27 Commits
Author SHA1 Message Date
Hervé BREDIN d381589409 setup: switch to lightning (from pytorch_lightning) 2025-03-27 11:16:49 +01:00
Hervé BREDIN 93ad8b90ca BREAKING: switch to uv, namespace package, Python 3.10+, remove old CLI 2025-02-12 13:07:16 +01:00
Hervé BREDIN 2623cf1980 BREAKING(io): update Audio behavior for multi-channel audio
- BREAKING(io): channels are now 0-indexed (used to be 1-indexed)
- BREAKING(io): multi-channel audio is no longer downmixed to mono by default.
  You should update how `pyannote.audio.core.io.Audio` is instantiated:
  * replace `Audio()` by `Audio(mono="downmix")`;
  * replace `Audio(mono=True)` by `Audio(mono="downmix")`;
  * replace `Audio(mono=False)` by `Audio()`.
2023-04-21 15:14:56 +02:00
Hervé BREDIN 7121a73f08 feat: add support for label scope and missing classes (#1296)
- setup: switch to pyannote.database 5.0
- feat(task): add support for label scope in speaker diarization task (from pyannote.database 5.0)
- feat(task): add support for missing classes in multi-label segmentation task (from pyannote.database 5.0)
- improve(task): load metadata as tensors/arrays rather than pyannote.core instances
- BREAKING(task): rename `Segmentation` task to `SpeakerDiarization`
- BREAKING(task): remove support for variable chunk duration
2023-04-13 21:15:57 +02:00
pajowuandHervé BREDIN 9cc8b75777 BREAKING: add progress hook to pipelines
Co-authored-by: Hervé BREDIN <hbredin@users.noreply.github.com>
2022-12-14 22:50:53 +01:00
Hervé BREDIN c69e92df77 chore: remove multi-task support (#675) 2021-06-04 16:35:32 +02:00
Hervé BREDIN 0b9be93ee1 chore: replace load_from_checkpoint by Model.from_pretrained (#588)
* chore: import Model from package root
* chore: import Inference from package root
2021-01-20 16:39:08 +01:00
Hervé BREDIN c2b4247f22 chore: remove Diarization alias and MultiTaskSegmentation import (#586) 2021-01-20 16:13:55 +01:00
Hervé BREDIN 8a86468d44 chore: simplify the API even more 2021-01-14 14:33:17 +01:00
Hervé BREDIN 2827a9c129 feat: add progress_hook option to inference 2021-01-08 22:59:01 +01:00
Hervé BREDIN a08cbe0d26 feat: estimate upper bound for number of speakers
* test: add segmentation to example notebook
2021-01-08 16:12:10 +01:00
Hervé BREDIN 3f17dad793 feat: add segmentation and speaker embedding 2020-12-17 14:05:38 +01:00
Hervé BREDIN 035368428a feat: reproduce original segmentation results 2020-12-07 14:42:50 +01:00
Hervé BREDIN 06d3b6116c feat: add Hydra-based "pyannote-audio-train" CLI
* closes #476 (CLI)
* closes #514 (support for AutoLR)
* closes #485 (hyper-parameter optimization with Hydra’s Ax sweeper)
* closes #412 (log graph to Tensorboard)
2020-11-18 10:59:39 +01:00
Hervé BREDIN a9778a4d62 feat: add pipelines (and tutorials) 2020-11-16 15:07:44 +01:00
Harry Coultas Blum f872cb881c feat: listen to an AudioFile 2020-11-08 22:43:53 +01:00
Hervé BREDINandHarry Coultas Blum 936a417863 feat: add augmentation registry
Co-authored-by: Harry Coultas Blum <himion0@gmail.com>
2020-11-08 22:43:53 +01:00
Hervé BREDIN b859bb450c feat: add multi-task support (and multi-task segmentation) 2020-11-08 22:43:33 +01:00
Hervé BREDIN 682b0a861b feat: add model introspection and a bunch of tasks 2020-11-08 22:43:13 +01:00
Hervé BREDINandHarry 4753cd342a feat: add methods to freeze/unfreeze layers
Co-authored-by: Harry <himion0@gmail.com>
2020-11-08 22:43:13 +01:00
Hervé BREDINandHarry 8ab9db137c feat: add first implementation of Inference wrapper (#18)
* feat: add first implementation of Inference wrapper

* feat: add Inference.crop

* feat: add "window" option (sliding vs. whole)

* fix: fix inference notebook

* fix: add support for corner case when waveform is shorter than chunk duration

* fix: fix inference notebook

* fix: fix on_{save | load}_checkpoint logic

* Clearning inference notebook

Co-authored-by: Harry <himion0@gmail.com>
2020-11-08 22:43:13 +01:00
Hervé BREDINandHarry 0501cdbf07 feat: add generic load_from_checkpoint and initial version checking (#12) + versioning fixes
* feat: add generic load_from_checkpoint and initial version checking

Addresses #8

* fix: remove FIXME as torch.hub takes care of caching downloads

* setup: switch to pytorch-lightning 1.0.0

* feat: check pytorch-lightning version

* feat: add version compatibility check

* test: add basic "sharing" notebook

* fix: remove references to local files

* fix: fix versioning until setuptools_scm really supports semver

* Revert "fix: fix versioning until setuptools_scm really supports semver"

This reverts commit a16ec7721aae4d1da42d86da6da908570d88b951.

* Using package after setup

* fix: make setuptools_scm version semver-compliant

* fix: remove setuptools_scm_git_semver hack

* fix: skip pyannote.audio version checking

* Copying torchaudio version

* Stripping whitespace from version

Co-authored-by: Harry <himion0@gmail.com>
2020-11-08 22:43:13 +01:00
Harry Coultas BlumandHervé BREDIN 5512a1dba7 feat: add Debug database and notebook testing
Co-authored-by: Hervé BREDIN <hbredin@users.noreply.github.com>
2020-11-08 22:43:13 +01:00
Hervé Bredin 4097af53d4 feat: initial import 2020-11-08 22:42:37 +01:00
Hervé Bredin b7078bd2be git: remove useless notebooks 2018-06-29 15:59:05 +02:00
Hervé Bredin 1e1bfde386 chore: improve "auto_lr" heuristics 2018-06-29 15:59:05 +02:00
Hervé Bredin c5beb5ceec fix: fix loss value in progress bar 2018-06-29 15:59:05 +02:00