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", "model-export/export-model-state-dict", "model-export/export-ncnn", "model-export/export-onnx", "model-export/export-with-torch-jit-script", "model-export/export-with-torch-jit-trace", "model-export/index", "recipes/Non-streaming-ASR/aishell/conformer_ctc", "recipes/Non-streaming-ASR/aishell/index", "recipes/Non-streaming-ASR/aishell/stateless_transducer", "recipes/Non-streaming-ASR/aishell/tdnn_lstm_ctc", "recipes/Non-streaming-ASR/index", "recipes/Non-streaming-ASR/librispeech/conformer_ctc", "recipes/Non-streaming-ASR/librispeech/index", "recipes/Non-streaming-ASR/librispeech/pruned_transducer_stateless", "recipes/Non-streaming-ASR/librispeech/tdnn_lstm_ctc", "recipes/Non-streaming-ASR/librispeech/zipformer_ctc_blankskip", "recipes/Non-streaming-ASR/librispeech/zipformer_mmi", "recipes/Non-streaming-ASR/timit/index", "recipes/Non-streaming-ASR/timit/tdnn_ligru_ctc", "recipes/Non-streaming-ASR/timit/tdnn_lstm_ctc", "recipes/Non-streaming-ASR/yesno/index", "recipes/Non-streaming-ASR/yesno/tdnn", "recipes/Streaming-ASR/index", "recipes/Streaming-ASR/introduction", "recipes/Streaming-ASR/librispeech/index", "recipes/Streaming-ASR/librispeech/lstm_pruned_stateless_transducer", "recipes/Streaming-ASR/librispeech/pruned_transducer_stateless", "recipes/Streaming-ASR/librispeech/zipformer_transducer", "recipes/index"], "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", "model-export/export-model-state-dict.rst", "model-export/export-ncnn.rst", "model-export/export-onnx.rst", "model-export/export-with-torch-jit-script.rst", "model-export/export-with-torch-jit-trace.rst", "model-export/index.rst", "recipes/Non-streaming-ASR/aishell/conformer_ctc.rst", "recipes/Non-streaming-ASR/aishell/index.rst", "recipes/Non-streaming-ASR/aishell/stateless_transducer.rst", "recipes/Non-streaming-ASR/aishell/tdnn_lstm_ctc.rst", "recipes/Non-streaming-ASR/index.rst", "recipes/Non-streaming-ASR/librispeech/conformer_ctc.rst", "recipes/Non-streaming-ASR/librispeech/index.rst", "recipes/Non-streaming-ASR/librispeech/pruned_transducer_stateless.rst", "recipes/Non-streaming-ASR/librispeech/tdnn_lstm_ctc.rst", "recipes/Non-streaming-ASR/librispeech/zipformer_ctc_blankskip.rst", "recipes/Non-streaming-ASR/librispeech/zipformer_mmi.rst", "recipes/Non-streaming-ASR/timit/index.rst", "recipes/Non-streaming-ASR/timit/tdnn_ligru_ctc.rst", "recipes/Non-streaming-ASR/timit/tdnn_lstm_ctc.rst", "recipes/Non-streaming-ASR/yesno/index.rst", "recipes/Non-streaming-ASR/yesno/tdnn.rst", "recipes/Streaming-ASR/index.rst", "recipes/Streaming-ASR/introduction.rst", "recipes/Streaming-ASR/librispeech/index.rst", "recipes/Streaming-ASR/librispeech/lstm_pruned_stateless_transducer.rst", "recipes/Streaming-ASR/librispeech/pruned_transducer_stateless.rst", "recipes/Streaming-ASR/librispeech/zipformer_transducer.rst", "recipes/index.rst"], "titles": ["Follow the code style", "Contributing to Documentation", "How to create a recipe", "Contributing", "Huggingface", "Pre-trained models", "Huggingface spaces", "Icefall", "Installation", "Export model.state_dict()", "Export to ncnn", "Export to ONNX", "Export model with torch.jit.script()", "Export model with torch.jit.trace()", "Model export", "Conformer CTC", "aishell", "Stateless Transducer", "TDNN-LSTM CTC", "Non Streaming ASR", "Conformer CTC", "LibriSpeech", "Pruned transducer statelessX", "TDNN-LSTM-CTC", "Zipformer CTC Blank Skip", "Zipformer MMI", "TIMIT", "TDNN-LiGRU-CTC", "TDNN-LSTM-CTC", "YesNo", "TDNN-CTC", "Streaming ASR", "Introduction", "LibriSpeech", "LSTM Transducer", "Pruned transducer statelessX", "Zipformer Transducer", "Recipes"], "terms": {"we": [0, 1, 2, 3, 5, 6, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36, 37], "us": [0, 1, 2, 4, 6, 7, 8, 10, 14, 15, 16, 17, 18, 20, 23, 27, 28, 30, 32], "tool": [0, 34], "make": [0, 1, 3, 15, 17, 20, 32, 34], "consist": [0, 17, 22, 34, 35, 36], "possibl": [0, 2, 3, 8, 15, 20], "black": 0, "format": [0, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "flake8": 0, "check": [0, 20], "qualiti": [0, 16], "isort": 0, "sort": [0, 8], "import": [0, 35, 36], "The": [0, 1, 2, 6, 8, 9, 10, 15, 16, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "version": [0, 8, 9, 15, 17, 18, 20, 22, 23, 27, 28, 34, 35], "abov": [0, 8, 9, 15, 16, 17, 18, 20, 22, 24, 25, 30, 32, 34, 35, 36], "ar": [0, 1, 3, 8, 9, 11, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36, 37], "22": [0, 20, 27, 28, 30], "3": [0, 7, 9, 18, 22, 23, 24, 25, 30, 34, 35, 36], "0": [0, 1, 7, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "5": [0, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "4": [0, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "10": [0, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "1": [0, 7, 9, 11, 12, 13, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "after": [0, 1, 6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "run": [0, 2, 6, 8, 11, 14, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "command": [0, 1, 8, 9, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "git": [0, 8, 9, 15, 17, 18, 20, 23, 27, 28, 30, 34], "clone": [0, 8, 9, 15, 17, 18, 20, 23, 27, 28, 30, 34], "http": [0, 1, 2, 5, 6, 8, 9, 10, 11, 12, 13, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "github": [0, 2, 5, 8, 9, 10, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "com": [0, 2, 5, 6, 8, 9, 10, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "k2": [0, 2, 5, 6, 7, 9, 10, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "fsa": [0, 2, 5, 6, 8, 9, 10, 11, 12, 13, 15, 17, 20, 22, 24, 25, 34, 35, 36], "icefal": [0, 2, 3, 5, 6, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36, 37], "cd": [0, 1, 2, 8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "pip": [0, 1, 8, 17, 34], "instal": [0, 1, 4, 6, 7, 9, 22, 24, 25, 30, 34, 35, 36], "pre": [0, 3, 4, 6, 7, 8], "commit": 0, "whenev": 0, "you": [0, 1, 2, 5, 6, 8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "automat": [0, 6], "hook": 0, "invok": 0, "fail": [0, 8], "If": [0, 2, 6, 8, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "ani": [0, 8, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "your": [0, 1, 2, 4, 6, 7, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "wa": [0, 8, 9, 20, 23], "success": [0, 8], "pleas": [0, 1, 2, 6, 8, 10, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "fix": [0, 8, 20], "issu": [0, 8, 20, 35, 36], "report": [0, 8], "some": [0, 1, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "i": [0, 1, 2, 6, 8, 9, 10, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "e": [0, 2, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "modifi": [0, 15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "file": [0, 2, 6, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "place": [0, 8, 9, 17, 20, 23], "so": [0, 6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "statu": 0, "failur": 0, "see": [0, 1, 6, 8, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "which": [0, 2, 6, 9, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 35, 36], "ha": [0, 2, 10, 11, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 32, 34, 35, 36], "been": [0, 10, 11, 17], "befor": [0, 1, 9, 12, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "further": 0, "chang": [0, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "all": [0, 5, 6, 9, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "again": [0, 30], "should": [0, 2, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "succe": 0, "thi": [0, 2, 3, 4, 8, 9, 11, 12, 13, 14, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36, 37], "time": [0, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "succeed": 0, "want": [0, 8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "can": [0, 1, 2, 5, 6, 8, 9, 10, 11, 12, 13, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "do": [0, 2, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "Or": 0, "without": [0, 4, 6, 15, 20, 34], "your_changed_fil": 0, "py": [0, 2, 8, 11, 12, 13, 14, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "sphinx": 1, "write": [1, 2, 3], "have": [1, 2, 5, 6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "prepar": [1, 3, 9], "environ": [1, 15, 16, 17, 18, 20, 22, 23, 27, 28, 30, 34, 35, 36], "doc": [1, 9], "r": [1, 8, 27, 28], "requir": [1, 8, 35, 36], "txt": [1, 8, 15, 17, 18, 20, 23, 27, 28, 30], "set": [1, 8, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "up": [1, 8, 9, 15, 18, 20, 22, 23, 24, 25, 35, 36], "readi": [1, 15, 20], "refer": [1, 2, 8, 9, 10, 12, 13, 15, 17, 18, 20, 22, 23, 24, 27, 28, 30, 32, 35, 36], "restructuredtext": 1, "primer": 1, "familiar": 1, "build": [1, 8, 9, 15, 17, 20, 34], "local": [1, 8, 22, 24, 25, 34, 35, 36], "preview": 1, "what": [1, 2, 8, 17, 32], "look": [1, 2, 5, 8, 15, 17, 18, 20], "like": [1, 2, 6, 8, 15, 17, 18, 20, 22, 24, 25, 30, 32, 34, 35], "publish": [1, 9, 16], "html": [1, 2, 8, 12, 13, 22, 34, 35, 36], "gener": [1, 9, 11, 12, 13, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "view": [1, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "follow": [1, 2, 3, 5, 6, 7, 8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "python3": [1, 8, 34], "m": [1, 17, 22, 24, 25, 27, 28, 34, 35, 36], "server": [1, 6, 8, 34], "It": [1, 2, 4, 8, 10, 11, 12, 13, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "print": [1, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "serv": [1, 22, 24, 25, 34, 35, 36], "port": [1, 22, 24, 25, 34, 35, 36], "8000": [1, 30], "open": [1, 9, 16, 17, 20], "browser": [1, 4, 6, 22, 24, 25, 34, 35, 36], "go": [1, 15, 17, 20, 22, 24, 25, 34, 35, 36], "read": [2, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "code": [2, 3, 7, 15, 20, 22, 23, 27, 28, 30, 35, 36], "style": [2, 3, 7], "adjust": 2, "sytl": 2, "design": 2, "python": [2, 8, 9, 12, 13, 15, 17, 20, 22, 24, 25, 34, 35, 36], "recommend": [2, 8, 15, 17, 18, 20, 22, 35, 36], "test": [2, 7, 9, 10, 11, 15, 17, 18, 20, 23, 24, 27, 28], "valid": [2, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "dataset": [2, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "lhots": [2, 7, 9, 15, 17, 20], "readthedoc": [2, 8], "io": [2, 8, 12, 13, 22, 34, 35, 36], "en": [2, 8], "latest": [2, 6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "index": [2, 8, 12, 13, 34, 35, 36], "yesno": [2, 7, 8, 19, 30, 37], "veri": [2, 3, 17, 27, 28, 30, 35, 36], "good": 2, "exampl": [2, 6, 7, 9, 11, 12, 13, 23, 27, 28, 30], "speech": [2, 6, 7, 8, 10, 11, 16, 17, 30, 37], "pull": [2, 15, 17, 20, 32], "380": [2, 28], "show": [2, 6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "add": [2, 15, 17, 18, 35, 37], "new": [2, 3, 6, 8, 15, 16, 17, 18, 20, 22, 23, 24, 25, 30, 34, 35, 36], "suppos": [2, 35, 36], "would": [2, 8, 9, 20, 23, 35, 36], "name": [2, 9, 15, 17, 22, 24, 25, 35, 36], "foo": [2, 11, 13, 15, 20, 22, 24, 25, 34, 35, 36], "eg": [2, 5, 8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "mkdir": [2, 15, 17, 18, 20, 23, 27, 28, 30, 34], "p": [2, 8, 17, 27, 28], "asr": [2, 5, 7, 8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36, 37], "touch": 2, "sh": [2, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "chmod": 2, "x": [2, 32], "simpl": [2, 17], "own": [2, 22, 35, 36], "otherwis": [2, 15, 17, 20, 22, 24, 25, 34, 35, 36], "librispeech": [2, 5, 7, 9, 11, 12, 13, 19, 20, 22, 23, 24, 25, 31, 32, 34, 35, 36, 37], "assum": [2, 8, 9, 15, 17, 18, 20, 22, 23, 27, 28, 30, 34, 35, 36], "fanci": 2, "call": 2, "bar": [2, 11, 13, 15, 20, 22, 24, 25, 34, 35, 36], "organ": 2, "wai": [2, 3, 14, 22, 24, 25, 32, 34, 35, 36], "readm": [2, 15, 17, 18, 20, 23, 27, 28, 30], "md": [2, 5, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "asr_datamodul": [2, 8], "pretrain": [2, 9, 11, 13, 15, 17, 18, 20, 23, 27, 28, 30], "For": [2, 5, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "instanc": [2, 5, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "tdnn": [2, 8, 16, 19, 21, 26, 29], "its": [2, 9, 13, 17, 24], "directori": [2, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "structur": 2, "descript": [2, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "contain": [2, 7, 9, 10, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36, 37], "inform": [2, 9, 15, 17, 18, 20, 22, 23, 24, 27, 28, 30, 32, 34, 35, 36], "g": [2, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "wer": [2, 8, 9, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "etc": [2, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "provid": [2, 6, 8, 9, 10, 11, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36, 37], "pytorch": [2, 7, 17], "dataload": [2, 8], "take": [2, 9, 22, 30, 35, 36], "input": [2, 9, 15, 17, 18, 20, 23, 27, 28, 30, 32], "checkpoint": [2, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "save": [2, 8, 9, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "dure": [2, 6, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "stage": [2, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "": [2, 8, 9, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "definit": 2, "neural": [2, 15, 20], "network": [2, 15, 17, 20, 22, 24, 25, 34, 35, 36], "script": [2, 7, 8, 13, 14, 15, 17, 18, 20, 23, 27, 28, 30, 34], "infer": [2, 9, 11], "tdnn_lstm_ctc": [2, 18, 23, 28], "conformer_ctc": [2, 15, 20], "get": [2, 6, 8, 15, 17, 18, 20, 22, 23, 24, 25, 30, 34, 35, 36], "feel": [2, 34], "result": [2, 5, 6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "everi": [2, 9, 22, 24, 25, 34, 35, 36], "kept": [2, 22, 35, 36], "self": [2, 10, 32], "toler": 2, "duplic": 2, "among": [2, 8], "differ": [2, 8, 15, 16, 20, 22, 32, 34, 35, 36], "invoc": 2, "help": [2, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "blob": [2, 5, 9, 13, 22, 24, 25, 34, 35, 36], "master": [2, 5, 9, 11, 12, 13, 17, 22, 24, 25, 34, 35, 36], "transform": [2, 15, 20, 34], "conform": [2, 11, 12, 16, 17, 19, 21, 22, 24, 34, 35, 36], "base": [2, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "lstm": [2, 10, 13, 16, 19, 21, 26, 31, 33], "attent": [2, 17, 18, 32, 35, 36], "lm": [2, 8, 17, 22, 23, 27, 28, 30, 35, 36], "rescor": [2, 18, 23, 25, 27, 28, 30], "demonstr": [2, 4, 6, 9], "consid": 2, "colab": 2, "notebook": 2, "welcom": 3, "There": [3, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "mani": [3, 35, 36], "two": [3, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "them": [3, 4, 5, 6, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "To": [3, 6, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "document": [3, 7, 9, 25], "repositori": 3, "recip": [3, 5, 7, 8, 9, 15, 17, 18, 20, 22, 23, 27, 28, 30, 32, 34, 35, 36], "In": [3, 6, 8, 9, 11, 12, 13, 14, 15, 17, 18, 20, 23, 27, 28, 30, 32], "page": [3, 6, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36, 37], "describ": [3, 4, 9, 11, 12, 13, 14, 15, 17, 18, 20, 22, 23, 27, 28, 35, 36], "how": [3, 4, 6, 7, 8, 14, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "creat": [3, 7, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35], "data": [3, 9, 11, 12, 13, 16], "train": [3, 4, 6, 7, 9, 12, 13, 32], "decod": [3, 6, 11, 13, 14], "model": [3, 4, 6, 7, 8, 10, 32], "section": [4, 8, 9, 11, 12, 13, 14, 15, 20], "find": [4, 5, 6, 8, 9, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "also": [4, 5, 8, 9, 10, 11, 13, 15, 17, 18, 20, 22, 24, 25, 30, 32, 34, 35, 36], "try": [4, 6, 22, 24, 25, 34, 35, 36], "from": [4, 6, 8, 9, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "within": [4, 6], "anyth": [4, 6], "space": [4, 7], "youtub": [4, 7, 20, 22, 23, 24, 25, 34, 35, 36], "video": [4, 7, 20, 22, 23, 24, 25, 34, 35, 36], "upload": [5, 6, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "huggingfac": [5, 7, 9, 15, 17, 18, 20, 23, 24, 25, 27, 28, 30, 34], "co": [5, 6, 9, 15, 16, 17, 18, 20, 23, 24, 25, 27, 28, 30, 34], "visit": [5, 6, 22, 24, 25, 34, 35, 36], "link": [5, 8, 9, 10, 22, 24, 25, 34, 35, 36], "search": [5, 6], "specif": [5, 17], "correspond": [5, 6], "aishel": [5, 7, 15, 17, 18, 19, 37], "gigaspeech": [5, 12, 34], "wenetspeech": [5, 12], "integr": 6, "framework": [6, 11, 22, 35], "sherpa": [6, 10, 11, 12, 13, 34], "need": [6, 8, 9, 10, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "download": [6, 7, 16], "window": [6, 10, 11, 34], "maco": [6, 10, 11, 34], "linux": [6, 10, 11, 34], "even": [6, 8], "ipad": 6, "phone": 6, "start": [6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "address": [6, 9, 17, 22, 25, 34, 35, 36], "recognit": [6, 7, 10, 11, 16, 17, 30, 37], "screenshot": [6, 15, 17, 18, 20, 22, 30, 34, 35], "select": [6, 22, 23, 27, 28, 30, 34, 35, 36], "languag": [6, 15, 17, 18], "current": [6, 8, 17, 32, 34, 35, 36, 37], "chines": [6, 16, 17], "english": [6, 30, 34], "target": 6, "method": [6, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "greedi": 6, "modified_beam_search": [6, 17, 22, 24, 34, 35, 36], "choos": [6, 8, 22, 24, 25, 34, 35, 36], "number": [6, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "activ": 6, "path": [6, 9, 11, 13, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "either": [6, 15, 17, 18, 20, 35, 36], "record": [6, 15, 16, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "click": [6, 8, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "button": 6, "submit": 6, "wait": 6, "moment": 6, "an": [6, 8, 9, 11, 12, 13, 15, 16, 17, 20, 22, 25, 30, 34, 35, 36], "when": [6, 14, 17, 20, 22, 24, 25, 35, 36], "bottom": [6, 22, 24, 25, 34, 35, 36], "part": [6, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "tabl": 6, "one": [6, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "next": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "gen": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "kaldi": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "subscrib": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "channel": [6, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "nadira": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "povei": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "www": [6, 8, 16, 20, 22, 23, 24, 25, 34, 35, 36], "uc_vaumpkminz1pnkfxan9mw": [6, 8, 20, 22, 23, 24, 25, 34, 35, 36], "torchaudio": [7, 32], "2": [7, 9, 11, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "export": [7, 8, 15, 17, 18, 20, 23, 27, 28, 30], "state_dict": [7, 14, 15, 17, 18, 20, 23, 27, 28, 30], "torch": [7, 8, 9, 14, 15, 17, 20], "jit": [7, 14, 20], "trace": [7, 12, 14], "onnx": [7, 9, 14], "ncnn": [7, 14], "non": [7, 11, 20, 32, 35, 37], "stream": [7, 11, 15, 20, 27, 28, 34, 37], "timit": [7, 19, 27, 28, 37], "introduct": [7, 31, 37], "contribut": 7, "depend": [8, 15, 20, 34], "step": [8, 9, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "order": [8, 15, 18, 20, 23, 27, 28], "matter": 8, "org": [8, 16, 17, 22, 34, 35, 36], "least": 8, "v1": [8, 15, 18, 20, 23, 27, 28], "9": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 30, 34, 35, 36], "alreadi": [8, 9], "don": [8, 12, 15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "t": [8, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "replac": 8, "compil": [8, 15, 17, 20], "against": 8, "strongli": 8, "collect": 8, "variabl": [8, 15, 18, 20, 22, 24, 25, 34, 35, 36], "pythonpath": 8, "point": [8, 9, 15, 18, 20, 22, 24, 25, 34, 35, 36], "folder": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "tmp": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "setup": [8, 15, 17, 18, 20, 22, 23, 27, 28, 30, 34, 35, 36], "put": [8, 24, 35], "sever": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "same": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "switch": [8, 15, 20, 25], "just": [8, 32], "about": [8, 17, 22, 25, 34, 35, 36], "virtualenv": 8, "8": [8, 9, 15, 17, 20, 22, 23, 24, 25, 30, 34, 35, 36], "cpython3": 8, "6": [8, 15, 17, 20, 22, 23, 27, 28, 34], "final": [8, 9, 20, 23], "64": [8, 9, 17, 35], "1540m": 8, "creator": 8, "cpython3posix": 8, "dest": 8, "ceph": [8, 9, 15, 17, 20], "fj": [8, 9, 17, 20], "fangjun": [8, 9, 17, 20], "clear": 8, "fals": [8, 9, 15, 17, 20], "no_vcs_ignor": 8, "global": 8, "seeder": 8, "fromappdata": 8, "bundl": 8, "setuptool": 8, "wheel": 8, "via": [8, 12, 13], "copi": [8, 32], "app_data_dir": 8, "root": 8, "share": 8, "v": [8, 20, 27, 28], "irtualenv": 8, "ad": [8, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "seed": 8, "packag": 8, "21": [8, 9, 15, 17, 20, 27, 28], "57": [8, 20, 23], "36": [8, 17, 20], "bashactiv": 8, "cshellactiv": 8, "fishactiv": 8, "powershellactiv": 8, "pythonactiv": 8, "xonshactiv": 8, "sourc": [8, 9, 15, 16, 17, 20], "bin": [8, 15, 20, 34], "dev20210822": 8, "cpu": [8, 9, 12, 15, 22, 24, 25, 30, 35, 36], "torch1": 8, "f": [8, 27, 28], "nightli": 8, "whl": [8, 34], "2bcpu": 8, "cp38": 8, "linux_x86_64": 8, "mb": 8, "________________________________": 8, "185": [8, 15, 20, 30], "kb": [8, 27, 28], "graphviz": 8, "17": [8, 9, 15, 20, 27, 28, 34], "py3": 8, "none": [8, 15, 20], "18": [8, 15, 17, 18, 20, 22, 23, 27, 28, 34, 35, 36], "cach": 8, "manylinux1_x86_64": 8, "831": [8, 17, 28], "type": [8, 9, 15, 17, 20, 22, 24, 25, 30, 32, 34, 35, 36], "extens": 8, "typing_extens": 8, "26": [8, 17, 20, 28], "successfulli": 8, "probabl": [8, 17, 22, 24, 34, 35, 36], "cuda": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "req": 8, "7b1b76ge": 8, "q": 8, "audioread": 8, "soundfil": 8, "post1": 8, "py2": 8, "7": [8, 9, 15, 18, 20, 22, 23, 27, 28, 34, 35], "97": [8, 15], "cytoolz": 8, "11": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "manylinux_2_17_x86_64": 8, "manylinux2014_x86_64": 8, "dataclass": 8, "14": [8, 9, 11, 12, 15, 20, 22, 23, 24, 27, 34, 35, 36], "h5py": 8, "manylinux_2_12_x86_64": 8, "manylinux2010_x86_64": 8, "684": [8, 15, 30], "intervaltre": 8, "lilcom": 8, "numpi": 8, "15": [8, 9, 17, 18, 20, 27, 30], "40": [8, 18, 20, 23, 27, 28], "pyyaml": 8, "662": 8, "tqdm": 8, "62": [8, 20, 23], "76": [8, 30], "73": 8, "satisfi": 8, "lib": 8, "site": 8, "dev": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "2a1410b": 8, "clean": [8, 15, 17, 20, 22, 23, 24, 25, 34, 35, 36], "toolz": 8, "55": [8, 18, 20, 27], "sortedcontain": 8, "29": [8, 15, 17, 18, 20, 23, 27, 28], "cffi": 8, "411": [8, 20], "pycpars": 8, "20": [8, 9, 15, 17, 18, 20, 22, 23, 27, 28, 30, 35], "112": 8, "pypars": 8, "67": 8, "done": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "filenam": [8, 11, 12, 13, 24, 25, 34, 36], "dev_2a1410b_clean": 8, "size": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "342242": 8, "sha256": 8, "f683444afa4dc0881133206b4646a": 8, "9d0f774224cc84000f55d0a67f6e4a37997": 8, "store": [8, 20], "ephem": 8, "ftu0qysz": 8, "7f": 8, "7a": 8, "8e": 8, "a0bf241336e2e3cb573e1e21e5600952d49f5162454f2e612f": 8, "warn": 8, "built": 8, "invalid": [8, 20], "metadata": [8, 27, 28], "mandat": 8, "pep": 8, "440": 8, "packa": 8, "ging": 8, "deprec": [8, 17], "legaci": 8, "becaus": 8, "could": [8, 15, 18], "A": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 34, 35, 36], "discuss": 8, "regard": 8, "pypa": 8, "sue": 8, "8368": 8, "inter": 8, "valtre": 8, "sor": 8, "tedcontain": 8, "remot": 8, "enumer": 8, "object": [8, 15, 17, 18, 22, 30, 34, 35], "500": [8, 9, 17, 20, 25, 34], "count": 8, "100": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "compress": 8, "308": [8, 15, 17, 18], "total": [8, 15, 17, 18, 20, 22, 23, 30, 34, 35], "delta": 8, "263": 8, "reus": 8, "307": 8, "102": [8, 15], "pack": [8, 35, 36], "receiv": 8, "172": 8, "49": [8, 20, 28, 30], "kib": 8, "385": 8, "00": [8, 15, 17, 18, 20, 23, 27, 28, 30], "resolv": 8, "kaldilm": 8, "tar": 8, "gz": 8, "48": [8, 15, 17], "574": 8, "kaldialign": 8, "sentencepiec": [8, 20], "96": 8, "tensorboard": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "41": [8, 15, 17, 27, 30], "line": [8, 22, 35, 36], "absl": 8, "absl_pi": 8, "13": [8, 9, 17, 18, 20, 23, 24, 27], "132": 8, "googl": [8, 22, 24, 25, 34, 35, 36], "auth": 8, "oauthlib": 8, "google_auth_oauthlib": 8, "grpcio": 8, "24": [8, 18, 23, 27, 28, 30], "39": [8, 17, 20, 23, 27], "ment": 8, "12": [8, 9, 15, 17, 18, 20, 22, 24, 25, 27, 30, 34, 35, 36], "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, 9, 17, 20, 34], "152": 8, "request": [8, 32], "plugin": 8, "wit": 8, "tensorboard_plugin_wit": 8, "781": 8, "markdown": 8, "six": 8, "16": [8, 9, 13, 15, 17, 18, 20, 22, 23, 27, 28, 30, 34, 35, 36], "cachetool": 8, "rsa": 8, "34": 8, "pyasn1": 8, "modul": [8, 24, 35], "pyasn1_modul": 8, "155": 8, "requests_oauthlib": 8, "23": [8, 15, 17, 18, 20, 27, 28, 30], "77": [8, 20], "urllib3": 8, "27": [8, 15, 17, 23, 28], "138": [8, 15, 17], "certifi": 8, "2017": 8, "2021": [8, 15, 18, 20, 23, 27, 28, 30], "30": [8, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "145": 8, "charset": 8, "normal": [8, 23, 27, 28, 30, 35], "charset_norm": 8, "idna": 8, "59": [8, 18, 20], "146": 8, "897233": 8, "eccb906cafcd45bf9a7e1a1718e4534254bfb": 8, "f4c0d0cbc66eee6c88d68a63862": 8, "85": 8, "7d": 8, "63": [8, 17], "f2dd586369b8797cb36d213bf3a84a789eeb92db93d2e723c9": 8, "etool": 8, "oaut": 8, "hlib": 8, "let": [8, 15, 20, 34], "u": [8, 15, 17, 18, 20, 30, 34], "log": [8, 23, 27, 28, 30], "08": [8, 20, 23, 25, 27, 28, 30, 34], "19": [8, 9, 15, 20, 23, 27, 28], "main": [8, 15, 20, 32], "dl_dir": [8, 15, 18, 20, 22, 24, 25, 34, 35, 36], "waves_yesno": 8, "49mb": 8, "03": [8, 9, 17, 20, 27, 28, 34], "39mb": 8, "manifest": 8, "31": [8, 20], "42": [8, 15, 20, 30], "comput": [8, 9, 15, 17, 18, 22, 23, 25, 27, 28, 30, 34, 35, 36], "fbank": [8, 9, 15, 17, 18, 20, 23, 27, 28, 30], "32": [8, 15, 17, 18, 36], "803": 8, "info": [8, 9, 15, 17, 18, 20, 23, 27, 28, 30], "compute_fbank_yesno": 8, "52": [8, 15, 20], "process": [8, 9, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "extract": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "featur": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "_______________________________________________________________": 8, "90": 8, "01": [8, 17, 18, 20], "80": [8, 9, 15, 17, 20], "57it": 8, "085": 8, "______________________________________________________________": 8, "248": [8, 17], "21it": 8, "lang": [8, 9, 17, 20, 25], "fcordre9": 8, "kaldilm_6899d26f2d684ad48f21025950cd2866": 8, "csrc": [8, 20], "arpa_file_pars": 8, "cc": 8, "void": 8, "arpafilepars": 8, "rea": 8, "d": [8, 27, 28], "std": 8, "istream": 8, "79": 8, "140": [8, 18], "gram": [8, 15, 17, 18, 22, 23, 25, 27, 28, 35, 36], "89": [8, 15], "hlg": [8, 23, 27, 28, 30], "928": 8, "compile_hlg": 8, "120": 8, "lang_phon": [8, 18, 23, 27, 28, 30], "929": [8, 17], "lexicon": [8, 15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "116": 8, "convert": [8, 20, 34], "l": [8, 17, 27, 28, 30, 34], "pt": [8, 9, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "linv": [8, 17, 20, 30], "931": 8, "ctc_topo": 8, "max_token_id": 8, "932": 8, "load": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "fst": [8, 17, 30], "intersect": [8, 22, 35, 36], "933": 8, "lg": [8, 22, 25, 35, 36], "shape": 8, "66": 8, "connect": [8, 9, 20, 22, 23, 34, 35, 36], "68": [8, 20], "70": 8, "class": [8, 20], "tensor": [8, 15, 17, 18, 20, 22, 30, 34, 35], "71": [8, 20, 23], "determin": 8, "934": 8, "74": [8, 9], "_k2": 8, "raggedint": 8, "remov": [8, 15, 17, 18, 20, 23, 27, 28], "disambigu": 8, "symbol": [8, 17, 22, 35, 36], "87": 8, "remove_epsilon": 8, "935": 8, "92": [8, 20], "arc": 8, "95": [8, 16], "compos": 8, "h": 8, "105": [8, 20], "936": 8, "107": [8, 23], "123": 8, "now": [8, 15, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "cuda_visible_devic": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "gpu": [8, 15, 17, 18, 20, 22, 24, 25, 27, 28, 30, 34, 35, 36], "avail": [8, 9, 15, 17, 20, 23, 27, 28, 30, 34], "case": [8, 9, 22, 24, 25, 34, 35, 36], "segment": 8, "fault": 8, "core": 8, "dump": 8, "error": [8, 20], "protocol_buffers_python_implement": 8, "more": [8, 15, 20, 30, 32, 34, 35], "674": 8, "interest": [8, 22, 24, 25, 34, 35, 36], "given": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 35, 36], "below": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35], "072": 8, "465": 8, "466": 8, "exp_dir": [8, 17, 20, 22, 24, 25, 35, 36], "posixpath": [8, 17, 20], "exp": [8, 9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "lang_dir": [8, 17, 20], "lr": [8, 17, 34], "feature_dim": [8, 9, 15, 17, 20, 30], "weight_decai": 8, "1e": 8, "06": [8, 9, 18, 20, 23, 30], "start_epoch": 8, "best_train_loss": [8, 9], "inf": [8, 9], "best_valid_loss": [8, 9], "best_train_epoch": [8, 9], "best_valid_epoch": [8, 9], "batch_idx_train": [8, 9], "log_interv": [8, 9], "valid_interv": [8, 9], "beam_siz": [8, 9, 17], "reduct": [8, 24], "sum": 8, "use_doub": 8, "le_scor": 8, "true": [8, 9, 15, 17, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "world_siz": 8, "master_port": 8, "12354": 8, "num_epoch": 8, "feature_dir": [8, 20], "max_dur": [8, 20], "bucketing_sampl": [8, 20], "num_bucket": [8, 20], "concatenate_cut": [8, 20], "duration_factor": [8, 20], "gap": [8, 20], "on_the_fly_feat": [8, 20], "shuffl": [8, 20], "return_cut": [8, 20], "num_work": [8, 20], "074": 8, "113": [8, 17, 20], "098": [8, 23], "cut": [8, 20], "240": [8, 15, 30], "149": [8, 20], "200": [8, 9, 15, 20, 27, 28, 30], "singlecutsampl": 8, "206": [8, 20], "219": [8, 17, 20], "246": [8, 17, 20, 27, 28], "357": 8, "416": 8, "epoch": [8, 9, 11, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "batch": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "avg": [8, 9, 11, 12, 13, 17, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "loss": [8, 15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "0789": 8, "848": 8, "5356": 8, "7556": 8, "301": [8, 9, 20], "432": [8, 20], "9972": 8, "best": [8, 15, 18, 20], "805": 8, "2436": 8, "5717": 8, "33": [8, 15, 16, 17, 20, 27], "109": [8, 15, 20], "4167": 8, "121": [8, 23], "325": 8, "2214": 8, "798": [8, 17], "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, 20], "0093": 8, "0096": 8, "50": [8, 9, 20, 22, 27, 34, 35, 36], "239": [8, 17], "0104": 8, "0101": 8, "569": 8, "0092": 8, "819": [8, 27], "835": 8, "51": [8, 15, 20, 30], "024": 8, "0105": 8, "317": 8, "0099": 8, "0097": 8, "552": 8, "0108": 8, "869": 8, "0102": 8, "126": [8, 20], "128": [8, 20], "537": [8, 20], "192": [8, 20], "249": 8, "250": [8, 17, 23], "lm_dir": [8, 20], "search_beam": [8, 15, 20, 30], "output_beam": [8, 15, 20, 30], "min_active_st": [8, 15, 20, 30], "max_active_st": [8, 15, 20, 30], "10000": [8, 15, 20, 30], "use_double_scor": [8, 15, 20, 30], "193": 8, "213": [8, 30], "259": [8, 15], "devic": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 35, 36], "217": [8, 15, 20], "279": [8, 20], "averag": [8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "userwarn": [8, 17], "floor_divid": 8, "futur": [8, 17, 37], "round": [8, 17], "toward": [8, 17], "trunc": [8, 17], "function": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "NOT": [8, 15, 17, 20, 30], "floor": [8, 17], "incorrect": [8, 17], "neg": [8, 17], "valu": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "keep": [8, 17, 22, 35, 36], "behavior": [8, 17], "div": [8, 17], "b": [8, 17, 20, 27, 28], "rounding_mod": [8, 17], "actual": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "divis": [8, 17], "trigger": 8, "intern": 8, "aten": 8, "src": [8, 34], "nativ": 8, "binaryop": 8, "cpp": [8, 12], "450": [8, 15, 17, 18], "k": [8, 22, 27, 28, 34, 35, 36], "n": [8, 15, 22, 24, 25, 27, 28, 34, 35, 36], "220": [8, 17, 18, 20], "409": 8, "190": [8, 23], "until": [8, 20, 24], "571": [8, 20], "228": [8, 20], "transcript": [8, 15, 16, 17, 18, 20, 22, 23, 27, 28, 34, 35, 36], "recog": [8, 17, 20], "test_set": [8, 30], "572": 8, "util": [8, 20], "ins": [8, 20, 30], "del": [8, 20, 30], "sub": [8, 20, 30], "573": 8, "236": 8, "wrote": [8, 20], "detail": [8, 10, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "stat": [8, 20], "err": [8, 17, 20], "299": 8, "congratul": [8, 15, 18, 20, 23, 27, 28, 30], "first": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "fun": 8, "debug": 8, "variou": [8, 14, 37], "problem": 8, "mai": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36, 37], "encount": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "while": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "period": 9, "disk": 9, "optim": [9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "other": [9, 17, 20, 22, 23, 27, 28, 30, 32, 35, 36, 37], "relat": [9, 15, 17, 20, 23, 27, 28, 30], "resum": [9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "howev": 9, "onli": [9, 11, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36, 37], "strip": 9, "except": 9, "reduc": [9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "each": [9, 15, 17, 18, 20, 22, 24, 25, 32, 34, 35, 36], "well": [9, 30, 37], "usag": [9, 12, 13, 23, 27, 28, 30], "pruned_transducer_stateless3": [9, 11, 12, 32], "almost": [9, 22, 32, 35, 36], "dir": [9, 11, 12, 13, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "bpe": [9, 11, 12, 13, 20, 22, 24, 25, 34, 35, 36], "lang_bpe_500": [9, 11, 12, 13, 20, 22, 24, 25, 34, 35, 36], "dict": 9, "host": 9, "csukuangfj": [9, 15, 17, 18, 20, 23, 27, 28, 30, 34], "prune": [9, 17, 19, 21, 31, 32, 33, 34, 36], "transduc": [9, 10, 11, 16, 19, 21, 31, 32, 33], "stateless3": 9, "2022": [9, 17, 22, 24, 25, 34, 35], "05": [9, 15, 17, 18, 20, 28], "lf": [9, 15, 17, 18, 20, 23, 25, 27, 28, 30], "repo": 9, "prefix": 9, "those": 9, "xxx": 9, "wave": [9, 11, 15, 20], "iter": [9, 13, 22, 24, 25, 34, 35, 36], "1224000": 9, "greedy_search": [9, 17, 22, 24, 34, 35, 36], "test_wav": [9, 15, 17, 18, 20, 23, 27, 28, 30], "1089": [9, 20, 23], "134686": [9, 20, 23], "0001": [9, 20, 23], "wav": [9, 11, 13, 15, 17, 18, 20, 22, 24, 25, 27, 28, 30, 34, 35, 36], "1221": [9, 20, 23], "135766": [9, 20, 23], "0002": [9, 20, 23], "multipl": [9, 15, 17, 18, 20, 23, 27, 28, 30], "sound": [9, 13, 15, 17, 18, 20, 23, 27, 28, 30], "Its": [9, 20], "output": [9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "09": [9, 15, 17, 18, 20, 34], "02": [9, 17, 20, 22, 28, 34, 35], "233": 9, "265": 9, "reset_interv": 9, "3000": 9, "subsampling_factor": [9, 15, 17, 20], "encoder_dim": 9, "512": [9, 15, 17, 20], "nhead": [9, 15, 17, 20, 22, 35], "dim_feedforward": [9, 17], "2048": [9, 17], "num_encoder_lay": [9, 17], "decoder_dim": 9, "joiner_dim": 9, "model_warm_step": 9, "env_info": [9, 15, 17, 20], "releas": [9, 15, 17, 20], "sha1": [9, 15, 17, 20], "4810e00d8738f1a21278b0156a42ff396a2d40ac": 9, "date": [9, 15, 17, 20], "fri": 9, "oct": [9, 20], "miss": [9, 17, 20], "cu102": 9, "branch": [9, 15, 17, 20, 24], "1013": 9, "c39cba5": 9, "dirti": [9, 15, 20], "thu": [9, 17, 20, 23], "__init__": [9, 15, 17, 20], "jsonl": 9, "hostnam": [9, 17], "de": [9, 17], "74279": [9, 17], "0324160024": 9, "65bfd8b584": 9, "jjlbn": 9, "ip": [9, 17], "177": [9, 17, 18, 20], "203": [9, 20], "bpe_model": [9, 20], "sound_fil": [9, 15, 17, 20, 30], "sample_r": [9, 15, 17, 20, 30], "16000": [9, 15, 17, 18, 20, 23, 24, 27, 28], "beam": [9, 34], "max_context": 9, "max_stat": 9, "context_s": [9, 17], "max_sym_per_fram": [9, 17], "simulate_stream": 9, "decode_chunk_s": 9, "left_context": 9, "dynamic_chunk_train": 9, "causal_convolut": 9, "short_chunk_s": [9, 35, 36], "25": [9, 15, 20, 22, 27, 28, 30, 35], "num_left_chunk": 9, "blank_id": [9, 17], "unk_id": 9, "vocab_s": [9, 17], "271": 9, "273": [9, 17], "612": 9, "458": 9, "disabl": 9, "giga": [9, 34], "623": 9, "277": 9, "paramet": [9, 12, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "78648040": 9, "951": [9, 20], "285": [9, 17, 20], "construct": [9, 15, 17, 18, 20, 23, 27, 28, 30], "952": 9, "295": [9, 15, 17, 18, 20], "957": 9, "700": 9, "329": [9, 20], "912": 9, "388": 9, "earli": [9, 20, 23], "nightfal": [9, 20, 23], "THE": [9, 20, 23], "yellow": [9, 20, 23], "lamp": [9, 20, 23], "light": [9, 20, 23], "here": [9, 15, 17, 18, 20, 23, 32, 35], "AND": [9, 20, 23], "THERE": [9, 20, 23], "squalid": [9, 20, 23], "quarter": [9, 20, 23], "OF": [9, 20, 23], "brothel": [9, 20, 23], "god": [9, 20, 23], "AS": [9, 20, 23], "direct": [9, 20, 23], "consequ": [9, 20, 23], "sin": [9, 20, 23], "man": [9, 20, 23], "punish": [9, 20, 23], "had": [9, 20, 23], "her": [9, 20, 23], "love": [9, 20, 23], "child": [9, 20, 23], "whose": [9, 17, 20, 23], "ON": [9, 20, 23], "THAT": [9, 20, 23], "dishonor": [9, 20, 23], "bosom": [9, 20, 23], "TO": [9, 20, 23], "parent": [9, 20, 23], "forev": [9, 20, 23], "WITH": [9, 20, 23], "race": [9, 20, 23], "descent": [9, 20, 23], "mortal": [9, 20, 23], "BE": [9, 20, 23], "bless": [9, 20, 23], "soul": [9, 20, 23], "IN": [9, 20, 23], "heaven": [9, 20, 23], "yet": [9, 20, 23], "THESE": [9, 20, 23], "thought": [9, 20, 23], "affect": [9, 20, 23], "hester": [9, 20, 23], "prynn": [9, 20, 23], "less": [9, 20, 23, 30, 35, 36], "hope": [9, 16, 20, 23], "than": [9, 15, 17, 18, 20, 22, 23, 24, 25, 30, 34, 35, 36], "apprehens": [9, 20, 23], "390": 9, "alwai": 9, "note": [9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "down": [9, 15, 20, 22, 24, 25, 34, 35, 36], "reproduc": [9, 20], "ln": [9, 15, 20, 22, 24, 25, 34, 35, 36], "9999": [9, 24, 25, 34], "symlink": 9, "pass": [9, 15, 17, 18, 20, 22, 24, 25, 32, 34, 35, 36], "max": [9, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "durat": [9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "600": [9, 20, 22, 24, 34, 35, 36], "reason": [9, 35], "support": [10, 11, 15, 17, 20, 22, 24, 25, 32, 34, 35, 36], "perform": [10, 11, 17, 35], "raspberri": [10, 34], "pi": [10, 34], "project": 10, "static": [10, 34], "produc": [10, 22, 24, 25, 34, 35, 36], "binari": [10, 15, 17, 18, 20, 22, 30, 34, 35], "everyth": 10, "tree": [11, 12, 13, 15, 17, 18, 20, 23, 27, 28, 30, 34], "insid": [11, 13], "encod": [11, 13, 15, 17, 18, 20, 22, 23, 24, 30, 32, 34, 35, 36], "joiner": [11, 13, 17, 22, 34, 35, 36], "joiner_encoder_proj": 11, "joiner_decoder_proj": 11, "onnx_pretrain": 11, "proj": 11, "baz": [11, 13], "onnxruntim": 11, "our": [12, 13, 20, 22, 32, 35, 36], "torchscript": [12, 13, 34], "cpu_jit": [12, 15, 20, 22, 24, 25, 35, 36], "confus": 12, "move": [12, 22, 24, 25, 35, 36], "why": 12, "streaming_asr": [12, 13, 34, 35, 36], "emform": 12, "conv_emform": 12, "offline_asr": [12, 22], "lstm_transducer_stateless2": [13, 34], "468000": [13, 34], "three": [13, 15, 17, 32], "encoder_jit_trac": [13, 34, 36], "decoder_jit_trac": [13, 34, 36], "joiner_jit_trac": [13, 34, 36], "jit_pretrain": [13, 24, 25, 34], "tutori": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "learn": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "singl": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "1best": [15, 18, 20, 23, 24, 25, 27, 28], "handl": [15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "automag": [15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "stop": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "control": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "By": [15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "default": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "execut": [15, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "mean": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 32, 34, 35, 36], "musan": [15, 18, 20, 22, 24, 25, 34, 35, 36], "sai": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "thei": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "won": [15, 18, 20, 22, 24, 25, 34, 35, 36], "re": [15, 18, 20, 22, 24, 25, 34, 35, 36], "intal": [15, 18], "initi": [15, 18], "sudo": [15, 18], "apt": [15, 18], "permiss": [15, 18], "commandlin": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "quit": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "often": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "experi": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "num": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "state": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "world": [15, 17, 18, 20, 22, 23, 24, 25, 34, 35, 36], "multi": [15, 17, 18, 20, 22, 24, 25, 32, 34, 35, 36], "machin": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "ddp": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "implement": [15, 17, 18, 20, 22, 24, 25, 32, 34, 35, 36], "present": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "later": [15, 18, 20, 22, 23, 24, 25, 27, 28, 34, 35, 36], "specifi": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "second": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "over": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "utter": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "pad": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "oom": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "v100": [15, 17, 18, 20], "nvidia": [15, 17, 18, 20], "due": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "usual": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "larger": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "caus": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "smaller": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "increas": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "tune": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "weight": [15, 18, 20, 24, 25, 34], "decai": [15, 18, 20, 24, 25, 34], "warmup": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "get_param": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "realli": [15, 18, 20, 22, 24, 25, 34, 35, 36], "directli": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "perturb": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "speed": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "factor": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "3x150": [15, 17, 18], "hour": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "These": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "rate": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "visual": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "logdir": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "labelsmooth": 15, "someth": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "tensorflow": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "found": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "continu": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "press": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "ctrl": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35, 36], "engw8ksktzqs24zbv5dgcg": 15, "22t11": 15, "scan": [15, 17, 18, 20, 22, 30, 34, 35], "116068": 15, "scalar": [15, 17, 18, 20, 22, 30, 34, 35], "listen": [15, 17, 18, 22, 30, 34, 35], "url": [15, 17, 18, 20, 22, 24, 25, 30, 34, 35], "xxxx": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "text": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "saw": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "consol": [15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "typic": [15, 17, 18, 20], "avoid": [15, 17, 20], "commonli": [15, 17, 18, 20, 23, 27, 28, 30], "nbest": [15, 20, 25], "scale": [15, 20, 23, 25, 27, 28], "lattic": [15, 18, 20, 22, 23, 27, 28, 35, 36], "score": [15, 20, 22, 35, 36], "uniqu": [15, 20, 22, 35, 36], "pkufool": [15, 18, 23], "icefall_asr_aishell_conformer_ctc": 15, "transcrib": [15, 17, 18, 20], "lang_char": [15, 17], "token": [15, 17, 18, 20, 23, 27, 28, 30], "word": [15, 17, 18, 20, 23, 27, 28, 30], "bac009s0764w0121": [15, 17, 18], "bac009s0764w0122": [15, 17, 18], "bac009s0764w0123": [15, 17, 18], "tran": [15, 18, 20, 23, 27, 28], "graph": [15, 18, 20, 22, 23, 27, 28, 35, 36], "id": [15, 18, 20, 23, 27, 28], "conveni": [15, 18, 20], "eo": [15, 18, 20], "easili": [15, 18, 20], "obtain": [15, 17, 18, 20, 23, 27, 28], "84": 15, "list": [15, 17, 18, 20, 23, 27, 28], "soxi": [15, 17, 18, 20, 23, 30], "sampl": [15, 17, 18, 20, 23, 24, 30, 35, 36], "precis": [15, 17, 18, 20, 22, 23, 30, 35, 36], "bit": [15, 17, 18, 20, 23, 30], "04": [15, 17, 18, 20, 23, 27, 28], "67263": [15, 17, 18], "315": [15, 17, 18, 20, 23], "cdda": [15, 17, 18, 20, 23, 30], "sector": [15, 17, 18, 20, 23, 30], "135k": [15, 17, 18], "256k": [15, 17, 18, 20], "sign": [15, 17, 18, 20, 30], "integ": [15, 17, 18, 20, 30], "pcm": [15, 17, 18, 20, 30], "65840": [15, 17, 18], "625": [15, 17, 18], "132k": [15, 17, 18], "64000": [15, 17, 18], "300": [15, 17, 18, 20, 22, 35], "128k": [15, 17, 18, 30], "displai": [15, 17, 18, 20], "topologi": [15, 20], "07": [15, 17, 18, 20], "53": [15, 22, 23, 28, 34, 35], "707": [15, 20], "229": 15, "attention_dim": [15, 17, 20], "num_decoder_lay": [15, 20], "vgg_frontend": [15, 17, 20], "use_feat_batchnorm": [15, 20], "f2fd997f752ed11bbef4c306652c433e83f9cf12": 15, "sun": 15, "sep": 15, "46": [15, 20], "33cfe45": 15, "d57a873": 15, "wed": [15, 17, 20], "nov": [15, 20], "hw": 15, "kangwei": 15, "icefall_aishell3": 15, "k2_releas": 15, "tokens_fil": 15, "words_fil": [15, 20, 30], "num_path": [15, 20, 22, 35, 36], "ngram_lm_scal": [15, 20], "attention_decoder_scal": [15, 20], "nbest_scal": [15, 20], "sos_id": [15, 20], "eos_id": [15, 20], "num_class": [15, 20, 30], "4336": [15, 17], "708": [15, 17, 20, 30], "242": [15, 20], "131": [15, 20], "134": 15, "269": [15, 27, 28], "275": 15, "241": 15, "293": [15, 20], "704": [15, 27], "369": [15, 20], "\u751a": [15, 17], "\u81f3": [15, 17], "\u51fa": [15, 17], "\u73b0": [15, 17], "\u4ea4": [15, 17], "\u6613": [15, 17], "\u51e0": [15, 17], "\u4e4e": [15, 17], "\u505c": [15, 17], "\u6b62": 15, "\u7684": [15, 17, 18], "\u60c5": [15, 17], "\u51b5": [15, 17], "\u4e00": [15, 17], "\u4e8c": [15, 17], "\u7ebf": [15, 17, 18], "\u57ce": [15, 17], "\u5e02": [15, 17], "\u867d": [15, 17], "\u7136": [15, 17], "\u4e5f": [15, 17, 18], "\u5904": [15, 17], "\u4e8e": [15, 17], "\u8c03": [15, 17], "\u6574": [15, 17], "\u4e2d": [15, 17, 18], "\u4f46": [15, 17, 18], "\u56e0": [15, 17], "\u4e3a": [15, 17], "\u805a": [15, 17], "\u96c6": [15, 17], "\u4e86": [15, 17, 18], "\u8fc7": [15, 17], "\u591a": [15, 17], "\u516c": [15, 17], "\u5171": [15, 17], "\u8d44": [15, 17], "\u6e90": [15, 17], "371": 15, "37": [15, 17, 20, 27], "38": [15, 17, 20, 27], "683": 15, "47": [15, 20], "651": [15, 30], "654": 15, "659": 15, "752": 15, "321": 15, "887": 15, "340": 15, "370": 15, "\u751a\u81f3": [15, 18], "\u51fa\u73b0": [15, 18], "\u4ea4\u6613": [15, 18], "\u51e0\u4e4e": [15, 18], "\u505c\u6b62": 15, "\u60c5\u51b5": [15, 18], "\u4e00\u4e8c": [15, 18], "\u57ce\u5e02": [15, 18], "\u867d\u7136": [15, 18], "\u5904\u4e8e": [15, 18], "\u8c03\u6574": [15, 18], "\u56e0\u4e3a": [15, 18], "\u805a\u96c6": [15, 18], "\u8fc7\u591a": [15, 18], "\u516c\u5171": [15, 18], "\u8d44\u6e90": [15, 18], "372": 15, "recor": [15, 20], "highest": [15, 20], "965": 15, "966": 15, "821": 15, "822": 15, "826": 15, "916": 15, "115": [15, 20], "345": 15, "888": 15, "889": 15, "limit": [15, 17, 20, 32, 35], "memori": [15, 17, 20, 32], "upgrad": [15, 20], "pro": [15, 20], "finish": [15, 17, 18, 20, 22, 23, 27, 28, 30, 35, 36], "deploi": [15, 20], "At": [15, 20], "doe": [15, 17, 20, 30], "home": [15, 20], "checkout": [15, 20], "v2": [15, 20], "cmake": [15, 20], "dcmake_build_typ": [15, 20], "j": [15, 20], "hlg_decod": [15, 20], "four": [15, 20], "messag": [15, 20, 22, 24, 25, 34, 35, 36], "nn_model": [15, 20], "use_gpu": [15, 20], "word_tabl": [15, 20], "caution": [15, 20], "forward": [15, 20, 24], "cu": [15, 20], "int": [15, 20], "char": [15, 20], "124": [15, 20], "98": 15, "142": [15, 18, 20], "150": [15, 20], "693": [15, 27], "165": [15, 20], "nnet_output": [15, 20], "182": [15, 23], "180": [15, 20], "489": 15, "45": [15, 17, 20], "216": [15, 20, 27, 28], "mandarin": 16, "corpu": 16, "beij": 16, "shell": 16, "technologi": 16, "ltd": 16, "400": 16, "peopl": 16, "accent": 16, "area": 16, "china": 16, "invit": 16, "particip": 16, "conduct": 16, "quiet": 16, "indoor": 16, "high": 16, "fidel": 16, "microphon": 16, "downsampl": 16, "16khz": 16, "manual": 16, "accuraci": 16, "through": 16, "profession": 16, "annot": 16, "strict": 16, "inspect": 16, "free": [16, 34], "academ": 16, "moder": 16, "amount": 16, "research": 16, "field": 16, "openslr": 16, "ctc": [16, 19, 21, 25, 26, 29], "stateless": [16, 19, 22, 34, 35, 36], "instead": [17, 35], "rnn": [17, 22, 24, 34, 35, 36], "As": [17, 20], "head": [17, 32], "dim": [17, 22, 35], "layer": [17, 22, 32, 34, 35, 36], "feedforward": [17, 22, 35], "embed": [17, 22, 34, 35, 36], "conv1d": [17, 22, 34, 35, 36], "kernel": 17, "left": [17, 35, 36], "context": [17, 22, 32, 34, 35, 36], "nn": [17, 22, 24, 25, 34, 35, 36], "tanh": 17, "linear": 17, "borrow": 17, "ieeexplor": 17, "ieee": 17, "stamp": 17, "jsp": 17, "arnumb": 17, "9054419": 17, "predict": [17, 22, 34, 35, 36], "modif": 17, "right": [17, 32, 35], "charact": 17, "unit": 17, "vocabulari": 17, "87939824": 17, "88": 17, "optimized_transduc": 17, "extra": [17, 32, 35], "technqiu": 17, "propos": [17, 32, 36], "improv": 17, "end": [17, 22, 24, 25, 30, 34, 35, 36], "furthermor": 17, "maximum": 17, "emit": 17, "per": [17, 22, 35, 36], "frame": [17, 22, 24, 35, 36], "simplifi": [17, 32], "significantli": 17, "degrad": 17, "exactli": 17, "benchmark": 17, "unprun": 17, "advantag": 17, "minim": 17, "pruned_transducer_stateless": [17, 22, 32, 35], "altern": 17, "though": 17, "transducer_stateless_modifi": 17, "option": [17, 23, 27, 28, 30], "pr": 17, "gb": 17, "ram": 17, "small": [17, 27, 28, 30], "tri": 17, "prob": [17, 34], "appli": [17, 32], "configur": [17, 23, 27, 28, 30], "c": [17, 18, 22, 24, 25, 30, 34, 35, 36], "lagz6hrcqxoigbfd5e0y3q": 17, "03t14": 17, "8477": 17, "sym": [17, 22, 35, 36], "beam_search": [17, 22, 35, 36], "decoding_method": 17, "beam_4": 17, "28": [17, 20, 23], "ensur": 17, "give": 17, "poor": 17, "531": [17, 18], "994": [17, 20], "176": [17, 20], "027": 17, "encoder_out_dim": 17, "f4fefe4882bc0ae59af951da3f47335d5495ef71": 17, "feb": 17, "50d2281": 17, "mar": 17, "0815224919": 17, "75d558775b": 17, "mmnv8": 17, "72": [17, 20], "878": [17, 28], "257": [17, 27, 28], "880": 17, "267": [17, 27, 28], "891": 17, "__floordiv__": 17, "length": [17, 35, 36], "x_len": 17, "163": [17, 20], "320": 17, "\u6ede": 17, "322": 17, "759": 17, "760": 17, "919": 17, "922": 17, "046": 17, "047": 17, "319": [17, 20], "214": [17, 20], "215": [17, 20, 23], "402": 17, "topk_hyp_index": 17, "topk_index": 17, "logit": 17, "583": [17, 28], "2000": 18, "lji9mwuorlow3jkdhxwk8a": 18, "13t11": 18, "4454": 18, "icefall_asr_aishell_tdnn_lstm_ctc": 18, "858": [18, 20], "389": [18, 20], "154": 18, "161": [18, 20], "536": 18, "171": [18, 20, 27, 28], "539": 18, "917": 18, "207": [18, 20], "129": 18, "\u505c\u6ede": 18, "222": [18, 20], "statelessx": [19, 21, 31, 32, 33], "zipform": [19, 21, 31, 33], "mmi": [19, 21], "blank": [19, 21], "skip": [19, 21, 22, 34, 35, 36], "ligru": [19, 26], "full": [20, 22, 24, 25, 34, 35, 36], "libri": [20, 22, 24, 25, 34, 35, 36], "960": [20, 22, 24, 25, 34, 35, 36], "subset": [20, 22, 24, 25, 34, 35, 36], "3x960": [20, 22, 24, 25, 34, 35, 36], "2880": [20, 22, 24, 25, 34, 35, 36], "lzgnetjwrxc3yghnmd4kpw": 20, "24t16": 20, "43": 20, "4540": 20, "sentenc": 20, "piec": 20, "And": [20, 22, 24, 25, 34, 35, 36], "neither": 20, "nor": 20, "vocab": 20, "work": 20, "5000": 20, "44": [20, 27, 28], "033": 20, "538": 20, "full_libri": 20, "406": 20, "464": 20, "548": 20, "776": 20, "652": [20, 30], "109226120": 20, "714": [20, 27], "473": 20, "944": 20, "1328": 20, "54": [20, 23, 27, 28], "443": [20, 23], "2563": 20, "56": [20, 27], "494": 20, "592": 20, "331": [20, 23], "1715": 20, "52576": 20, "1424": 20, "807": 20, "506": 20, "808": [20, 27], "522": 20, "362": 20, "565": 20, "1477": 20, "106": 20, "2922": 20, "208": 20, "4295": 20, "52343": 20, "396": 20, "3584": 20, "433": 20, "680": [20, 27], "_pickl": 20, "unpicklingerror": 20, "kei": 20, "hlg_modifi": 20, "g_4_gram": [20, 23, 27, 28], "106000": [20, 23], "496": [20, 23], "875": [20, 23], "212k": 20, "267440": [20, 23], "1253": [20, 23], "535k": 20, "83": [20, 23], "77200": [20, 23], "361": [20, 23], "154k": 20, "554": 20, "260": 20, "7178d67e594bc7fa89c2b331ad7bd1c62a6a9eb4": 20, "tue": 20, "8d93169": 20, "266": [20, 23], "268": [20, 23], "601": 20, "758": 20, "025": 20, "204": 20, "425": 20, "broffel": 20, "osom": 20, "427": 20, "723": 20, "775": 20, "881": 20, "352": 20, "234": 20, "384": 20, "whole": [20, 23, 27, 28, 35, 36], "ngram": [20, 23, 27, 28], "857": 20, "979": 20, "980": 20, "055": 20, "117": 20, "051": 20, "363": 20, "959": [20, 28], "546": 20, "598": 20, "599": [20, 23], "833": 20, "834": 20, "915": 20, "076": 20, "110": 20, "397": 20, "999": [20, 22, 35, 36], "concaten": 20, "bucket": 20, "sampler": 20, "1000": 20, "ctc_decod": 20, "ngram_lm_rescor": 20, "attention_rescor": 20, "kind": [20, 22, 24, 25, 34, 35, 36], "316": 20, "118": 20, "58": 20, "221": 20, "125": [20, 30], "136": 20, "144": 20, "159": [20, 30], "543": 20, "174": 20, "topo": 20, "547": 20, "729": 20, "111": 20, "702": 20, "703": 20, "545": 20, "122": 20, "280": 20, "135": [20, 30], "153": [20, 30], "945": 20, "475": 20, "191": [20, 27, 28], "398": 20, "199": [20, 23], "515": 20, "205": 20, "w": [20, 27, 28], "deseri": 20, "441": 20, "fsaclass": 20, "loadfsa": 20, "const": 20, "string": 20, "c10": 20, "ignor": 20, "attribut": 20, "dummi": 20, "589": 20, "attention_scal": 20, "656": 20, "162": 20, "169": [20, 27, 28], "188": 20, "624": 20, "519": [20, 28], "632": 20, "645": [20, 30], "243": 20, "970": 20, "303": 20, "179": 20, "suitabl": [22, 34, 35, 36], "pruned_transducer_stateless2": [22, 32, 35], "pruned_transducer_stateless4": [22, 32, 35], "pruned_transducer_stateless5": [22, 32, 35], "scroll": [22, 24, 25, 34, 35, 36], "scratch": [22, 24, 25, 34, 35, 36], "paper": [22, 34, 35, 36], "arxiv": [22, 34, 35, 36], "ab": [22, 34, 35, 36], "2206": [22, 34, 35, 36], "13236": [22, 34, 35, 36], "rework": [22, 32, 35], "daniel": [22, 35, 36], "joint": [22, 34, 35, 36], "contrari": [22, 34, 35, 36], "convent": [22, 34, 35, 36], "That": [22, 34, 35, 36], "recurr": [22, 34, 35, 36], "fp16": [22, 24, 25, 34, 35, 36], "half": [22, 35, 36], "2x": [22, 35, 36], "dimens": [22, 35, 36], "littl": [22, 35], "allow": [22, 35], "436000": [22, 24, 25, 34, 35, 36], "438000": [22, 24, 25, 34, 35, 36], "qogspbgsr8kzcrmmie9jgw": 22, "20t15": [22, 34, 35], "4468": [22, 34, 35], "210171": [22, 34, 35], "access": [22, 24, 25, 34, 35, 36], "6008": [22, 24, 25, 34, 35, 36], "localhost": [22, 24, 25, 34, 35, 36], "expos": [22, 24, 25, 34, 35, 36], "proxi": [22, 24, 25, 34, 35, 36], "bind_al": [22, 24, 25, 34, 35, 36], "suggest": [22, 24, 25, 34, 35, 36], "both": [22, 24, 25, 34, 35, 36], "lowest": [22, 24, 25, 34, 35, 36], "fast_beam_search": [22, 24, 34, 35, 36], "474000": [22, 34, 35, 36], "largest": [22, 35, 36], "posterior": [22, 24, 35, 36], "algorithm": [22, 35, 36], "pdf": [22, 25, 35, 36], "1211": [22, 35, 36], "3711": [22, 35, 36], "espnet": [22, 35, 36], "net": [22, 35, 36], "beam_search_transduc": [22, 35, 36], "basicli": [22, 35, 36], "topk": [22, 35, 36], "expand": [22, 35, 36], "mode": [22, 35, 36], "being": [22, 35, 36], "hardcod": [22, 35, 36], "composit": [22, 35, 36], "between": [22, 35, 36], "log_prob": [22, 35, 36], "hard": [22, 32, 35, 36], "2211": [22, 35, 36], "00484": [22, 35, 36], "rnnt": [22, 35, 36], "effici": [22, 35, 36], "fast_beam_search_lg": [22, 35, 36], "trivial": [22, 35, 36], "fast_beam_search_nbest": [22, 35, 36], "random_path": [22, 35, 36], "shortest": [22, 35, 36], "fast_beam_search_nbest_lg": [22, 35, 36], "logic": [22, 35, 36], "includ": [22, 24, 25, 34, 35, 36], "But": [22, 24, 25, 34, 35, 36], "smallest": [22, 34, 35, 36], "icefall_asr_librispeech_tdnn": 23, "lstm_ctc": 23, "flac": 23, "116k": 23, "140k": 23, "343k": 23, "164k": 23, "105k": 23, "174k": 23, "pretraind": 23, "168": 23, "170": 23, "581": 23, "584": [23, 28], "209": 23, "791": 23, "245": 23, "099": 23, "methond": [23, 27, 28], "725": 23, "403": 23, "631": 23, "010": 23, "guidanc": 24, "calcul": [24, 35, 36], "bigger": 24, "threshold": 24, "simpli": 24, "discard": 24, "prevent": 24, "convolut": [24, 32, 35], "similar": [24, 35, 36], "lconv": 24, "encourag": [24, 25, 34], "stabil": [24, 25], "doesn": 24, "warm": [24, 25], "pruned_transducer_stateless7_ctc_b": 24, "xyozukpeqm62hbilud4upa": [24, 25], "ctc_guild_decode_b": 24, "pretrained_ctc": 24, "jit_pretrained_ctc": 24, "yfyeung": 24, "wechat": 25, "zipformer_mmi": 25, "worker": [25, 34], "hp": 25, "zengwei": [25, 34], "tdnn_ligru_ctc": 27, "enough": [27, 28, 30], "luomingshuang": [27, 28], "icefall_asr_timit_tdnn_ligru_ctc": 27, "pretrained_average_9_25": 27, "fdhc0_si1559": [27, 28], "felc0_si756": [27, 28], "fmgd0_si1564": [27, 28], "ffprobe": [27, 28], "show_format": [27, 28], "nistspher": [27, 28], "database_id": [27, 28], "database_vers": [27, 28], "utterance_id": [27, 28], "dhc0_si1559": [27, 28], "sample_min": [27, 28], "4176": [27, 28], "sample_max": [27, 28], "5984": [27, 28], "bitrat": [27, 28], "258": [27, 28], "audio": [27, 28], "pcm_s16le": [27, 28], "hz": [27, 28], "s16": [27, 28], "256": [27, 28], "elc0_si756": [27, 28], "1546": [27, 28], "1989": [27, 28], "mgd0_si1564": [27, 28], "7626": [27, 28], "10573": [27, 28], "660": 27, "183": [27, 28], "695": 27, "697": 27, "210": [27, 28], "829": 27, "sil": [27, 28], "dh": [27, 28], "ih": [27, 28], "uw": [27, 28], "ah": [27, 28], "ii": [27, 28], "z": [27, 28], "aa": [27, 28], "ei": [27, 28], "dx": [27, 28], "uh": [27, 28], "ng": [27, 28], "th": [27, 28], "eh": [27, 28], "jh": [27, 28], "er": [27, 28], "ai": [27, 28], "hh": [27, 28], "aw": 27, "ae": [27, 28], "705": 27, "715": 27, "720": 27, "251": [27, 28], "348": 27, "ch": 27, "icefall_asr_timit_tdnn_lstm_ctc": 28, "pretrained_average_16_25": 28, "816": 28, "827": 28, "387": 28, "unk": 28, "739": 28, "971": 28, "977": 28, "978": 28, "981": 28, "ow": 28, "ykubhb5wrmosxykid1z9eg": 30, "23t23": 30, "sinc": [30, 34], "icefall_asr_yesno_tdnn": 30, "l_disambig": 30, "lexicon_disambig": 30, "arpa": 30, "0_0_0_1_0_0_0_1": 30, "0_0_1_0_0_0_1_0": 30, "0_0_1_0_0_1_1_1": 30, "0_0_1_0_1_0_0_1": 30, "0_0_1_1_0_0_0_1": 30, "0_0_1_1_0_1_1_0": 30, "0_0_1_1_1_0_0_0": 30, "0_0_1_1_1_1_0_0": 30, "0_1_0_0_0_1_0_0": 30, "0_1_0_0_1_0_1_0": 30, "0_1_0_1_0_0_0_0": 30, "0_1_0_1_1_1_0_0": 30, "0_1_1_0_0_1_1_1": 30, "0_1_1_1_0_0_1_0": 30, "0_1_1_1_1_0_1_0": 30, "1_0_0_0_0_0_0_0": 30, "1_0_0_0_0_0_1_1": 30, "1_0_0_1_0_1_1_1": 30, "1_0_1_1_0_1_1_1": 30, "1_0_1_1_1_1_0_1": 30, "1_1_0_0_0_1_1_1": 30, "1_1_0_0_1_0_1_1": 30, "1_1_0_1_0_1_0_0": 30, "1_1_0_1_1_0_0_1": 30, "1_1_0_1_1_1_1_0": 30, "1_1_1_0_0_1_0_1": 30, "1_1_1_0_1_0_1_0": 30, "1_1_1_1_0_0_1_0": 30, "1_1_1_1_1_0_0_0": 30, "1_1_1_1_1_1_1_1": 30, "54080": 30, "507": 30, "108k": 30, "No": 30, "ye": 30, "hebrew": 30, "NO": 30, "621": 30, "119": 30, "127": 30, "650": 30, "139": 30, "143": 30, "198": 30, "181": 30, "186": 30, "187": 30, "287": 30, "correctli": 30, "simplest": 30, "former": 32, "idea": 32, "achiev": 32, "mask": [32, 35, 36], "wenet": 32, "did": 32, "argument": 32, "adapt": 32, "complic": 32, "techniqu": 32, "bank": 32, "compon": 32, "memor": 32, "histori": 32, "introduc": 32, "variant": 32, "pruned_stateless_emformer_rnnt2": 32, "conv_emformer_transducer_stateless": 32, "convemform": 32, "ourself": 32, "mechan": 32, "conv_emformer_transducer_stateless2": 32, "onlin": 34, "lstm_transducer_stateless": 34, "architectur": 34, "lower": 34, "prepare_giga_speech": 34, "cj2vtpiwqhkn9q1tx6ptpg": 34, "hidden": 34, "1024": 34, "pnnx": 34, "submodul": 34, "updat": 34, "recurs": 34, "init": 34, "bdist_wheel": 34, "lh": 34, "dist": 34, "j4": 34, "pwd": 34, "third": 34, "param": 34, "abl": 34, "extern": 34, "stateless2": 34, "compar": 35, "dynam": [35, 36], "chunk": [35, 36], "causal": 35, "short": [35, 36], "2012": 35, "05481": 35, "flag": 35, "indic": [35, 36], "whether": 35, "must": 35, "sequenc": [35, 36], "uniformli": [35, 36], "most": [35, 36], "seen": [35, 36], "97vkxf80ru61cnp2alwzzg": 35, "streaming_decod": [35, 36], "acoust": [35, 36], "wise": [35, 36], "subsampl": [35, 36], "equal": [35, 36], "where": 35, "parallel": [35, 36], "bath": [35, 36], "parallelli": [35, 36], "seem": 35, "benefit": 35, "might": [35, 36], "mismatch": 35, "yourself": [35, 36], "mdoel": 35, "pruned_transducer_stateless7_stream": 36, "len": 36, "320m": 36, "550": 36, "scriptmodul": 36, "jit_trace_export": 36, "jit_trace_pretrain": 36, "task": 37}, "objects": {}, "objtypes": {}, "objnames": {}, "titleterms": {"follow": 0, "code": 0, "style": 0, "contribut": [1, 3], "document": 1, "how": [2, 9, 11, 12, 13], "creat": [2, 8], "recip": [2, 37], "data": [2, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "prepar": [2, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "train": [2, 5, 8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "decod": [2, 8, 9, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "pre": [2, 5, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "model": [2, 5, 9, 11, 12, 13, 14, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "huggingfac": [4, 6], "space": 6, "youtub": [6, 8], "video": [6, 8], "icefal": [7, 8], "content": [7, 37], "instal": [8, 15, 17, 18, 20, 23, 27, 28], "0": 8, "pytorch": 8, "torchaudio": 8, "1": [8, 15, 17, 18, 20], "k2": 8, "2": [8, 15, 17, 18, 20], "lhots": 8, "3": [8, 15, 17, 20], "download": [8, 15, 17, 18, 20, 22, 23, 24, 25, 27, 28, 30, 34, 35, 36], "exampl": [8, 15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "virtual": 8, "environ": 8, "activ": 8, "your": 8, "4": 8, "5": 8, "test": 8, "export": [9, 10, 11, 12, 13, 14, 22, 24, 25, 34, 35, 36], "state_dict": [9, 22, 24, 25, 34, 35, 36], "when": [9, 11, 12, 13], "us": [9, 11, 12, 13, 22, 24, 25, 34, 35, 36], "run": 9, "py": 9, "ncnn": [10, 34], "onnx": 11, "torch": [12, 13, 22, 24, 25, 34, 35, 36], "jit": [12, 13, 22, 24, 25, 34, 35, 36], "script": [12, 22, 24, 25, 35, 36], "trace": [13, 34, 36], "conform": [15, 20, 32], "ctc": [15, 18, 20, 23, 24, 27, 28, 30], "configur": [15, 18, 20, 22, 24, 25, 34, 35, 36], "option": [15, 18, 20, 22, 24, 25, 34, 35, 36], "log": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "usag": [15, 17, 18, 20, 22, 24, 25, 34, 35, 36], "case": [15, 17, 18, 20], "kaldifeat": [15, 17, 18, 20, 23, 27, 28, 30], "hlg": [15, 18, 20], "attent": [15, 20], "rescor": [15, 20], "colab": [15, 17, 18, 20, 23, 27, 28, 30], "notebook": [15, 17, 18, 20, 23, 27, 28, 30], "deploy": [15, 20], "c": [15, 20], "aishel": 16, "stateless": 17, "transduc": [17, 22, 34, 35, 36], "The": 17, "loss": 17, "todo": 17, "greedi": 17, "search": 17, "beam": 17, "modifi": 17, "tdnn": [18, 23, 27, 28, 30], "lstm": [18, 23, 28, 34], "non": 19, "stream": [19, 31, 32, 35, 36], "asr": [19, 31], "lm": 20, "comput": 20, "wer": 20, "n": 20, "gram": 20, "librispeech": [21, 33], "prune": [22, 35], "statelessx": [22, 35], "pretrain": [22, 24, 25, 34, 35, 36], "deploi": [22, 35, 36], "sherpa": [22, 35, 36], "infer": [23, 27, 28, 30], "zipform": [24, 25, 36], "blank": 24, "skip": 24, "mmi": 25, "timit": 26, "ligru": 27, "yesno": 29, "introduct": 32, "emform": 32, "which": 34, "simul": [35, 36], "real": [35, 36], "tabl": 37}, "envversion": {"sphinx.domains.c": 2, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 8, "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": 57}, "alltitles": {"Follow the code style": [[0, "follow-the-code-style"]], "Contributing to Documentation": [[1, "contributing-to-documentation"]], "How to create a recipe": [[2, "how-to-create-a-recipe"]], "Data Preparation": [[2, "data-preparation"], [17, "data-preparation"]], "Training": [[2, "training"], [8, "training"], [15, "training"], [17, "training"], [18, "training"], [20, "training"], [22, "training"], [23, "training"], [24, "training"], [25, "training"], [27, "training"], [28, "training"], [30, "training"], [34, "training"], [35, "training"], [36, "training"]], "Decoding": [[2, "decoding"], [8, "decoding"], [15, "decoding"], [17, "decoding"], [18, "decoding"], [20, "decoding"], [22, "decoding"], [23, "decoding"], [24, "decoding"], [25, "decoding"], [27, "decoding"], [28, "decoding"], [30, "decoding"], [34, "decoding"], [35, "decoding"], [36, "decoding"]], "Pre-trained model": [[2, "pre-trained-model"]], "Contributing": [[3, "contributing"]], "Huggingface": [[4, "huggingface"]], "Pre-trained models": [[5, "pre-trained-models"]], "Huggingface spaces": [[6, "huggingface-spaces"]], "YouTube Video": [[6, "youtube-video"], [8, "youtube-video"]], "Icefall": [[7, "icefall"]], "Contents:": [[7, null]], "Installation": [[8, "installation"]], "(0) Install PyTorch and torchaudio": [[8, "install-pytorch-and-torchaudio"]], "(1) Install k2": [[8, "install-k2"]], "(2) Install lhotse": [[8, "install-lhotse"]], "(3) Download icefall": [[8, "download-icefall"]], "Installation example": [[8, "installation-example"]], "(1) Create a virtual environment": [[8, "create-a-virtual-environment"]], "(2) Activate your virtual environment": [[8, "activate-your-virtual-environment"]], "(3) Install k2": [[8, "id1"]], "(4) Install lhotse": [[8, "id2"]], "(5) Download icefall": [[8, "id3"]], "Test Your Installation": [[8, "test-your-installation"]], "Data preparation": [[8, "data-preparation"], [15, "data-preparation"], [18, "data-preparation"], [20, "data-preparation"], [22, "data-preparation"], [23, "data-preparation"], [24, "data-preparation"], [25, "data-preparation"], [27, "data-preparation"], [28, "data-preparation"], [30, "data-preparation"], [34, "data-preparation"], [35, "data-preparation"], [36, "data-preparation"]], "Export model.state_dict()": [[9, "export-model-state-dict"], [22, "export-model-state-dict"], [24, "export-model-state-dict"], [25, "export-model-state-dict"], [34, "export-model-state-dict"], [35, "export-model-state-dict"], [36, "export-model-state-dict"]], "When to use it": [[9, "when-to-use-it"], [11, "when-to-use-it"], [12, "when-to-use-it"], [13, "when-to-use-it"]], "How to export": [[9, "how-to-export"], [11, "how-to-export"], [12, "how-to-export"], [13, "how-to-export"]], "How to use the exported model": [[9, "how-to-use-the-exported-model"], [11, "how-to-use-the-exported-model"], [12, "how-to-use-the-exported-model"]], "Use the exported model to run decode.py": [[9, "use-the-exported-model-to-run-decode-py"]], "Export to ncnn": [[10, "export-to-ncnn"]], "Export to ONNX": [[11, "export-to-onnx"]], "Export model with torch.jit.script()": [[12, "export-model-with-torch-jit-script"]], "Export model with torch.jit.trace()": [[13, "export-model-with-torch-jit-trace"]], "How to use the exported models": [[13, "how-to-use-the-exported-models"]], "Model export": [[14, "model-export"]], "Conformer CTC": [[15, "conformer-ctc"], [20, "conformer-ctc"]], "Configurable options": [[15, "configurable-options"], [18, "configurable-options"], [20, "configurable-options"], [22, "configurable-options"], [24, "configurable-options"], [25, "configurable-options"], [34, "configurable-options"], [35, "configurable-options"], [36, "configurable-options"]], "Pre-configured options": [[15, "pre-configured-options"], [18, "pre-configured-options"], [20, "pre-configured-options"], [22, "pre-configured-options"], [24, "pre-configured-options"], [25, "pre-configured-options"], [34, "pre-configured-options"], [35, "pre-configured-options"], [36, "pre-configured-options"]], "Training logs": [[15, "training-logs"], [17, "training-logs"], [18, "training-logs"], [20, "training-logs"], [22, "training-logs"], [24, "training-logs"], [25, "training-logs"], [34, "training-logs"], [35, "training-logs"], [36, "training-logs"]], "Usage examples": [[15, "usage-examples"], [17, "usage-examples"], [18, "usage-examples"], [20, "usage-examples"]], "Case 1": [[15, "case-1"], [17, "case-1"], [18, "case-1"], [20, "case-1"]], "Case 2": [[15, "case-2"], [17, "case-2"], [18, "case-2"], [20, "case-2"]], "Case 3": [[15, "case-3"], [17, "case-3"], [20, "case-3"]], "Pre-trained Model": [[15, "pre-trained-model"], [17, "pre-trained-model"], [18, "pre-trained-model"], [20, "pre-trained-model"], [23, "pre-trained-model"], [27, "pre-trained-model"], [28, "pre-trained-model"], [30, "pre-trained-model"]], "Install kaldifeat": [[15, "install-kaldifeat"], [17, "install-kaldifeat"], [18, "install-kaldifeat"], [20, "install-kaldifeat"], [23, "install-kaldifeat"], [27, "install-kaldifeat"], [28, "install-kaldifeat"]], "Download the pre-trained model": [[15, "download-the-pre-trained-model"], [17, "download-the-pre-trained-model"], [18, "download-the-pre-trained-model"], [20, "download-the-pre-trained-model"], [23, "download-the-pre-trained-model"], [27, "download-the-pre-trained-model"], [28, "download-the-pre-trained-model"], [30, "download-the-pre-trained-model"]], "Usage": [[15, "usage"], [17, "usage"], [18, "usage"], [20, "usage"]], "CTC decoding": [[15, "ctc-decoding"], [20, "ctc-decoding"], [20, "id2"]], "HLG decoding": [[15, "hlg-decoding"], [15, "id2"], [18, "hlg-decoding"], [20, "hlg-decoding"], [20, "id3"]], "HLG decoding + attention decoder rescoring": [[15, "hlg-decoding-attention-decoder-rescoring"]], "Colab notebook": [[15, "colab-notebook"], [17, "colab-notebook"], [18, "colab-notebook"], [20, "colab-notebook"], [23, "colab-notebook"], [27, "colab-notebook"], [28, "colab-notebook"], [30, "colab-notebook"]], "Deployment with C++": [[15, "deployment-with-c"], [20, "deployment-with-c"]], "aishell": [[16, "aishell"]], "Stateless Transducer": [[17, "stateless-transducer"]], "The Model": [[17, "the-model"]], "The Loss": [[17, "the-loss"]], "Todo": [[17, "id1"]], "Greedy search": [[17, "greedy-search"]], "Beam search": [[17, "beam-search"]], "Modified Beam search": [[17, "modified-beam-search"]], "TDNN-LSTM CTC": [[18, "tdnn-lstm-ctc"]], "Non Streaming ASR": [[19, "non-streaming-asr"]], "HLG decoding + LM rescoring": [[20, "hlg-decoding-lm-rescoring"]], "HLG decoding + LM rescoring + attention decoder rescoring": [[20, "hlg-decoding-lm-rescoring-attention-decoder-rescoring"]], "Compute WER with the pre-trained model": [[20, "compute-wer-with-the-pre-trained-model"]], "HLG decoding + n-gram LM rescoring": [[20, "hlg-decoding-n-gram-lm-rescoring"]], "HLG decoding + n-gram LM rescoring + attention decoder rescoring": [[20, "hlg-decoding-n-gram-lm-rescoring-attention-decoder-rescoring"]], "LibriSpeech": [[21, "librispeech"], [33, "librispeech"]], "Pruned transducer statelessX": [[22, "pruned-transducer-statelessx"], [35, "pruned-transducer-statelessx"]], "Usage example": [[22, "usage-example"], [24, "usage-example"], [25, "usage-example"], [34, "usage-example"], [35, "usage-example"], [36, "usage-example"]], "Export Model": [[22, "export-model"], [35, "export-model"], [36, "export-model"]], "Export model using torch.jit.script()": [[22, "export-model-using-torch-jit-script"], [24, "export-model-using-torch-jit-script"], [25, "export-model-using-torch-jit-script"], [35, "export-model-using-torch-jit-script"], [36, "export-model-using-torch-jit-script"]], "Download pretrained models": [[22, "download-pretrained-models"], [24, "download-pretrained-models"], [25, "download-pretrained-models"], [34, "download-pretrained-models"], [35, "download-pretrained-models"], [36, "download-pretrained-models"]], "Deploy with Sherpa": [[22, "deploy-with-sherpa"], [35, "deploy-with-sherpa"], [36, "deploy-with-sherpa"]], "TDNN-LSTM-CTC": [[23, "tdnn-lstm-ctc"], [28, "tdnn-lstm-ctc"]], "Inference with a pre-trained model": [[23, "inference-with-a-pre-trained-model"], [27, "inference-with-a-pre-trained-model"], [28, "inference-with-a-pre-trained-model"], [30, "inference-with-a-pre-trained-model"]], "Zipformer CTC Blank Skip": [[24, "zipformer-ctc-blank-skip"]], "Export models": [[24, "export-models"], [25, "export-models"], [34, "export-models"]], "Zipformer MMI": [[25, "zipformer-mmi"]], "TIMIT": [[26, "timit"]], "TDNN-LiGRU-CTC": [[27, "tdnn-ligru-ctc"]], "YesNo": [[29, "yesno"]], "TDNN-CTC": [[30, "tdnn-ctc"]], "Download kaldifeat": [[30, "download-kaldifeat"]], "Streaming ASR": [[31, "streaming-asr"]], "Introduction": [[32, "introduction"]], "Streaming Conformer": [[32, "streaming-conformer"]], "Streaming Emformer": [[32, "streaming-emformer"]], "LSTM Transducer": [[34, "lstm-transducer"]], "Which model to use": [[34, "which-model-to-use"]], "Export model using torch.jit.trace()": [[34, "export-model-using-torch-jit-trace"], [36, "export-model-using-torch-jit-trace"]], "Export model for ncnn": [[34, "export-model-for-ncnn"]], "Simulate streaming decoding": [[35, "simulate-streaming-decoding"], [36, "simulate-streaming-decoding"]], "Real streaming decoding": [[35, "real-streaming-decoding"], [36, "real-streaming-decoding"]], "Zipformer Transducer": [[36, "zipformer-transducer"]], "Recipes": [[37, "recipes"]], "Table of Contents": [[37, null]]}, "indexentries": {}})