I have disabled speech recognition by deleting some code that does transcription, and then providing some quality metrics on the audio files. This needed modification of the schema and tables as well. I have attached evidence below of testing for a successful run.
These are the logs from my backend code:
flask run --debug
* Serving Flask app '__init__.py'
* Debug mode: on
Level: INFO
Time: 2023-08-10 10:17:19,191
Logger: werkzeug
Path: _internal:224
Function :_log
Message: WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
* Running on http://127.0.0.1:5000
Level: INFO
Time: 2023-08-10 10:17:19,191
Logger: werkzeug
Path: _internal:224
Function :_log
Message: Press CTRL+C to quit
Level: INFO
Time: 2023-08-10 10:17:19,192
Logger: werkzeug
Path: _internal:224
Function :_log
Message: * Restarting with stat
Level: WARNING
Time: 2023-08-10 10:17:19,431
Logger: werkzeug
Path: _internal:224
Function :_log
Message: * Debugger is active!
Level: INFO
Time: 2023-08-10 10:17:19,438
Logger: werkzeug
Path: _internal:224
Function :_log
Message: * Debugger PIN: 140-917-433
Level: INFO
Time: 2023-08-10 10:18:33,018
Logger: werkzeug
Path: _internal:224
Function :_log
Message: 127.0.0.1 - - [10/Aug/2023 10:18:33] "GET /api/v1/speak/get_reference HTTP/1.1" 200 -
Level: INFO
Time: 2023-08-10 10:18:33,021
Logger: werkzeug
Path: _internal:224
Function :_log
Message: 127.0.0.1 - - [10/Aug/2023 10:18:33] "GET /api/v1/speak/get_reference HTTP/1.1" 200 -
Level: INFO
Time: 2023-08-10 10:18:58,321
Logger: speak
Path: length_check:9
Function :get_audio_length
Message: duration (s): 3.54
0000000000000000000000000000000000000001111111111+(1.1700000000000008)11111111111111111111111111111111111111111111111111111111111111111110-(3.509999999999992)
Level: INFO
Time: 2023-08-10 10:18:58,324
Logger: speak
Path: silence_check:165
Function :get_silence_ratio
Message: Voiced frames count: 77
Level: INFO
Time: 2023-08-10 10:18:58,324
Logger: speak
Path: silence_check:166
Function :get_silence_ratio
Message: Total frames count: 117
Level: INFO
Time: 2023-08-10 10:18:58,324
Logger: speak
Path: silence_check:168
Function :get_silence_ratio
Message: Silence ratio: 0.34
Level: INFO
Time: 2023-08-10 10:18:59,449
Logger: werkzeug
Path: _internal:224
Function :_log
Message: 127.0.0.1 - - [10/Aug/2023 10:18:59] "POST /api/v1/speak/submit/test10.874510967740258.webm HTTP/1.1" 200 -
This is my screenshot from the front end
This is the file metadata being recorded in the sqlite database.
Aug 11, 2023: Edit this code now depends on Minio
Refer to this comment and look at the associated PR to get Minio installed:
TODO: README needs updating.
I have disabled speech recognition by deleting some code that does transcription, and then providing some quality metrics on the audio files. This needed modification of the schema and tables as well. I have attached evidence below of testing for a successful run.
These are the logs from my backend code:
This is my screenshot from the front end
This is the file metadata being recorded in the sqlite database.
These are the sqlite tables:
Audio file uploaded to S3: