A full spaCy pipeline and models for scientific/biomedical documents.

LLM training code for Databricks foundation models

Supercharge Your Model Training

A Data Streaming Library for Efficient Neural Network Training

Fast and flexible reference benchmarks

Scalene: a high-performance, high-precision CPU, GPU, and memory profiler for Python with AI-powered optimization proposals

A concise but complete full-attention transformer with a set of promising experimental features from various papers

Semantic Scholar's Author Disambiguation Algorithm & Evaluation Suite

The Semantic Scholar Search Reranker

https://arxiv.org/pdf/1909.04054