Orchestra Research
- 114 skills
- 0 followers
- 10k repo stars
- 2 weeks ago last updated
- ▌ Audiocraft Audio Generation · orchestra-research bundleGenerate music and sound effects from text descriptions using Meta's AudioCraft library, with support for melody conditioning, stereo output, and style transfer.
- ▌ Openrlhf Training · orchestra-research bundleTrain large language models (7B-70B+) with RLHF using PPO, GRPO, DPO, and other algorithms, accelerated by Ray and vLLM for distributed multi-GPU setups.
- ▌ Serving Llms Vllm · orchestra-research bundleDeploy and serve LLMs with high throughput using vLLM's PagedAttention and continuous batching. Supports OpenAI-compatible endpoints, quantization (GPTQ/AWQ/FP8), and tensor parallelism for production inference.
- ▌ Fine Tuning Openvla Oft · orchestra-research bundleFine-tunes and evaluates OpenVLA-OFT and OpenVLA-OFT+ policies for robot action generation with continuous action heads, LoRA adaptation, and FiLM conditioning on LIBERO simulation and ALOHA real-world setups.
- ▌ Rwkv Architecture · orchestra-research bundleUse RWKV, a linear-time RNN-Transformer hybrid, for efficient long-context inference and training with constant memory usage.
- ▌ Skypilot Multi Cloud Orchestration · orchestra-research bundleRun ML training and batch jobs across multiple clouds with automatic cost optimization, spot instance recovery, and unified orchestration.
- ▌ Dspy · orchestra-research bundleBuild complex AI systems with declarative programming, optimize prompts automatically, and create modular RAG systems and agents using Stanford NLP's DSPy framework.
- ▌ Langsmith Observability · orchestra-research bundleDebug, evaluate, and monitor LLM applications with tracing, datasets, and built-in evaluators.
- ▌ Mamba Architecture · orchestra-research bundleTrain and run Mamba state-space models with O(n) complexity, achieving faster inference and longer context than Transformers.
- ▌ Ray Data · orchestra-research bundleProcess large-scale ML datasets with distributed streaming execution across CPU/GPU, supporting Parquet, CSV, JSON, images, and integration with PyTorch, TensorFlow, and Ray Train.
- ▌ Torchforge Rl Training · orchestra-research bundleTrain reinforcement learning models using torchforge, Meta's PyTorch-native RL library for scalable, algorithm-focused experimentation with GRPO, DAPO, and custom loss functions.
- ▌ Sglang · orchestra-research bundleServe LLMs and VLMs with structured outputs, prefix caching, and high throughput using RadixAttention.
- ▌ Weights And Biases · orchestra-research bundleTrack ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, and manage model registry with W&B.
- ▌ Evaluating Cosmos Policy · orchestra-research bundleEvaluate NVIDIA Cosmos Policy on LIBERO and RoboCasa simulation environments with headless GPU evaluation and inference profiling.