From d0bbf09eff92183b2777bcc1e8b5c5df83135f70 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E9=83=91=E7=89=B9=E9=A9=B9?= Date: Mon, 26 May 2025 15:28:44 +0800 Subject: [PATCH] =?UTF-8?q?=E6=9B=B4=E6=96=B0DBnet=E3=80=81CLIP=E3=80=81ES?= =?UTF-8?q?Pnet2=E4=BE=9D=E8=B5=96=E5=BA=93=E7=89=88=E6=9C=AC?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md | 8 ++++++++ .../audio/ESPnet2_for_PyTorch/requirements.txt | 3 ++- .../DB_ID0706_for_PyTorch/2.1_requirements.txt | 2 +- PyTorch/built-in/mm/CLIP_for_PyTorch/requirements.txt | 10 +++++----- 4 files changed, 16 insertions(+), 7 deletions(-) diff --git a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md index 37a69d343f..c161f6424e 100644 --- a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md +++ b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/README.md @@ -70,6 +70,12 @@ ESPNet是一套基于E2E的开源工具包，可进行语音识别等任务。 ``` pip3 install -r requirements.txt + + git clone https://github.com/lumaku/ctc-segmentation + cd ctc-segmentation + cythonize -3 ctc_segmentation/ctc_segmentation_dyn.pyx + python setup.py build + python setup.py install --optimize=1 --skip-build ``` @@ -86,6 +92,8 @@ ESPNet是一套基于E2E的开源工具包，可进行语音识别等任务。 - 跳过安装kaldi； - 安装espnet时，步骤1中的git clone ESPnet代码替换为下载本modelzoo中ESPnet的代码；步骤2跳过；步骤3中设置python环境，若当前已有可用的python环境，可以选择D选项执行；步骤4中进入tools目录后，需要增加installers文件夹的执行权限```chmod +x -R installers/```，然后直接使用make命令进行安装，不需要指定PyTorch版本； + + - make完成安装后，重新安装typeguard: pip install typeguard==2.13.3 - custom tool installation这一步可以选择不安装。check installation步骤在make时已执行，可跳过； diff --git a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/requirements.txt b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/requirements.txt index fdb2f39861..a70e104980 100644 --- a/PyTorch/built-in/audio/ESPnet2_for_PyTorch/requirements.txt +++ b/PyTorch/built-in/audio/ESPnet2_for_PyTorch/requirements.txt @@ -1,4 +1,5 @@ sympy==1.10.1 librosa==0.9.2 protobuf==3.20.3 -typeguard==2.13.3 \ No newline at end of file +typeguard==2.13.3 +pyworld==0.3.2 \ No newline at end of file diff --git a/PyTorch/built-in/cv/detection/DB_ID0706_for_PyTorch/2.1_requirements.txt b/PyTorch/built-in/cv/detection/DB_ID0706_for_PyTorch/2.1_requirements.txt index b4ad068b0a..be84dc2193 100644 --- a/PyTorch/built-in/cv/detection/DB_ID0706_for_PyTorch/2.1_requirements.txt +++ b/PyTorch/built-in/cv/detection/DB_ID0706_for_PyTorch/2.1_requirements.txt @@ -4,7 +4,7 @@ tensorboardX==1.8 opencv-python-headless anyconfig munch -scipy==1.2.7 +scipy==1.8.1 sortedcontainers shapely==1.3.0 pyclipper diff --git a/PyTorch/built-in/mm/CLIP_for_PyTorch/requirements.txt b/PyTorch/built-in/mm/CLIP_for_PyTorch/requirements.txt index 013ce07f39..69e41202d7 100644 --- a/PyTorch/built-in/mm/CLIP_for_PyTorch/requirements.txt +++ b/PyTorch/built-in/mm/CLIP_for_PyTorch/requirements.txt @@ -1,11 +1,11 @@ accelerate -datasets == 2.7.1 +datasets == 3.1.0 tokenizers -sentencepiece == 0.1.96 -protobuf == 3.14 +sentencepiece +protobuf wikiextractor -sklearn -torchvision == 0.9.1 +scikit-learn +torchvision == 0.16.0 Pillow decorator sympy -- Gitee