freeliuzc 22a4f6019d [Speculative Decoding][BugFix] Fix apply repeat times penalty kernel and change spec default verify strategy (#7467) ...

* fix repeat_time kernel and change default spec verify strategy

* fix unit_test

2026-04-18 00:38:01 +08:00

..

batch_invariant

[Optimization] enable trtllm_all_reduce fusion kernel in glm model (#6660)

2026-04-16 14:10:19 +08:00

benchmarks

…

cache_manager

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

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

ce

…

ci_use

[Others] Fix typo (#7280)

2026-04-14 17:28:22 +08:00

deterministic

…

distributed

[Optimization][DeepSeekV3.2]Reducing slot_mapping compute frequency from twice per layer to a single pre-processing step. (#7367)

2026-04-16 19:54:12 +08:00

e2e

[CI] Add pytest failure log collection and persistence (#7405)

2026-04-16 22:56:17 +08:00

engine

[Optimization] enable trtllm_all_reduce fusion kernel in glm model (#6660)

2026-04-16 14:10:19 +08:00

entrypoints

[Feature] implement log channel separation and request log level system (#7190)

2026-04-16 15:13:05 +08:00

eplb

[Others] Fix typo (#7280)

2026-04-14 17:28:22 +08:00

graph_optimization

[RL] Add clear_graph_opt_backend for glm4_mtp (#7378)

2026-04-15 19:44:15 +08:00

input

[Feature] implement log channel separation and request log level system (#7190)

2026-04-16 15:13:05 +08:00

inter_communicator

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

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

layers

[Speculative Decoding][BugFix] Fix apply repeat times penalty kernel and change spec default verify strategy (#7467)

2026-04-18 00:38:01 +08:00

logger

[Feature] implement log channel separation and request log level system (#7190)

2026-04-16 15:13:05 +08:00

metax_ci

[Others] Fix typo (#7280)

2026-04-14 17:28:22 +08:00

metrics

…

model_executor

[Optimization] enable trtllm_all_reduce fusion kernel in glm model (#6660)

2026-04-16 14:10:19 +08:00

model_loader

…

multimodal

…

operators

[Speculative Decoding][BugFix] Fix apply repeat times penalty kernel and change spec default verify strategy (#7467)

2026-04-18 00:38:01 +08:00

output

[Feature] implement log channel separation and request log level system (#7190)

2026-04-16 15:13:05 +08:00

platforms

…

plugins

…

pooling

…

quantization

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

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

reasoning

…

rl

…

router

…

scheduler

[Feature] implement log channel separation and request log level system (#7190)

2026-04-16 15:13:05 +08:00

spec_decode

…

splitwise

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

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

trace

…

usage

…

utils

…

v1

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

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

worker

[Optimization][DeepSeekV3.2]Reducing slot_mapping compute frequency from twice per layer to a single pre-processing step. (#7367)

2026-04-16 19:54:12 +08:00

xpu_ci

[XPU] Split the block_attn operator into smaller operators (#6798)

2026-04-16 14:28:40 +08:00

conftest.py

[CI] Add pytest failure log collection and persistence (#7405)

2026-04-16 22:56:17 +08:00

cov_pytest.ini

…

test_envs.py

…

utils.py

…