vllm-omni

duplex

[Frontend] Add server-side VAD for turn-based omni models (#6618 )

2026-09-08 03:17:00 -07:00

__init__.py

[Test] L2 & L3 Test Case Stratification Design for Omni Model (#1272 )

2026-02-12 09:56:35 +08:00

test_arg_utils_shared_fields.py

[Refactor][CI] Reorganize test tree, and move slow e2e to weekly (#4951 )

2026-07-29 23:53:03 +08:00

test_arg_utils.py

Add vLLM-Omni library info to Hugging Face Hub requests (#5381 )

2026-09-08 10:08:38 -06:00

test_async_omni_engine_abort_ack.py

[Bugfix] Fix colocate-async sleep admission race and deliver AR abort tokens (#6367 )

2026-08-29 09:11:39 +00:00

test_async_omni_engine_abort.py

[Refactor] Remove internal stage_configs_path plumbing (#5741 )

2026-08-13 21:29:03 +08:00

test_async_omni_engine_input.py

[Bugfix][Engine] Make abort idempotent during shutdown (#6327 )

2026-08-28 19:51:25 -07:00

test_async_omni_engine_outputs.py

[Core] Graduate MiniCPM-o 4.5 and PersonaPlex full-duplex serving out of experimental (#6196 )

2026-09-05 08:20:21 +00:00

test_async_omni_engine_stage_init.py

[Rebase] Rebase to vLLM 0.29.0 (#7230 )

2026-09-10 12:08:28 +08:00

test_build_engine_args_no_mutate.py

[Bugfix] Prevent build_engine_args_dict from mutating stage_config.engine_args (#6783 )

2026-09-01 16:45:28 +00:00

test_cfg_companion_lifecycle.py

[Bugfix] Per-replica fault isolation: keep API server alive on single-stage death (#4285 ) (#4583 )

2026-08-10 23:00:18 -04:00

test_cfg_companion_tracker.py

[FEAT] support multi-stage deployment (#2396 )

2026-05-06 19:33:18 +08:00

test_correlated_rpc_client.py

[Full Duplex] Feat: Support Full-Duplex realtime runtime & add MiniCPM-o 4.5 demo (#3907 )

2026-07-23 16:45:18 +08:00

test_cross_stage_lora.py

[Feature] End-to-end LoRA support for BAGEL (#2494 )

2026-04-09 01:33:55 +00:00

test_data_entry_keys.py

[Feature] Support text encoder Disaggregation for H3 (#5885 )

2026-08-28 14:27:05 +08:00

test_diffusion_streaming_output_kwarg.py

[Feature][LingBot World]stepwise execution (#6844 )

2026-09-07 13:38:23 +08:00

test_duplex_import_boundary.py

[Realtime] Emit current (non-beta) OpenAI audio/transcript event names (#7339 )

2026-09-11 08:50:12 +08:00

test_ensure_tensor_values.py

[Refactor] Output Processor Phase 2: separate multimodal output channel (#1601 ) (#2744 )

2026-06-12 15:19:14 +08:00

test_membership_controller.py

text and image metric (#6150 )

2026-08-26 02:54:45 +00:00

test_multimodal_accumulation.py

[Bugfix][Minicpm] Support full-duplex three-stage serving (#5380 )

2026-07-29 01:58:34 +08:00

test_omni_request_output.py

[Bugfix] Carry ec_transfer_params and num_cache_creation_tokens on OmniRequestOutput (#6152 )

2026-08-15 08:37:27 +00:00

test_omni_request.py

[New Model] Support IndexTTS 2.5 (#5957 )

2026-08-11 23:00:35 +00:00

test_orchestrator_error_handling.py

[Model][JoyAI-VL-Interaction] Add native multi-stage pipeline (#5352 )

2026-09-08 05:07:08 -07:00

test_orchestrator_event_driven.py

[Perf][Engine] Event-driven orchestration loop (opt-in) — S1 of #4855 (#5221 )

2026-08-26 22:15:25 +08:00

test_orchestrator_kv_sender_info.py

[Core] Add NIXL omni connector (#6093 )

2026-09-13 16:35:02 +08:00

test_orchestrator_monitor.py

[Core] Performance fix for orchestrator bottleneck: separates inter-stage from client outputs (#4527 )

2026-06-29 17:43:36 +02:00

test_orchestrator_segment_stage_keying.py

[Core] Graduate MiniCPM-o 4.5 and PersonaPlex full-duplex serving out of experimental (#6196 )

2026-09-05 08:20:21 +00:00

test_orchestrator_stage_input_bridge.py

[Core] Add NIXL omni connector (#6093 )

2026-09-13 16:35:02 +08:00

test_orchestrator.py

[CI/Build] Wait for duplex reaper recovery in retry test (#7225 )

2026-09-08 17:10:51 +08:00

test_output_modality.py

[Refactor][OutputProcessor 1/8]: clean up multimodal payload accumulation and move payload logic into MultimodalPayload & directory refactor (#4980 )

2026-07-15 15:50:19 +08:00

test_output_processor.py

[Bugfix] Fix colocate-async sleep admission race and deliver AR abort tokens (#6367 )

2026-08-29 09:11:39 +00:00

test_parallel_stage_init.py

[Bugfix][Engine] Group overlapping device stages into one sequential init component (#7328 )

2026-09-09 20:25:41 +08:00

test_rpc_result_router.py

[Feature]: Support pause / resume and sleep / wake for AR stages in AsyncOmni (#6084 )

2026-08-24 16:21:32 +08:00

test_single_stage_mode.py

[Frontend] Add server-side VAD for turn-based omni models (#6618 )

2026-09-08 03:17:00 -07:00

test_stage_device_layout.py

[Bugfix] Validate per-stage device layout before spawning workers (#5003 ) (#5742 )

2026-08-26 19:25:22 +00:00

test_stage_engine_args.py

[Perf][Diffusion] Optimize HSDP startup via Rank-0 shared weight loading and accelerated LoRA delta computation (#7005 )

2026-09-11 08:28:42 +00:00

test_stage_engine_core_client.py

[Refactor] Omni Stage Runtime and Distributed Replica Control Plane (#3855 )

2026-06-17 14:13:20 +08:00

test_stage_engine_core_proc.py

[perf] Minicpm o4.5 optimization (#5228 )

2026-07-24 17:20:39 +08:00

test_stage_engine_startup_cache_env.py

[Rebase] Rebase to vllm 0.27.0 (#5976 )

2026-08-11 15:23:46 +00:00

test_stage_metadata.py

[Bugfix] Default missing legacy stage input sources (#5498 )

2026-07-31 03:00:14 -07:00

test_stage_pool_collective_rpc.py

[Bugfix] Fix colocate-async sleep admission race and deliver AR abort tokens (#6367 )

2026-08-29 09:11:39 +00:00

test_stage_request_token_budget.py

[Bugfix][MiniCPM-o] Stop Talker mid-utterance truncation and load cod… (#6346 )

2026-08-20 11:50:07 +08:00

test_stage_runtime_launch_concurrency.py

[Bugfix] Allow LLM replicas on different GPUs to initialize concurrently (#7292 )

2026-09-10 09:54:20 +08:00

test_strategy_lb_policy.py

[Core] Composable parallel: Phase 1 strategy overlay (TP/DP/stage-replica) (RFC #4084 ) (#4281 )

2026-07-10 21:29:24 -06:00

test_wire_multimodal_output.py

[Refactor] Output Processor Phase 2: separate multimodal output channel (#1601 ) (#2744 )

2026-06-12 15:19:14 +08:00