| | import pathlib |
| | import re |
| | import logging |
| | import json |
| |
|
| | DEBUG = False |
| | LOG_LEVEL = logging.DEBUG if DEBUG else logging.WARNING |
| |
|
| | logging.getLogger("pywhispercpp").setLevel(logging.WARNING) |
| | logging.basicConfig( |
| | level=LOG_LEVEL, |
| | format="%(asctime)s - %(levelname)s - %(message)s", |
| | filename='translator.log', |
| | datefmt="%H:%M:%S" |
| | ) |
| |
|
| | SAVE_DATA_SAVE = False |
| |
|
| | console_handler = logging.StreamHandler() |
| | console_handler.setLevel(LOG_LEVEL) |
| | console_formatter = logging.Formatter("%(asctime)s - %(levelname)s - %(message)s") |
| | console_handler.setFormatter(console_formatter) |
| | logging.getLogger().addHandler(console_handler) |
| |
|
| | |
| | FRAME_SCOPE_TIME_THRESHOLD = 6 |
| | |
| | MAX_SPEECH_DURATION_S = 15 |
| |
|
| | BASE_DIR = pathlib.Path(__file__).parent.parent |
| | MODEL_DIR = BASE_DIR / "moyoyo_asr_models" |
| | ASSERT_DIR = BASE_DIR / "assets" |
| | CONFIG_DIR = BASE_DIR / "config" |
| |
|
| | SAMPLE_RATE = 16000 |
| | |
| | SENTENCE_END_MARKERS = ['.', '!', '?', '。', '!', '?', ';', ';', ':', ':'] |
| | PAUSE_END_MARKERS = [',', ',', '、'] |
| | |
| | ALL_MARKERS = SENTENCE_END_MARKERS + PAUSE_END_MARKERS |
| | |
| | REGEX_MARKERS = re.compile(r'[' + re.escape(''.join(ALL_MARKERS)) + r']$') |
| |
|
| | sentence_end_chars = ''.join([re.escape(char) for char in SENTENCE_END_MARKERS]) |
| | SENTENCE_END_PATTERN = re.compile(f'[{sentence_end_chars}]') |
| |
|
| | |
| | pattern_string = '[' + ''.join([re.escape(char) for char in PAUSE_END_MARKERS]) + r']$' |
| | PAUSE_END_PATTERN = re.compile(pattern_string) |
| | |
| | WHISPER_PROMPT_ZH = "以下是简体中文普通话的句子。" |
| | MAX_LENGTH_ZH = 4 |
| |
|
| | WHISPER_PROMPT_EN = "" |
| | MAX_LENGTH_EN = 8 |
| |
|
| | |
| | WHISPER_MODEL_EN = 'large-v3-turbo-q5_0' |
| | |
| | WHISPER_MODEL_ZH = 'large-v3-turbo-q5_0' |
| | |
| | LLM_MODEL_PATH = (MODEL_DIR / "qwen2.5-1.5b-instruct-q5_0.gguf").as_posix() |
| | LLM_LARGE_MODEL_PATH = (MODEL_DIR / "qwen2.5-1.5b-instruct-q5_0.gguf").as_posix() |
| | |
| |
|
| | |
| | VAD_MODEL_PATH = (MODEL_DIR / "silero-vad" / "silero_vad.onnx").as_posix() |
| |
|