lightsofapollo commited on
Commit
3b6574b
·
verified ·
1 Parent(s): b2d12be

Initial upload: 8-bit MLX quantized Nemotron streaming ASR (CC-BY-4.0 derivative of nvidia/nemotron-speech-streaming-en-0.6b)

Browse files
Files changed (7) hide show
  1. README.md +82 -0
  2. config.json +2267 -0
  3. model.safetensors +3 -0
  4. model_config.yaml +2227 -0
  5. tokenizer.model +3 -0
  6. tokenizer.vocab +1024 -0
  7. vocab.txt +1023 -0
README.md ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ track_downloads: true
4
+ language:
5
+ - en
6
+ pipeline_tag: automatic-speech-recognition
7
+ library_name: mlx
8
+ datasets:
9
+ - nvidia/Granary
10
+ - YTC
11
+ - Yodas2
12
+ - LibriLight
13
+ - librispeech_asr
14
+ - fisher_corpus
15
+ - Switchboard-1
16
+ - WSJ-0
17
+ - WSJ-1
18
+ - National-Singapore-Corpus-Part-1
19
+ - National-Singapore-Corpus-Part-6
20
+ - vctk
21
+ - voxpopuli
22
+ - europarl
23
+ - multilingual_librispeech
24
+ - fleurs
25
+ - mozilla-foundation/common_voice_8_0
26
+ - MLCommons/peoples_speech
27
+ - google/speech_commands
28
+ tags:
29
+ - quantized
30
+ - speech-recognition
31
+ - cache-aware ASR
32
+ - automatic-speech-recognition
33
+ - streaming-asr
34
+ - speech
35
+ - audio
36
+ - FastConformer
37
+ - RNNT
38
+ - Parakeet
39
+ - ASR
40
+ - pytorch
41
+ - NeMo
42
+ - mlx
43
+ base_model: nvidia/nemotron-speech-streaming-en-0.6b
44
+ base_model_relation: quantized
45
+ ---
46
+
47
+ # **animaslabs/nemotron-speech-streaming-en-0.6b-mlx-8bit**
48
+
49
+ This model was converted to MLX format, 8-bit quantized from [nvidia/nemotron-speech-streaming-en-0.6b](https://huggingface.co/nvidia/nemotron-speech-streaming-en-0.6b) using the scripts in this [github repo](https://github.com/animaslabs/mlx-models). Please refer to [original model card](https://huggingface.co/nvidia/nemotron-speech-streaming-en-0.6b) for more details on the model.
50
+
51
+ ## Usage
52
+
53
+ Quantized models require calling `mlx.nn.quantize()` before loading weights.
54
+
55
+ ```python
56
+ import json
57
+ import mlx.nn as nn
58
+ from huggingface_hub import hf_hub_download
59
+ from parakeet_mlx.utils import from_config
60
+
61
+ # Download and load config
62
+ config_path = hf_hub_download("animaslabs/nemotron-speech-streaming-en-0.6b-mlx-8bit", "config.json")
63
+ with open(config_path) as f:
64
+ config = json.load(f)
65
+
66
+ # Build model and apply quantization structure
67
+ model = from_config(config)
68
+ nn.quantize(
69
+ model,
70
+ bits=config["quantization"]["bits"],
71
+ group_size=config["quantization"]["group_size"],
72
+ )
73
+
74
+ # Load quantized weights
75
+ weights_path = hf_hub_download("animaslabs/nemotron-speech-streaming-en-0.6b-mlx-8bit", "model.safetensors")
76
+ model.load_weights(weights_path)
77
+
78
+ # Transcribe
79
+ result = model.transcribe("audio.wav")
80
+ print(result.text)
81
+ ```
82
+
config.json ADDED
@@ -0,0 +1,2267 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "sample_rate": 16000,
3
+ "compute_eval_loss": false,
4
+ "log_prediction": true,
5
+ "rnnt_reduction": "mean_volume",
6
+ "skip_nan_grad": false,
7
+ "model_defaults": {
8
+ "enc_hidden": 1024,
9
+ "pred_hidden": 640,
10
+ "joint_hidden": 640
11
+ },
12
+ "train_ds": {
13
+ "use_lhotse": true,
14
+ "skip_missing_manifest_entries": true,
15
+ "input_cfg": null,
16
+ "tarred_audio_filepaths": null,
17
+ "manifest_filepath": null,
18
+ "sample_rate": 16000,
19
+ "shuffle": true,
20
+ "num_workers": 2,
21
+ "pin_memory": true,
22
+ "max_duration": 40.0,
23
+ "min_duration": 0.1,
24
+ "text_field": "answer",
25
+ "batch_duration": null,
26
+ "use_bucketing": true,
27
+ "max_tps": null,
28
+ "bucket_duration_bins": null,
29
+ "bucket_batch_size": null,
30
+ "num_buckets": null,
31
+ "bucket_buffer_size": null,
32
+ "shuffle_buffer_size": null,
33
+ "augmentor": null
34
+ },
35
+ "validation_ds": {
36
+ "use_lhotse": true,
37
+ "manifest_filepath": "/data/ASR/en/librispeech/test-other.json",
38
+ "sample_rate": 16000,
39
+ "batch_size": 32,
40
+ "shuffle": false,
41
+ "max_duration": 40.0,
42
+ "min_duration": 0.1,
43
+ "num_workers": 2,
44
+ "pin_memory": true,
45
+ "text_field": "answer",
46
+ "tarred_audio_filepaths": null
47
+ },
48
+ "tokenizer": {
49
+ "dir": null,
50
+ "type": "bpe",
51
+ "model_path": "nemo:tokenizer.model",
52
+ "vocab_path": "nemo:vocab.txt",
53
+ "spe_tokenizer_vocab": "nemo:tokenizer.vocab"
54
+ },
55
+ "preprocessor": {
56
+ "_target_": "nemo.collections.asr.modules.AudioToMelSpectrogramPreprocessor",
57
+ "sample_rate": 16000,
58
+ "normalize": "NA",
59
+ "window_size": 0.025,
60
+ "window_stride": 0.01,
61
+ "window": "hann",
62
+ "features": 128,
63
+ "n_fft": 512,
64
+ "log": true,
65
+ "frame_splicing": 1,
66
+ "dither": 1e-05,
67
+ "pad_to": 0,
68
+ "pad_value": 0.0
69
+ },
70
+ "spec_augment": {
71
+ "_target_": "nemo.collections.asr.modules.SpectrogramAugmentation",
72
+ "freq_masks": 2,
73
+ "time_masks": 10,
74
+ "freq_width": 27,
75
+ "time_width": 0.05
76
+ },
77
+ "encoder": {
78
+ "_target_": "nemo.collections.asr.modules.ConformerEncoder",
79
+ "feat_in": 128,
80
+ "feat_out": -1,
81
+ "n_layers": 24,
82
+ "d_model": 1024,
83
+ "use_bias": false,
84
+ "subsampling": "dw_striding",
85
+ "subsampling_factor": 8,
86
+ "subsampling_conv_channels": 256,
87
+ "causal_downsampling": true,
88
+ "reduction": null,
89
+ "reduction_position": null,
90
+ "reduction_factor": 1,
91
+ "ff_expansion_factor": 4,
92
+ "self_attention_model": "rel_pos",
93
+ "n_heads": 8,
94
+ "att_context_size": [
95
+ [
96
+ 70,
97
+ 13
98
+ ],
99
+ [
100
+ 70,
101
+ 6
102
+ ],
103
+ [
104
+ 70,
105
+ 1
106
+ ],
107
+ [
108
+ 70,
109
+ 0
110
+ ]
111
+ ],
112
+ "att_context_style": "chunked_limited",
113
+ "xscaling": false,
114
+ "untie_biases": true,
115
+ "pos_emb_max_len": 5000,
116
+ "conv_kernel_size": 9,
117
+ "conv_norm_type": "layer_norm",
118
+ "conv_context_size": "causal",
119
+ "dropout": 0.1,
120
+ "dropout_pre_encoder": 0.1,
121
+ "dropout_emb": 0.0,
122
+ "dropout_att": 0.1,
123
+ "stochastic_depth_drop_prob": 0.0,
124
+ "stochastic_depth_mode": "linear",
125
+ "stochastic_depth_start_layer": 1
126
+ },
127
+ "decoder": {
128
+ "_target_": "nemo.collections.asr.modules.RNNTDecoder",
129
+ "normalization_mode": null,
130
+ "random_state_sampling": false,
131
+ "blank_as_pad": true,
132
+ "prednet": {
133
+ "pred_hidden": 640,
134
+ "pred_rnn_layers": 2,
135
+ "t_max": null,
136
+ "dropout": 0.2
137
+ },
138
+ "vocab_size": 1024
139
+ },
140
+ "joint": {
141
+ "_target_": "nemo.collections.asr.modules.RNNTJoint",
142
+ "log_softmax": null,
143
+ "preserve_memory": false,
144
+ "fuse_loss_wer": true,
145
+ "fused_batch_size": 4,
146
+ "jointnet": {
147
+ "joint_hidden": 640,
148
+ "activation": "relu",
149
+ "dropout": 0.2,
150
+ "encoder_hidden": 1024,
151
+ "pred_hidden": 640
152
+ },
153
+ "num_classes": 1024,
154
+ "vocabulary": [
155
+ "<unk>",
156
+ "\u2581t",
157
+ "\u2581th",
158
+ "\u2581a",
159
+ "in",
160
+ "\u2581the",
161
+ "re",
162
+ "\u2581w",
163
+ "\u2581o",
164
+ "\u2581s",
165
+ "er",
166
+ "at",
167
+ "ou",
168
+ "nd",
169
+ "it",
170
+ "is",
171
+ "\u2581h",
172
+ "\u2581b",
173
+ "on",
174
+ "\u2581c",
175
+ "ing",
176
+ "en",
177
+ "\u2581to",
178
+ "\u2581m",
179
+ "\u2581f",
180
+ "\u2581p",
181
+ "or",
182
+ "an",
183
+ "es",
184
+ "\u2581of",
185
+ "\u2581d",
186
+ "ed",
187
+ "ll",
188
+ "\u2581and",
189
+ "\u2581I",
190
+ "\u2581in",
191
+ "\u2581l",
192
+ "ar",
193
+ "\u2581y",
194
+ "\u2581g",
195
+ "as",
196
+ "\u2581you",
197
+ "om",
198
+ "\u2581n",
199
+ "ic",
200
+ "ve",
201
+ "al",
202
+ "ion",
203
+ "us",
204
+ "\u2581be",
205
+ "ow",
206
+ "le",
207
+ "\u2581wh",
208
+ "\u2581e",
209
+ "ot",
210
+ "ut",
211
+ "\u2581it",
212
+ "\u2581is",
213
+ "\u2581we",
214
+ "\u2581T",
215
+ "\u2581re",
216
+ "et",
217
+ "\u2581A",
218
+ "ent",
219
+ "\u2581on",
220
+ "\u2581ha",
221
+ "ay",
222
+ "\u2581S",
223
+ "ct",
224
+ "\u2581Th",
225
+ "ver",
226
+ "id",
227
+ "ig",
228
+ "im",
229
+ "ro",
230
+ "\u2581for",
231
+ "ly",
232
+ "\u2581he",
233
+ "ke",
234
+ "ld",
235
+ "se",
236
+ "st",
237
+ "ch",
238
+ "\u2581st",
239
+ "all",
240
+ "ce",
241
+ "ur",
242
+ "ith",
243
+ "am",
244
+ "if",
245
+ "ir",
246
+ "\u2581go",
247
+ "\u2581u",
248
+ "\u2581as",
249
+ "\u2581was",
250
+ "ad",
251
+ "\u2581W",
252
+ "\u2581k",
253
+ "\u2581an",
254
+ "ht",
255
+ "th",
256
+ "\u2581r",
257
+ "\u2581are",
258
+ "ere",
259
+ "\u2581se",
260
+ "\u2581do",
261
+ "\u2581B",
262
+ "\u2581so",
263
+ "\u2581sh",
264
+ "\u2581not",
265
+ "\u2581li",
266
+ "od",
267
+ "\u2581C",
268
+ "ust",
269
+ "ill",
270
+ "ight",
271
+ "ally",
272
+ "\u2581And",
273
+ "ter",
274
+ "\u2581or",
275
+ "\u2581me",
276
+ "\u2581M",
277
+ "ome",
278
+ "op",
279
+ "\u2581at",
280
+ "il",
281
+ "\u2581The",
282
+ "ould",
283
+ "\u2581j",
284
+ "ant",
285
+ "\u2581So",
286
+ "\u2581H",
287
+ "ol",
288
+ "ain",
289
+ "\u2581can",
290
+ "\u2581de",
291
+ "\u2581ne",
292
+ "ore",
293
+ "\u2581con",
294
+ "\u2581kn",
295
+ "ck",
296
+ "ul",
297
+ "\u2581fr",
298
+ "\u2581ab",
299
+ "ers",
300
+ "ess",
301
+ "ge",
302
+ "\u2581pro",
303
+ "pe",
304
+ "ate",
305
+ "\u2581su",
306
+ "\u2581com",
307
+ "\u2581but",
308
+ "\u2581all",
309
+ "est",
310
+ "qu",
311
+ "\u2581ex",
312
+ "\u2581al",
313
+ "ra",
314
+ "\u2581O",
315
+ "out",
316
+ "use",
317
+ "very",
318
+ "pp",
319
+ "\u2581Y",
320
+ "\u2581ch",
321
+ "ri",
322
+ "ist",
323
+ "\u2581v",
324
+ "\u2581lo",
325
+ "ment",
326
+ "art",
327
+ "\u2581P",
328
+ "nt",
329
+ "ab",
330
+ "\u2581one",
331
+ "\u2581N",
332
+ "ive",
333
+ "\u2581wor",
334
+ "ions",
335
+ "ort",
336
+ "\u2581L",
337
+ "\u2581by",
338
+ "ich",
339
+ "\u2581my",
340
+ "ity",
341
+ "ok",
342
+ "\u2581G",
343
+ "res",
344
+ "\u2581up",
345
+ "un",
346
+ "um",
347
+ "ea",
348
+ "ind",
349
+ "and",
350
+ "ink",
351
+ "el",
352
+ "\u2581D",
353
+ "em",
354
+ "\u2581E",
355
+ "os",
356
+ "oug",
357
+ "\u2581if",
358
+ "ca",
359
+ "\u2581out",
360
+ "\u2581int",
361
+ "ie",
362
+ "\u2581F",
363
+ "\u2581It",
364
+ "\u2581his",
365
+ "ard",
366
+ "\u2581had",
367
+ "\u2581tr",
368
+ "her",
369
+ "our",
370
+ "ies",
371
+ "ake",
372
+ "\u2581R",
373
+ "\u2581We",
374
+ "\u2581get",
375
+ "\u2581don",
376
+ "\u2581us",
377
+ "ak",
378
+ "\u2581pl",
379
+ "ect",
380
+ "ure",
381
+ "ame",
382
+ "ast",
383
+ "\u2581who",
384
+ "ack",
385
+ "\u2581le",
386
+ "\u2581sa",
387
+ "iv",
388
+ "ci",
389
+ "ide",
390
+ "\u2581tim",
391
+ "\u2581our",
392
+ "ound",
393
+ "ous",
394
+ "\u2581co",
395
+ "\u2581pe",
396
+ "ose",
397
+ "ud",
398
+ "\u2581see",
399
+ "ough",
400
+ "\u2581man",
401
+ "\u2581qu",
402
+ "\u2581You",
403
+ "so",
404
+ "ople",
405
+ "\u2581Wh",
406
+ "ong",
407
+ "ap",
408
+ "ther",
409
+ "\u2581J",
410
+ "are",
411
+ "ine",
412
+ "\u2581say",
413
+ "\u2581im",
414
+ "\u2581But",
415
+ "ings",
416
+ "\u2581has",
417
+ "\u2581ag",
418
+ "ff",
419
+ "\u2581her",
420
+ "itt",
421
+ "one",
422
+ "\u2581en",
423
+ "\u2581ar",
424
+ "\u2581fe",
425
+ "ven",
426
+ "\u2581any",
427
+ "\u2581mo",
428
+ "reat",
429
+ "ag",
430
+ "\u2581how",
431
+ "\u2581cl",
432
+ "pt",
433
+ "\u2581now",
434
+ "own",
435
+ "ber",
436
+ "\u2581him",
437
+ "\u2581act",
438
+ "hing",
439
+ "ice",
440
+ "\u2581no",
441
+ "ans",
442
+ "iz",
443
+ "\u2581fa",
444
+ "per",
445
+ "pl",
446
+ "\u2581te",
447
+ "\u2581ad",
448
+ "age",
449
+ "ree",
450
+ "\u2581tw",
451
+ "ank",
452
+ "\u2581He",
453
+ "ple",
454
+ "ite",
455
+ "ry",
456
+ "\u2581U",
457
+ "ish",
458
+ "ire",
459
+ "ue",
460
+ "\u2581In",
461
+ "\u2581she",
462
+ "ble",
463
+ "cc",
464
+ "nder",
465
+ "\u2581way",
466
+ "\u2581pr",
467
+ "ear",
468
+ "\u2581did",
469
+ "\u2581po",
470
+ "eah",
471
+ "\u2581un",
472
+ "omet",
473
+ "ence",
474
+ "ep",
475
+ "uch",
476
+ "\u2581sp",
477
+ "ach",
478
+ "og",
479
+ "ance",
480
+ "able",
481
+ "iff",
482
+ "sel",
483
+ "\u2581got",
484
+ "way",
485
+ "\u2581gr",
486
+ "alk",
487
+ "\u2581res",
488
+ "ated",
489
+ "irst",
490
+ "ick",
491
+ "ass",
492
+ "\u2581two",
493
+ "\u2581dis",
494
+ "ord",
495
+ "\u2581pre",
496
+ "ount",
497
+ "ase",
498
+ "ip",
499
+ "ult",
500
+ "ical",
501
+ "orm",
502
+ "ary",
503
+ "ace",
504
+ "\u2581spe",
505
+ "\u2581Ch",
506
+ "\u2581thr",
507
+ "\u2581imp",
508
+ "int",
509
+ "\u2581am",
510
+ "\u2581off",
511
+ "act",
512
+ "ia",
513
+ "\u2581ro",
514
+ "ress",
515
+ "\u2581per",
516
+ "\u2581fo",
517
+ "\u2581br",
518
+ "\u2581K",
519
+ "vel",
520
+ "\u2581gu",
521
+ "\u2581bo",
522
+ "ang",
523
+ "kay",
524
+ "ub",
525
+ "ign",
526
+ "\u2581may",
527
+ "ving",
528
+ "ces",
529
+ "ens",
530
+ "cl",
531
+ "\u2581lot",
532
+ "ru",
533
+ "ade",
534
+ "\u2581bet",
535
+ "\u2581bl",
536
+ "\u2581let",
537
+ "fore",
538
+ "co",
539
+ "ild",
540
+ "ning",
541
+ "xt",
542
+ "ile",
543
+ "ark",
544
+ "self",
545
+ "\u2581app",
546
+ "ory",
547
+ "du",
548
+ "\u2581day",
549
+ "\u2581St",
550
+ "ater",
551
+ "\u2581use",
552
+ "ys",
553
+ "fter",
554
+ "\u2581new",
555
+ "ious",
556
+ "ial",
557
+ "he",
558
+ "wn",
559
+ "ved",
560
+ "red",
561
+ "\u2581fl",
562
+ "iss",
563
+ "ody",
564
+ "form",
565
+ "ian",
566
+ "tain",
567
+ "\u2581bu",
568
+ "\u2581V",
569
+ "\u2581rec",
570
+ "ty",
571
+ "be",
572
+ "\u2581sc",
573
+ "ors",
574
+ "vers",
575
+ "\u2581put",
576
+ "ife",
577
+ "\u2581If",
578
+ "we",
579
+ "te",
580
+ "ject",
581
+ "ath",
582
+ "ting",
583
+ "\u2581rem",
584
+ "\u2581acc",
585
+ "ull",
586
+ "ons",
587
+ "\u2581ind",
588
+ "\u2581ser",
589
+ "\u2581ke",
590
+ "ates",
591
+ "ves",
592
+ "na",
593
+ "lic",
594
+ "\u2581des",
595
+ "\u2581its",
596
+ "ful",
597
+ "ents",
598
+ "erm",
599
+ "ac",
600
+ "ered",
601
+ "ise",
602
+ "\u2581sy",
603
+ "urn",
604
+ "\u2581em",
605
+ "oth",
606
+ "ual",
607
+ "ne",
608
+ "ward",
609
+ "ib",
610
+ "\u2581try",
611
+ "\u2581pos",
612
+ "nds",
613
+ "ft",
614
+ "get",
615
+ "ph",
616
+ "\u2581ob",
617
+ "ady",
618
+ "igh",
619
+ "ood",
620
+ "\u2581rel",
621
+ "\u2581wr",
622
+ "ug",
623
+ "ears",
624
+ "ail",
625
+ "\u2581Now",
626
+ "\u2581bit",
627
+ "ng",
628
+ "\u2581Oh",
629
+ "\u2581hel",
630
+ "ange",
631
+ "\u2581reg",
632
+ "\u2581rep",
633
+ "\u2581bel",
634
+ "\u2581sm",
635
+ "ost",
636
+ "tern",
637
+ "gr",
638
+ "\u2581own",
639
+ "\u2581end",
640
+ "pect",
641
+ "ily",
642
+ "day",
643
+ "ied",
644
+ "ific",
645
+ "ower",
646
+ "\u2581add",
647
+ "cess",
648
+ "ict",
649
+ "ible",
650
+ "\u2581bas",
651
+ "\u2581i",
652
+ "\u2581op",
653
+ "cial",
654
+ "ular",
655
+ "\u2581Be",
656
+ "ced",
657
+ "\u2581too",
658
+ "ks",
659
+ "ew",
660
+ "mer",
661
+ "\u2581ph",
662
+ "ob",
663
+ "==",
664
+ "\u2581la",
665
+ "\u2581set",
666
+ "\u2581min",
667
+ "\u2581sub",
668
+ "\u2581gen",
669
+ "atch",
670
+ "..",
671
+ "\u2581inv",
672
+ "\u2581As",
673
+ "\u2581nat",
674
+ "\u2581sl",
675
+ "\u2581num",
676
+ "av",
677
+ "ways",
678
+ "\u2581God",
679
+ "stem",
680
+ "\u2581ac",
681
+ "\u2581att",
682
+ "\u2581ev",
683
+ "\u2581def",
684
+ "llow",
685
+ "\u2581str",
686
+ "lect",
687
+ "ars",
688
+ "\u2581cr",
689
+ "\u2581Is",
690
+ "olog",
691
+ "les",
692
+ "oy",
693
+ "\u2581ask",
694
+ "\u2581inc",
695
+ "body",
696
+ "\u2581ent",
697
+ "\u2581pol",
698
+ "ness",
699
+ "ix",
700
+ "\u2581why",
701
+ "onna",
702
+ "\u2581ear",
703
+ "\u2581tak",
704
+ "\u2581Un",
705
+ "ited",
706
+ "mun",
707
+ "li",
708
+ "ute",
709
+ "ract",
710
+ "\u2581dec",
711
+ "uro",
712
+ "\u2581mak",
713
+ "\u2581fin",
714
+ "ween",
715
+ "\u2581No",
716
+ "arch",
717
+ "\u2581bec",
718
+ "gan",
719
+ "old",
720
+ "cy",
721
+ "\u2581big",
722
+ "\u2581For",
723
+ "ren",
724
+ "als",
725
+ "und",
726
+ "\u2581Al",
727
+ "\u2581All",
728
+ "ss",
729
+ "ows",
730
+ "\u2581mod",
731
+ "ock",
732
+ "\u2581id",
733
+ "ism",
734
+ "cus",
735
+ "\u2581gl",
736
+ "ably",
737
+ "\u2581ass",
738
+ "\u2581car",
739
+ "ata",
740
+ "ppen",
741
+ "led",
742
+ "\u2581sim",
743
+ "\u2581mon",
744
+ "ics",
745
+ "\u2581giv",
746
+ "cept",
747
+ "\u2581Mr",
748
+ "pan",
749
+ "\u2581pub",
750
+ "\u2581eff",
751
+ "\u2581How",
752
+ "ps",
753
+ "vern",
754
+ "end",
755
+ "hip",
756
+ "iew",
757
+ "ope",
758
+ "\u2581An",
759
+ "\u2581She",
760
+ "\u2581Com",
761
+ "ee",
762
+ "ures",
763
+ "ell",
764
+ "ouse",
765
+ "cond",
766
+ "king",
767
+ "oc",
768
+ "ues",
769
+ "ever",
770
+ "\u2581To",
771
+ "clud",
772
+ "\u2581ins",
773
+ "\u2581exp",
774
+ "\u2581old",
775
+ "\u2581mem",
776
+ "\u2581ref",
777
+ "\u2581tra",
778
+ "\u2581far",
779
+ "ave",
780
+ "rat",
781
+ "\u2581sur",
782
+ "ruct",
783
+ "rib",
784
+ "duct",
785
+ "uff",
786
+ "\u2581met",
787
+ "\u2581sch",
788
+ "ince",
789
+ "\u2581run",
790
+ "ense",
791
+ "\u2581cle",
792
+ "\u2581==",
793
+ "mon",
794
+ "ize",
795
+ "\u2581ord",
796
+ "blem",
797
+ "tin",
798
+ "\u2581Let",
799
+ "ner",
800
+ "ond",
801
+ "its",
802
+ "\u2581cor",
803
+ "land",
804
+ "\u2581cur",
805
+ "\u2581Re",
806
+ "\u2581bus",
807
+ "\u2581uh",
808
+ "air",
809
+ "ote",
810
+ "ants",
811
+ "ason",
812
+ "ric",
813
+ "\u2581el",
814
+ "\u2581cer",
815
+ "nce",
816
+ "\u2581fam",
817
+ "\u2581cap",
818
+ "uck",
819
+ "ool",
820
+ "ried",
821
+ "\u2581cou",
822
+ "\u2581fun",
823
+ "\u2581wom",
824
+ "\u2581hum",
825
+ "\u2581ty",
826
+ "\u2581ap",
827
+ "ike",
828
+ "\u2581few",
829
+ "oney",
830
+ "\u2581inf",
831
+ "ont",
832
+ "ese",
833
+ "ook",
834
+ "gy",
835
+ "uth",
836
+ "ulat",
837
+ "ieve",
838
+ "ized",
839
+ "ross",
840
+ "\u2581ple",
841
+ "\u2581um",
842
+ "\u2581val",
843
+ "\u2581equ",
844
+ "\u2581lea",
845
+ "\u2581lar",
846
+ "ah",
847
+ "eral",
848
+ "\u2581ed",
849
+ "ared",
850
+ "lish",
851
+ "arn",
852
+ "ds",
853
+ "esn",
854
+ "\u2581iss",
855
+ "\u2581ca",
856
+ "ted",
857
+ "ices",
858
+ "\u2581wee",
859
+ "ash",
860
+ "\u2581top",
861
+ "ten",
862
+ "up",
863
+ "ts",
864
+ "gin",
865
+ "con",
866
+ "ari",
867
+ "\u2581opp",
868
+ "osed",
869
+ "\u2581eas",
870
+ "\u2581ext",
871
+ "gg",
872
+ "az",
873
+ "\u2581Fr",
874
+ "ideo",
875
+ "izat",
876
+ "\u2581men",
877
+ "\u2581mom",
878
+ "\u2581ret",
879
+ "tty",
880
+ "rist",
881
+ "\u2581gra",
882
+ "alth",
883
+ "ef",
884
+ "\u2581det",
885
+ "ax",
886
+ "\u2581mat",
887
+ "chn",
888
+ "ern",
889
+ "peri",
890
+ "\u2581bre",
891
+ "\u2581Sh",
892
+ "sw",
893
+ "erat",
894
+ "\u2581sit",
895
+ "ters",
896
+ "ale",
897
+ "man",
898
+ "\u2581sol",
899
+ "ork",
900
+ "\u2581adv",
901
+ "ety",
902
+ "\u2581vis",
903
+ "\u2581med",
904
+ "uc",
905
+ "less",
906
+ "\u2581unt",
907
+ "gram",
908
+ "ets",
909
+ "ists",
910
+ "\u2581ey",
911
+ "\u2581col",
912
+ "imes",
913
+ "\u2581law",
914
+ "\u2581pri",
915
+ "sid",
916
+ "\u2581On",
917
+ "\u2581mot",
918
+ "ield",
919
+ "\u2581Do",
920
+ "\u2581At",
921
+ "ages",
922
+ "amp",
923
+ "\u2581art",
924
+ "miss",
925
+ "\u2581sk",
926
+ "alf",
927
+ "pr",
928
+ "ier",
929
+ "\u2581beh",
930
+ "\u2581Yes",
931
+ "ural",
932
+ "ime",
933
+ "\u2581wa",
934
+ "oks",
935
+ "bers",
936
+ "ger",
937
+ "ient",
938
+ "ries",
939
+ "...",
940
+ "\u2581che",
941
+ "\u2581Br",
942
+ "ird",
943
+ "\u2581Ar",
944
+ "\u2581war",
945
+ "inat",
946
+ "\u2581My",
947
+ "ital",
948
+ "wh",
949
+ "med",
950
+ "\u2581pur",
951
+ "ully",
952
+ "\u2581One",
953
+ "\u2581rat",
954
+ "ines",
955
+ "\u2581Of",
956
+ "io",
957
+ "\u2581loc",
958
+ "ret",
959
+ "ctor",
960
+ "\u2581leg",
961
+ "stit",
962
+ "ined",
963
+ "ught",
964
+ "\u2581dur",
965
+ "\u2581es",
966
+ "vent",
967
+ "aj",
968
+ "\u2581bro",
969
+ "\u2581saw",
970
+ "\u2581sec",
971
+ "ream",
972
+ "\u2581pop",
973
+ "reen",
974
+ "\u2581Ind",
975
+ "els",
976
+ "\u2581yet",
977
+ "ired",
978
+ "\u2581sw",
979
+ "tro",
980
+ "oup",
981
+ "most",
982
+ "pean",
983
+ "eds",
984
+ "ush",
985
+ "oh",
986
+ "\u2581Se",
987
+ "\u2581tea",
988
+ "ann",
989
+ "ilit",
990
+ "err",
991
+ "pend",
992
+ "ton",
993
+ "ased",
994
+ "\u2581aff",
995
+ "\u2581mor",
996
+ "\u2581dra",
997
+ "put",
998
+ "\u2581dr",
999
+ "ins",
1000
+ "uat",
1001
+ "nect",
1002
+ "cri",
1003
+ "outh",
1004
+ "\u2581ra",
1005
+ "\u2581pay",
1006
+ "ms",
1007
+ "\u2581av",
1008
+ "bs",
1009
+ "ling",
1010
+ "\u2581De",
1011
+ "\u2581Or",
1012
+ "ove",
1013
+ "\u2581Can",
1014
+ "\u2581eng",
1015
+ "ames",
1016
+ "ided",
1017
+ "\u2581Go",
1018
+ "mitt",
1019
+ "ode",
1020
+ "\u2581cre",
1021
+ "par",
1022
+ "ides",
1023
+ "pos",
1024
+ "\u2581fav",
1025
+ "\u2581air",
1026
+ "\u2581New",
1027
+ "\u2581bad",
1028
+ "\u2581six",
1029
+ "vat",
1030
+ "\u2581pat",
1031
+ "not",
1032
+ "\u2581di",
1033
+ "rop",
1034
+ "ral",
1035
+ "orn",
1036
+ "\u2581par",
1037
+ "cing",
1038
+ "\u2581aw",
1039
+ "orts",
1040
+ "ox",
1041
+ "\u2581yes",
1042
+ "cuss",
1043
+ "eng",
1044
+ "ives",
1045
+ "erms",
1046
+ "\u2581job",
1047
+ "mand",
1048
+ "ying",
1049
+ "\u2581occ",
1050
+ "aps",
1051
+ "ases",
1052
+ "\u2581Not",
1053
+ "rent",
1054
+ "ency",
1055
+ "att",
1056
+ "ised",
1057
+ "vice",
1058
+ "\u2581Eng",
1059
+ "\u2581est",
1060
+ "oked",
1061
+ "\u2581Q",
1062
+ "iron",
1063
+ "idd",
1064
+ "me",
1065
+ "unch",
1066
+ "ane",
1067
+ "\u2581z",
1068
+ "br",
1069
+ "arts",
1070
+ "\u2581fat",
1071
+ "ery",
1072
+ "anks",
1073
+ "\u2581jo",
1074
+ "\u2581mar",
1075
+ "aw",
1076
+ "ott",
1077
+ "ards",
1078
+ "\u2581oh",
1079
+ "ians",
1080
+ "\u2581sci",
1081
+ "row",
1082
+ "unt",
1083
+ "ury",
1084
+ "\u2581abs",
1085
+ "ergy",
1086
+ "\u2581Z",
1087
+ "ump",
1088
+ "\u2581Am",
1089
+ "ened",
1090
+ "angu",
1091
+ "\u2581Pro",
1092
+ "icat",
1093
+ "itch",
1094
+ "\u2581dri",
1095
+ "iat",
1096
+ "\u2581",
1097
+ "e",
1098
+ "t",
1099
+ "o",
1100
+ "a",
1101
+ "n",
1102
+ "i",
1103
+ "s",
1104
+ "r",
1105
+ "h",
1106
+ "l",
1107
+ "d",
1108
+ "u",
1109
+ "c",
1110
+ "m",
1111
+ "y",
1112
+ "g",
1113
+ "w",
1114
+ "f",
1115
+ "p",
1116
+ ",",
1117
+ ".",
1118
+ "b",
1119
+ "v",
1120
+ "k",
1121
+ "'",
1122
+ "I",
1123
+ "T",
1124
+ "A",
1125
+ "S",
1126
+ "x",
1127
+ "W",
1128
+ "j",
1129
+ "C",
1130
+ "B",
1131
+ "M",
1132
+ "?",
1133
+ "H",
1134
+ "O",
1135
+ "0",
1136
+ "P",
1137
+ "q",
1138
+ "Y",
1139
+ "N",
1140
+ "L",
1141
+ "D",
1142
+ "1",
1143
+ "E",
1144
+ "G",
1145
+ "z",
1146
+ "F",
1147
+ "R",
1148
+ "-",
1149
+ "2",
1150
+ "J",
1151
+ "U",
1152
+ "9",
1153
+ "K",
1154
+ "5",
1155
+ "3",
1156
+ "V",
1157
+ "=",
1158
+ "4",
1159
+ "8",
1160
+ "6",
1161
+ "7",
1162
+ "!",
1163
+ "%",
1164
+ ":",
1165
+ "Q",
1166
+ "Z",
1167
+ "$",
1168
+ "X",
1169
+ "\"",
1170
+ "&",
1171
+ "*",
1172
+ "/",
1173
+ "\u00a3",
1174
+ "+",
1175
+ "\u20ac",
1176
+ "_",
1177
+ "^",
1178
+ "\u00a5"
1179
+ ]
1180
+ },
1181
+ "decoding": {
1182
+ "strategy": "greedy_batch",
1183
+ "greedy": {
1184
+ "max_symbols": 10
1185
+ },
1186
+ "beam": {
1187
+ "beam_size": 2,
1188
+ "return_best_hypothesis": false,
1189
+ "score_norm": true,
1190
+ "tsd_max_sym_exp": 50,
1191
+ "alsd_max_target_len": 2.0
1192
+ }
1193
+ },
1194
+ "aux_ctc": {
1195
+ "ctc_loss_weight": 0.3,
1196
+ "use_cer": false,
1197
+ "ctc_reduction": "mean_batch",
1198
+ "decoder": {
1199
+ "_target_": "nemo.collections.asr.modules.ConvASRDecoder",
1200
+ "feat_in": null,
1201
+ "num_classes": -1,
1202
+ "vocabulary": []
1203
+ },
1204
+ "decoding": {
1205
+ "strategy": "greedy"
1206
+ }
1207
+ },
1208
+ "interctc": {
1209
+ "loss_weights": [],
1210
+ "apply_at_layers": []
1211
+ },
1212
+ "loss": {
1213
+ "loss_name": "default",
1214
+ "warprnnt_numba_kwargs": {
1215
+ "fastemit_lambda": 0.005,
1216
+ "clamp": -1.0
1217
+ }
1218
+ },
1219
+ "optim": {
1220
+ "name": "adamw",
1221
+ "lr": 3.0,
1222
+ "betas": [
1223
+ 0.9,
1224
+ 0.98
1225
+ ],
1226
+ "weight_decay": 0.001,
1227
+ "sched": {
1228
+ "name": "NoamAnnealing",
1229
+ "d_model": 1024,
1230
+ "warmup_steps": 8000,
1231
+ "warmup_ratio": null,
1232
+ "min_lr": 1e-06
1233
+ }
1234
+ },
1235
+ "labels": [
1236
+ "<unk>",
1237
+ "\u2581t",
1238
+ "\u2581th",
1239
+ "\u2581a",
1240
+ "in",
1241
+ "\u2581the",
1242
+ "re",
1243
+ "\u2581w",
1244
+ "\u2581o",
1245
+ "\u2581s",
1246
+ "er",
1247
+ "at",
1248
+ "ou",
1249
+ "nd",
1250
+ "it",
1251
+ "is",
1252
+ "\u2581h",
1253
+ "\u2581b",
1254
+ "on",
1255
+ "\u2581c",
1256
+ "ing",
1257
+ "en",
1258
+ "\u2581to",
1259
+ "\u2581m",
1260
+ "\u2581f",
1261
+ "\u2581p",
1262
+ "or",
1263
+ "an",
1264
+ "es",
1265
+ "\u2581of",
1266
+ "\u2581d",
1267
+ "ed",
1268
+ "ll",
1269
+ "\u2581and",
1270
+ "\u2581I",
1271
+ "\u2581in",
1272
+ "\u2581l",
1273
+ "ar",
1274
+ "\u2581y",
1275
+ "\u2581g",
1276
+ "as",
1277
+ "\u2581you",
1278
+ "om",
1279
+ "\u2581n",
1280
+ "ic",
1281
+ "ve",
1282
+ "al",
1283
+ "ion",
1284
+ "us",
1285
+ "\u2581be",
1286
+ "ow",
1287
+ "le",
1288
+ "\u2581wh",
1289
+ "\u2581e",
1290
+ "ot",
1291
+ "ut",
1292
+ "\u2581it",
1293
+ "\u2581is",
1294
+ "\u2581we",
1295
+ "\u2581T",
1296
+ "\u2581re",
1297
+ "et",
1298
+ "\u2581A",
1299
+ "ent",
1300
+ "\u2581on",
1301
+ "\u2581ha",
1302
+ "ay",
1303
+ "\u2581S",
1304
+ "ct",
1305
+ "\u2581Th",
1306
+ "ver",
1307
+ "id",
1308
+ "ig",
1309
+ "im",
1310
+ "ro",
1311
+ "\u2581for",
1312
+ "ly",
1313
+ "\u2581he",
1314
+ "ke",
1315
+ "ld",
1316
+ "se",
1317
+ "st",
1318
+ "ch",
1319
+ "\u2581st",
1320
+ "all",
1321
+ "ce",
1322
+ "ur",
1323
+ "ith",
1324
+ "am",
1325
+ "if",
1326
+ "ir",
1327
+ "\u2581go",
1328
+ "\u2581u",
1329
+ "\u2581as",
1330
+ "\u2581was",
1331
+ "ad",
1332
+ "\u2581W",
1333
+ "\u2581k",
1334
+ "\u2581an",
1335
+ "ht",
1336
+ "th",
1337
+ "\u2581r",
1338
+ "\u2581are",
1339
+ "ere",
1340
+ "\u2581se",
1341
+ "\u2581do",
1342
+ "\u2581B",
1343
+ "\u2581so",
1344
+ "\u2581sh",
1345
+ "\u2581not",
1346
+ "\u2581li",
1347
+ "od",
1348
+ "\u2581C",
1349
+ "ust",
1350
+ "ill",
1351
+ "ight",
1352
+ "ally",
1353
+ "\u2581And",
1354
+ "ter",
1355
+ "\u2581or",
1356
+ "\u2581me",
1357
+ "\u2581M",
1358
+ "ome",
1359
+ "op",
1360
+ "\u2581at",
1361
+ "il",
1362
+ "\u2581The",
1363
+ "ould",
1364
+ "\u2581j",
1365
+ "ant",
1366
+ "\u2581So",
1367
+ "\u2581H",
1368
+ "ol",
1369
+ "ain",
1370
+ "\u2581can",
1371
+ "\u2581de",
1372
+ "\u2581ne",
1373
+ "ore",
1374
+ "\u2581con",
1375
+ "\u2581kn",
1376
+ "ck",
1377
+ "ul",
1378
+ "\u2581fr",
1379
+ "\u2581ab",
1380
+ "ers",
1381
+ "ess",
1382
+ "ge",
1383
+ "\u2581pro",
1384
+ "pe",
1385
+ "ate",
1386
+ "\u2581su",
1387
+ "\u2581com",
1388
+ "\u2581but",
1389
+ "\u2581all",
1390
+ "est",
1391
+ "qu",
1392
+ "\u2581ex",
1393
+ "\u2581al",
1394
+ "ra",
1395
+ "\u2581O",
1396
+ "out",
1397
+ "use",
1398
+ "very",
1399
+ "pp",
1400
+ "\u2581Y",
1401
+ "\u2581ch",
1402
+ "ri",
1403
+ "ist",
1404
+ "\u2581v",
1405
+ "\u2581lo",
1406
+ "ment",
1407
+ "art",
1408
+ "\u2581P",
1409
+ "nt",
1410
+ "ab",
1411
+ "\u2581one",
1412
+ "\u2581N",
1413
+ "ive",
1414
+ "\u2581wor",
1415
+ "ions",
1416
+ "ort",
1417
+ "\u2581L",
1418
+ "\u2581by",
1419
+ "ich",
1420
+ "\u2581my",
1421
+ "ity",
1422
+ "ok",
1423
+ "\u2581G",
1424
+ "res",
1425
+ "\u2581up",
1426
+ "un",
1427
+ "um",
1428
+ "ea",
1429
+ "ind",
1430
+ "and",
1431
+ "ink",
1432
+ "el",
1433
+ "\u2581D",
1434
+ "em",
1435
+ "\u2581E",
1436
+ "os",
1437
+ "oug",
1438
+ "\u2581if",
1439
+ "ca",
1440
+ "\u2581out",
1441
+ "\u2581int",
1442
+ "ie",
1443
+ "\u2581F",
1444
+ "\u2581It",
1445
+ "\u2581his",
1446
+ "ard",
1447
+ "\u2581had",
1448
+ "\u2581tr",
1449
+ "her",
1450
+ "our",
1451
+ "ies",
1452
+ "ake",
1453
+ "\u2581R",
1454
+ "\u2581We",
1455
+ "\u2581get",
1456
+ "\u2581don",
1457
+ "\u2581us",
1458
+ "ak",
1459
+ "\u2581pl",
1460
+ "ect",
1461
+ "ure",
1462
+ "ame",
1463
+ "ast",
1464
+ "\u2581who",
1465
+ "ack",
1466
+ "\u2581le",
1467
+ "\u2581sa",
1468
+ "iv",
1469
+ "ci",
1470
+ "ide",
1471
+ "\u2581tim",
1472
+ "\u2581our",
1473
+ "ound",
1474
+ "ous",
1475
+ "\u2581co",
1476
+ "\u2581pe",
1477
+ "ose",
1478
+ "ud",
1479
+ "\u2581see",
1480
+ "ough",
1481
+ "\u2581man",
1482
+ "\u2581qu",
1483
+ "\u2581You",
1484
+ "so",
1485
+ "ople",
1486
+ "\u2581Wh",
1487
+ "ong",
1488
+ "ap",
1489
+ "ther",
1490
+ "\u2581J",
1491
+ "are",
1492
+ "ine",
1493
+ "\u2581say",
1494
+ "\u2581im",
1495
+ "\u2581But",
1496
+ "ings",
1497
+ "\u2581has",
1498
+ "\u2581ag",
1499
+ "ff",
1500
+ "\u2581her",
1501
+ "itt",
1502
+ "one",
1503
+ "\u2581en",
1504
+ "\u2581ar",
1505
+ "\u2581fe",
1506
+ "ven",
1507
+ "\u2581any",
1508
+ "\u2581mo",
1509
+ "reat",
1510
+ "ag",
1511
+ "\u2581how",
1512
+ "\u2581cl",
1513
+ "pt",
1514
+ "\u2581now",
1515
+ "own",
1516
+ "ber",
1517
+ "\u2581him",
1518
+ "\u2581act",
1519
+ "hing",
1520
+ "ice",
1521
+ "\u2581no",
1522
+ "ans",
1523
+ "iz",
1524
+ "\u2581fa",
1525
+ "per",
1526
+ "pl",
1527
+ "\u2581te",
1528
+ "\u2581ad",
1529
+ "age",
1530
+ "ree",
1531
+ "\u2581tw",
1532
+ "ank",
1533
+ "\u2581He",
1534
+ "ple",
1535
+ "ite",
1536
+ "ry",
1537
+ "\u2581U",
1538
+ "ish",
1539
+ "ire",
1540
+ "ue",
1541
+ "\u2581In",
1542
+ "\u2581she",
1543
+ "ble",
1544
+ "cc",
1545
+ "nder",
1546
+ "\u2581way",
1547
+ "\u2581pr",
1548
+ "ear",
1549
+ "\u2581did",
1550
+ "\u2581po",
1551
+ "eah",
1552
+ "\u2581un",
1553
+ "omet",
1554
+ "ence",
1555
+ "ep",
1556
+ "uch",
1557
+ "\u2581sp",
1558
+ "ach",
1559
+ "og",
1560
+ "ance",
1561
+ "able",
1562
+ "iff",
1563
+ "sel",
1564
+ "\u2581got",
1565
+ "way",
1566
+ "\u2581gr",
1567
+ "alk",
1568
+ "\u2581res",
1569
+ "ated",
1570
+ "irst",
1571
+ "ick",
1572
+ "ass",
1573
+ "\u2581two",
1574
+ "\u2581dis",
1575
+ "ord",
1576
+ "\u2581pre",
1577
+ "ount",
1578
+ "ase",
1579
+ "ip",
1580
+ "ult",
1581
+ "ical",
1582
+ "orm",
1583
+ "ary",
1584
+ "ace",
1585
+ "\u2581spe",
1586
+ "\u2581Ch",
1587
+ "\u2581thr",
1588
+ "\u2581imp",
1589
+ "int",
1590
+ "\u2581am",
1591
+ "\u2581off",
1592
+ "act",
1593
+ "ia",
1594
+ "\u2581ro",
1595
+ "ress",
1596
+ "\u2581per",
1597
+ "\u2581fo",
1598
+ "\u2581br",
1599
+ "\u2581K",
1600
+ "vel",
1601
+ "\u2581gu",
1602
+ "\u2581bo",
1603
+ "ang",
1604
+ "kay",
1605
+ "ub",
1606
+ "ign",
1607
+ "\u2581may",
1608
+ "ving",
1609
+ "ces",
1610
+ "ens",
1611
+ "cl",
1612
+ "\u2581lot",
1613
+ "ru",
1614
+ "ade",
1615
+ "\u2581bet",
1616
+ "\u2581bl",
1617
+ "\u2581let",
1618
+ "fore",
1619
+ "co",
1620
+ "ild",
1621
+ "ning",
1622
+ "xt",
1623
+ "ile",
1624
+ "ark",
1625
+ "self",
1626
+ "\u2581app",
1627
+ "ory",
1628
+ "du",
1629
+ "\u2581day",
1630
+ "\u2581St",
1631
+ "ater",
1632
+ "\u2581use",
1633
+ "ys",
1634
+ "fter",
1635
+ "\u2581new",
1636
+ "ious",
1637
+ "ial",
1638
+ "he",
1639
+ "wn",
1640
+ "ved",
1641
+ "red",
1642
+ "\u2581fl",
1643
+ "iss",
1644
+ "ody",
1645
+ "form",
1646
+ "ian",
1647
+ "tain",
1648
+ "\u2581bu",
1649
+ "\u2581V",
1650
+ "\u2581rec",
1651
+ "ty",
1652
+ "be",
1653
+ "\u2581sc",
1654
+ "ors",
1655
+ "vers",
1656
+ "\u2581put",
1657
+ "ife",
1658
+ "\u2581If",
1659
+ "we",
1660
+ "te",
1661
+ "ject",
1662
+ "ath",
1663
+ "ting",
1664
+ "\u2581rem",
1665
+ "\u2581acc",
1666
+ "ull",
1667
+ "ons",
1668
+ "\u2581ind",
1669
+ "\u2581ser",
1670
+ "\u2581ke",
1671
+ "ates",
1672
+ "ves",
1673
+ "na",
1674
+ "lic",
1675
+ "\u2581des",
1676
+ "\u2581its",
1677
+ "ful",
1678
+ "ents",
1679
+ "erm",
1680
+ "ac",
1681
+ "ered",
1682
+ "ise",
1683
+ "\u2581sy",
1684
+ "urn",
1685
+ "\u2581em",
1686
+ "oth",
1687
+ "ual",
1688
+ "ne",
1689
+ "ward",
1690
+ "ib",
1691
+ "\u2581try",
1692
+ "\u2581pos",
1693
+ "nds",
1694
+ "ft",
1695
+ "get",
1696
+ "ph",
1697
+ "\u2581ob",
1698
+ "ady",
1699
+ "igh",
1700
+ "ood",
1701
+ "\u2581rel",
1702
+ "\u2581wr",
1703
+ "ug",
1704
+ "ears",
1705
+ "ail",
1706
+ "\u2581Now",
1707
+ "\u2581bit",
1708
+ "ng",
1709
+ "\u2581Oh",
1710
+ "\u2581hel",
1711
+ "ange",
1712
+ "\u2581reg",
1713
+ "\u2581rep",
1714
+ "\u2581bel",
1715
+ "\u2581sm",
1716
+ "ost",
1717
+ "tern",
1718
+ "gr",
1719
+ "\u2581own",
1720
+ "\u2581end",
1721
+ "pect",
1722
+ "ily",
1723
+ "day",
1724
+ "ied",
1725
+ "ific",
1726
+ "ower",
1727
+ "\u2581add",
1728
+ "cess",
1729
+ "ict",
1730
+ "ible",
1731
+ "\u2581bas",
1732
+ "\u2581i",
1733
+ "\u2581op",
1734
+ "cial",
1735
+ "ular",
1736
+ "\u2581Be",
1737
+ "ced",
1738
+ "\u2581too",
1739
+ "ks",
1740
+ "ew",
1741
+ "mer",
1742
+ "\u2581ph",
1743
+ "ob",
1744
+ "==",
1745
+ "\u2581la",
1746
+ "\u2581set",
1747
+ "\u2581min",
1748
+ "\u2581sub",
1749
+ "\u2581gen",
1750
+ "atch",
1751
+ "..",
1752
+ "\u2581inv",
1753
+ "\u2581As",
1754
+ "\u2581nat",
1755
+ "\u2581sl",
1756
+ "\u2581num",
1757
+ "av",
1758
+ "ways",
1759
+ "\u2581God",
1760
+ "stem",
1761
+ "\u2581ac",
1762
+ "\u2581att",
1763
+ "\u2581ev",
1764
+ "\u2581def",
1765
+ "llow",
1766
+ "\u2581str",
1767
+ "lect",
1768
+ "ars",
1769
+ "\u2581cr",
1770
+ "\u2581Is",
1771
+ "olog",
1772
+ "les",
1773
+ "oy",
1774
+ "\u2581ask",
1775
+ "\u2581inc",
1776
+ "body",
1777
+ "\u2581ent",
1778
+ "\u2581pol",
1779
+ "ness",
1780
+ "ix",
1781
+ "\u2581why",
1782
+ "onna",
1783
+ "\u2581ear",
1784
+ "\u2581tak",
1785
+ "\u2581Un",
1786
+ "ited",
1787
+ "mun",
1788
+ "li",
1789
+ "ute",
1790
+ "ract",
1791
+ "\u2581dec",
1792
+ "uro",
1793
+ "\u2581mak",
1794
+ "\u2581fin",
1795
+ "ween",
1796
+ "\u2581No",
1797
+ "arch",
1798
+ "\u2581bec",
1799
+ "gan",
1800
+ "old",
1801
+ "cy",
1802
+ "\u2581big",
1803
+ "\u2581For",
1804
+ "ren",
1805
+ "als",
1806
+ "und",
1807
+ "\u2581Al",
1808
+ "\u2581All",
1809
+ "ss",
1810
+ "ows",
1811
+ "\u2581mod",
1812
+ "ock",
1813
+ "\u2581id",
1814
+ "ism",
1815
+ "cus",
1816
+ "\u2581gl",
1817
+ "ably",
1818
+ "\u2581ass",
1819
+ "\u2581car",
1820
+ "ata",
1821
+ "ppen",
1822
+ "led",
1823
+ "\u2581sim",
1824
+ "\u2581mon",
1825
+ "ics",
1826
+ "\u2581giv",
1827
+ "cept",
1828
+ "\u2581Mr",
1829
+ "pan",
1830
+ "\u2581pub",
1831
+ "\u2581eff",
1832
+ "\u2581How",
1833
+ "ps",
1834
+ "vern",
1835
+ "end",
1836
+ "hip",
1837
+ "iew",
1838
+ "ope",
1839
+ "\u2581An",
1840
+ "\u2581She",
1841
+ "\u2581Com",
1842
+ "ee",
1843
+ "ures",
1844
+ "ell",
1845
+ "ouse",
1846
+ "cond",
1847
+ "king",
1848
+ "oc",
1849
+ "ues",
1850
+ "ever",
1851
+ "\u2581To",
1852
+ "clud",
1853
+ "\u2581ins",
1854
+ "\u2581exp",
1855
+ "\u2581old",
1856
+ "\u2581mem",
1857
+ "\u2581ref",
1858
+ "\u2581tra",
1859
+ "\u2581far",
1860
+ "ave",
1861
+ "rat",
1862
+ "\u2581sur",
1863
+ "ruct",
1864
+ "rib",
1865
+ "duct",
1866
+ "uff",
1867
+ "\u2581met",
1868
+ "\u2581sch",
1869
+ "ince",
1870
+ "\u2581run",
1871
+ "ense",
1872
+ "\u2581cle",
1873
+ "\u2581==",
1874
+ "mon",
1875
+ "ize",
1876
+ "\u2581ord",
1877
+ "blem",
1878
+ "tin",
1879
+ "\u2581Let",
1880
+ "ner",
1881
+ "ond",
1882
+ "its",
1883
+ "\u2581cor",
1884
+ "land",
1885
+ "\u2581cur",
1886
+ "\u2581Re",
1887
+ "\u2581bus",
1888
+ "\u2581uh",
1889
+ "air",
1890
+ "ote",
1891
+ "ants",
1892
+ "ason",
1893
+ "ric",
1894
+ "\u2581el",
1895
+ "\u2581cer",
1896
+ "nce",
1897
+ "\u2581fam",
1898
+ "\u2581cap",
1899
+ "uck",
1900
+ "ool",
1901
+ "ried",
1902
+ "\u2581cou",
1903
+ "\u2581fun",
1904
+ "\u2581wom",
1905
+ "\u2581hum",
1906
+ "\u2581ty",
1907
+ "\u2581ap",
1908
+ "ike",
1909
+ "\u2581few",
1910
+ "oney",
1911
+ "\u2581inf",
1912
+ "ont",
1913
+ "ese",
1914
+ "ook",
1915
+ "gy",
1916
+ "uth",
1917
+ "ulat",
1918
+ "ieve",
1919
+ "ized",
1920
+ "ross",
1921
+ "\u2581ple",
1922
+ "\u2581um",
1923
+ "\u2581val",
1924
+ "\u2581equ",
1925
+ "\u2581lea",
1926
+ "\u2581lar",
1927
+ "ah",
1928
+ "eral",
1929
+ "\u2581ed",
1930
+ "ared",
1931
+ "lish",
1932
+ "arn",
1933
+ "ds",
1934
+ "esn",
1935
+ "\u2581iss",
1936
+ "\u2581ca",
1937
+ "ted",
1938
+ "ices",
1939
+ "\u2581wee",
1940
+ "ash",
1941
+ "\u2581top",
1942
+ "ten",
1943
+ "up",
1944
+ "ts",
1945
+ "gin",
1946
+ "con",
1947
+ "ari",
1948
+ "\u2581opp",
1949
+ "osed",
1950
+ "\u2581eas",
1951
+ "\u2581ext",
1952
+ "gg",
1953
+ "az",
1954
+ "\u2581Fr",
1955
+ "ideo",
1956
+ "izat",
1957
+ "\u2581men",
1958
+ "\u2581mom",
1959
+ "\u2581ret",
1960
+ "tty",
1961
+ "rist",
1962
+ "\u2581gra",
1963
+ "alth",
1964
+ "ef",
1965
+ "\u2581det",
1966
+ "ax",
1967
+ "\u2581mat",
1968
+ "chn",
1969
+ "ern",
1970
+ "peri",
1971
+ "\u2581bre",
1972
+ "\u2581Sh",
1973
+ "sw",
1974
+ "erat",
1975
+ "\u2581sit",
1976
+ "ters",
1977
+ "ale",
1978
+ "man",
1979
+ "\u2581sol",
1980
+ "ork",
1981
+ "\u2581adv",
1982
+ "ety",
1983
+ "\u2581vis",
1984
+ "\u2581med",
1985
+ "uc",
1986
+ "less",
1987
+ "\u2581unt",
1988
+ "gram",
1989
+ "ets",
1990
+ "ists",
1991
+ "\u2581ey",
1992
+ "\u2581col",
1993
+ "imes",
1994
+ "\u2581law",
1995
+ "\u2581pri",
1996
+ "sid",
1997
+ "\u2581On",
1998
+ "\u2581mot",
1999
+ "ield",
2000
+ "\u2581Do",
2001
+ "\u2581At",
2002
+ "ages",
2003
+ "amp",
2004
+ "\u2581art",
2005
+ "miss",
2006
+ "\u2581sk",
2007
+ "alf",
2008
+ "pr",
2009
+ "ier",
2010
+ "\u2581beh",
2011
+ "\u2581Yes",
2012
+ "ural",
2013
+ "ime",
2014
+ "\u2581wa",
2015
+ "oks",
2016
+ "bers",
2017
+ "ger",
2018
+ "ient",
2019
+ "ries",
2020
+ "...",
2021
+ "\u2581che",
2022
+ "\u2581Br",
2023
+ "ird",
2024
+ "\u2581Ar",
2025
+ "\u2581war",
2026
+ "inat",
2027
+ "\u2581My",
2028
+ "ital",
2029
+ "wh",
2030
+ "med",
2031
+ "\u2581pur",
2032
+ "ully",
2033
+ "\u2581One",
2034
+ "\u2581rat",
2035
+ "ines",
2036
+ "\u2581Of",
2037
+ "io",
2038
+ "\u2581loc",
2039
+ "ret",
2040
+ "ctor",
2041
+ "\u2581leg",
2042
+ "stit",
2043
+ "ined",
2044
+ "ught",
2045
+ "\u2581dur",
2046
+ "\u2581es",
2047
+ "vent",
2048
+ "aj",
2049
+ "\u2581bro",
2050
+ "\u2581saw",
2051
+ "\u2581sec",
2052
+ "ream",
2053
+ "\u2581pop",
2054
+ "reen",
2055
+ "\u2581Ind",
2056
+ "els",
2057
+ "\u2581yet",
2058
+ "ired",
2059
+ "\u2581sw",
2060
+ "tro",
2061
+ "oup",
2062
+ "most",
2063
+ "pean",
2064
+ "eds",
2065
+ "ush",
2066
+ "oh",
2067
+ "\u2581Se",
2068
+ "\u2581tea",
2069
+ "ann",
2070
+ "ilit",
2071
+ "err",
2072
+ "pend",
2073
+ "ton",
2074
+ "ased",
2075
+ "\u2581aff",
2076
+ "\u2581mor",
2077
+ "\u2581dra",
2078
+ "put",
2079
+ "\u2581dr",
2080
+ "ins",
2081
+ "uat",
2082
+ "nect",
2083
+ "cri",
2084
+ "outh",
2085
+ "\u2581ra",
2086
+ "\u2581pay",
2087
+ "ms",
2088
+ "\u2581av",
2089
+ "bs",
2090
+ "ling",
2091
+ "\u2581De",
2092
+ "\u2581Or",
2093
+ "ove",
2094
+ "\u2581Can",
2095
+ "\u2581eng",
2096
+ "ames",
2097
+ "ided",
2098
+ "\u2581Go",
2099
+ "mitt",
2100
+ "ode",
2101
+ "\u2581cre",
2102
+ "par",
2103
+ "ides",
2104
+ "pos",
2105
+ "\u2581fav",
2106
+ "\u2581air",
2107
+ "\u2581New",
2108
+ "\u2581bad",
2109
+ "\u2581six",
2110
+ "vat",
2111
+ "\u2581pat",
2112
+ "not",
2113
+ "\u2581di",
2114
+ "rop",
2115
+ "ral",
2116
+ "orn",
2117
+ "\u2581par",
2118
+ "cing",
2119
+ "\u2581aw",
2120
+ "orts",
2121
+ "ox",
2122
+ "\u2581yes",
2123
+ "cuss",
2124
+ "eng",
2125
+ "ives",
2126
+ "erms",
2127
+ "\u2581job",
2128
+ "mand",
2129
+ "ying",
2130
+ "\u2581occ",
2131
+ "aps",
2132
+ "ases",
2133
+ "\u2581Not",
2134
+ "rent",
2135
+ "ency",
2136
+ "att",
2137
+ "ised",
2138
+ "vice",
2139
+ "\u2581Eng",
2140
+ "\u2581est",
2141
+ "oked",
2142
+ "\u2581Q",
2143
+ "iron",
2144
+ "idd",
2145
+ "me",
2146
+ "unch",
2147
+ "ane",
2148
+ "\u2581z",
2149
+ "br",
2150
+ "arts",
2151
+ "\u2581fat",
2152
+ "ery",
2153
+ "anks",
2154
+ "\u2581jo",
2155
+ "\u2581mar",
2156
+ "aw",
2157
+ "ott",
2158
+ "ards",
2159
+ "\u2581oh",
2160
+ "ians",
2161
+ "\u2581sci",
2162
+ "row",
2163
+ "unt",
2164
+ "ury",
2165
+ "\u2581abs",
2166
+ "ergy",
2167
+ "\u2581Z",
2168
+ "ump",
2169
+ "\u2581Am",
2170
+ "ened",
2171
+ "angu",
2172
+ "\u2581Pro",
2173
+ "icat",
2174
+ "itch",
2175
+ "\u2581dri",
2176
+ "iat",
2177
+ "\u2581",
2178
+ "e",
2179
+ "t",
2180
+ "o",
2181
+ "a",
2182
+ "n",
2183
+ "i",
2184
+ "s",
2185
+ "r",
2186
+ "h",
2187
+ "l",
2188
+ "d",
2189
+ "u",
2190
+ "c",
2191
+ "m",
2192
+ "y",
2193
+ "g",
2194
+ "w",
2195
+ "f",
2196
+ "p",
2197
+ ",",
2198
+ ".",
2199
+ "b",
2200
+ "v",
2201
+ "k",
2202
+ "'",
2203
+ "I",
2204
+ "T",
2205
+ "A",
2206
+ "S",
2207
+ "x",
2208
+ "W",
2209
+ "j",
2210
+ "C",
2211
+ "B",
2212
+ "M",
2213
+ "?",
2214
+ "H",
2215
+ "O",
2216
+ "0",
2217
+ "P",
2218
+ "q",
2219
+ "Y",
2220
+ "N",
2221
+ "L",
2222
+ "D",
2223
+ "1",
2224
+ "E",
2225
+ "G",
2226
+ "z",
2227
+ "F",
2228
+ "R",
2229
+ "-",
2230
+ "2",
2231
+ "J",
2232
+ "U",
2233
+ "9",
2234
+ "K",
2235
+ "5",
2236
+ "3",
2237
+ "V",
2238
+ "=",
2239
+ "4",
2240
+ "8",
2241
+ "6",
2242
+ "7",
2243
+ "!",
2244
+ "%",
2245
+ ":",
2246
+ "Q",
2247
+ "Z",
2248
+ "$",
2249
+ "X",
2250
+ "\"",
2251
+ "&",
2252
+ "*",
2253
+ "/",
2254
+ "\u00a3",
2255
+ "+",
2256
+ "\u20ac",
2257
+ "_",
2258
+ "^",
2259
+ "\u00a5"
2260
+ ],
2261
+ "target": "nemo.collections.asr.models.rnnt_bpe_models.EncDecRNNTBPEModel",
2262
+ "nemo_version": "2.4.0rc0",
2263
+ "quantization": {
2264
+ "bits": 8,
2265
+ "group_size": 64
2266
+ }
2267
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e87a607c1fcdd81a3b8f27c721d88469b586bad91c0401aca2f22ed6601ded0
3
+ size 663931242
model_config.yaml ADDED
@@ -0,0 +1,2227 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ sample_rate: 16000
2
+ compute_eval_loss: false
3
+ log_prediction: true
4
+ rnnt_reduction: mean_volume
5
+ skip_nan_grad: false
6
+ model_defaults:
7
+ enc_hidden: 1024
8
+ pred_hidden: 640
9
+ joint_hidden: 640
10
+ train_ds:
11
+ use_lhotse: true
12
+ skip_missing_manifest_entries: true
13
+ input_cfg: null
14
+ tarred_audio_filepaths: null
15
+ manifest_filepath: null
16
+ sample_rate: 16000
17
+ shuffle: true
18
+ num_workers: 2
19
+ pin_memory: true
20
+ max_duration: 40.0
21
+ min_duration: 0.1
22
+ text_field: answer
23
+ batch_duration: null
24
+ use_bucketing: true
25
+ max_tps: null
26
+ bucket_duration_bins: null
27
+ bucket_batch_size: null
28
+ num_buckets: null
29
+ bucket_buffer_size: null
30
+ shuffle_buffer_size: null
31
+ augmentor: null
32
+ validation_ds:
33
+ use_lhotse: true
34
+ manifest_filepath: /data/ASR/en/librispeech/test-other.json
35
+ sample_rate: 16000
36
+ batch_size: 32
37
+ shuffle: false
38
+ max_duration: 40.0
39
+ min_duration: 0.1
40
+ num_workers: 2
41
+ pin_memory: true
42
+ text_field: answer
43
+ tarred_audio_filepaths: null
44
+ tokenizer:
45
+ dir: null
46
+ type: bpe
47
+ model_path: nemo:e8dc9587bf16427491071501d9080e4e_tokenizer.model
48
+ vocab_path: nemo:5f463a58a6f9401796aae466cbf12f7d_vocab.txt
49
+ spe_tokenizer_vocab: nemo:26b131bdef564f2cbe451e9219efcbab_tokenizer.vocab
50
+ preprocessor:
51
+ _target_: nemo.collections.asr.modules.AudioToMelSpectrogramPreprocessor
52
+ sample_rate: 16000
53
+ normalize: NA
54
+ window_size: 0.025
55
+ window_stride: 0.01
56
+ window: hann
57
+ features: 128
58
+ n_fft: 512
59
+ log: true
60
+ frame_splicing: 1
61
+ dither: 1.0e-05
62
+ pad_to: 0
63
+ pad_value: 0.0
64
+ spec_augment:
65
+ _target_: nemo.collections.asr.modules.SpectrogramAugmentation
66
+ freq_masks: 2
67
+ time_masks: 10
68
+ freq_width: 27
69
+ time_width: 0.05
70
+ encoder:
71
+ _target_: nemo.collections.asr.modules.ConformerEncoder
72
+ feat_in: 128
73
+ feat_out: -1
74
+ n_layers: 24
75
+ d_model: 1024
76
+ use_bias: false
77
+ subsampling: dw_striding
78
+ subsampling_factor: 8
79
+ subsampling_conv_channels: 256
80
+ causal_downsampling: true
81
+ reduction: null
82
+ reduction_position: null
83
+ reduction_factor: 1
84
+ ff_expansion_factor: 4
85
+ self_attention_model: rel_pos
86
+ n_heads: 8
87
+ att_context_size:
88
+ - - 70
89
+ - 13
90
+ - - 70
91
+ - 6
92
+ - - 70
93
+ - 1
94
+ - - 70
95
+ - 0
96
+ att_context_style: chunked_limited
97
+ xscaling: false
98
+ untie_biases: true
99
+ pos_emb_max_len: 5000
100
+ conv_kernel_size: 9
101
+ conv_norm_type: layer_norm
102
+ conv_context_size: causal
103
+ dropout: 0.1
104
+ dropout_pre_encoder: 0.1
105
+ dropout_emb: 0.0
106
+ dropout_att: 0.1
107
+ stochastic_depth_drop_prob: 0.0
108
+ stochastic_depth_mode: linear
109
+ stochastic_depth_start_layer: 1
110
+ decoder:
111
+ _target_: nemo.collections.asr.modules.RNNTDecoder
112
+ normalization_mode: null
113
+ random_state_sampling: false
114
+ blank_as_pad: true
115
+ prednet:
116
+ pred_hidden: 640
117
+ pred_rnn_layers: 2
118
+ t_max: null
119
+ dropout: 0.2
120
+ vocab_size: 1024
121
+ joint:
122
+ _target_: nemo.collections.asr.modules.RNNTJoint
123
+ log_softmax: null
124
+ preserve_memory: false
125
+ fuse_loss_wer: true
126
+ fused_batch_size: 4
127
+ jointnet:
128
+ joint_hidden: 640
129
+ activation: relu
130
+ dropout: 0.2
131
+ encoder_hidden: 1024
132
+ pred_hidden: 640
133
+ num_classes: 1024
134
+ vocabulary:
135
+ - <unk>
136
+ - ▁t
137
+ - ▁th
138
+ - ▁a
139
+ - in
140
+ - ▁the
141
+ - re
142
+ - ▁w
143
+ - ▁o
144
+ - ▁s
145
+ - er
146
+ - at
147
+ - ou
148
+ - nd
149
+ - it
150
+ - is
151
+ - ▁h
152
+ - ▁b
153
+ - 'on'
154
+ - ▁c
155
+ - ing
156
+ - en
157
+ - ▁to
158
+ - ▁m
159
+ - ▁f
160
+ - ▁p
161
+ - or
162
+ - an
163
+ - es
164
+ - ▁of
165
+ - ▁d
166
+ - ed
167
+ - ll
168
+ - ▁and
169
+ - ▁I
170
+ - ▁in
171
+ - ▁l
172
+ - ar
173
+ - ▁y
174
+ - ▁g
175
+ - as
176
+ - ▁you
177
+ - om
178
+ - ▁n
179
+ - ic
180
+ - ve
181
+ - al
182
+ - ion
183
+ - us
184
+ - ▁be
185
+ - ow
186
+ - le
187
+ - ▁wh
188
+ - ▁e
189
+ - ot
190
+ - ut
191
+ - ▁it
192
+ - ▁is
193
+ - ▁we
194
+ - ▁T
195
+ - ▁re
196
+ - et
197
+ - ▁A
198
+ - ent
199
+ - ▁on
200
+ - ▁ha
201
+ - ay
202
+ - ▁S
203
+ - ct
204
+ - ▁Th
205
+ - ver
206
+ - id
207
+ - ig
208
+ - im
209
+ - ro
210
+ - ▁for
211
+ - ly
212
+ - ▁he
213
+ - ke
214
+ - ld
215
+ - se
216
+ - st
217
+ - ch
218
+ - ▁st
219
+ - all
220
+ - ce
221
+ - ur
222
+ - ith
223
+ - am
224
+ - if
225
+ - ir
226
+ - ▁go
227
+ - ▁u
228
+ - ▁as
229
+ - ▁was
230
+ - ad
231
+ - ▁W
232
+ - ▁k
233
+ - ▁an
234
+ - ht
235
+ - th
236
+ - ▁r
237
+ - ▁are
238
+ - ere
239
+ - ▁se
240
+ - ▁do
241
+ - ▁B
242
+ - ▁so
243
+ - ▁sh
244
+ - ▁not
245
+ - ▁li
246
+ - od
247
+ - ▁C
248
+ - ust
249
+ - ill
250
+ - ight
251
+ - ally
252
+ - ▁And
253
+ - ter
254
+ - ▁or
255
+ - ▁me
256
+ - ▁M
257
+ - ome
258
+ - op
259
+ - ▁at
260
+ - il
261
+ - ▁The
262
+ - ould
263
+ - ▁j
264
+ - ant
265
+ - ▁So
266
+ - ▁H
267
+ - ol
268
+ - ain
269
+ - ▁can
270
+ - ▁de
271
+ - ▁ne
272
+ - ore
273
+ - ▁con
274
+ - ▁kn
275
+ - ck
276
+ - ul
277
+ - ▁fr
278
+ - ▁ab
279
+ - ers
280
+ - ess
281
+ - ge
282
+ - ▁pro
283
+ - pe
284
+ - ate
285
+ - ▁su
286
+ - ▁com
287
+ - ▁but
288
+ - ▁all
289
+ - est
290
+ - qu
291
+ - ▁ex
292
+ - ▁al
293
+ - ra
294
+ - ▁O
295
+ - out
296
+ - use
297
+ - very
298
+ - pp
299
+ - ▁Y
300
+ - ▁ch
301
+ - ri
302
+ - ist
303
+ - ▁v
304
+ - ▁lo
305
+ - ment
306
+ - art
307
+ - ▁P
308
+ - nt
309
+ - ab
310
+ - ▁one
311
+ - ▁N
312
+ - ive
313
+ - ▁wor
314
+ - ions
315
+ - ort
316
+ - ▁L
317
+ - ▁by
318
+ - ich
319
+ - ▁my
320
+ - ity
321
+ - ok
322
+ - ▁G
323
+ - res
324
+ - ▁up
325
+ - un
326
+ - um
327
+ - ea
328
+ - ind
329
+ - and
330
+ - ink
331
+ - el
332
+ - ▁D
333
+ - em
334
+ - ▁E
335
+ - os
336
+ - oug
337
+ - ▁if
338
+ - ca
339
+ - ▁out
340
+ - ▁int
341
+ - ie
342
+ - ▁F
343
+ - ▁It
344
+ - ▁his
345
+ - ard
346
+ - ▁had
347
+ - ▁tr
348
+ - her
349
+ - our
350
+ - ies
351
+ - ake
352
+ - ▁R
353
+ - ▁We
354
+ - ▁get
355
+ - ▁don
356
+ - ▁us
357
+ - ak
358
+ - ▁pl
359
+ - ect
360
+ - ure
361
+ - ame
362
+ - ast
363
+ - ▁who
364
+ - ack
365
+ - ▁le
366
+ - ▁sa
367
+ - iv
368
+ - ci
369
+ - ide
370
+ - ▁tim
371
+ - ▁our
372
+ - ound
373
+ - ous
374
+ - ▁co
375
+ - ▁pe
376
+ - ose
377
+ - ud
378
+ - ▁see
379
+ - ough
380
+ - ▁man
381
+ - ▁qu
382
+ - ▁You
383
+ - so
384
+ - ople
385
+ - ▁Wh
386
+ - ong
387
+ - ap
388
+ - ther
389
+ - ▁J
390
+ - are
391
+ - ine
392
+ - ▁say
393
+ - ▁im
394
+ - ▁But
395
+ - ings
396
+ - ▁has
397
+ - ▁ag
398
+ - ff
399
+ - ▁her
400
+ - itt
401
+ - one
402
+ - ▁en
403
+ - ▁ar
404
+ - ▁fe
405
+ - ven
406
+ - ▁any
407
+ - ▁mo
408
+ - reat
409
+ - ag
410
+ - ▁how
411
+ - ▁cl
412
+ - pt
413
+ - ▁now
414
+ - own
415
+ - ber
416
+ - ▁him
417
+ - ▁act
418
+ - hing
419
+ - ice
420
+ - ▁no
421
+ - ans
422
+ - iz
423
+ - ▁fa
424
+ - per
425
+ - pl
426
+ - ▁te
427
+ - ▁ad
428
+ - age
429
+ - ree
430
+ - ▁tw
431
+ - ank
432
+ - ▁He
433
+ - ple
434
+ - ite
435
+ - ry
436
+ - ▁U
437
+ - ish
438
+ - ire
439
+ - ue
440
+ - ▁In
441
+ - ▁she
442
+ - ble
443
+ - cc
444
+ - nder
445
+ - ▁way
446
+ - ▁pr
447
+ - ear
448
+ - ▁did
449
+ - ▁po
450
+ - eah
451
+ - ▁un
452
+ - omet
453
+ - ence
454
+ - ep
455
+ - uch
456
+ - ▁sp
457
+ - ach
458
+ - og
459
+ - ance
460
+ - able
461
+ - iff
462
+ - sel
463
+ - ▁got
464
+ - way
465
+ - ▁gr
466
+ - alk
467
+ - ▁res
468
+ - ated
469
+ - irst
470
+ - ick
471
+ - ass
472
+ - ▁two
473
+ - ▁dis
474
+ - ord
475
+ - ▁pre
476
+ - ount
477
+ - ase
478
+ - ip
479
+ - ult
480
+ - ical
481
+ - orm
482
+ - ary
483
+ - ace
484
+ - ▁spe
485
+ - ▁Ch
486
+ - ▁thr
487
+ - ▁imp
488
+ - int
489
+ - ▁am
490
+ - ▁off
491
+ - act
492
+ - ia
493
+ - ▁ro
494
+ - ress
495
+ - ▁per
496
+ - ▁fo
497
+ - ▁br
498
+ - ▁K
499
+ - vel
500
+ - ▁gu
501
+ - ▁bo
502
+ - ang
503
+ - kay
504
+ - ub
505
+ - ign
506
+ - ▁may
507
+ - ving
508
+ - ces
509
+ - ens
510
+ - cl
511
+ - ▁lot
512
+ - ru
513
+ - ade
514
+ - ▁bet
515
+ - ▁bl
516
+ - ▁let
517
+ - fore
518
+ - co
519
+ - ild
520
+ - ning
521
+ - xt
522
+ - ile
523
+ - ark
524
+ - self
525
+ - ▁app
526
+ - ory
527
+ - du
528
+ - ▁day
529
+ - ▁St
530
+ - ater
531
+ - ▁use
532
+ - ys
533
+ - fter
534
+ - ▁new
535
+ - ious
536
+ - ial
537
+ - he
538
+ - wn
539
+ - ved
540
+ - red
541
+ - ▁fl
542
+ - iss
543
+ - ody
544
+ - form
545
+ - ian
546
+ - tain
547
+ - ▁bu
548
+ - ▁V
549
+ - ▁rec
550
+ - ty
551
+ - be
552
+ - ▁sc
553
+ - ors
554
+ - vers
555
+ - ▁put
556
+ - ife
557
+ - ▁If
558
+ - we
559
+ - te
560
+ - ject
561
+ - ath
562
+ - ting
563
+ - ▁rem
564
+ - ▁acc
565
+ - ull
566
+ - ons
567
+ - ▁ind
568
+ - ▁ser
569
+ - ▁ke
570
+ - ates
571
+ - ves
572
+ - na
573
+ - lic
574
+ - ▁des
575
+ - ▁its
576
+ - ful
577
+ - ents
578
+ - erm
579
+ - ac
580
+ - ered
581
+ - ise
582
+ - ▁sy
583
+ - urn
584
+ - ▁em
585
+ - oth
586
+ - ual
587
+ - ne
588
+ - ward
589
+ - ib
590
+ - ▁try
591
+ - ▁pos
592
+ - nds
593
+ - ft
594
+ - get
595
+ - ph
596
+ - ▁ob
597
+ - ady
598
+ - igh
599
+ - ood
600
+ - ▁rel
601
+ - ▁wr
602
+ - ug
603
+ - ears
604
+ - ail
605
+ - ▁Now
606
+ - ▁bit
607
+ - ng
608
+ - ▁Oh
609
+ - ▁hel
610
+ - ange
611
+ - ▁reg
612
+ - ▁rep
613
+ - ▁bel
614
+ - ▁sm
615
+ - ost
616
+ - tern
617
+ - gr
618
+ - ▁own
619
+ - ▁end
620
+ - pect
621
+ - ily
622
+ - day
623
+ - ied
624
+ - ific
625
+ - ower
626
+ - ▁add
627
+ - cess
628
+ - ict
629
+ - ible
630
+ - ▁bas
631
+ - ▁i
632
+ - ▁op
633
+ - cial
634
+ - ular
635
+ - ▁Be
636
+ - ced
637
+ - ▁too
638
+ - ks
639
+ - ew
640
+ - mer
641
+ - ▁ph
642
+ - ob
643
+ - ==
644
+ - ▁la
645
+ - ▁set
646
+ - ▁min
647
+ - ▁sub
648
+ - ▁gen
649
+ - atch
650
+ - ..
651
+ - ▁inv
652
+ - ▁As
653
+ - ▁nat
654
+ - ▁sl
655
+ - ▁num
656
+ - av
657
+ - ways
658
+ - ▁God
659
+ - stem
660
+ - ▁ac
661
+ - ▁att
662
+ - ▁ev
663
+ - ▁def
664
+ - llow
665
+ - ▁str
666
+ - lect
667
+ - ars
668
+ - ▁cr
669
+ - ▁Is
670
+ - olog
671
+ - les
672
+ - oy
673
+ - ▁ask
674
+ - ▁inc
675
+ - body
676
+ - ▁ent
677
+ - ▁pol
678
+ - ness
679
+ - ix
680
+ - ▁why
681
+ - onna
682
+ - ▁ear
683
+ - ▁tak
684
+ - ▁Un
685
+ - ited
686
+ - mun
687
+ - li
688
+ - ute
689
+ - ract
690
+ - ▁dec
691
+ - uro
692
+ - ▁mak
693
+ - ▁fin
694
+ - ween
695
+ - ▁No
696
+ - arch
697
+ - ▁bec
698
+ - gan
699
+ - old
700
+ - cy
701
+ - ▁big
702
+ - ▁For
703
+ - ren
704
+ - als
705
+ - und
706
+ - ▁Al
707
+ - ▁All
708
+ - ss
709
+ - ows
710
+ - ▁mod
711
+ - ock
712
+ - ▁id
713
+ - ism
714
+ - cus
715
+ - ▁gl
716
+ - ably
717
+ - ▁ass
718
+ - ▁car
719
+ - ata
720
+ - ppen
721
+ - led
722
+ - ▁sim
723
+ - ▁mon
724
+ - ics
725
+ - ▁giv
726
+ - cept
727
+ - ▁Mr
728
+ - pan
729
+ - ▁pub
730
+ - ▁eff
731
+ - ▁How
732
+ - ps
733
+ - vern
734
+ - end
735
+ - hip
736
+ - iew
737
+ - ope
738
+ - ▁An
739
+ - ▁She
740
+ - ▁Com
741
+ - ee
742
+ - ures
743
+ - ell
744
+ - ouse
745
+ - cond
746
+ - king
747
+ - oc
748
+ - ues
749
+ - ever
750
+ - ▁To
751
+ - clud
752
+ - ▁ins
753
+ - ▁exp
754
+ - ▁old
755
+ - ▁mem
756
+ - ▁ref
757
+ - ▁tra
758
+ - ▁far
759
+ - ave
760
+ - rat
761
+ - ▁sur
762
+ - ruct
763
+ - rib
764
+ - duct
765
+ - uff
766
+ - ▁met
767
+ - ▁sch
768
+ - ince
769
+ - ▁run
770
+ - ense
771
+ - ▁cle
772
+ - ▁==
773
+ - mon
774
+ - ize
775
+ - ▁ord
776
+ - blem
777
+ - tin
778
+ - ▁Let
779
+ - ner
780
+ - ond
781
+ - its
782
+ - ▁cor
783
+ - land
784
+ - ▁cur
785
+ - ▁Re
786
+ - ▁bus
787
+ - ▁uh
788
+ - air
789
+ - ote
790
+ - ants
791
+ - ason
792
+ - ric
793
+ - ▁el
794
+ - ▁cer
795
+ - nce
796
+ - ▁fam
797
+ - ▁cap
798
+ - uck
799
+ - ool
800
+ - ried
801
+ - ▁cou
802
+ - ▁fun
803
+ - ▁wom
804
+ - ▁hum
805
+ - ▁ty
806
+ - ▁ap
807
+ - ike
808
+ - ▁few
809
+ - oney
810
+ - ▁inf
811
+ - ont
812
+ - ese
813
+ - ook
814
+ - gy
815
+ - uth
816
+ - ulat
817
+ - ieve
818
+ - ized
819
+ - ross
820
+ - ▁ple
821
+ - ▁um
822
+ - ▁val
823
+ - ▁equ
824
+ - ▁lea
825
+ - ▁lar
826
+ - ah
827
+ - eral
828
+ - ▁ed
829
+ - ared
830
+ - lish
831
+ - arn
832
+ - ds
833
+ - esn
834
+ - ▁iss
835
+ - ▁ca
836
+ - ted
837
+ - ices
838
+ - ▁wee
839
+ - ash
840
+ - ▁top
841
+ - ten
842
+ - up
843
+ - ts
844
+ - gin
845
+ - con
846
+ - ari
847
+ - ▁opp
848
+ - osed
849
+ - ▁eas
850
+ - ▁ext
851
+ - gg
852
+ - az
853
+ - ▁Fr
854
+ - ideo
855
+ - izat
856
+ - ▁men
857
+ - ▁mom
858
+ - ▁ret
859
+ - tty
860
+ - rist
861
+ - ▁gra
862
+ - alth
863
+ - ef
864
+ - ▁det
865
+ - ax
866
+ - ▁mat
867
+ - chn
868
+ - ern
869
+ - peri
870
+ - ▁bre
871
+ - ▁Sh
872
+ - sw
873
+ - erat
874
+ - ▁sit
875
+ - ters
876
+ - ale
877
+ - man
878
+ - ▁sol
879
+ - ork
880
+ - ▁adv
881
+ - ety
882
+ - ▁vis
883
+ - ▁med
884
+ - uc
885
+ - less
886
+ - ▁unt
887
+ - gram
888
+ - ets
889
+ - ists
890
+ - ▁ey
891
+ - ▁col
892
+ - imes
893
+ - ▁law
894
+ - ▁pri
895
+ - sid
896
+ - ▁On
897
+ - ▁mot
898
+ - ield
899
+ - ▁Do
900
+ - ▁At
901
+ - ages
902
+ - amp
903
+ - ▁art
904
+ - miss
905
+ - ▁sk
906
+ - alf
907
+ - pr
908
+ - ier
909
+ - ▁beh
910
+ - ▁Yes
911
+ - ural
912
+ - ime
913
+ - ▁wa
914
+ - oks
915
+ - bers
916
+ - ger
917
+ - ient
918
+ - ries
919
+ - '...'
920
+ - ▁che
921
+ - ▁Br
922
+ - ird
923
+ - ▁Ar
924
+ - ▁war
925
+ - inat
926
+ - ▁My
927
+ - ital
928
+ - wh
929
+ - med
930
+ - ▁pur
931
+ - ully
932
+ - ▁One
933
+ - ▁rat
934
+ - ines
935
+ - ▁Of
936
+ - io
937
+ - ▁loc
938
+ - ret
939
+ - ctor
940
+ - ▁leg
941
+ - stit
942
+ - ined
943
+ - ught
944
+ - ▁dur
945
+ - ▁es
946
+ - vent
947
+ - aj
948
+ - ▁bro
949
+ - ▁saw
950
+ - ▁sec
951
+ - ream
952
+ - ▁pop
953
+ - reen
954
+ - ▁Ind
955
+ - els
956
+ - ▁yet
957
+ - ired
958
+ - ▁sw
959
+ - tro
960
+ - oup
961
+ - most
962
+ - pean
963
+ - eds
964
+ - ush
965
+ - oh
966
+ - ▁Se
967
+ - ▁tea
968
+ - ann
969
+ - ilit
970
+ - err
971
+ - pend
972
+ - ton
973
+ - ased
974
+ - ▁aff
975
+ - ▁mor
976
+ - ▁dra
977
+ - put
978
+ - ▁dr
979
+ - ins
980
+ - uat
981
+ - nect
982
+ - cri
983
+ - outh
984
+ - ▁ra
985
+ - ▁pay
986
+ - ms
987
+ - ▁av
988
+ - bs
989
+ - ling
990
+ - ▁De
991
+ - ▁Or
992
+ - ove
993
+ - ▁Can
994
+ - ▁eng
995
+ - ames
996
+ - ided
997
+ - ▁Go
998
+ - mitt
999
+ - ode
1000
+ - ▁cre
1001
+ - par
1002
+ - ides
1003
+ - pos
1004
+ - ▁fav
1005
+ - ▁air
1006
+ - ▁New
1007
+ - ▁bad
1008
+ - ▁six
1009
+ - vat
1010
+ - ▁pat
1011
+ - not
1012
+ - ▁di
1013
+ - rop
1014
+ - ral
1015
+ - orn
1016
+ - ▁par
1017
+ - cing
1018
+ - ▁aw
1019
+ - orts
1020
+ - ox
1021
+ - ▁yes
1022
+ - cuss
1023
+ - eng
1024
+ - ives
1025
+ - erms
1026
+ - ▁job
1027
+ - mand
1028
+ - ying
1029
+ - ▁occ
1030
+ - aps
1031
+ - ases
1032
+ - ▁Not
1033
+ - rent
1034
+ - ency
1035
+ - att
1036
+ - ised
1037
+ - vice
1038
+ - ▁Eng
1039
+ - ▁est
1040
+ - oked
1041
+ - ▁Q
1042
+ - iron
1043
+ - idd
1044
+ - me
1045
+ - unch
1046
+ - ane
1047
+ - ▁z
1048
+ - br
1049
+ - arts
1050
+ - ▁fat
1051
+ - ery
1052
+ - anks
1053
+ - ▁jo
1054
+ - ▁mar
1055
+ - aw
1056
+ - ott
1057
+ - ards
1058
+ - ▁oh
1059
+ - ians
1060
+ - ▁sci
1061
+ - row
1062
+ - unt
1063
+ - ury
1064
+ - ▁abs
1065
+ - ergy
1066
+ - ▁Z
1067
+ - ump
1068
+ - ▁Am
1069
+ - ened
1070
+ - angu
1071
+ - ▁Pro
1072
+ - icat
1073
+ - itch
1074
+ - ▁dri
1075
+ - iat
1076
+ - ▁
1077
+ - e
1078
+ - t
1079
+ - o
1080
+ - a
1081
+ - 'n'
1082
+ - i
1083
+ - s
1084
+ - r
1085
+ - h
1086
+ - l
1087
+ - d
1088
+ - u
1089
+ - c
1090
+ - m
1091
+ - 'y'
1092
+ - g
1093
+ - w
1094
+ - f
1095
+ - p
1096
+ - ','
1097
+ - .
1098
+ - b
1099
+ - v
1100
+ - k
1101
+ - ''''
1102
+ - I
1103
+ - T
1104
+ - A
1105
+ - S
1106
+ - x
1107
+ - W
1108
+ - j
1109
+ - C
1110
+ - B
1111
+ - M
1112
+ - '?'
1113
+ - H
1114
+ - O
1115
+ - '0'
1116
+ - P
1117
+ - q
1118
+ - 'Y'
1119
+ - 'N'
1120
+ - L
1121
+ - D
1122
+ - '1'
1123
+ - E
1124
+ - G
1125
+ - z
1126
+ - F
1127
+ - R
1128
+ - '-'
1129
+ - '2'
1130
+ - J
1131
+ - U
1132
+ - '9'
1133
+ - K
1134
+ - '5'
1135
+ - '3'
1136
+ - V
1137
+ - '='
1138
+ - '4'
1139
+ - '8'
1140
+ - '6'
1141
+ - '7'
1142
+ - '!'
1143
+ - '%'
1144
+ - ':'
1145
+ - Q
1146
+ - Z
1147
+ - $
1148
+ - X
1149
+ - '"'
1150
+ - '&'
1151
+ - '*'
1152
+ - /
1153
+ - £
1154
+ - +
1155
+ - €
1156
+ - _
1157
+ - ^
1158
+ - ¥
1159
+ decoding:
1160
+ strategy: greedy_batch
1161
+ greedy:
1162
+ max_symbols: 10
1163
+ beam:
1164
+ beam_size: 2
1165
+ return_best_hypothesis: false
1166
+ score_norm: true
1167
+ tsd_max_sym_exp: 50
1168
+ alsd_max_target_len: 2.0
1169
+ aux_ctc:
1170
+ ctc_loss_weight: 0.3
1171
+ use_cer: false
1172
+ ctc_reduction: mean_batch
1173
+ decoder:
1174
+ _target_: nemo.collections.asr.modules.ConvASRDecoder
1175
+ feat_in: null
1176
+ num_classes: -1
1177
+ vocabulary: []
1178
+ decoding:
1179
+ strategy: greedy
1180
+ interctc:
1181
+ loss_weights: []
1182
+ apply_at_layers: []
1183
+ loss:
1184
+ loss_name: default
1185
+ warprnnt_numba_kwargs:
1186
+ fastemit_lambda: 0.005
1187
+ clamp: -1.0
1188
+ optim:
1189
+ name: adamw
1190
+ lr: 3.0
1191
+ betas:
1192
+ - 0.9
1193
+ - 0.98
1194
+ weight_decay: 0.001
1195
+ sched:
1196
+ name: NoamAnnealing
1197
+ d_model: 1024
1198
+ warmup_steps: 8000
1199
+ warmup_ratio: null
1200
+ min_lr: 1.0e-06
1201
+ labels:
1202
+ - <unk>
1203
+ - ▁t
1204
+ - ▁th
1205
+ - ▁a
1206
+ - in
1207
+ - ▁the
1208
+ - re
1209
+ - ▁w
1210
+ - ▁o
1211
+ - ▁s
1212
+ - er
1213
+ - at
1214
+ - ou
1215
+ - nd
1216
+ - it
1217
+ - is
1218
+ - ▁h
1219
+ - ▁b
1220
+ - 'on'
1221
+ - ▁c
1222
+ - ing
1223
+ - en
1224
+ - ▁to
1225
+ - ▁m
1226
+ - ▁f
1227
+ - ▁p
1228
+ - or
1229
+ - an
1230
+ - es
1231
+ - ▁of
1232
+ - ▁d
1233
+ - ed
1234
+ - ll
1235
+ - ▁and
1236
+ - ▁I
1237
+ - ▁in
1238
+ - ▁l
1239
+ - ar
1240
+ - ▁y
1241
+ - ▁g
1242
+ - as
1243
+ - ▁you
1244
+ - om
1245
+ - ▁n
1246
+ - ic
1247
+ - ve
1248
+ - al
1249
+ - ion
1250
+ - us
1251
+ - ▁be
1252
+ - ow
1253
+ - le
1254
+ - ▁wh
1255
+ - ▁e
1256
+ - ot
1257
+ - ut
1258
+ - ▁it
1259
+ - ▁is
1260
+ - ▁we
1261
+ - ▁T
1262
+ - ▁re
1263
+ - et
1264
+ - ▁A
1265
+ - ent
1266
+ - ▁on
1267
+ - ▁ha
1268
+ - ay
1269
+ - ▁S
1270
+ - ct
1271
+ - ▁Th
1272
+ - ver
1273
+ - id
1274
+ - ig
1275
+ - im
1276
+ - ro
1277
+ - ▁for
1278
+ - ly
1279
+ - ▁he
1280
+ - ke
1281
+ - ld
1282
+ - se
1283
+ - st
1284
+ - ch
1285
+ - ▁st
1286
+ - all
1287
+ - ce
1288
+ - ur
1289
+ - ith
1290
+ - am
1291
+ - if
1292
+ - ir
1293
+ - ▁go
1294
+ - ▁u
1295
+ - ▁as
1296
+ - ▁was
1297
+ - ad
1298
+ - ▁W
1299
+ - ▁k
1300
+ - ▁an
1301
+ - ht
1302
+ - th
1303
+ - ▁r
1304
+ - ▁are
1305
+ - ere
1306
+ - ▁se
1307
+ - ▁do
1308
+ - ▁B
1309
+ - ▁so
1310
+ - ▁sh
1311
+ - ▁not
1312
+ - ▁li
1313
+ - od
1314
+ - ▁C
1315
+ - ust
1316
+ - ill
1317
+ - ight
1318
+ - ally
1319
+ - ▁And
1320
+ - ter
1321
+ - ▁or
1322
+ - ▁me
1323
+ - ▁M
1324
+ - ome
1325
+ - op
1326
+ - ▁at
1327
+ - il
1328
+ - ▁The
1329
+ - ould
1330
+ - ▁j
1331
+ - ant
1332
+ - ▁So
1333
+ - ▁H
1334
+ - ol
1335
+ - ain
1336
+ - ▁can
1337
+ - ▁de
1338
+ - ▁ne
1339
+ - ore
1340
+ - ▁con
1341
+ - ▁kn
1342
+ - ck
1343
+ - ul
1344
+ - ▁fr
1345
+ - ▁ab
1346
+ - ers
1347
+ - ess
1348
+ - ge
1349
+ - ▁pro
1350
+ - pe
1351
+ - ate
1352
+ - ▁su
1353
+ - ▁com
1354
+ - ▁but
1355
+ - ▁all
1356
+ - est
1357
+ - qu
1358
+ - ▁ex
1359
+ - ▁al
1360
+ - ra
1361
+ - ▁O
1362
+ - out
1363
+ - use
1364
+ - very
1365
+ - pp
1366
+ - ���Y
1367
+ - ▁ch
1368
+ - ri
1369
+ - ist
1370
+ - ▁v
1371
+ - ▁lo
1372
+ - ment
1373
+ - art
1374
+ - ▁P
1375
+ - nt
1376
+ - ab
1377
+ - ▁one
1378
+ - ▁N
1379
+ - ive
1380
+ - ▁wor
1381
+ - ions
1382
+ - ort
1383
+ - ▁L
1384
+ - ▁by
1385
+ - ich
1386
+ - ▁my
1387
+ - ity
1388
+ - ok
1389
+ - ▁G
1390
+ - res
1391
+ - ▁up
1392
+ - un
1393
+ - um
1394
+ - ea
1395
+ - ind
1396
+ - and
1397
+ - ink
1398
+ - el
1399
+ - ▁D
1400
+ - em
1401
+ - ▁E
1402
+ - os
1403
+ - oug
1404
+ - ▁if
1405
+ - ca
1406
+ - ▁out
1407
+ - ▁int
1408
+ - ie
1409
+ - ▁F
1410
+ - ▁It
1411
+ - ▁his
1412
+ - ard
1413
+ - ▁had
1414
+ - ▁tr
1415
+ - her
1416
+ - our
1417
+ - ies
1418
+ - ake
1419
+ - ▁R
1420
+ - ▁We
1421
+ - ▁get
1422
+ - ▁don
1423
+ - ▁us
1424
+ - ak
1425
+ - ▁pl
1426
+ - ect
1427
+ - ure
1428
+ - ame
1429
+ - ast
1430
+ - ▁who
1431
+ - ack
1432
+ - ▁le
1433
+ - ▁sa
1434
+ - iv
1435
+ - ci
1436
+ - ide
1437
+ - ▁tim
1438
+ - ▁our
1439
+ - ound
1440
+ - ous
1441
+ - ▁co
1442
+ - ▁pe
1443
+ - ose
1444
+ - ud
1445
+ - ▁see
1446
+ - ough
1447
+ - ▁man
1448
+ - ▁qu
1449
+ - ▁You
1450
+ - so
1451
+ - ople
1452
+ - ▁Wh
1453
+ - ong
1454
+ - ap
1455
+ - ther
1456
+ - ▁J
1457
+ - are
1458
+ - ine
1459
+ - ▁say
1460
+ - ▁im
1461
+ - ▁But
1462
+ - ings
1463
+ - ▁has
1464
+ - ▁ag
1465
+ - ff
1466
+ - ▁her
1467
+ - itt
1468
+ - one
1469
+ - ▁en
1470
+ - ▁ar
1471
+ - ▁fe
1472
+ - ven
1473
+ - ▁any
1474
+ - ▁mo
1475
+ - reat
1476
+ - ag
1477
+ - ▁how
1478
+ - ▁cl
1479
+ - pt
1480
+ - ▁now
1481
+ - own
1482
+ - ber
1483
+ - ▁him
1484
+ - ▁act
1485
+ - hing
1486
+ - ice
1487
+ - ▁no
1488
+ - ans
1489
+ - iz
1490
+ - ▁fa
1491
+ - per
1492
+ - pl
1493
+ - ▁te
1494
+ - ▁ad
1495
+ - age
1496
+ - ree
1497
+ - ▁tw
1498
+ - ank
1499
+ - ▁He
1500
+ - ple
1501
+ - ite
1502
+ - ry
1503
+ - ▁U
1504
+ - ish
1505
+ - ire
1506
+ - ue
1507
+ - ▁In
1508
+ - ▁she
1509
+ - ble
1510
+ - cc
1511
+ - nder
1512
+ - ▁way
1513
+ - ▁pr
1514
+ - ear
1515
+ - ▁did
1516
+ - ▁po
1517
+ - eah
1518
+ - ▁un
1519
+ - omet
1520
+ - ence
1521
+ - ep
1522
+ - uch
1523
+ - ▁sp
1524
+ - ach
1525
+ - og
1526
+ - ance
1527
+ - able
1528
+ - iff
1529
+ - sel
1530
+ - ▁got
1531
+ - way
1532
+ - ▁gr
1533
+ - alk
1534
+ - ▁res
1535
+ - ated
1536
+ - irst
1537
+ - ick
1538
+ - ass
1539
+ - ▁two
1540
+ - ▁dis
1541
+ - ord
1542
+ - ▁pre
1543
+ - ount
1544
+ - ase
1545
+ - ip
1546
+ - ult
1547
+ - ical
1548
+ - orm
1549
+ - ary
1550
+ - ace
1551
+ - ▁spe
1552
+ - ▁Ch
1553
+ - ▁thr
1554
+ - ▁imp
1555
+ - int
1556
+ - ▁am
1557
+ - ▁off
1558
+ - act
1559
+ - ia
1560
+ - ▁ro
1561
+ - ress
1562
+ - ▁per
1563
+ - ▁fo
1564
+ - ▁br
1565
+ - ▁K
1566
+ - vel
1567
+ - ▁gu
1568
+ - ▁bo
1569
+ - ang
1570
+ - kay
1571
+ - ub
1572
+ - ign
1573
+ - ▁may
1574
+ - ving
1575
+ - ces
1576
+ - ens
1577
+ - cl
1578
+ - ▁lot
1579
+ - ru
1580
+ - ade
1581
+ - ▁bet
1582
+ - ▁bl
1583
+ - ▁let
1584
+ - fore
1585
+ - co
1586
+ - ild
1587
+ - ning
1588
+ - xt
1589
+ - ile
1590
+ - ark
1591
+ - self
1592
+ - ▁app
1593
+ - ory
1594
+ - du
1595
+ - ▁day
1596
+ - ▁St
1597
+ - ater
1598
+ - ▁use
1599
+ - ys
1600
+ - fter
1601
+ - ▁new
1602
+ - ious
1603
+ - ial
1604
+ - he
1605
+ - wn
1606
+ - ved
1607
+ - red
1608
+ - ▁fl
1609
+ - iss
1610
+ - ody
1611
+ - form
1612
+ - ian
1613
+ - tain
1614
+ - ▁bu
1615
+ - ▁V
1616
+ - ▁rec
1617
+ - ty
1618
+ - be
1619
+ - ▁sc
1620
+ - ors
1621
+ - vers
1622
+ - ▁put
1623
+ - ife
1624
+ - ▁If
1625
+ - we
1626
+ - te
1627
+ - ject
1628
+ - ath
1629
+ - ting
1630
+ - ▁rem
1631
+ - ▁acc
1632
+ - ull
1633
+ - ons
1634
+ - ▁ind
1635
+ - ▁ser
1636
+ - ▁ke
1637
+ - ates
1638
+ - ves
1639
+ - na
1640
+ - lic
1641
+ - ▁des
1642
+ - ▁its
1643
+ - ful
1644
+ - ents
1645
+ - erm
1646
+ - ac
1647
+ - ered
1648
+ - ise
1649
+ - ▁sy
1650
+ - urn
1651
+ - ▁em
1652
+ - oth
1653
+ - ual
1654
+ - ne
1655
+ - ward
1656
+ - ib
1657
+ - ▁try
1658
+ - ▁pos
1659
+ - nds
1660
+ - ft
1661
+ - get
1662
+ - ph
1663
+ - ▁ob
1664
+ - ady
1665
+ - igh
1666
+ - ood
1667
+ - ▁rel
1668
+ - ▁wr
1669
+ - ug
1670
+ - ears
1671
+ - ail
1672
+ - ▁Now
1673
+ - ▁bit
1674
+ - ng
1675
+ - ▁Oh
1676
+ - ▁hel
1677
+ - ange
1678
+ - ▁reg
1679
+ - ▁rep
1680
+ - ▁bel
1681
+ - ▁sm
1682
+ - ost
1683
+ - tern
1684
+ - gr
1685
+ - ▁own
1686
+ - ▁end
1687
+ - pect
1688
+ - ily
1689
+ - day
1690
+ - ied
1691
+ - ific
1692
+ - ower
1693
+ - ▁add
1694
+ - cess
1695
+ - ict
1696
+ - ible
1697
+ - ▁bas
1698
+ - ▁i
1699
+ - ▁op
1700
+ - cial
1701
+ - ular
1702
+ - ▁Be
1703
+ - ced
1704
+ - ▁too
1705
+ - ks
1706
+ - ew
1707
+ - mer
1708
+ - ▁ph
1709
+ - ob
1710
+ - ==
1711
+ - ▁la
1712
+ - ▁set
1713
+ - ▁min
1714
+ - ▁sub
1715
+ - ▁gen
1716
+ - atch
1717
+ - ..
1718
+ - ▁inv
1719
+ - ▁As
1720
+ - ▁nat
1721
+ - ▁sl
1722
+ - ▁num
1723
+ - av
1724
+ - ways
1725
+ - ▁God
1726
+ - stem
1727
+ - ▁ac
1728
+ - ▁att
1729
+ - ▁ev
1730
+ - ▁def
1731
+ - llow
1732
+ - ▁str
1733
+ - lect
1734
+ - ars
1735
+ - ▁cr
1736
+ - ▁Is
1737
+ - olog
1738
+ - les
1739
+ - oy
1740
+ - ▁ask
1741
+ - ▁inc
1742
+ - body
1743
+ - ▁ent
1744
+ - ▁pol
1745
+ - ness
1746
+ - ix
1747
+ - ▁why
1748
+ - onna
1749
+ - ▁ear
1750
+ - ▁tak
1751
+ - ▁Un
1752
+ - ited
1753
+ - mun
1754
+ - li
1755
+ - ute
1756
+ - ract
1757
+ - ▁dec
1758
+ - uro
1759
+ - ▁mak
1760
+ - ▁fin
1761
+ - ween
1762
+ - ▁No
1763
+ - arch
1764
+ - ▁bec
1765
+ - gan
1766
+ - old
1767
+ - cy
1768
+ - ▁big
1769
+ - ▁For
1770
+ - ren
1771
+ - als
1772
+ - und
1773
+ - ▁Al
1774
+ - ▁All
1775
+ - ss
1776
+ - ows
1777
+ - ▁mod
1778
+ - ock
1779
+ - ▁id
1780
+ - ism
1781
+ - cus
1782
+ - ▁gl
1783
+ - ably
1784
+ - ▁ass
1785
+ - ▁car
1786
+ - ata
1787
+ - ppen
1788
+ - led
1789
+ - ▁sim
1790
+ - ▁mon
1791
+ - ics
1792
+ - ▁giv
1793
+ - cept
1794
+ - ▁Mr
1795
+ - pan
1796
+ - ▁pub
1797
+ - ▁eff
1798
+ - ▁How
1799
+ - ps
1800
+ - vern
1801
+ - end
1802
+ - hip
1803
+ - iew
1804
+ - ope
1805
+ - ▁An
1806
+ - ▁She
1807
+ - ▁Com
1808
+ - ee
1809
+ - ures
1810
+ - ell
1811
+ - ouse
1812
+ - cond
1813
+ - king
1814
+ - oc
1815
+ - ues
1816
+ - ever
1817
+ - ▁To
1818
+ - clud
1819
+ - ▁ins
1820
+ - ▁exp
1821
+ - ▁old
1822
+ - ▁mem
1823
+ - ▁ref
1824
+ - ▁tra
1825
+ - ▁far
1826
+ - ave
1827
+ - rat
1828
+ - ▁sur
1829
+ - ruct
1830
+ - rib
1831
+ - duct
1832
+ - uff
1833
+ - ▁met
1834
+ - ▁sch
1835
+ - ince
1836
+ - ▁run
1837
+ - ense
1838
+ - ▁cle
1839
+ - ▁==
1840
+ - mon
1841
+ - ize
1842
+ - ▁ord
1843
+ - blem
1844
+ - tin
1845
+ - ▁Let
1846
+ - ner
1847
+ - ond
1848
+ - its
1849
+ - ▁cor
1850
+ - land
1851
+ - ▁cur
1852
+ - ▁Re
1853
+ - ▁bus
1854
+ - ▁uh
1855
+ - air
1856
+ - ote
1857
+ - ants
1858
+ - ason
1859
+ - ric
1860
+ - ▁el
1861
+ - ▁cer
1862
+ - nce
1863
+ - ▁fam
1864
+ - ▁cap
1865
+ - uck
1866
+ - ool
1867
+ - ried
1868
+ - ▁cou
1869
+ - ▁fun
1870
+ - ▁wom
1871
+ - ▁hum
1872
+ - ▁ty
1873
+ - ▁ap
1874
+ - ike
1875
+ - ▁few
1876
+ - oney
1877
+ - ▁inf
1878
+ - ont
1879
+ - ese
1880
+ - ook
1881
+ - gy
1882
+ - uth
1883
+ - ulat
1884
+ - ieve
1885
+ - ized
1886
+ - ross
1887
+ - ▁ple
1888
+ - ▁um
1889
+ - ▁val
1890
+ - ▁equ
1891
+ - ▁lea
1892
+ - ▁lar
1893
+ - ah
1894
+ - eral
1895
+ - ▁ed
1896
+ - ared
1897
+ - lish
1898
+ - arn
1899
+ - ds
1900
+ - esn
1901
+ - ▁iss
1902
+ - ▁ca
1903
+ - ted
1904
+ - ices
1905
+ - ▁wee
1906
+ - ash
1907
+ - ▁top
1908
+ - ten
1909
+ - up
1910
+ - ts
1911
+ - gin
1912
+ - con
1913
+ - ari
1914
+ - ▁opp
1915
+ - osed
1916
+ - ▁eas
1917
+ - ▁ext
1918
+ - gg
1919
+ - az
1920
+ - ▁Fr
1921
+ - ideo
1922
+ - izat
1923
+ - ▁men
1924
+ - ▁mom
1925
+ - ▁ret
1926
+ - tty
1927
+ - rist
1928
+ - ▁gra
1929
+ - alth
1930
+ - ef
1931
+ - ▁det
1932
+ - ax
1933
+ - ▁mat
1934
+ - chn
1935
+ - ern
1936
+ - peri
1937
+ - ▁bre
1938
+ - ▁Sh
1939
+ - sw
1940
+ - erat
1941
+ - ▁sit
1942
+ - ters
1943
+ - ale
1944
+ - man
1945
+ - ▁sol
1946
+ - ork
1947
+ - ▁adv
1948
+ - ety
1949
+ - ▁vis
1950
+ - ▁med
1951
+ - uc
1952
+ - less
1953
+ - ▁unt
1954
+ - gram
1955
+ - ets
1956
+ - ists
1957
+ - ▁ey
1958
+ - ▁col
1959
+ - imes
1960
+ - ▁law
1961
+ - ▁pri
1962
+ - sid
1963
+ - ▁On
1964
+ - ▁mot
1965
+ - ield
1966
+ - ▁Do
1967
+ - ▁At
1968
+ - ages
1969
+ - amp
1970
+ - ▁art
1971
+ - miss
1972
+ - ▁sk
1973
+ - alf
1974
+ - pr
1975
+ - ier
1976
+ - ▁beh
1977
+ - ▁Yes
1978
+ - ural
1979
+ - ime
1980
+ - ▁wa
1981
+ - oks
1982
+ - bers
1983
+ - ger
1984
+ - ient
1985
+ - ries
1986
+ - '...'
1987
+ - ▁che
1988
+ - ▁Br
1989
+ - ird
1990
+ - ▁Ar
1991
+ - ▁war
1992
+ - inat
1993
+ - ▁My
1994
+ - ital
1995
+ - wh
1996
+ - med
1997
+ - ▁pur
1998
+ - ully
1999
+ - ▁One
2000
+ - ▁rat
2001
+ - ines
2002
+ - ▁Of
2003
+ - io
2004
+ - ▁loc
2005
+ - ret
2006
+ - ctor
2007
+ - ▁leg
2008
+ - stit
2009
+ - ined
2010
+ - ught
2011
+ - ▁dur
2012
+ - ▁es
2013
+ - vent
2014
+ - aj
2015
+ - ▁bro
2016
+ - ▁saw
2017
+ - ▁sec
2018
+ - ream
2019
+ - ▁pop
2020
+ - reen
2021
+ - ▁Ind
2022
+ - els
2023
+ - ▁yet
2024
+ - ired
2025
+ - ▁sw
2026
+ - tro
2027
+ - oup
2028
+ - most
2029
+ - pean
2030
+ - eds
2031
+ - ush
2032
+ - oh
2033
+ - ▁Se
2034
+ - ▁tea
2035
+ - ann
2036
+ - ilit
2037
+ - err
2038
+ - pend
2039
+ - ton
2040
+ - ased
2041
+ - ▁aff
2042
+ - ▁mor
2043
+ - ▁dra
2044
+ - put
2045
+ - ▁dr
2046
+ - ins
2047
+ - uat
2048
+ - nect
2049
+ - cri
2050
+ - outh
2051
+ - ▁ra
2052
+ - ▁pay
2053
+ - ms
2054
+ - ▁av
2055
+ - bs
2056
+ - ling
2057
+ - ▁De
2058
+ - ▁Or
2059
+ - ove
2060
+ - ▁Can
2061
+ - ▁eng
2062
+ - ames
2063
+ - ided
2064
+ - ▁Go
2065
+ - mitt
2066
+ - ode
2067
+ - ▁cre
2068
+ - par
2069
+ - ides
2070
+ - pos
2071
+ - ▁fav
2072
+ - ▁air
2073
+ - ▁New
2074
+ - ▁bad
2075
+ - ▁six
2076
+ - vat
2077
+ - ▁pat
2078
+ - not
2079
+ - ▁di
2080
+ - rop
2081
+ - ral
2082
+ - orn
2083
+ - ▁par
2084
+ - cing
2085
+ - ▁aw
2086
+ - orts
2087
+ - ox
2088
+ - ▁yes
2089
+ - cuss
2090
+ - eng
2091
+ - ives
2092
+ - erms
2093
+ - ▁job
2094
+ - mand
2095
+ - ying
2096
+ - ▁occ
2097
+ - aps
2098
+ - ases
2099
+ - ▁Not
2100
+ - rent
2101
+ - ency
2102
+ - att
2103
+ - ised
2104
+ - vice
2105
+ - ▁Eng
2106
+ - ▁est
2107
+ - oked
2108
+ - ▁Q
2109
+ - iron
2110
+ - idd
2111
+ - me
2112
+ - unch
2113
+ - ane
2114
+ - ▁z
2115
+ - br
2116
+ - arts
2117
+ - ▁fat
2118
+ - ery
2119
+ - anks
2120
+ - ▁jo
2121
+ - ▁mar
2122
+ - aw
2123
+ - ott
2124
+ - ards
2125
+ - ▁oh
2126
+ - ians
2127
+ - ▁sci
2128
+ - row
2129
+ - unt
2130
+ - ury
2131
+ - ▁abs
2132
+ - ergy
2133
+ - ▁Z
2134
+ - ump
2135
+ - ▁Am
2136
+ - ened
2137
+ - angu
2138
+ - ▁Pro
2139
+ - icat
2140
+ - itch
2141
+ - ▁dri
2142
+ - iat
2143
+ - ▁
2144
+ - e
2145
+ - t
2146
+ - o
2147
+ - a
2148
+ - 'n'
2149
+ - i
2150
+ - s
2151
+ - r
2152
+ - h
2153
+ - l
2154
+ - d
2155
+ - u
2156
+ - c
2157
+ - m
2158
+ - 'y'
2159
+ - g
2160
+ - w
2161
+ - f
2162
+ - p
2163
+ - ','
2164
+ - .
2165
+ - b
2166
+ - v
2167
+ - k
2168
+ - ''''
2169
+ - I
2170
+ - T
2171
+ - A
2172
+ - S
2173
+ - x
2174
+ - W
2175
+ - j
2176
+ - C
2177
+ - B
2178
+ - M
2179
+ - '?'
2180
+ - H
2181
+ - O
2182
+ - '0'
2183
+ - P
2184
+ - q
2185
+ - 'Y'
2186
+ - 'N'
2187
+ - L
2188
+ - D
2189
+ - '1'
2190
+ - E
2191
+ - G
2192
+ - z
2193
+ - F
2194
+ - R
2195
+ - '-'
2196
+ - '2'
2197
+ - J
2198
+ - U
2199
+ - '9'
2200
+ - K
2201
+ - '5'
2202
+ - '3'
2203
+ - V
2204
+ - '='
2205
+ - '4'
2206
+ - '8'
2207
+ - '6'
2208
+ - '7'
2209
+ - '!'
2210
+ - '%'
2211
+ - ':'
2212
+ - Q
2213
+ - Z
2214
+ - $
2215
+ - X
2216
+ - '"'
2217
+ - '&'
2218
+ - '*'
2219
+ - /
2220
+ - £
2221
+ - +
2222
+ - €
2223
+ - _
2224
+ - ^
2225
+ - ¥
2226
+ target: nemo.collections.asr.models.rnnt_bpe_models.EncDecRNNTBPEModel
2227
+ nemo_version: 2.4.0rc0
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07d4e5a63840a53ab2d4d106d2874768143fb3fbdd47938b3910d2da05bfb0a9
3
+ size 251056
tokenizer.vocab ADDED
@@ -0,0 +1,1024 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ <unk> 0
2
+ ▁t -0
3
+ ▁th -1
4
+ ▁a -2
5
+ in -3
6
+ ▁the -4
7
+ re -5
8
+ ▁w -6
9
+ ▁o -7
10
+ ▁s -8
11
+ er -9
12
+ at -10
13
+ ou -11
14
+ nd -12
15
+ it -13
16
+ is -14
17
+ ▁h -15
18
+ ▁b -16
19
+ on -17
20
+ ▁c -18
21
+ ing -19
22
+ en -20
23
+ ▁to -21
24
+ ▁m -22
25
+ ▁f -23
26
+ ▁p -24
27
+ or -25
28
+ an -26
29
+ es -27
30
+ ▁of -28
31
+ ▁d -29
32
+ ed -30
33
+ ll -31
34
+ ▁and -32
35
+ ▁I -33
36
+ ▁in -34
37
+ ▁l -35
38
+ ar -36
39
+ ▁y -37
40
+ ▁g -38
41
+ as -39
42
+ ▁you -40
43
+ om -41
44
+ ▁n -42
45
+ ic -43
46
+ ve -44
47
+ al -45
48
+ ion -46
49
+ us -47
50
+ ▁be -48
51
+ ow -49
52
+ le -50
53
+ ▁wh -51
54
+ ▁e -52
55
+ ot -53
56
+ ut -54
57
+ ▁it -55
58
+ ▁is -56
59
+ ▁we -57
60
+ ▁T -58
61
+ ▁re -59
62
+ et -60
63
+ ▁A -61
64
+ ent -62
65
+ ▁on -63
66
+ ▁ha -64
67
+ ay -65
68
+ ▁S -66
69
+ ct -67
70
+ ▁Th -68
71
+ ver -69
72
+ id -70
73
+ ig -71
74
+ im -72
75
+ ro -73
76
+ ▁for -74
77
+ ly -75
78
+ ▁he -76
79
+ ke -77
80
+ ld -78
81
+ se -79
82
+ st -80
83
+ ch -81
84
+ ▁st -82
85
+ all -83
86
+ ce -84
87
+ ur -85
88
+ ith -86
89
+ am -87
90
+ if -88
91
+ ir -89
92
+ ▁go -90
93
+ ▁u -91
94
+ ▁as -92
95
+ ▁was -93
96
+ ad -94
97
+ ▁W -95
98
+ ▁k -96
99
+ ▁an -97
100
+ ht -98
101
+ th -99
102
+ ▁r -100
103
+ ▁are -101
104
+ ere -102
105
+ ▁se -103
106
+ ▁do -104
107
+ ▁B -105
108
+ ▁so -106
109
+ ▁sh -107
110
+ ▁not -108
111
+ ▁li -109
112
+ od -110
113
+ ▁C -111
114
+ ust -112
115
+ ill -113
116
+ ight -114
117
+ ally -115
118
+ ▁And -116
119
+ ter -117
120
+ ▁or -118
121
+ ▁me -119
122
+ ▁M -120
123
+ ome -121
124
+ op -122
125
+ ▁at -123
126
+ il -124
127
+ ▁The -125
128
+ ould -126
129
+ ▁j -127
130
+ ant -128
131
+ ▁So -129
132
+ ▁H -130
133
+ ol -131
134
+ ain -132
135
+ ▁can -133
136
+ ▁de -134
137
+ ▁ne -135
138
+ ore -136
139
+ ▁con -137
140
+ ▁kn -138
141
+ ck -139
142
+ ul -140
143
+ ▁fr -141
144
+ ▁ab -142
145
+ ers -143
146
+ ess -144
147
+ ge -145
148
+ ▁pro -146
149
+ pe -147
150
+ ate -148
151
+ ▁su -149
152
+ ▁com -150
153
+ ▁but -151
154
+ ▁all -152
155
+ est -153
156
+ qu -154
157
+ ▁ex -155
158
+ ▁al -156
159
+ ra -157
160
+ ▁O -158
161
+ out -159
162
+ use -160
163
+ very -161
164
+ pp -162
165
+ ▁Y -163
166
+ ▁ch -164
167
+ ri -165
168
+ ist -166
169
+ ▁v -167
170
+ ▁lo -168
171
+ ment -169
172
+ art -170
173
+ ▁P -171
174
+ nt -172
175
+ ab -173
176
+ ▁one -174
177
+ ▁N -175
178
+ ive -176
179
+ ▁wor -177
180
+ ions -178
181
+ ort -179
182
+ ▁L -180
183
+ ▁by -181
184
+ ich -182
185
+ ▁my -183
186
+ ity -184
187
+ ok -185
188
+ ▁G -186
189
+ res -187
190
+ ▁up -188
191
+ un -189
192
+ um -190
193
+ ea -191
194
+ ind -192
195
+ and -193
196
+ ink -194
197
+ el -195
198
+ ▁D -196
199
+ em -197
200
+ ▁E -198
201
+ os -199
202
+ oug -200
203
+ ▁if -201
204
+ ca -202
205
+ ▁out -203
206
+ ▁int -204
207
+ ie -205
208
+ ▁F -206
209
+ ▁It -207
210
+ ▁his -208
211
+ ard -209
212
+ ▁had -210
213
+ ▁tr -211
214
+ her -212
215
+ our -213
216
+ ies -214
217
+ ake -215
218
+ ▁R -216
219
+ ▁We -217
220
+ ▁get -218
221
+ ▁don -219
222
+ ▁us -220
223
+ ak -221
224
+ ▁pl -222
225
+ ect -223
226
+ ure -224
227
+ ame -225
228
+ ast -226
229
+ ▁who -227
230
+ ack -228
231
+ ▁le -229
232
+ ▁sa -230
233
+ iv -231
234
+ ci -232
235
+ ide -233
236
+ ▁tim -234
237
+ ▁our -235
238
+ ound -236
239
+ ous -237
240
+ ▁co -238
241
+ ▁pe -239
242
+ ose -240
243
+ ud -241
244
+ ▁see -242
245
+ ough -243
246
+ ▁man -244
247
+ ▁qu -245
248
+ ▁You -246
249
+ so -247
250
+ ople -248
251
+ ▁Wh -249
252
+ ong -250
253
+ ap -251
254
+ ther -252
255
+ ▁J -253
256
+ are -254
257
+ ine -255
258
+ ▁say -256
259
+ ▁im -257
260
+ ▁But -258
261
+ ings -259
262
+ ▁has -260
263
+ ▁ag -261
264
+ ff -262
265
+ ▁her -263
266
+ itt -264
267
+ one -265
268
+ ▁en -266
269
+ ▁ar -267
270
+ ▁fe -268
271
+ ven -269
272
+ ▁any -270
273
+ ▁mo -271
274
+ reat -272
275
+ ag -273
276
+ ▁how -274
277
+ ▁cl -275
278
+ pt -276
279
+ ▁now -277
280
+ own -278
281
+ ber -279
282
+ ▁him -280
283
+ ▁act -281
284
+ hing -282
285
+ ice -283
286
+ ▁no -284
287
+ ans -285
288
+ iz -286
289
+ ▁fa -287
290
+ per -288
291
+ pl -289
292
+ ▁te -290
293
+ ▁ad -291
294
+ age -292
295
+ ree -293
296
+ ▁tw -294
297
+ ank -295
298
+ ▁He -296
299
+ ple -297
300
+ ite -298
301
+ ry -299
302
+ ▁U -300
303
+ ish -301
304
+ ire -302
305
+ ue -303
306
+ ▁In -304
307
+ ▁she -305
308
+ ble -306
309
+ cc -307
310
+ nder -308
311
+ ▁way -309
312
+ ▁pr -310
313
+ ear -311
314
+ ▁did -312
315
+ ▁po -313
316
+ eah -314
317
+ ▁un -315
318
+ omet -316
319
+ ence -317
320
+ ep -318
321
+ uch -319
322
+ ▁sp -320
323
+ ach -321
324
+ og -322
325
+ ance -323
326
+ able -324
327
+ iff -325
328
+ sel -326
329
+ ▁got -327
330
+ way -328
331
+ ▁gr -329
332
+ alk -330
333
+ ▁res -331
334
+ ated -332
335
+ irst -333
336
+ ick -334
337
+ ass -335
338
+ ▁two -336
339
+ ▁dis -337
340
+ ord -338
341
+ ▁pre -339
342
+ ount -340
343
+ ase -341
344
+ ip -342
345
+ ult -343
346
+ ical -344
347
+ orm -345
348
+ ary -346
349
+ ace -347
350
+ ▁spe -348
351
+ ▁Ch -349
352
+ ▁thr -350
353
+ ▁imp -351
354
+ int -352
355
+ ▁am -353
356
+ ▁off -354
357
+ act -355
358
+ ia -356
359
+ ▁ro -357
360
+ ress -358
361
+ ▁per -359
362
+ ▁fo -360
363
+ ▁br -361
364
+ ▁K -362
365
+ vel -363
366
+ ▁gu -364
367
+ ▁bo -365
368
+ ang -366
369
+ kay -367
370
+ ub -368
371
+ ign -369
372
+ ▁may -370
373
+ ving -371
374
+ ces -372
375
+ ens -373
376
+ cl -374
377
+ ▁lot -375
378
+ ru -376
379
+ ade -377
380
+ ▁bet -378
381
+ ▁bl -379
382
+ ▁let -380
383
+ fore -381
384
+ co -382
385
+ ild -383
386
+ ning -384
387
+ xt -385
388
+ ile -386
389
+ ark -387
390
+ self -388
391
+ ▁app -389
392
+ ory -390
393
+ du -391
394
+ ▁day -392
395
+ ▁St -393
396
+ ater -394
397
+ ▁use -395
398
+ ys -396
399
+ fter -397
400
+ ▁new -398
401
+ ious -399
402
+ ial -400
403
+ he -401
404
+ wn -402
405
+ ved -403
406
+ red -404
407
+ ▁fl -405
408
+ iss -406
409
+ ody -407
410
+ form -408
411
+ ian -409
412
+ tain -410
413
+ ▁bu -411
414
+ ▁V -412
415
+ ▁rec -413
416
+ ty -414
417
+ be -415
418
+ ▁sc -416
419
+ ors -417
420
+ vers -418
421
+ ▁put -419
422
+ ife -420
423
+ ▁If -421
424
+ we -422
425
+ te -423
426
+ ject -424
427
+ ath -425
428
+ ting -426
429
+ ▁rem -427
430
+ ▁acc -428
431
+ ull -429
432
+ ons -430
433
+ ▁ind -431
434
+ ▁ser -432
435
+ ▁ke -433
436
+ ates -434
437
+ ves -435
438
+ na -436
439
+ lic -437
440
+ ▁des -438
441
+ ▁its -439
442
+ ful -440
443
+ ents -441
444
+ erm -442
445
+ ac -443
446
+ ered -444
447
+ ise -445
448
+ ▁sy -446
449
+ urn -447
450
+ ▁em -448
451
+ oth -449
452
+ ual -450
453
+ ne -451
454
+ ward -452
455
+ ib -453
456
+ ▁try -454
457
+ ▁pos -455
458
+ nds -456
459
+ ft -457
460
+ get -458
461
+ ph -459
462
+ ▁ob -460
463
+ ady -461
464
+ igh -462
465
+ ood -463
466
+ ▁rel -464
467
+ ▁wr -465
468
+ ug -466
469
+ ears -467
470
+ ail -468
471
+ ▁Now -469
472
+ ▁bit -470
473
+ ng -471
474
+ ▁Oh -472
475
+ ▁hel -473
476
+ ange -474
477
+ ▁reg -475
478
+ ▁rep -476
479
+ ▁bel -477
480
+ ▁sm -478
481
+ ost -479
482
+ tern -480
483
+ gr -481
484
+ ▁own -482
485
+ ▁end -483
486
+ pect -484
487
+ ily -485
488
+ day -486
489
+ ied -487
490
+ ific -488
491
+ ower -489
492
+ ▁add -490
493
+ cess -491
494
+ ict -492
495
+ ible -493
496
+ ▁bas -494
497
+ ▁i -495
498
+ ▁op -496
499
+ cial -497
500
+ ular -498
501
+ ▁Be -499
502
+ ced -500
503
+ ▁too -501
504
+ ks -502
505
+ ew -503
506
+ mer -504
507
+ ▁ph -505
508
+ ob -506
509
+ == -507
510
+ ▁la -508
511
+ ▁set -509
512
+ ▁min -510
513
+ ▁sub -511
514
+ ▁gen -512
515
+ atch -513
516
+ .. -514
517
+ ▁inv -515
518
+ ▁As -516
519
+ ▁nat -517
520
+ ▁sl -518
521
+ ▁num -519
522
+ av -520
523
+ ways -521
524
+ ▁God -522
525
+ stem -523
526
+ ▁ac -524
527
+ ▁att -525
528
+ ▁ev -526
529
+ ▁def -527
530
+ llow -528
531
+ ▁str -529
532
+ lect -530
533
+ ars -531
534
+ ▁cr -532
535
+ ▁Is -533
536
+ olog -534
537
+ les -535
538
+ oy -536
539
+ ▁ask -537
540
+ ▁inc -538
541
+ body -539
542
+ ▁ent -540
543
+ ▁pol -541
544
+ ness -542
545
+ ix -543
546
+ ▁why -544
547
+ onna -545
548
+ ▁ear -546
549
+ ▁tak -547
550
+ ▁Un -548
551
+ ited -549
552
+ mun -550
553
+ li -551
554
+ ute -552
555
+ ract -553
556
+ ▁dec -554
557
+ uro -555
558
+ ▁mak -556
559
+ ▁fin -557
560
+ ween -558
561
+ ▁No -559
562
+ arch -560
563
+ ▁bec -561
564
+ gan -562
565
+ old -563
566
+ cy -564
567
+ ▁big -565
568
+ ▁For -566
569
+ ren -567
570
+ als -568
571
+ und -569
572
+ ▁Al -570
573
+ ▁All -571
574
+ ss -572
575
+ ows -573
576
+ ▁mod -574
577
+ ock -575
578
+ ▁id -576
579
+ ism -577
580
+ cus -578
581
+ ▁gl -579
582
+ ably -580
583
+ ▁ass -581
584
+ ▁car -582
585
+ ata -583
586
+ ppen -584
587
+ led -585
588
+ ▁sim -586
589
+ ▁mon -587
590
+ ics -588
591
+ ▁giv -589
592
+ cept -590
593
+ ▁Mr -591
594
+ pan -592
595
+ ▁pub -593
596
+ ▁eff -594
597
+ ▁How -595
598
+ ps -596
599
+ vern -597
600
+ end -598
601
+ hip -599
602
+ iew -600
603
+ ope -601
604
+ ▁An -602
605
+ ▁She -603
606
+ ▁Com -604
607
+ ee -605
608
+ ures -606
609
+ ell -607
610
+ ouse -608
611
+ cond -609
612
+ king -610
613
+ oc -611
614
+ ues -612
615
+ ever -613
616
+ ▁To -614
617
+ clud -615
618
+ ▁ins -616
619
+ ▁exp -617
620
+ ▁old -618
621
+ ▁mem -619
622
+ ▁ref -620
623
+ ▁tra -621
624
+ ▁far -622
625
+ ave -623
626
+ rat -624
627
+ ▁sur -625
628
+ ruct -626
629
+ rib -627
630
+ duct -628
631
+ uff -629
632
+ ▁met -630
633
+ ▁sch -631
634
+ ince -632
635
+ ▁run -633
636
+ ense -634
637
+ ▁cle -635
638
+ ▁== -636
639
+ mon -637
640
+ ize -638
641
+ ▁ord -639
642
+ blem -640
643
+ tin -641
644
+ ▁Let -642
645
+ ner -643
646
+ ond -644
647
+ its -645
648
+ ▁cor -646
649
+ land -647
650
+ ▁cur -648
651
+ ▁Re -649
652
+ ▁bus -650
653
+ ▁uh -651
654
+ air -652
655
+ ote -653
656
+ ants -654
657
+ ason -655
658
+ ric -656
659
+ ▁el -657
660
+ ▁cer -658
661
+ nce -659
662
+ ▁fam -660
663
+ ▁cap -661
664
+ uck -662
665
+ ool -663
666
+ ried -664
667
+ ▁cou -665
668
+ ▁fun -666
669
+ ▁wom -667
670
+ ▁hum -668
671
+ ▁ty -669
672
+ ▁ap -670
673
+ ike -671
674
+ ▁few -672
675
+ oney -673
676
+ ▁inf -674
677
+ ont -675
678
+ ese -676
679
+ ook -677
680
+ gy -678
681
+ uth -679
682
+ ulat -680
683
+ ieve -681
684
+ ized -682
685
+ ross -683
686
+ ▁ple -684
687
+ ▁um -685
688
+ ▁val -686
689
+ ▁equ -687
690
+ ▁lea -688
691
+ ▁lar -689
692
+ ah -690
693
+ eral -691
694
+ ▁ed -692
695
+ ared -693
696
+ lish -694
697
+ arn -695
698
+ ds -696
699
+ esn -697
700
+ ▁iss -698
701
+ ▁ca -699
702
+ ted -700
703
+ ices -701
704
+ ▁wee -702
705
+ ash -703
706
+ ▁top -704
707
+ ten -705
708
+ up -706
709
+ ts -707
710
+ gin -708
711
+ con -709
712
+ ari -710
713
+ ▁opp -711
714
+ osed -712
715
+ ▁eas -713
716
+ ▁ext -714
717
+ gg -715
718
+ az -716
719
+ ▁Fr -717
720
+ ideo -718
721
+ izat -719
722
+ ▁men -720
723
+ ▁mom -721
724
+ ▁ret -722
725
+ tty -723
726
+ rist -724
727
+ ▁gra -725
728
+ alth -726
729
+ ef -727
730
+ ▁det -728
731
+ ax -729
732
+ ▁mat -730
733
+ chn -731
734
+ ern -732
735
+ peri -733
736
+ ▁bre -734
737
+ ▁Sh -735
738
+ sw -736
739
+ erat -737
740
+ ▁sit -738
741
+ ters -739
742
+ ale -740
743
+ man -741
744
+ ▁sol -742
745
+ ork -743
746
+ ▁adv -744
747
+ ety -745
748
+ ▁vis -746
749
+ ▁med -747
750
+ uc -748
751
+ less -749
752
+ ▁unt -750
753
+ gram -751
754
+ ets -752
755
+ ists -753
756
+ ▁ey -754
757
+ ▁col -755
758
+ imes -756
759
+ ▁law -757
760
+ ▁pri -758
761
+ sid -759
762
+ ▁On -760
763
+ ▁mot -761
764
+ ield -762
765
+ ▁Do -763
766
+ ▁At -764
767
+ ages -765
768
+ amp -766
769
+ ▁art -767
770
+ miss -768
771
+ ▁sk -769
772
+ alf -770
773
+ pr -771
774
+ ier -772
775
+ ▁beh -773
776
+ ▁Yes -774
777
+ ural -775
778
+ ime -776
779
+ ▁wa -777
780
+ oks -778
781
+ bers -779
782
+ ger -780
783
+ ient -781
784
+ ries -782
785
+ ... -783
786
+ ▁che -784
787
+ ▁Br -785
788
+ ird -786
789
+ ▁Ar -787
790
+ ▁war -788
791
+ inat -789
792
+ ▁My -790
793
+ ital -791
794
+ wh -792
795
+ med -793
796
+ ▁pur -794
797
+ ully -795
798
+ ▁One -796
799
+ ▁rat -797
800
+ ines -798
801
+ ▁Of -799
802
+ io -800
803
+ ▁loc -801
804
+ ret -802
805
+ ctor -803
806
+ ▁leg -804
807
+ stit -805
808
+ ined -806
809
+ ught -807
810
+ ▁dur -808
811
+ ▁es -809
812
+ vent -810
813
+ aj -811
814
+ ▁bro -812
815
+ ▁saw -813
816
+ ▁sec -814
817
+ ream -815
818
+ ▁pop -816
819
+ reen -817
820
+ ▁Ind -818
821
+ els -819
822
+ ▁yet -820
823
+ ired -821
824
+ ▁sw -822
825
+ tro -823
826
+ oup -824
827
+ most -825
828
+ pean -826
829
+ eds -827
830
+ ush -828
831
+ oh -829
832
+ ▁Se -830
833
+ ▁tea -831
834
+ ann -832
835
+ ilit -833
836
+ err -834
837
+ pend -835
838
+ ton -836
839
+ ased -837
840
+ ▁aff -838
841
+ ▁mor -839
842
+ ▁dra -840
843
+ put -841
844
+ ▁dr -842
845
+ ins -843
846
+ uat -844
847
+ nect -845
848
+ cri -846
849
+ outh -847
850
+ ▁ra -848
851
+ ▁pay -849
852
+ ms -850
853
+ ▁av -851
854
+ bs -852
855
+ ling -853
856
+ ▁De -854
857
+ ▁Or -855
858
+ ove -856
859
+ ▁Can -857
860
+ ▁eng -858
861
+ ames -859
862
+ ided -860
863
+ ▁Go -861
864
+ mitt -862
865
+ ode -863
866
+ ▁cre -864
867
+ par -865
868
+ ides -866
869
+ pos -867
870
+ ▁fav -868
871
+ ▁air -869
872
+ ▁New -870
873
+ ▁bad -871
874
+ ▁six -872
875
+ vat -873
876
+ ▁pat -874
877
+ not -875
878
+ ▁di -876
879
+ rop -877
880
+ ral -878
881
+ orn -879
882
+ ▁par -880
883
+ cing -881
884
+ ▁aw -882
885
+ orts -883
886
+ ox -884
887
+ ▁yes -885
888
+ cuss -886
889
+ eng -887
890
+ ives -888
891
+ erms -889
892
+ ▁job -890
893
+ mand -891
894
+ ying -892
895
+ ▁occ -893
896
+ aps -894
897
+ ases -895
898
+ ▁Not -896
899
+ rent -897
900
+ ency -898
901
+ att -899
902
+ ised -900
903
+ vice -901
904
+ ▁Eng -902
905
+ ▁est -903
906
+ oked -904
907
+ ▁Q -905
908
+ iron -906
909
+ idd -907
910
+ me -908
911
+ unch -909
912
+ ane -910
913
+ ▁z -911
914
+ br -912
915
+ arts -913
916
+ ▁fat -914
917
+ ery -915
918
+ anks -916
919
+ ▁jo -917
920
+ ▁mar -918
921
+ aw -919
922
+ ott -920
923
+ ards -921
924
+ ▁oh -922
925
+ ians -923
926
+ ▁sci -924
927
+ row -925
928
+ unt -926
929
+ ury -927
930
+ ▁abs -928
931
+ ergy -929
932
+ ▁Z -930
933
+ ump -931
934
+ ▁Am -932
935
+ ened -933
936
+ angu -934
937
+ ▁Pro -935
938
+ icat -936
939
+ itch -937
940
+ ▁dri -938
941
+ iat -939
942
+ ▁ -940
943
+ e -941
944
+ t -942
945
+ o -943
946
+ a -944
947
+ n -945
948
+ i -946
949
+ s -947
950
+ r -948
951
+ h -949
952
+ l -950
953
+ d -951
954
+ u -952
955
+ c -953
956
+ m -954
957
+ y -955
958
+ g -956
959
+ w -957
960
+ f -958
961
+ p -959
962
+ , -960
963
+ . -961
964
+ b -962
965
+ v -963
966
+ k -964
967
+ ' -965
968
+ I -966
969
+ T -967
970
+ A -968
971
+ S -969
972
+ x -970
973
+ W -971
974
+ j -972
975
+ C -973
976
+ B -974
977
+ M -975
978
+ ? -976
979
+ H -977
980
+ O -978
981
+ 0 -979
982
+ P -980
983
+ q -981
984
+ Y -982
985
+ N -983
986
+ L -984
987
+ D -985
988
+ 1 -986
989
+ E -987
990
+ G -988
991
+ z -989
992
+ F -990
993
+ R -991
994
+ - -992
995
+ 2 -993
996
+ J -994
997
+ U -995
998
+ 9 -996
999
+ K -997
1000
+ 5 -998
1001
+ 3 -999
1002
+ V -1000
1003
+ = -1001
1004
+ 4 -1002
1005
+ 8 -1003
1006
+ 6 -1004
1007
+ 7 -1005
1008
+ ! -1006
1009
+ % -1007
1010
+ : -1008
1011
+ Q -1009
1012
+ Z -1010
1013
+ $ -1011
1014
+ X -1012
1015
+ " -1013
1016
+ & -1014
1017
+ * -1015
1018
+ / -1016
1019
+ £ -1017
1020
+ + -1018
1021
+ € -1019
1022
+ _ -1020
1023
+ ^ -1021
1024
+ ¥ -1022
vocab.txt ADDED
@@ -0,0 +1,1023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ t
2
+ th
3
+ a
4
+ ##in
5
+ the
6
+ ##re
7
+ w
8
+ o
9
+ s
10
+ ##er
11
+ ##at
12
+ ##ou
13
+ ##nd
14
+ ##it
15
+ ##is
16
+ h
17
+ b
18
+ ##on
19
+ c
20
+ ##ing
21
+ ##en
22
+ to
23
+ m
24
+ f
25
+ p
26
+ ##or
27
+ ##an
28
+ ##es
29
+ of
30
+ d
31
+ ##ed
32
+ ##ll
33
+ and
34
+ I
35
+ in
36
+ l
37
+ ##ar
38
+ y
39
+ g
40
+ ##as
41
+ you
42
+ ##om
43
+ n
44
+ ##ic
45
+ ##ve
46
+ ##al
47
+ ##ion
48
+ ##us
49
+ be
50
+ ##ow
51
+ ##le
52
+ wh
53
+ e
54
+ ##ot
55
+ ##ut
56
+ it
57
+ is
58
+ we
59
+ T
60
+ re
61
+ ##et
62
+ A
63
+ ##ent
64
+ on
65
+ ha
66
+ ##ay
67
+ S
68
+ ##ct
69
+ Th
70
+ ##ver
71
+ ##id
72
+ ##ig
73
+ ##im
74
+ ##ro
75
+ for
76
+ ##ly
77
+ he
78
+ ##ke
79
+ ##ld
80
+ ##se
81
+ ##st
82
+ ##ch
83
+ st
84
+ ##all
85
+ ##ce
86
+ ##ur
87
+ ##ith
88
+ ##am
89
+ ##if
90
+ ##ir
91
+ go
92
+ u
93
+ as
94
+ was
95
+ ##ad
96
+ W
97
+ k
98
+ an
99
+ ##ht
100
+ ##th
101
+ r
102
+ are
103
+ ##ere
104
+ se
105
+ do
106
+ B
107
+ so
108
+ sh
109
+ not
110
+ li
111
+ ##od
112
+ C
113
+ ##ust
114
+ ##ill
115
+ ##ight
116
+ ##ally
117
+ And
118
+ ##ter
119
+ or
120
+ me
121
+ M
122
+ ##ome
123
+ ##op
124
+ at
125
+ ##il
126
+ The
127
+ ##ould
128
+ j
129
+ ##ant
130
+ So
131
+ H
132
+ ##ol
133
+ ##ain
134
+ can
135
+ de
136
+ ne
137
+ ##ore
138
+ con
139
+ kn
140
+ ##ck
141
+ ##ul
142
+ fr
143
+ ab
144
+ ##ers
145
+ ##ess
146
+ ##ge
147
+ pro
148
+ ##pe
149
+ ##ate
150
+ su
151
+ com
152
+ but
153
+ all
154
+ ##est
155
+ ##qu
156
+ ex
157
+ al
158
+ ##ra
159
+ O
160
+ ##out
161
+ ##use
162
+ ##very
163
+ ##pp
164
+ Y
165
+ ch
166
+ ##ri
167
+ ##ist
168
+ v
169
+ lo
170
+ ##ment
171
+ ##art
172
+ P
173
+ ##nt
174
+ ##ab
175
+ one
176
+ N
177
+ ##ive
178
+ wor
179
+ ##ions
180
+ ##ort
181
+ L
182
+ by
183
+ ##ich
184
+ my
185
+ ##ity
186
+ ##ok
187
+ G
188
+ ##res
189
+ up
190
+ ##un
191
+ ##um
192
+ ##ea
193
+ ##ind
194
+ ##and
195
+ ##ink
196
+ ##el
197
+ D
198
+ ##em
199
+ E
200
+ ##os
201
+ ##oug
202
+ if
203
+ ##ca
204
+ out
205
+ int
206
+ ##ie
207
+ F
208
+ It
209
+ his
210
+ ##ard
211
+ had
212
+ tr
213
+ ##her
214
+ ##our
215
+ ##ies
216
+ ##ake
217
+ R
218
+ We
219
+ get
220
+ don
221
+ us
222
+ ##ak
223
+ pl
224
+ ##ect
225
+ ##ure
226
+ ##ame
227
+ ##ast
228
+ who
229
+ ##ack
230
+ le
231
+ sa
232
+ ##iv
233
+ ##ci
234
+ ##ide
235
+ tim
236
+ our
237
+ ##ound
238
+ ##ous
239
+ co
240
+ pe
241
+ ##ose
242
+ ##ud
243
+ see
244
+ ##ough
245
+ man
246
+ qu
247
+ You
248
+ ##so
249
+ ##ople
250
+ Wh
251
+ ##ong
252
+ ##ap
253
+ ##ther
254
+ J
255
+ ##are
256
+ ##ine
257
+ say
258
+ im
259
+ But
260
+ ##ings
261
+ has
262
+ ag
263
+ ##ff
264
+ her
265
+ ##itt
266
+ ##one
267
+ en
268
+ ar
269
+ fe
270
+ ##ven
271
+ any
272
+ mo
273
+ ##reat
274
+ ##ag
275
+ how
276
+ cl
277
+ ##pt
278
+ now
279
+ ##own
280
+ ##ber
281
+ him
282
+ act
283
+ ##hing
284
+ ##ice
285
+ no
286
+ ##ans
287
+ ##iz
288
+ fa
289
+ ##per
290
+ ##pl
291
+ te
292
+ ad
293
+ ##age
294
+ ##ree
295
+ tw
296
+ ##ank
297
+ He
298
+ ##ple
299
+ ##ite
300
+ ##ry
301
+ U
302
+ ##ish
303
+ ##ire
304
+ ##ue
305
+ In
306
+ she
307
+ ##ble
308
+ ##cc
309
+ ##nder
310
+ way
311
+ pr
312
+ ##ear
313
+ did
314
+ po
315
+ ##eah
316
+ un
317
+ ##omet
318
+ ##ence
319
+ ##ep
320
+ ##uch
321
+ sp
322
+ ##ach
323
+ ##og
324
+ ##ance
325
+ ##able
326
+ ##iff
327
+ ##sel
328
+ got
329
+ ##way
330
+ gr
331
+ ##alk
332
+ res
333
+ ##ated
334
+ ##irst
335
+ ##ick
336
+ ##ass
337
+ two
338
+ dis
339
+ ##ord
340
+ pre
341
+ ##ount
342
+ ##ase
343
+ ##ip
344
+ ##ult
345
+ ##ical
346
+ ##orm
347
+ ##ary
348
+ ##ace
349
+ spe
350
+ Ch
351
+ thr
352
+ imp
353
+ ##int
354
+ am
355
+ off
356
+ ##act
357
+ ##ia
358
+ ro
359
+ ##ress
360
+ per
361
+ fo
362
+ br
363
+ K
364
+ ##vel
365
+ gu
366
+ bo
367
+ ##ang
368
+ ##kay
369
+ ##ub
370
+ ##ign
371
+ may
372
+ ##ving
373
+ ##ces
374
+ ##ens
375
+ ##cl
376
+ lot
377
+ ##ru
378
+ ##ade
379
+ bet
380
+ bl
381
+ let
382
+ ##fore
383
+ ##co
384
+ ##ild
385
+ ##ning
386
+ ##xt
387
+ ##ile
388
+ ##ark
389
+ ##self
390
+ app
391
+ ##ory
392
+ ##du
393
+ day
394
+ St
395
+ ##ater
396
+ use
397
+ ##ys
398
+ ##fter
399
+ new
400
+ ##ious
401
+ ##ial
402
+ ##he
403
+ ##wn
404
+ ##ved
405
+ ##red
406
+ fl
407
+ ##iss
408
+ ##ody
409
+ ##form
410
+ ##ian
411
+ ##tain
412
+ bu
413
+ V
414
+ rec
415
+ ##ty
416
+ ##be
417
+ sc
418
+ ##ors
419
+ ##vers
420
+ put
421
+ ##ife
422
+ If
423
+ ##we
424
+ ##te
425
+ ##ject
426
+ ##ath
427
+ ##ting
428
+ rem
429
+ acc
430
+ ##ull
431
+ ##ons
432
+ ind
433
+ ser
434
+ ke
435
+ ##ates
436
+ ##ves
437
+ ##na
438
+ ##lic
439
+ des
440
+ its
441
+ ##ful
442
+ ##ents
443
+ ##erm
444
+ ##ac
445
+ ##ered
446
+ ##ise
447
+ sy
448
+ ##urn
449
+ em
450
+ ##oth
451
+ ##ual
452
+ ##ne
453
+ ##ward
454
+ ##ib
455
+ try
456
+ pos
457
+ ##nds
458
+ ##ft
459
+ ##get
460
+ ##ph
461
+ ob
462
+ ##ady
463
+ ##igh
464
+ ##ood
465
+ rel
466
+ wr
467
+ ##ug
468
+ ##ears
469
+ ##ail
470
+ Now
471
+ bit
472
+ ##ng
473
+ Oh
474
+ hel
475
+ ##ange
476
+ reg
477
+ rep
478
+ bel
479
+ sm
480
+ ##ost
481
+ ##tern
482
+ ##gr
483
+ own
484
+ end
485
+ ##pect
486
+ ##ily
487
+ ##day
488
+ ##ied
489
+ ##ific
490
+ ##ower
491
+ add
492
+ ##cess
493
+ ##ict
494
+ ##ible
495
+ bas
496
+ i
497
+ op
498
+ ##cial
499
+ ##ular
500
+ Be
501
+ ##ced
502
+ too
503
+ ##ks
504
+ ##ew
505
+ ##mer
506
+ ph
507
+ ##ob
508
+ ##==
509
+ la
510
+ set
511
+ min
512
+ sub
513
+ gen
514
+ ##atch
515
+ ##..
516
+ inv
517
+ As
518
+ nat
519
+ sl
520
+ num
521
+ ##av
522
+ ##ways
523
+ God
524
+ ##stem
525
+ ac
526
+ att
527
+ ev
528
+ def
529
+ ##llow
530
+ str
531
+ ##lect
532
+ ##ars
533
+ cr
534
+ Is
535
+ ##olog
536
+ ##les
537
+ ##oy
538
+ ask
539
+ inc
540
+ ##body
541
+ ent
542
+ pol
543
+ ##ness
544
+ ##ix
545
+ why
546
+ ##onna
547
+ ear
548
+ tak
549
+ Un
550
+ ##ited
551
+ ##mun
552
+ ##li
553
+ ##ute
554
+ ##ract
555
+ dec
556
+ ##uro
557
+ mak
558
+ fin
559
+ ##ween
560
+ No
561
+ ##arch
562
+ bec
563
+ ##gan
564
+ ##old
565
+ ##cy
566
+ big
567
+ For
568
+ ##ren
569
+ ##als
570
+ ##und
571
+ Al
572
+ All
573
+ ##ss
574
+ ##ows
575
+ mod
576
+ ##ock
577
+ id
578
+ ##ism
579
+ ##cus
580
+ gl
581
+ ##ably
582
+ ass
583
+ car
584
+ ##ata
585
+ ##ppen
586
+ ##led
587
+ sim
588
+ mon
589
+ ##ics
590
+ giv
591
+ ##cept
592
+ Mr
593
+ ##pan
594
+ pub
595
+ eff
596
+ How
597
+ ##ps
598
+ ##vern
599
+ ##end
600
+ ##hip
601
+ ##iew
602
+ ##ope
603
+ An
604
+ She
605
+ Com
606
+ ##ee
607
+ ##ures
608
+ ##ell
609
+ ##ouse
610
+ ##cond
611
+ ##king
612
+ ##oc
613
+ ##ues
614
+ ##ever
615
+ To
616
+ ##clud
617
+ ins
618
+ exp
619
+ old
620
+ mem
621
+ ref
622
+ tra
623
+ far
624
+ ##ave
625
+ ##rat
626
+ sur
627
+ ##ruct
628
+ ##rib
629
+ ##duct
630
+ ##uff
631
+ met
632
+ sch
633
+ ##ince
634
+ run
635
+ ##ense
636
+ cle
637
+ ==
638
+ ##mon
639
+ ##ize
640
+ ord
641
+ ##blem
642
+ ##tin
643
+ Let
644
+ ##ner
645
+ ##ond
646
+ ##its
647
+ cor
648
+ ##land
649
+ cur
650
+ Re
651
+ bus
652
+ uh
653
+ ##air
654
+ ##ote
655
+ ##ants
656
+ ##ason
657
+ ##ric
658
+ el
659
+ cer
660
+ ##nce
661
+ fam
662
+ cap
663
+ ##uck
664
+ ##ool
665
+ ##ried
666
+ cou
667
+ fun
668
+ wom
669
+ hum
670
+ ty
671
+ ap
672
+ ##ike
673
+ few
674
+ ##oney
675
+ inf
676
+ ##ont
677
+ ##ese
678
+ ##ook
679
+ ##gy
680
+ ##uth
681
+ ##ulat
682
+ ##ieve
683
+ ##ized
684
+ ##ross
685
+ ple
686
+ um
687
+ val
688
+ equ
689
+ lea
690
+ lar
691
+ ##ah
692
+ ##eral
693
+ ed
694
+ ##ared
695
+ ##lish
696
+ ##arn
697
+ ##ds
698
+ ##esn
699
+ iss
700
+ ca
701
+ ##ted
702
+ ##ices
703
+ wee
704
+ ##ash
705
+ top
706
+ ##ten
707
+ ##up
708
+ ##ts
709
+ ##gin
710
+ ##con
711
+ ##ari
712
+ opp
713
+ ##osed
714
+ eas
715
+ ext
716
+ ##gg
717
+ ##az
718
+ Fr
719
+ ##ideo
720
+ ##izat
721
+ men
722
+ mom
723
+ ret
724
+ ##tty
725
+ ##rist
726
+ gra
727
+ ##alth
728
+ ##ef
729
+ det
730
+ ##ax
731
+ mat
732
+ ##chn
733
+ ##ern
734
+ ##peri
735
+ bre
736
+ Sh
737
+ ##sw
738
+ ##erat
739
+ sit
740
+ ##ters
741
+ ##ale
742
+ ##man
743
+ sol
744
+ ##ork
745
+ adv
746
+ ##ety
747
+ vis
748
+ med
749
+ ##uc
750
+ ##less
751
+ unt
752
+ ##gram
753
+ ##ets
754
+ ##ists
755
+ ey
756
+ col
757
+ ##imes
758
+ law
759
+ pri
760
+ ##sid
761
+ On
762
+ mot
763
+ ##ield
764
+ Do
765
+ At
766
+ ##ages
767
+ ##amp
768
+ art
769
+ ##miss
770
+ sk
771
+ ##alf
772
+ ##pr
773
+ ##ier
774
+ beh
775
+ Yes
776
+ ##ural
777
+ ##ime
778
+ wa
779
+ ##oks
780
+ ##bers
781
+ ##ger
782
+ ##ient
783
+ ##ries
784
+ ##...
785
+ che
786
+ Br
787
+ ##ird
788
+ Ar
789
+ war
790
+ ##inat
791
+ My
792
+ ##ital
793
+ ##wh
794
+ ##med
795
+ pur
796
+ ##ully
797
+ One
798
+ rat
799
+ ##ines
800
+ Of
801
+ ##io
802
+ loc
803
+ ##ret
804
+ ##ctor
805
+ leg
806
+ ##stit
807
+ ##ined
808
+ ##ught
809
+ dur
810
+ es
811
+ ##vent
812
+ ##aj
813
+ bro
814
+ saw
815
+ sec
816
+ ##ream
817
+ pop
818
+ ##reen
819
+ Ind
820
+ ##els
821
+ yet
822
+ ##ired
823
+ sw
824
+ ##tro
825
+ ##oup
826
+ ##most
827
+ ##pean
828
+ ##eds
829
+ ##ush
830
+ ##oh
831
+ Se
832
+ tea
833
+ ##ann
834
+ ##ilit
835
+ ##err
836
+ ##pend
837
+ ##ton
838
+ ##ased
839
+ aff
840
+ mor
841
+ dra
842
+ ##put
843
+ dr
844
+ ##ins
845
+ ##uat
846
+ ##nect
847
+ ##cri
848
+ ##outh
849
+ ra
850
+ pay
851
+ ##ms
852
+ av
853
+ ##bs
854
+ ##ling
855
+ De
856
+ Or
857
+ ##ove
858
+ Can
859
+ eng
860
+ ##ames
861
+ ##ided
862
+ Go
863
+ ##mitt
864
+ ##ode
865
+ cre
866
+ ##par
867
+ ##ides
868
+ ##pos
869
+ fav
870
+ air
871
+ New
872
+ bad
873
+ six
874
+ ##vat
875
+ pat
876
+ ##not
877
+ di
878
+ ##rop
879
+ ##ral
880
+ ##orn
881
+ par
882
+ ##cing
883
+ aw
884
+ ##orts
885
+ ##ox
886
+ yes
887
+ ##cuss
888
+ ##eng
889
+ ##ives
890
+ ##erms
891
+ job
892
+ ##mand
893
+ ##ying
894
+ occ
895
+ ##aps
896
+ ##ases
897
+ Not
898
+ ##rent
899
+ ##ency
900
+ ##att
901
+ ##ised
902
+ ##vice
903
+ Eng
904
+ est
905
+ ##oked
906
+ Q
907
+ ##iron
908
+ ##idd
909
+ ##me
910
+ ##unch
911
+ ##ane
912
+ z
913
+ ##br
914
+ ##arts
915
+ fat
916
+ ##ery
917
+ ##anks
918
+ jo
919
+ mar
920
+ ##aw
921
+ ##ott
922
+ ##ards
923
+ oh
924
+ ##ians
925
+ sci
926
+ ##row
927
+ ##unt
928
+ ##ury
929
+ abs
930
+ ##ergy
931
+ Z
932
+ ##ump
933
+ Am
934
+ ##ened
935
+ ##angu
936
+ Pro
937
+ ##icat
938
+ ##itch
939
+ dri
940
+ ##iat
941
+ ▁
942
+ ##e
943
+ ##t
944
+ ##o
945
+ ##a
946
+ ##n
947
+ ##i
948
+ ##s
949
+ ##r
950
+ ##h
951
+ ##l
952
+ ##d
953
+ ##u
954
+ ##c
955
+ ##m
956
+ ##y
957
+ ##g
958
+ ##w
959
+ ##f
960
+ ##p
961
+ ##,
962
+ ##.
963
+ ##b
964
+ ##v
965
+ ##k
966
+ ##'
967
+ ##I
968
+ ##T
969
+ ##A
970
+ ##S
971
+ ##x
972
+ ##W
973
+ ##j
974
+ ##C
975
+ ##B
976
+ ##M
977
+ ##?
978
+ ##H
979
+ ##O
980
+ ##0
981
+ ##P
982
+ ##q
983
+ ##Y
984
+ ##N
985
+ ##L
986
+ ##D
987
+ ##1
988
+ ##E
989
+ ##G
990
+ ##z
991
+ ##F
992
+ ##R
993
+ ##-
994
+ ##2
995
+ ##J
996
+ ##U
997
+ ##9
998
+ ##K
999
+ ##5
1000
+ ##3
1001
+ ##V
1002
+ ##=
1003
+ ##4
1004
+ ##8
1005
+ ##6
1006
+ ##7
1007
+ ##!
1008
+ ##%
1009
+ ##:
1010
+ ##Q
1011
+ ##Z
1012
+ ##$
1013
+ ##X
1014
+ ##"
1015
+ ##&
1016
+ ##*
1017
+ ##/
1018
+ ##£
1019
+ ##+
1020
+ ##€
1021
+ ##_
1022
+ ##^
1023
+ ##¥