Zheng Wengang 2d120f8b18 [Feature][Multimodal] Implement LRU cache for multimodal embeddings (#8292) ...

Signed-off-by: Xinyuan Tong <xinyuantong.cs@gmail.com>
Co-authored-by: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com>
Co-authored-by: Xinyuan Tong <xinyuantong.cs@gmail.com>

2025-08-06 23:21:40 -07:00

..

configs

…

cpu

[CPU][sgl-kernel] biased_grouped_topk: fix correction_bias dtype to float32 (#8212)

2025-08-04 18:28:31 -07:00

models

[Feature][Multimodal] Implement LRU cache for multimodal embeddings (#8292)

2025-08-06 23:21:40 -07:00

openai_server

GLM-4.5 Model Support Follow-up (#8445)

2025-07-27 23:35:20 -07:00

double-sparsity-config-Llama-3.1-8B-Instruct.json

…

experiment_runner.py

…

kv_cache_scales_llama3_1_8b.json

…

kv_cache_scales_llama3_8b.json

…

kv_cache_scales_qwen2_1_5b.json

…

parse_results.py

…

run_suite.py

Support B200 in CI (#8861)

2025-08-06 21:42:44 +08:00

test_abort.py

Support updating weights at once by stopping all requests (#6698)

2025-07-02 22:26:06 -07:00

test_ascend_mla_w8a8int8.py

[CI] Ascend NPU CI enhancement (#8294)

2025-08-03 22:16:38 -07:00

test_ascend_tp1_bf16.py

[CI] Ascend NPU CI enhancement (#8294)

2025-08-03 22:16:38 -07:00

test_ascend_tp2_bf16.py

[CI] Ascend NPU CI enhancement (#8294)

2025-08-03 22:16:38 -07:00

test_awq_dequant.py

[AMD] Add triton awq_dequantize kernel to support AWQ on ROCm (#7661)

2025-07-18 14:27:25 -07:00

test_awq.py

…

test_bench_one_batch.py

Add V2-lite model test (#7390)

2025-07-03 22:25:50 -07:00

test_bench_serving.py

Support overlapped lora updates (#8213)

2025-07-27 13:00:44 -07:00

test_block_int8.py

[1/N] MoE Refactor: refactor select_experts (#7966)

2025-07-19 00:51:15 -07:00

test_bnb.py

Implement Siglip Vision model, and support BNB quantization for gemma3-mm (#5339)

2025-05-20 23:53:46 -07:00

test_chunked_prefill.py

…

test_cpp_radix_cache.py

[Feature] Radix Tree in C++ (#7369)

2025-08-02 19:50:14 -07:00

test_create_kvindices.py

…

test_custom_allreduce.py

[AMD] Add unit-test-sgl-kernel-amd to AMD CI (#7539)

2025-06-29 15:50:09 -07:00

test_data_parallelism.py

…

test_deepep_internode.py

…

test_deepep_intranode.py

…

test_deepep_large.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_deepep_low_latency.py

…

test_deepep_small.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_disaggregation_different_tp.py

[CI] Fix setup of disaggregation with different tp (#6706)

2025-05-28 11:17:27 -07:00

test_disaggregation.py

[PD] Support decode retract and update decode.py (#7196)

2025-06-14 19:48:05 -07:00

test_double_sparsity.py

…

test_dp_attention.py

[feat] support minimum token load balance in dp attention (#7379)

2025-08-03 00:46:47 -07:00

test_eagle_infer_a.py

Improve test cases for eagle infer (#7173)

2025-06-13 22:25:13 -07:00

test_eagle_infer_b.py

[EAGLE] Refactor code for page size > 1 & more simplifications (#7213)

2025-06-16 03:04:29 -07:00

test_ebnf_constrained.py

ci: Fix test_ebnf_generate_all_optional_function_params (#7288)

2025-06-17 16:39:42 -07:00

test_eplb.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_eval_accuracy_large.py

Fix GPU OOM (#6564)

2025-05-24 16:38:39 -07:00

test_eval_fp8_accuracy.py

…

test_expert_distribution.py

Support gathering expert distribution details (#6665)

2025-05-27 15:32:59 -07:00

test_expert_location_updater.py

Move files related to EPLB (#7580)

2025-06-29 15:39:38 -07:00

test_fa3.py

[Eagle] Fix kernel call after updating speculative sampling kernels (#7231)

2025-06-16 07:25:59 -07:00

test_fim_completion.py

…

test_flashmla.py

Add V2-lite model test (#7390)

2025-07-03 22:25:50 -07:00

test_forward_split_prefill.py

[Feature] Add a test for Layer-wise Prefill (#8231)

2025-07-21 22:06:15 +08:00

test_fp8_kernel.py

…

test_fp8_kvcache.py

…

test_full_deepseek_v3.py

Simplify prepare_extend_after_decode (#6987)

2025-06-09 16:39:21 -07:00

test_function_call_parser.py

GLM-4.5 Model Support Follow-up (#8445)

2025-07-27 23:35:20 -07:00

test_fused_moe.py

[1/N] MoE Refactor: refactor select_experts (#7966)

2025-07-19 00:51:15 -07:00

test_get_weights_by_name.py

…

test_gguf.py

…

test_gptqmodel_dynamic.py

[3/n] chore: decouple AWQ implementation from vLLM dependency (#8113)

2025-07-18 11:45:22 -07:00

test_health_check.py

…

test_hicache_mla.py

…

test_hicache_page.py

Move host memory pools into a separate file (#7200)

2025-06-14 21:31:42 -07:00

test_hicache_storage.py

Hicache Storage Layer Prototype (#7704)

2025-07-18 15:20:19 +08:00

test_hicache.py

…

test_hidden_states.py

Open AI API hidden states (#6716)

2025-06-10 14:37:29 -07:00

test_hybrid_attn_backend.py

[feat] Support different attention backends for prefill and decode (#6338)

2025-07-28 11:42:29 +08:00

test_hybrid_dp_ep_tp_mtp.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_input_embeddings.py

…

test_int8_kernel.py

[1/N] MoE Refactor: refactor select_experts (#7966)

2025-07-19 00:51:15 -07:00

test_intel_amx_attention_backend.py

Add V2-lite model test (#7390)

2025-07-03 22:25:50 -07:00

test_io_struct.py

fix: Handles input_embeds in GenerateReqInput when n>1 (#7830)

2025-07-08 14:00:42 -07:00

test_jinja_template_utils.py

[feat] Add detail in image_data (#8596)

2025-08-05 14:01:38 +08:00

test_kv_events.py

feat: add dp-rank to KV events (#6852)

2025-06-04 15:29:34 -07:00

test_local_attn.py

…

test_metrics.py

…

test_mla_deepseek_v3.py

Fix FP8 KV Cache Support in FA3 Backend (#7148)

2025-06-25 02:14:40 -07:00

test_mla_flashinfer.py

Fix CI and triton moe Configs (#6974)

2025-06-08 05:06:46 -07:00

test_mla_fp8.py

…

test_mla_int8_deepseek_v3.py

…

test_mla_tp.py

…

test_mla.py

Improve profiler and integrate profiler in bench_one_batch_server (#6787)

2025-05-31 15:53:55 -07:00

test_modelopt_fp8kvcache.py

…

test_modelopt.py

…

test_models_from_modelscope.py

…

test_moe_deepep_eval_accuracy_large.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_moe_deepep.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_moe_ep.py

[5/N] MoE Refactor: Update MoE parallelism arguments (#8658)

2025-08-01 01:20:03 -07:00

test_moe_eval_accuracy_large.py

…

test_mscclpp.py

support 1 shot allreduce in 1-node and 2-node using mscclpp (#6277)

2025-06-04 22:11:24 -07:00

test_multi_instance_release_memory_occupation.py

Add e2e test for multi instance multi stage memory release/resume occupuation (#7208)

2025-06-26 17:40:38 -07:00

test_nightly_gsm8k_eval_amd.py

Update amd nightly models CI. (#6992)

2025-06-09 10:54:08 -07:00

test_nightly_gsm8k_eval.py

…

test_no_chunked_prefill.py

…

test_no_overlap_scheduler.py

…

test_page_size.py

…

test_patch_torch.py

…

test_penalty.py

…

test_pp_single_node.py

Improve profiler and integrate profiler in bench_one_batch_server (#6787)

2025-05-31 15:53:55 -07:00

test_pytorch_sampling_backend.py

…

test_quick_allreduce.py

[Feature] Integrate quick allreduce and select the best allreduce implementation (#6619)

2025-07-24 20:48:42 -07:00

test_radix_attention.py

Improve profiler and integrate profiler in bench_one_batch_server (#6787)

2025-05-31 15:53:55 -07:00

test_reasoning_parser.py

Enables force reasoning based on chat template for Qwen3-Thinking (#8369)

2025-08-06 20:02:47 -07:00

test_regex_constrained.py

…

test_release_memory_occupation.py

Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (#7099)

2025-06-19 00:56:37 -07:00

test_request_queue_validation.py

feat: throttle requests at scheduler based on --max_queued_requests (#7565)

2025-07-28 22:32:33 +08:00

test_retract_decode.py

…

test_rope_rocm.py

[AMD] Add test_fused_moe.py and test_rope_rocm.py to AMD CI (#5246)

2025-07-06 01:47:16 -07:00

test_sagemaker_server.py

…

test_schedule_policy.py

…

test_score_api.py

Decoder-only Scoring API (#6460)

2025-06-04 14:14:54 -07:00

test_server_args.py

ci: fix port args (#7792)

2025-07-05 15:06:42 -07:00

test_session_control.py

Move mem_fraction_static adjustment for multimodal models to server_args.py & Fix session control & Other cleanups (#7748)

2025-07-04 16:33:33 -07:00

test_skip_tokenizer_init.py

Support precomputed multimodal features for Qwen-VL and Gemma3 models. (#6136)

2025-05-16 12:26:15 -07:00

test_srt_endpoint.py

[Feature] Add Logit Bias (#6579)

2025-06-10 15:39:25 -07:00

test_srt_engine_with_quant_args.py

Revert "fix some typos" (#6244)

2025-05-12 12:53:26 -07:00

test_srt_engine.py

chore: upgrade flashinfer v0.2.6.post1 jit (#6958)

2025-06-09 09:22:39 -07:00

test_start_profile.py

[feature] Add start step profile argument in /start_profile (#7608)

2025-07-09 18:42:15 -07:00

test_swa_unittest.py

SWA Prefix Cache (#7367)

2025-07-13 12:31:07 -07:00

test_torch_compile_moe.py

Fix throughput threshold for amd ci test (#6414)

2025-05-19 14:17:41 -07:00

test_torch_compile.py

Improve profiler and integrate profiler in bench_one_batch_server (#6787)

2025-05-31 15:53:55 -07:00

test_torch_native_attention_backend.py

…

test_torch_tp.py

…

test_torchao.py

…

test_triton_attention_backend.py

…

test_triton_attention_kernels.py

Add unit test for triton swa kernel (#8853)

2025-08-06 16:10:38 +08:00

test_triton_attention_rocm_mla.py

…

test_triton_fused_moe.py

Integrate triton moe kernel (#7689)

2025-07-06 20:05:49 -07:00

test_triton_moe_channel_fp8_kernel.py

[1/N] MoE Refactor: refactor select_experts (#7966)

2025-07-19 00:51:15 -07:00

test_triton_moe_wna16.py

[1/N] MoE Refactor: refactor select_experts (#7966)

2025-07-19 00:51:15 -07:00

test_triton_sliding_window.py

Fix triton sliding window test case (#6981)

2025-06-08 17:20:46 -07:00

test_two_batch_overlap.py

[Feature] improve TBO: two chunk overlap (#8144)

2025-08-05 21:11:01 -07:00

test_update_weights_from_disk.py

Support updating weights at once by stopping all requests (#6698)

2025-07-02 22:26:06 -07:00

test_update_weights_from_distributed.py

[RL] support update_weights_from_distributed with different group and multiple weights (#7292)

2025-07-02 19:29:11 -07:00

test_update_weights_from_tensor.py

support custom weight loader for model runner (#7122)

2025-06-16 16:28:15 -07:00

test_utils_update_weights.py

Add tensor.detach() back to update weight util (#8691)

2025-08-02 00:41:05 -07:00

test_verl_engine_2_gpu.py

Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (#7099)

2025-06-19 00:56:37 -07:00

test_verl_engine_4_gpu.py

Multi-Stage Awake: Support Resume and Pause KV Cache and Weights separately (#7099)

2025-06-19 00:56:37 -07:00

test_vertex_endpoint.py

…

test_vision_chunked_prefill.py

Improve streaming, log_level, memory report, weight loading, and benchmark script (#7632)

2025-06-29 23:16:19 -07:00

test_vision_openai_server_a.py

Fix chat template handling for OpenAI serving (#8635)

2025-07-31 21:49:45 -07:00

test_vision_openai_server_b.py

Fix chat template handling for OpenAI serving (#8635)

2025-07-31 21:49:45 -07:00

test_vision_openai_server_common.py

Fix chat template handling for OpenAI serving (#8635)

2025-07-31 21:49:45 -07:00

test_vllm_dependency.py

[CI] Fix flaky threshold (#8370)

2025-07-25 16:41:56 -07:00

test_vlm_accuracy.py

refactor: unify names of the feature field of MultimodalDataItem (#8075)

2025-07-16 17:52:38 -07:00

test_vlm_input_format.py

bring back kimi vl ci (#8537)

2025-07-29 13:14:18 -07:00

test_w8a8_quantization.py

…