pogmommy/rebble-asr
Archived
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Use the latest_short model where applicable.

Browse source
This commit is contained in:
Katharine Berry 2023-08-02 19:02:15 -07:00
parent 37302ebed4
commit 141910548a
2 changed files with 49 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
asr/__init__.py
View file

@ -1,10 +1,9 @@
import uuid
import gevent.monkey import gevent.monkey
gevent.monkey.patch_all() gevent.monkey.patch_all()
import base64 import base64
from email.mime.multipart import MIMEMultipart from email.mime.multipart import MIMEMultipart
from email.message import Message from email.message import Message
from .model_map import get_model_for_lang
import json import json
import struct import struct
import os import os
@ -69,6 +68,7 @@ def recognise():
'sample_rate_hertz': 16000, 'sample_rate_hertz': 16000,
'max_alternatives': 1, 'max_alternatives': 1,
'enableAutomaticPunctuation': True, 'enableAutomaticPunctuation': True,
'model': get_model_for_lang(lang),
# 'metadata': { # 'metadata': {
# 'interaction_type': 'DICTATION', # 'interaction_type': 'DICTATION',
# 'microphone_distance': 'NEARFIELD', # 'microphone_distance': 'NEARFIELD',

47
asr/model_map.py Normal file
View file

@ -0,0 +1,47 @@
MODEL_MAP = {
'af-za': 'default', # latest_short not supported
'cs-cz': 'latest_short',
'da-dk': 'latest_short',
'de-de': 'latest_short',
'en-au': 'latest_short',
'en-us': 'latest_short',
'en-gb': 'latest_short',
'en-in': 'latest_short',
'en-ca': 'default', # latest_short not supported
'fi-fi': 'latest_short',
'fil-ph': 'default', # latest_short not supported
'fr-ca': 'latest_short',
'fr-fr': 'latest_short',
'gl-es': 'default', # latest_short not supported
'id-id': 'latest_short',
'is-is': 'default', # latest_short not supported
'it-it': 'latest_short',
'ko-kr': 'latest_short',
'lv-lv': 'default', # latest_short not supported
'lt-lt': 'default', # latest_short not supported
'hr-hr': 'default', # latest_short not supported
'hu-hu': 'default', # latest_short not supported
'ms-my': 'default', # latest_short not supported
'nl-nl': 'latest_short',
'nb-no': 'default', # this isn't listed at all. Did we mean no_NO?
'no-no': 'latest_short',
'pt-pt': 'latest_short',
'pl-pl': 'latest_short',
'ro-ro': 'latest_short',
'ru-ru': 'latest_short',
'es-es': 'latest_short',
'es-mx': 'default', # latest_short not supported
'es-us': 'latest_short',
'sk-sk': 'default', # latest_short not supported
'sl-sl': 'default', # this doesn't seem to exist
'sl-si': 'default', # latest_short not supported
'sv-se': 'latest_short',
'sw-tz': 'default', # latest_short not supported
'sw-ke': 'default', # latest_short not supported
'tr-tr': 'default', # latest_short lacks automatic punctuation
'zu-za': 'default', # latest_short not supported
}
def get_model_for_lang(code: str) -> str:
return MODEL_MAP.get(code.lower(), 'default')