4 people 576bf75d0e [AMD][EPLB] Enable EPLB for Quark OCP MXFP4 MoE (#47220) ...

Signed-off-by: okorzh-amd <okorzh-amd@users.noreply.github.com>
Co-authored-by: okorzh-amd <okorzh-amd@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: Andreas Karatzas <akaratza@amd.com>

2026-07-03 14:41:52 -05:00

..

kernels

[Attention][DSA] support dcp for FLASHINFER_MLA_SPARSE (#46076)

2026-07-01 00:32:20 -04:00

layers

[AMD][EPLB] Enable EPLB for Quark OCP MXFP4 MoE (#47220)

2026-07-03 14:41:52 -05:00

model_loader

Remove more unnecessary load_weights methods (#47058)

2026-06-30 15:22:16 +01:00

models

Move Roberta remaining nn.Embedding to VocabParallelEmbedding (#47452)

2026-07-03 09:47:50 +00:00

offloader

[Platform] Replace torch.cuda.Event with torch.Event (#47140)

2026-06-30 08:39:59 -07:00

warmup

[CI Bugfix] Lazily import Qwen warmup dependencies (#47539)

2026-07-03 23:10:49 +08:00

__init__.py

[Platform] Deprecate seed_everything (#31659)

2026-01-04 18:34:04 -08:00

custom_op.py

Add ability to replace oot ops when using lora (#37181)

2026-03-16 18:04:15 -07:00

parameter.py

Speed up docs build (#44635)

2026-06-05 14:51:44 +00:00

utils.py

Remove more unnecessary load_weights methods (#47058)

2026-06-30 15:22:16 +01:00