-
i'm trying to convert tensorflow model to onnex model, but i get this error
RuntimeError: Shape inference error(s): (op_type:Cast, node name: InceptionResnetV1/Bottleneck/BatchNorm/Reshape_1__503): …
-
Hi.
When I tried to export pre-trained ICNet, an error occurred.
How can I export it??
```
>>> import gluoncv as gcv
>>> from gluoncv.utils import export_block
>>> net = gcv.model_zoo.get_mode…
-
嵌入式的ARM平台上、PC 和 Caffe CPU推理时间相比 速度慢了许多 有没有优化建议啊
-
# System Information
- Windows 10 Version 2004 / Build 19041.508)
- WSL2 cat /etc/debian_version > 10.0
- Python 3.7.7
- TensorFlow-DirectML Version tensorflow_directml-1.15.3.dev200911-cp37-cp3…
-
在windows平台利用deeplabv3p做遥感影像地块分割训练，出现一些问题：
配置：windows10, RTX 3070, CUDA 10.2, cuDNN 7.6.5, paddlepaddle2.0rc. paddle.fluid.install_check.run_check() 输出Your Paddle Fluid is installed succesfully!
问题…
-
cuda10.1， cudnn8.0.4， nccl都安装正常，但是报了一个比较奇怪的错误，看不到原因
--------------------------------------
C++ Traceback (most recent call last):
--------------------------------------
0 paddle::framework…
-
Hello,
I just want some support to inference my model.
I transformed one model I found on the internet to ONNX and I want to use the 0.4 runtime by a C API (I already tested the model with pyth…
SiR0N updated
4 years ago
-
environment： gluoncv 0.7 or 0.8， mxnet-cu90mkl 1.6， gtx1080Ti x 2
The training segmentation model icnet does not use syncbn when the batchsize reaches 16 or less can also be trained normally. After…
-
Hi,
Thanks for sharing.
The link to the Resnet50 model is unavailable now, could you provide a new link?
Need the pretrained model for ICNet training.
-
Hello,
When I run `python train.py --arch icnet --dataset pascal --n_epoch 500`
I get the following output (checked on Windows and Linux, with PyTorch 0.3.1 and 0.4.1)
The dataset is the one [her…