diff --git a/CHANGELOG.md b/CHANGELOG.md index d7b3adcf..959785ed 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -6,17 +6,17 @@ #### Improved speaker assignment and counting -`pyannote/speaker-diarization-4.0` pretrained pipeline relies on VBx clustering instead of agglomerative hierarchical clustering (as suggested by [BUT Speech@FIT](https://speech.fit.vut.cz/) researchers [Petr Pálka](https://github.com/Selesnyan) and [Jiangyu Han](https://github.com/jyhan03)). +`pyannote/speaker-diarization-community-1` pretrained pipeline relies on VBx clustering instead of agglomerative hierarchical clustering (as suggested by [BUT Speech@FIT](https://speech.fit.vut.cz/) researchers [Petr Pálka](https://github.com/Selesnyan) and [Jiangyu Han](https://github.com/jyhan03)). #### *Exclusive* speaker diarization -`pyannote/speaker-diarization-4.0` pretrained pipeline returns a new *exclusive* speaker diarization, on top of the regular speaker diarization. +`pyannote/speaker-diarization-community-1` pretrained pipeline returns a new *exclusive* speaker diarization, on top of the regular speaker diarization. This is a feature which is [backported from our latest commercial model](https://www.pyannote.ai/blog/precision-2) that simplifies the reconciliation between fine-grained speaker diarization timestamps and (sometimes not so precise) transcription timestamps. ```python from pyannote.audio import Pipeline pipeline = Pipeline.from_pretrained( - "pyannote/speaker-diarization-4.0", token="huggingface-access-token") + "pyannote/speaker-diarization-community-1", token="huggingface-access-token") output = pipeline("/path/to/conversation.wav") print(output.speaker_diarization) # regular speaker diarization print(output.exclusive_speaker_diarization) # exclusive speaker diarization @@ -34,7 +34,7 @@ Change one line of code to use [pyannoteAI](https://docs.pyannote.ai) premium mo ```diff from pyannote.audio import Pipeline pipeline = Pipeline.from_pretrained( -- "pyannote/speaker-diarization-4.0", token="huggingface-access-token") +- "pyannote/speaker-diarization-community-1", token="huggingface-access-token") + "pyannote/speaker-diarization-precision-2, token="pyannoteAI-api-key") diarization = pipeline("/path/to/conversation.wav") ``` @@ -43,12 +43,12 @@ diarization = pipeline("/path/to/conversation.wav") Pipelines can now be stored alongside their internal models in the same repository, streamlining fully offline use. -1. Accept `pyannote/speaker-diarization-4.0` pipeline [user agreement](https://hf.co/pyannote/speaker-diarization-4.0) +1. Accept `pyannote/speaker-diarization-community-1` pipeline [user agreement](https://hf.co/pyannote/speaker-diarization-community-1) 2. Clone the pipeline repository from Huggingface (if prompted for a password, use a Huggingface access token with correct permissions) ```bash $ git lfs install - $ git clone https://hf.co/pyannote/speaker-diarization-4.0 /path/to/directory/pyannote-speaker-diarization-4.0 + $ git clone https://hf.co/pyannote/speaker-diarization-community-1 /path/to/directory/pyannote-speaker-diarization-community-1 ``` 3. Enjoy! @@ -56,7 +56,7 @@ Pipelines can now be stored alongside their internal models in the same reposito ```python # load pipeline from disk (works without internet connection) from pyannote.audio import Pipeline - pipeline = Pipeline.from_pretrained('/path/to/directory/pyannote-speaker-diarization-4.0') + pipeline = Pipeline.from_pretrained('/path/to/directory/pyannote-speaker-diarization-community-1') # run the pipeline locally on your computer diarization = pipeline("audio.wav") diff --git a/README.md b/README.md index db3757b8..9f5eae4e 100644 --- a/README.md +++ b/README.md @@ -21,7 +21,7 @@ Consider switching to [ `pyannote.audio` speaker diarization [pipeline v4.0](https://hf.co/pyannote/speaker-diarization-4.0) is expected to be much better than v3.1. +Out of the box, `pyannote.audio` speaker diarization [pipeline v4.0](https://hf.co/pyannote/speaker-diarization-community-1) is expected to be much better than v3.1. `pyannoteAI` premium models are even better (and also 2x faster). `labs` model is currently in private beta. -| Benchmark (last updated in 2025-08) |
v3.1
|
v4.0
|
API
|
labs
| +| Benchmark (last updated in 2025-08) |
v3.1
|
v4.0
|
API
|
labs
| | --------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------ | -------------------------------------------------| ------------------------------------------------ | --- | | [AISHELL-4](https://arxiv.org/abs/2104.03603) | 12.2 | 11.7 | 11.8 | 11.4 | | [AliMeeting](https://www.openslr.org/119/) (channel 1) | 24.5 | 20.3 | 16.3 | 15.2 | diff --git a/src/pyannote/audio/pipelines/speaker_diarization.py b/src/pyannote/audio/pipelines/speaker_diarization.py index 02453c67..e84533bb 100644 --- a/src/pyannote/audio/pipelines/speaker_diarization.py +++ b/src/pyannote/audio/pipelines/speaker_diarization.py @@ -194,17 +194,17 @@ class SpeakerDiarization(SpeakerDiarizationMixin, Pipeline): self, legacy: bool = False, segmentation: PipelineModel = { - "checkpoint": "pyannote/speaker-diarization-4.0", + "checkpoint": "pyannote/speaker-diarization-community-1", "subfolder": "segmentation", }, segmentation_step: float = 0.1, embedding: PipelineModel = { - "checkpoint": "pyannote/speaker-diarization-4.0", + "checkpoint": "pyannote/speaker-diarization-community-1", "subfolder": "embedding", }, embedding_exclude_overlap: bool = False, plda: PipelinePLDA = { - "checkpoint": "pyannote/speaker-diarization-4.0", + "checkpoint": "pyannote/speaker-diarization-community-1", "subfolder": "plda", }, clustering: str = "VBxClustering",