Will you be at
@CollisionHQ this week? Our team will be at the AWS Startup Theatre today, Wednesday, June 28, 12:00-12:45pm ET. We will be discussing how to navigate the frontiers of foundation models. #StabilityAI https://
collisionconf.com
LLMS
-
Stability AI explores foundation models frontier at Collision 2026
By
–
-
Cohere Launches New LLM Deployment Module
By
–
We're thrilled about the new possibilities with LLMs and can't wait for you to embark on this journey with us! To get started with the Deployment module, visit: https://
docs.cohere.com/docs/intro-dep
loyment
… -
Deploy LLM Applications as Extensions in Google Sheets
By
–
LLM applications can also be deployed as extensions to existing applications. You'll learn about two such extensions, the first one being Google Apps Script for Google Sheets. You'll learn how to build a text classifier and summarizer right inside Google Sheets.
-
Building LLM-Powered APIs with FastAPI for Sentiment Analysis
By
–
For building more serious applications, you’ll want to set up an LLM-powered back-end as endpoints. FastAPI lets you do that. It is a web framework for building APIs with Python. In this next module, you'll learn how to create a sentiment analysis classifier API with FastAPI.
-
Deploying LLM-Powered Applications: Platform Options Guide
By
–
Are you wondering what options are available out there for deploying LLM-powered applications? In http://
llm.university’s latest module, Deployment, you’ll learn how to deploy applications on several platforms and frameworks. Let’s take a look at what you’ll learn. -
Model Distillation Performance Trade-offs in Large Language Models
By
–
But yeah, modeling performance-wise you'll probably take a hit. Same thing that happened when they distilled ChatGPT GPT-3.5 and later GPT-4
-
Making LLM serving affordable through pruning and quantization
By
–
Fair point, but I think the appeal is more in trying to make running LLM more affordable (via pruning and quantization) to reduce the number of GPUs required for serving.
-
MLPerf GPT-3 Training Results Clarification Explained
By
–
Context: People are misunderstanding the GPT-3 – 11 minutes training result from the latest MLPerf.
reference: -
GPT-3 Training Speed: Clarifying the 11-Minute Claim
By
–
No, GPT-3 wasn't trained in 11 minutes. The GPT-3 architecture was trained on the C4 dataset to 2.69 log-probability in 11 minutes on 3584 H100 GPUs. Don't focus on the "11 minutes" — because it's like saying "ResNet-50 was trained in 5 seconds on MNIST to 80% accuracy"
-
Pruning Large Language Models Without Retraining Using Activation Norms
By
–
LLMs are usually too large for most contexts, but creating pruned versions of a model usually require retraining. Here's a new straightforward alternative based on computing element-wise product between the weight magnitude and norm of input activations: https://
arxiv.org/abs/2306.11695