Create fast graph language models from converted PDF documents for knowledge extraction and Q&A.

Simple package to extract text with coordinates from programmatic PDFs

🪞PRIMP (Python Requests IMPersonate). The fastest python HTTP client that can impersonate web browsers

BoringSSL bindings for the Rust programming language.

InstructLab Core package. Use this to chat with a model and execute the InstructLab workflow to train a model using custom taxonomy data.

Bazel Python Rules

self-contained, portable perl binaries

Python bindings to PDFium

📰 Binary distribution of PDFium

Raw bindings to platform APIs for Rust

Private Open AI on Kubernetes

Accessible large language models via k-bit quantization for PyTorch.

Environment for creating a docker image running cph performance tests for Persistent and Non-Persistent messaging against an MQ Queue Manager.

Representation and Reference Lowering of ONNX Models in MLIR Compiler Infrastructure

Triton backend that enables pre-process, post-processing and other logic to be implemented in Python.

The Triton backend for the ONNX Runtime.

The core library and APIs implementing the Triton Inference Server.

Generative AI extensions for onnxruntime

A high-throughput and memory-efficient inference and serving engine for LLMs

Development repository for the Triton language and compiler

🚀 Accelerate training and inference of 🤗 Transformers and 🤗 Diffusers with easy to use hardware optimization tools

A cargo-subcommand to speed up Rust Docker builds using Docker layer caching.

Embeddings Microservice for use in various projects

OpenShift on IBM PowerVM servers managed using HMC