5 people d2d649e674 [DS V4.1][Engram] Support async prefetch for offloaded engram lookups and engram DP sharding (#56512) ...

Signed-off-by: Juntian Liu <Juntianl777@gmail.com>
Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
Co-authored-by: Roger Wang <hey@rogerw.io>
Co-authored-by: ywang96 <>
Co-authored-by: Woosuk Kwon <woosuk@inferact.ai>
Co-authored-by: Codex <noreply@openai.com>

2026-09-13 02:44:23 -07:00

..

__init__.py

[CI/Build] Move test_utils.py to tests/utils.py (#4425)

2024-05-13 23:50:09 +09:00

conftest.py

fix: include topic frame in KV events replay response (#45177)

2026-07-07 14:48:23 -04:00

eplb_utils.py

[EPLB] Add alternative communication for EPLB weight exchange (#33176)

2026-03-31 08:17:12 -04:00

test_ca_buffer_sharing.py

[Hardware] Replace torch.cuda.synchronize() api with torch.accelerator.synchronize (#36085)

2026-03-05 10:36:39 +00:00

test_comm_ops.py

[Perf][ROCm] Add AITER custom AG/RS (DP only) (#48247)

2026-09-11 13:19:57 +08:00

test_context_parallel.py

DCP supports hybrid attention (#40996)

2026-07-09 21:34:45 -07:00

test_custom_all_gather_reduce_scatter.py

Fix MNNVL Lamport mailbox publication and cleanup (#53000)

2026-08-23 23:56:54 -07:00

test_custom_all_reduce.py

[Model] Support Qwen3.8-Flash-Next (#53896)

2026-08-31 05:57:56 +00:00

test_dcp_a2a.py

[DCP] Default query replication for GLM sparse attention (#50382)

2026-08-21 09:07:05 -07:00

test_dcp_direct_a2a_lse_reduce.py

[refactor] consolidate cp attn ops (#52839)

2026-08-19 18:04:11 -07:00

test_distributed_oot.py

[CI] Reorganize entrypoints CI (#44947)

2026-06-09 00:46:11 -07:00

test_elastic_ep.py

[Elastic EP] Reuse CUDA graphs across reconfiguration (#54985)

2026-09-12 10:26:39 -04:00

test_engram_dp_shard.py

[DS V4.1][Engram] Support async prefetch for offloaded engram lookups and engram DP sharding (#56512)

2026-09-13 02:44:23 -07:00

test_eplb_algo.py

fix and test EPLB balancedness calculation (#51813)

2026-08-12 20:27:22 -07:00

test_eplb_events.py

[EPLB] Refactor Async EPLB synchronization logic (#37601)

2026-04-20 17:05:41 +00:00

test_eplb_execute.py

[Bugfix] Pin EPLB and MLA host-to-device transfer buffers (#56138)

2026-09-11 01:57:47 -07:00

test_eplb_fused_moe_layer_dep_nvfp4.py

[MoE Refactor] Rename FusedMoE to FusedMoEFactory (#44941)

2026-07-31 11:28:18 +01:00

test_eplb_fused_moe_layer.py

[MoE Refactor] Rename FusedMoE to FusedMoEFactory (#44941)

2026-07-31 11:28:18 +01:00

test_eplb_quant_scale_consistency.py

feat: extended EPLB support for Mistral Large 3 and additional MoE backends (#48355)

2026-08-07 05:32:52 -07:00

test_eplb_spec_decode.py

[EPLB] Make async EPLB default (#43219)

2026-05-29 18:07:16 +00:00

test_eplb_utils.py

[Elastic EP] Reuse CUDA graphs across reconfiguration (#54985)

2026-09-12 10:26:39 -04:00

test_events.py

feat(grpc): add KV event source discovery (#50033)

2026-07-29 13:46:40 -07:00

test_expert_parallel.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

test_expert_placement.py

[MoE Refactor] Move expert map related code into ExpertMapManager class (#41046)

2026-05-12 09:18:27 -04:00

test_file_store.py

[BugFix] Use file:// rendezvous for single-node executors to eliminate startup port races (#50999)

2026-08-09 22:00:04 -04:00

test_flashinfer_pcie_ipc_all_reduce.py

[Distributed] Add opt-in FlashInfer PCIe IPC all-reduce backend (#53576)

2026-09-01 03:34:56 +00:00

test_kimi_linear_context_parallel.py

[Kimi-K3] support DCP partial prefix cache hit (#50493)

2026-08-17 17:05:51 -07:00

test_kv_cache_events.py

[Core][KV Events] Echo session_id on GPU BlockStored events (#51381)

2026-09-04 02:22:50 +00:00

test_kvlayout.py

[6/N][KV-Cache Layout Refactor] Standardize KV cache layout (#51718)

2026-08-21 17:47:49 -07:00

test_mnnvl_alltoall.py

[MoE] Refine FlashInfer one-sided All2All integration (#51924)

2026-08-18 00:18:14 +00:00

test_mq_connect_ip.py

[Bugfix] Fix MessageQueue connect_ip for cross-node data parallelism (#35429)

2026-02-26 22:08:16 +00:00

test_multi_node_assignment.py

[CI][Ray] Fix flaky multi-node assignment test after placement-group teardown (#53621)

2026-08-29 23:48:22 +00:00

test_multiproc_executor.py

[Bugfix] Account for local DP workers in startup thread allocation (#52385)

2026-08-17 12:10:34 -07:00

test_nccl_symm_mem.py

[PERF] Extend NCCL symmetric memory to AllGather and ReduceScatter (#46703)

2026-06-30 11:38:18 -07:00

test_node_count.py

[Chore] Separate out vllm.utils.network_utils (#27164)

2025-10-19 03:06:32 -07:00

test_packed_tensor.py

[rl] Stateful Trainer Send: NCCL + Sparse NCCL [3/N] (#50902)

2026-08-07 00:35:59 -07:00

test_pipeline_parallel.py

[Model] Remove ten deprecated model architectures (#53608)

2026-08-24 17:06:00 -07:00

test_pipeline_partition.py

Convert formatting to use ruff instead of yapf + isort (#26247)

2025-10-05 07:06:22 -07:00

test_pp_cudagraph.py

[AMD][CI] Fix Pipeline + Context Parallelism test group (#46650)

2026-06-24 22:23:42 -05:00

test_pynccl.py

[Core] Release NCCL communicator memory in sleep mode (#51485)

2026-09-01 14:31:16 +08:00

test_quick_all_reduce.py

[ROCm][CI] Use larger atol value for INT3 in test_quick_all_reduce.py (#50450)

2026-07-30 21:41:44 -05:00

test_ray_v2_executor_e2e.py

[Feat][Executor] Introduce RayExecutorV2 (#36836)

2026-04-01 14:34:29 -07:00

test_ray_v2_executor.py

[Test] Assert co-located RayExecutorV2 stores publish distinct ports (#54310)

2026-08-28 22:13:31 -07:00

test_rocm_aiter_custom_ar.py

[Perf][ROCm] Add AITER custom AG/RS (DP only) (#48247)

2026-09-11 13:19:57 +08:00

test_rocm_quick_reduce.py

[communication] [bugfix] fix quickreduce acc error in cudagraph mode (#46913)

2026-07-27 01:38:45 -07:00

test_same_node.py

create is_in_the_same_node on cpu (#26832)

2025-10-21 02:04:14 +00:00

test_sharded_rdt_plan.py

[RL] P2P RDT weight sync (#43375)

2026-08-23 09:42:22 -07:00

test_sharded_rdt_producer.py

[RL] P2P RDT weight sync (#43375)

2026-08-23 09:42:22 -07:00

test_sharded_rdt_trainer.py

[RL] P2P RDT weight sync (#43375)

2026-08-23 09:42:22 -07:00

test_shm_broadcast.py

[Bugfix] Fix ZMQ port TOCTOU race in shm_broadcast MessageQueue (#50960)

2026-08-07 18:19:02 -07:00

test_shm_buffer.py

[Core] Cleanup shm based object store on engine shutdown (#32429)

2026-01-20 08:53:37 +00:00

test_shm_storage.py

[Multimodal] Simplify MM input definitions (#33331)

2026-01-29 13:32:04 +00:00

test_split_group.py

use split_group for pytorch process group creation (#41980)

2026-06-04 14:36:07 -04:00

test_symm_mem_allreduce.py

[Hardware] Replace torch.cuda.device_count/current_device/set_device API (#36145)

2026-03-12 07:57:47 -07:00

test_torchrun_example_moe.py

use split_group for pytorch process group creation (#41980)

2026-06-04 14:36:07 -04:00

test_torchrun_example.py

use split_group for pytorch process group creation (#41980)

2026-06-04 14:36:07 -04:00

test_utils.py

replace cuda_device_count_stateless() to current_platform.device_count() (#37841)

2026-03-31 22:32:54 +08:00

test_weight_transfer_nccl_uid.py

[Distributed] Support pre-shared ncclUniqueId rendezvous for weight transfer (#53784)

2026-09-01 14:55:57 +08:00

test_weight_transfer.py

[RL] Support checkpoint-coordinate sparse NCCL weight updates (#53751)

2026-08-27 18:19:51 +08:00