Posts in: Blog

How alpha, rank, and learning rate affect model accuracy, and whether rank-stabilized LoRA helps.

Highlights from CVPR and ICML 2024.

Sora, Make-a-Video and Imagen Video explained.

Tensor parallelism, from beginner to expert using PyTorch.

How we implemented Retrieval Augmented Generation for the Antonio Nearly hologram, including the data sources and ingestion, vector DB, embedding model, reranker, and query rephrasing.

Announcing GenAI studio general availability for Determined Enterprise users.

How to measure activation memory in PyTorch, and why the choice of activation function matters.

Let’s compare how Mistral-7b-instruct-v0.2 and Llama 2-7B-chat perform on some basic LLM prompts.

An introduction to activation memory, and how it affects GPU memory consumption during model training.

Using HPE’s AI platform to develop an early brain cancer detection machine learning model.