Revert "1)Modified code to accomodate microsoft asr_models also which require their custom code to run, hence set trust_remote_code=True in pipe, 2)Added extra error handling, 3) Now, pipe is loaded in main thread instead of background thread."

1)Modified code to accomodate microsoft asr_models also which require their custom code to run, hence set trust_remote_code=True in pipe, 2)Added extra error handling, 3) Now, pipe is loaded in main thread instead of background thread.

minor changes

Added email notification whenever transcription for any new model is starts

1) added model id sanitization i.e. removing invaid character as per hugging face \

removed return_timestamps=True parameter from pipe

bug fix : ASR_model = openai/whisper-tiny, csv_transcript, csv_result values were hardcoded, making them query string dependent inside generateTranscript

minor bug : double quotes inside double quotes not allowed in csv_path string

small bug fix

Major Update : created all the routes reqired, add utils folder containing helper fxns

added cpu score check to see the performance of allocated cpu in free tier

update output_dir = data instead of /data which was wrong and removed GPU available check because transformer library automatically uses GPU if available

minor bug {idx+1} is a set so can't multiply by 100

write files to /data directory becuase ./ directory is not allowed to write in HF spaces

small bug fixed : df used before initialization

corrected status route and variable job_status

added code to track transcription process running in background

fixing the error - WORKER TIMEOUT says gunicorn because it's taking too long to respond, hence using threading to run generateTranscription fxn in background

bug fix try : ./cache write permission not given

added generate transcript fxn

minor changes

minor changes

Add application file