From eb1574b813e230b156fc09eaaf03227b1b0b4134 Mon Sep 17 00:00:00 2001
From: weilikai <jasper@talkus.fun>
Date: 星期六, 20 九月 2025 22:41:05 +0800
Subject: [PATCH] fix: support loading .pcm (16k 1c 16bit) audio files in load_utils.py (#2667) (#2668)
---
funasr/models/whisper/model.py | 8 +++++---
1 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/funasr/models/whisper/model.py b/funasr/models/whisper/model.py
index a332100..398eea3 100644
--- a/funasr/models/whisper/model.py
+++ b/funasr/models/whisper/model.py
@@ -9,6 +9,7 @@
from torch import nn
import whisper
+
# import whisper_timestamped as whisper
from funasr.utils.load_utils import load_audio_text_image_video, extract_fbank
@@ -27,6 +28,7 @@
@tables.register("model_classes", "Whisper-large-v1")
@tables.register("model_classes", "Whisper-large-v2")
@tables.register("model_classes", "Whisper-large-v3")
+@tables.register("model_classes", "Whisper-large-v3-turbo")
@tables.register("model_classes", "WhisperWarp")
class WhisperWarp(nn.Module):
def __init__(self, *args, **kwargs):
@@ -111,10 +113,10 @@
# decode the audio
options = whisper.DecodingOptions(**kwargs.get("DecodingOptions", {}))
-
- result = whisper.decode(self.model, speech, language='english')
+
+ result = whisper.decode(self.model, speech, options=options)
# result = whisper.transcribe(self.model, speech)
-
+
results = []
result_i = {"key": key[0], "text": result.text}
--
Gitblit v1.9.1