admin-valentin/WhisperLiveKit
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix --direct-english-translation not setting task=translate for localagreement backends

Browse source 
the flag was only used for tokenizer language selection but never
actually passed to whisper/faster-whisper transcribe calls. also init
OpenaiApiASR.task and read from transcribe_kargs.

fixes #306
This commit is contained in:
Quentin Fuxa 2026-02-11 22:10:00 +01:00
parent aded1649ae
commit 8923337380
2 changed files with 6 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
whisperlivekit/local_agreement/backends.py
View file

@ -249,6 +249,7 @@ class OpenaiApiASR(ASRBase):
self.load_model() self.load_model()
self.use_vad_opt = False self.use_vad_opt = False
self.direct_english_translation = False self.direct_english_translation = False
self.task = "transcribe"
def load_model(self, *args, **kwargs): def load_model(self, *args, **kwargs):
from openai import OpenAI from openai import OpenAI
@ -294,7 +295,8 @@ class OpenaiApiASR(ASRBase):
params["language"] = self.original_language params["language"] = self.original_language
if prompt: if prompt:
params["prompt"] = prompt params["prompt"] = prompt
proc = self.client.audio.translations if self.task == "translate" else self.client.audio.transcriptions task = self.transcribe_kargs.get("task", self.task)
proc = self.client.audio.translations if task == "translate" else self.client.audio.transcriptions
transcript = proc.create(**params) transcript = proc.create(**params)
logger.debug(f"OpenAI API processed accumulated {self.transcribed_seconds} seconds") logger.debug(f"OpenAI API processed accumulated {self.transcribed_seconds} seconds")
return transcript return transcript

5
whisperlivekit/local_agreement/whisper_online.py
View file

@ -146,6 +146,7 @@ def backend_factory(
if direct_english_translation: if direct_english_translation:
tgt_language = "en" # Whisper translates into English tgt_language = "en" # Whisper translates into English
asr.transcribe_kargs["task"] = "translate"
else: else:
tgt_language = lan # Whisper transcribes in this language tgt_language = lan # Whisper transcribes in this language
@ -154,9 +155,9 @@ def backend_factory(
tokenizer = create_tokenizer(tgt_language) tokenizer = create_tokenizer(tgt_language)
else: else:
tokenizer = None tokenizer = None
warmup_asr(asr, warmup_file) warmup_asr(asr, warmup_file)
asr.confidence_validation = confidence_validation asr.confidence_validation = confidence_validation
asr.tokenizer = tokenizer asr.tokenizer = tokenizer
asr.buffer_trimming = buffer_trimming asr.buffer_trimming = buffer_trimming