paddlepaddle和sklearn存在冲突，退出码-1073741819 (0xC0000005)
收藏
- 问题描述
在使用paddle.io.Dataloader时，如果之前导入过sklearn，数据加载器在读取最后一个batch时，会导致异常退出Process finished with exit code -1073741819 (0xC0000005)，而且内存占用有逐步异常增加现象 - 环境信息
1. Windows10 64bit
2. Python 3.8（Anaconda虚拟环境）
3. 安装的包# Name Version Build Channel astor 0.8.1 pypi_0 pypi babel 2.9.1 pypi_0 pypi backports-entry-points-selectable 1.1.0 pypi_0 pypi bce-python-sdk 0.8.62 pypi_0 pypi ca-certificates 2021.5.30 h5b45459_0 conda-forge certifi 2021.5.30 py38haa244fe_0 conda-forge cfgv 3.3.0 pypi_0 pypi charset-normalizer 2.0.4 pypi_0 pypi click 8.0.1 pypi_0 pypi colorama 0.4.4 pypi_0 pypi colorlog 5.0.1 pypi_0 pypi decorator 5.0.9 pypi_0 pypi dill 0.3.4 pypi_0 pypi distlib 0.3.2 pypi_0 pypi easydict 1.9 pypi_0 pypi filelock 3.0.12 pypi_0 pypi flake8 3.9.2 pypi_0 pypi flask 2.0.1 pypi_0 pypi flask-babel 2.0.0 pypi_0 pypi future 0.18.2 pypi_0 pypi gast 0.3.3 pypi_0 pypi gitdb 4.0.7 pypi_0 pypi gitpython 3.1.20 pypi_0 pypi h5py 3.3.0 pypi_0 pypi identify 2.2.13 pypi_0 pypi idna 3.2 pypi_0 pypi intel-openmp 2021.3.0 h57928b3_3372 conda-forge itsdangerous 2.0.1 pypi_0 pypi jieba 0.42.1 pypi_0 pypi jinja2 3.0.1 pypi_0 pypi joblib 1.0.1 pypi_0 pypi libblas 3.9.0 11_win64_mkl conda-forge libcblas 3.9.0 11_win64_mkl conda-forge liblapack 3.9.0 11_win64_mkl conda-forge markupsafe 2.0.1 pypi_0 pypi mccabe 0.6.1 pypi_0 pypi mkl 2021.3.0 hb70f87d_564 conda-forge multiprocess 0.70.12.2 pypi_0 pypi nodeenv 1.6.0 pypi_0 pypi numpy 1.21.1 py38h09042cb_0 conda-forge opencv-python 4.5.3.56 pypi_0 pypi openssl 1.1.1k h8ffe710_0 conda-forge packaging 21.0 pypi_0 pypi paddle2onnx 0.7 pypi_0 pypi paddlehub 2.1.0 pypi_0 pypi paddlenlp 2.0.7 pypi_0 pypi paddlepaddle-gpu 2.1.2.post110 pypi_0 pypi pandas 1.3.1 py38h60cbd38_0 conda-forge pip 21.2.2 py38haa95532_0 platformdirs 2.2.0 pypi_0 pypi pre-commit 2.14.0 pypi_0 pypi protobuf 3.17.3 pypi_0 pypi pycodestyle 2.7.0 pypi_0 pypi pycryptodome 3.10.1 pypi_0 pypi pyflakes 2.3.1 pypi_0 pypi python 3.8.11 h6244533_1 python-dateutil 2.8.2 pyhd8ed1ab_0 conda-forge python_abi 3.8 2_cp38 conda-forge pytz 2021.1 pyhd8ed1ab_0 conda-forge pyyaml 5.4.1 pypi_0 pypi pyzmq 22.2.1 pypi_0 pypi rarfile 4.0 pypi_0 pypi requests 2.26.0 pypi_0 pypi scikit-learn 0.24.2 pypi_0 pypi seqeval 1.2.2 pypi_0 pypi setuptools 52.0.0 py38haa95532_0 shellcheck-py 0.7.2.1 pypi_0 pypi six 1.16.0 pyh6c4a22f_0 conda-forge smmap 4.0.0 pypi_0 pypi sqlite 3.36.0 h2bbff1b_0 tbb 2021.3.0 h2d74725_0 conda-forge threadpoolctl 2.2.0 pypi_0 pypi toml 0.10.2 pypi_0 pypi tqdm 4.62.0 pypi_0 pypi typing-extensions 3.10.0.0 pypi_0 pypi urllib3 1.26.6 pypi_0 pypi vc 14.2 h21ff451_1 virtualenv 20.7.1 pypi_0 pypi visualdl 2.2.0 pypi_0 pypi vs2015_runtime 14.27.29016 h5e58377_2 werkzeug 2.0.1 pypi_0 pypi wheel 0.36.2 pyhd3eb1b0_0 wincertstore 0.2 py38_0 - 已尝试过的方法
回退scikit-learn版本0.22.2，无效 - 复现代码
import numpy as np import paddle from paddle.io import Dataset, DataLoader import sklearn # 注释掉这一行即可正常运行 class TestDataset(Dataset): def __init__(self, num): super(TestDataset, self).__init__() self.num = num self.images = [] self.labels = [] for _ in range(num): image = paddle.to_tensor(np.random.rand(3, 224, 224)) label = paddle.to_tensor(np.random.rand(1)) self.images.append(image) self.labels.append(label) def __getitem__(self, index): return self.images[index], self.labels[index] def __len__(self): return self.num train_dataset = TestDataset(num=50) train_loader = DataLoader(train_dataset, batch_size=8) for i, (input, target) in enumerate(train_loader): print('batch.num: {}, input.shape: {}, target.shape: {}'.format(i + 1, input.shape, target.shape)) print('success') - 截图
0
收藏
请登录后评论
建议提issue
Process finished with exit code-1073740791(0xC0000409)您好请问这是什么问题，在开始训练的时候出现这个问题
好呢好呢，请问您的GitHub账号是多少可以关注一下，我的问题真的找不到人能解决了