-
Hi, Thanks for this great work!
I'm wondering if you have training hyperparameters for all the self-trained models listed below:
https://github.com/rwightman/pytorch-image-models#self-trained-weig…
-
Environment:
Ubuntu Version: 18.04
OpenCV Version: 4.3.0
Vulkan Version: 1.1.114.0
NVIDIA Driver Version: 440.33.01
CUDA Driver Version: 10.2
The speed is slow compared to [nVIDIA RTX2080 …
-
@nihui 您好,我使用您给的https://github.com/nihui/ncnn-assets/blob/master/ncnntest-20190323.zip官方代码跑出来结果:
[0 GeForce RTX 2080] queueC=2 queueT=1 memU=4294967295 memDL=7 memHV=9
[0 GeForce RTX 2080] fp1…
-
## 背景
目前OpenCL Image2D高优Kernel已经合如develop完成。需要对5个高优模型尝试以Image2D形式跑通,这5个模型分别是:
1. MobileNetV1;
2. MobileNetV2;
3. MnasNet;
4. yolonano;
5. video-sr。
## 工作量
1. 【已完成】对OpenCL NCHW形式的Kernel…
-
## torchvision c++ api build from source, in master branch, Build error about C++14 and C++11
### Environment
PyTorch Version: 1.5.1 build from source
OS (e.g., Linux): Ubuntu 16
How you insta…
-
Hi rwightman, do you have any plans to release the training logs? especially for Mnasnet and efficientnet-b0. Because they are so tricky.
-
@iamhankai Hi,
Great work!
1. Why did you exclude EfficientNetB0 (0.390 BFlops - 76.3% Top1) from Accuracy-Latency chart?
2. Also what `mini_batch_size` did you use for training GhostNet?
…
-
When I build the latest torchvision with cuda 9.2, errors occur:
```
-- Caffe2: CUDA detected: 9.2
-- Caffe2: CUDA nvcc is: /usr/local/cuda/bin/nvcc
-- Caffe2: CUDA toolkit directory: /usr/local/c…
-
I'm trying to use ExponentialMovingAverage with MirroredStrategy training the model on several GPUs, but getting an error:
```
RuntimeError: Tried to create variable dense/kernel/replica_1/Exponenti…
-
如题,同一个项目,升级ncnn版本后,结果不正确。
旧版本使用的ncnn版本时间为:去年6月份左右
新版本:今天6月份vulkan版本的ncnn
代码说明:修改点只有引入新ncnn和openmp外。原项目代码没有设置过option。
新版本或vulkan版本,是不是还要专门要设置什么参数才能兼容低版本?