icefall/searchindex.js

Search.setIndex({"docnames": ["contributing/code-style", "contributing/doc", "contributing/how-to-create-a-recipe", "contributing/index", "huggingface/index", "huggingface/pretrained-models", "huggingface/spaces", "index", "installation/index", "recipes/aishell/conformer_ctc", "recipes/aishell/index", "recipes/aishell/stateless_transducer", "recipes/aishell/tdnn_lstm_ctc", "recipes/index", "recipes/librispeech/conformer_ctc", "recipes/librispeech/index", "recipes/librispeech/tdnn_lstm_ctc", "recipes/timit/index", "recipes/timit/tdnn_ligru_ctc", "recipes/timit/tdnn_lstm_ctc", "recipes/yesno/index", "recipes/yesno/tdnn"], "filenames": ["contributing/code-style.rst", "contributing/doc.rst", "contributing/how-to-create-a-recipe.rst", "contributing/index.rst", "huggingface/index.rst", "huggingface/pretrained-models.rst", "huggingface/spaces.rst", "index.rst", "installation/index.rst", "recipes/aishell/conformer_ctc.rst", "recipes/aishell/index.rst", "recipes/aishell/stateless_transducer.rst", "recipes/aishell/tdnn_lstm_ctc.rst", "recipes/index.rst", "recipes/librispeech/conformer_ctc.rst", "recipes/librispeech/index.rst", "recipes/librispeech/tdnn_lstm_ctc.rst", "recipes/timit/index.rst", "recipes/timit/tdnn_ligru_ctc.rst", "recipes/timit/tdnn_lstm_ctc.rst", "recipes/yesno/index.rst", "recipes/yesno/tdnn.rst"], "titles": ["Follow the code style", "Contributing to Documentation", "How to create a recipe", "Contributing", "Huggingface", "Pre-trained models", "Huggingface spaces", "Icefall", "Installation", "Conformer CTC", "aishell", "Stateless Transducer", "TDNN-LSTM CTC", "Recipes", "Conformer CTC", "LibriSpeech", "TDNN-LSTM-CTC", "TIMIT", "TDNN-LiGRU-CTC", "TDNN-LSTM-CTC", "YesNo", "TDNN-CTC"], "terms": {"we": [0, 1, 2, 3, 5, 6, 8, 9, 10, 11, 12, 13, 14, 16, 18, 19, 21], "us": [0, 1, 2, 4, 6, 7, 8, 9, 10, 11, 12, 14, 16, 18, 19, 21], "tool": 0, "make": [0, 1, 3, 9, 11, 14], "consist": [0, 11], "possibl": [0, 2, 3, 8, 9, 14], "black": 0, "format": [0, 9, 11, 12, 14, 16, 18, 19, 21], "flake8": 0, "check": [0, 14], "qualiti": [0, 10], "isort": 0, "sort": [0, 8], "import": 0, "The": [0, 1, 2, 6, 8, 9, 10, 12, 14, 16, 18, 19, 21], "version": [0, 8, 9, 11, 12, 14, 16, 18, 19], "abov": [0, 8, 9, 10, 11, 12, 14, 21], "ar": [0, 1, 3, 8, 9, 10, 11, 12, 13, 14, 16, 18, 19, 21], "12": [0, 8, 9, 11, 12, 14, 18, 21], "6b0": 0, "3": [0, 7, 12, 16, 21], "9": [0, 8, 9, 11, 12, 14, 16, 18, 21], "2": [0, 7, 16, 18, 19, 21], "5": [0, 9, 11, 12, 14, 16, 18, 19, 21], "after": [0, 1, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "run": [0, 2, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "command": [0, 1, 8, 9, 11, 12, 14, 16, 18, 19, 21], "git": [0, 8, 9, 11, 12, 14, 16, 18, 19, 21], "clone": [0, 8, 9, 11, 12, 14, 16, 18, 19, 21], "http": [0, 1, 2, 5, 6, 8, 9, 10, 11, 12, 14, 16, 18, 19, 21], "github": [0, 2, 5, 8, 9, 11, 12, 14, 16, 18, 19, 21], "com": [0, 2, 5, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "k2": [0, 2, 5, 6, 7, 9, 11, 12, 14, 16, 18, 19], "fsa": [0, 2, 5, 6, 8, 9, 11, 14], "icefal": [0, 2, 3, 5, 6, 9, 11, 12, 13, 14, 16, 18, 19, 21], "cd": [0, 1, 2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "pip": [0, 1, 8, 11], "instal": [0, 1, 4, 6, 7, 21], "pre": [0, 3, 4, 6, 7, 8], "commit": 0, "whenev": 0, "you": [0, 1, 2, 5, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "automat": [0, 6], "hook": 0, "invok": 0, "fail": [0, 8], "If": [0, 2, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "ani": [0, 8, 9, 11, 12, 14, 21], "your": [0, 1, 2, 4, 6, 7, 9, 11, 12, 14, 16, 18, 19, 21], "wa": [0, 8, 14, 16], "success": [0, 8], "pleas": [0, 1, 2, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "fix": [0, 8, 14], "issu": [0, 8, 14], "report": [0, 8], "some": [0, 1, 9, 11, 12, 14, 16, 18, 19, 21], "i": [0, 1, 2, 6, 8, 9, 10, 11, 12, 14, 16, 18, 19, 21], "e": [0, 2, 9, 11, 12, 14, 16, 18, 19, 21], "modifi": [0, 9, 12, 14, 16, 18, 19, 21], "file": [0, 2, 6, 9, 11, 12, 14, 16, 18, 19, 21], "place": [0, 8, 11, 14, 16], "so": [0, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "statu": 0, "failur": 0, "see": [0, 1, 6, 9, 11, 12, 14, 16, 18, 19, 21], "which": [0, 2, 6, 9, 10, 11, 12, 14, 16, 18, 19, 21], "ha": [0, 2, 9, 11, 12, 14, 16, 18, 19], "been": [0, 11], "befor": [0, 1, 9, 11, 12, 14], "further": 0, "chang": [0, 9, 11, 12, 14, 16, 18, 19, 21], "all": [0, 5, 6, 9, 11, 12, 14, 16, 18, 19, 21], "again": [0, 21], "should": [0, 2, 9, 11, 12, 14, 16, 18, 19, 21], "succe": 0, "thi": [0, 2, 3, 4, 8, 9, 11, 12, 13, 14, 16, 18, 19, 21], "time": [0, 9, 11, 12, 14, 16, 18, 19, 21], "succeed": 0, "want": [0, 8, 9, 11, 12, 14, 16, 18, 19, 21], "can": [0, 1, 2, 5, 6, 8, 9, 10, 11, 12, 14, 16, 18, 19, 21], "do": [0, 2, 9, 11, 12, 14, 16, 18, 19, 21], "21": [0, 8, 9, 11, 14, 18, 19], "your_changed_fil": 0, "py": [0, 2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "sphinx": 1, "write": [1, 2, 3], "have": [1, 2, 5, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "prepar": [1, 3], "environ": [1, 9, 10, 11, 12, 14, 16, 18, 19, 21], "doc": 1, "r": [1, 8, 18, 19], "requir": [1, 8], "txt": [1, 8, 9, 11, 12, 14, 16, 18, 19, 21], "set": [1, 8, 9, 11, 12, 14, 21], "up": [1, 8, 9, 12, 14, 16], "readi": [1, 9, 14], "refer": [1, 2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "restructuredtext": 1, "primer": 1, "familiar": 1, "build": [1, 8, 9, 11, 14], "local": [1, 8], "preview": 1, "what": [1, 2, 8, 11], "look": [1, 2, 5, 8, 9, 11, 12, 14], "like": [1, 2, 6, 8, 9, 11, 12, 14, 21], "publish": [1, 10], "html": [1, 2, 8], "gener": [1, 9, 11, 12, 14], "view": [1, 9, 11, 12, 14, 21], "follow": [1, 2, 3, 5, 6, 7, 8, 9, 11, 12, 14, 16, 18, 19, 21], "python3": [1, 8], "m": [1, 11, 18, 19], "server": [1, 6, 8], "It": [1, 2, 4, 8, 9, 10, 11, 12, 14, 16, 18, 19, 21], "print": [1, 9, 11, 12, 14, 16, 18, 19, 21], "serv": 1, "0": [1, 7, 9, 11, 12, 14, 16, 18, 19, 21], "port": 1, "8000": [1, 21], "open": [1, 10, 11, 14], "browser": [1, 4, 6], "go": [1, 9, 11, 14], "read": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "code": [2, 3, 7, 9, 14, 16, 18, 19, 21], "style": [2, 3, 7], "adjust": 2, "sytl": 2, "design": 2, "python": [2, 8, 9, 11, 14], "recommend": [2, 8, 9, 11, 12, 14], "test": [2, 7, 9, 11, 12, 14, 16, 18, 19], "valid": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "dataset": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "lhots": [2, 7, 9, 11, 14], "readthedoc": [2, 8], "io": [2, 8], "en": [2, 8], "latest": [2, 6, 8, 14, 16], "index": [2, 8], "yesno": [2, 7, 8, 13, 21], "veri": [2, 3, 11, 18, 19, 21], "good": 2, "exampl": [2, 6, 7, 16, 18, 19, 21], "speech": [2, 6, 7, 8, 10, 11, 13, 21], "pull": [2, 9, 11, 14], "380": [2, 19], "show": [2, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "add": [2, 9, 11, 12, 13], "new": [2, 3, 6, 8, 9, 10, 11, 12, 14, 16, 21], "suppos": 2, "would": [2, 8, 14, 16], "name": [2, 9, 11], "foo": [2, 9, 14], "eg": [2, 5, 8, 9, 11, 12, 14, 16, 18, 19, 21], "mkdir": [2, 9, 11, 12, 14, 16, 18, 19, 21], "p": [2, 8, 11, 18, 19], "asr": [2, 5, 8, 9, 11, 12, 14, 16, 18, 19, 21], "touch": 2, "sh": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "chmod": 2, "x": 2, "simpl": [2, 11], "own": 2, "otherwis": [2, 9, 11, 14], "librispeech": [2, 5, 7, 13, 14, 16], "assum": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "fanci": 2, "call": 2, "bar": [2, 9, 14], "organ": 2, "wai": [2, 3], "readm": [2, 9, 11, 12, 14, 16, 18, 19, 21], "md": [2, 5, 9, 11, 12, 14, 16, 18, 19, 21], "asr_datamodul": [2, 8], "pretrain": [2, 9, 11, 12, 14, 16, 18, 19, 21], "For": [2, 5, 9, 11, 12, 14, 16, 18, 19, 21], "instanc": [2, 5, 9, 11, 12, 14, 16, 18, 19, 21], "tdnn": [2, 8, 10, 13, 15, 17, 20], "its": [2, 11], "directori": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "structur": 2, "descript": [2, 9, 11, 12, 14, 16, 18, 19, 21], "contain": [2, 7, 9, 11, 12, 13, 14, 16, 18, 19, 21], "inform": [2, 9, 11, 12, 14, 16, 18, 19, 21], "g": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "wer": [2, 8, 16, 18, 19, 21], "etc": [2, 9, 11, 12, 14, 16, 18, 19, 21], "provid": [2, 6, 8, 9, 10, 11, 12, 13, 14, 16, 18, 19, 21], "pytorch": [2, 7, 11], "dataload": [2, 8], "take": [2, 21], "input": [2, 9, 11, 12, 14, 16, 18, 19, 21], "checkpoint": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "save": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "dure": [2, 6, 9, 11, 12, 14, 16, 18, 19, 21], "stage": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "definit": 2, "neural": [2, 9, 14], "network": [2, 9, 11, 14], "script": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "infer": 2, "tdnn_lstm_ctc": [2, 12, 16, 19], "conformer_ctc": [2, 9, 14], "get": [2, 6, 8, 9, 11, 12, 14, 16, 21], "feel": 2, "result": [2, 5, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "everi": 2, "kept": 2, "self": 2, "toler": 2, "duplic": 2, "among": [2, 8], "differ": [2, 8, 9, 10, 14], "invoc": 2, "help": [2, 9, 11, 12, 14, 16, 18, 19, 21], "blob": [2, 5], "master": [2, 5, 11], "transform": [2, 9, 14], "conform": [2, 10, 11, 13, 15], "base": [2, 9, 11, 12, 14], "lstm": [2, 10, 13, 15, 17], "attent": [2, 11, 12], "lm": [2, 8, 11, 16, 18, 19, 21], "rescor": [2, 12, 16, 18, 19, 21], "demonstr": [2, 4, 6], "consid": 2, "colab": 2, "notebook": 2, "welcom": 3, "There": [3, 9, 11, 12, 14], "mani": 3, "two": [3, 9, 11, 12, 14, 16, 18, 19, 21], "them": [3, 4, 5, 6, 9, 11, 12, 14, 16, 18, 19, 21], "To": [3, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "document": [3, 7], "repositori": 3, "recip": [3, 5, 7, 8, 9, 11, 12, 14, 16, 18, 19, 21], "In": [3, 6, 9, 11, 12, 14, 16, 18, 19, 21], "page": [3, 6, 9, 11, 12, 13, 14, 16, 18, 19, 21], "describ": [3, 4, 9, 11, 12, 14, 16, 18, 19], "how": [3, 4, 6, 7, 8, 9, 11, 12, 14, 16, 18, 19, 21], "creat": [3, 7, 9, 11, 12, 14, 16, 18, 19, 21], "data": [3, 10], "train": [3, 4, 6, 7], "decod": [3, 6], "model": [3, 4, 6, 7, 8], "section": [4, 8, 9, 14], "find": [4, 5, 6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "also": [4, 5, 8, 9, 11, 12, 14, 21], "try": [4, 6], "from": [4, 6, 8, 9, 10, 11, 12, 14, 16, 18, 19, 21], "within": [4, 6], "without": [4, 6, 9, 14], "anyth": [4, 6], "space": [4, 7], "youtub": [4, 7, 14, 16], "video": [4, 7, 14, 16], "upload": [5, 6, 9, 11, 12, 14, 16, 18, 19, 21], "huggingfac": [5, 7, 9, 11, 12, 14, 16, 18, 19, 21], "co": [5, 6, 9, 10, 11, 12, 14, 16, 18, 19, 21], "visit": [5, 6], "link": [5, 8], "search": [5, 6], "specif": [5, 11], "correspond": [5, 6], "aishel": [5, 7, 9, 11, 12, 13], "gigaspeech": 5, "wenetspeech": 5, "integr": 6, "framework": 6, "sherpa": 6, "need": [6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "download": [6, 7, 10], "window": 6, "maco": 6, "linux": 6, "even": [6, 8], "ipad": 6, "phone": 6, "start": [6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "address": [6, 11], "recognit": [6, 7, 10, 11, 13, 21], "screenshot": [6, 9, 11, 12, 14, 21], "select": [6, 16, 18, 19, 21], "languag": [6, 9, 11, 12], "current": [6, 8, 11, 13], "chines": [6, 10, 11], "english": [6, 21], "target": 6, "method": [6, 8, 9, 11, 12, 14, 16, 18, 19], "greedi": 6, "modified_beam_search": [6, 11], "choos": [6, 8], "number": [6, 9, 11, 12, 14, 16, 18, 19, 21], "activ": 6, "path": [6, 9, 11, 12, 14], "either": [6, 9, 11, 12, 14], "record": [6, 9, 10, 11, 12, 14, 16, 18, 19, 21], "click": [6, 8, 9, 11, 12, 14, 21], "button": 6, "submit": 6, "wait": 6, "moment": 6, "an": [6, 8, 9, 10, 11, 14, 21], "when": [6, 11, 14], "bottom": 6, "part": [6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "tabl": 6, "one": [6, 9, 11, 12, 14, 16, 18, 19, 21], "next": [6, 8, 14, 16], "gen": [6, 8, 14, 16], "kaldi": [6, 8, 14, 16], "subscrib": [6, 8, 14, 16], "channel": [6, 8, 9, 11, 12, 14, 16, 18, 19, 21], "nadira": [6, 8, 14, 16], "povei": [6, 8, 14, 16], "www": [6, 8, 10, 14, 16], "uc_vaumpkminz1pnkfxan9mw": [6, 8, 14, 16], "torchaudio": 7, "1": [7, 16, 18, 19, 21], "timit": [7, 13, 18, 19], "contribut": 7, "depend": [8, 9, 14], "step": [8, 9, 11, 12, 14, 21], "order": [8, 9, 12, 14, 16, 18, 19], "matter": 8, "org": [8, 10, 11], "least": 8, "v1": [8, 9, 12, 14, 16, 18, 19], "alreadi": 8, "don": [8, 9, 12, 14, 16, 18, 19, 21], "t": [8, 9, 11, 12, 14, 16, 18, 19, 21], "replac": 8, "compil": [8, 9, 11, 14], "against": 8, "strongli": 8, "collect": 8, "variabl": [8, 9, 12, 14], "pythonpath": 8, "point": [8, 9, 12, 14], "folder": [8, 9, 11, 12, 14, 16, 18, 19, 21], "tmp": [8, 9, 11, 12, 14, 16, 18, 19, 21], "setup": [8, 9, 11, 12, 14, 16, 18, 19, 21], "export": [8, 9, 11, 12, 14, 16, 18, 19, 21], "put": 8, "sever": [8, 9, 11, 12, 14, 16, 18, 19, 21], "same": [8, 9, 11, 12, 14, 16, 18, 19, 21], "switch": [8, 9, 14], "just": 8, "about": [8, 11], "virtualenv": 8, "8": [8, 9, 11, 14, 16, 21], "cpython3": 8, "6": [8, 9, 11, 14, 16, 18, 19], "final": [8, 14, 16], "64": [8, 11], "1540m": 8, "creator": 8, "cpython3posix": 8, "dest": 8, "ceph": [8, 9, 11, 14], "fj": [8, 11, 14], "fangjun": [8, 11, 14], "clear": 8, "fals": [8, 9, 11, 14], "no_vcs_ignor": 8, "global": 8, "seeder": 8, "fromappdata": 8, "bundl": 8, "setuptool": 8, "wheel": 8, "via": 8, "copi": 8, "app_data_dir": 8, "root": 8, "share": 8, "v": [8, 14, 18, 19], "irtualenv": 8, "ad": [8, 9, 11, 12, 14, 21], "seed": 8, "packag": 8, "57": [8, 14, 16], "36": [8, 11, 14], "bashactiv": 8, "cshellactiv": 8, "fishactiv": 8, "powershellactiv": 8, "pythonactiv": 8, "xonshactiv": 8, "sourc": [8, 9, 10, 11, 14], "bin": [8, 9, 14], "dev20210822": 8, "cpu": [8, 9, 21], "torch1": 8, "f": [8, 18, 19], "nightli": 8, "whl": 8, "2bcpu": 8, "cp38": 8, "linux_x86_64": 8, "mb": 8, "________________________________": 8, "185": [8, 9, 14, 21], "kb": [8, 18, 19], "graphviz": 8, "17": [8, 9, 14, 18, 19], "py3": 8, "none": [8, 9, 14], "18": [8, 9, 11, 12, 14, 16, 18, 19], "torch": [8, 9, 11, 14], "cach": 8, "manylinux1_x86_64": 8, "831": [8, 11, 19], "type": [8, 9, 11, 14, 21], "extens": 8, "typing_extens": 8, "10": [8, 9, 11, 12, 14, 16, 18, 19, 21], "26": [8, 11, 14, 19], "successfulli": 8, "probabl": [8, 11], "cuda": [8, 9, 11, 12, 14, 16, 18, 19], "req": 8, "7b1b76ge": 8, "q": 8, "audioread": 8, "soundfil": 8, "post1": 8, "py2": 8, "7": [8, 9, 12, 14, 16, 18, 19], "97": [8, 9], "cytoolz": 8, "11": [8, 9, 11, 12, 14, 16, 18, 19, 21], "manylinux_2_17_x86_64": 8, "manylinux2014_x86_64": 8, "dataclass": 8, "14": [8, 9, 14, 16, 18], "h5py": 8, "manylinux_2_12_x86_64": 8, "manylinux2010_x86_64": 8, "684": [8, 9, 21], "intervaltre": 8, "lilcom": 8, "numpi": 8, "15": [8, 11, 12, 14, 18, 21], "40": [8, 12, 14, 16, 18, 19], "pyyaml": 8, "662": 8, "tqdm": 8, "62": [8, 14, 16], "76": [8, 21], "73": 8, "satisfi": 8, "lib": 8, "site": 8, "dev": [8, 9, 11, 12, 14, 16, 18, 19, 21], "2a1410b": 8, "clean": [8, 9, 11, 14, 16], "toolz": 8, "55": [8, 12, 14, 18], "sortedcontain": 8, "29": [8, 9, 11, 12, 14, 16, 18, 19], "cffi": 8, "411": [8, 14], "pycpars": 8, "20": [8, 9, 11, 12, 14, 16, 18, 19, 21], "112": 8, "pypars": 8, "67": 8, "done": [8, 9, 11, 12, 14, 16, 18, 19, 21], "filenam": 8, "dev_2a1410b_clean": 8, "size": [8, 9, 11, 12, 14, 16, 18, 19, 21], "342242": 8, "sha256": 8, "f683444afa4dc0881133206b4646a": 8, "9d0f774224cc84000f55d0a67f6e4a37997": 8, "store": [8, 14], "ephem": 8, "ftu0qysz": 8, "7f": 8, "7a": 8, "8e": 8, "a0bf241336e2e3cb573e1e21e5600952d49f5162454f2e612f": 8, "warn": 8, "built": 8, "invalid": [8, 14], "metadata": [8, 18, 19], "mandat": 8, "pep": 8, "440": 8, "packa": 8, "ging": 8, "deprec": [8, 11], "legaci": 8, "becaus": 8, "could": [8, 9, 12], "A": [8, 9, 11, 12, 14, 16], "discuss": 8, "regard": 8, "pypa": 8, "sue": 8, "8368": 8, "inter": 8, "valtre": 8, "sor": 8, "tedcontain": 8, "remot": 8, "enumer": 8, "object": [8, 9, 11, 12, 21], "500": [8, 11, 14], "count": 8, "100": [8, 9, 11, 12, 14], "compress": 8, "308": [8, 9, 11, 12], "total": [8, 9, 11, 12, 14, 16, 21], "delta": 8, "263": 8, "reus": 8, "307": 8, "102": [8, 9], "pack": 8, "receiv": 8, "172": 8, "49": [8, 14, 19, 21], "kib": 8, "385": 8, "00": [8, 9, 11, 12, 14, 16, 18, 19, 21], "resolv": 8, "kaldilm": 8, "tar": 8, "gz": 8, "48": [8, 9, 11], "574": 8, "kaldialign": 8, "sentencepiec": [8, 14], "96": 8, "tensorboard": [8, 9, 11, 12, 14, 16, 18, 19, 21], "41": [8, 9, 11, 18, 21], "line": 8, "absl": 8, "absl_pi": 8, "13": [8, 11, 12, 14, 16, 18], "132": 8, "googl": 8, "auth": 8, "oauthlib": 8, "google_auth_oauthlib": 8, "grpcio": 8, "24": [8, 12, 16, 18, 19, 21], "39": [8, 11, 14, 16, 18], "ment": 8, "requi": 8, "rement": 8, "protobuf": 8, "manylinux_2_5_x86_64": 8, "werkzeug": 8, "288": 8, "tensorboard_data_serv": 8, "google_auth": 8, "35": [8, 11, 14], "152": 8, "request": 8, "plugin": 8, "wit": 8, "tensorboard_plugin_wit": 8, "781": 8, "markdown": 8, "six": 8, "16": [8, 9, 11, 12, 14, 16, 18, 19, 21], "cachetool": 8, "rsa": 8, "34": 8, "pyasn1": 8, "modul": 8, "pyasn1_modul": 8, "155": 8, "requests_oauthlib": 8, "23": [8, 9, 11, 12, 14, 18, 19, 21], "77": [8, 14], "urllib3": 8, "27": [8, 9, 11, 16, 19], "138": [8, 9, 11], "certifi": 8, "2017": 8, "2021": [8, 9, 12, 14, 16, 18, 19, 21], "30": [8, 9, 11, 12, 14, 21], "145": 8, "charset": 8, "normal": [8, 16, 18, 19, 21], "charset_norm": 8, "idna": 8, "59": [8, 12, 14], "146": 8, "897233": 8, "eccb906cafcd45bf9a7e1a1718e4534254bfb": 8, "f4c0d0cbc66eee6c88d68a63862": 8, "85": 8, "7d": 8, "63": [8, 11], "f2dd586369b8797cb36d213bf3a84a789eeb92db93d2e723c9": 8, "etool": 8, "oaut": 8, "hlib": 8, "let": [8, 9, 14], "u": [8, 9, 11, 12, 14, 21], "log": [8, 16, 18, 19, 21], "08": [8, 14, 16, 18, 19, 21], "19": [8, 9, 14, 16, 18, 19], "main": [8, 9, 14], "dl_dir": [8, 9, 12, 14], "waves_yesno": 8, "49mb": 8, "03": [8, 11, 14, 18, 19], "39mb": 8, "manifest": 8, "31": [8, 14], "42": [8, 9, 14, 21], "comput": [8, 9, 11, 12, 16, 18, 19, 21], "fbank": [8, 9, 11, 12, 14, 16, 18, 19, 21], "32": [8, 9, 11, 12], "803": 8, "info": [8, 9, 11, 12, 14, 16, 18, 19, 21], "compute_fbank_yesno": 8, "52": [8, 9, 14], "process": [8, 9, 11, 12, 14], "extract": [8, 9, 11, 12, 14, 16, 18, 19, 21], "featur": [8, 9, 11, 12, 14, 16, 18, 19, 21], "_______________________________________________________________": 8, "90": 8, "01": [8, 11, 12, 14], "80": [8, 9, 11, 14], "57it": 8, "085": 8, "______________________________________________________________": 8, "248": [8, 11], "21it": 8, "lang": [8, 11, 14], "fcordre9": 8, "kaldilm_6899d26f2d684ad48f21025950cd2866": 8, "csrc": [8, 14], "arpa_file_pars": 8, "cc": 8, "void": 8, "arpafilepars": 8, "rea": 8, "d": [8, 18, 19], "std": 8, "istream": 8, "79": 8, "140": [8, 12], "gram": [8, 9, 11, 12, 14, 16, 18, 19], "89": [8, 9], "hlg": [8, 16, 18, 19, 21], "928": 8, "compile_hlg": 8, "120": 8, "lang_phon": [8, 12, 16, 18, 19, 21], "929": [8, 11], "lexicon": [8, 9, 11, 12, 14, 21], "116": 8, "convert": [8, 14], "l": [8, 11, 18, 19, 21], "pt": [8, 9, 11, 12, 14, 16, 18, 19, 21], "linv": [8, 11, 14, 21], "931": 8, "ctc_topo": 8, "max_token_id": 8, "932": 8, "load": [8, 9, 11, 12, 14, 16, 18, 19, 21], "fst": [8, 11, 21], "intersect": 8, "933": 8, "lg": 8, "shape": 8, "66": 8, "connect": [8, 14, 16], "68": [8, 14], "70": 8, "class": [8, 14], "tensor": [8, 9, 11, 12, 14, 21], "71": [8, 14, 16], "determin": 8, "934": 8, "74": 8, "_k2": 8, "raggedint": 8, "remov": [8, 9, 11, 12, 14, 16, 18, 19], "disambigu": 8, "symbol": [8, 11], "87": 8, "remove_epsilon": 8, "935": 8, "92": [8, 14], "arc": 8, "95": [8, 10], "compos": 8, "h": 8, "105": [8, 14], "936": 8, "107": [8, 16], "123": 8, "now": [8, 9, 14, 16, 18, 19], "cuda_visible_devic": [8, 9, 11, 12, 14, 16, 18, 19, 21], "gpu": [8, 9, 11, 12, 14, 18, 19, 21], "avail": [8, 9, 11, 14, 16, 18, 19, 21], "given": [8, 9, 11, 12, 14, 16], "below": [8, 9, 11, 12, 14, 16, 18, 19, 21], "072": 8, "465": 8, "466": 8, "exp_dir": [8, 11, 14], "posixpath": [8, 11, 14], "exp": [8, 9, 11, 12, 14, 16, 18, 19, 21], "lang_dir": [8, 11, 14], "lr": [8, 11], "feature_dim": [8, 9, 11, 14, 21], "weight_decai": 8, "1e": 8, "06": [8, 12, 14, 16, 21], "start_epoch": 8, "best_train_loss": 8, "inf": 8, "best_valid_loss": 8, "best_train_epoch": 8, "best_valid_epoch": 8, "batch_idx_train": 8, "log_interv": 8, "valid_interv": 8, "beam_siz": [8, 11], "reduct": 8, "sum": 8, "use_doub": 8, "le_scor": 8, "true": [8, 9, 11, 14, 16, 18, 19, 21], "world_siz": 8, "master_port": 8, "12354": 8, "num_epoch": 8, "feature_dir": [8, 14], "max_dur": [8, 14], "bucketing_sampl": [8, 14], "num_bucket": [8, 14], "concatenate_cut": [8, 14], "duration_factor": [8, 14], "gap": [8, 14], "on_the_fly_feat": [8, 14], "shuffl": [8, 14], "return_cut": [8, 14], "num_work": [8, 14], "074": 8, "113": [8, 11, 14], "098": [8, 16], "cut": [8, 14], "240": [8, 9, 21], "149": [8, 14], "200": [8, 9, 14, 18, 19, 21], "singlecutsampl": 8, "206": [8, 14], "219": [8, 11, 14], "246": [8, 11, 14, 18, 19], "357": 8, "416": 8, "epoch": [8, 9, 11, 12, 14, 16, 18, 19, 21], "batch": [8, 9, 11, 12, 14], "avg": [8, 11, 14, 16, 18, 19, 21], "loss": [8, 9, 12, 14, 16, 18, 19, 21], "0789": 8, "848": 8, "5356": 8, "7556": 8, "301": [8, 14], "432": [8, 14], "9972": 8, "best": [8, 9, 12, 14], "805": 8, "2436": 8, "5717": 8, "33": [8, 9, 10, 11, 14, 18], "109": [8, 9, 14], "4167": 8, "121": [8, 16], "325": 8, "2214": 8, "798": [8, 11], "0781": 8, "1343": 8, "065": 8, "0859": 8, "556": 8, "0421": 8, "0975": 8, "810": 8, "0431": 8, "824": 8, "657": 8, "0109": 8, "984": [8, 14], "0093": 8, "0096": 8, "50": [8, 14, 18], "239": [8, 11], "0104": 8, "0101": 8, "569": 8, "0092": 8, "819": [8, 18], "835": 8, "51": [8, 9, 14, 21], "024": 8, "0105": 8, "317": 8, "0099": 8, "0097": 8, "552": 8, "0108": 8, "869": 8, "0102": 8, "126": [8, 14], "128": [8, 14], "537": [8, 14], "192": [8, 14], "249": 8, "250": [8, 11, 16], "lm_dir": [8, 14], "search_beam": [8, 9, 14, 21], "output_beam": [8, 9, 14, 21], "min_active_st": [8, 9, 14, 21], "max_active_st": [8, 9, 14, 21], "10000": [8, 9, 14, 21], "use_double_scor": [8, 9, 14, 21], "193": 8, "213": [8, 21], "259": [8, 9], "devic": [8, 9, 11, 12, 14, 16, 18, 19, 21], "217": [8, 9, 14], "279": [8, 14], "averag": [8, 9, 11, 12, 14, 16, 18, 19, 21], "userwarn": [8, 11], "floor_divid": 8, "futur": [8, 11, 13], "round": [8, 11], "toward": [8, 11], "trunc": [8, 11], "function": [8, 9, 11, 12, 14, 16, 18, 19, 21], "NOT": [8, 9, 11, 14, 21], "floor": [8, 11], "incorrect": [8, 11], "neg": [8, 11], "valu": [8, 9, 11, 12, 14], "keep": [8, 11], "behavior": [8, 11], "div": [8, 11], "b": [8, 11, 14, 18, 19], "rounding_mod": [8, 11], "actual": [8, 9, 11, 12, 14], "divis": [8, 11], "trigger": 8, "intern": 8, "aten": 8, "src": 8, "nativ": 8, "binaryop": 8, "cpp": 8, "450": [8, 9, 11, 12], "k": [8, 18, 19], "n": [8, 9, 14, 18, 19], "220": [8, 11, 12, 14], "409": 8, "190": [8, 16], "until": [8, 14], "571": [8, 14], "228": [8, 14], "transcript": [8, 9, 10, 11, 12, 14, 16, 18, 19], "recog": [8, 11, 14], "test_set": [8, 21], "572": 8, "util": [8, 14], "ins": [8, 14, 21], "del": [8, 14, 21], "sub": [8, 14, 21], "573": 8, "236": 8, "wrote": [8, 14], "detail": [8, 9, 11, 12, 14, 16, 18, 19, 21], "error": [8, 14], "stat": [8, 14], "err": [8, 11, 14], "299": 8, "congratul": [8, 9, 12, 14, 16, 18, 19, 21], "first": [8, 9, 11, 12, 14, 16, 18, 19, 21], "fun": 8, "debug": 8, "variou": [8, 13], "problem": 8, "mai": [8, 9, 11, 12, 13, 14], "encount": [8, 9, 11, 12, 14], "while": [8, 9, 11, 12, 14], "tutori": [9, 11, 12, 14, 16, 18, 19], "learn": [9, 11, 12, 14, 16, 18, 19, 21], "singl": [9, 11, 12, 14, 16, 18, 19, 21], "multipl": [9, 11, 12, 14, 16, 18, 19, 21], "1best": [9, 12, 14, 16, 18, 19], "handl": [9, 12, 14, 16, 18, 19, 21], "automag": [9, 12, 14, 16, 18, 19, 21], "stop": [9, 11, 12, 14, 16, 18, 19, 21], "control": [9, 11, 12, 14, 16, 18, 19, 21], "By": [9, 12, 14, 16, 18, 19, 21], "default": [9, 11, 12, 14, 16, 18, 19, 21], "execut": [9, 12, 14, 16, 18, 19, 21], "mean": [9, 11, 12, 14, 16, 18, 19, 21], "onli": [9, 11, 12, 13, 14, 16, 18, 19, 21], "musan": [9, 12, 14], "sai": [9, 11, 12, 14, 16, 18, 19, 21], "thei": [9, 11, 12, 14], "won": [9, 12, 14], "re": [9, 12, 14], "intal": [9, 12], "initi": [9, 12], "lf": [9, 11, 12, 14, 16, 18, 19, 21], "sudo": [9, 12], "apt": [9, 12], "permiss": [9, 12], "binari": [9, 11, 12, 14, 21], "here": [9, 11, 12, 14, 16], "pass": [9, 11, 12, 14], "commandlin": [9, 11, 12, 14], "quit": [9, 11, 12, 14], "often": [9, 11, 12, 14], "dir": [9, 11, 12, 14, 16, 18, 19, 21], "experi": [9, 11, 12, 14, 21], "num": [9, 11, 12, 14], "resum": [9, 11, 12, 14, 16, 18, 19, 21], "state": [9, 11, 12, 14], "world": [9, 11, 12, 14, 16], "multi": [9, 11, 12, 14], "machin": [9, 11, 12, 14], "ddp": [9, 11, 12, 14], "4": [9, 11, 12, 14, 16, 18, 19, 21], "implement": [9, 11, 12, 14], "present": [9, 11, 12, 14], "later": [9, 12, 14, 16, 18, 19], "max": [9, 11, 12, 14], "durat": [9, 11, 12, 14, 16, 18, 19, 21], "specifi": [9, 11, 12, 14, 16, 18, 19, 21], "second": [9, 11, 12, 14, 21], "over": [9, 11, 12, 14], "utter": [9, 11, 12, 14], "pad": [9, 11, 12, 14], "oom": [9, 11, 12, 14], "reduc": [9, 11, 12, 14, 16, 18, 19, 21], "v100": [9, 11, 12, 14], "nvidia": [9, 11, 12, 14], "due": [9, 11, 12, 14], "usual": [9, 11, 12, 14], "larger": [9, 11, 12, 14], "than": [9, 11, 12, 14, 16, 21], "caus": [9, 11, 12, 14], "smaller": [9, 11, 12, 14], "increas": [9, 11, 12, 14], "tune": [9, 11, 12, 14], "weight": [9, 12, 14], "decai": [9, 12, 14], "warmup": [9, 11, 12, 14], "get_param": [9, 11, 12, 14, 16, 18, 19, 21], "paramet": [9, 11, 12, 14, 16, 18, 19], "realli": [9, 12, 14], "directli": [9, 11, 12, 14], "perturb": [9, 11, 12, 14], "speed": [9, 11, 12, 14], "factor": [9, 11, 12, 14], "each": [9, 11, 12, 14], "3x150": [9, 11, 12], "hour": [9, 11, 12, 14], "These": [9, 11, 12, 14, 16, 18, 19, 21], "state_dict": [9, 11, 12, 14, 16, 18, 19, 21], "optim": [9, 11, 12, 14, 16, 18, 19, 21], "rate": [9, 11, 12, 14, 16, 18, 19, 21], "visual": [9, 11, 12, 14, 16, 18, 19, 21], "logdir": [9, 11, 12, 14, 16, 18, 19, 21], "labelsmooth": 9, "someth": [9, 11, 12, 14, 21], "tensorflow": [9, 11, 12, 14, 21], "found": [9, 11, 12, 14, 21], "continu": [9, 11, 12, 14, 21], "press": [9, 11, 12, 14, 21], "ctrl": [9, 11, 12, 14, 21], "engw8ksktzqs24zbv5dgcg": 9, "22t11": 9, "09": [9, 11, 12, 14], "scan": [9, 11, 12, 14, 21], "116068": 9, "scalar": [9, 11, 12, 14, 21], "listen": [9, 11, 12, 21], "note": [9, 11, 12, 14, 16, 18, 19, 21], "url": [9, 11, 12, 14, 21], "output": [9, 11, 12, 14, 16, 18, 19, 21], "xxxx": [9, 11, 12, 14, 16, 18, 19, 21], "text": [9, 11, 12, 14, 16, 18, 19, 21], "saw": [9, 11, 12, 14, 16, 18, 19, 21], "consol": [9, 11, 12, 14, 16, 18, 19, 21], "typic": [9, 11, 12, 14], "avoid": [9, 11, 14], "commonli": [9, 11, 12, 14, 16, 18, 19, 21], "nbest": [9, 14], "scale": [9, 14, 16, 18, 19], "down": [9, 14], "lattic": [9, 12, 14, 16, 18, 19], "score": [9, 14], "more": [9, 14, 21], "uniqu": [9, 14], "pkufool": [9, 12, 16], "icefall_asr_aishell_conformer_ctc": 9, "transcrib": [9, 11, 12, 14], "sound": [9, 11, 12, 14, 16, 18, 19, 21], "csukuangfj": [9, 11, 12, 14, 16, 18, 19, 21], "tree": [9, 11, 12, 14, 16, 18, 19, 21], "lang_char": [9, 11], "token": [9, 11, 12, 14, 16, 18, 19, 21], "word": [9, 11, 12, 14, 16, 18, 19, 21], "test_wav": [9, 11, 12, 14, 16, 18, 19, 21], "bac009s0764w0121": [9, 11, 12], "wav": [9, 11, 12, 14, 18, 19, 21], "bac009s0764w0122": [9, 11, 12], "bac009s0764w0123": [9, 11, 12], "tran": [9, 12, 14, 16, 18, 19], "graph": [9, 12, 14, 16, 18, 19], "id": [9, 12, 14, 16, 18, 19], "conveni": [9, 12, 14], "eo": [9, 12, 14], "easili": [9, 12, 14], "obtain": [9, 11, 12, 14, 16, 18, 19], "25": [9, 14, 18, 19, 21], "84": 9, "list": [9, 11, 12, 14, 16, 18, 19], "soxi": [9, 11, 12, 14, 16, 21], "sampl": [9, 11, 12, 14, 16, 21], "16000": [9, 11, 12, 14, 16, 18, 19], "precis": [9, 11, 12, 14, 16, 21], "bit": [9, 11, 12, 14, 16, 21], "04": [9, 11, 12, 14, 16, 18, 19], "67263": [9, 11, 12], "315": [9, 11, 12, 14, 16], "295": [9, 11, 12, 14], "cdda": [9, 11, 12, 14, 16, 21], "sector": [9, 11, 12, 14, 16, 21], "135k": [9, 11, 12], "256k": [9, 11, 12, 14], "encod": [9, 11, 12, 14, 16, 21], "sign": [9, 11, 12, 14, 21], "integ": [9, 11, 12, 14, 21], "pcm": [9, 11, 12, 14, 21], "65840": [9, 11, 12], "625": [9, 11, 12], "132k": [9, 11, 12], "64000": [9, 11, 12], "300": [9, 11, 12, 14], "128k": [9, 11, 12, 21], "displai": [9, 11, 12, 14], "support": [9, 11, 14], "three": [9, 11], "topologi": [9, 14], "07": [9, 11, 12, 14], "53": [9, 16, 19], "707": [9, 14], "229": 9, "sample_r": [9, 11, 14, 21], "subsampling_factor": [9, 11, 14], "nhead": [9, 11, 14], "attention_dim": [9, 11, 14], "512": [9, 11, 14], "num_decoder_lay": [9, 14], "vgg_frontend": [9, 11, 14], "use_feat_batchnorm": [9, 14], "env_info": [9, 11, 14], "releas": [9, 11, 14], "sha1": [9, 11, 14], "f2fd997f752ed11bbef4c306652c433e83f9cf12": 9, "date": [9, 11, 14], "sun": 9, "sep": 9, "46": [9, 14], "33cfe45": 9, "branch": [9, 11, 14], "d57a873": 9, "dirti": [9, 14], "wed": [9, 11, 14], "nov": [9, 14], "hw": 9, "kangwei": 9, "icefall_aishell3": 9, "k2_releas": 9, "__init__": [9, 11, 14], "tokens_fil": 9, "words_fil": [9, 14, 21], "num_path": [9, 14], "ngram_lm_scal": [9, 14], "attention_decoder_scal": [9, 14], "nbest_scal": [9, 14], "sos_id": [9, 14], "eos_id": [9, 14], "num_class": [9, 14, 21], "4336": [9, 11], "sound_fil": [9, 11, 14, 21], "708": [9, 11, 14, 21], "242": [9, 14], "131": [9, 14], "construct": [9, 11, 12, 14, 16, 18, 19, 21], "134": 9, "269": [9, 18, 19], "275": 9, "241": 9, "293": [9, 14], "704": [9, 18], "369": [9, 14], "\u751a": [9, 11], "\u81f3": [9, 11], "\u51fa": [9, 11], "\u73b0": [9, 11], "\u4ea4": [9, 11], "\u6613": [9, 11], "\u51e0": [9, 11], "\u4e4e": [9, 11], "\u505c": [9, 11], "\u6b62": 9, "\u7684": [9, 11, 12], "\u60c5": [9, 11], "\u51b5": [9, 11], "\u4e00": [9, 11], "\u4e8c": [9, 11], "\u7ebf": [9, 11, 12], "\u57ce": [9, 11], "\u5e02": [9, 11], "\u867d": [9, 11], "\u7136": [9, 11], "\u4e5f": [9, 11, 12], "\u5904": [9, 11], "\u4e8e": [9, 11], "\u8c03": [9, 11], "\u6574": [9, 11], "\u4e2d": [9, 11, 12], "\u4f46": [9, 11, 12], "\u56e0": [9, 11], "\u4e3a": [9, 11], "\u805a": [9, 11], "\u96c6": [9, 11], "\u4e86": [9, 11, 12], "\u8fc7": [9, 11], "\u591a": [9, 11], "\u516c": [9, 11], "\u5171": [9, 11], "\u8d44": [9, 11], "\u6e90": [9, 11], "371": 9, "37": [9, 11, 14, 18], "38": [9, 11, 14, 18], "683": 9, "47": [9, 14], "651": [9, 21], "654": 9, "659": 9, "752": 9, "321": 9, "887": 9, "340": 9, "370": 9, "\u751a\u81f3": [9, 12], "\u51fa\u73b0": [9, 12], "\u4ea4\u6613": [9, 12], "\u51e0\u4e4e": [9, 12], "\u505c\u6b62": 9, "\u60c5\u51b5": [9, 12], "\u4e00\u4e8c": [9, 12], "\u57ce\u5e02": [9, 12], "\u867d\u7136": [9, 12], "\u5904\u4e8e": [9, 12], "\u8c03\u6574": [9, 12], "\u56e0\u4e3a": [9, 12], "\u805a\u96c6": [9, 12], "\u8fc7\u591a": [9, 12], "\u516c\u5171": [9, 12], "\u8d44\u6e90": [9, 12], "372": 9, "recor": [9, 14], "highest": [9, 14], "05": [9, 11, 12, 14, 19], "965": 9, "966": 9, "821": 9, "822": 9, "826": 9, "916": 9, "115": [9, 14], "345": 9, "888": 9, "889": 9, "limit": [9, 11, 14], "memori": [9, 11, 14], "upgrad": [9, 14], "pro": [9, 14], "finish": [9, 11, 12, 14, 16, 18, 19, 21], "deploi": [9, 14], "At": [9, 14], "doe": [9, 11, 14, 21], "stream": [9, 14, 18, 19], "home": [9, 14], "checkout": [9, 14], "v2": [9, 14], "cmake": [9, 14], "dcmake_build_typ": [9, 14], "j": [9, 14], "hlg_decod": [9, 14], "four": [9, 14], "ln": [9, 14], "messag": [9, 14], "nn_model": [9, 14], "use_gpu": [9, 14], "word_tabl": [9, 14], "wave": [9, 14], "caution": [9, 14], "relat": [9, 11, 14, 16, 18, 19, 21], "forward": [9, 14], "cpu_jit": [9, 14], "cu": [9, 14], "int": [9, 14], "char": [9, 14], "124": [9, 14], "98": 9, "142": [9, 12, 14], "150": [9, 14], "693": [9, 18], "165": [9, 14], "nnet_output": [9, 14], "182": [9, 16], "180": [9, 14], "489": 9, "45": [9, 11, 14], "216": [9, 14, 18, 19], "mandarin": 10, "corpu": 10, "beij": 10, "shell": 10, "technologi": 10, "ltd": 10, "400": 10, "peopl": 10, "accent": 10, "area": 10, "china": 10, "invit": 10, "particip": 10, "conduct": 10, "quiet": 10, "indoor": 10, "high": 10, "fidel": 10, "microphon": 10, "downsampl": 10, "16khz": 10, "manual": 10, "accuraci": 10, "through": 10, "profession": 10, "annot": 10, "strict": 10, "inspect": 10, "free": 10, "academ": 10, "hope": [10, 14, 16], "moder": 10, "amount": 10, "research": 10, "field": 10, "openslr": 10, "ctc": [10, 13, 15, 17, 20], "stateless": [10, 13], "transduc": [10, 13], "instead": 11, "rnn": 11, "As": [11, 14], "head": 11, "dim": 11, "layer": 11, "feedforward": 11, "2048": 11, "embed": 11, "conv1d": 11, "kernel": 11, "left": 11, "context": 11, "joiner": 11, "nn": 11, "tanh": 11, "linear": 11, "borrow": 11, "ieeexplor": 11, "ieee": 11, "stamp": 11, "jsp": 11, "arnumb": 11, "9054419": 11, "predict": 11, "modif": 11, "right": 11, "charact": 11, "unit": 11, "whose": [11, 14, 16], "vocabulari": 11, "87939824": 11, "88": 11, "optimized_transduc": 11, "extra": 11, "technqiu": 11, "propos": 11, "improv": 11, "end": [11, 21], "furthermor": 11, "maximum": 11, "emit": 11, "per": 11, "frame": 11, "simplifi": 11, "significantli": 11, "degrad": 11, "perform": 11, "exactli": 11, "benchmark": 11, "unprun": 11, "advantag": 11, "minim": 11, "pruned_transducer_stateless": 11, "prune": 11, "other": [11, 13, 14, 16, 18, 19, 21], "altern": 11, "though": 11, "transducer_stateless_modifi": 11, "option": [11, 16, 18, 19, 21], "pr": 11, "gb": 11, "ram": 11, "small": [11, 18, 19, 21], "tri": 11, "prob": 11, "appli": 11, "configur": [11, 16, 18, 19, 21], "c": [11, 12, 21], "lagz6hrcqxoigbfd5e0y3q": 11, "2022": 11, "03t14": 11, "8477": 11, "greedy_search": 11, "sym": 11, "beam_search": 11, "decoding_method": 11, "beam_4": 11, "02": [11, 14, 19], "28": [11, 14, 16], "ensur": 11, "give": 11, "poor": 11, "531": [11, 12], "994": [11, 14], "176": [11, 14], "027": 11, "encoder_out_dim": 11, "dim_feedforward": 11, "num_encoder_lay": 11, "f4fefe4882bc0ae59af951da3f47335d5495ef71": 11, "thu": [11, 14, 16], "feb": 11, "miss": [11, 14], "50d2281": 11, "mar": 11, "hostnam": 11, "de": 11, "74279": 11, "0815224919": 11, "75d558775b": 11, "mmnv8": 11, "ip": 11, "177": [11, 12, 14], "72": [11, 14], "context_s": 11, "max_sym_per_fram": 11, "blank_id": 11, "vocab_s": 11, "878": [11, 19], "257": [11, 18, 19], "880": 11, "267": [11, 18, 19], "891": 11, "273": 11, "__floordiv__": 11, "length": 11, "x_len": 11, "163": [11, 14], "320": 11, "\u6ede": 11, "322": 11, "285": [11, 14], "759": 11, "760": 11, "919": 11, "922": 11, "046": 11, "047": 11, "319": [11, 14], "214": [11, 14], "215": [11, 14, 16], "402": 11, "topk_hyp_index": 11, "topk_index": 11, "logit": 11, "583": [11, 19], "2000": 12, "lji9mwuorlow3jkdhxwk8a": 12, "13t11": 12, "4454": 12, "icefall_asr_aishell_tdnn_lstm_ctc": 12, "858": [12, 14], "389": [12, 14], "154": 12, "161": [12, 14], "536": 12, "171": [12, 14, 18, 19], "539": 12, "917": 12, "207": [12, 14], "129": 12, "\u505c\u6ede": 12, "222": [12, 14], "task": 13, "well": [13, 21], "ligru": [13, 17], "full": 14, "libri": 14, "960": 14, "subset": 14, "3x960": 14, "2880": 14, "lzgnetjwrxc3yghnmd4kpw": 14, "24t16": 14, "43": 14, "4540": 14, "sentenc": 14, "piec": 14, "And": 14, "neither": 14, "nor": 14, "vocab": 14, "work": 14, "5000": 14, "lang_bpe_500": 14, "44": [14, 18, 19], "033": 14, "538": 14, "full_libri": 14, "406": 14, "464": 14, "548": 14, "776": 14, "652": [14, 21], "109226120": 14, "714": [14, 18], "473": 14, "944": 14, "1328": 14, "54": [14, 16, 18, 19], "443": [14, 16], "2563": 14, "56": [14, 18], "494": 14, "592": 14, "331": [14, 16], "1715": 14, "52576": 14, "1424": 14, "807": 14, "506": 14, "808": [14, 18], "522": 14, "362": 14, "203": 14, "565": 14, "1477": 14, "106": 14, "2922": 14, "208": 14, "4295": 14, "52343": 14, "396": 14, "3584": 14, "433": 14, "680": [14, 18], "jit": 14, "bpe": 14, "_pickl": 14, "unpicklingerror": 14, "kei": 14, "hlg_modifi": 14, "g_4_gram": [14, 16, 18, 19], "1089": [14, 16], "134686": [14, 16], "0001": [14, 16], "1221": [14, 16], "135766": [14, 16], "0002": [14, 16], "reproduc": 14, "our": 14, "106000": [14, 16], "496": [14, 16], "875": [14, 16], "212k": 14, "267440": [14, 16], "1253": [14, 16], "535k": 14, "83": [14, 16], "77200": [14, 16], "361": [14, 16], "154k": 14, "554": 14, "260": 14, "bpe_model": 14, "7178d67e594bc7fa89c2b331ad7bd1c62a6a9eb4": 14, "tue": 14, "oct": 14, "22": [14, 18, 19, 21], "8d93169": 14, "266": [14, 16], "268": [14, 16], "600": 14, "601": 14, "758": 14, "025": 14, "204": 14, "425": 14, "earli": [14, 16], "nightfal": [14, 16], "THE": [14, 16], "yellow": [14, 16], "lamp": [14, 16], "light": [14, 16], "AND": [14, 16], "THERE": [14, 16], "squalid": [14, 16], "quarter": [14, 16], "OF": [14, 16], "broffel": 14, "god": [14, 16], "AS": [14, 16], "direct": [14, 16], "consequ": [14, 16], "sin": [14, 16], "man": [14, 16], "punish": [14, 16], "had": [14, 16], "her": [14, 16], "love": [14, 16], "child": [14, 16], "ON": [14, 16], "THAT": [14, 16], "dishonor": [14, 16], "osom": 14, "TO": [14, 16], "parent": [14, 16], "forev": [14, 16], "WITH": [14, 16], "race": [14, 16], "descent": [14, 16], "mortal": [14, 16], "BE": [14, 16], "bless": [14, 16], "soul": [14, 16], "IN": [14, 16], "heaven": [14, 16], "yet": [14, 16], "THESE": [14, 16], "thought": [14, 16], "affect": [14, 16], "hester": [14, 16], "prynn": [14, 16], "less": [14, 16, 21], "apprehens": [14, 16], "427": 14, "723": 14, "775": 14, "881": 14, "951": 14, "352": 14, "234": 14, "384": 14, "brothel": [14, 16], "bosom": [14, 16], "whole": [14, 16, 18, 19], "ngram": [14, 16, 18, 19], "Its": 14, "857": 14, "979": 14, "980": 14, "055": 14, "117": 14, "051": 14, "363": 14, "959": [14, 19], "546": 14, "598": 14, "599": [14, 16], "833": 14, "834": 14, "915": 14, "076": 14, "110": 14, "329": 14, "397": 14, "999": 14, "concaten": 14, "bucket": 14, "sampler": 14, "1000": 14, "ctc_decod": 14, "ngram_lm_rescor": 14, "attention_rescor": 14, "kind": 14, "316": 14, "118": 14, "58": 14, "221": 14, "125": [14, 21], "136": 14, "144": 14, "159": [14, 21], "543": 14, "174": 14, "topo": 14, "547": 14, "729": 14, "111": 14, "702": 14, "703": 14, "545": 14, "122": 14, "280": 14, "135": [14, 21], "153": [14, 21], "945": 14, "475": 14, "191": [14, 18, 19], "398": 14, "199": [14, 16], "515": 14, "205": 14, "w": [14, 18, 19], "deseri": 14, "441": 14, "fsaclass": 14, "loadfsa": 14, "const": 14, "string": 14, "c10": 14, "ignor": 14, "non": 14, "attribut": 14, "dummi": 14, "589": 14, "attention_scal": 14, "656": 14, "162": 14, "169": [14, 18, 19], "188": 14, "624": 14, "519": [14, 19], "632": 14, "645": [14, 21], "243": 14, "970": 14, "303": 14, "179": 14, "icefall_asr_librispeech_tdnn": 16, "lstm_ctc": 16, "flac": 16, "116k": 16, "140k": 16, "343k": 16, "164k": 16, "105k": 16, "174k": 16, "usag": [16, 18, 19, 21], "pretraind": 16, "168": 16, "170": 16, "581": 16, "584": [16, 19], "209": 16, "791": 16, "245": 16, "099": 16, "methond": [16, 18, 19], "725": 16, "403": 16, "631": 16, "010": 16, "tdnn_ligru_ctc": 18, "enough": [18, 19, 21], "luomingshuang": [18, 19], "icefall_asr_timit_tdnn_ligru_ctc": 18, "pretrained_average_9_25": 18, "fdhc0_si1559": [18, 19], "felc0_si756": [18, 19], "fmgd0_si1564": [18, 19], "ffprobe": [18, 19], "show_format": [18, 19], "nistspher": [18, 19], "database_id": [18, 19], "database_vers": [18, 19], "utterance_id": [18, 19], "dhc0_si1559": [18, 19], "sample_min": [18, 19], "4176": [18, 19], "sample_max": [18, 19], "5984": [18, 19], "bitrat": [18, 19], "258": [18, 19], "audio": [18, 19], "pcm_s16le": [18, 19], "hz": [18, 19], "s16": [18, 19], "256": [18, 19], "elc0_si756": [18, 19], "1546": [18, 19], "1989": [18, 19], "mgd0_si1564": [18, 19], "7626": [18, 19], "10573": [18, 19], "660": 18, "183": [18, 19], "695": 18, "697": 18, "210": [18, 19], "829": 18, "sil": [18, 19], "dh": [18, 19], "ih": [18, 19], "uw": [18, 19], "ah": [18, 19], "ii": [18, 19], "z": [18, 19], "aa": [18, 19], "ei": [18, 19], "dx": [18, 19], "uh": [18, 19], "ng": [18, 19], "th": [18, 19], "eh": [18, 19], "jh": [18, 19], "er": [18, 19], "ai": [18, 19], "hh": [18, 19], "aw": 18, "ae": [18, 19], "705": 18, "715": 18, "720": 18, "251": [18, 19], "348": 18, "ch": 18, "icefall_asr_timit_tdnn_lstm_ctc": 19, "pretrained_average_16_25": 19, "816": 19, "827": 19, "387": 19, "unk": 19, "739": 19, "971": 19, "977": 19, "978": 19, "981": 19, "ow": 19, "ykubhb5wrmosxykid1z9eg": 21, "23t23": 21, "sinc": 21, "icefall_asr_yesno_tdnn": 21, "l_disambig": 21, "lexicon_disambig": 21, "arpa": 21, "0_0_0_1_0_0_0_1": 21, "0_0_1_0_0_0_1_0": 21, "0_0_1_0_0_1_1_1": 21, "0_0_1_0_1_0_0_1": 21, "0_0_1_1_0_0_0_1": 21, "0_0_1_1_0_1_1_0": 21, "0_0_1_1_1_0_0_0": 21, "0_0_1_1_1_1_0_0": 21, "0_1_0_0_0_1_0_0": 21, "0_1_0_0_1_0_1_0": 21, "0_1_0_1_0_0_0_0": 21, "0_1_0_1_1_1_0_0": 21, "0_1_1_0_0_1_1_1": 21, "0_1_1_1_0_0_1_0": 21, "0_1_1_1_1_0_1_0": 21, "1_0_0_0_0_0_0_0": 21, "1_0_0_0_0_0_1_1": 21, "1_0_0_1_0_1_1_1": 21, "1_0_1_1_0_1_1_1": 21, "1_0_1_1_1_1_0_1": 21, "1_1_0_0_0_1_1_1": 21, "1_1_0_0_1_0_1_1": 21, "1_1_0_1_0_1_0_0": 21, "1_1_0_1_1_0_0_1": 21, "1_1_0_1_1_1_1_0": 21, "1_1_1_0_0_1_0_1": 21, "1_1_1_0_1_0_1_0": 21, "1_1_1_1_0_0_1_0": 21, "1_1_1_1_1_0_0_0": 21, "1_1_1_1_1_1_1_1": 21, "54080": 21, "507": 21, "108k": 21, "No": 21, "ye": 21, "hebrew": 21, "NO": 21, "621": 21, "119": 21, "127": 21, "650": 21, "139": 21, "143": 21, "198": 21, "181": 21, "186": 21, "187": 21, "287": 21, "correctli": 21, "simplest": 21}, "objects": {}, "objtypes": {}, "objnames": {}, "titleterms": {"follow": 0, "code": 0, "style": 0, "contribut": [1, 3], "document": 1, "how": 2, "creat": [2, 8], "recip": [2, 13], "data": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "prepar": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "train": [2, 5, 8, 9, 11, 12, 14, 16, 18, 19, 21], "decod": [2, 8, 9, 11, 12, 14, 16, 18, 19, 21], "pre": [2, 5, 9, 11, 12, 14, 16, 18, 19, 21], "model": [2, 5, 9, 11, 12, 14, 16, 18, 19, 21], "huggingfac": [4, 6], "space": 6, "youtub": [6, 8], "video": [6, 8], "icefal": [7, 8], "content": [7, 13], "instal": [8, 9, 11, 12, 14, 16, 18, 19], "0": 8, "pytorch": 8, "torchaudio": 8, "1": [8, 9, 11, 12, 14], "k2": 8, "2": [8, 9, 11, 12, 14], "lhots": 8, "3": [8, 9, 11, 14], "download": [8, 9, 11, 12, 14, 16, 18, 19, 21], "exampl": [8, 9, 11, 12, 14], "virtual": 8, "environ": 8, "activ": 8, "your": 8, "4": 8, "5": 8, "test": 8, "conform": [9, 14], "ctc": [9, 12, 14, 16, 18, 19, 21], "configur": [9, 12, 14], "option": [9, 12, 14], "log": [9, 11, 12, 14], "usag": [9, 11, 12, 14], "case": [9, 11, 12, 14], "kaldifeat": [9, 11, 12, 14, 16, 18, 19, 21], "hlg": [9, 12, 14], "attent": [9, 14], "rescor": [9, 14], "colab": [9, 11, 12, 14, 16, 18, 19, 21], "notebook": [9, 11, 12, 14, 16, 18, 19, 21], "deploy": [9, 14], "c": [9, 14], "aishel": 10, "stateless": 11, "transduc": 11, "The": 11, "loss": 11, "todo": 11, "greedi": 11, "search": 11, "beam": 11, "modifi": 11, "tdnn": [12, 16, 18, 19, 21], "lstm": [12, 16, 19], "tabl": 13, "lm": 14, "comput": 14, "wer": 14, "librispeech": 15, "infer": [16, 18, 19, 21], "timit": 17, "ligru": 18, "yesno": 20}, "envversion": {"sphinx.domains.c": 2, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 6, "sphinx.domains.index": 1, "sphinx.domains.javascript": 2, "sphinx.domains.math": 2, "sphinx.domains.python": 3, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx.ext.todo": 2, "sphinx": 56}})