Background of error
1, voice recognition model generation, test is no problem!
2, but with their own audio files, because the file size is 21 KB, online training sets are 32 KB; Don't know how to let oneself of the file size is 32 KB, don't know how to grasp the size
3, the error below
ValueError: always reshape a array of size 10399 into shape (64000, 1)


Request you, it is how to do the test file is the same size?