: str):
# Getting Media Formats for User
media_formats = users_context[user_id]['downloader'].get_media_formats()
# User Folder Path
user_folder_path = os.path.join(output_folder, user_id)
# Storing User's Media Formats to Directory
media_formats_path = os.path.join(user_folder_path, 'media_formats.json')
with open(media_formats_path, "w") as outfile:
json.dump(media_formats, outfile)
# Storing User's Media Formats to S3
s3_path = s3.upload_file(user_id, 'media_formats.json', media_formats_path)
# Getting Status
status = 1 if media_formats else 0
if status:
# Storing Media Info in the context for this user's session
users_context[user_id]['media_formats'] = media_formats
return {'status': status,
'media_formats': media_formats,
'media_formats_path': s3_path}
@app.get("/download_media")
async def download_media(user_id: str, media_type: str, media_format: str, media_quality: str):
# Downloading Media for User
media_path = users_context[user_id]['downloader'].download(media_type, media_format, media_quality)
# Storing User's Downloaded Media to S3
media_file = f"{media_type.lower()}_{media_quality.lower()}.{media_format.lower()}"
s3_path = s3.upload_file(user_id, media_file, media_path)
# Getting Status
status = 1 if media_path else 0
if status:
# Storing Media Info in the context for this user's session
users_context[user_id]['media_path'] = media_path
users_context[user_id]['media_type'] = media_type
return {'status': status, 'media_path': s3_path}
@app.get("/get_transcript")
async def get_transcript(user_id: str, subtitle_format: str = 'srt', word_level: bool = False):
# If Video Already Downloaded
if 'media_path' in users_context[user_id].keys():
# Retrieving the media_path from the context for this user's session
media_path = users_context[user_id]['media_path']
# Checking if the media_type is Video, then extract it's audio
media_type = users_context[user_id]['media_type']
if media_type == 'video':
media_path = extract_audio(media_path)
else:
# Downloading Audio for Transcription
media_path = users_context[user_id]['downloader'].download('audio', 'mp3', '128kbps')
# Whisper based transcription
user_folder_path = os.path.join(output_folder, user_id)
stable_whisper_transcript = StableWhisper(model=MODELS['transcription'],
media_path=media_path,
output_path=user_folder_path,
subtitle_format=subtitle_format,
word_level=word_level)
transcript = stable_whisper_transcript.generate_transcript()
transcript_path = stable_whisper_transcript.save_transcript()
subtitles_path = stable_whisper_transcript.save_subtitles()
# Storing User's Transcripts to S3
s3_transcript_path = s3.upload_file(user_id, 'transcript.txt', transcript_path)
s3_subtitles_path = s3.upload_file(user_id, f'subtitles.{subtitle_format}', subtitles_path)
# Getting Status
status = 1 if transcript and s3_transcript_path and s3_subtitles_path else 0
if status:
# Storing Transcript Info in the context for this user's session
users_context[user_id]['transcript'] = transcript
users_context[user_id]['transcript_path'] = transcript_path
users_context[user_id]['subtitles_path'] = subtitles_path
return {'status': status,
'transcript': transcript,
'transcript_path': s3_transcript_path,
'subtitles_path': s3_subtitles_path}
@app.get("/get_languages")
async def get_translation(user_id: str):
return CODE2LANG
@app.get("/get_translation")
async def get_translation(user_id: str, target_language: str = 'en'):
# If Transcript Available
if 'transcript' in users_context[user_id].keys():
# Retrieving the transcript from the context for this user's session
transcript = users_context[user_id]['transcript']
else:
return {'status': 0, 'message': 'Transcript not generated yet'}
# NLLB based Translation
user_folder_path = os.path.join(output_folder, user_id)
nllb_translator = Translation(model=MODELS['translation'],
transcript_dict=transcript,
source_lang=transcript['language'],
target_lang=target_language,
output_path=user_folder_path)
translated_transcript = nllb_translator.get_translated_transcript()
translated_subtitles = nllb_translator.get_translated_subtitles()
# Storing Translated Transcript as TXT file in UTF-8 format
translated_transcript_path = os.path.join(user_folder_path, 'translated_transcript.txt')
with open(translated_transcript_path, 'w', encoding='utf-8') as file:
file.write(translated_transcript)
# Storing Translated Transcript to S3
s3_transcript_path = s3.upload_file(user_id, 'translated_transcript.txt', translated_transcript_path)
# TODO: Write Translated Transcript as SRT, VTT, ASS files
# Storing Translated Subtitles as JSON file (For Now)
translated_subtitles_path = os.path.join(user_folder_path, 'translated_subtitles.json')
with open(translated_subtitles_path, "w", encoding='utf-8') as file:
json.dump(media_formats, file)
# Storing Translated Subtitles to S3
s3_subtitles_path = s3.upload_file(user_id, 'translated_subtitles.json', translated_subtitles_path)
# Getting Status
status = 1 if translated_transcript and translated_subtitles else 0
if status:
# Storing Translated Transcript Info in the context for this user's session
users_context[user_id]['translated_transcript'] = translated_transcript
users_context[user_id]['translated_transcript_path'] = translated_transcript_path
users_context[user_id]['translated_subtitles'] = translated_subtitles
users_context[user_id]['translated_subtitles_path'] = translated_subtitles_path
return {'status': status,
'transcript': translated_transcript,
'subtitles': translated_subtitles,
'transcript_path': s3_transcript_path,
'subtitles_path': s3_subtitles_path}
@app.get("/get_summary")
async def get_summary(user_id: str, Summary_type: str, Summary_strategy: str, Target_Person_type: str,
Response_length: str, Writing_style: str):
# If Transcript Available
if 'transcript' in users_context[user_id].keys():
# Retrieving the transcript from the context for this user's session
text_input = users_context[user_id]['transcript']
else:
return {'status': 0, 'message': 'Transcript not generated yet'}
# Extracting Summary
summary_extractor = Extract_Summary(text_input=text_input)
output = summary_extractor.define_chain(Summary_type=Summary_type,
Summary_strategy=Summary_strategy,
Target_Person_type=Target_Person_type,
Response_length=Response_length,
Writing_style=Writing_style,
key_information=False)
# Getting Status
status = 1 if output else 0
if status:
# Storing Summary Info in the context for this user's session
users_context[user_id]['summary'] = output
return {'status': status, "summary": output}
@app.get("/get_key_info")
async def get_key_info(user_id: str, Summary_type: str, Summary_strategy: str, Target_Person_type: str,
Response_length: str, Writing_style: str):
# If Transcript Available
if 'transcript' in users_context[user_id].keys():
# Retrieving the transcript from the context for this user's session
text_input = users_context[user_id]['transcript']
else:
return {'status': 0, 'message': 'Transcript not generated yet'}
# Extracting Key Value Info
summary_extractor = Extract_Summary(text_input=text_input)
output = summary_extractor.define_chain(Summary_type=Summary_type,
Summary_strategy=Summary_strategy,
Target_Person_type=Target_Person_type,
Response_length=Response_length,
Writing_style=Writing_style,
key_information=True)
# Getting Status
status = 1 if output else 0
if status:
# Storing Key Info in the context for this user's session
users_context[user_id]['key_info'] = output
return {'status': status, "key_info": output}
@app.get("/get_audiobook")
async def get_audiobook(user_id: str, narration_style: str, speaker: str = "male",
audio_format: str = "mp3", audio_quality: str = "128kbps"):
# If Transcript Available
if 'transcript' in users_context[user_id].keys():
# Retrieving the transcript from the context for this user's session
text_input = users_context[user_id]['transcript']
else:
return {'status': 0, 'message': 'Transcript not generated yet'}
# Extracting Narration
narrator = AudioBookNarration(text_input=text_input)
output = narrator.define_chain(narration_style=narration_style)
# Generating Audiobook
audiobook = AudioBook(output_folder=output_folder)
audio_path = audiobook.generate_audio_from_text(output, speaker=speaker, filename="output_audio")
# Converting the Audio to Required Audio Parameters
audio_path = convert_audio(audio_path, audio_format, audio_quality)
# Storing User's Audiobook to S3
media_file = f"audiobook_{audio_quality.lower()}.{audio_format.lower()}"
s3_path = s3.upload_file(user_id, media_file, audio_path)
# Getting Status
status = 1 if audio_path else 0
if status:
# Storing Audiobook path in the context for this user's session
users_context[user_id]['audiobook_path'] = audio_path
return {'status': status, "audiobook_path": s3_path}
@app.get("/get_rendered_video")
async def get_rendered_video(user_id: str, video_format: str, video_quality: str, subtitles_type: str = 'original'):
# # Retrieving the media_path from the context for this user's session
# media_path = users_context[user_id]['media_path']
# Downloading Video with Required Video Parameters for User
media_path = users_context[user_id]['downloader'].download('video', video_format, video_quality)
# Getting Required Subtitles
if 'original' in subtitles_type.lower():
subtitles_path = users_context[user_id]['subtitles_path']
elif 'translated' in subtitles_type.lower():
# Getting Translated Subtitles from the context for this user's session
translated_subtitles = users_context[user_id]['translated_subtitles_path']
# Saving Translated Subtitles
subtitles_path = save_translated_subtitles(translated_subtitles, media_path)
# Burning Subtitles & Rendering Video
rendered_video_path = burn_subtitles(media_path, subtitles_path)
# Storing User's Rendered Video to S3
media_file = f"subtitles_video_{video_quality.lower()}.{video_format.lower()}"
s3_path = s3.upload_file(user_id, media_file, media_path)
# Getting Status
status = 1 if rendered_video_path else 0
return {'status': status, "rendered_video_path": s3_path}
if __name__ == "__main__":
uvicorn.run(app, host="127.0.0.1", port=8000)