Spaces:
Runtime error
Runtime error
| """ | |
| Author: Tanmay Jain | |
| Email: csetanmayjain@gmail.com | |
| """ | |
| import os | |
| import threading | |
| import uuid | |
| from flask import Flask, jsonify, request, send_file | |
| from flask_cors import CORS | |
| import nemo.collections.asr as nemo_asr | |
| import subprocess | |
| import config | |
| hi_asr_model = None | |
| en_asr_model = None | |
| en_upload_files_path=None | |
| hi_upload_files_path=None | |
| def initialize_path(): | |
| global hi_upload_files_path, en_upload_files_path, hi_am_model_path, en_am_model_path | |
| hi_am_model_path = config.hi_am_model_path | |
| en_am_model_path = config.en_am_model_path | |
| hi_upload_files_path = os.path.join(config.upload_files_path, "hindi") | |
| en_upload_files_path = os.path.join(config.upload_files_path, "english") | |
| os.makedirs(hi_upload_files_path, exist_ok=True) | |
| os.makedirs(en_upload_files_path, exist_ok=True) | |
| def load_model(lang): | |
| global hi_asr_model, en_asr_model | |
| if lang == "hi" and hi_asr_model == None: | |
| print("Loading Hindi Model") | |
| if not os.path.exists(hi_am_model_path): | |
| download=subprocess.run(["wget","-P",hi_am_model_path,"https://storage.googleapis.com/vakyansh-open-models/conformer_models/hindi/filtered_v1_ssl_2022-07-08_19-43-25/Conformer-CTC-BPE-Large.nemo"],capture_output=True, text=True) | |
| print('done Downloading hindi asr') | |
| if download.returncode != 0: | |
| raise Exception(f"wget download failed: {download.stderr}") | |
| #hi_asr_model = nemo_asr.models.EncDecCTCModelBPE.restore_from(hi_am_model_path) | |
| # path for hi_asr_model for local sever running | |
| hi_asr_model = nemo_asr.models.EncDecCTCModelBPE.restore_from('./hi_am_model/Conformer-CTC-BPE-Large.nemo') | |
| if lang == "en" and en_asr_model == None: | |
| if not os.path.exists(en_am_model_path): | |
| print(en_am_model_path) | |
| print('not found downloading english model') | |
| download=subprocess.run(["wget","-P",en_am_model_path,"https://storage.googleapis.com/vakyansh-open-models/conformer_models/english/2022-09-13_15-50-48/Conformer-CTC-BPE-Large.nemo"],capture_output=True, text=True) | |
| print('done Downloading') | |
| print("Loading English Model") | |
| en_asr_model = nemo_asr.models.EncDecCTCModelBPE.restore_from('./en_am_model/Conformer-CTC-BPE-Large.nemo') | |
| def transcribe(audio_file_path, transcription_file_path, lang, logprobs=False): | |
| #load_model(lang) | |
| if lang == "hi": | |
| transcription = hi_asr_model.transcribe([audio_file_path], logprobs=logprobs)[0] | |
| if lang == "en": | |
| transcription = en_asr_model.transcribe([audio_file_path], logprobs=logprobs)[0] | |
| f = open(transcription_file_path, "w") | |
| f.write(transcription) | |
| f.close() | |
| def asr(audio_file_path,lang): | |
| initialize_path() | |
| global en_upload_files_path,hi_upload_files_path | |
| if lang == None: | |
| return jsonify({'error': 'no language code provided'}), 400 | |
| if lang != "hi" and lang != "en": | |
| return jsonify({'error': 'Invalid language code'}), 400 | |
| client_id = str(uuid.uuid4()) | |
| if lang == "hi": | |
| client_id_dir = os.path.join(hi_upload_files_path, client_id) | |
| if lang == "en": | |
| client_id_dir = os.path.join(en_upload_files_path, client_id) | |
| transcription_file_path = os.path.join(client_id_dir, "transcript.txt") | |
| os.makedirs(client_id_dir) | |
| transcribe(audio_file_path,transcription_file_path,lang) | |
| return client_id | |
| def get_transcription(client_id,lang): | |
| global en_upload_files_path | |
| if client_id == None: | |
| return 'No Client ID Provided' | |
| if lang == None: | |
| return 'no language code provided' | |
| if lang != "hi" and lang != "en": | |
| return 'Invalid language code' | |
| print(type(lang),en_upload_files_path,"PATHS is thisss eherreb") | |
| if lang == "hi": | |
| print('looking in hindi file path') | |
| client_dir = os.path.join('dependency/audio_logs/hindi', client_id) | |
| if lang == "en": | |
| print('looking in english file path') | |
| client_dir = os.path.join('dependency/audio_logs/english', client_id) | |
| transcript_file_path = os.path.join(client_dir,'transcript.txt') | |
| if not os.path.exists(client_dir): | |
| return "Invalid Client ID" | |
| if os.path.exists(transcript_file_path): | |
| print(transcript_file_path,"transcript file path") | |
| #return send_file(transcript_file_path, as_attachment=True, download_name='transcript.txt'), 201 | |
| return transcript_file_path | |
| else: | |
| return 'your transcription file is not ready' | |
| # if __name__ == '__main__': | |
| # initialize_path() | |
| # app.run(host='0.0.0.0', port=5000) | |