Orchestra-Research

Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime.

Battle-tested PyTorch training recipes for all domains — LLMs, vision, diffusion, medical imaging, protein/drug discovery, spatial omics…

Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments with MLflow…

Serverless GPU cloud platform for running ML workloads.

Merge multiple fine-tuned models using mergekit to combine capabilities without retraining.

Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT.

Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace.

Educational GPT implementation in ~300 lines.

GPU-accelerated data curation for LLM training.

NVIDIA's runtime safety framework for LLM applications.

Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution.

Fine-tune and serve Physical Intelligence OpenPI models (pi0, pi0-fast, pi0.5) using JAX or PyTorch backends for robot policy inference…