Loading pkgs/development/python-modules/whisperx/default.nix +10 −18 Original line number Diff line number Diff line Loading @@ -13,6 +13,7 @@ huggingface-hub, nltk, numpy, omegaconf, pandas, pyannote-audio, torch, Loading @@ -36,16 +37,16 @@ let }; }; in buildPythonPackage rec { buildPythonPackage (finalAttrs: { pname = "whisperx"; version = "3.7.6"; version = "3.8.1"; pyproject = true; src = fetchFromGitHub { owner = "m-bain"; repo = "whisperX"; tag = "v${version}"; hash = "sha256-ZHPGQP5HIuFafHGS6ykiSNtHY6QHh0o8DUE2lV41lUI="; tag = "v${finalAttrs.version}"; hash = "sha256-2HjQtb8k3px0kqXowKtCXkiG2GuKLCuCtDOPYYa/tbc="; }; # As `makeWrapperArgs` does not apply to the module, and whisperx depends on `ffmpeg`, Loading @@ -60,9 +61,6 @@ buildPythonPackage rec { build-system = [ setuptools ]; pythonRelaxDeps = [ "numpy" "pandas" "pyannote-audio" "torch" "torchaudio" ]; Loading @@ -72,6 +70,7 @@ buildPythonPackage rec { huggingface-hub nltk numpy omegaconf pandas pyannote-audio torch Loading @@ -85,21 +84,14 @@ buildPythonPackage rec { # No tests in repository doCheck = false; pythonImportsCheck = [ "whisperx" ]; meta = { mainProgram = "whisperx"; description = "Automatic Speech Recognition with Word-level Timestamps (& Diarization)"; homepage = "https://github.com/m-bain/whisperX"; changelog = "https://github.com/m-bain/whisperX/releases/tag/${src.tag}"; changelog = "https://github.com/m-bain/whisperX/releases/tag/${finalAttrs.src.tag}"; license = lib.licenses.bsd2; maintainers = [ lib.maintainers.bengsparks ]; # nixpkgs has `pyannote-audio` >= 4.0.0, but `whisperx`'s `pyproject.toml` specifies <4.0.0. # # See https://github.com/m-bain/whisperX/issues/1240 for a serious discussion, # and a potential upgrade in https://github.com/m-bain/whisperX/pull/1243. # Alternatively read https://github.com/m-bain/whisperX/issues/1336 if you prefer a more humorous perspective. # # Failure was first documented in nixpkgs under https://github.com/NixOS/nixpkgs/issues/460172. broken = true; }; } }) Loading
pkgs/development/python-modules/whisperx/default.nix +10 −18 Original line number Diff line number Diff line Loading @@ -13,6 +13,7 @@ huggingface-hub, nltk, numpy, omegaconf, pandas, pyannote-audio, torch, Loading @@ -36,16 +37,16 @@ let }; }; in buildPythonPackage rec { buildPythonPackage (finalAttrs: { pname = "whisperx"; version = "3.7.6"; version = "3.8.1"; pyproject = true; src = fetchFromGitHub { owner = "m-bain"; repo = "whisperX"; tag = "v${version}"; hash = "sha256-ZHPGQP5HIuFafHGS6ykiSNtHY6QHh0o8DUE2lV41lUI="; tag = "v${finalAttrs.version}"; hash = "sha256-2HjQtb8k3px0kqXowKtCXkiG2GuKLCuCtDOPYYa/tbc="; }; # As `makeWrapperArgs` does not apply to the module, and whisperx depends on `ffmpeg`, Loading @@ -60,9 +61,6 @@ buildPythonPackage rec { build-system = [ setuptools ]; pythonRelaxDeps = [ "numpy" "pandas" "pyannote-audio" "torch" "torchaudio" ]; Loading @@ -72,6 +70,7 @@ buildPythonPackage rec { huggingface-hub nltk numpy omegaconf pandas pyannote-audio torch Loading @@ -85,21 +84,14 @@ buildPythonPackage rec { # No tests in repository doCheck = false; pythonImportsCheck = [ "whisperx" ]; meta = { mainProgram = "whisperx"; description = "Automatic Speech Recognition with Word-level Timestamps (& Diarization)"; homepage = "https://github.com/m-bain/whisperX"; changelog = "https://github.com/m-bain/whisperX/releases/tag/${src.tag}"; changelog = "https://github.com/m-bain/whisperX/releases/tag/${finalAttrs.src.tag}"; license = lib.licenses.bsd2; maintainers = [ lib.maintainers.bengsparks ]; # nixpkgs has `pyannote-audio` >= 4.0.0, but `whisperx`'s `pyproject.toml` specifies <4.0.0. # # See https://github.com/m-bain/whisperX/issues/1240 for a serious discussion, # and a potential upgrade in https://github.com/m-bain/whisperX/pull/1243. # Alternatively read https://github.com/m-bain/whisperX/issues/1336 if you prefer a more humorous perspective. # # Failure was first documented in nixpkgs under https://github.com/NixOS/nixpkgs/issues/460172. broken = true; }; } })