AlpinDale 5b0c11d190 support pipeline parallel pynccl groups преди 5 месеца
..
__init__.py f8dfac6372 chore: attention refactor and upstream sync apr01 (#365) преди 9 месеца
cpu_executor.py ef733aee43 implement ExecuteModelData to reduce executor complexity преди 6 месеца
distributed_gpu_executor.py de62ceb18c refactor: eliminate parallel worker per-step task scheduling overhead преди 5 месеца
executor_base.py de62ceb18c refactor: eliminate parallel worker per-step task scheduling overhead преди 5 месеца
gpu_executor.py 236be273e5 feat: tensor parallel speculative decoding (#554) преди 6 месеца
multiproc_gpu_executor.py 5b0c11d190 support pipeline parallel pynccl groups преди 5 месеца
multiproc_worker_utils.py eaa06fdd14 fix some f-strings преди 6 месеца
neuron_executor.py ef733aee43 implement ExecuteModelData to reduce executor complexity преди 6 месеца
ray_gpu_executor.py de62ceb18c refactor: eliminate parallel worker per-step task scheduling overhead преди 5 месеца
ray_utils.py c6a501f682 add multiprocessing executor; make ray optional преди 6 месеца