From 188444f77841725fd720cb1115fd700bc6363615 Mon Sep 17 00:00:00 2001
From: zxcd <228587199@qq.com>
Date: Sun, 29 Sep 2024 11:22:13 +0800
Subject: [PATCH 1/3] fix (#3856)
---
README.md | 2 +-
paddlespeech/cls/exps/panns/deploy/predict.py | 7 +++++--
2 files changed, 6 insertions(+), 3 deletions(-)
diff --git a/README.md b/README.md
index 19ec61cb0..5197ff4c7 100644
--- a/README.md
+++ b/README.md
@@ -7,7 +7,7 @@
- 
+ 
diff --git a/paddlespeech/cls/exps/panns/deploy/predict.py b/paddlespeech/cls/exps/panns/deploy/predict.py
index f14b44215..866a669e8 100644
--- a/paddlespeech/cls/exps/panns/deploy/predict.py
+++ b/paddlespeech/cls/exps/panns/deploy/predict.py
@@ -17,7 +17,7 @@ import os
import numpy as np
from paddle import inference
from paddle.audio.datasets import ESC50
-from paddle.audio.features import MelSpectrogram
+from paddle.audio.features import LogMelSpectrogram
from paddleaudio.backends import soundfile_load as load_audio
from scipy.special import softmax
@@ -53,7 +53,10 @@ def extract_features(files: str, **kwargs):
pad_width = max_length - len(waveforms[i])
waveforms[i] = np.pad(waveforms[i], pad_width=(0, pad_width))
- feat = MelSpectrogram(waveforms[i], sr, **kwargs).transpose()
+ feature_extractor = LogMelSpectrogram(sr, **kwargs)
+ feat = feature_extractor(paddle.to_tensor(waveforms[i]))
+ feat = paddle.transpose(feat, perm=[1, 0]).unsqueeze(0)
+
feats.append(feat)
return np.stack(feats, axis=0)
From bf03c9a620463962b79a5ed3b9f1dbb8c0340fb1 Mon Sep 17 00:00:00 2001
From: tianshuo78520a <707759223@qq.com>
Date: Tue, 8 Oct 2024 14:30:39 +0800
Subject: [PATCH 2/3] Add server ci (#3857)
* Add server ci
* fix
---
tests/unit/ci.sh | 5 +++++
tools/Dockerfile | 2 +-
2 files changed, 6 insertions(+), 1 deletion(-)
diff --git a/tests/unit/ci.sh b/tests/unit/ci.sh
index 9342a2685..72b4678d6 100644
--- a/tests/unit/ci.sh
+++ b/tests/unit/ci.sh
@@ -26,6 +26,11 @@ function main(){
cd ${speech_ci_path}/cli
bash test_cli.sh
echo "End cli"
+
+ echo "Start server"
+ cd ${speech_ci_path}/server/offline
+ bash test_server_client.sh
+ echo "End server"
}
main
diff --git a/tools/Dockerfile b/tools/Dockerfile
index 18596f32c..b2f2b3125 100644
--- a/tools/Dockerfile
+++ b/tools/Dockerfile
@@ -1,4 +1,4 @@
FROM registry.baidubce.com/paddlepaddle/paddle:latest-dev-cuda11.2-cudnn8.2-trt8.0-gcc82
RUN apt-get update -y
-RUN apt-get -y install libsndfile1
+RUN apt-get -y install libsndfile1 lsof
RUN pip3.8 install pytest-runner
From 658d19a73e2c8af06f5e17efd5b3885eb0689018 Mon Sep 17 00:00:00 2001
From: zxcd <228587199@qq.com>
Date: Wed, 9 Oct 2024 17:38:49 +0800
Subject: [PATCH 3/3] =?UTF-8?q?=E3=80=90doc=E3=80=91add=20Squeezeformer=20?=
=?UTF-8?q?info=20in=20README=20(#3860)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
* Update README.md
* Update README_cn.md
* Update README_cn.md
---
README.md | 1 +
README_cn.md | 1 +
2 files changed, 2 insertions(+)
diff --git a/README.md b/README.md
index 5197ff4c7..00367d787 100644
--- a/README.md
+++ b/README.md
@@ -179,6 +179,7 @@ Via the easy-to-use, efficient, flexible and scalable implementation, our vision
### Recent Update
- 👑 2023.05.31: Add [WavLM ASR-en](https://github.com/PaddlePaddle/PaddleSpeech/blob/develop/examples/librispeech/asr5), WavLM fine-tuning for ASR on LibriSpeech.
+- 🎉 2023.05.18: Add [Squeezeformer](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/aishell/asr1), Squeezeformer training for ASR on Aishell.
- 👑 2023.05.04: Add [HuBERT ASR-en](https://github.com/PaddlePaddle/PaddleSpeech/blob/develop/examples/librispeech/asr4), HuBERT fine-tuning for ASR on LibriSpeech.
- ⚡ 2023.04.28: Fix [0-d tensor](https://github.com/PaddlePaddle/PaddleSpeech/pull/3214), with the upgrade of paddlepaddle==2.5, the problem of modifying 0-d tensor has been solved.
- 👑 2023.04.25: Add [AMP for U2 conformer](https://github.com/PaddlePaddle/PaddleSpeech/pull/3167).
diff --git a/README_cn.md b/README_cn.md
index 7aef30871..d70940dd2 100644
--- a/README_cn.md
+++ b/README_cn.md
@@ -184,6 +184,7 @@
### 近期更新
- 👑 2023.05.31: 新增 [WavLM ASR-en](https://github.com/PaddlePaddle/PaddleSpeech/blob/develop/examples/librispeech/asr5), 基于WavLM的英语识别微调，使用LibriSpeech数据集
+- 🎉 2023.05.18: 新增 [Squeezeformer](https://github.com/PaddlePaddle/PaddleSpeech/tree/develop/examples/aishell/asr1), 使用Squeezeformer进行训练，使用Aishell数据集
- 👑 2023.05.04: 新增 [HuBERT ASR-en](https://github.com/PaddlePaddle/PaddleSpeech/blob/develop/examples/librispeech/asr4), 基于HuBERT的英语识别微调，使用LibriSpeech数据集
- ⚡ 2023.04.28: 修正 [0-d tensor](https://github.com/PaddlePaddle/PaddleSpeech/pull/3214), 配合PaddlePaddle2.5升级修改了0-d tensor的问题。
- 👑 2023.04.25: 新增 [U2 conformer 的 AMP 训练](https://github.com/PaddlePaddle/PaddleSpeech/pull/3167).