Yuhao Yang 37a830b667 [Feature] Add DWDP (Distributed Weight Data Parallelism) for MoE prefill (#29778)

2026-07-20 23:59:54 -07:00

..

scheduler_components

[Feature] Add DWDP (Distributed Weight Data Parallelism) for MoE prefill (#29778)

2026-07-20 23:59:54 -07:00

async_dynamic_batch_tokenizer.py

[tokenizer] improve non streaming request processing + some small fixes. (#20310)

2026-04-10 15:46:12 -07:00

cache_controller.py

[HiCache] Optimize L2 mem allocation when cache miss in L3 (#19320)

2026-07-16 15:00:53 -07:00

communicator.py

[1/N] elastic-ep: Add runtime EP scale-up (#30164)

2026-07-16 15:53:44 -07:00

configure_logging.py

[Utils] Support configure log level at runtime (#26583)

2026-05-29 14:49:06 -07:00

data_parallel_controller.py

[1/N] elastic-ep: Add runtime EP scale-up (#30164)

2026-07-16 15:53:44 -07:00

detokenizer_manager.py

Add Inkling model support (#31681)

2026-07-19 22:57:37 -07:00

disagg_service.py

[PD] Cleanup BootstrapServer init and ready check (#19551)

2026-02-28 16:41:42 +08:00

embed_types.py

Convert IPC dataclasses to msgspec.Struct with opt-in msgpack transport (#28688)

2026-06-26 12:04:03 -07:00

hisparse_coordinator.py

Introduce req.kv container for coupled owned kv field lifecycle (#29427)

2026-07-15 14:40:38 +08:00

io_struct.py

Add Inkling model support (#31681)

2026-07-19 22:57:37 -07:00

load_snapshot.py

refactor(load-snapshot): build LoadSnapshot directly, drop legacy get_loads IPC (#30525)

2026-07-09 05:58:10 -07:00

min_free_slots_delayer.py

[Spec] Add DSpark: confidence-scheduled speculative decoding (#30261)

2026-07-12 17:25:26 -05:00

mm_utils.py

[NPU] Fix vit graph tnd cu seqlens (#31701)

2026-07-20 16:20:33 +08:00

multi_tokenizer_mixin.py

[Fix] Wire the detokenizer soft watchdog into the multi-http-worker event loop (#31392)

2026-07-15 21:55:46 -07:00

multimodal_processor.py

[Feature] Stronger transformers modeling backend with TP, PP, MoE, VLMs, and torch compile (#19163)

2026-04-02 16:02:33 -07:00

overlap_utils.py

[bug-fix] Stabilize GLM-5.2 MTP IndexShare across PD and CUDA graph replay (#30839)

2026-07-13 19:37:07 -07:00

prefill_delayer.py

Fix prefill delayer wait histograms always observing 0 (#25975)

2026-06-14 00:35:29 -07:00

schedule_batch.py

Fix stop boundaries for grammar-constrained speculative decoding (#31738)

2026-07-20 17:55:27 -07:00

schedule_policy.py

Add Inkling model support (#31681)

2026-07-19 22:57:37 -07:00

scheduler_input_blocker.py

[Cleanup] IPC struct renames, better typing, and SenderWrapper removal (#29214)

2026-06-24 17:25:24 -07:00

scheduler_pp_mixin.py

[Bugfix] Release Mamba cache after PP dynamic chunk profiling (#31321)

2026-07-16 02:32:07 +08:00

scheduler.py

Overlap grammar (constrained decoding) with speculative decode verify (#31488)

2026-07-20 17:36:05 -07:00

tokenizer_control_mixin.py

[1/N] elastic-ep: Add runtime EP scale-up (#30164)

2026-07-16 15:53:44 -07:00

tokenizer_manager_score_mixin.py

Replace single-line defensive getattrs with direct access (#25435)

2026-05-16 09:18:43 +08:00

tokenizer_manager.py

[1/N] elastic-ep: Add runtime EP scale-up (#30164)

2026-07-16 15:53:44 -07:00

tp_worker.py

Add Inkling model support (#31681)

2026-07-19 22:57:37 -07:00

utils.py

Overlap grammar (constrained decoding) with speculative decode verify (#31488)

2026-07-20 17:36:05 -07:00