Hi I would like to know about the ideal bit rate and frame rate for training and later testing with deepspeech 0.7.0 checkpoint.
Presently I am converting all the files into 256k bit rate and 16000 frame rate through the ffmpeg command -
os.system(command = f"ffmpeg -i {os.path.join(raw_wav_dir,file)} -b:a 256k -ar 16000 {dest}")
Should I opt for some other bit rate or frame rate? Kindly let me know.