Liangsheng Yin 0c00220795 tiny unify environ usage (#15335)

2025-12-17 23:31:43 +08:00

..

async_dynamic_batch_tokenizer.py

…

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

[Auto Sync] Update data_parallel_controller.py, detokenizer... (20251209) (#14759)

2025-12-09 18:38:38 -08:00

detokenizer_manager.py

fix(function_call): fallback to decode when batch decode options differ (#15155)

2025-12-16 20:21:08 -08:00

disagg_service.py

…

io_struct.py

feat: support EPD disaggregation (#12263)

2025-12-14 22:30:08 +08:00

mm_utils.py

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

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

multi_tokenizer_mixin.py

refine stdout logging codes (#13015)

2025-11-12 00:16:14 +08:00

multimodal_processor.py

feat: support external custom models (#13429)

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

overlap_utils.py

Minor style fixes to the scheduler.py (#15218)

2025-12-16 17:09:44 -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

Minor style fixes to the scheduler.py (#15218)

2025-12-16 17:09:44 -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

…

scheduler_metrics_mixin.py

Add realtime token counter metrics (#15198)

2025-12-17 21:20:24 +08:00

scheduler_output_processor_mixin.py

Add cuda_graph_forward_passes_total and num_retracted_reqs_total (#15189)

2025-12-17 21:19:42 +08:00

scheduler_pp_mixin.py

[PP] Minor code cleanup for Pipeline Parallelism (#15329)

2025-12-17 23:31:25 +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

Minor style fixes to the scheduler.py (#15218)

2025-12-16 17:09:44 -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

Add cuda_graph_forward_passes_total and num_retracted_reqs_total (#15189)

2025-12-17 21:19:42 +08:00

session_controller.py

…

template_manager.py

…

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

tiny unify environ usage (#15335)

2025-12-17 23:31:43 +08:00

tp_worker.py

Minor style fixes to the scheduler.py (#15218)

2025-12-16 17:09:44 -08:00

utils.py

[Qwen3-next] support mamba radix cache for overlap scheduler (#14792)

2025-12-14 18:54:16 -08:00