Hugging Face's logo

Spaces:
fedirz
/
faster-whisper-server
Configuration error

App Files Community
faster-whisper-server
File size: 3,385 Bytes 
313814b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
 
import enum

from pydantic import BaseModel, Field
from pydantic_settings import BaseSettings, SettingsConfigDict

SAMPLES_PER_SECOND = 16000
BYTES_PER_SAMPLE = 2
BYTES_PER_SECOND = SAMPLES_PER_SECOND * BYTES_PER_SAMPLE
# 2 BYTES = 16 BITS = 1 SAMPLE
# 1 SECOND OF AUDIO = 32000 BYTES = 16000 SAMPLES


# TODO: confirm names
class Model(enum.StrEnum):
    TINY_EN = "tiny.en"
    TINY = "tiny"
    BASE_EN = "base.en"
    BASE = "base"
    SMALL_EN = "small.en"
    SMALL = "small"
    MEDIUM_EN = "medium.en"
    MEDIUM = "medium"
    LARGE = "large"
    LARGE_V1 = "large-v1"
    LARGE_V2 = "large-v2"
    LARGE_V3 = "large-v3"
    DISTIL_SMALL_EN = "distil-small.en"
    DISTIL_MEDIUM_EN = "distil-medium.en"
    DISTIL_LARGE_V2 = "distil-large-v2"
    DISTIL_LARGE_V3 = "distil-large-v3"


class Device(enum.StrEnum):
    CPU = "cpu"
    CUDA = "cuda"
    AUTO = "auto"


# https://github.com/OpenNMT/CTranslate2/blob/master/docs/quantization.md
class Quantization(enum.StrEnum):
    INT8 = "int8"
    INT8_FLOAT16 = "int8_float16"
    INT8_BFLOAT16 = "int8_bfloat16"
    INT8_FLOAT32 = "int8_float32"
    INT16 = "int16"
    FLOAT16 = "float16"
    BFLOAT16 = "bfloat16"
    FLOAT32 = "float32"
    DEFAULT = "default"


class Language(enum.StrEnum):
    AF = "af"
    AM = "am"
    AR = "ar"
    AS = "as"
    AZ = "az"
    BA = "ba"
    BE = "be"
    BG = "bg"
    BN = "bn"
    BO = "bo"
    BR = "br"
    BS = "bs"
    CA = "ca"
    CS = "cs"
    CY = "cy"
    DA = "da"
    DE = "de"
    EL = "el"
    EN = "en"
    ES = "es"
    ET = "et"
    EU = "eu"
    FA = "fa"
    FI = "fi"
    FO = "fo"
    FR = "fr"
    GL = "gl"
    GU = "gu"
    HA = "ha"
    HAW = "haw"
    HE = "he"
    HI = "hi"
    HR = "hr"
    HT = "ht"
    HU = "hu"
    HY = "hy"
    ID = "id"
    IS = "is"
    IT = "it"
    JA = "ja"
    JW = "jw"
    KA = "ka"
    KK = "kk"
    KM = "km"
    KN = "kn"
    KO = "ko"
    LA = "la"
    LB = "lb"
    LN = "ln"
    LO = "lo"
    LT = "lt"
    LV = "lv"
    MG = "mg"
    MI = "mi"
    MK = "mk"
    ML = "ml"
    MN = "mn"
    MR = "mr"
    MS = "ms"
    MT = "mt"
    MY = "my"
    NE = "ne"
    NL = "nl"
    NN = "nn"
    NO = "no"
    OC = "oc"
    PA = "pa"
    PL = "pl"
    PS = "ps"
    PT = "pt"
    RO = "ro"
    RU = "ru"
    SA = "sa"
    SD = "sd"
    SI = "si"
    SK = "sk"
    SL = "sl"
    SN = "sn"
    SO = "so"
    SQ = "sq"
    SR = "sr"
    SU = "su"
    SV = "sv"
    SW = "sw"
    TA = "ta"
    TE = "te"
    TG = "tg"
    TH = "th"
    TK = "tk"
    TL = "tl"
    TR = "tr"
    TT = "tt"
    UK = "uk"
    UR = "ur"
    UZ = "uz"
    VI = "vi"
    YI = "yi"
    YO = "yo"
    YUE = "yue"
    ZH = "zh"


class WhisperConfig(BaseModel):
    model: Model = Field(default=Model.DISTIL_SMALL_EN)
    inference_device: Device = Field(default=Device.AUTO)
    compute_type: Quantization = Field(default=Quantization.DEFAULT)


class Config(BaseSettings):
    model_config = SettingsConfigDict(env_nested_delimiter="_")

    log_level: str = "info"
    whisper: WhisperConfig = WhisperConfig()
    """
    Max duration to for the next audio chunk before finilizing the transcription and closing the connection.
    """
    max_no_data_seconds: float = 1.0
    min_duration: float = 1.0
    word_timestamp_error_margin: float = 0.2
    inactivity_window_seconds: float = 3.0
    max_inactivity_seconds: float = 1.5


config = Config()