From 1af68ba6ffc21d4dc3bd5f01cda656def97e361c Mon Sep 17 00:00:00 2001
From: Nixon <2465004358@qq.com>
Date: 星期六, 14 九月 2024 10:13:23 +0800
Subject: [PATCH] fix bug, 1 fix cuda oom, 2 fix choose a window size 400 that is [2, 0] (#2075)
---
funasr/models/whisper/model.py | 7 +++----
1 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/funasr/models/whisper/model.py b/funasr/models/whisper/model.py
index 4710b9c..791fddd 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
@@ -111,12 +112,10 @@
# 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)
- import pdb; pdb.set_trace()
-
results = []
result_i = {"key": key[0], "text": result.text}
--
Gitblit v1.9.1