-
I have conf inputs in yaml:
```
model:
name: origin_model
framework: tensorflow
inputs: dense_input,seq_50_input,sparse_ids_input,sparse_wgt_inp…
-
1、我使用了`.whl`安装了`nncase1.5`,我该如何使用`ncc`命令行呢,还是说这个版本的nncase只能自己写py脚本来编译模型(`.tflite`)
2、我参考`Usage`里边的示例自己写了个py脚本,代码如下:
```python
import nncase
def read_model_file(model_file):
with open(mo…
-
## Bug Description
incompatible function arguments.
## To Reproduce
Steps to reproduce the behavior:
https://github.com/NVIDIA/Torch-TensorRT/blob/master/tests/py/test_ptq_trt_calibrator.…
-
When I compile and run the 20class_yolo_example on k210. It gives me an error.
============================================
LCD init
system start
[02m[..m/runtime_module.cpp:65 (shape_reg)] [0…
-
Thanks for your team's work. Can you post an **example** of **AdaRound** with tensorRT?
Since the doc[http://mqbench.tech/assets/docs/html/user_guide/PTQ/adaround.html](url) is hard to start and the…
-
Hi @jcwchen,
I converted a model from keras to onnx and converted that float32 to float16 model using float16.py and noticed that the inference time it took was more than the float32 model on CPU. So…
-
Hi,
Although others like `efficientdet_lite2_448_ptq.tflite` works well, I met a trouble when I tried to segment `efficientdet_lite3_512_ptq.tflite` that grabbed from
https://github.com/googl…
-
Description
I am very confused about the design concept of the Q/DQ node during QAT. now that TensorRT has the implementation of int8 conv op,
why we cannot quantize weights and activations to int…
-
## 🐛 Bug
https://github.com/pytorch/pytorch/pull/50748 added Linear-Bn1d fusion for PTQ. We should also have it for QAT. This is useful for vision transformers.
high level steps to enable this…
vkuzo updated
2 years ago
-
## ❓ Question
I'm wondering if there is a way to get the TRT model after compilation and visualize it. I trying to compare a PTQ model to a QAT model. I know I might have to do some further optim…