-
it looks for me that your are down sampling along x-aixs only twice (so the final feature map along x-axis is 1/4 of the image width), but you are down sampling y-axis five times (so the final featur…
-
## 🐛 Bug
When I use a multi-threaded server to run rnn, it will segment fault. So I dig into it with gdb and find it crashes in ```torch/nn/modles/rnn.py```. I am new to PyTorch and cuda programmin…
-
hello friends, I met the following problem:
"from warpctc_pytorch import CTCLoss"
get a error:
> Traceback (most recent call last):
File "crnn_new.py", line 20, in
from warpctc_pytorch i…
-
Traceback (most recent call last):
File "demo.py", line 3, in
import model
File "/home/abc/github/chinese-ocr/model.py", line 5, in
from crnn.crnn import crnnOcr
File "/home/abc/g…
-
`torch/install/include/thpp/Storage.h:22:43: fatal error: thpp/if/gen-cpp2/Tensor_types.h: No such file or directory
compilation terminated.
CMakeFiles/crnn.dir/build.make:86: recipe for target 'CMa…
-
-
在编译安装warp_ctc的时候设置了CUDA_HOME,运行时依然提示no CUDA-capable device is detected
torch.cuda.is_available() 返回true
-
line 63行代码,为什么 recog_map = None ?
-
作者您好,我用3700多的中文字符训练集 训练模型,发现 训练几个小时后 cost 到50多就不下降了,然后把batch 改大 ,然后训练到cost 为30之后又不下降了。请问你出现过类似的问题吗?3700个类别会不会太多了,所以会产生这个问题? 我之前之用了100个中文做实验,效果挺好的。
-
请问这个问题怎么解决?
`Traceback (most recent call last):
File "crnn_main.py", line 211, in
cost = trainBatch(crnn, criterion, optimizer)
File "crnn_main.py", line 196, in trainBatch
cost.back…