Alison Shao e2fccb2ee0 Fix flaky Qwen3-Next KL divergence tests by reverting mamba slot release (#18910)

2026-02-19 07:55:16 +08:00

..

async_dynamic_batch_tokenizer.py

…

async_mm_data_processor.py

…

cache_controller.py

[HiCache][HA 1/N] Support HiCache storage runtime attach/detach (#15892)

2026-01-26 19:33:19 -08:00

configure_logging.py

…

data_parallel_controller.py

refactor FAKE transfer backend and remove --disaggregation-decode-enable-fake-auto parameter (#18345)

2026-02-16 17:27:02 +03:00

detokenizer_manager.py

Add spec_accept_histogram request statistic (#18332)

2026-02-12 21:09:21 +08:00

disagg_service.py

refactor(codespell): corrects typos covered up by whitelist (#17601)

2026-01-22 11:45:47 -08:00

io_struct.py

[misc] adding metadata field in UpdateWeightFromDiskReqInput (#18821)

2026-02-17 12:14:15 +08:00

mm_utils.py

model: support Qwen3.5 (#18489)

2026-02-10 00:27:59 +08:00

multi_tokenizer_mixin.py

Add spec_accept_histogram request statistic (#18332)

2026-02-12 21:09:21 +08:00

multimodal_processor.py

Skip mm feature pool init to avoid EPD OOM (#16388)

2026-01-23 10:53:45 +08:00

overlap_utils.py

Fix Bug on dsv3.2 (#18553)

2026-02-11 14:39:01 +08:00

prefill_delayer.py

Support token low usage watermark in prefill delayer (#16814)

2026-01-09 23:07:18 +08:00

request_metrics_exporter.py

[Auto Sync] Update request_metrics_exporter.py (20251230) (#16183)

2025-12-30 11:09:41 -08:00

schedule_batch.py

Add spec_accept_histogram request statistic (#18332)

2026-02-12 21:09:21 +08:00

schedule_policy.py

fix double-free kv cache for requests that have already finished and been freed during preemption (#18694)

2026-02-13 13:17:44 -08:00

scheduler_dp_attn_mixin.py

cleanup prefill metrics logging to fix dp-attn metrics (#18778)

2026-02-17 14:35:35 -08:00

scheduler_input_blocker.py

…

scheduler_metrics_mixin.py

cleanup prefill metrics logging to fix dp-attn metrics (#18778)

2026-02-17 14:35:35 -08:00

scheduler_output_processor_mixin.py

cleanup prefill metrics logging to fix dp-attn metrics (#18778)

2026-02-17 14:35:35 -08:00

scheduler_pp_mixin.py

cleanup prefill metrics logging to fix dp-attn metrics (#18778)

2026-02-17 14:35:35 -08:00

scheduler_profiler_mixin.py

[NPU] torch_npu profiler tensorboard path type fix (#17545)

2026-01-24 22:55:49 +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

Refactor prefix cache type checking (#17028)

2026-01-15 11:28:13 +08:00

scheduler_update_weights_mixin.py

refactor: add type hints to scheduler mixins (#15913)

2025-12-26 16:50:07 -08:00

scheduler.py

Fix flaky Qwen3-Next KL divergence tests by reverting mamba slot release (#18910)

2026-02-19 07:55:16 +08:00

session_controller.py

refactor(codespell): corrects typos covered up by whitelist (#17601)

2026-01-22 11:45:47 -08:00

template_manager.py

fix: Handle multiple named chat templates in HuggingFace tokenizers (#17236)

2026-01-18 17:20:04 +08:00

tokenizer_communicator_mixin.py

[HiCache][HA 1/N] Support HiCache storage runtime attach/detach (#15892)

2026-01-26 19:33:19 -08:00

tokenizer_manager_multiitem_mixin.py

feat: support qwen3(-VL) rerank scoring&chat template (#16403)

2026-01-15 00:45:46 +08:00

tokenizer_manager.py

[Auto Sync] Update grpc_request_manager.py, tokenizer_manag... (20260214) (#18838)

2026-02-14 18:12:32 -08:00

tp_worker.py

update pre-commit config (#18860)

2026-02-16 00:18:31 +08:00

utils.py

Introduce global alloc_len_per_decode & clean check decode memory (#15115)

2026-01-26 10:26:20 -08:00