Research platform for model training, evaluation, and experimentation across architectures, benchmarks, and recipes.
benchmarking machine-learning research deep-learning evaluation transformer experimentation lora ddp knowledge-distillation distributed-training finetuning sft dpo mixture-of-experts pretraining llm instruction-tuning fsdp causal-lm
-
Updated
Jul 20, 2026 - Python