Helmet Detector based on the CenterNet.
Comprehensive image classification for training multilayer perceptron (MLP), LeNet, LeNet5, conv2, conv4, conv6, VGG11, VGG13, VGG16, VGG19 with batch normalization, ResNet18, ResNet34, ResNet50, MobilNetV2 on MNIST, CIFAR10, CIFAR100, and ImageNet1K.
This repository demonstrates the efficiency difference between training with AMP and without AMP.
An end-to-end Deep Learning pipeline for Digital Image Processing: Image restoration (denoising & deblurring via U-Net/ResNet) followed by robust image classification using PreActResNet-18 on CIFAR-10 with PyTorch AMP & TorchScript export.
Multi-Engine (PyTorch & JAX/XLA) Zero-Branching Geometric Acceleration Core. Enforces 0% Graph Breaks & Real-time Fault-Isolation via hardware-native bitwise MUX operations (torch.where / jax.lax.select) to permanently eliminate 'jmp' instructions and host-device synchronization fences.
Experimentations with optimising techniques for pytorch modelling - Pruning, Automatic Mixed Precision, Knowledge Distillation, Quantization, Profiling, Forward Automatic Differentiation
Add a description, image, and links to the automatic-mixed-precision topic page so that developers can more easily learn about it.
To associate your repository with the automatic-mixed-precision topic, visit your repo's landing page and select "manage topics."