- How to Speed Up XGBoost Model Training - Dec 20, 2021.
XGBoost is an open-source implementation of gradient boosting designed for speed and performance. However, even XGBoost training can sometimes be slow. This article will review the advantages and disadvantages of each approach as well as go over how to get started.
Machine Learning, Performance, Training, XGBoost
- AI Infinite Training & Maintaining Loop - Nov 4, 2021.
Productizing AI is an infrastructure orchestration problem. In planning your solution design, you should use continuous monitoring, retraining, and feedback to ensure stability and sustainability.
AI, Deployment, Machine Learning, Production, Training
- Speeding up Neural Network Training With Multiple GPUs and Dask - Sep 14, 2021.
A common moment when training a neural network is when you realize the model isn’t training quickly enough on a CPU and you need to switch to using a GPU. It turns out multi-GPU model training across multiple machines is pretty easy with Dask. This blog post is about my first experiment in using multiple GPUs with Dask and the results.
Dask, GPU, Neural Networks, Training
- How to Train a BERT Model From Scratch - Aug 13, 2021.
Meet BERT’s Italian cousin, FiliBERTo.
BERT, Hugging Face, NLP, Python, Training
- 10 Machine Learning Model Training Mistakes - Jul 30, 2021.
These common ML model training mistakes are easy to overlook but costly to redeem.
Machine Learning, Modeling, Training
- Exploring the SwAV Method - Jul 9, 2021.
This post discusses the SwAV (Swapping Assignments between multiple Views of the same image) method from the paper “Unsupervised Learning of Visual Features by Contrasting Cluster Assignments” by M. Caron et al.
Feature Extraction, Image Classification, Modeling, Training
- Write and train your own custom machine learning models using PyCaret - May 25, 2021.
A step-by-step, beginner-friendly tutorial on how to write and train custom machine learning models in PyCaret.
Machine Learning, Modeling, PyCaret, Python, Training
- How to Determine if Your Machine Learning Model is Overtrained - May 20, 2021.
WeightWatcher is based on theoretical research (done injoint with UC Berkeley) into Why Deep Learning Works, based on our Theory of Heavy Tailed Self-Regularization (HT-SR). It uses ideas from Random Matrix Theory (RMT), Statistical Mechanics, and Strongly Correlated Systems.
Learning, Modeling, Python, Training
- IBM Uses Continual Learning to Avoid The Amnesia Problem in Neural Networks - Feb 15, 2021.
Using continual learning might avoid the famous catastrophic forgetting problem in neural networks.
IBM, Learning, Neural Networks, Training
- How to Speed up Scikit-Learn Model Training - Feb 11, 2021.
Scikit-Learn is an easy to use a Python library for machine learning. However, sometimes scikit-learn models can take a long time to train. The question becomes, how do you create the best scikit-learn model in the least amount of time?
Distributed Systems, Hyperparameter, Machine Learning, Optimization, Parallelism, Python, scikit-learn, Training
- Building Deep Learning Projects with fastai — From Model Training to Deployment - Nov 4, 2020.
A getting started guide to develop computer vision application with fastai.
Deep Learning, Deployment, fast.ai, Modeling, Python, Training
- Microsoft and Google Open Sourced These Frameworks Based on Their Work Scaling Deep Learning Training - Nov 2, 2020.
Google and Microsoft have recently released new frameworks for distributed deep learning training.
Deep Learning, Google, Microsoft, Open Source, Scalability, Training
- Introduction to Federated Learning - Aug 20, 2020.
Federated learning means enabling on-device training, model personalization, and more. Read more about it in this article.
Data Labeling, Federated Learning, Mobile, Privacy, Training
- Stop training more models, start deploying them - Jun 30, 2020.
We are hardly living up to the promises of AI in healthcare. It’s not because of our training, it’s because of our deployment.
Deployment, Modeling, Training
- TensorFlow 2.0 Tutorial: Optimizing Training Time Performance - Mar 5, 2020.
Tricks to improve TensorFlow training time with tf.data pipeline optimizations, mixed precision training and multi-GPU strategies.
Neural Networks, Optimization, Python, TensorFlow, Training
- Uber Creates Generative Teaching Networks to Better Train Deep Neural Networks - Jan 13, 2020.
The new technique can really improve how deep learning models are trained at scale.
Generative Adversarial Network, Neural Networks, Training, Uber
- This Microsoft Neural Network can Answer Questions About Scenic Images with Minimum Training - Oct 21, 2019.
Recently, a group of AI experts from Microsoft Research published a paper proposing a method for scene understanding that combines two key tasks: image captioning and visual question answering (VQA).
Image Recognition, Microsoft, Neural Networks, Question answering, Training
- Train sklearn 100x Faster - Sep 11, 2019.
As compute gets cheaper and time to market for machine learning solutions becomes more critical, we’ve explored options for speeding up model training. One of those solutions is to combine elements from Spark and scikit-learn into our own hybrid solution.
Distributed Systems, Machine Learning, Python, scikit-learn, Training
- Pre-training, Transformers, and Bi-directionality - Jul 12, 2019.
Bidirectional Encoder Representations from Transformers BERT (Devlin et al., 2018) is a language representation model that combines the power of pre-training with the bi-directionality of the Transformer’s encoder (Vaswani et al., 2017). BERT improves the state-of-the-art performance on a wide array of downstream NLP tasks with minimal additional task-specific training.
AISC, BERT, NLP, Training, Transformer
- Checklist for Debugging Neural Networks - Mar 22, 2019.
Check out these tangible steps you can take to identify and fix issues with training, generalization, and optimization for machine learning models.
Checklist, Modeling, Neural Networks, Optimization, Tips, Training
- 10 Tools to Help You Learn R - Jan 4, 2018.
There are several tools to help you grasp the foundational principles and more. The list below gives you an idea of what’s available and how much it costs.
R, Tools, Training
- 8 Ways to Improve Your Data Science Skills in 2 Years - Nov 17, 2017.
Two years. Two years is the maximum amount of time you should spend focused on your learning, education and training. That’s exactly why this guide is focused on honing the most beneficial skills in two years.
Data Science, Data Science Skills, Skills, Training
- Applying Deep Learning to Real-world Problems - Jun 30, 2017.
In this blog post I shared three learnings that are important to us at Merantix when applying deep learning to real-world problems. I hope that these ideas are helpful for other people who plan to use deep learning in their business.
Balancing Classes, Deep Learning, Neural Networks, Training, Unbalanced
- Adversarial Validation, Explained - Oct 7, 2016.
This post proposes and outlines adversarial validation, a method for selecting training examples most similar to test examples and using them as a validation set, and provides a practical scenario for its usefulness.
Pages: 1 2
Adversarial, Explained, Training, Validation
- Apache Spark: O’Reilly Certification, EU Training, University Program - Sep 26, 2014.
Recent news on Apache Spark includes developer certification from O'Reilly, upcoming training workshops in EU by Databricks, and Spark tutorial events at major universities.
Academics, Apache Spark, Big Data, Certification, Databricks, Paco Nathan, Strata, Training