mkurman

OpenAI CLIP is a contrastive language-image pre-training model. It enables zero-shot image classification, image-text similarity, concept search, and cross-modal retrieval by embedding images and text into a shared space.

EconML (Microsoft) is a library for heterogeneous treatment effect estimation, utilizing methods like Double ML, Causal Forest, Deep IV, and various metalearners. It applies orthogonal learning to estimate causal effects from observational data.

Weights & Biases provides ML experiment tracking and visualization, allowing you to log metrics, hyperparameters, model checkpoints, and artifacts. It offers collaborative dashboards, hyperparameter search, and a model registry.