
报错解决
余俊晖
余俊晖,NLP炼丹师,目前专注自然语言处理领域研究。曾获得国内外自然语言处理算法竞赛TOP奖项近二十项。在国内外SCI、顶会等发表多篇论文。
展开
-
【flash-attention】Building wheel for flash-attn (pyproject.toml) did not run successfully
【flash-attention】Building wheel for flash-attn (pyproject.toml) did not run successfully。原创 2024-01-09 16:37:43 · 3500 阅读 · 0 评论 -
【qwen】DeepSpeed Zero-3 is not compatible with `low_cpu_mem_usage=True` or with passing a `device_map
DeepSpeed Zero-3的核心功能就是在显存不足的情况下,使用CPU内存。原创 2023-12-28 17:22:00 · 1913 阅读 · 0 评论 -
cannot import name ‘DEFAULT_CIPHERS‘ from ‘urllib3.util.ssl_‘解决
urllib3版本冲突问题。原创 2023-12-11 11:18:26 · 5651 阅读 · 0 评论 -
paddleocr关闭log日志打印输出
可以看到paddleocr输出logging主要有两种,DEBUG和WARNING,因此关闭这两种打印日志即可。原创 2023-09-12 14:32:31 · 1574 阅读 · 0 评论 -
【报错解决】:DataGrip连接Mysql报08S01解决方法
如图修改:将useSsl设置为false。原创 2023-08-23 11:05:01 · 1546 阅读 · 1 评论 -
tensorboard报错:AttributeError: module ‘distutils‘ has no attribute ‘version‘
tensorboard报错:AttributeError: module ‘distutils‘ has no attribute ‘version‘环境:pytorch 1.10。原创 2023-08-17 17:40:23 · 302 阅读 · 0 评论 -
pytorch_pretrained_bert换成transformers踩坑
本文以文本分类为例叙述步骤1、前向传播时,pytorch_pretrained_bert是以下内容 _, pooled = self.bert(context, token_type_ids=types, attention_mask=mask, output_all_encoded_layers=False)报错: result = self.forwar原创 2021-06-13 15:44:03 · 3064 阅读 · 2 评论 -
To disable this warning, please explicitly set TOKENIZERS_PARALLELISM=(true | false)
pytorch transformers训练模型时报错:To disable this warning, please explicitly set TOKENIZERS_PARALLELISM=(true | false)解决方法:代码中导入:import osos.environ["TOKENIZERS_PARALLELISM"] = "false"原创 2021-04-23 15:48:22 · 4083 阅读 · 0 评论 -
ModuleNotFoundError: No module named ‘werkzeug.contrib‘错误
flask在服务器部署时出现:**ModuleNotFoundError: No module named ‘werkzeug.contrib’**错误。1、原因werkzeug在1.0以上版本已经移除了from werkzeug.contrib.fixers import ProxyFix2、解决办法from werkzeug.contrib.fixers import ProxyFix替换成from werkzeug.middleware.proxy_fix import ProxyFix原创 2021-04-11 21:41:31 · 1562 阅读 · 0 评论 -
jupyter notebook找不到conda的虚拟环境
步骤1、在默认的环境:conda install nb_conda_kernels2、然后分别激活不同的虚拟环境,安装ipykernelconda install ipykernel3、重启Jupyter Notebook原创 2021-02-10 14:51:43 · 837 阅读 · 0 评论 -
RuntimeError: Integer division of tensors using div or / is no longer supported, and in a future rel
RuntimeError: Integer division of tensors using div or / is no longer supported, and in a future release div will perform true division as in Python 3. Use true_divide or floor_divide (// in Python) instead.在高版本的pytorch中,会报这个错误。例子:norm_img = transforms原创 2020-11-19 21:32:34 · 372 阅读 · 0 评论 -
win10下cuda10.2配置tensorflow2.x
目前,tensorflow官方还未发布cuda10.2对应的tensorflow版本亲测有效的配置方法:windows下需要把cuda安装目录下的cudart64_102.dll复制一份重命名为cudart64_101.dll,具体目录是C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.2\bin\cudart64_101.dll,这个根据个人电脑安装可能有不同...原创 2020-11-10 13:32:39 · 695 阅读 · 0 评论 -
The “freeze_support()” line can be omitted if the program is not going to be frozen to produce an ex
PyTorch错误:The “freeze_support()” line can be omitted if the program is not going to be frozen to produce an executable.错误代码如下:RuntimeError: An attempt has been made to start a new process before the current process has finished its boots原创 2020-10-25 21:29:46 · 2161 阅读 · 0 评论 -
pip is configured with locations that require TLS/SSL, however the ssl modul win10+anaconda解决方案
pip is configured with locations that require TLS/SSL, however the ssl modul win10+anaconda解决方案#只需在环境变量中加入如下内容:D:\software\Anaconda3.7\ScriptsD:\software\Anaconda3.7\D:\software\Anaconda3.7\Librar...原创 2020-04-25 16:03:19 · 2623 阅读 · 2 评论 -
解决LINK : fatal error LNK1168: 无法打开 无法写入
LINK : fatal error LNK1168: 无法打开xxxxx 进行写入遇到这种情况,一般是vs进行图片读写时上一次程序占用没有彻底关闭, 此时需要做的是 打开任务管理器,找到该进程,杀死他,再编译,就会发现编译通过。 以上就是解决方法,或许不单单是进行图片操作是出现,比如:mfc界面操作时也会遇到这种情况,解决方法统一。原创 2017-11-12 17:37:45 · 15970 阅读 · 4 评论 -
django.core.exceptions.ImproperlyConfigured: mysqlclient 1.3.13 or newer is required; you have 0.9.3
解决方法1、找到Python安装路径下(我的是虚拟环境)~/.virtualenvs/web/lib/python3.6/sitepackages/django/db/backends/mysqlbase.py文件2、将文件中的如下代码注释if version < (1, 3, 3): raise ImproperlyConfigured("mysqlclient 1.3.3...原创 2020-03-31 17:02:13 · 210 阅读 · 0 评论