sglang

async_dynamic_batch_tokenizer.py

[Performance] Dynamic Batch Tokenizer (#9382 )

2025-09-14 01:56:04 +08:00

async_mm_data_processor.py

[VLM] Optimize async mm data process mechanism (#12066 )

2025-11-01 01:24:53 +08:00

cache_controller.py

[BugFix] fix prefixcache performance and accuracy on ascend (#13573 )

2025-12-08 02:16:20 +08:00

configure_logging.py

…

data_parallel_controller.py

Support soft watchdog for tokenizer/detokenizer/dp-controller processes (#15607 )

2025-12-22 16:59:16 +08:00

detokenizer_manager.py

Support soft watchdog for tokenizer/detokenizer/dp-controller processes (#15607 )

2025-12-22 16:59:16 +08:00

disagg_service.py

[1/2] Refactor multi-tokenizer manager (#10074 )

2025-09-07 19:13:34 +08:00

io_struct.py

[Feature] Enable return routed experts (#12162 )

2025-12-21 15:16:43 +08:00

mm_utils.py

[VLM] feat: support chunked vit attention (#14907 )

2025-12-15 12:11:02 +08:00

multi_tokenizer_mixin.py

[Feature] Enable return routed experts (#12162 )

2025-12-21 15:16:43 +08:00

multimodal_processor.py

feat: support external custom models (#13429 )

2025-11-20 11:16:58 +08:00

overlap_utils.py

[Feature] Xiaomi MiMo-V2-Flash day0 support (#15207 )

2025-12-19 11:40:07 +08:00

request_metrics_exporter.py

Add RequestMetricsExporter utility to export request-level metrics (#10973 )

2025-11-12 16:33:37 -08:00

schedule_batch.py

[Feature] Enable return routed experts (#12162 )

2025-12-21 15:16:43 +08:00

schedule_policy.py

[PP] Add pp support for Qwen3-VL (#12333 )

2025-12-17 16:03:58 +08:00

scheduler_dp_attn_mixin.py

[scheduler] remove scheduler allgather for best throughout (#14294 )

2025-12-14 16:37:49 +08:00

scheduler_enhancer.py

[scheduler] enhance scheduler in dp_attention mixed case with spec (#14201 )

2025-12-14 02:52:26 +08:00

scheduler_input_blocker.py

Tiny move files to utils folder (#11166 )

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

scheduler_metrics_mixin.py

Support EPLB balancedness prometheus metric without GPU->CPU synchronize (#15401 )

2025-12-18 22:24:23 +08:00

scheduler_output_processor_mixin.py

[Feature] Enable return routed experts (#12162 )

2025-12-21 15:16:43 +08:00

scheduler_pp_mixin.py

Tiny extract ModelRunnerOutput (#15400 )

2025-12-18 22:18:45 +08:00

scheduler_profiler_mixin.py

Tiny call cudaProfilerStart only on first rank in node (#14211 )

2025-12-01 18:18:45 +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_runtime_checker_mixin.py

Support soft watchdog for tokenizer/detokenizer/dp-controller processes (#15607 )

2025-12-22 16:59:16 +08:00

scheduler_update_weights_mixin.py

Support sanity checking weight consistency especially for RL (#13854 )

2025-11-27 20:25:12 +08:00

scheduler.py

Support soft watchdog for tokenizer/detokenizer/dp-controller processes (#15607 )

2025-12-22 16:59:16 +08:00

session_controller.py

[Test] Fix session control test (#12336 )

2025-10-29 18:28:04 +08:00

template_manager.py

Move parsers under a single folder (#9912 )

2025-09-02 18:25:04 -07:00

tokenizer_communicator_mixin.py

[fix] Fix issues for in-flight weight updates (#14064 )

2025-12-09 18:12:19 -08:00

tokenizer_manager_multiitem_mixin.py

move multi-item scoring functions in tokenizer manager into a separate file (#14740 )

2025-12-09 14:47:06 -08:00

tokenizer_manager.py

Support soft watchdog for tokenizer/detokenizer/dp-controller processes (#15607 )

2025-12-22 16:59:16 +08:00

tp_worker.py

Fix type mismatch in LoRA batch validation causing assertion failures (#15427 )

2025-12-21 15:54:41 -08:00

utils.py

Support EPLB balancedness prometheus metric without GPU->CPU synchronize (#15401 )

2025-12-18 22:24:23 +08:00