fzyzcjy dd176387c6 Add scripted-runtime harness core and wire scheduler/IPC hooks (#27411)

2026-06-06 09:07:01 +08:00

..

scheduler_components

Add scripted-runtime harness core and wire scheduler/IPC hooks (#27411)

2026-06-06 09:07:01 +08:00

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

[UnifiedTree]: Support l3 storage for swa and deepseek v4 (#26881)

2026-06-04 10:17:34 +08:00

communicator.py

refactor: extract FanOutCommunicator and use declarative spec table (#22967)

2026-04-16 15:37:19 -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

fix(load-snapshot): avoid duplicate zmq bind in multi-tokenizer mode (#27145)

2026-06-03 18:24:56 -07:00

detokenizer_manager.py

Optimize streaming detokenizer updates (#24659)

2026-06-02 23:44:21 -07:00

disagg_service.py

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

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

embed_types.py

[Score API] Hoist query placeholder scan and specialize PositionalEmbeds stacking (#23513)

2026-04-29 13:51:53 -07:00

hisparse_coordinator.py

[PD & HiSparse] Add DeepSeek V4 support for HiSparse direct Prefill-to-Decode DRAM (#24880)

2026-06-05 15:39:48 +08:00

io_struct.py

[3/N][Sync sglang-miles] TITO Support (#23751)

2026-06-03 21:45:33 -04:00

load_snapshot.py

Add num_waiting_uncached_tokens load metric (#27174)

2026-06-03 18:29:49 -07:00

mm_utils.py

Fix weights_checker checksum for 0-dim tensors and multi-GPU (#26863)

2026-05-31 21:27:03 -07:00

multi_tokenizer_mixin.py

fix(load-snapshot): avoid duplicate zmq bind in multi-tokenizer mode (#27145)

2026-06-03 18:24:56 -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

[Scheduler] Defer prefill input_ids H2D to forward stream, unify resolve via future_map (#25945)

2026-05-30 02:58:13 -07:00

prefill_delayer.py

[PrefillDelayer] support NCCL all-gather for cross-DP info sync (#24768)

2026-05-09 21:20:03 -07:00

schedule_batch.py

Add num_waiting_uncached_tokens load metric (#27174)

2026-06-03 18:29:49 -07:00

schedule_policy.py

[mem_cache][2/N] refactor: move SWATokenToKVPoolAllocator to allocator/swa.py (#26676)

2026-06-04 15:10:37 +08:00

scheduler_input_blocker.py

Tiny move files to utils folder (#11166)

2025-10-03 22:40:06 +08:00

scheduler_pp_mixin.py

[HiSparse PD & PP]Fix HiSparse compatibility with PP decode (#27258)

2026-06-04 21:37:42 +08:00

scheduler_recv_skipper.py

[Feat]Add scheduler recv skipper weights to environment configuration (#13855)

2025-11-27 18:16:11 +08:00

scheduler.py

Add scripted-runtime harness core and wire scheduler/IPC hooks (#27411)

2026-06-06 09:07:01 +08:00

template_detection.py

Clean up server startup log noise (#26205)

2026-05-24 14:35:15 -07:00

template_manager.py

Clean up server startup log noise (#26205)

2026-05-24 14:35:15 -07:00

tokenizer_control_mixin.py

Fix weights_checker checksum for 0-dim tensors and multi-GPU (#26863)

2026-05-31 21:27:03 -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

Don't write crash dump on graceful exit (#27405)

2026-06-05 17:06:39 -07:00

tp_worker.py

Replace skip_attn_backend_init with a batch-carried attention plan marker (+ staleness re-plan) (#27193)

2026-06-04 17:13:18 -07:00

utils.py

[SPEC] fix: use effective max draft tokens for adaptive spec initiali… (#26354)

2026-05-28 13:33:11 -07:00