深度学习bug
zycongzz
这个作者很懒,什么都没留下…
展开
-
transformers
transformertransformerVision TrasnformersSwin Transformerstransformer对于标准Transformer 要求输入token序列 即二位矩阵[num_token][token_dim]Vision Trasnformers在在代码实现中,通过一个卷积层来实现ViT-B/16 ,使用卷积核大小为16x16,stiide为16卷积核个数为768[224,224,3] ->[14,14,768]->[196,768]mod原创 2022-05-05 14:55:25 · 661 阅读 · 0 评论 -
修复初始化时 lr_scheduler 的 last_epoch 值
2022-04-28 18:36:09,822 - PCN - INFO - [Epoch 1/150][Batch 3601/3621] BatchTime = 0.418 (s) DataTime = 0.001 (s) Losses = ['3.2145', '2.9331'] lr = 0.000100/home/zyc/miniconda3/lib/python3.7/site-packages/torch/optim/lr_scheduler.py:154: UserWarning: The原创 2022-04-28 20:52:35 · 2724 阅读 · 0 评论 -
RuntimeError: CUDA error: an illegal memory access was encountered
问题描述Traceback (most recent call last): File "/home/zyc/code/code/mypointcloudnet/pointnet/pointnet/loss.py", line 53, in <module> print(loss) File "/home/zyc/miniconda3/lib/python3.7/site-packages/torch/_tensor.py", line 305, in __repr__原创 2022-04-28 19:53:44 · 2486 阅读 · 0 评论 -
error: command ‘/usr/local/cuda-11.6/bin/nvcc‘ failed with exit code
pytorch c++扩展 python setup install 错误解决方法将/home/zyc/miniconda3/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch 下cloneable。h 文件 中 改为下图所示TORCH_CHECK( copy->parameters_.size() == this->parameters_.size(), "Th原创 2022-04-27 16:31:48 · 4906 阅读 · 2 评论