A Survey on Image Generation and Generative Image Editing

The ability to generate and edit images using AI has evolved dramatically in the past few years. While diffusion models such as Stable Diffusion and Flux currently dominate the image generation landscape, recent hybrid architectures combining autoregressive and diffusion approaches have demonstrated comparable quality with significantly better computational efficiency. In

Integrating Linear Webhooks with Discord Using AWS Lambda

Workflow automation tools such as Zapier and Make are extremely popular and make life easier. However, if you are a small team trying to automate small tasks such as syncing updates from Jira or Linear for task management, paying $10-20 or more per month can feel excessive. That's

Object Detection Simplified: Fine Tune Faster R-CNN to Detect Trash

Introduction Trash and marine debris pollution pose significant environmental challenges, that impact ecosystems, wildlife, and human health. Detecting and classifying trash in diverse and complex environments is a significant challenge. The manual process of identifying and cataloging trash is not only labor-intensive and time-consuming but also prone to errors and

Protein-Ligand Binding Affinity Prediction: How Machine Learning Accelerates Drug Discovery

Accurately predicting the binding affinity and configuration between a ligand and its target protein are crucial in drug discovery, which directly affects the drug's effectiveness and selectivity. A high binding affinity is critical not only for the effectiveness of a drug but also for minimizing side effects. When

Deploying LLMs on AWS EC2 Using Cog: A Complete Guide

In our previous post, we introduced the key elements of LLM training, focusing particularly on Mistral-7B and we showed the process of dataset preparation and parameter-efficient fine-tuning Mistral-7B on a novel downstream task. Moving forward, we're shifting gears from model training to deployment. This post will guide you

Instruct Fine-Tuning Mistral 7B with PEFT: A Cookbook

Large language models have swept through academia and industry alike in the last 2 years. Close-source chat and general-purpose language models such as ChatGPT, GPT-4 and Claude have quickly been rivaled by open source alternatives such as Llama, Llama 2, Falcon and many others. However, it is worth noting that

Deploying an ML Model on Google Compute Engine

Introduction Deploying and maintaining machine learning models as a web service can be expensive and tricky. I’ve recently started using Compute Engine for my projects and found that Google Cloud, very much like AWS, has a rather steep learning curve. In this post, I will be discussing how to