Orchestra-Research

PyTorch library for audio generation including text-to-music (MusicGen) and text-to-sound (AudioGen).

Autonomous AI agent platform for building and deploying continuous agents.

Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss.

Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support

Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss.

Vision-language pre-training framework bridging frozen image encoders and LLMs.

Open-source embedding database for AI applications.

OpenAI's model connecting vision and language.

Anthropic's method for training harmless AI through self-improvement.

Evaluates NVIDIA Cosmos Policy on LIBERO and RoboCasa simulation environments.

Applies cognitive science frameworks for creative thinking to CS and AI research ideation.

Multi-agent orchestration framework for autonomous AI collaboration.