I’d like to share 2 different ways to save and load a model using Pytorch. Saving The Entire Model #save model torch.save(model, PATH) #load model model = torch.load(PATH) model.eval() This save/load process has the least amount of code to implement.…
Here is one way to combine losses using PyTorch. Let’s say we want to train a binary semantic segmentation model using Binary Cross Entropy Loss. When you look at your data, you’ve noticed that the data is quite imbalanced. So…
I’ve learned about S3D(Separable 3D CNN) today so I like to share it here. S3D helps solve three challenges for video analysis. How to understand spatial information. (Recognizing the appearance of an object) How to understand temporal information. (Such as…
I found these losses quite hard to distinguish between each other. So, I created an intuitive explanation of the differences.
Here is one way to visualize your model’s training process using tensorboard with Pytorch #Inside Code for Training from torch.utils.tensorboard import SummaryWriter() tb = SummaryWriter() #Add Metrics you want to evaluate tb.add_scalar(‘Loss’,loss,epoch) tb.add_scalar(‘Accuracy’,accuracy,epoch) tb.add_scalar(‘F1Score’,fa_score,epoch) To start a tensorboard server go…
I did an experiment to compare how model performance change when converted to TensorRT engine. These are the steps I took. 1. Train Several Binary Segmentation Model 2. Randomly Choose 30 images from the dataset 3. Run inference with the…
There are several ways to convert arrays to torch tensors. import torch import numpy as np data = np.array([1,2,3,4,5,6]) option1 = torch.Tensor(arr) option2 = torch.tensor(arr) option3 = torch.as_tensor(arr) option4 = torch.from_numpy(arr) Option Description 1 Make a Copy and convert array…
I found out why my TensorRT engine model was not working as expected. I messed up with configuring the preprocessing step for Deepstream. When you use Deepstream to run inference, there is a property called net-scale-factor and offsets which you…
I’ve compared the performance between the model before and after converting to TensorRT engine. The output from the TensorRT engine is flattened to a 1d array, so I had to convert that back to a 2d array to display the…
A semantic segmentation model outputs a tensor shaped [Batch_size,Channel(Number of Classes),Img_Height,Img_Width] (If using Pytorch), but if you convert that to a TensorRT engine for faster inference, the output is flattened to a 1d array. Therefore the shaping being, [(Batch_size)X(Channel)X(Img_Height)X(Img_Width),] Considering…