freeliuzc 31e2a8bbad [Speculative Decoding] Support mtp super ultra overlap in pd-split mode with insert_task overlap (#7323) ...

* support mtp overlap in pd-split mode with insert_task overlap

2026-04-13 19:41:17 +08:00

..

__init__.py

[LLM] First commit the llm deployment code

2025-06-09 19:20:15 +08:00

dcu_model_runner.py

[FDConfig]Remove splitwise_role and engine_worker_queue_port in FDConfig (#4147)

2025-09-19 17:01:52 +08:00

dcu_worker.py

Support setting communication groups in custom_allreduce and the all-to-all\transpose fused operator during the decoding phase. (#5917)

2026-01-12 14:09:39 +08:00

eplb.py

polish code with new pre-commit rule (#2923)

2025-07-19 23:19:27 +08:00

experts_manager.py

[Feature] Support redundant expert for eplb (#5918)

2026-01-09 17:13:24 +08:00

gcu_model_runner.py

Split enable_mm (#7183)

2026-04-08 11:25:41 +08:00

gcu_worker.py

[Feature]Report FD statistical information (#5646)

2026-01-14 17:54:01 +08:00

gpu_model_runner.py

[Speculative Decoding] Support mtp super ultra overlap in pd-split mode with insert_task overlap (#7323)

2026-04-13 19:41:17 +08:00

gpu_worker.py

[FDConfig] Support CLI args for quantization params and add cudagraph validation (#7281)

2026-04-10 14:13:42 +08:00

hpu_model_runner.py

[BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533)

2026-03-16 21:32:43 +08:00

hpu_worker.py

[Feature]Report FD statistical information (#5646)

2026-01-14 17:54:01 +08:00

iluvatar_model_runner.py

[Iluvatar] Fix cuda graph error for tp > 1 in ernie models (#7126)

2026-04-01 19:13:34 +08:00

iluvatar_worker.py

Split enable_mm (#7183)

2026-04-08 11:25:41 +08:00

input_batch.py

Split enable_mm (#7183)

2026-04-08 11:25:41 +08:00

metax_model_runner.py

Split enable_mm (#7183)

2026-04-08 11:25:41 +08:00

metax_worker.py

[RL] Adapt async rollout checkpoint update flow (#7042)

2026-03-30 19:19:34 +08:00

model_runner_base.py

[Graph Optimization][CINN] Use CINN in PaddleOCR-VL ViT part (#5223)

2025-12-09 14:37:00 +08:00

output.py

[Feature] GPU Memory Optimization and Retirement of V0 Scheduler (#6407)

2026-02-28 15:07:43 +08:00

tbo.py

[TBO] Apply tbo to gpu_model_runner (#7165)

2026-04-08 16:55:17 +08:00

worker_base.py

[FDConfig]Remove splitwise_role and engine_worker_queue_port in FDConfig (#4147)

2025-09-19 17:01:52 +08:00

worker_process.py

[Cleanup] Replace torch proxy alias with public compat API (#7348)

2026-04-13 11:43:26 +08:00

xpu_model_runner.py

[XPU] Add TP broadcast after sampling in XPU model runner to ensure consistent results across ranks. (#7096)

2026-04-08 19:26:53 +08:00

xpu_worker.py

[BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533)

2026-03-16 21:32:43 +08:00