AlpinDale b3f9ab3b72 quant: add tensor parallel support for bitsandbytes (#1052) 2 nedēļas atpakaļ
..
device_communicators 239a8cae25 torch.compile: register all-reduce operations as custom ops (#1050) 2 nedēļas atpakaļ
__init__.py 9d81716bfd [v0.5.3] Release Candidate (#388) 8 mēneši atpakaļ
communication_op.py 0e558e9b2f fix: loading chameleon model with TP>1 (#695) 4 mēneši atpakaļ
parallel_state.py b3f9ab3b72 quant: add tensor parallel support for bitsandbytes (#1052) 2 nedēļas atpakaļ
utils.py 22a4cd4595 core: fix spec decode metrics and envs circular import (#889) 3 nedēļas atpakaļ