Orchestra-Research

Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam…

Facebook's library for efficient similarity search and clustering of dense vectors.

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction.

GGUF format and llama.cpp quantization for efficient CPU/GPU inference.

Post-training 4-bit quantization for LLMs with minimal accuracy loss.

Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training

Control LLM output with regex and grammars, guarantee valid JSON/XML/code generation, enforce structured formats, and build multi-step…

Half-Quadratic Quantization for LLMs without calibration data.

Simplest distributed training API. 4 lines to add distributed support to any PyTorch script.

Fast tokenizers optimized for research and production.

Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse complex JSON with type…

Compress large language models using knowledge distillation from teacher to student models.