github/faster-whisper
1
1
Fork 0
mirror of https://github.com/SYSTRAN/faster-whisper.git synced 2026-01-09 21:48:08 -05:00
Code Issues Packages Projects Releases Wiki Activity

Fix list index out of range in word timestamps (#1157)

Browse Source
This commit is contained in:
Mahmoud Ashraf
2024-11-20 12:36:58 +02:00
committed by GitHub
parent bcd8ce0fc7
commit f830c6f241
1 changed files with 8 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
faster_whisper/transcribe.py
View File

@@ -1699,12 +1699,14 @@ class WhisperModel:
# array([0.]) # array([0.])
# This results in crashes when we lookup jump_times with float, like # This results in crashes when we lookup jump_times with float, like
# IndexError: arrays used as indices must be of integer (or boolean) type # IndexError: arrays used as indices must be of integer (or boolean) type
return [] return_list.append([])
continue
word_boundaries = np.pad( word_boundaries = np.pad(
np.cumsum([len(t) for t in word_tokens[:-1]]), (1, 0) np.cumsum([len(t) for t in word_tokens[:-1]]), (1, 0)
) )
if len(word_boundaries) <= 1: if len(word_boundaries) <= 1:
return [] return_list.append([])
continue
jumps = np.pad(np.diff(text_indices), (1, 0), constant_values=1).astype( jumps = np.pad(np.diff(text_indices), (1, 0), constant_values=1).astype(
bool bool
@@ -1884,11 +1886,9 @@ def merge_punctuations(alignment: List[dict], prepended: str, appended: str) ->
if previous["word"].startswith(" ") and previous["word"].strip() in prepended: if previous["word"].startswith(" ") and previous["word"].strip() in prepended:
# prepend it to the following word # prepend it to the following word
following["word"] = previous["word"] + following["word"] following["word"] = previous["word"] + following["word"]
if "tokens" in alignment[0].keys(): following["tokens"] = previous["tokens"] + following["tokens"]
following["tokens"] = previous["tokens"] + following["tokens"]
previous["tokens"] = []
previous["word"] = "" previous["word"] = ""
previous["tokens"] = []
else: else:
j = i j = i
i -= 1 i -= 1
@@ -1902,11 +1902,9 @@ def merge_punctuations(alignment: List[dict], prepended: str, appended: str) ->
if not previous["word"].endswith(" ") and following["word"] in appended: if not previous["word"].endswith(" ") and following["word"] in appended:
# append it to the previous word # append it to the previous word
previous["word"] = previous["word"] + following["word"] previous["word"] = previous["word"] + following["word"]
if "tokens" in alignment[0].keys(): previous["tokens"] = previous["tokens"] + following["tokens"]
previous["tokens"] = previous["tokens"] + following["tokens"]
following["tokens"] = []
following["word"] = "" following["word"] = ""
following["tokens"] = []
else: else:
i = j i = j
j += 1 j += 1