Upload folder using huggingface_hub

README.md

@@ -0,0 +1,30 @@
+---
+library_name: mlx
+tags:
+- mlx
+- automatic-speech-recognition
+- speech
+- audio
+- FastConformer
+- Conformer
+- Parakeet
+license: cc-by-4.0
+pipeline_tag: automatic-speech-recognition
+base_model: nvidia/parakeet-ctc-1.1b
+---
+
+# mlx-community/parakeet-ctc-1.1b
+
+This model was converted to MLX format from [nvidia/parakeet-ctc-1.1b](https://huggingface.co/nvidia/parakeet-ctc-1.1b) using [the conversion script](https://gist.github.com/senstella/77178bb5d6ec67bf8c54705a5f490bed). Please refer to [original model card](https://huggingface.co/nvidia/parakeet-ctc-1.1b) for more details on the model.
+
+## Use with mlx
+
+### parakeet-mlx
+
+```bash
+pip install -U parakeet-mlx
+```
+
+```bash
+parakeet-mlx audio.wav --model mlx-community/parakeet-ctc-1.1b
+```

config.json

@@ -0,0 +1,1212 @@
+{
+  "sample_rate": 16000,
+  "log_prediction": true,
+  "ctc_reduction": "mean_volume",
+  "skip_nan_grad": false,
+  "model_defaults": {
+    "enc_hidden": 1024,
+    "pred_hidden": 640,
+    "joint_hidden": 640
+  },
+  "train_ds": {
+    "manifest_filepath": "/disk1/NVIDIA/datasets/LibriSpeech_NeMo/librivox-train-all.json",
+    "sample_rate": 16000,
+    "batch_size": 16,
+    "shuffle": true,
+    "num_workers": 8,
+    "pin_memory": true,
+    "use_start_end_token": false,
+    "trim_silence": false,
+    "max_duration": 16.7,
+    "min_duration": 0.1,
+    "is_tarred": false,
+    "tarred_audio_filepaths": null,
+    "shuffle_n": 2048,
+    "bucketing_strategy": "fully_randomized",
+    "bucketing_batch_size": null
+  },
+  "validation_ds": {
+    "manifest_filepath": "/disk1/NVIDIA/datasets/LibriSpeech_NeMo/librivox-dev-clean.json",
+    "sample_rate": 16000,
+    "batch_size": 16,
+    "shuffle": false,
+    "use_start_end_token": false,
+    "num_workers": 8,
+    "pin_memory": true
+  },
+  "test_ds": {
+    "manifest_filepath": null,
+    "sample_rate": 16000,
+    "batch_size": 16,
+    "shuffle": false,
+    "use_start_end_token": false,
+    "num_workers": 8,
+    "pin_memory": true
+  },
+  "tokenizer": {
+    "dir": "/disk3/datasets/suno_asr_set3_tokenizer/tokenizer_spe_bpe_v1024/",
+    "type": "bpe",
+    "model_path": "nemo:dd0d316f51614e1ca87ff4c94ba2babf_tokenizer.model",
+    "vocab_path": "nemo:d60196a1e3454ddeb787f9e67be6c509_vocab.txt",
+    "spe_tokenizer_vocab": "nemo:053e7b77f4bb4ae1bb197a0ad4fc6670_tokenizer.vocab"
+  },
+  "preprocessor": {
+    "_target_": "nemo.collections.asr.modules.AudioToMelSpectrogramPreprocessor",
+    "sample_rate": 16000,
+    "normalize": "per_feature",
+    "window_size": 0.025,
+    "window_stride": 0.01,
+    "window": "hann",
+    "features": 80,
+    "n_fft": 512,
+    "log": true,
+    "frame_splicing": 1,
+    "dither": 1e-05,
+    "pad_to": 0,
+    "pad_value": 0.0
+  },
+  "spec_augment": {
+    "_target_": "nemo.collections.asr.modules.SpectrogramAugmentation",
+    "freq_masks": 2,
+    "time_masks": 10,
+    "freq_width": 27,
+    "time_width": 0.05
+  },
+  "encoder": {
+    "_target_": "nemo.collections.asr.modules.ConformerEncoder",
+    "feat_in": 80,
+    "feat_out": -1,
+    "n_layers": 42,
+    "d_model": 1024,
+    "subsampling": "dw_striding",
+    "subsampling_factor": 8,
+    "subsampling_conv_channels": 256,
+    "causal_downsampling": false,
+    "ff_expansion_factor": 4,
+    "self_attention_model": "rel_pos",
+    "n_heads": 8,
+    "att_context_size": [
+      -1,
+      -1
+    ],
+    "att_context_style": "regular",
+    "xscaling": true,
+    "untie_biases": true,
+    "pos_emb_max_len": 5000,
+    "conv_kernel_size": 9,
+    "conv_norm_type": "batch_norm",
+    "conv_context_size": null,
+    "dropout": 0.1,
+    "dropout_pre_encoder": 0.1,
+    "dropout_emb": 0.0,
+    "dropout_att": 0.1,
+    "stochastic_depth_drop_prob": 0.0,
+    "stochastic_depth_mode": "linear",
+    "stochastic_depth_start_layer": 1
+  },
+  "decoder": {
+    "_target_": "nemo.collections.asr.modules.ConvASRDecoder",
+    "feat_in": 1024,
+    "num_classes": 1024,
+    "vocabulary": [
+      "<unk>",
+      "\u2581t",
+      "\u2581th",
+      "\u2581a",
+      "\u2581i",
+      "\u2581the",
+      "re",
+      "\u2581w",
+      "\u2581s",
+      "\u2581o",
+      "in",
+      "at",
+      "er",
+      "ou",
+      "nd",
+      "\u2581c",
+      "\u2581b",
+      "\u2581h",
+      "on",
+      "\u2581m",
+      "\u2581f",
+      "ing",
+      "\u2581to",
+      "en",
+      "\u2581p",
+      "\u2581and",
+      "\u2581d",
+      "es",
+      "or",
+      "an",
+      "ll",
+      "\u2581y",
+      "\u2581l",
+      "ed",
+      "\u2581of",
+      "\u2581in",
+      "it",
+      "is",
+      "\u2581you",
+      "\u2581that",
+      "ar",
+      "\u2581g",
+      "\u2581n",
+      "as",
+      "om",
+      "\u2581it",
+      "ic",
+      "ve",
+      "\u2581e",
+      "\u2581wh",
+      "\u2581be",
+      "us",
+      "le",
+      "al",
+      "ion",
+      "ow",
+      "\u2581we",
+      "\u2581re",
+      "\u2581is",
+      "ut",
+      "ot",
+      "ent",
+      "\u2581on",
+      "et",
+      "\u2581ha",
+      "ay",
+      "ct",
+      "\u2581he",
+      "id",
+      "\u2581for",
+      "\u2581st",
+      "ver",
+      "ly",
+      "ro",
+      "ig",
+      "\u2581so",
+      "ld",
+      "\u2581this",
+      "ke",
+      "\u2581u",
+      "se",
+      "all",
+      "st",
+      "ur",
+      "ce",
+      "ch",
+      "im",
+      "ith",
+      "\u2581as",
+      "\u2581k",
+      "\u2581an",
+      "\u2581was",
+      "\u2581j",
+      "\u2581with",
+      "ir",
+      "\u2581go",
+      "ra",
+      "\u2581do",
+      "\u2581have",
+      "\u2581li",
+      "\u2581sh",
+      "\u2581se",
+      "\u2581they",
+      "\u2581are",
+      "am",
+      "ht",
+      "\u2581but",
+      "ation",
+      "\u2581not",
+      "th",
+      "\u2581r",
+      "ally",
+      "ad",
+      "ust",
+      "\u2581or",
+      "\u2581com",
+      "ould",
+      "\u2581can",
+      "ill",
+      "\u2581ne",
+      "ight",
+      "\u2581ch",
+      "\u2581de",
+      "\u2581con",
+      "\u2581at",
+      "\u2581mo",
+      "ant",
+      "oo",
+      "il",
+      "\u2581me",
+      "\u2581what",
+      "\u2581there",
+      "ter",
+      "pe",
+      "\u2581ab",
+      "\u2581su",
+      "ere",
+      "ck",
+      "\u2581pro",
+      "\u2581al",
+      "\u2581fr",
+      "\u2581kn",
+      "\u2581all",
+      "ers",
+      "\u2581like",
+      "ge",
+      "\u2581ex",
+      "\u2581som",
+      "ul",
+      "\u2581your",
+      "\u2581v",
+      "pp",
+      "use",
+      "\u2581if",
+      "ess",
+      "ate",
+      "est",
+      "\u2581know",
+      "out",
+      "if",
+      "\u2581just",
+      "ment",
+      "qu",
+      "op",
+      "ain",
+      "\u2581one",
+      "ol",
+      "ri",
+      "art",
+      "very",
+      "\u2581wor",
+      "ive",
+      "ist",
+      "\u2581my",
+      "nt",
+      "ab",
+      "\u2581from",
+      "ort",
+      "\u2581ma",
+      "\u2581about",
+      "res",
+      "ity",
+      "\u2581out",
+      "\u2581bec",
+      "\u2581le",
+      "our",
+      "od",
+      "and",
+      "ink",
+      "ie",
+      "\u2581up",
+      "ind",
+      "os",
+      "un",
+      "ause",
+      "oug",
+      "um",
+      "\u2581some",
+      "\u2581int",
+      "\u2581by",
+      "\u2581pl",
+      "\u2581get",
+      "el",
+      "ard",
+      "\u2581when",
+      "\u2581don",
+      "her",
+      "\u2581will",
+      "\u2581us",
+      "\u2581would",
+      "ook",
+      "ies",
+      "ich",
+      "\u2581because",
+      "\u2581think",
+      "em",
+      "\u2581pe",
+      "\u2581his",
+      "ack",
+      "\u2581then",
+      "\u2581our",
+      "ide",
+      "\u2581tim",
+      "\u2581how",
+      "ven",
+      "\u2581tr",
+      "\u2581who",
+      "\u2581them",
+      "ure",
+      "\u2581ar",
+      "\u2581ye",
+      "\u2581more",
+      "\u2581going",
+      "ect",
+      "\u2581sa",
+      "\u2581cl",
+      "\u2581had",
+      "\u2581now",
+      "\u2581which",
+      "\u2581here",
+      "ous",
+      "\u2581their",
+      "\u2581tw",
+      "so",
+      "\u2581has",
+      "ud",
+      "\u2581co",
+      "\u2581ta",
+      "ound",
+      "\u2581were",
+      "ast",
+      "\u2581peop",
+      "ough",
+      "\u2581no",
+      "\u2581really",
+      "\u2581any",
+      "\u2581people",
+      "\u2581want",
+      "\u2581she",
+      "\u2581en",
+      "\u2581fa",
+      "\u2581te",
+      "ame",
+      "ine",
+      "\u2581qu",
+      "red",
+      "\u2581im",
+      "\u2581right",
+      "ther",
+      "\u2581act",
+      "\u2581thing",
+      "king",
+      "ose",
+      "\u2581ad",
+      "\u2581see",
+      "\u2581time",
+      "\u2581these",
+      "ci",
+      "one",
+      "\u2581say",
+      "\u2581also",
+      "\u2581fe",
+      "per",
+      "\u2581ag",
+      "\u2581man",
+      "ore",
+      "\u2581un",
+      "pt",
+      "\u2581her",
+      "\u2581look",
+      "ong",
+      "ice",
+      "\u2581very",
+      "ff",
+      "ions",
+      "\u2581comp",
+      "\u2581did",
+      "itt",
+      "\u2581well",
+      "\u2581other",
+      "iv",
+      "ase",
+      "ree",
+      "hing",
+      "\u2581lo",
+      "reat",
+      "\u2581cont",
+      "\u2581part",
+      "\u2581into",
+      "nder",
+      "\u2581been",
+      "are",
+      "\u2581am",
+      "ans",
+      "\u2581sp",
+      "\u2581two",
+      "ue",
+      "\u2581way",
+      "age",
+      "\u2581where",
+      "ite",
+      "\u2581dis",
+      "\u2581than",
+      "\u2581every",
+      "\u2581pr",
+      "\u2581po",
+      "ag",
+      "\u2581need",
+      "ach",
+      "iff",
+      "ence",
+      "pl",
+      "own",
+      "\u2581ac",
+      "ble",
+      "\u2581over",
+      "iz",
+      "\u2581work",
+      "\u2581res",
+      "\u2581make",
+      "\u2581could",
+      "\u2581off",
+      "ually",
+      "\u2581ro",
+      "\u2581back",
+      "able",
+      "ip",
+      "ry",
+      "\u2581him",
+      "\u2581cour",
+      "ber",
+      "\u2581pre",
+      "\u2581fir",
+      "\u2581spe",
+      "ap",
+      "ars",
+      "\u2581diff",
+      "ire",
+      "\u2581somet",
+      "\u2581imp",
+      "\u2581those",
+      "\u2581comm",
+      "ance",
+      "ick",
+      "\u2581even",
+      "ated",
+      "way",
+      "sel",
+      "\u2581let",
+      "\u2581br",
+      "ty",
+      "\u2581per",
+      "int",
+      "\u2581first",
+      "\u2581thr",
+      "\u2581under",
+      "ah",
+      "\u2581may",
+      "\u2581cou",
+      "\u2581new",
+      "ress",
+      "act",
+      "\u2581gr",
+      "ep",
+      "\u2581said",
+      "ations",
+      "\u2581good",
+      "ace",
+      "ass",
+      "\u2581does",
+      "orm",
+      "ish",
+      "\u2581af",
+      "ving",
+      "co",
+      "\u2581app",
+      "\u2581lot",
+      "\u2581things",
+      "\u2581tra",
+      "ittle",
+      "\u2581bl",
+      "\u2581little",
+      "\u2581mu",
+      "cess",
+      "fe",
+      "ome",
+      "\u2581inc",
+      "\u2581differe",
+      "ary",
+      "ical",
+      "\u2581only",
+      "ult",
+      "\u2581again",
+      "\u2581got",
+      "ens",
+      "\u2581gu",
+      "\u2581kind",
+      "\u2581much",
+      "ord",
+      "\u2581through",
+      "ition",
+      "ild",
+      "\u2581down",
+      "\u2581actually",
+      "\u2581something",
+      "ang",
+      "ru",
+      "ces",
+      "\u2581fl",
+      "ile",
+      "ater",
+      "\u2581ra",
+      "\u2581take",
+      "ict",
+      "ign",
+      "\u2581sc",
+      "vel",
+      "\u2581bet",
+      "\u2581tal",
+      "\u2581yeah",
+      "\u2581use",
+      "fore",
+      "\u2581bu",
+      "\u2581start",
+      "ory",
+      "be",
+      "\u2581day",
+      "wn",
+      "xt",
+      "ia",
+      "ak",
+      "\u2581after",
+      "\u2581should",
+      "\u2581fo",
+      "\u2581ho",
+      "\u2581hel",
+      "\u2581ind",
+      "\u2581uh",
+      "na",
+      "ial",
+      "other",
+      "\u2581ke",
+      "\u2581call",
+      "\u2581most",
+      "\u2581ok",
+      "\u2581different",
+      "\u2581em",
+      "ting",
+      "ple",
+      "\u2581being",
+      "\u2581bo",
+      "ning",
+      "\u2581too",
+      "ors",
+      "\u2581happ",
+      "ark",
+      "og",
+      "\u2581help",
+      "\u2581rem",
+      "du",
+      "ction",
+      "ood",
+      "\u2581ser",
+      "ether",
+      "ious",
+      "\u2581mean",
+      "\u2581many",
+      "\u2581court",
+      "\u2581bel",
+      "ade",
+      "\u2581la",
+      "ved",
+      "\u2581des",
+      "\u2581rec",
+      "\u2581jo",
+      "\u2581dec",
+      "ves",
+      "\u2581before",
+      "\u2581put",
+      "self",
+      "\u2581point",
+      "te",
+      "\u2581ev",
+      "form",
+      "ents",
+      "\u2581add",
+      "ody",
+      "thing",
+      "\u2581case",
+      "\u2581pers",
+      "\u2581cons",
+      "iss",
+      "\u2581three",
+      "oth",
+      "\u2581ph",
+      "\u2581come",
+      "\u2581find",
+      "\u2581why",
+      "ull",
+      "\u2581show",
+      "\u2581bas",
+      "\u2581great",
+      "ily",
+      "\u2581rel",
+      "\u2581sm",
+      "\u2581its",
+      "\u2581fact",
+      "\u2581pos",
+      "ool",
+      "ments",
+      "ise",
+      "nds",
+      "ys",
+      "\u2581try",
+      "ual",
+      "ful",
+      "erm",
+      "\u2581inter",
+      "ons",
+      "\u2581quest",
+      "\u2581sub",
+      "we",
+      "vers",
+      "\u2581supp",
+      "\u2581feel",
+      "\u2581same",
+      "ub",
+      "ates",
+      "urn",
+      "ert",
+      "\u2581inv",
+      "day",
+      "\u2581rep",
+      "igh",
+      "\u2581sy",
+      "\u2581inst",
+      "\u2581long",
+      "\u2581still",
+      "\u2581okay",
+      "ft",
+      "ific",
+      "atch",
+      "ought",
+      "ath",
+      "\u2581own",
+      "\u2581made",
+      "ix",
+      "ced",
+      "ks",
+      "lic",
+      "\u2581wr",
+      "de",
+      "\u2581cr",
+      "\u2581att",
+      "\u2581ob",
+      "\u2581world",
+      "\u2581sure",
+      "ward",
+      "\u2581bit",
+      "\u2581life",
+      "\u2581person",
+      "\u2581pres",
+      "ph",
+      "\u2581vide",
+      "\u2581reg",
+      "\u2581end",
+      "ject",
+      "ange",
+      "\u2581fin",
+      "ied",
+      "pect",
+      "\u2581didn",
+      "\u2581around",
+      "ian",
+      "\u2581car",
+      "ible",
+      "\u2581sim",
+      "ever",
+      "\u2581sch",
+      "ating",
+      "\u2581pol",
+      "\u2581set",
+      "\u2581oh",
+      "cy",
+      "\u2581real",
+      "\u2581import",
+      "\u2581count",
+      "\u2581um",
+      "\u2581next",
+      "cial",
+      "les",
+      "\u2581hu",
+      "\u2581acc",
+      "\u2581might",
+      "\u2581ent",
+      "\u2581doing",
+      "\u2581ins",
+      "\u2581gen",
+      "\u2581play",
+      "\u2581cle",
+      "\u2581another",
+      "ady",
+      "ular",
+      "ib",
+      "ways",
+      "ered",
+      "ility",
+      "ities",
+      "\u2581op",
+      "\u2581def",
+      "\u2581years",
+      "\u2581never",
+      "ower",
+      "ram",
+      "\u2581tell",
+      "\u2581sl",
+      "onna",
+      "ail",
+      "ren",
+      "ute",
+      "\u2581gonna",
+      "\u2581big",
+      "\u2581give",
+      "der",
+      "ount",
+      "\u2581ap",
+      "kes",
+      "\u2581state",
+      "\u2581cor",
+      "\u2581min",
+      "ically",
+      "\u2581mon",
+      "\u2581fam",
+      "\u2581important",
+      "\u2581always",
+      "\u2581high",
+      "\u2581four",
+      "\u2581gra",
+      "\u2581ca",
+      "\u2581stud",
+      "\u2581dist",
+      "\u2581talk",
+      "\u2581num",
+      "\u2581str",
+      "\u2581today",
+      "ract",
+      "\u2581while",
+      "ason",
+      "\u2581iss",
+      "\u2581sur",
+      "\u2581char",
+      "\u2581last",
+      "oy",
+      "ited",
+      "\u2581exper",
+      "\u2581place",
+      "\u2581tri",
+      "\u2581ear",
+      "\u2581belie",
+      "\u2581able",
+      "\u2581underst",
+      "\u2581che",
+      "\u2581both",
+      "ug",
+      "\u2581doesn",
+      "\u2581keep",
+      "\u2581happen",
+      "ings",
+      "iew",
+      "ather",
+      "\u2581ass",
+      "\u2581love",
+      "ative",
+      "av",
+      "\u2581yes",
+      "\u2581ele",
+      "\u2581year",
+      "\u2581such",
+      "\u2581video",
+      "ness",
+      "\u2581el",
+      "\u2581trans",
+      "\u2581five",
+      "\u2581produ",
+      "ave",
+      "erest",
+      "als",
+      "body",
+      "cus",
+      "\u2581found",
+      "atter",
+      "\u2581eff",
+      "\u2581god",
+      "\u2581used",
+      "llow",
+      "\u2581interest",
+      "\u2581question",
+      "hip",
+      "\u2581bus",
+      "\u2581ask",
+      "\u2581exam",
+      "\u2581prov",
+      "lud",
+      "\u2581form",
+      "\u2581law",
+      "ense",
+      "\u2581child",
+      "\u2581gl",
+      "ne",
+      "\u2581each",
+      "\u2581understand",
+      "\u2581care",
+      "stem",
+      "\u2581med",
+      "\u2581maybe",
+      "ably",
+      "\u2581det",
+      "\u2581coll",
+      "its",
+      "\u2581commun",
+      "\u2581hand",
+      "\u2581'",
+      "\u2581ref",
+      "\u2581lear",
+      "\u2581done",
+      "\u2581gener",
+      "vern",
+      "\u2581mr",
+      "ween",
+      "\u2581better",
+      "\u2581between",
+      "li",
+      "blem",
+      "\u2581system",
+      "ertain",
+      "\u2581school",
+      "\u2581eas",
+      "\u2581exp",
+      "\u2581war",
+      "ention",
+      "\u2581ty",
+      "\u2581govern",
+      "ues",
+      "\u2581problem",
+      "\u2581plan",
+      "ac",
+      "\u2581conf",
+      "\u2581course",
+      "ouse",
+      "\u2581mar",
+      "\u2581stand",
+      "\u2581sk",
+      "\u2581seco",
+      "uring",
+      "\u2581ed",
+      "\u2581mem",
+      "ros",
+      "cri",
+      "\u2581thought",
+      "cept",
+      "\u2581partic",
+      "\u2581test",
+      "olog",
+      "iness",
+      "\u2581far",
+      "led",
+      "\u2581col",
+      "\u2581looking",
+      "\u2581read",
+      "\u2581whether",
+      "\u2581word",
+      "me",
+      "\u2581once",
+      "ize",
+      "\u2581home",
+      "\u2581requ",
+      "gg",
+      "\u2581ide",
+      "\u2581thank",
+      "ures",
+      "\u2581called",
+      "\u2581cur",
+      "\u2581water",
+      "\u2581frie",
+      "\u2581side",
+      "\u2581best",
+      "\u2581number",
+      "oney",
+      "\u2581turn",
+      "ock",
+      "\u2581eng",
+      "\u2581top",
+      "\u2581open",
+      "ead",
+      "\u2581everything",
+      "\u2581term",
+      "\u2581prob",
+      "\u2581hard",
+      "\u2581fun",
+      "\u2581spec",
+      "\u2581dire",
+      "\u2581second",
+      "\u2581pa",
+      "\u2581build",
+      "\u2581run",
+      "\u2581sign",
+      "\u2581reason",
+      "\u2581inform",
+      "\u2581watch",
+      "ution",
+      "\u2581few",
+      "mo",
+      "\u2581hum",
+      "ision",
+      "\u2581ext",
+      "\u2581tog",
+      "\u2581conc",
+      "\u2581thous",
+      "\u2581thousand",
+      "\u2581support",
+      "\u2581together",
+      "\u2581six",
+      "ps",
+      "\u2581mark",
+      "ics",
+      "\u2581includ",
+      "ef",
+      "\u2581opp",
+      "ident",
+      "\u2581anything",
+      "\u2581met",
+      "\u2581bre",
+      "\u2581jud",
+      "\u2581away",
+      "\u2581old",
+      "\u2581prog",
+      "ten",
+      "\u2581book",
+      "\u2581says",
+      "\u2581seem",
+      "\u2581contin",
+      "\u2581process",
+      "\u2581sing",
+      "\u2581money",
+      "\u2581having",
+      "\u2581beg",
+      "\u2581comple",
+      "\u2581thir",
+      "\u2581using",
+      "\u2581ret",
+      "ger",
+      "\u2581head",
+      "\u2581cre",
+      "\u2581poss",
+      "enty",
+      "\u2581certain",
+      "\u2581clear",
+      "ines",
+      "\u2581wee",
+      "arch",
+      "\u2581inf",
+      "ont",
+      "\u2581sit",
+      "\u2581lead",
+      "alth",
+      "\u2581art",
+      "ross",
+      "\u2581pub",
+      "\u2581without",
+      "\u2581pret",
+      "\u2581getting",
+      "ient",
+      "\u2581z",
+      "\u2581wom",
+      "\u2581power",
+      "ational",
+      "ner",
+      "\u2581rest",
+      "\u2581believe",
+      "\u2581wa",
+      "\u2581aut",
+      "\u2581move",
+      "aim",
+      "\u2581sort",
+      "idence",
+      "\u2581creat",
+      "\u2581expl",
+      "\u2581name",
+      "\u2581went",
+      "\u2581eu",
+      "\u2581change",
+      "\u2581came",
+      "\u2581pay",
+      "ices",
+      "\u2581sin",
+      "\u2581pur",
+      "\u2581pass",
+      "\u2581whole",
+      "\u2581house",
+      "\u2581hund",
+      "\u2581hundred",
+      "\u2581pretty",
+      "\u2581trying",
+      "\u2581ple",
+      "\u2581allow",
+      "\u2581compan",
+      "\u2581government",
+      "\u2581small",
+      "\u2581light",
+      "\u2581bra",
+      "\u2581stu",
+      "aint",
+      "\u2581ah",
+      "\u2581prot",
+      "ets",
+      "\u2581cent",
+      "velop",
+      "\u2581family",
+      "\u2581business",
+      "ety",
+      "\u2581making",
+      "\u2581list",
+      "\u2581experi",
+      "eric",
+      "\u2581follow",
+      "ately",
+      "\u2581probably",
+      "\u2581appe",
+      "\u2581serv",
+      "\u2581val",
+      "\u2581leg",
+      "\u2581resp",
+      "\u2581develop",
+      "ready",
+      "\u2581already",
+      "\u2581sec",
+      "ell",
+      "\u2581saying",
+      "ash",
+      "\u2581hear",
+      "\u2581loc",
+      "\u2581adv",
+      "\u2581pri",
+      "ret",
+      "\u2581lar",
+      "\u2581beh",
+      "\u2581must",
+      "\u2581hon",
+      "\u2581means",
+      "ew",
+      "\u2581par",
+      "\u2581order",
+      "\u2581mom",
+      "gn",
+      "\u2581though",
+      "\u2581record",
+      "\u2581miss",
+      "\u2581dr",
+      "\u2581es",
+      "\u2581eight",
+      "\u2581ever",
+      "\u2581left",
+      "\u2581example",
+      "\u2581enough",
+      "osed",
+      "\u2581claim",
+      "ank",
+      "con",
+      "\u2581americ",
+      "\u2581information",
+      "\u2581arg",
+      "\u2581full",
+      "nce",
+      "\u2581consid",
+      "\u2581working",
+      "ature",
+      "\u2581",
+      "e",
+      "t",
+      "a",
+      "o",
+      "i",
+      "n",
+      "s",
+      "r",
+      "h",
+      "l",
+      "d",
+      "u",
+      "c",
+      "m",
+      "y",
+      "w",
+      "g",
+      "f",
+      "p",
+      "b",
+      "v",
+      "k",
+      "'",
+      "j",
+      "x",
+      "q",
+      "z"
+    ]
+  },
+  "interctc": {
+    "loss_weights": [],
+    "apply_at_layers": []
+  },
+  "optim": {
+    "name": "adamw",
+    "lr": 0.001,
+    "betas": [
+      0.9,
+      0.98
+    ],
+    "weight_decay": 0.001,
+    "sched": {
+      "name": "CosineAnnealing",
+      "warmup_steps": 15000,
+      "warmup_ratio": null,
+      "min_lr": 0.0001
+    }
+  },
+  "target": "nemo.collections.asr.models.ctc_bpe_models.EncDecCTCModelBPE",
+  "nemo_version": "1.19.0rc0",
+  "decoding": {
+    "strategy": "greedy",
+    "preserve_alignments": null,
+    "compute_timestamps": null,
+    "word_seperator": " ",
+    "ctc_timestamp_type": "all",
+    "batch_dim_index": 0,
+    "greedy": {
+      "preserve_alignments": false,
+      "compute_timestamps": false,
+      "preserve_frame_confidence": false,
+      "confidence_method_cfg": null
+    },
+    "beam": {
+      "beam_size": 4,
+      "search_type": "default",
+      "preserve_alignments": false,
+      "compute_timestamps": false,
+      "return_best_hypothesis": true,
+      "beam_alpha": 1.0,
+      "beam_beta": 0.0,
+      "kenlm_path": null,
+      "flashlight_cfg": {
+        "lexicon_path": null,
+        "boost_path": null,
+        "beam_size_token": 16,
+        "beam_threshold": 20.0,
+        "unk_weight": -Infinity,
+        "sil_weight": 0.0
+      },
+      "pyctcdecode_cfg": {
+        "beam_prune_logp": -10.0,
+        "token_min_logp": -5.0,
+        "prune_history": false,
+        "hotwords": null,
+        "hotword_weight": 10.0
+      }
+    },
+    "confidence_cfg": {
+      "preserve_frame_confidence": false,
+      "preserve_token_confidence": false,
+      "preserve_word_confidence": false,
+      "exclude_blank": true,
+      "aggregation": "min",
+      "method_cfg": {
+        "name": "entropy",
+        "entropy_type": "tsallis",
+        "temperature": 0.33,
+        "entropy_norm": "exp"
+      }
+    },
+    "temperature": 1.0
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:730ec02e84879638c30834b8d451e1c5a321f0f6ed6796f88520755e739bb20f
+size 4250695964

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f279c3951a1c280a99a108d07b429bd98c03c94ecd0270748bd7affca9c0817
+size 259162

tokenizer.vocab

@@ -0,0 +1,1024 @@
+<unk>	0
+▁t	-0
+▁th	-1
+▁a	-2
+▁i	-3
+▁the	-4
+re	-5
+▁w	-6
+▁s	-7
+▁o	-8
+in	-9
+at	-10
+er	-11
+ou	-12
+nd	-13
+▁c	-14
+▁b	-15
+▁h	-16
+on	-17
+▁m	-18
+▁f	-19
+ing	-20
+▁to	-21
+en	-22
+▁p	-23
+▁and	-24
+▁d	-25
+es	-26
+or	-27
+an	-28
+ll	-29
+▁y	-30
+▁l	-31
+ed	-32
+▁of	-33
+▁in	-34
+it	-35
+is	-36
+▁you	-37
+▁that	-38
+ar	-39
+▁g	-40
+▁n	-41
+as	-42
+om	-43
+▁it	-44
+ic	-45
+ve	-46
+▁e	-47
+▁wh	-48
+▁be	-49
+us	-50
+le	-51
+al	-52
+ion	-53
+ow	-54
+▁we	-55
+▁re	-56
+▁is	-57
+ut	-58
+ot	-59
+ent	-60
+▁on	-61
+et	-62
+▁ha	-63
+ay	-64
+ct	-65
+▁he	-66
+id	-67
+▁for	-68
+▁st	-69
+ver	-70
+ly	-71
+ro	-72
+ig	-73
+▁so	-74
+ld	-75
+▁this	-76
+ke	-77
+▁u	-78
+se	-79
+all	-80
+st	-81
+ur	-82
+ce	-83
+ch	-84
+im	-85
+ith	-86
+▁as	-87
+▁k	-88
+▁an	-89
+▁was	-90
+▁j	-91
+▁with	-92
+ir	-93
+▁go	-94
+ra	-95
+▁do	-96
+▁have	-97
+▁li	-98
+▁sh	-99
+▁se	-100
+▁they	-101
+▁are	-102
+am	-103
+ht	-104
+▁but	-105
+ation	-106
+▁not	-107
+th	-108
+▁r	-109
+ally	-110
+ad	-111
+ust	-112
+▁or	-113
+▁com	-114
+ould	-115
+▁can	-116
+ill	-117
+▁ne	-118
+ight	-119
+▁ch	-120
+▁de	-121
+▁con	-122
+▁at	-123
+▁mo	-124
+ant	-125
+oo	-126
+il	-127
+▁me	-128
+▁what	-129
+▁there	-130
+ter	-131
+pe	-132
+▁ab	-133
+▁su	-134
+ere	-135
+ck	-136
+▁pro	-137
+▁al	-138
+▁fr	-139
+▁kn	-140
+▁all	-141
+ers	-142
+▁like	-143
+ge	-144
+▁ex	-145
+▁som	-146
+ul	-147
+▁your	-148
+▁v	-149
+pp	-150
+use	-151
+▁if	-152
+ess	-153
+ate	-154
+est	-155
+▁know	-156
+out	-157
+if	-158
+▁just	-159
+ment	-160
+qu	-161
+op	-162
+ain	-163
+▁one	-164
+ol	-165
+ri	-166
+art	-167
+very	-168
+▁wor	-169
+ive	-170
+ist	-171
+▁my	-172
+nt	-173
+ab	-174
+▁from	-175
+ort	-176
+▁ma	-177
+▁about	-178
+res	-179
+ity	-180
+▁out	-181
+▁bec	-182
+▁le	-183
+our	-184
+od	-185
+and	-186
+ink	-187
+ie	-188
+▁up	-189
+ind	-190
+os	-191
+un	-192
+ause	-193
+oug	-194
+um	-195
+▁some	-196
+▁int	-197
+▁by	-198
+▁pl	-199
+▁get	-200
+el	-201
+ard	-202
+▁when	-203
+▁don	-204
+her	-205
+▁will	-206
+▁us	-207
+▁would	-208
+ook	-209
+ies	-210
+ich	-211
+▁because	-212
+▁think	-213
+em	-214
+▁pe	-215
+▁his	-216
+ack	-217
+▁then	-218
+▁our	-219
+ide	-220
+▁tim	-221
+▁how	-222
+ven	-223
+▁tr	-224
+▁who	-225
+▁them	-226
+ure	-227
+▁ar	-228
+▁ye	-229
+▁more	-230
+▁going	-231
+ect	-232
+▁sa	-233
+▁cl	-234
+▁had	-235
+▁now	-236
+▁which	-237
+▁here	-238
+ous	-239
+▁their	-240
+▁tw	-241
+so	-242
+▁has	-243
+ud	-244
+▁co	-245
+▁ta	-246
+ound	-247
+▁were	-248
+ast	-249
+▁peop	-250
+ough	-251
+▁no	-252
+▁really	-253
+▁any	-254
+▁people	-255
+▁want	-256
+▁she	-257
+▁en	-258
+▁fa	-259
+▁te	-260
+ame	-261
+ine	-262
+▁qu	-263
+red	-264
+▁im	-265
+▁right	-266
+ther	-267
+▁act	-268
+▁thing	-269
+king	-270
+ose	-271
+▁ad	-272
+▁see	-273
+▁time	-274
+▁these	-275
+ci	-276
+one	-277
+▁say	-278
+▁also	-279
+▁fe	-280
+per	-281
+▁ag	-282
+▁man	-283
+ore	-284
+▁un	-285
+pt	-286
+▁her	-287
+▁look	-288
+ong	-289
+ice	-290
+▁very	-291
+ff	-292
+ions	-293
+▁comp	-294
+▁did	-295
+itt	-296
+▁well	-297
+▁other	-298
+iv	-299
+ase	-300
+ree	-301
+hing	-302
+▁lo	-303
+reat	-304
+▁cont	-305
+▁part	-306
+▁into	-307
+nder	-308
+▁been	-309
+are	-310
+▁am	-311
+ans	-312
+▁sp	-313
+▁two	-314
+ue	-315
+▁way	-316
+age	-317
+▁where	-318
+ite	-319
+▁dis	-320
+▁than	-321
+▁every	-322
+▁pr	-323
+▁po	-324
+ag	-325
+▁need	-326
+ach	-327
+iff	-328
+ence	-329
+pl	-330
+own	-331
+▁ac	-332
+ble	-333
+▁over	-334
+iz	-335
+▁work	-336
+▁res	-337
+▁make	-338
+▁could	-339
+▁off	-340
+ually	-341
+▁ro	-342
+▁back	-343
+able	-344
+ip	-345
+ry	-346
+▁him	-347
+▁cour	-348
+ber	-349
+▁pre	-350
+▁fir	-351
+▁spe	-352
+ap	-353
+ars	-354
+▁diff	-355
+ire	-356
+▁somet	-357
+▁imp	-358
+▁those	-359
+▁comm	-360
+ance	-361
+ick	-362
+▁even	-363
+ated	-364
+way	-365
+sel	-366
+▁let	-367
+▁br	-368
+ty	-369
+▁per	-370
+int	-371
+▁first	-372
+▁thr	-373
+▁under	-374
+ah	-375
+▁may	-376
+▁cou	-377
+▁new	-378
+ress	-379
+act	-380
+▁gr	-381
+ep	-382
+▁said	-383
+ations	-384
+▁good	-385
+ace	-386
+ass	-387
+▁does	-388
+orm	-389
+ish	-390
+▁af	-391
+ving	-392
+co	-393
+▁app	-394
+▁lot	-395
+▁things	-396
+▁tra	-397
+ittle	-398
+▁bl	-399
+▁little	-400
+▁mu	-401
+cess	-402
+fe	-403
+ome	-404
+▁inc	-405
+▁differe	-406
+ary	-407
+ical	-408
+▁only	-409
+ult	-410
+▁again	-411
+▁got	-412
+ens	-413
+▁gu	-414
+▁kind	-415
+▁much	-416
+ord	-417
+▁through	-418
+ition	-419
+ild	-420
+▁down	-421
+▁actually	-422
+▁something	-423
+ang	-424
+ru	-425
+ces	-426
+▁fl	-427
+ile	-428
+ater	-429
+▁ra	-430
+▁take	-431
+ict	-432
+ign	-433
+▁sc	-434
+vel	-435
+▁bet	-436
+▁tal	-437
+▁yeah	-438
+▁use	-439
+fore	-440
+▁bu	-441
+▁start	-442
+ory	-443
+be	-444
+▁day	-445
+wn	-446
+xt	-447
+ia	-448
+ak	-449
+▁after	-450
+▁should	-451
+▁fo	-452
+▁ho	-453
+▁hel	-454
+▁ind	-455
+▁uh	-456
+na	-457
+ial	-458
+other	-459
+▁ke	-460
+▁call	-461
+▁most	-462
+▁ok	-463
+▁different	-464
+▁em	-465
+ting	-466
+ple	-467
+▁being	-468
+▁bo	-469
+ning	-470
+▁too	-471
+ors	-472
+▁happ	-473
+ark	-474
+og	-475
+▁help	-476
+▁rem	-477
+du	-478
+ction	-479
+ood	-480
+▁ser	-481
+ether	-482
+ious	-483
+▁mean	-484
+▁many	-485
+▁court	-486
+▁bel	-487
+ade	-488
+▁la	-489
+ved	-490
+▁des	-491
+▁rec	-492
+▁jo	-493
+▁dec	-494
+ves	-495
+▁before	-496
+▁put	-497
+self	-498
+▁point	-499
+te	-500
+▁ev	-501
+form	-502
+ents	-503
+▁add	-504
+ody	-505
+thing	-506
+▁case	-507
+▁pers	-508
+▁cons	-509
+iss	-510
+▁three	-511
+oth	-512
+▁ph	-513
+▁come	-514
+▁find	-515
+▁why	-516
+ull	-517
+▁show	-518
+▁bas	-519
+▁great	-520
+ily	-521
+▁rel	-522
+▁sm	-523
+▁its	-524
+▁fact	-525
+▁pos	-526
+ool	-527
+ments	-528
+ise	-529
+nds	-530
+ys	-531
+▁try	-532
+ual	-533
+ful	-534
+erm	-535
+▁inter	-536
+ons	-537
+▁quest	-538
+▁sub	-539
+we	-540
+vers	-541
+▁supp	-542
+▁feel	-543
+▁same	-544
+ub	-545
+ates	-546
+urn	-547
+ert	-548
+▁inv	-549
+day	-550
+▁rep	-551
+igh	-552
+▁sy	-553
+▁inst	-554
+▁long	-555
+▁still	-556
+▁okay	-557
+ft	-558
+ific	-559
+atch	-560
+ought	-561
+ath	-562
+▁own	-563
+▁made	-564
+ix	-565
+ced	-566
+ks	-567
+lic	-568
+▁wr	-569
+de	-570
+▁cr	-571
+▁att	-572
+▁ob	-573
+▁world	-574
+▁sure	-575
+ward	-576
+▁bit	-577
+▁life	-578
+▁person	-579
+▁pres	-580
+ph	-581
+▁vide	-582
+▁reg	-583
+▁end	-584
+ject	-585
+ange	-586
+▁fin	-587
+ied	-588
+pect	-589
+▁didn	-590
+▁around	-591
+ian	-592
+▁car	-593
+ible	-594
+▁sim	-595
+ever	-596
+▁sch	-597
+ating	-598
+▁pol	-599
+▁set	-600
+▁oh	-601
+cy	-602
+▁real	-603
+▁import	-604
+▁count	-605
+▁um	-606
+▁next	-607
+cial	-608
+les	-609
+▁hu	-610
+▁acc	-611
+▁might	-612
+▁ent	-613
+▁doing	-614
+▁ins	-615
+▁gen	-616
+▁play	-617
+▁cle	-618
+▁another	-619
+ady	-620
+ular	-621
+ib	-622
+ways	-623
+ered	-624
+ility	-625
+ities	-626
+▁op	-627
+▁def	-628
+▁years	-629
+▁never	-630
+ower	-631
+ram	-632
+▁tell	-633
+▁sl	-634
+onna	-635
+ail	-636
+ren	-637
+ute	-638
+▁gonna	-639
+▁big	-640
+▁give	-641
+der	-642
+ount	-643
+▁ap	-644
+kes	-645
+▁state	-646
+▁cor	-647
+▁min	-648
+ically	-649
+▁mon	-650
+▁fam	-651
+▁important	-652
+▁always	-653
+▁high	-654
+▁four	-655
+▁gra	-656
+▁ca	-657
+▁stud	-658
+▁dist	-659
+▁talk	-660
+▁num	-661
+▁str	-662
+▁today	-663
+ract	-664
+▁while	-665
+ason	-666
+▁iss	-667
+▁sur	-668
+▁char	-669
+▁last	-670
+oy	-671
+ited	-672
+▁exper	-673
+▁place	-674
+▁tri	-675
+▁ear	-676
+▁belie	-677
+▁able	-678
+▁underst	-679
+▁che	-680
+▁both	-681
+ug	-682
+▁doesn	-683
+▁keep	-684
+▁happen	-685
+ings	-686
+iew	-687
+ather	-688
+▁ass	-689
+▁love	-690
+ative	-691
+av	-692
+▁yes	-693
+▁ele	-694
+▁year	-695
+▁such	-696
+▁video	-697
+ness	-698
+▁el	-699
+▁trans	-700
+▁five	-701
+▁produ	-702
+ave	-703
+erest	-704
+als	-705
+body	-706
+cus	-707
+▁found	-708
+atter	-709
+▁eff	-710
+▁god	-711
+▁used	-712
+llow	-713
+▁interest	-714
+▁question	-715
+hip	-716
+▁bus	-717
+▁ask	-718
+▁exam	-719
+▁prov	-720
+lud	-721
+▁form	-722
+▁law	-723
+ense	-724
+▁child	-725
+▁gl	-726
+ne	-727
+▁each	-728
+▁understand	-729
+▁care	-730
+stem	-731
+▁med	-732
+▁maybe	-733
+ably	-734
+▁det	-735
+▁coll	-736
+its	-737
+▁commun	-738
+▁hand	-739
+▁'	-740
+▁ref	-741
+▁lear	-742
+▁done	-743
+▁gener	-744
+vern	-745
+▁mr	-746
+ween	-747
+▁better	-748
+▁between	-749
+li	-750
+blem	-751
+▁system	-752
+ertain	-753
+▁school	-754
+▁eas	-755
+▁exp	-756
+▁war	-757
+ention	-758
+▁ty	-759
+▁govern	-760
+ues	-761
+▁problem	-762
+▁plan	-763
+ac	-764
+▁conf	-765
+▁course	-766
+ouse	-767
+▁mar	-768
+▁stand	-769
+▁sk	-770
+▁seco	-771
+uring	-772
+▁ed	-773
+▁mem	-774
+ros	-775
+cri	-776
+▁thought	-777
+cept	-778
+▁partic	-779
+▁test	-780
+olog	-781
+iness	-782
+▁far	-783
+led	-784
+▁col	-785
+▁looking	-786
+▁read	-787
+▁whether	-788
+▁word	-789
+me	-790
+▁once	-791
+ize	-792
+▁home	-793
+▁requ	-794
+gg	-795
+▁ide	-796
+▁thank	-797
+ures	-798
+▁called	-799
+▁cur	-800
+▁water	-801
+▁frie	-802
+▁side	-803
+▁best	-804
+▁number	-805
+oney	-806
+▁turn	-807
+ock	-808
+▁eng	-809
+▁top	-810
+▁open	-811
+ead	-812
+▁everything	-813
+▁term	-814
+▁prob	-815
+▁hard	-816
+▁fun	-817
+▁spec	-818
+▁dire	-819
+▁second	-820
+▁pa	-821
+▁build	-822
+▁run	-823
+▁sign	-824
+▁reason	-825
+▁inform	-826
+▁watch	-827
+ution	-828
+▁few	-829
+mo	-830
+▁hum	-831
+ision	-832
+▁ext	-833
+▁tog	-834
+▁conc	-835
+▁thous	-836
+▁thousand	-837
+▁support	-838
+▁together	-839
+▁six	-840
+ps	-841
+▁mark	-842
+ics	-843
+▁includ	-844
+ef	-845
+▁opp	-846
+ident	-847
+▁anything	-848
+▁met	-849
+▁bre	-850
+▁jud	-851
+▁away	-852
+▁old	-853
+▁prog	-854
+ten	-855
+▁book	-856
+▁says	-857
+▁seem	-858
+▁contin	-859
+▁process	-860
+▁sing	-861
+▁money	-862
+▁having	-863
+▁beg	-864
+▁comple	-865
+▁thir	-866
+▁using	-867
+▁ret	-868
+ger	-869
+▁head	-870
+▁cre	-871
+▁poss	-872
+enty	-873
+▁certain	-874
+▁clear	-875
+ines	-876
+▁wee	-877
+arch	-878
+▁inf	-879
+ont	-880
+▁sit	-881
+▁lead	-882
+alth	-883
+▁art	-884
+ross	-885
+▁pub	-886
+▁without	-887
+▁pret	-888
+▁getting	-889
+ient	-890
+▁z	-891
+▁wom	-892
+▁power	-893
+ational	-894
+ner	-895
+▁rest	-896
+▁believe	-897
+▁wa	-898
+▁aut	-899
+▁move	-900
+aim	-901
+▁sort	-902
+idence	-903
+▁creat	-904
+▁expl	-905
+▁name	-906
+▁went	-907
+▁eu	-908
+▁change	-909
+▁came	-910
+▁pay	-911
+ices	-912
+▁sin	-913
+▁pur	-914
+▁pass	-915
+▁whole	-916
+▁house	-917
+▁hund	-918
+▁hundred	-919
+▁pretty	-920
+▁trying	-921
+▁ple	-922
+▁allow	-923
+▁compan	-924
+▁government	-925
+▁small	-926
+▁light	-927
+▁bra	-928
+▁stu	-929
+aint	-930
+▁ah	-931
+▁prot	-932
+ets	-933
+▁cent	-934
+velop	-935
+▁family	-936
+▁business	-937
+ety	-938
+▁making	-939
+▁list	-940
+▁experi	-941
+eric	-942
+▁follow	-943
+ately	-944
+▁probably	-945
+▁appe	-946
+▁serv	-947
+▁val	-948
+▁leg	-949
+▁resp	-950
+▁develop	-951
+ready	-952
+▁already	-953
+▁sec	-954
+ell	-955
+▁saying	-956
+ash	-957
+▁hear	-958
+▁loc	-959
+▁adv	-960
+▁pri	-961
+ret	-962
+▁lar	-963
+▁beh	-964
+▁must	-965
+▁hon	-966
+▁means	-967
+ew	-968
+▁par	-969
+▁order	-970
+▁mom	-971
+gn	-972
+▁though	-973
+▁record	-974
+▁miss	-975
+▁dr	-976
+▁es	-977
+▁eight	-978
+▁ever	-979
+▁left	-980
+▁example	-981
+▁enough	-982
+osed	-983
+▁claim	-984
+ank	-985
+con	-986
+▁americ	-987
+▁information	-988
+▁arg	-989
+▁full	-990
+nce	-991
+▁consid	-992
+▁working	-993
+ature	-994
+▁	-995
+e	-996
+t	-997
+a	-998
+o	-999
+i	-1000
+n	-1001
+s	-1002
+r	-1003
+h	-1004
+l	-1005
+d	-1006
+u	-1007
+c	-1008
+m	-1009
+y	-1010
+w	-1011
+g	-1012
+f	-1013
+p	-1014
+b	-1015
+v	-1016
+k	-1017
+'	-1018
+j	-1019
+x	-1020
+q	-1021
+z	-1022

vocab.txt

@@ -0,0 +1,1023 @@
+t
+th
+a
+i
+the
+##re
+w
+s
+o
+##in
+##at
+##er
+##ou
+##nd
+c
+b
+h
+##on
+m
+f
+##ing
+to
+##en
+p
+and
+d
+##es
+##or
+##an
+##ll
+y
+l
+##ed
+of
+in
+##it
+##is
+you
+that
+##ar
+g
+n
+##as
+##om
+it
+##ic
+##ve
+e
+wh
+be
+##us
+##le
+##al
+##ion
+##ow
+we
+re
+is
+##ut
+##ot
+##ent
+on
+##et
+ha
+##ay
+##ct
+he
+##id
+for
+st
+##ver
+##ly
+##ro
+##ig
+so
+##ld
+this
+##ke
+u
+##se
+##all
+##st
+##ur
+##ce
+##ch
+##im
+##ith
+as
+k
+an
+was
+j
+with
+##ir
+go
+##ra
+do
+have
+li
+sh
+se
+they
+are
+##am
+##ht
+but
+##ation
+not
+##th
+r
+##ally
+##ad
+##ust
+or
+com
+##ould
+can
+##ill
+ne
+##ight
+ch
+de
+con
+at
+mo
+##ant
+##oo
+##il
+me
+what
+there
+##ter
+##pe
+ab
+su
+##ere
+##ck
+pro
+al
+fr
+kn
+all
+##ers
+like
+##ge
+ex
+som
+##ul
+your
+v
+##pp
+##use
+if
+##ess
+##ate
+##est
+know
+##out
+##if
+just
+##ment
+##qu
+##op
+##ain
+one
+##ol
+##ri
+##art
+##very
+wor
+##ive
+##ist
+my
+##nt
+##ab
+from
+##ort
+ma
+about
+##res
+##ity
+out
+bec
+le
+##our
+##od
+##and
+##ink
+##ie
+up
+##ind
+##os
+##un
+##ause
+##oug
+##um
+some
+int
+by
+pl
+get
+##el
+##ard
+when
+don
+##her
+will
+us
+would
+##ook
+##ies
+##ich
+because
+think
+##em
+pe
+his
+##ack
+then
+our
+##ide
+tim
+how
+##ven
+tr
+who
+them
+##ure
+ar
+ye
+more
+going
+##ect
+sa
+cl
+had
+now
+which
+here
+##ous
+their
+tw
+##so
+has
+##ud
+co
+ta
+##ound
+were
+##ast
+peop
+##ough
+no
+really
+any
+people
+want
+she
+en
+fa
+te
+##ame
+##ine
+qu
+##red
+im
+right
+##ther
+act
+thing
+##king
+##ose
+ad
+see
+time
+these
+##ci
+##one
+say
+also
+fe
+##per
+ag
+man
+##ore
+un
+##pt
+her
+look
+##ong
+##ice
+very
+##ff
+##ions
+comp
+did
+##itt
+well
+other
+##iv
+##ase
+##ree
+##hing
+lo
+##reat
+cont
+part
+into
+##nder
+been
+##are
+am
+##ans
+sp
+two
+##ue
+way
+##age
+where
+##ite
+dis
+than
+every
+pr
+po
+##ag
+need
+##ach
+##iff
+##ence
+##pl
+##own
+ac
+##ble
+over
+##iz
+work
+res
+make
+could
+off
+##ually
+ro
+back
+##able
+##ip
+##ry
+him
+cour
+##ber
+pre
+fir
+spe
+##ap
+##ars
+diff
+##ire
+somet
+imp
+those
+comm
+##ance
+##ick
+even
+##ated
+##way
+##sel
+let
+br
+##ty
+per
+##int
+first
+thr
+under
+##ah
+may
+cou
+new
+##ress
+##act
+gr
+##ep
+said
+##ations
+good
+##ace
+##ass
+does
+##orm
+##ish
+af
+##ving
+##co
+app
+lot
+things
+tra
+##ittle
+bl
+little
+mu
+##cess
+##fe
+##ome
+inc
+differe
+##ary
+##ical
+only
+##ult
+again
+got
+##ens
+gu
+kind
+much
+##ord
+through
+##ition
+##ild
+down
+actually
+something
+##ang
+##ru
+##ces
+fl
+##ile
+##ater
+ra
+take
+##ict
+##ign
+sc
+##vel
+bet
+tal
+yeah
+use
+##fore
+bu
+start
+##ory
+##be
+day
+##wn
+##xt
+##ia
+##ak
+after
+should
+fo
+ho
+hel
+ind
+uh
+##na
+##ial
+##other
+ke
+call
+most
+ok
+different
+em
+##ting
+##ple
+being
+bo
+##ning
+too
+##ors
+happ
+##ark
+##og
+help
+rem
+##du
+##ction
+##ood
+ser
+##ether
+##ious
+mean
+many
+court
+bel
+##ade
+la
+##ved
+des
+rec
+jo
+dec
+##ves
+before
+put
+##self
+point
+##te
+ev
+##form
+##ents
+add
+##ody
+##thing
+case
+pers
+cons
+##iss
+three
+##oth
+ph
+come
+find
+why
+##ull
+show
+bas
+great
+##ily
+rel
+sm
+its
+fact
+pos
+##ool
+##ments
+##ise
+##nds
+##ys
+try
+##ual
+##ful
+##erm
+inter
+##ons
+quest
+sub
+##we
+##vers
+supp
+feel
+same
+##ub
+##ates
+##urn
+##ert
+inv
+##day
+rep
+##igh
+sy
+inst
+long
+still
+okay
+##ft
+##ific
+##atch
+##ought
+##ath
+own
+made
+##ix
+##ced
+##ks
+##lic
+wr
+##de
+cr
+att
+ob
+world
+sure
+##ward
+bit
+life
+person
+pres
+##ph
+vide
+reg
+end
+##ject
+##ange
+fin
+##ied
+##pect
+didn
+around
+##ian
+car
+##ible
+sim
+##ever
+sch
+##ating
+pol
+set
+oh
+##cy
+real
+import
+count
+um
+next
+##cial
+##les
+hu
+acc
+might
+ent
+doing
+ins
+gen
+play
+cle
+another
+##ady
+##ular
+##ib
+##ways
+##ered
+##ility
+##ities
+op
+def
+years
+never
+##ower
+##ram
+tell
+sl
+##onna
+##ail
+##ren
+##ute
+gonna
+big
+give
+##der
+##ount
+ap
+##kes
+state
+cor
+min
+##ically
+mon
+fam
+important
+always
+high
+four
+gra
+ca
+stud
+dist
+talk
+num
+str
+today
+##ract
+while
+##ason
+iss
+sur
+char
+last
+##oy
+##ited
+exper
+place
+tri
+ear
+belie
+able
+underst
+che
+both
+##ug
+doesn
+keep
+happen
+##ings
+##iew
+##ather
+ass
+love
+##ative
+##av
+yes
+ele
+year
+such
+video
+##ness
+el
+trans
+five
+produ
+##ave
+##erest
+##als
+##body
+##cus
+found
+##atter
+eff
+god
+used
+##llow
+interest
+question
+##hip
+bus
+ask
+exam
+prov
+##lud
+form
+law
+##ense
+child
+gl
+##ne
+each
+understand
+care
+##stem
+med
+maybe
+##ably
+det
+coll
+##its
+commun
+hand
+'
+ref
+lear
+done
+gener
+##vern
+mr
+##ween
+better
+between
+##li
+##blem
+system
+##ertain
+school
+eas
+exp
+war
+##ention
+ty
+govern
+##ues
+problem
+plan
+##ac
+conf
+course
+##ouse
+mar
+stand
+sk
+seco
+##uring
+ed
+mem
+##ros
+##cri
+thought
+##cept
+partic
+test
+##olog
+##iness
+far
+##led
+col
+looking
+read
+whether
+word
+##me
+once
+##ize
+home
+requ
+##gg
+ide
+thank
+##ures
+called
+cur
+water
+frie
+side
+best
+number
+##oney
+turn
+##ock
+eng
+top
+open
+##ead
+everything
+term
+prob
+hard
+fun
+spec
+dire
+second
+pa
+build
+run
+sign
+reason
+inform
+watch
+##ution
+few
+##mo
+hum
+##ision
+ext
+tog
+conc
+thous
+thousand
+support
+together
+six
+##ps
+mark
+##ics
+includ
+##ef
+opp
+##ident
+anything
+met
+bre
+jud
+away
+old
+prog
+##ten
+book
+says
+seem
+contin
+process
+sing
+money
+having
+beg
+comple
+thir
+using
+ret
+##ger
+head
+cre
+poss
+##enty
+certain
+clear
+##ines
+wee
+##arch
+inf
+##ont
+sit
+lead
+##alth
+art
+##ross
+pub
+without
+pret
+getting
+##ient
+z
+wom
+power
+##ational
+##ner
+rest
+believe
+wa
+aut
+move
+##aim
+sort
+##idence
+creat
+expl
+name
+went
+eu
+change
+came
+pay
+##ices
+sin
+pur
+pass
+whole
+house
+hund
+hundred
+pretty
+trying
+ple
+allow
+compan
+government
+small
+light
+bra
+stu
+##aint
+ah
+prot
+##ets
+cent
+##velop
+family
+business
+##ety
+making
+list
+experi
+##eric
+follow
+##ately
+probably
+appe
+serv
+val
+leg
+resp
+develop
+##ready
+already
+sec
+##ell
+saying
+##ash
+hear
+loc
+adv
+pri
+##ret
+lar
+beh
+must
+hon
+means
+##ew
+par
+order
+mom
+##gn
+though
+record
+miss
+dr
+es
+eight
+ever
+left
+example
+enough
+##osed
+claim
+##ank
+##con
+americ
+information
+arg
+full
+##nce
+consid
+working
+##ature
+▁
+##e
+##t
+##a
+##o
+##i
+##n
+##s
+##r
+##h
+##l
+##d
+##u
+##c
+##m
+##y
+##w
+##g
+##f
+##p
+##b
+##v
+##k
+##'
+##j
+##x
+##q
+##z