whisper/CHANGELOG.md

CHANGELOG

v20230918

• Add .pre-commit-config.yaml (#1528)
• fix doc of TextDecoder (#1526)
• Update model-card.md (#1643)
• word timing tweaks (#1559)
• Avoid rearranging all caches (#1483)
• Improve timestamp heuristics. (#1461)
• fix condition_on_previous_text (#1224)
• Fix numba depreceation notice (#1233)
• Updated README.md to provide more insight on BLEU and specific appendices (#1236)
• Avoid computing higher temperatures on no_speech segments (#1279)
• Dropped unused execute bit from mel_filters.npz. (#1254)
• Drop ffmpeg-python dependency and call ffmpeg directly. (#1242)
• Python 3.11 (#1171)
• Update decoding.py (#1219)
• Update decoding.py (#1155)
• Update README.md to reference tiktoken (#1105)
• Implement max line width and max line count, and make word highlighting optional (#1184)
• Squash long words at window and sentence boundaries. (#1114)
• python-publish.yml: bump actions version to fix node warning (#1211)
• Update tokenizer.py (#1163)

v20230314

• abort find_alignment on empty input (#1090)
• Fix truncated words list when the replacement character is decoded (#1089)
• fix github language stats getting dominated by jupyter notebook (#1076)
• Fix alignment between the segments and the list of words (#1087)
• Use tiktoken (#1044)

v20230308

• kwargs in decode() for convenience (#1061)
• fix all_tokens handling that caused more repetitions and discrepancy in JSON (#1060)
• fix typo in CHANGELOG.md

v20230307

• Fix the repetition/hallucination issue identified in #1046 (#1052)
• Use triton==2.0.0 (#1053)
• Install triton in x86_64 linux only (#1051)
• update setup.py to specify python >= 3.8 requirement

v20230306

• remove auxiliary audio extension (#1021)
• apply formatting with black, isort, and flake8 (#1038)
• word-level timestamps in transcribe() (#869)
• Decoding improvements (#1033)
• Update README.md (#894)
• Fix infinite loop caused by incorrect timestamp tokens prediction (#914)
• drop python 3.7 support (#889)

v20230124

• handle printing even if sys.stdout.buffer is not available (#887)
• Add TSV formatted output in transcript, using integer start/end time in milliseconds (#228)
• Added --output_format option (#333)
• Handle XDG_CACHE_HOME properly for download_root (#864)
• use stdout for printing transcription progress (#867)
• Fix bug where mm is mistakenly replaced with hmm in e.g. 20mm (#659)
• print '?' if a letter can't be encoded using the system default encoding (#859)

v20230117

The first versioned release available on PyPI