Gili NachumLLM domain adaptation using continued pre-training — Part 1/3An FAQ style three part deep diveMay 8May 8
Gili NachumHow to one click open Code Editor in SageMaker StudioCode Editor in Amazon SageMaker studio is a web IDE that runs on the cloud, yet looks and feels just like your VSCode desktop IDE.Mar 28Mar 28
Gili Nachum$ Cost of LLM continued pre-trainingHow much will it cost you to do continued pre-training for a small (7B) LLM?Feb 13Feb 13
Gili NachumDeploy ANY opensource LLM on AWS? (Easily)Try these options (with growing complexity): 1. As API on Amazon Bedrock. 2/ Deploy from Jumpstart 3/ Deploy To sagemaker from example…Jan 81Jan 81
Gili NachumWhat’s the Best LLM for Hebrew Classification?tl;dr: Claude2, GPT4 and Claude Instant came at the top. When also accounting for cost and latency, Claude Instant is the winner.Aug 13, 2023Aug 13, 2023
Gili Nachumמה מודל השפה הטוב ביותר לקלסיפיקציה בעברית?המודלים של Anthropic ו OpenAI במבחןJul 17, 2023Jul 17, 2023
Gili NachumLLM use cases for Enterprise companiesThe list below is large language models use cases with high relevancy for Enterprise companies. (Also see use cases per vertical by…Jan 31, 2023Jan 31, 2023
Gili NachumUsing vs.code on SageMaker notebookSetting up SSH to access gitlab, on vs.code in a SageMaker notebook instance, using this guide.Jul 31, 2022Jul 31, 2022
Gili NachumVerifying EFA is enabled on a SageMaker Training jobElastic Fabric Adapter (EFA) is a network interface for Amazon EC2 instances that enables customers to run applications requiring high…Jun 21, 2022Jun 21, 2022
Gili NachuminCloudy-MLBenchmarking TensorFlow on AWS DeepLearning AMIHere’s a simple TensorFlow benchmark on AWS DeepLearning AMI instances that focuses on convolutional network architecture (GPUs helps):Nov 11, 2018Nov 11, 2018