speechtotext.functions

Module with functions for the speechtotext package.

Use this module like this:

# Imports
from speechtotext.functions import *

# Force torch use for cuda
force_cudnn_initialization()

# Clean string
string_cleaning("this has.//./8 to be cleaned::@")

Module attributes

REGEX_STRING_PARSE	Regex used to clean the transcripts.
DEFAULT_DATETIME_FORMAT	Default datetime format.
DEFAULT_REPORTS_FOLDER	Default folder to save the reports.
DEFAULT_CSV_NAME	Default path to save Benchmark results.

Functions

benchmark_results_to_csv	Creates csv from benchmark results.
force_cudnn_initialization	Force torch use for cuda if available.
get_extention_of_file_name	Get extention of file name.
get_file_name_without_extention	Get extention of file name.
join_benchmark_results	Join Benchmark results.
load_env_variable	Loads and returns env variable.
multidispatch	Allow for Method overloading for classes.
save_folder_name	Makes folder path.
save_sub_folder_name	Creates subfolder path.
separate_benchmark_results_by_model	Seperate benchmark results for each model.
string_cleaning	Cleaning of string for STT.
timing	Functions used to time duration of function.
uppercase_for_first_character_in_string	Return string where first character is uppercase.

Classes

BaseResult

Parent class for results.

Exceptions

NoTranscriptReturned()	Exception when API does not return a transcript.
RequiredEnvVariablesMissing(env_name)	Exception when an required env variable is missing.