From 0c7f32f55c36776f780fcb48ec35a2f4bca84ca0 Mon Sep 17 00:00:00 2001 From: Simon Sorg Date: Fri, 3 Nov 2023 19:47:00 +0100 Subject: [PATCH] no align based on space --- whisperx/alignment.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/whisperx/alignment.py b/whisperx/alignment.py index 68465f9..340e4a9 100644 --- a/whisperx/alignment.py +++ b/whisperx/alignment.py @@ -283,7 +283,8 @@ def align( sentence_text = text[sstart:send] sentence_start = curr_chars["start"].min() - sentence_end = curr_chars["end"].max() + end_chars = curr_chars[curr_chars["char"] != ' '] + sentence_end = end_chars["end"].max() sentence_words = [] for word_idx in curr_chars["word-idx"].unique():