Siddharth BedekarandOpenAI Codex 266b9d9c64 [Frontend][Core] Add sparse NCCL weight transfer support for in-place updates (#40096) ...

Signed-off-by: Siddharth Bedekar <bedeksid@gmail.com>
Co-authored-by: OpenAI Codex <codex@openai.com>

2026-06-01 15:37:30 -04:00

..

attention

[ROCm] Add attention sink support to AITer flash attention backend (#43817)

2026-05-30 18:13:18 +08:00

core

[Bugfix] Corrupted MLA + linear attention (#43961)

2026-05-29 05:00:51 -07:00

engine

[Bugfix] Fix Ray placement group allocation with grouped nodes (#43998)

2026-05-29 12:51:05 -06:00

executor

[Feat] Add support for per GPU worker RDMA NIC selection (#42083)

2026-05-28 12:45:23 -07:00

kv_offload

[kv_offload] Skip decode-phase blocks in CPU offload (#43797)

2026-05-29 06:39:43 +03:00

metrics

[Metrics] Exclude KV transfer tokens from iteration_tokens_total (#43346)

2026-05-29 19:56:44 +00:00

pool

[Misc] Added missing return type annotations to improve mypy and IDE tooling (#43383)

2026-05-23 13:28:22 +08:00

sample

[CPU Backend] CPU top-k and top-p sampling kernels using Triton (#43633)

2026-05-29 15:02:39 +08:00

simple_kv_offload

[Feat][KVConnector] Support DSV4 in SimpleCPUOffloadBackend (#42296)

2026-05-25 14:04:30 +08:00

spec_decode

[Model]Support Step-3.7-Flash (#43859)

2026-05-28 17:01:48 -07:00

structured_output

[Bug][Structured Outputs] Fix bug that leads to unconstrained generations with structural tags (#42452)

2026-05-20 07:08:58 -07:00

worker

[Frontend][Core] Add sparse NCCL weight transfer support for in-place updates (#40096)

2026-06-01 15:37:30 -04:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

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

cudagraph_dispatcher.py

[Experimental] Breakable CUDA graph (#42304)

2026-05-16 22:04:12 +08:00

kv_cache_interface.py

[Bugfix][Hybrid][NemotronH] Fix mamba_cache_mode=all + speculative decoding crash (#41233)

2026-05-18 14:54:00 +03:00

outputs.py

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

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

request.py

[bug] AsyncScheduler drops first post-resume token after pause_generation + clear_cache (#42117)

2026-05-19 01:06:21 -07:00

serial_utils.py

Add tensor IPC transfer mechanism for multimodal data (#32104)

2026-03-21 20:10:20 +00:00

utils.py

[Bug] Fix tests/distributed/test_elastic_ep.py - assert False (#43813)

2026-05-28 11:00:56 -04:00