atlas news
    
Neptune Ai
28  mars     12h00
How to Optimize GPU Usage During Model Training With neptune.ai
Mirza Mujtaba    As data scientists or machine learning engineers, one of the routine tasks we tackle involves rapid experimentation and training multiple models under different settings to identify the most effective ones. This iterative process is usually one of the most costly and time consuming phases, so any...
22  mars     15h00
Zero-Shot and Few-Shot Learning with LLMs
MichaÅ‚ Oleszak    Chatbots based on Large Language Models LLMs , such as OpenAI’s ChatGPT, show an astonishing capability to perform tasks for which they have not been explicitly trained. In some cases, they can do it out of the box. In others, the user must specify a few labeled examples for the model to pick up...
12  mars     11h36
LLMOps: What It Is, Why It Matters, and How to Implement It
Stephen Oladele    Large Language Models LLMs like Meta AI’s LLaMA models, MISTRAL AI’s open models, and OpenAI’s GPT series have improved language based AI. These models excel at various tasks, such as translating languages with remarkable accuracy, generating creative writing, and even coding software. A...
08  mars     20h00
The Real Cost of Self-Hosting MLflow
Aurimas Griciunas    MLflow is well regarded as an experiment tracking platform. Since it’s open source, you can download it for free and host as many instances as you want without incurring license fees. This, and the extendability of MLflow, sees data science teams gravitating towards adopting it as their end to end...
01  mars     09h24
Deep Learning Model Optimization Methods
Alessandro Lamberti    Deep learning models continue to dominate the machine learning landscape. Whether it’s the original fully connected neural networks, recurrent or convolutional architectures, or the transformer behemoths of the early s, their performance across tasks is unparalleled. However, these capabilities...
22  février     11h17
Continual Learning: Methods and Application
Mateusz Wojcik    At the beginning of my machine learning journey, I was convinced that creating an ML model always looks similar. You start with a business problem, prepare a dataset, and finally train the model, which is evaluated and deployed. Then, you repeat this process until you are satisfied with the results...
15  février     15h17
2024 Layoffs and LLMs: Pivoting for Success
Aurimas Griciunas    If Oxford declared that the word of the year for was layoff,’ it wouldn’t surprise tens of thousands of people across the globe. In a time where economic challenges force companies to streamline operations, machine learning ML specialists and adjacent roles are not immune to the trend of...
26  janvier     15h30
Mikiko Bazeley: What I Learned Building the ML Platform at Mailchimp
Mikiko Bazeley    I started my ML journey as an analyst back in . Since then, I’ve worked as a data scientist for a multinational company and an MLOps engineer for an early stage startup before moving to Mailchimp in May . I joined just before its billion acquisition by Intuit. It was an exciting time to...
    14h23
How to Build Machine Learning Systems With a Feature Store
Jim Dowling    Training and evaluating models is just the first step toward machine learning success. To generate value from your model, it should make many predictions, and these predictions should improve a product or lead to better decisions. For this, we have to build an entire machine learning system around...
24  janvier     15h43
Logging PyMC and Arviz Artifacts on Neptune
Alessandro Angioi    When dealing with limited data or uncertain scenarios, one of the most potent methods is Bayesian inference. At its core, it is a formulation of statistics that enables one to incorporate prior knowledge and update beliefs systematically and coherently. Its power lies in the flexibility in model...
19  janvier     13h24
LLM Fine-Tuning and Model Selection Using Neptune and Transformers
Pedro Gabriel Gengo Lourenço    Imagine you’re facing the following challenge: you want to develop a Large Language Model LLM that can proficiently respond to inquiries in Portuguese. You have a valuable dataset and can choose from various base models. But here’s the catch you’re working with limited computational resources...
14  novembre     15h30
How to Visualize Deep Learning Models
Nilesh Barla    Deep learning models are typically highly complex. While many traditional machine learning models make do with just a couple of hundreds of parameters, deep learning models have millions or billions of parameters. The large language model GPT that OpenAI released in the spring of is rumored...