3 people b0dec2a11b [ROCM][DSV32][Perf][MTP] Enable UNIFORM_BATCH CG mode in rocm_aiter_mla_sparse (#45149) ...

Signed-off-by: Teemu Virolainen <teemu.virolainen@amd.com>
Co-authored-by: TJian <tunjian.tan@embeddedllm.com>
Co-authored-by: Andreas Karatzas <akaratza@amd.com>

2026-07-09 14:35:10 -05:00

..

attention

[ROCM][DSV32][Perf][MTP] Enable UNIFORM_BATCH CG mode in rocm_aiter_mla_sparse (#45149)

2026-07-09 14:35:10 -05:00

core

[P/D][Bugfix] Fix PD async KV load lookahead handling for MTP spec decode (#46694)

2026-07-09 10:01:22 +00:00

engine

Bump Transformers version to 5.10.4 (#41359)

2026-07-07 05:13:28 -07:00

executor

[UX] Log worker exit code when process dies unexpectedly (#38641)

2026-07-07 12:36:29 +00:00

kv_offload

[docs] Fix the docs build (#48008)

2026-07-08 15:47:22 -07:00

metrics

[Migration] Migrate GGUF quantization support to plugin (#39612)

2026-06-12 12:02:21 -07:00

pool

[Core] Move MRV1 late_interaction_runner.py out of MRV2 subtree (#48014)

2026-07-08 18:30:11 +00:00

sample

[Bugfix][Reasoning] Fix thinking_token_budget not enforced on re-entry after forced end (#43757)

2026-06-30 01:04:25 +00:00

simple_kv_offload

[KV Offloading] Add free block iterator for CPU offload scheduling (#47849)

2026-07-09 06:59:54 +00:00

spec_decode

Revert "[Platform] Replace torch.cuda.Event with torch.Event (#47140)" (#47668)

2026-07-06 14:18:33 +01:00

structured_output

Bump Transformers version to 5.10.4 (#41359)

2026-07-07 05:13:28 -07:00

worker

[Bugfix][MRV2] Reset num_accepted_tokens on add_request in all modes (#48132)

2026-07-09 18:17:11 +00:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

2024-11-11 23:05:38 +00:00

cudagraph_dispatcher.py

[Model] Add DiffusionGemma Support (#45163)

2026-06-11 22:17:35 -07:00

kv_cache_interface.py

[Bugfix][V1] Free out-of-window blocks on the processed-token basis under async scheduling (#47728)

2026-07-08 13:34:19 +01:00

kv_cache_spec_registry.py

[KVCache] Support Pluggable KVCacheSpec (#37505)

2026-06-03 09:05:16 -07:00

outputs.py

[Core] Cleanup KVConnector handling with PP + fix MRV2 (#43732)

2026-05-28 13:12:03 -07:00

request.py

[Bugfix][V1] Free out-of-window blocks on the processed-token basis under async scheduling (#47728)

2026-07-08 13:34:19 +01:00

serial_utils.py

[Core] Ensure memory is pinned prior to async h2d copy (#45424)

2026-06-20 20:02:24 -07:00

utils.py

Fix Transformers modeling backend usage stats (#47472)

2026-07-02 12:51:23 -07:00