vLLM/vllm/model_executor at 24fab45d96a91b491db338ee02cd24e55b7fbb5f - vLLM

iHore/vLLM

Fork 0

Files

T

History

63400259d0 [Performance] Move apply_w8a8_block_fp8_linear to an op class (#24666 )

Signed-off-by: ElizaWszola <ewszola@redhat.com>
Signed-off-by: ElizaWszola <elizaw.9289@gmail.com>
Signed-off-by: Luka Govedič <lgovedic@redhat.com>
Signed-off-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Luka Govedič <lgovedic@redhat.com>

2025-09-23 12:03:10 -07:00

layers

[Performance] Move apply_w8a8_block_fp8_linear to an op class (#24666 )

2025-09-23 12:03:10 -07:00

model_loader

[TPU] Deprecate xm.mark_step in favor of `torch_xla.sync (#25254 )

2025-09-22 10:12:57 +00:00

models

[V1] Remove V0 code paths for Hybrid models (#25400 )

2025-09-23 08:26:13 -07:00

warmup

[Kernels] Support blocked fp8 quantization for compressed tensors MoE (#25219 )