forked from aivo0/rust-asr-server
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathconfig.toml
More file actions
148 lines (131 loc) · 4.98 KB
/
Copy pathconfig.toml
File metadata and controls
148 lines (131 loc) · 4.98 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
bind_address = "127.0.0.1:8081"
[tts]
default_voice = "et-cv"
provider = "cpu"
num_threads = 4
sample_rate = 22050
[tts.voices.et-cv]
model_path = "./models/tts/vits-coqui-et-cv/model.onnx"
tokens_path = "./models/tts/vits-coqui-et-cv/tokens.txt"
name = "Estonian (CommonVoice)"
language_code = "et"
sample_rate = 22050
[asr]
default_language = "fastconformer_transducer_en_1040ms"
provider = "cpu"
num_threads = 6
sample_rate = 16000
[asr.pseudo_streaming]
enabled = true
chunk_duration_seconds = 1.6
total_buffer_seconds = 4.0
left_context_seconds = 1.2
right_context_seconds = 1.2
merge_algorithm = "WordBoundary"
max_delay_seconds = 10.0
[asr.punctuation]
enabled = true
model_path = "./models/sherpa-onnx-online-punct-en-2024-08-06/model.onnx"
bpe_vocab_path = "./models/sherpa-onnx-online-punct-en-2024-08-06/bpe.vocab"
num_threads = 2
[asr.models.et]
model_path = "./models/streaming-zipformer-large.et-en/encoder.onnx"
tokens_path = "./models/streaming-zipformer-large.et-en/tokens.txt"
name = "Estonian/English Zipformer"
language_code = "et"
model_type = "Transducer"
# [asr.models.en]
# model_path = "./models/parakeet-realtime-eou-120m-v1-onnx"
# name = "Parakeet Realtime EOU 120M v1 (English)"
# language_code = "en"
# model_type = "ParakeetEOU"
# sample_rate = 16000
# Note: Disabled to free up memory - using FastConformer CTC instead
# [asr.models.ru]
# model_path = "./models/sherpa-onnx-streaming-t-one-russian-2025-09-08/model.onnx"
# tokens_path = "./models/sherpa-onnx-streaming-t-one-russian-2025-09-08/tokens.txt"
# name = "Russian T-one Streaming (CTC, telephony)"
# language_code = "ru"
# model_type = "Ctc"
# sample_rate = 8000
# Note: Temporarily disabled due to 'encoder_dims' metadata issue causing server crash
[asr.models.fastconformer_transducer_en_1040ms]
model_path = "./models/fastconformer-en-1040ms-transducer/encoder.onnx"
tokens_path = "./models/fastconformer-en-1040ms-transducer/tokens.txt"
encoder_path = "./models/fastconformer-en-1040ms-transducer/encoder.onnx"
decoder_path = "./models/fastconformer-en-1040ms-transducer/decoder.onnx"
joiner_path = "./models/fastconformer-en-1040ms-transducer/joiner.onnx"
name = "FastConformer English Streaming 1040ms (RNNT)"
language_code = "en"
model_type = "Transducer"
sample_rate = 16000
punctuation = true # Enable punctuation post-processing for this English model
# Disabled: CTC 1040ms (replaced by RNNT above)
# [asr.models.fastconformer_ctc_en_1040ms]
# model_path = "./models/fastconformer-en-1040ms-ctc/model.onnx"
# tokens_path = "./models/fastconformer-en-1040ms-ctc/tokens.txt"
# name = "FastConformer English Streaming 1040ms (CTC)"
# language_code = "en"
# model_type = "NemoCtc"
# sample_rate = 16000
# punctuation = true
[asr.models.fastconformer_transducer_en_80ms]
model_path = "./models/fastconformer-en-80ms-transducer/encoder.onnx"
tokens_path = "./models/fastconformer-en-80ms-transducer/tokens.txt"
encoder_path = "./models/fastconformer-en-80ms-transducer/encoder.onnx"
decoder_path = "./models/fastconformer-en-80ms-transducer/decoder.onnx"
joiner_path = "./models/fastconformer-en-80ms-transducer/joiner.onnx"
name = "FastConformer English Streaming 80ms (RNNT)"
language_code = "en"
model_type = "Transducer"
sample_rate = 16000
punctuation = false # Model has native punctuation & capitalization (pc)
[asr.models.parakeet_tdt_v3]
model_path = "./models/sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8/encoder.int8.onnx"
tokens_path = "./models/sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8/tokens.txt"
encoder_path = "./models/sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8/encoder.int8.onnx"
decoder_path = "./models/sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8/decoder.int8.onnx"
joiner_path = "./models/sherpa-onnx-nemo-parakeet-tdt-0.6b-v3-int8/joiner.int8.onnx"
name = "Parakeet TDT v3 (25 European languages)"
language_code = "auto"
model_type = "NemoTransducer"
sample_rate = 16000
[audio]
buffer_size = 8192
max_chunk_size = 4096
silence_threshold = 0.01
min_speech_duration_ms = 500
[processing]
max_concurrent_sessions = 100
processing_threads = 6
session_timeout_seconds = 300
[vad]
enabled = false
model_path = "./models/silero_vad.onnx"
threshold = 0.5
min_silence_duration = 0.3
min_speech_duration = 0.2
max_speech_duration = 20.0
window_size = 512
buffer_size_seconds = 30.0
# Real-time speaker diarization (opt-in per session via {"type":"start","diarize":true}).
# Model produced by offline-asr/scripts/export_redimnet2_onnx.py; the model_version
# must match the roster's model_version exactly.
[diarization]
enabled = false
model_path = "./models/speaker/redimnet2-b6-vb2-vox2-lm.onnx"
model_version = "redimnet2-b6-vb2-vox2-lm"
num_threads = 2
vad_threshold = 0.5
vad_min_silence_duration = 0.35
vad_min_speech_duration = 0.25
window_min_seconds = 1.0
window_target_seconds = 3.0
window_max_seconds = 5.0
clustering_interval_seconds = 5.0
clustering_distance_threshold = 0.4
max_segments = 1000
# Known-speaker identification against an offline-asr roster.json (optional).
[speaker_id]
# roster_path = "./models/speaker/roster.json"
match_threshold = 0.6