-
i have three code snippet: A, B, C
A.
```
model = alexnet(pretrained=True).eval().cuda()
image_cv = cv2.imread(filename)
image_cv = cv2.resize(image_cv, (224, 224))
image_cv = torch.from_n…
-
Hi, when I convert [BiSeNet](https://github.com/ooooverflow/BiSeNet) model to trt,
```
from model.build_BiSeNet import BiSeNet
model = BiSeNet(12, 'resnet18').cuda()
model.module.load_state_dic…
-
There is a difference between FP32 and tensorrt FP32.
The data set is CIFAR100. Am I doing the wrong conversion?
~~~
def TensorRT(epoch):
# make_sample_image
sample_img = np.random.rand(a…
-
The default of torch2trt Linear input`s shape is [Batch,feature_in], the most of situation is [Batch,*,feature_in] in [pytorch api](https://pytorch.org/docs/stable/nn.functional.html#torch.nn.function…
-
Thank you for developing&maintaining TensorRT.
I'm currently trying to deploy my model which is trained with pytorch 1.3.1.
So, I made onnx file with (opset 7, ir_version 0.0.4) and made trt eng…
-
there is IResizeLayer in trt could do interpolate, why make interpolate plugin here?
just for demonstrating plugin usage?
-
Hi, I got 2 questions:
1. What's this trt different with tensorrt engine? (which commonly used when convert onnx model to tensorrt engine)
2. How using C++ api?
-
Hi John
While working on the pull request, I figured out that if I use interpolate plugin on `F.interpolate` it works fine otherwise there is an error when `nn.Upsample` is used. This happens for `…
-
## 🐛 Bug
## To Reproduce
Steps to reproduce the behavior:
```
In [1]: import torch …
-
With the latest master running Python 2.7, saving TRT model does not work.
```
# Convert model to TRT
print("Running torch2trt ...")
model = torch2trt(model, [input_t])
…