2025-09-13 19:20:10 +00:00
28 changed files with 2308 additions and 7823 deletions
--- a/.github/dependabot.yml
+++ b/.github/dependabot.yml
@ -1,13 +0,0 @@
 # Keep GitHub Actions up to date with GitHub's Dependabot...
 # https://docs.github.com/en/code-security/dependabot/working-with-dependabot/keeping-your-actions-up-to-date-with-dependabot
 # https://docs.github.com/en/code-security/dependabot/dependabot-version-updates/configuration-options-for-the-dependabot.yml-file#package-ecosystem
 version: 2
 updates:
  - package-ecosystem: github-actions
    directory: /
    groups:
      github-actions:
        patterns:
          - "*"  # Group all Actions updates into a single larger pull request
    schedule:
      interval: weekly
--- a/.github/workflows/python-publish.yml
+++ b/.github/workflows/python-publish.yml
@ -8,23 +8,23 @@ jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v2
    - uses: actions-ecosystem/action-regex-match@v2
      id: regex-match
      with:
        text: ${{ github.event.head_commit.message }}
        regex: '^Release ([^ ]+)'
    - name: Set up Python
-      uses: actions/setup-python@v5
+      uses: actions/setup-python@v2
      with:
-        python-version: '3.12'
+        python-version: '3.8'
    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
-        pip install setuptools wheel twine build
+        pip install setuptools wheel twine
    - name: Release
      if: ${{ steps.regex-match.outputs.match != '' }}
-      uses: softprops/action-gh-release@v2
+      uses: softprops/action-gh-release@v1
      with:
        tag_name: v${{ steps.regex-match.outputs.group1 }}
    - name: Build and publish
@ -33,5 +33,5 @@ jobs:
        TWINE_USERNAME: __token__
        TWINE_PASSWORD: ${{ secrets.PYPI_API_TOKEN }}
      run: |
-        python -m build --sdist
+        python setup.py sdist
        twine upload dist/*
--- a/.github/workflows/test.yml
+++ b/.github/workflows/test.yml
@ -6,74 +6,23 @@ on:
  pull_request:
    branches:
      - main
 jobs:
  pre-commit:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Fetch base branch
        run: git fetch origin ${{ github.base_ref }}
      - uses: actions/setup-python@v5
        with:
          python-version: "3.9"
          architecture: x64
      - name: Get pip cache dir
        id: pip-cache
        run: |
          echo "dir=$(pip cache dir)" >> $GITHUB_OUTPUT
      - name: pip/pre-commit cache
        uses: actions/cache@v4
        with:
          path: |
            ${{ steps.pip-cache.outputs.dir }}
            ~/.cache/pre-commit
          key: ${{ runner.os }}-pip-pre-commit-${{ hashFiles('**/.pre-commit-config.yaml') }}
          restore-keys: |
            ${{ runner.os }}-pip-pre-commit
      - name: pre-commit
        run: |
          pip install --upgrade pre-commit
          pre-commit install --install-hooks
          pre-commit run --all-files
  whisper-test:
    needs: pre-commit
    runs-on: ubuntu-latest
    strategy:
      fail-fast: false
      matrix:
-        include:
+        python-version: ['3.8', '3.9', '3.10']
-          - python-version: '3.8'
+        pytorch-version: [1.10.2, 1.13.1]
-            pytorch-version: 1.10.1
+        exclude:
            numpy-requirement: "'numpy<2'"
          - python-version: '3.8'
            pytorch-version: 1.13.1
            numpy-requirement: "'numpy<2'"
          - python-version: '3.8'
            pytorch-version: 2.0.1
            numpy-requirement: "'numpy<2'"
          - python-version: '3.9'
            pytorch-version: 2.1.2
            numpy-requirement: "'numpy<2'"
          - python-version: '3.10'
-            pytorch-version: 2.2.2
+            pytorch-version: 1.10.2
            numpy-requirement: "'numpy<2'"
          - python-version: '3.11'
            pytorch-version: 2.3.1
            numpy-requirement: "'numpy'"
          - python-version: '3.12'
            pytorch-version: 2.4.1
            numpy-requirement: "'numpy'"
          - python-version: '3.12'
            pytorch-version: 2.5.1
            numpy-requirement: "'numpy'"
          - python-version: '3.13'
            pytorch-version: 2.5.1
            numpy-requirement: "'numpy'"
    steps:
-      - uses: conda-incubator/setup-miniconda@v3
+      - uses: conda-incubator/setup-miniconda@v2
-      - run: conda install -n test ffmpeg python=${{ matrix.python-version }}
+      - run: conda install -n test ffmpeg python=${{ matrix.python-version }} pytorch=${{ matrix.pytorch-version }} cpuonly -c pytorch
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v2
      - run: echo "$CONDA/envs/test/bin" >> $GITHUB_PATH
-      - run: pip3 install .["dev"] ${{ matrix.numpy-requirement }} torch==${{ matrix.pytorch-version }}+cpu --index-url https://download.pytorch.org/whl/cpu --extra-index-url https://pypi.org/simple
+      - run: pip install .["dev"]
      - run: black --check --diff -t py38 --include '(\.pyi?)$' .
      - run: isort --check --diff .
      - run: flake8 --ignore E203,W503,W504,E501,E731,E741 .
      - run: pytest --durations=0 -vv -k 'not test_transcribe or test_transcribe[tiny] or test_transcribe[tiny.en]' -m 'not requires_cuda'
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@ -1,28 +0,0 @@
 repos:
  - repo: https://github.com/pre-commit/pre-commit-hooks
    rev: v5.0.0
    hooks:
      - id: check-json
      - id: end-of-file-fixer
        types: [file, python]
      - id: trailing-whitespace
        types: [file, python]
      - id: mixed-line-ending
      - id: check-added-large-files
        args: [--maxkb=4096]
  - repo: https://github.com/psf/black
    rev: 25.1.0
    hooks:
      - id: black
  - repo: https://github.com/pycqa/isort
    rev: 6.0.0
    hooks:
      - id: isort
        name: isort (python)
        args: ["--profile", "black", "-l", "88", "--trailing-comma", "--multi-line", "3"]
  - repo: https://github.com/pycqa/flake8.git
    rev: 7.1.1
    hooks:
      - id: flake8
        types: [python]
        args: ["--max-line-length", "88", "--ignore", "E203,E501,W503,W504"]
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@ -1,80 +1,5 @@
 # CHANGELOG
 ## [v20250625](https://github.com/openai/whisper/releases/tag/v20250625)
 * Fix: Update torch.load to use weights_only=True to prevent security w… ([#2451](https://github.com/openai/whisper/pull/2451))
 * Fix: Ensure DTW cost tensor is on the same device as input tensor ([#2561](https://github.com/openai/whisper/pull/2561))
 * docs: updated README to specify translation model limitation ([#2547](https://github.com/openai/whisper/pull/2547))
 * Fixed triton kernel update to support latest triton versions ([#2588](https://github.com/openai/whisper/pull/2588))
 * Fix: GitHub display errors for Jupyter notebooks ([#2589](https://github.com/openai/whisper/pull/2589))
 * Bump the github-actions group with 3 updates ([#2592](https://github.com/openai/whisper/pull/2592))
 * Keep GitHub Actions up to date with GitHub's Dependabot ([#2486](https://github.com/openai/whisper/pull/2486))
 * pre-commit: Upgrade black v25.1.0 and isort v6.0.0 ([#2514](https://github.com/openai/whisper/pull/2514))
 * GitHub Actions: Add Python 3.13 to the testing ([#2487](https://github.com/openai/whisper/pull/2487))
 * PEP 621: Migrate from setup.py to pyproject.toml ([#2435](https://github.com/openai/whisper/pull/2435))
 * pre-commit autoupdate && pre-commit run --all-files ([#2484](https://github.com/openai/whisper/pull/2484))
 * Upgrade GitHub Actions ([#2430](https://github.com/openai/whisper/pull/2430))
 * Bugfix: Illogical "Avoid computing higher temperatures on no_speech" ([#1903](https://github.com/openai/whisper/pull/1903))
 * Updating README and doc strings to reflect that n_mels can now be 128 ([#2049](https://github.com/openai/whisper/pull/2049))
 * fix typo data/README.md ([#2433](https://github.com/openai/whisper/pull/2433))
 * Update README.md ([#2379](https://github.com/openai/whisper/pull/2379))
 * Add option to carry initial_prompt with the sliding window ([#2343](https://github.com/openai/whisper/pull/2343))
 * more pytorch versions in tests ([#2408](https://github.com/openai/whisper/pull/2408))
 ## [v20240930](https://github.com/openai/whisper/releases/tag/v20240930)
 * allowing numpy 2 in tests ([#2362](https://github.com/openai/whisper/pull/2362))
 * large-v3-turbo model ([#2361](https://github.com/openai/whisper/pull/2361))
 * test on python/pytorch versions up to 3.12 and 2.4.1 ([#2360](https://github.com/openai/whisper/pull/2360))
 * using sdpa if available ([#2359](https://github.com/openai/whisper/pull/2359))
 ## [v20240927](https://github.com/openai/whisper/releases/tag/v20240927)
 * pinning numpy<2 in tests ([#2332](https://github.com/openai/whisper/pull/2332))
 * Relax triton requirements for compatibility with pytorch 2.4 and newer ([#2307](https://github.com/openai/whisper/pull/2307))
 * Skip silence around hallucinations ([#1838](https://github.com/openai/whisper/pull/1838))
 * Fix triton env marker ([#1887](https://github.com/openai/whisper/pull/1887))
 ## [v20231117](https://github.com/openai/whisper/releases/tag/v20231117)
 * Relax triton requirements for compatibility with pytorch 2.1 and newer ([#1802](https://github.com/openai/whisper/pull/1802))
 ## [v20231106](https://github.com/openai/whisper/releases/tag/v20231106)
 * large-v3 ([#1761](https://github.com/openai/whisper/pull/1761))
 ## [v20231105](https://github.com/openai/whisper/releases/tag/v20231105)
 * remove tiktoken pin ([#1759](https://github.com/openai/whisper/pull/1759))
 * docs: Disambiguation of the term "relative speed" in the README ([#1751](https://github.com/openai/whisper/pull/1751))
 * allow_pickle=False while loading of mel matrix IN audio.py ([#1511](https://github.com/openai/whisper/pull/1511))
 * handling transcribe exceptions. ([#1682](https://github.com/openai/whisper/pull/1682))
 * Add new option to generate subtitles by a specific number of words ([#1729](https://github.com/openai/whisper/pull/1729))
 * Fix exception when an audio file with no speech is provided ([#1396](https://github.com/openai/whisper/pull/1396))
 ## [v20230918](https://github.com/openai/whisper/releases/tag/v20230918)
 * Add .pre-commit-config.yaml ([#1528](https://github.com/openai/whisper/pull/1528))
 * fix doc of TextDecoder ([#1526](https://github.com/openai/whisper/pull/1526))
 * Update model-card.md ([#1643](https://github.com/openai/whisper/pull/1643))
 * word timing tweaks ([#1559](https://github.com/openai/whisper/pull/1559))
 * Avoid rearranging all caches ([#1483](https://github.com/openai/whisper/pull/1483))
 * Improve timestamp heuristics. ([#1461](https://github.com/openai/whisper/pull/1461))
 * fix condition_on_previous_text ([#1224](https://github.com/openai/whisper/pull/1224))
 * Fix numba depreceation notice ([#1233](https://github.com/openai/whisper/pull/1233))
 * Updated README.md to provide more insight on BLEU and specific appendices ([#1236](https://github.com/openai/whisper/pull/1236))
 * Avoid computing higher temperatures on no_speech segments ([#1279](https://github.com/openai/whisper/pull/1279))
 * Dropped unused execute bit from mel_filters.npz. ([#1254](https://github.com/openai/whisper/pull/1254))
 * Drop ffmpeg-python dependency and call ffmpeg directly. ([#1242](https://github.com/openai/whisper/pull/1242))
 * Python 3.11 ([#1171](https://github.com/openai/whisper/pull/1171))
 * Update decoding.py ([#1219](https://github.com/openai/whisper/pull/1219))
 * Update decoding.py ([#1155](https://github.com/openai/whisper/pull/1155))
 * Update README.md to reference tiktoken ([#1105](https://github.com/openai/whisper/pull/1105))
 * Implement max line width and max line count, and make word highlighting optional ([#1184](https://github.com/openai/whisper/pull/1184))
 * Squash long words at window and sentence boundaries. ([#1114](https://github.com/openai/whisper/pull/1114))
 * python-publish.yml: bump actions version to fix node warning ([#1211](https://github.com/openai/whisper/pull/1211))
 * Update tokenizer.py ([#1163](https://github.com/openai/whisper/pull/1163))
 ## [v20230314](https://github.com/openai/whisper/releases/tag/v20230314)
 * abort find_alignment on empty input ([#1090](https://github.com/openai/whisper/pull/1090))
--- a/README.md
+++ b/README.md
@ -17,7 +17,7 @@ A Transformer sequence-to-sequence model is trained on various speech processing
 ## Setup
-We used Python 3.9.9 and [PyTorch](https://pytorch.org/) 1.10.1 to train and test our models, but the codebase is expected to be compatible with Python 3.8-3.11 and recent PyTorch versions. The codebase also depends on a few Python packages, most notably [OpenAI's tiktoken](https://github.com/openai/tiktoken) for their fast tokenizer implementation. You can download and install (or update to) the latest release of Whisper with the following command:
+We used Python 3.9.9 and [PyTorch](https://pytorch.org/) 1.10.1 to train and test our models, but the codebase is expected to be compatible with Python 3.8-3.10 and recent PyTorch versions. The codebase also depends on a few Python packages, most notably [HuggingFace Transformers](https://huggingface.co/docs/transformers/index) for their fast tokenizer implementation and [ffmpeg-python](https://github.com/kkroening/ffmpeg-python) for reading audio files. You can download and install (or update to) the latest release of Whisper with the following command:
    pip install -U openai-whisper
@ -48,7 +48,7 @@ choco install ffmpeg
 scoop install ffmpeg
 ```
-You may need [`rust`](http://rust-lang.org) installed as well, in case [tiktoken](https://github.com/openai/tiktoken) does not provide a pre-built wheel for your platform. If you see installation errors during the `pip install` command above, please follow the [Getting started page](https://www.rust-lang.org/learn/get-started) to install Rust development environment. Additionally, you may need to configure the `PATH` environment variable, e.g. `export PATH="$HOME/.cargo/bin:$PATH"`. If the installation fails with `No module named 'setuptools_rust'`, you need to install `setuptools_rust`, e.g. by running:
+You may need [`rust`](http://rust-lang.org) installed as well, in case [tokenizers](https://pypi.org/project/tokenizers/) does not provide a pre-built wheel for your platform. If you see installation errors during the `pip install` command above, please follow the [Getting started page](https://www.rust-lang.org/learn/get-started) to install Rust development environment. Additionally, you may need to configure the `PATH` environment variable, e.g. `export PATH="$HOME/.cargo/bin:$PATH"`. If the installation fails with `No module named 'setuptools_rust'`, you need to install `setuptools_rust`, e.g. by running:
 ```bash
 pip install setuptools-rust
@ -57,55 +57,42 @@ pip install setuptools-rust
 ## Available models and languages
-There are six model sizes, four with English-only versions, offering speed and accuracy tradeoffs.
+There are five model sizes, four with English-only versions, offering speed and accuracy tradeoffs. Below are the names of the available models and their approximate memory requirements and relative speed. 
-Below are the names of the available models and their approximate memory requirements and inference speed relative to the large model.
+
 The relative speeds below are measured by transcribing English speech on a A100, and the real-world speed may vary significantly depending on many factors including the language, the speaking speed, and the available hardware.
 |  Size  | Parameters | English-only model | Multilingual model | Required VRAM | Relative speed |
 |:------:|:----------:|:------------------:|:------------------:|:-------------:|:--------------:|
-|  tiny  |    39 M    |     `tiny.en`      |       `tiny`       |     ~1 GB     |      ~10x      |
+|  tiny  |    39 M    |     `tiny.en`      |       `tiny`       |     ~1 GB     |      ~32x      |
-|  base  |    74 M    |     `base.en`      |       `base`       |     ~1 GB     |      ~7x       |
+|  base  |    74 M    |     `base.en`      |       `base`       |     ~1 GB     |      ~16x      |
-| small  |   244 M    |     `small.en`     |      `small`       |     ~2 GB     |      ~4x       |
+| small  |   244 M    |     `small.en`     |      `small`       |     ~2 GB     |      ~6x       |
 | medium |   769 M    |    `medium.en`     |      `medium`      |     ~5 GB     |      ~2x       |
 | large  |   1550 M   |        N/A         |      `large`       |    ~10 GB     |       1x       |
 | turbo  |   809 M    |        N/A         |      `turbo`       |     ~6 GB     |      ~8x       |
 The `.en` models for English-only applications tend to perform better, especially for the `tiny.en` and `base.en` models. We observed that the difference becomes less significant for the `small.en` and `medium.en` models.
 Additionally, the `turbo` model is an optimized version of `large-v3` that offers faster transcription speed with a minimal degradation in accuracy.
-Whisper's performance varies widely depending on the language. The figure below shows a performance breakdown of `large-v3` and `large-v2` models by language, using WERs (word error rates) or CER (character error rates, shown in *Italic*) evaluated on the Common Voice 15 and Fleurs datasets. Additional WER/CER metrics corresponding to the other models and datasets can be found in Appendix D.1, D.2, and D.4 of [the paper](https://arxiv.org/abs/2212.04356), as well as the BLEU (Bilingual Evaluation Understudy) scores for translation in Appendix D.3.
+Whisper's performance varies widely depending on the language. The figure below shows a WER (Word Error Rate) breakdown by languages of the Fleurs dataset using the `large-v2` model. More WER and BLEU scores corresponding to the other models and datasets can be found in Appendix D in [the paper](https://arxiv.org/abs/2212.04356). The smaller, the better.
 ![WER breakdown by language](https://raw.githubusercontent.com/openai/whisper/main/language-breakdown.svg)
 ![WER breakdown by language](https://github.com/openai/whisper/assets/266841/f4619d66-1058-4005-8f67-a9d811b77c62)
 ## Command-line usage
-The following command will transcribe speech in audio files, using the `turbo` model:
+The following command will transcribe speech in audio files, using the `medium` model:
-```bash
+    whisper audio.flac audio.mp3 audio.wav --model medium
 whisper audio.flac audio.mp3 audio.wav --model turbo
 ```
-The default setting (which selects the `turbo` model) works well for transcribing English. However, **the `turbo` model is not trained for translation tasks**. If you need to **translate non-English speech into English**, use one of the **multilingual models** (`tiny`, `base`, `small`, `medium`, `large`) instead of `turbo`. 
+The default setting (which selects the `small` model) works well for transcribing English. To transcribe an audio file containing non-English speech, you can specify the language using the `--language` option:
-For example, to transcribe an audio file containing non-English speech, you can specify the language:
+    whisper japanese.wav --language Japanese
-```bash
+Adding `--task translate` will translate the speech into English:
 whisper japanese.wav --language Japanese
 ```
-To **translate** speech into English, use:
+    whisper japanese.wav --language Japanese --task translate
 ```bash
 whisper japanese.wav --model medium --language Japanese --task translate
 ```
 > **Note:** The `turbo` model will return the original language even if `--task translate` is specified. Use `medium` or `large` for the best translation results.
 Run the following to view all available options:
-```bash
+    whisper --help
 whisper --help
 ```
 See [tokenizer.py](https://github.com/openai/whisper/blob/main/whisper/tokenizer.py) for the list of all available languages.
@ -117,7 +104,7 @@ Transcription can also be performed within Python:
 ```python
 import whisper
-model = whisper.load_model("turbo")
+model = whisper.load_model("base")
 result = model.transcribe("audio.mp3")
 print(result["text"])
 ```
@ -129,14 +116,14 @@ Below is an example usage of `whisper.detect_language()` and `whisper.decode()`
 ```python
 import whisper
-model = whisper.load_model("turbo")
+model = whisper.load_model("base")
 # load audio and pad/trim it to fit 30 seconds
 audio = whisper.load_audio("audio.mp3")
 audio = whisper.pad_or_trim(audio)
 # make log-Mel spectrogram and move to the same device as the model
-mel = whisper.log_mel_spectrogram(audio, n_mels=model.dims.n_mels).to(model.device)
+mel = whisper.log_mel_spectrogram(audio).to(model.device)
 # detect the spoken language
 _, probs = model.detect_language(mel)
--- a/data/README.md
+++ b/data/README.md
@ -45,7 +45,7 @@ We downloaded the [CHiME-5 dataset](https://spandh.dcs.shef.ac.uk//chime_challen
 ### AMI-IHM, AMI-SDM1
-We preprocessed the [AMI Corpus](https://groups.inf.ed.ac.uk/ami/corpus/overview.shtml) by following the stage 0 and 2 of the [s5b recipe](https://github.com/kaldi-asr/kaldi/tree/master/egs/ami/s5b).
+We preprocessed the [AMI Corpus](https://groups.inf.ed.ac.uk/ami/corpus/overview.shtml) by following the stage 0 ad 2 of the [s5b recipe](https://github.com/kaldi-asr/kaldi/tree/master/egs/ami/s5b).
 ## Long-form English-only datasets
--- a/language-breakdown.svg
+++ b/language-breakdown.svg
--- a/model-card.md
+++ b/model-card.md
@ -16,15 +16,13 @@ The Whisper models are trained for speech recognition and translation tasks, cap
 | small  |   244 M    |         ✓          |         ✓          |
 | medium |   769 M    |         ✓          |         ✓          |
 | large  |   1550 M   |                    |         ✓          |
 | turbo  |   798 M    |                    |         ✓          |
-In December 2022, we [released an improved large model named `large-v2`](https://github.com/openai/whisper/discussions/661), and `large-v3` in November 2023.
+In December 2022, we [released an improved large model named `large-v2`](https://github.com/openai/whisper/discussions/661).
 Additionally, we've added a `turbo` model in September 2024 which is optimized for inference speed.
 ### Release date
-September 2022 (original series), December 2022 (`large-v2`), November 2023 (`large-v3`), September 2024 (`large-v3-turbo`)
+September 2022 (original series) and December 2022 (`large-v2`)
 ### Model type
@ -39,7 +37,7 @@ Sequence-to-sequence ASR (automatic speech recognition) and speech translation m
 ### Evaluated Use
-The primary intended users of these models are AI researchers studying the robustness, generalization, capabilities, biases, and constraints of the current model. However, Whisper is also potentially quite useful as an ASR solution for developers, especially for English speech recognition. We recognize that once models are released, it is impossible to restrict access to only “intended” uses or to draw reasonable guidelines around what is or is not research.
+The primary intended users of these models are AI researchers studying robustness, generalization, capabilities, biases, and constraints of the current model. However, Whisper is also potentially quite useful as an ASR solution for developers, especially for English speech recognition. We recognize that once models are released, it is impossible to restrict access to only “intended” uses or to draw reasonable guidelines around what is or is not research.
 The models are primarily trained and evaluated on ASR and speech translation to English tasks. They show strong ASR results in ~10 languages. They may exhibit additional capabilities, particularly if fine-tuned on certain tasks like voice activity detection, speaker classification, or speaker diarization but have not been robustly evaluated in these areas. We strongly recommend that users perform robust evaluations of the models in a particular context and domain before deploying them.
@ -55,17 +53,17 @@ As discussed in [the accompanying paper](https://arxiv.org/abs/2212.04356), we s
 ## Performance and Limitations
-Our studies show that, over many existing ASR systems, the models exhibit improved robustness to accents, background noise, and technical language, as well as zero-shot translation from multiple languages into English; and that accuracy on speech recognition and translation is near the state-of-the-art level. 
+Our studies show that, over many existing ASR systems, the models exhibit improved robustness to accents, background noise, technical language, as well as zero shot translation from multiple languages into English; and that accuracy on speech recognition and translation is near the state-of-the-art level. 
 However, because the models are trained in a weakly supervised manner using large-scale noisy data, the predictions may include texts that are not actually spoken in the audio input (i.e. hallucination). We hypothesize that this happens because, given their general knowledge of language, the models combine trying to predict the next word in audio with trying to transcribe the audio itself.
-Our models perform unevenly across languages, and we observe lower accuracy on low-resource and/or low-discoverability languages or languages where we have less training data. The models also exhibit disparate performance on different accents and dialects of particular languages, which may include a higher word error rate across speakers of different genders, races, ages, or other demographic criteria. Our full evaluation results are presented in [the paper accompanying this release](https://arxiv.org/abs/2212.04356).
+Our models perform unevenly across languages, and we observe lower accuracy on low-resource and/or low-discoverability languages or languages where we have less training data. The models also exhibit disparate performance on different accents and dialects of particular languages, which may include higher word error rate across speakers of different genders, races, ages, or other demographic criteria. Our full evaluation results are presented in [the paper accompanying this release](https://arxiv.org/abs/2212.04356).
-In addition, the sequence-to-sequence architecture of the model makes it prone to generating repetitive texts, which can be mitigated to some degree by beam search and temperature scheduling but not perfectly. Further analysis of these limitations is provided in [the paper](https://arxiv.org/abs/2212.04356). It is likely that this behavior and hallucinations may be worse in lower-resource and/or lower-discoverability languages.
+In addition, the sequence-to-sequence architecture of the model makes it prone to generating repetitive texts, which can be mitigated to some degree by beam search and temperature scheduling but not perfectly. Further analysis on these limitations are provided in [the paper](https://arxiv.org/abs/2212.04356). It is likely that this behavior and hallucinations may be worse on lower-resource and/or lower-discoverability languages.
 ## Broader Implications
 We anticipate that Whisper models’ transcription capabilities may be used for improving accessibility tools. While Whisper models cannot be used for real-time transcription out of the box – their speed and size suggest that others may be able to build applications on top of them that allow for near-real-time speech recognition and translation. The real value of beneficial applications built on top of Whisper models suggests that the disparate performance of these models may have real economic implications.
-There are also potential dual-use concerns that come with releasing Whisper. While we hope the technology will be used primarily for beneficial purposes, making ASR technology more accessible could enable more actors to build capable surveillance technologies or scale up existing surveillance efforts, as the speed and accuracy allow for affordable automatic transcription and translation of large volumes of audio communication. Moreover, these models may have some capabilities to recognize specific individuals out of the box, which in turn presents safety concerns related both to dual use and disparate performance. In practice, we expect that the cost of transcription is not the limiting factor of scaling up surveillance projects.
+There are also potential dual use concerns that come with releasing Whisper. While we hope the technology will be used primarily for beneficial purposes, making ASR technology more accessible could enable more actors to build capable surveillance technologies or scale up existing surveillance efforts, as the speed and accuracy allow for affordable automatic transcription and translation of large volumes of audio communication. Moreover, these models may have some capabilities to recognize specific individuals out of the box, which in turn presents safety concerns related both to dual use and disparate performance. In practice, we expect that the cost of transcription is not the limiting factor of scaling up surveillance projects.
--- a/notebooks/LibriSpeech.ipynb
+++ b/notebooks/LibriSpeech.ipynb
@ -949,8 +949,7 @@
      "style": "IPY_MODEL_039b53f2702c4179af7e0548018d0588",
      "value": " 164/164 [05:08&lt;00:00,  1.86s/it]"
     }
-    },
+    }
    "state": {}
   }
  }
 },
--- a/notebooks/Multilingual_ASR.ipynb
+++ b/notebooks/Multilingual_ASR.ipynb
@ -4219,8 +4219,7 @@
            "_view_name": "StyleView",
            "description_width": ""
          }
-        },
+        }
        "state": {}
      }
    }
  },
--- a/pyproject.toml
+++ b/pyproject.toml
@ -1,50 +1,3 @@
 [build-system]
 build-backend = "setuptools.build_meta"
 requires = [ "setuptools>=61.2" ]
 [project]
 name = "openai-whisper"
 description = "Robust Speech Recognition via Large-Scale Weak Supervision"
 readme.content-type = "text/markdown"
 readme.file = "README.md"
 license = { text = "MIT" }
 authors = [ { name = "OpenAI" } ]
 requires-python = ">=3.8"
 classifiers = [
  "Programming Language :: Python :: 3 :: Only",
  "Programming Language :: Python :: 3.8",
  "Programming Language :: Python :: 3.9",
  "Programming Language :: Python :: 3.10",
  "Programming Language :: Python :: 3.11",
  "Programming Language :: Python :: 3.12",
  "Programming Language :: Python :: 3.13",
 ]
 dynamic = [ "version" ]
 dependencies = [
  "more-itertools",
  "numba",
  "numpy",
  "tiktoken",
  "torch",
  "tqdm",
  "triton>=2; (platform_machine=='x86_64' and sys_platform=='linux') or sys_platform=='linux2'",
 ]
 optional-dependencies.dev = [ "black", "flake8", "isort", "pytest", "scipy" ]
 urls = { Homepage = "https://github.com/openai/whisper" }
 scripts.whisper = "whisper.transcribe:cli"
 [tool.setuptools]
 py-modules = [ "whisper" ]
 include-package-data = true
 [tool.setuptools.dynamic]
 version = { attr = "whisper.version.__version__" }
 [tool.setuptools.packages.find]
 exclude = [ "tests*" ]
 namespaces = false
 [tool.black]
 [tool.isort]
@ -52,3 +5,4 @@ profile = "black"
 include_trailing_comma = true
 line_length = 88
 multi_line_output = 3
--- a/requirements.txt
+++ b/requirements.txt
@ -3,5 +3,5 @@ numpy
 torch
 tqdm
 more-itertools
-tiktoken
+tiktoken==0.3.1
-triton>=2.0.0;platform_machine=="x86_64" and sys_platform=="linux" or sys_platform=="linux2"
+ffmpeg-python==0.2.0
--- a/setup.py
+++ b/setup.py
@ -0,0 +1,43 @@
 import os
 import platform
 import sys
 import pkg_resources
 from setuptools import find_packages, setup
 def read_version(fname="whisper/version.py"):
    exec(compile(open(fname, encoding="utf-8").read(), fname, "exec"))
    return locals()["__version__"]
 requirements = []
 if sys.platform.startswith("linux") and platform.machine() == "x86_64":
    requirements.append("triton==2.0.0")
 setup(
    name="openai-whisper",
    py_modules=["whisper"],
    version=read_version(),
    description="Robust Speech Recognition via Large-Scale Weak Supervision",
    long_description=open("README.md", encoding="utf-8").read(),
    long_description_content_type="text/markdown",
    readme="README.md",
    python_requires=">=3.8",
    author="OpenAI",
    url="https://github.com/openai/whisper",
    license="MIT",
    packages=find_packages(exclude=["tests*"]),
    install_requires=requirements
    + [
        str(r)
        for r in pkg_resources.parse_requirements(
            open(os.path.join(os.path.dirname(__file__), "requirements.txt"))
        )
    ],
    entry_points={
        "console_scripts": ["whisper=whisper.transcribe:cli"],
    },
    include_package_data=True,
    extras_require={"dev": ["pytest", "scipy", "black", "flake8", "isort"]},
 )
--- a/tests/test_tokenizer.py
+++ b/tests/test_tokenizer.py
@ -1,17 +1,7 @@
 import pytest
 from whisper.tokenizer import get_tokenizer
-@pytest.mark.parametrize("multilingual", [True, False])
+def test_tokenizer():
 def test_tokenizer(multilingual):
    tokenizer = get_tokenizer(multilingual=False)
    assert tokenizer.sot in tokenizer.sot_sequence
    assert len(tokenizer.all_language_codes) == len(tokenizer.all_language_tokens)
    assert all(c < tokenizer.timestamp_begin for c in tokenizer.all_language_tokens)
 def test_multilingual_tokenizer():
    gpt2_tokenizer = get_tokenizer(multilingual=False)
    multilingual_tokenizer = get_tokenizer(multilingual=True)
@ -30,5 +20,5 @@ def test_split_on_unicode():
    tokens = [8404, 871, 287, 6, 246, 526, 3210, 20378]
    words, word_tokens = multilingual_tokenizer.split_tokens_on_unicode(tokens)
-    assert words == [" elle", " est", " l", "'", "\ufffd", "é", "rit", "oire"]
+    assert words == [" elle", " est", " l", "'", "<EFBFBD>", "é", "rit", "oire"]
    assert word_tokens == [[8404], [871], [287], [6], [246], [526], [3210], [20378]]
--- a/tests/test_transcribe.py
+++ b/tests/test_transcribe.py
@ -25,7 +25,7 @@ def test_transcribe(model_name: str):
    assert "your country" in transcription
    assert "do for you" in transcription
-    tokenizer = get_tokenizer(model.is_multilingual, num_languages=model.num_languages)
+    tokenizer = get_tokenizer(model.is_multilingual)
    all_tokens = [t for s in result["segments"] for t in s["tokens"]]
    assert tokenizer.decode(all_tokens) == result["text"]
    assert tokenizer.decode_with_timestamps(all_tokens).startswith("<|0.00|>")
--- a/whisper/init.py
+++ b/whisper/init.py
@ -25,10 +25,7 @@ _MODELS = {
    "medium": "https://openaipublic.azureedge.net/main/whisper/models/345ae4da62f9b3d59415adc60127b97c714f32e89e936602e85993674d08dcb1/medium.pt",
    "large-v1": "https://openaipublic.azureedge.net/main/whisper/models/e4b87e7e0bf463eb8e6956e646f1e277e901512310def2c24bf0e11bd3c28e9a/large-v1.pt",
    "large-v2": "https://openaipublic.azureedge.net/main/whisper/models/81f7c96c852ee8fc832187b0132e569d6c3065a3252ed18e56effd0b6a73e524/large-v2.pt",
-    "large-v3": "https://openaipublic.azureedge.net/main/whisper/models/e5b1a55b89c1367dacf97e3e19bfd829a01529dbfdeefa8caeb59b3f1b81dadb/large-v3.pt",
+    "large": "https://openaipublic.azureedge.net/main/whisper/models/81f7c96c852ee8fc832187b0132e569d6c3065a3252ed18e56effd0b6a73e524/large-v2.pt",
    "large": "https://openaipublic.azureedge.net/main/whisper/models/e5b1a55b89c1367dacf97e3e19bfd829a01529dbfdeefa8caeb59b3f1b81dadb/large-v3.pt",
    "large-v3-turbo": "https://openaipublic.azureedge.net/main/whisper/models/aff26ae408abcba5fbf8813c21e62b0941638c5f6eebfb145be0c9839262a19a/large-v3-turbo.pt",
    "turbo": "https://openaipublic.azureedge.net/main/whisper/models/aff26ae408abcba5fbf8813c21e62b0941638c5f6eebfb145be0c9839262a19a/large-v3-turbo.pt",
 }
 # base85-encoded (n_layers, n_heads) boolean arrays indicating the cross-attention heads that are
@ -44,10 +41,7 @@ _ALIGNMENT_HEADS = {
    "medium": b"ABzY8B0Jh+0{>%R7}kK1fFL7w6%<-Pf*t^=N)Qr&0RR9",
    "large-v1": b"ABzY8r9j$a0{>%R7#4sLmoOs{s)o3~84-RPdcFk!JR<kSfC2yj",
    "large-v2": b"ABzY8zd+h!0{>%R7=D0pU<_bnWW*tkYAhobTNnu$jnkEkXqp)j;w1Tzk)UH3X%SZd&fFZ2fC2yj",
-    "large-v3": b"ABzY8gWO1E0{>%R7(9S+Kn!D~%ngiGaR?*L!iJG9p-nab0JQ=-{D1-g00",
+    "large": b"ABzY8zd+h!0{>%R7=D0pU<_bnWW*tkYAhobTNnu$jnkEkXqp)j;w1Tzk)UH3X%SZd&fFZ2fC2yj",
    "large": b"ABzY8gWO1E0{>%R7(9S+Kn!D~%ngiGaR?*L!iJG9p-nab0JQ=-{D1-g00",
    "large-v3-turbo": b"ABzY8j^C+e0{>%RARaKHP%t(lGR*)0g!tONPyhe`",
    "turbo": b"ABzY8j^C+e0{>%RARaKHP%t(lGR*)0g!tONPyhe`",
 }
@ -147,8 +141,7 @@ def load_model(
    with (
        io.BytesIO(checkpoint_file) if in_memory else open(checkpoint_file, "rb")
    ) as fp:
-        kwargs = {"weights_only": True} if torch.__version__ >= "1.13" else {}
+        checkpoint = torch.load(fp, map_location=device)
        checkpoint = torch.load(fp, map_location=device, **kwargs)
    del checkpoint_file
    dims = ModelDimensions(**checkpoint["dims"])
--- a/whisper/assets/mel_filters.npz
+++ b/whisper/assets/mel_filters.npz
--- a/whisper/audio.py
+++ b/whisper/audio.py
@ -1,8 +1,8 @@
 import os
 from functools import lru_cache
 from subprocess import CalledProcessError, run
 from typing import Optional, Union
 import ffmpeg
 import numpy as np
 import torch
 import torch.nn.functional as F
@ -12,6 +12,7 @@ from .utils import exact_div
 # hard-coded audio hyperparameters
 SAMPLE_RATE = 16000
 N_FFT = 400
 N_MELS = 80
 HOP_LENGTH = 160
 CHUNK_LENGTH = 30
 N_SAMPLES = CHUNK_LENGTH * SAMPLE_RATE  # 480000 samples in a 30-second chunk
@ -38,25 +39,15 @@ def load_audio(file: str, sr: int = SAMPLE_RATE):
    -------
    A NumPy array containing the audio waveform, in float32 dtype.
    """
    # This launches a subprocess to decode audio while down-mixing
    # and resampling as necessary.  Requires the ffmpeg CLI in PATH.
    # fmt: off
    cmd = [
        "ffmpeg",
        "-nostdin",
        "-threads", "0",
        "-i", file,
        "-f", "s16le",
        "-ac", "1",
        "-acodec", "pcm_s16le",
        "-ar", str(sr),
        "-"
    ]
    # fmt: on
    try:
-        out = run(cmd, capture_output=True, check=True).stdout
+        # This launches a subprocess to decode audio while down-mixing and resampling as necessary.
-    except CalledProcessError as e:
+        # Requires the ffmpeg CLI and `ffmpeg-python` package to be installed.
        out, _ = (
            ffmpeg.input(file, threads=0)
            .output("-", format="s16le", acodec="pcm_s16le", ac=1, ar=sr)
            .run(cmd=["ffmpeg", "-nostdin"], capture_stdout=True, capture_stderr=True)
        )
    except ffmpeg.Error as e:
        raise RuntimeError(f"Failed to load audio: {e.stderr.decode()}") from e
    return np.frombuffer(out, np.int16).flatten().astype(np.float32) / 32768.0
@ -89,7 +80,7 @@ def pad_or_trim(array, length: int = N_SAMPLES, *, axis: int = -1):
@lru_cache(maxsize=None)
-def mel_filters(device, n_mels: int) -> torch.Tensor:
+def mel_filters(device, n_mels: int = N_MELS) -> torch.Tensor:
    """
    load the mel filterbank matrix for projecting STFT into a Mel spectrogram.
    Allows decoupling librosa dependency; saved using:
@ -97,19 +88,18 @@ def mel_filters(device, n_mels: int) -> torch.Tensor:
        np.savez_compressed(
            "mel_filters.npz",
            mel_80=librosa.filters.mel(sr=16000, n_fft=400, n_mels=80),
            mel_128=librosa.filters.mel(sr=16000, n_fft=400, n_mels=128),
        )
    """
-    assert n_mels in {80, 128}, f"Unsupported n_mels: {n_mels}"
+    assert n_mels == 80, f"Unsupported n_mels: {n_mels}"
-
+    with np.load(
-    filters_path = os.path.join(os.path.dirname(__file__), "assets", "mel_filters.npz")
+        os.path.join(os.path.dirname(__file__), "assets", "mel_filters.npz")
-    with np.load(filters_path, allow_pickle=False) as f:
+    ) as f:
        return torch.from_numpy(f[f"mel_{n_mels}"]).to(device)
 def log_mel_spectrogram(
    audio: Union[str, np.ndarray, torch.Tensor],
-    n_mels: int = 80,
+    n_mels: int = N_MELS,
    padding: int = 0,
    device: Optional[Union[str, torch.device]] = None,
 ):
@ -122,7 +112,7 @@ def log_mel_spectrogram(
        The path to audio or either a NumPy array or Tensor containing the audio waveform in 16 kHz
    n_mels: int
-        The number of Mel-frequency filters, only 80 and 128 are supported
+        The number of Mel-frequency filters, only 80 is supported
    padding: int
        Number of zero samples to pad to the right
@ -132,7 +122,7 @@ def log_mel_spectrogram(
    Returns
    -------
-    torch.Tensor, shape = (n_mels, n_frames)
+    torch.Tensor, shape = (80, n_frames)
        A Tensor that contains the Mel spectrogram
    """
    if not torch.is_tensor(audio):
--- a/whisper/decoding.py
+++ b/whisper/decoding.py
@ -32,9 +32,7 @@ def detect_language(
        list of dictionaries containing the probability distribution over all languages.
    """
    if tokenizer is None:
-        tokenizer = get_tokenizer(
+        tokenizer = get_tokenizer(model.is_multilingual)
            model.is_multilingual, num_languages=model.num_languages
        )
    if (
        tokenizer.language is None
        or tokenizer.language_token not in tokenizer.sot_sequence
@ -148,10 +146,6 @@ class PyTorchInference(Inference):
        self.kv_cache = {}
        self.hooks = []
        key_modules = [block.attn.key for block in self.model.decoder.blocks]
        value_modules = [block.attn.value for block in self.model.decoder.blocks]
        self.kv_modules = key_modules + value_modules
    def logits(self, tokens: Tensor, audio_features: Tensor) -> Tensor:
        if not self.kv_cache:
            self.kv_cache, self.hooks = self.model.install_kv_cache_hooks()
@ -170,10 +164,9 @@ class PyTorchInference(Inference):
        self.hooks = []
    def rearrange_kv_cache(self, source_indices):
-        if source_indices != list(range(len(source_indices))):
+        for module, tensor in self.kv_cache.items():
-            for module in self.kv_modules:
+            # update the key/value cache to contain the selected sequences
-                # update the key/value cache to contain the selected sequences
+            self.kv_cache[module] = tensor[source_indices].detach()
                self.kv_cache[module] = self.kv_cache[module][source_indices].detach()
 class SequenceRanker:
@ -476,12 +469,7 @@ class ApplyTimestampRules(LogitFilter):
            ]
            if timestamps.numel() > 0:
                # timestamps shouldn't decrease; forbid timestamp tokens smaller than the last
-                # also force each segment to have a nonzero length, to prevent infinite looping
+                logits[k, self.tokenizer.timestamp_begin : timestamps[-1]] = -np.inf
                if last_was_timestamp and not penultimate_was_timestamp:
                    timestamp_last = timestamps[-1]
                else:
                    timestamp_last = timestamps[-1] + 1
                logits[k, self.tokenizer.timestamp_begin : timestamp_last] = -np.inf
        if tokens.shape[1] == self.sample_begin:
            # suppress generating non-timestamp tokens at the beginning
@ -516,10 +504,7 @@ class DecodingTask:
        language = options.language or "en"
        tokenizer = get_tokenizer(
-            model.is_multilingual,
+            model.is_multilingual, language=language, task=options.task
            num_languages=model.num_languages,
            language=language,
            task=options.task,
        )
        self.tokenizer: Tokenizer = tokenizer
        self.options: DecodingOptions = self._verify_options(options)
@ -678,6 +663,7 @@ class DecodingTask:
        return languages, lang_probs
    def _main_loop(self, audio_features: Tensor, tokens: Tensor):
        assert audio_features.shape[0] == tokens.shape[0]
        n_batch = tokens.shape[0]
        sum_logprobs: Tensor = torch.zeros(n_batch, device=audio_features.device)
        no_speech_probs = [np.nan] * n_batch
@ -730,7 +716,8 @@ class DecodingTask:
                )
            ]
-        # repeat text tensors by the group size, for beam search or best-of-n sampling
+        # repeat the audio & text tensors by the group size, for beam search or best-of-n sampling
        audio_features = audio_features.repeat_interleave(self.n_group, dim=0)
        tokens = tokens.repeat_interleave(self.n_group, dim=0).to(audio_features.device)
        # call the main sampling loop
--- a/whisper/model.py
+++ b/whisper/model.py
@ -1,8 +1,7 @@
 import base64
 import gzip
 from contextlib import contextmanager
 from dataclasses import dataclass
-from typing import Dict, Iterable, Optional, Tuple
+from typing import Dict, Iterable, Optional
 import numpy as np
 import torch
@ -13,14 +12,6 @@ from .decoding import decode as decode_function
 from .decoding import detect_language as detect_language_function
 from .transcribe import transcribe as transcribe_function
 try:
    from torch.nn.functional import scaled_dot_product_attention
    SDPA_AVAILABLE = True
 except (ImportError, RuntimeError, OSError):
    scaled_dot_product_attention = None
    SDPA_AVAILABLE = False
@dataclass
 class ModelDimensions:
@ -68,19 +59,7 @@ def sinusoids(length, channels, max_timescale=10000):
    return torch.cat([torch.sin(scaled_time), torch.cos(scaled_time)], dim=1)
@contextmanager
 def disable_sdpa():
    prev_state = MultiHeadAttention.use_sdpa
    try:
        MultiHeadAttention.use_sdpa = False
        yield
    finally:
        MultiHeadAttention.use_sdpa = prev_state
 class MultiHeadAttention(nn.Module):
    use_sdpa = True
    def __init__(self, n_state: int, n_head: int):
        super().__init__()
        self.n_head = n_head
@ -113,30 +92,20 @@ class MultiHeadAttention(nn.Module):
    def qkv_attention(
        self, q: Tensor, k: Tensor, v: Tensor, mask: Optional[Tensor] = None
-    ) -> Tuple[torch.Tensor, Optional[torch.Tensor]]:
+    ):
        n_batch, n_ctx, n_state = q.shape
        scale = (n_state // self.n_head) ** -0.25
-        q = q.view(*q.shape[:2], self.n_head, -1).permute(0, 2, 1, 3)
+        q = q.view(*q.shape[:2], self.n_head, -1).permute(0, 2, 1, 3) * scale
-        k = k.view(*k.shape[:2], self.n_head, -1).permute(0, 2, 1, 3)
+        k = k.view(*k.shape[:2], self.n_head, -1).permute(0, 2, 3, 1) * scale
        v = v.view(*v.shape[:2], self.n_head, -1).permute(0, 2, 1, 3)
-        if SDPA_AVAILABLE and MultiHeadAttention.use_sdpa:
+        qk = q @ k
-            a = scaled_dot_product_attention(
+        if mask is not None:
-                q, k, v, is_causal=mask is not None and n_ctx > 1
+            qk = qk + mask[:n_ctx, :n_ctx]
-            )
+        qk = qk.float()
            out = a.permute(0, 2, 1, 3).flatten(start_dim=2)
            qk = None
        else:
            qk = (q * scale) @ (k * scale).transpose(-1, -2)
            if mask is not None:
                qk = qk + mask[:n_ctx, :n_ctx]
            qk = qk.float()
-            w = F.softmax(qk, dim=-1).to(q.dtype)
+        w = F.softmax(qk, dim=-1).to(q.dtype)
-            out = (w @ v).permute(0, 2, 1, 3).flatten(start_dim=2)
+        return (w @ v).permute(0, 2, 1, 3).flatten(start_dim=2), qk.detach()
            qk = qk.detach()
        return out, qk
 class ResidualAttentionBlock(nn.Module):
@ -228,7 +197,7 @@ class TextDecoder(nn.Module):
        """
        x : torch.LongTensor, shape = (batch_size, <= n_ctx)
            the text tokens
-        xa : torch.Tensor, shape = (batch_size, n_audio_ctx, n_audio_state)
+        xa : torch.Tensor, shape = (batch_size, n_mels, n_audio_ctx)
            the encoded audio features to be attended on
        """
        offset = next(iter(kv_cache.values())).shape[1] if kv_cache else 0
@ -267,8 +236,7 @@ class Whisper(nn.Module):
            self.dims.n_text_head,
            self.dims.n_text_layer,
        )
-        # use the last half among the decoder layers for time alignment by default;
+        # use the last half layers for alignment by default; see `set_alignment_heads()` below
        # to use a specific set of heads, see `set_alignment_heads()` below.
        all_heads = torch.zeros(
            self.dims.n_text_layer, self.dims.n_text_head, dtype=torch.bool
        )
@ -301,11 +269,7 @@ class Whisper(nn.Module):
    @property
    def is_multilingual(self):
-        return self.dims.n_vocab >= 51865
+        return self.dims.n_vocab == 51865
    @property
    def num_languages(self):
        return self.dims.n_vocab - 51765 - int(self.is_multilingual)
    def install_kv_cache_hooks(self, cache: Optional[dict] = None):
        """
--- a/whisper/normalizers/basic.py
+++ b/whisper/normalizers/basic.py
@ -30,19 +30,15 @@ def remove_symbols_and_diacritics(s: str, keep=""):
    and drop any diacritics (category 'Mn' and some manual mappings)
    """
    return "".join(
-        (
+        c
-            c
+        if c in keep
-            if c in keep
+        else ADDITIONAL_DIACRITICS[c]
-            else (
+        if c in ADDITIONAL_DIACRITICS
-                ADDITIONAL_DIACRITICS[c]
+        else ""
-                if c in ADDITIONAL_DIACRITICS
+        if unicodedata.category(c) == "Mn"
-                else (
+        else " "
-                    ""
+        if unicodedata.category(c)[0] in "MSP"
-                    if unicodedata.category(c) == "Mn"
+        else c
                    else " " if unicodedata.category(c)[0] in "MSP" else c
                )
            )
        )
        for c in unicodedata.normalize("NFKD", s)
    )
--- a/whisper/timing.py
+++ b/whisper/timing.py
@ -54,7 +54,7 @@ def median_filter(x: torch.Tensor, filter_width: int):
    return result
-@numba.jit(nopython=True)
+@numba.jit
 def backtrace(trace: np.ndarray):
    i = trace.shape[0] - 1
    j = trace.shape[1] - 1
@ -117,7 +117,7 @@ def dtw_cuda(x, BLOCK_SIZE=1024):
    x_skew = x_skew.T.contiguous()
    cost = torch.ones(N + M + 2, M + 2) * np.inf
    cost[0, 0] = 0
-    cost = cost.to(x.device)
+    cost = cost.cuda()
    trace = torch.zeros_like(cost, dtype=torch.int32)
    dtw_kernel[(1,)](
@ -191,9 +191,7 @@ def find_alignment(
        for i, block in enumerate(model.decoder.blocks)
    ]
-    from .model import disable_sdpa
+    with torch.no_grad():
    with torch.no_grad(), disable_sdpa():
        logits = model(mel.unsqueeze(0), tokens.unsqueeze(0))[0]
        sampled_logits = logits[len(tokenizer.sot_sequence) :, : tokenizer.eot]
        token_probs = sampled_logits.softmax(dim=-1)
@ -204,7 +202,7 @@ def find_alignment(
        hook.remove()
    # heads * tokens * frames
-    weights = torch.stack([QKs[_l][_h] for _l, _h in model.alignment_heads.indices().T])
+    weights = torch.stack([QKs[l][h] for l, h in model.alignment_heads.indices().T])
    weights = weights[:, :, : num_frames // 2]
    weights = (weights * qk_scale).softmax(dim=-1)
    std, mean = torch.std_mean(weights, dim=-2, keepdim=True, unbiased=False)
@ -216,13 +214,6 @@ def find_alignment(
    text_indices, time_indices = dtw(-matrix)
    words, word_tokens = tokenizer.split_to_word_tokens(text_tokens + [tokenizer.eot])
    if len(word_tokens) <= 1:
        # return on eot only
        # >>> np.pad([], (1, 0))
        # array([0.])
        # This results in crashes when we lookup jump_times with float, like
        # IndexError: arrays used as indices must be of integer (or boolean) type
        return []
    word_boundaries = np.pad(np.cumsum([len(t) for t in word_tokens[:-1]]), (1, 0))
    jumps = np.pad(np.diff(text_indices), (1, 0), constant_values=1).astype(bool)
@ -234,6 +225,19 @@ def find_alignment(
        for i, j in zip(word_boundaries[:-1], word_boundaries[1:])
    ]
    # hack: ensure the first and second word is not longer than twice the median word duration.
    # a better segmentation algorithm based on VAD should be able to replace this.
    word_durations = end_times - start_times
    word_durations = word_durations[word_durations.nonzero()]
    if len(word_durations) > 0:
        median_duration = np.median(word_durations)
        max_duration = median_duration * 2
        if len(word_durations) >= 2 and word_durations[1] > max_duration:
            boundary = max(end_times[2] / 2, end_times[2] - max_duration)
            end_times[0] = start_times[1] = boundary
        if len(word_durations) >= 1 and end_times[0] - start_times[0] > max_duration:
            start_times[0] = max(0, end_times[0] - max_duration)
    return [
        WordTiming(word, tokens, start, end, probability)
        for word, tokens, start, end, probability in zip(
@ -285,7 +289,6 @@ def add_word_timestamps(
    num_frames: int,
    prepend_punctuations: str = "\"'“¿([{-",
    append_punctuations: str = "\"'.。,，!！?？:：”)]}、",
    last_speech_timestamp: float,
    **kwargs,
 ):
    if len(segments) == 0:
@ -298,24 +301,6 @@ def add_word_timestamps(
    text_tokens = list(itertools.chain.from_iterable(text_tokens_per_segment))
    alignment = find_alignment(model, tokenizer, text_tokens, mel, num_frames, **kwargs)
    word_durations = np.array([t.end - t.start for t in alignment])
    word_durations = word_durations[word_durations.nonzero()]
    median_duration = np.median(word_durations) if len(word_durations) > 0 else 0.0
    median_duration = min(0.7, float(median_duration))
    max_duration = median_duration * 2
    # hack: truncate long words at sentence boundaries.
    # a better segmentation algorithm based on VAD should be able to replace this.
    if len(word_durations) > 0:
        sentence_end_marks = ".。!！?？"
        # ensure words at sentence boundaries are not longer than twice the median word duration.
        for i in range(1, len(alignment)):
            if alignment[i].end - alignment[i].start > max_duration:
                if alignment[i].word in sentence_end_marks:
                    alignment[i].end = alignment[i].start + max_duration
                elif alignment[i - 1].word in sentence_end_marks:
                    alignment[i].start = alignment[i].end - max_duration
    merge_punctuations(alignment, prepend_punctuations, append_punctuations)
    time_offset = segments[0]["seek"] * HOP_LENGTH / SAMPLE_RATE
@ -341,48 +326,9 @@ def add_word_timestamps(
            saved_tokens += len(timing.tokens)
            word_index += 1
        # hack: truncate long words at segment boundaries.
        # a better segmentation algorithm based on VAD should be able to replace this.
        if len(words) > 0:
-            # ensure the first and second word after a pause is not longer than
+            # adjust the segment-level timestamps based on the word-level timestamps
-            # twice the median word duration.
+            segment["start"] = words[0]["start"]
-            if words[0]["end"] - last_speech_timestamp > median_duration * 4 and (
+            segment["end"] = words[-1]["end"]
                words[0]["end"] - words[0]["start"] > max_duration
                or (
                    len(words) > 1
                    and words[1]["end"] - words[0]["start"] > max_duration * 2
                )
            ):
                if (
                    len(words) > 1
                    and words[1]["end"] - words[1]["start"] > max_duration
                ):
                    boundary = max(words[1]["end"] / 2, words[1]["end"] - max_duration)
                    words[0]["end"] = words[1]["start"] = boundary
                words[0]["start"] = max(0, words[0]["end"] - max_duration)
            # prefer the segment-level start timestamp if the first word is too long.
            if (
                segment["start"] < words[0]["end"]
                and segment["start"] - 0.5 > words[0]["start"]
            ):
                words[0]["start"] = max(
                    0, min(words[0]["end"] - median_duration, segment["start"])
                )
            else:
                segment["start"] = words[0]["start"]
            # prefer the segment-level end timestamp if the last word is too long.
            if (
                segment["end"] > words[-1]["start"]
                and segment["end"] + 0.5 < words[-1]["end"]
            ):
                words[-1]["end"] = max(
                    words[-1]["start"] + median_duration, segment["end"]
                )
            else:
                segment["end"] = words[-1]["end"]
            last_speech_timestamp = segment["end"]
        segment["words"] = words
--- a/whisper/tokenizer.py
+++ b/whisper/tokenizer.py
@ -6,6 +6,7 @@ from functools import cached_property, lru_cache
 from typing import Dict, List, Optional, Tuple
 import tiktoken
 from tiktoken_ext.openai_public import gpt2
 LANGUAGES = {
    "en": "english",
@ -107,7 +108,6 @@ LANGUAGES = {
    "ba": "bashkir",
    "jw": "javanese",
    "su": "sundanese",
    "yue": "cantonese",
 }
 # language code lookup by name, with a few language aliases
@ -124,7 +124,6 @@ TO_LANGUAGE_CODE = {
    "moldovan": "ro",
    "sinhalese": "si",
    "castilian": "es",
    "mandarin": "zh",
 }
@ -133,7 +132,6 @@ class Tokenizer:
    """A thin wrapper around `tiktoken` providing quick access to special tokens"""
    encoding: tiktoken.Encoding
    num_languages: int
    language: Optional[str] = None
    task: Optional[str] = None
    sot_sequence: Tuple[int] = ()
@ -148,7 +146,7 @@ class Tokenizer:
        translate: int = self.special_tokens["<|translate|>"]
        transcribe: int = self.special_tokens["<|transcribe|>"]
-        langs = tuple(LANGUAGES.keys())[: self.num_languages]
+        langs = tuple(LANGUAGES.keys())
        sot_sequence = [sot]
        if self.language is not None:
            sot_sequence.append(sot + 1 + langs.index(self.language))
@ -214,13 +212,10 @@ class Tokenizer:
        if self.language is None:
            raise ValueError("This tokenizer does not have language token configured")
-        return self.to_language_token(self.language)
+        if token := self.special_tokens.get(f"<|{self.language}|>", None):
    def to_language_token(self, language):
        if token := self.special_tokens.get(f"<|{language}|>", None):
            return token
-        raise KeyError(f"Language {language} not found in tokenizer.")
+        raise KeyError(f"Language {self.language} not found in tokenizer.")
    @cached_property
    def all_language_tokens(self) -> Tuple[int]:
@ -228,11 +223,11 @@ class Tokenizer:
        for token, token_id in self.special_tokens.items():
            if token.strip("<|>") in LANGUAGES:
                result.append(token_id)
-        return tuple(result)[: self.num_languages]
+        return tuple(result)
    @cached_property
    def all_language_codes(self) -> Tuple[str]:
-        return tuple(self.decode([_l]).strip("<|>") for _l in self.all_language_tokens)
+        return tuple(self.decode([l]).strip("<|>") for l in self.all_language_tokens)
    @cached_property
    def sot_sequence_including_notimestamps(self) -> Tuple[int]:
@ -275,7 +270,7 @@ class Tokenizer:
        return tuple(sorted(result))
    def split_to_word_tokens(self, tokens: List[int]):
-        if self.language in {"zh", "ja", "th", "lo", "my", "yue"}:
+        if self.language in {"zh", "ja", "th", "lo", "my"}:
            # These languages don't typically use spaces, so it is difficult to split words
            # without morpheme analysis. Here, we instead split words at any
            # position where the tokens are decoded as valid unicode points
@ -328,7 +323,7 @@ class Tokenizer:
@lru_cache(maxsize=None)
-def get_encoding(name: str = "gpt2", num_languages: int = 99):
+def get_encoding(name: str = "gpt2"):
    vocab_path = os.path.join(os.path.dirname(__file__), "assets", f"{name}.tiktoken")
    ranks = {
        base64.b64decode(token): int(rank)
@ -340,7 +335,7 @@ def get_encoding(name: str = "gpt2", num_languages: int = 99):
    specials = [
        "<|endoftext|>",
        "<|startoftranscript|>",
-        *[f"<|{lang}|>" for lang in list(LANGUAGES.keys())[:num_languages]],
+        *[f"<|{lang}|>" for lang in LANGUAGES.keys()],
        "<|translate|>",
        "<|transcribe|>",
        "<|startoflm|>",
@ -357,7 +352,7 @@ def get_encoding(name: str = "gpt2", num_languages: int = 99):
    return tiktoken.Encoding(
        name=os.path.basename(vocab_path),
        explicit_n_vocab=n_vocab,
-        pat_str=r"""'s|'t|'re|'ve|'m|'ll|'d| ?\p{L}+| ?\p{N}+| ?[^\s\p{L}\p{N}]+|\s+(?!\S)|\s+""",
+        pat_str=gpt2()["pat_str"],
        mergeable_ranks=ranks,
        special_tokens=special_tokens,
    )
@ -367,7 +362,6 @@ def get_encoding(name: str = "gpt2", num_languages: int = 99):
 def get_tokenizer(
    multilingual: bool,
    *,
    num_languages: int = 99,
    language: Optional[str] = None,
    task: Optional[str] = None,  # Literal["transcribe", "translate", None]
 ) -> Tokenizer:
@ -388,8 +382,6 @@ def get_tokenizer(
        language = None
        task = None
-    encoding = get_encoding(name=encoding_name, num_languages=num_languages)
+    encoding = get_encoding(name=encoding_name)
-    return Tokenizer(
+    return Tokenizer(encoding=encoding, language=language, task=task)
        encoding=encoding, num_languages=num_languages, language=language, task=task
    )
--- a/whisper/transcribe.py
+++ b/whisper/transcribe.py
@ -1,8 +1,7 @@
 import argparse
 import os
 import traceback
 import warnings
-from typing import TYPE_CHECKING, List, Optional, Tuple, Union
+from typing import TYPE_CHECKING, Optional, Tuple, Union
 import numpy as np
 import torch
@ -23,7 +22,6 @@ from .tokenizer import LANGUAGES, TO_LANGUAGE_CODE, get_tokenizer
 from .utils import (
    exact_div,
    format_timestamp,
    get_end,
    get_writer,
    make_safe,
    optional_float,
@ -46,12 +44,9 @@ def transcribe(
    no_speech_threshold: Optional[float] = 0.6,
    condition_on_previous_text: bool = True,
    initial_prompt: Optional[str] = None,
    carry_initial_prompt: bool = False,
    word_timestamps: bool = False,
    prepend_punctuations: str = "\"'“¿([{-",
    append_punctuations: str = "\"'.。,，!！?？:：”)]}、",
    clip_timestamps: Union[str, List[float]] = "0",
    hallucination_silence_threshold: Optional[float] = None,
    **decode_options,
 ):
    """
@ -103,22 +98,9 @@ def transcribe(
        "prompt-engineer" a context for transcription, e.g. custom vocabularies or proper nouns
        to make it more likely to predict those word correctly.
    carry_initial_prompt: bool
        If carry_initial_prompt is True, `initial_prompt` is prepended to the prompt of each internal
        `decode()` call. If there is not enough context space at the start of the prompt, it is
        left-sliced to make space.
    decode_options: dict
        Keyword arguments to construct `DecodingOptions` instances
    clip_timestamps: Union[str, List[float]]
        Comma-separated list start,end,start,end,... timestamps (in seconds) of clips to process.
        The last end timestamp defaults to the end of the file.
    hallucination_silence_threshold: Optional[float]
        When word_timestamps is True, skip silent periods longer than this threshold (in seconds)
        when a possible hallucination is detected
    Returns
    -------
    A dictionary containing the resulting text ("text") and segment-level details ("segments"), and
@ -136,9 +118,8 @@ def transcribe(
        decode_options["fp16"] = False
    # Pad 30-seconds of silence to the input audio, for slicing
-    mel = log_mel_spectrogram(audio, model.dims.n_mels, padding=N_SAMPLES)
+    mel = log_mel_spectrogram(audio, padding=N_SAMPLES)
    content_frames = mel.shape[-1] - N_FRAMES
    content_duration = float(content_frames * HOP_LENGTH / SAMPLE_RATE)
    if decode_options.get("language", None) is None:
        if not model.is_multilingual:
@ -158,25 +139,7 @@ def transcribe(
    language: str = decode_options["language"]
    task: str = decode_options.get("task", "transcribe")
-    tokenizer = get_tokenizer(
+    tokenizer = get_tokenizer(model.is_multilingual, language=language, task=task)
        model.is_multilingual,
        num_languages=model.num_languages,
        language=language,
        task=task,
    )
    if isinstance(clip_timestamps, str):
        clip_timestamps = [
            float(ts) for ts in (clip_timestamps.split(",") if clip_timestamps else [])
        ]
    seek_points: List[int] = [round(ts * FRAMES_PER_SECOND) for ts in clip_timestamps]
    if len(seek_points) == 0:
        seek_points.append(0)
    if len(seek_points) % 2 == 1:
        seek_points.append(content_frames)
    seek_clips: List[Tuple[int, int]] = list(zip(seek_points[::2], seek_points[1::2]))
    punctuation = "\"'“¿([{-\"'.。,，!！?？:：”)]}、"
    if word_timestamps and task == "translate":
        warnings.warn("Word-level timestamps on translations may not be reliable.")
@ -211,20 +174,13 @@ def transcribe(
                and decode_result.avg_logprob < logprob_threshold
            ):
                needs_fallback = True  # average log probability is too low
-            if (
+
                no_speech_threshold is not None
                and decode_result.no_speech_prob > no_speech_threshold
                and logprob_threshold is not None
                and decode_result.avg_logprob < logprob_threshold
            ):
                needs_fallback = False  # silence
            if not needs_fallback:
                break
        return decode_result
-    clip_idx = 0
+    seek = 0
    seek = seek_clips[clip_idx][0]
    input_stride = exact_div(
        N_FRAMES, model.dims.n_audio_ctx
    )  # mel frames per output token: 2
@ -235,11 +191,9 @@ def transcribe(
    all_segments = []
    prompt_reset_since = 0
    remaining_prompt_length = model.dims.n_text_ctx // 2 - 1
    if initial_prompt is not None:
        initial_prompt_tokens = tokenizer.encode(" " + initial_prompt.strip())
        all_tokens.extend(initial_prompt_tokens)
        remaining_prompt_length -= len(initial_prompt_tokens)
    else:
        initial_prompt_tokens = []
@ -264,34 +218,14 @@ def transcribe(
    with tqdm.tqdm(
        total=content_frames, unit="frames", disable=verbose is not False
    ) as pbar:
-        last_speech_timestamp = 0.0
+        while seek < content_frames:
        # NOTE: This loop is obscurely flattened to make the diff readable.
        # A later commit should turn this into a simpler nested loop.
        # for seek_clip_start, seek_clip_end in seek_clips:
        #     while seek < seek_clip_end
        while clip_idx < len(seek_clips):
            seek_clip_start, seek_clip_end = seek_clips[clip_idx]
            if seek < seek_clip_start:
                seek = seek_clip_start
            if seek >= seek_clip_end:
                clip_idx += 1
                if clip_idx < len(seek_clips):
                    seek = seek_clips[clip_idx][0]
                continue
            time_offset = float(seek * HOP_LENGTH / SAMPLE_RATE)
-            window_end_time = float((seek + N_FRAMES) * HOP_LENGTH / SAMPLE_RATE)
+            mel_segment = mel[:, seek : seek + N_FRAMES]
-            segment_size = min(N_FRAMES, content_frames - seek, seek_clip_end - seek)
+            segment_size = min(N_FRAMES, content_frames - seek)
            mel_segment = mel[:, seek : seek + segment_size]
            segment_duration = segment_size * HOP_LENGTH / SAMPLE_RATE
            mel_segment = pad_or_trim(mel_segment, N_FRAMES).to(model.device).to(dtype)
-            if carry_initial_prompt:
+            decode_options["prompt"] = all_tokens[prompt_reset_since:]
                nignored = max(len(initial_prompt_tokens), prompt_reset_since)
                remaining_prompt = all_tokens[nignored:][-remaining_prompt_length:]
                decode_options["prompt"] = initial_prompt_tokens + remaining_prompt
            else:
                decode_options["prompt"] = all_tokens[prompt_reset_since:]
            result: DecodingResult = decode_with_fallback(mel_segment)
            tokens = torch.tensor(result.tokens)
@ -312,30 +246,6 @@ def transcribe(
            previous_seek = seek
            current_segments = []
            # anomalous words are very long/short/improbable
            def word_anomaly_score(word: dict) -> float:
                probability = word.get("probability", 0.0)
                duration = word["end"] - word["start"]
                score = 0.0
                if probability < 0.15:
                    score += 1.0
                if duration < 0.133:
                    score += (0.133 - duration) * 15
                if duration > 2.0:
                    score += duration - 2.0
                return score
            def is_segment_anomaly(segment: Optional[dict]) -> bool:
                if segment is None or not segment["words"]:
                    return False
                words = [w for w in segment["words"] if w["word"] not in punctuation]
                words = words[:8]
                score = sum(word_anomaly_score(w) for w in words)
                return score >= 3 or score + 0.01 >= len(words)
            def next_words_segment(segments: List[dict]) -> Optional[dict]:
                return next((s for s in segments if s["words"]), None)
            timestamp_tokens: torch.Tensor = tokens.ge(tokenizer.timestamp_begin)
            single_timestamp_ending = timestamp_tokens[-2:].tolist() == [False, True]
@ -398,6 +308,10 @@ def transcribe(
                )
                seek += segment_size
            if not condition_on_previous_text or result.temperature > 0.5:
                # do not feed the prompt tokens if a high temperature was used
                prompt_reset_since = len(all_tokens)
            if word_timestamps:
                add_word_timestamps(
                    segments=current_segments,
@ -407,73 +321,16 @@ def transcribe(
                    num_frames=segment_size,
                    prepend_punctuations=prepend_punctuations,
                    append_punctuations=append_punctuations,
                    last_speech_timestamp=last_speech_timestamp,
                )
-
+                word_end_timestamps = [
-                if not single_timestamp_ending:
+                    w["end"] for s in current_segments for w in s["words"]
-                    last_word_end = get_end(current_segments)
+                ]
-                    if last_word_end is not None and last_word_end > time_offset:
+                if not single_timestamp_ending and len(word_end_timestamps) > 0:
-                        seek = round(last_word_end * FRAMES_PER_SECOND)
+                    seek_shift = round(
-
+                        (word_end_timestamps[-1] - time_offset) * FRAMES_PER_SECOND
-                # skip silence before possible hallucinations
+                    )
-                if hallucination_silence_threshold is not None:
+                    if seek_shift > 0:
-                    threshold = hallucination_silence_threshold
+                        seek = previous_seek + seek_shift
                    if not single_timestamp_ending:
                        last_word_end = get_end(current_segments)
                        if last_word_end is not None and last_word_end > time_offset:
                            remaining_duration = window_end_time - last_word_end
                            if remaining_duration > threshold:
                                seek = round(last_word_end * FRAMES_PER_SECOND)
                            else:
                                seek = previous_seek + segment_size
                    # if first segment might be a hallucination, skip leading silence
                    first_segment = next_words_segment(current_segments)
                    if first_segment is not None and is_segment_anomaly(first_segment):
                        gap = first_segment["start"] - time_offset
                        if gap > threshold:
                            seek = previous_seek + round(gap * FRAMES_PER_SECOND)
                            continue
                    # skip silence before any possible hallucination that is surrounded
                    # by silence or more hallucinations
                    hal_last_end = last_speech_timestamp
                    for si in range(len(current_segments)):
                        segment = current_segments[si]
                        if not segment["words"]:
                            continue
                        if is_segment_anomaly(segment):
                            next_segment = next_words_segment(
                                current_segments[si + 1 :]
                            )
                            if next_segment is not None:
                                hal_next_start = next_segment["words"][0]["start"]
                            else:
                                hal_next_start = time_offset + segment_duration
                            silence_before = (
                                segment["start"] - hal_last_end > threshold
                                or segment["start"] < threshold
                                or segment["start"] - time_offset < 2.0
                            )
                            silence_after = (
                                hal_next_start - segment["end"] > threshold
                                or is_segment_anomaly(next_segment)
                                or window_end_time - segment["end"] < 2.0
                            )
                            if silence_before and silence_after:
                                seek = round(
                                    max(time_offset + 1, segment["start"])
                                    * FRAMES_PER_SECOND
                                )
                                if content_duration - segment["end"] < threshold:
                                    seek = content_frames
                                current_segments[si:] = []
                                break
                        hal_last_end = segment["end"]
                last_word_end = get_end(current_segments)
                if last_word_end is not None:
                    last_speech_timestamp = last_word_end
            if verbose:
                for segment in current_segments:
@ -500,10 +357,6 @@ def transcribe(
                [token for segment in current_segments for token in segment["tokens"]]
            )
            if not condition_on_previous_text or result.temperature > 0.5:
                # do not feed the prompt tokens if a high temperature was used
                prompt_reset_since = len(all_tokens)
            # update progress bar
            pbar.update(min(content_frames, seek) - previous_seek)
@ -517,17 +370,10 @@ def transcribe(
 def cli():
    from . import available_models
    def valid_model_name(name):
        if name in available_models() or os.path.exists(name):
            return name
        raise ValueError(
            f"model should be one of {available_models()} or path to a model checkpoint"
        )
    # fmt: off
    parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
    parser.add_argument("audio", nargs="+", type=str, help="audio file(s) to transcribe")
-    parser.add_argument("--model", default="turbo", type=valid_model_name, help="name of the Whisper model to use")
+    parser.add_argument("--model", default="small", choices=available_models(), help="name of the Whisper model to use")
    parser.add_argument("--model_dir", type=str, default=None, help="the path to save model files; uses ~/.cache/whisper by default")
    parser.add_argument("--device", default="cuda" if torch.cuda.is_available() else "cpu", help="device to use for PyTorch inference")
    parser.add_argument("--output_dir", "-o", type=str, default=".", help="directory to save the outputs")
@ -545,8 +391,6 @@ def cli():
    parser.add_argument("--suppress_tokens", type=str, default="-1", help="comma-separated list of token ids to suppress during sampling; '-1' will suppress most special characters except common punctuations")
    parser.add_argument("--initial_prompt", type=str, default=None, help="optional text to provide as a prompt for the first window.")
    parser.add_argument("--carry_initial_prompt", type=str2bool, default=False, help="if True, prepend initial_prompt to every internal decode() call. May reduce the effectiveness of condition_on_previous_text")
    parser.add_argument("--condition_on_previous_text", type=str2bool, default=True, help="if True, provide the previous output of the model as a prompt for the next window; disabling may make the text inconsistent across windows, but the model becomes less prone to getting stuck in a failure loop")
    parser.add_argument("--fp16", type=str2bool, default=True, help="whether to perform inference in fp16; True by default")
@ -557,13 +401,7 @@ def cli():
    parser.add_argument("--word_timestamps", type=str2bool, default=False, help="(experimental) extract word-level timestamps and refine the results based on them")
    parser.add_argument("--prepend_punctuations", type=str, default="\"\'“¿([{-", help="if word_timestamps is True, merge these punctuation symbols with the next word")
    parser.add_argument("--append_punctuations", type=str, default="\"\'.。,，!！?？:：”)]}、", help="if word_timestamps is True, merge these punctuation symbols with the previous word")
    parser.add_argument("--highlight_words", type=str2bool, default=False, help="(requires --word_timestamps True) underline each word as it is spoken in srt and vtt")
    parser.add_argument("--max_line_width", type=optional_int, default=None, help="(requires --word_timestamps True) the maximum number of characters in a line before breaking the line")
    parser.add_argument("--max_line_count", type=optional_int, default=None, help="(requires --word_timestamps True) the maximum number of lines in a segment")
    parser.add_argument("--max_words_per_line", type=optional_int, default=None, help="(requires --word_timestamps True, no effect with --max_line_width) the maximum number of words in a segment")
    parser.add_argument("--threads", type=optional_int, default=0, help="number of threads used by torch for CPU inference; supercedes MKL_NUM_THREADS/OMP_NUM_THREADS")
    parser.add_argument("--clip_timestamps", type=str, default="0", help="comma-separated list start,end,start,end,... timestamps (in seconds) of clips to process, where the last end timestamp defaults to the end of the file")
    parser.add_argument("--hallucination_silence_threshold", type=optional_float, help="(requires --word_timestamps True) skip silent periods longer than this threshold (in seconds) when a possible hallucination is detected")
    # fmt: on
    args = parser.parse_args().__dict__
@ -595,28 +433,9 @@ def cli():
    model = load_model(model_name, device=device, download_root=model_dir)
    writer = get_writer(output_format, output_dir)
    word_options = [
        "highlight_words",
        "max_line_count",
        "max_line_width",
        "max_words_per_line",
    ]
    if not args["word_timestamps"]:
        for option in word_options:
            if args[option]:
                parser.error(f"--{option} requires --word_timestamps True")
    if args["max_line_count"] and not args["max_line_width"]:
        warnings.warn("--max_line_count has no effect without --max_line_width")
    if args["max_words_per_line"] and args["max_line_width"]:
        warnings.warn("--max_words_per_line has no effect with --max_line_width")
    writer_args = {arg: args.pop(arg) for arg in word_options}
    for audio_path in args.pop("audio"):
-        try:
+        result = transcribe(model, audio_path, temperature=temperature, **args)
-            result = transcribe(model, audio_path, temperature=temperature, **args)
+        writer(result, audio_path)
            writer(result, audio_path, **writer_args)
        except Exception as e:
            traceback.print_exc()
            print(f"Skipping {audio_path} due to {type(e).__name__}: {str(e)}")
 if __name__ == "__main__":
--- a/whisper/triton_ops.py
+++ b/whisper/triton_ops.py
@ -60,7 +60,7 @@ def median_kernel(filter_width: int):
        tl.store(y_ptr + offsets, MIDDLE_ROW_HERE, mask=mask)  # noqa: F821
    kernel = triton.JITFunction(kernel.fn)
-    new_kernel = kernel.src.replace(
+    kernel.src = kernel.src.replace(
        "    LOAD_ALL_ROWS_HERE",
        "\n".join(
            [
@ -69,8 +69,7 @@ def median_kernel(filter_width: int):
            ]
        ),
    )
-
+    kernel.src = kernel.src.replace(
    new_kernel = new_kernel.replace(
        "    BUBBLESORT_HERE",
        "\n\n".join(
            [
@ -91,14 +90,7 @@ def median_kernel(filter_width: int):
            ]
        ),
    )
-
+    kernel.src = kernel.src.replace("MIDDLE_ROW_HERE", f"row{filter_width // 2}")
    new_kernel = new_kernel.replace("MIDDLE_ROW_HERE", f"row{filter_width // 2}")
    if hasattr(kernel, "_unsafe_update_src") is True:
        kernel._unsafe_update_src(new_kernel)
        kernel.hash = None
    else:
        kernel.src = new_kernel
    return kernel
--- a/whisper/utils.py
+++ b/whisper/utils.py
@ -1,9 +1,8 @@
 import json
 import os
 import re
 import sys
 import zlib
-from typing import Callable, List, Optional, TextIO
+from typing import Callable, TextIO
 system_encoding = sys.getdefaultencoding()
@ -68,29 +67,13 @@ def format_timestamp(
    )
 def get_start(segments: List[dict]) -> Optional[float]:
    return next(
        (w["start"] for s in segments for w in s["words"]),
        segments[0]["start"] if segments else None,
    )
 def get_end(segments: List[dict]) -> Optional[float]:
    return next(
        (w["end"] for s in reversed(segments) for w in reversed(s["words"])),
        segments[-1]["end"] if segments else None,
    )
 class ResultWriter:
    extension: str
    def __init__(self, output_dir: str):
        self.output_dir = output_dir
-    def __call__(
+    def __call__(self, result: dict, audio_path: str):
        self, result: dict, audio_path: str, options: Optional[dict] = None, **kwargs
    ):
        audio_basename = os.path.basename(audio_path)
        audio_basename = os.path.splitext(audio_basename)[0]
        output_path = os.path.join(
@ -98,20 +81,16 @@ class ResultWriter:
        )
        with open(output_path, "w", encoding="utf-8") as f:
-            self.write_result(result, file=f, options=options, **kwargs)
+            self.write_result(result, file=f)
-    def write_result(
+    def write_result(self, result: dict, file: TextIO):
        self, result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
    ):
        raise NotImplementedError
 class WriteTXT(ResultWriter):
    extension: str = "txt"
-    def write_result(
+    def write_result(self, result: dict, file: TextIO):
        self, result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
    ):
        for segment in result["segments"]:
            print(segment["text"].strip(), file=file, flush=True)
@ -120,111 +99,33 @@ class SubtitlesWriter(ResultWriter):
    always_include_hours: bool
    decimal_marker: str
-    def iterate_result(
+    def iterate_result(self, result: dict):
-        self,
+        for segment in result["segments"]:
-        result: dict,
+            segment_start = self.format_timestamp(segment["start"])
-        options: Optional[dict] = None,
+            segment_end = self.format_timestamp(segment["end"])
-        *,
+            segment_text = segment["text"].strip().replace("-->", "->")
        max_line_width: Optional[int] = None,
        max_line_count: Optional[int] = None,
        highlight_words: bool = False,
        max_words_per_line: Optional[int] = None,
    ):
        options = options or {}
        max_line_width = max_line_width or options.get("max_line_width")
        max_line_count = max_line_count or options.get("max_line_count")
        highlight_words = highlight_words or options.get("highlight_words", False)
        max_words_per_line = max_words_per_line or options.get("max_words_per_line")
        preserve_segments = max_line_count is None or max_line_width is None
        max_line_width = max_line_width or 1000
        max_words_per_line = max_words_per_line or 1000
-        def iterate_subtitles():
+            if word_timings := segment.get("words", None):
-            line_len = 0
+                all_words = [timing["word"] for timing in word_timings]
-            line_count = 1
+                all_words[0] = all_words[0].strip()  # remove the leading space, if any
-            # the next subtitle to yield (a list of word timings with whitespace)
+                last = segment_start
-            subtitle: List[dict] = []
+                for i, this_word in enumerate(word_timings):
-            last: float = get_start(result["segments"]) or 0.0
+                    start = self.format_timestamp(this_word["start"])
-            for segment in result["segments"]:
+                    end = self.format_timestamp(this_word["end"])
-                chunk_index = 0
+                    if last != start:
-                words_count = max_words_per_line
+                        yield last, start, segment_text
                while chunk_index < len(segment["words"]):
                    remaining_words = len(segment["words"]) - chunk_index
                    if max_words_per_line > len(segment["words"]) - chunk_index:
                        words_count = remaining_words
                    for i, original_timing in enumerate(
                        segment["words"][chunk_index : chunk_index + words_count]
                    ):
                        timing = original_timing.copy()
                        long_pause = (
                            not preserve_segments and timing["start"] - last > 3.0
                        )
                        has_room = line_len + len(timing["word"]) <= max_line_width
                        seg_break = i == 0 and len(subtitle) > 0 and preserve_segments
                        if (
                            line_len > 0
                            and has_room
                            and not long_pause
                            and not seg_break
                        ):
                            # line continuation
                            line_len += len(timing["word"])
                        else:
                            # new line
                            timing["word"] = timing["word"].strip()
                            if (
                                len(subtitle) > 0
                                and max_line_count is not None
                                and (long_pause or line_count >= max_line_count)
                                or seg_break
                            ):
                                # subtitle break
                                yield subtitle
                                subtitle = []
                                line_count = 1
                            elif line_len > 0:
                                # line break
                                line_count += 1
                                timing["word"] = "\n" + timing["word"]
                            line_len = len(timing["word"].strip())
                        subtitle.append(timing)
                        last = timing["start"]
                    chunk_index += max_words_per_line
            if len(subtitle) > 0:
                yield subtitle
-        if len(result["segments"]) > 0 and "words" in result["segments"][0]:
+                    yield start, end, "".join(
-            for subtitle in iterate_subtitles():
+                        [
-                subtitle_start = self.format_timestamp(subtitle[0]["start"])
+                            f"<u>{word}</u>" if j == i else word
-                subtitle_end = self.format_timestamp(subtitle[-1]["end"])
+                            for j, word in enumerate(all_words)
-                subtitle_text = "".join([word["word"] for word in subtitle])
+                        ]
-                if highlight_words:
+                    )
-                    last = subtitle_start
+                    last = end
                    all_words = [timing["word"] for timing in subtitle]
                    for i, this_word in enumerate(subtitle):
                        start = self.format_timestamp(this_word["start"])
                        end = self.format_timestamp(this_word["end"])
                        if last != start:
                            yield last, start, subtitle_text
-                        yield start, end, "".join(
+                if last != segment_end:
-                            [
+                    yield last, segment_end, segment_text
-                                (
+            else:
                                    re.sub(r"^(\s*)(.*)$", r"\1<u>\2</u>", word)
                                    if j == i
                                    else word
                                )
                                for j, word in enumerate(all_words)
                            ]
                        )
                        last = end
                else:
                    yield subtitle_start, subtitle_end, subtitle_text
        else:
            for segment in result["segments"]:
                segment_start = self.format_timestamp(segment["start"])
                segment_end = self.format_timestamp(segment["end"])
                segment_text = segment["text"].strip().replace("-->", "->")
                yield segment_start, segment_end, segment_text
    def format_timestamp(self, seconds: float):
@ -240,11 +141,9 @@ class WriteVTT(SubtitlesWriter):
    always_include_hours: bool = False
    decimal_marker: str = "."
-    def write_result(
+    def write_result(self, result: dict, file: TextIO):
        self, result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
    ):
        print("WEBVTT\n", file=file)
-        for start, end, text in self.iterate_result(result, options, **kwargs):
+        for start, end, text in self.iterate_result(result):
            print(f"{start} --> {end}\n{text}\n", file=file, flush=True)
@ -253,12 +152,8 @@ class WriteSRT(SubtitlesWriter):
    always_include_hours: bool = True
    decimal_marker: str = ","
-    def write_result(
+    def write_result(self, result: dict, file: TextIO):
-        self, result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
+        for i, (start, end, text) in enumerate(self.iterate_result(result), start=1):
    ):
        for i, (start, end, text) in enumerate(
            self.iterate_result(result, options, **kwargs), start=1
        ):
            print(f"{i}\n{start} --> {end}\n{text}\n", file=file, flush=True)
@ -274,9 +169,7 @@ class WriteTSV(ResultWriter):
    extension: str = "tsv"
-    def write_result(
+    def write_result(self, result: dict, file: TextIO):
        self, result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
    ):
        print("start", "end", "text", sep="\t", file=file)
        for segment in result["segments"]:
            print(round(1000 * segment["start"]), file=file, end="\t")
@ -287,15 +180,11 @@ class WriteTSV(ResultWriter):
 class WriteJSON(ResultWriter):
    extension: str = "json"
-    def write_result(
+    def write_result(self, result: dict, file: TextIO):
        self, result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
    ):
        json.dump(result, file)
-def get_writer(
+def get_writer(output_format: str, output_dir: str) -> Callable[[dict, TextIO], None]:
    output_format: str, output_dir: str
 ) -> Callable[[dict, TextIO, dict], None]:
    writers = {
        "txt": WriteTXT,
        "vtt": WriteVTT,
@ -307,11 +196,9 @@ def get_writer(
    if output_format == "all":
        all_writers = [writer(output_dir) for writer in writers.values()]
-        def write_all(
+        def write_all(result: dict, file: TextIO):
            result: dict, file: TextIO, options: Optional[dict] = None, **kwargs
        ):
            for writer in all_writers:
-                writer(result, file, options, **kwargs)
+                writer(result, file)
        return write_all
--- a/whisper/version.py
+++ b/whisper/version.py
@ -1 +1 @@
-__version__ = "20250625"
+__version__ = "20230314"
`@ -1 +1 @@`
	`__version__ = "20250625"`	`__version__ = "20230314"`