Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

Tips

1. Training a model requires significantly more CPU/GPU memories than running inference using the model.

2. torch.cuda.empty_cache() doesn't help if memory is not enough

3. It is suggested that you …

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

Data Modeling with RDF(S)

Learn OWL and RDFS

RDF: Resource Description Framework OWL: Web Ontology Language

Products

neo4j

RDFox

Unique features of RDFox include the very powerful rule language (roughly …

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

1. Cross validation in scikit-learn supports pipeline in addition to vanilla models. Please refer to Cross Validation Pipeline for more details.

2. Label encoding is an easy way to convert a categorical response …

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

GPU Related Issues and Solutions

Input type (torch.FloatTensor) and weight type (torch.cuda.FloatTensor) should be the same

This means that the input data and the model are on different …

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

1. torchvision pretrained models

2. Model Zoo

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

Memory

1. ResNet152 inferencing requies about (a little bit less than) 2G GPU memory and training can be done with 8G GPU memory (need to set a small batch size). Deeper ResNet …