sglang

ascend

[Feature] npu support enable_torch_compile for torchair backend (#13410 )

2025-12-16 09:23:51 +08:00

configs

Add test config yamls for Deepseek v3 (#5433 )

2025-04-20 17:28:52 -07:00

cpu

[CPU] Add 4D input support for ROPE in sgl-kernel (#9337 )

2025-12-16 17:27:39 +08:00

debug_utils

Test reorganization: Move tests to manual/ (#13610 )

2025-11-20 13:41:58 -08:00

dllm

[DLLM] Fix dLLM regression (#15371 )

2025-12-19 10:55:46 +08:00

ep

[bugfix] fix TBO crashes when attn_tp_size > 1 (#13730 )

2025-12-11 18:18:40 -08:00

external_models

feat: support external custom models (#13429 )

2025-11-20 11:16:58 +08:00

hicache

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

layers/attention/mamba

[CI] Move nightly tests to test/nightly/ (#13683 )

2025-11-20 18:00:02 -08:00

models

[AMD] Clear pre-built AITER kernels and warmup to prevent segfaults and test timeouts (#15318 )

2025-12-18 14:00:09 -08:00

nightly

[AMD] Add TP=8 models to nightly test and make TP=2 test stable (#15296 )

2025-12-19 13:15:19 -08:00

openai_server

Clean up __init__ function of the scheduler and event loop for PD (#15298 )

2025-12-18 01:35:14 -08:00

ops

[VLM] Replace torch.repeat_interleave with faster np.repeat for Qwen-VL series (#13736 )

2025-11-22 22:45:32 +08:00

quant

[kernel][moe] add moe topk fast (#13969 )

2025-12-14 22:26:40 +08:00

rl

[Feature] Enable return routed experts (#12162 )

2025-12-21 15:16:43 +08:00

rotary_embedding

fix: small changes to enable test_mrope.py (#14082 )

2025-11-27 22:58:17 -08:00

xpu

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

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

Rename double sparsity config file (#2188 )

2024-11-25 17:12:08 -08:00

experiment_runner.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

kv_cache_scales_llama3_1_8b.json

Support FP8 E4M3 KV Cache (#2786 )

2025-01-12 21:17:11 -08:00

kv_cache_scales_llama3_8b.json

support e4m3 kvcache in qwen2 & add kv scaling facotr json (#2894 )

2025-01-18 11:43:22 +08:00

kv_cache_scales_qwen2_1_5b.json

support e4m3 kvcache in qwen2 & add kv scaling facotr json (#2894 )

2025-01-18 11:43:22 +08:00

parse_results.py

misc: parse bench_serving result as markdown table (#9377 )

2025-08-20 16:44:20 -07:00

run_suite.py

[Feature] Enable return routed experts (#12162 )

2025-12-21 15:16:43 +08:00

test_abort.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_bench_one_batch.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_bench_serving.py

[Embeddings Performance Testing] Add performance test for embedding models (#12359 )

2025-11-17 15:35:18 -08:00

test_bench_typebaseddispatcher.py

perf: optimize TypeBasedDispatcher using dict for O(1) lookup (#12001 )

2025-11-15 22:02:42 +08:00

test_bnb.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_chunked_prefill.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_constrained_decoding_spec_reasoning.py

Support grammar + spec + reasoning (#14163 )

2025-11-30 21:19:57 +08:00

test_constrained_decoding.py

[Test] Merge all constrained decoding tests. (#12633 )

2025-11-05 00:43:06 +08:00

test_create_kvindices.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_cutedsl_moe.py

Replace [silu_and_mul_]scaled_fp4_group_quant by Flashinfer equivalent (#12376 )

2025-11-13 00:26:00 -08:00

test_data_parallelism.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_deepseek_v3_basic.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_deepseek_v3_cutedsl_4gpu.py

[CI]Add gb200 runner back (#15024 )

2025-12-12 20:19:34 -08:00

test_deepseek_v3_fp4_4gpu.py

Support piecewise cuda graph for dsv3 fp4 (#15531 )

2025-12-21 14:50:32 +08:00

test_deepseek_v3_mtp.py

[AMD] add unit-test-backend-8-gpu-amd back (#15253 )

2025-12-19 00:59:28 -08:00

test_deepseek_v32_basic.py

[Deepseek V3.2] Fix accuracy bug in the Indexer (#12583 )

2025-11-11 16:15:26 -08:00

test_deepseek_v32_mtp.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_deterministic.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_disaggregation_basic.py

Introduce server_fixtures in sglang.test (#14899 )

2025-12-11 22:30:33 +09:00

test_disaggregation_different_tp.py

Introduce server_fixtures in sglang.test (#14899 )

2025-12-11 22:30:33 +09:00

test_disaggregation_dp_attention.py

Introduce server_fixtures in sglang.test (#14899 )

2025-12-11 22:30:33 +09:00

test_disaggregation_hybrid_attention.py

[Qwen3-next] Add PD disaggregation support for mamba with extra_buffer (#15180 )

2025-12-16 14:36:00 +08:00

test_disaggregation_pp.py

[PP] Add dynamic chunking PP test (#15395 )

2025-12-19 01:09:40 +08:00

test_dp_attention.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_eagle_dp_attention.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_eagle_infer_beta_dp_attention.py

[CI] fix UT success check in test_eagle_infer_beta_dp_attention.py (#14831 )

2025-12-11 01:00:50 +09:00

test_epd_disaggregation.py

fix: moving decorator to header (#15297 )

2025-12-16 17:05:45 -08:00

test_eval_accuracy_large.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_eval_fp8_accuracy.py

[CI] Fix ci tests (#5769 )

2025-04-27 07:18:10 -07:00

test_external_models.py

tiny unify environ usage (#15335 )

2025-12-17 23:31:43 +08:00

test_fa3.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_flash_attention_4.py

fix b200 fa4 ci (#14788 )

2025-12-10 00:03:43 -08:00

test_flashmla.py

Reserve more memory for DeepSeekOCR model and adjust server start timeout for DeepGEMM to reduce flakiness (#15277 )

2025-12-17 13:13:05 -08:00

test_fp8_blockwise_gemm.py

Add FP8 Blockwise GEMM Backend Flag --fp8-gemm-backend (#14379 )

2025-12-09 12:05:56 -08:00

test_fp8_utils.py

Support inverse transform ue8m0 scale (#13285 )

2025-11-15 16:34:32 +08:00

test_fused_moe.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_gguf.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_gpt_oss_1gpu.py

Fix CI (#9013 )

2025-08-09 16:00:10 -07:00

test_gpt_oss_4gpu.py

feat: add gpt oss b200 ci (#9988 )

2025-09-03 17:26:38 -07:00

test_gpt_oss_common.py

Fix condition for streaming output_ids in tokenizer manager (#13759 )

2025-11-29 13:56:15 -08:00

test_gptqmodel_dynamic.py

Depreate global_server_args_dict (#11528 )

2025-10-13 19:34:43 +08:00

test_harmony_parser.py

Move parsers under a single folder (#9912 )

2025-09-02 18:25:04 -07:00

test_hidden_states.py

Open AI API hidden states (#6716 )

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

test_hybrid_attn_backend.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_input_embeddings.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -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

Move parsers under a single folder (#9912 )

2025-09-02 18:25:04 -07:00

test_llama31_fp4.py

fix b200 ci (#14786 )

2025-12-09 23:08:41 -08:00

test_load_weights_from_remote_instance.py

support non disturbing remote instance weight loader v2 (#14997 )

2025-12-16 14:39:56 -08:00

test_local_attn.py

Simplify FA3 tests (#5779 )

2025-04-27 01:30:17 -07:00

test_mamba_unittest.py

[Qwen3-next] support mamba radix cache for overlap scheduler (#14792 )

2025-12-14 18:54:16 -08:00

test_metrics_utils.py

Rename customer label -> custom label (#10899 )

2025-09-25 16:19:53 -07:00

test_metrics.py

Improve structured outputs: fix race condition, server crash, metrics and style (#6188 )

2025-05-11 08:36:16 -07:00

test_mistral_large3_basic.py

Add nightly accuracy test for DeepSeek V3.2 (#14935 )

2025-12-13 12:11:16 -08:00

test_mla_deepseek_v3.py

Increase timeout for TestDeepseekV3MTP for potential DeepGEMM cold start (#15239 )

2025-12-16 00:26:10 -08:00

test_mla_flashinfer.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_mla_fp8.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_mla_int8_deepseek_v3.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_mla.py

Improve profiler and integrate profiler in bench_one_batch_server (#6787 )

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

test_model_hooks.py

Rename: --hooks to --forward-hooks (#13994 )

2025-11-26 22:26:28 +08:00

test_modelopt_export.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_modelopt_loader.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_moe_eval_accuracy_large.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_multi_instance_release_memory_occupation.py

Fix mem fraction static for nightly tests (#11076 )

2025-09-29 12:57:41 -07:00

test_multi_tokenizer.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_ngram_speculative_decoding.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_no_chunked_prefill.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_no_overlap_scheduler.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_original_logprobs.py

Extend lint test to test/ directory (#13247 )

2025-11-14 00:01:48 -08:00

test_page_size.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_patch_torch.py

Tiny move files to utils folder (#11166 )

2025-10-03 22:40:06 +08:00

test_penalty.py

[Test] Add Functional Tests for Penalty Parameters (#11931 )

2025-11-01 00:10:17 +08:00

test_pp_single_node.py

[PP] Add pp support for Qwen3-VL (#12333 )

2025-12-17 16:03:58 +08:00

test_priority_scheduling.py

[Fix]: Adjust FutureMap's token_id_bufs Size to Prevent ChunkedPrefill's next_token_ids from Overwriting Previous Prefill Requests' next_token_id (#13713 )

2025-11-25 01:08:52 +08:00

test_profile_merger_http_api.py

[FEATURE] Add Profile Trace Merger for Distributed Traces (#11413 )

2025-10-14 09:20:17 +08:00

test_profile_merger.py

[FEATURE] Add Profile Trace Merger for Distributed Traces (#11413 )

2025-10-14 09:20:17 +08:00

test_profile_v2.py

Implement profiler v2 and fix stage mixture bug (#14148 )

2025-11-30 16:59:52 +08:00

test_pytorch_sampling_backend.py

Add retry for flaky tests in CI (#4755 )

2025-03-25 16:53:12 -07:00

test_quantization.py

Fix import warnings (#15144 )

2025-12-14 21:25:18 -08:00

test_radix_attention.py

Enhance runtime memory check in CI (#15192 )

2025-12-16 21:40:38 +08:00

test_radix_cache_unit.py

fix: malformed KV events for NVIDIA Dynamo (#13488 )

2025-11-28 14:55:20 -08:00

test_reasoning_parser.py

Move parsers under a single folder (#9912 )

2025-09-02 18:25:04 -07:00

test_release_memory_occupation.py

Support releasing CUDA graph memory when paused (#7873 )

2025-10-28 14:40:50 +08:00

test_request_queue_validation.py

Simplify the BatchMultimodalOutput in io_struct.py (#12993 )

2025-11-10 13:54:56 -08:00

test_retract_decode.py

Enhance runtime memory check in CI (#15192 )

2025-12-16 21:40:38 +08:00

test_rope_rocm.py

[AMD] Clear pre-built AITER kernels and warmup to prevent segfaults and test timeouts (#15318 )

2025-12-18 14:00:09 -08:00

test_score_api.py

[Generative Score API] Multi-Item scoring with custom attention mask. (#10979 )

2025-10-08 18:47:32 -07:00

test_server_args.py

Do not use MagicMock to mock server_args in tests (#12154 )

2025-10-26 16:00:46 +08:00

test_skip_tokenizer_init.py

[CI fix] Fix image download failures in VLM CI tests (#13613 )

2025-11-20 11:18:06 +08:00

test_speculative_registry.py

[Auto Sync] Update scheduler.py, spec_info.py, run_suite.py... (20251027) (#12235 )

2025-10-27 17:21:08 -07:00

test_srt_endpoint.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_srt_engine.py

Tiny move files to utils folder (#11166 )

2025-10-03 22:40:06 +08:00

test_standalone_speculative_decoding.py

Revert "tiny remove deprecated endpoint call" (#14533 )

2025-12-05 23:48:54 -08:00

test_start_profile.py

[feature] Add layerwise NVTX support (#11870 )

2025-11-15 19:20:56 -08:00

test_swa_unittest.py

Refactor cache init logic (#13800 )

2025-11-24 11:41:46 +08:00

test_torch_compile_moe.py

[UT for RL] Add UT to cover release/resume memory case for moe model (#8803 )

2025-09-09 19:25:12 -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

[CI] test chunked prefill more (#5798 )

2025-04-28 10:57:17 -07:00

test_torchao.py

Fix TorchAO quant in VLM (#13508 )

2025-11-24 22:15:40 +08:00

test_triton_attention_backend.py

[CI] test chunked prefill more (#5798 )

2025-04-28 10:57:17 -07:00

test_triton_attention_kernels.py

fixes hardcoded "cuda" device references in unit tests to use a dynamic device selection (#12761 )

2025-11-07 11:38:35 +08:00

test_triton_fused_moe.py

Fix B200 Nightly tests and move one manual test back to unit test to prevent the same issue (#13746 )

2025-11-21 17:41:12 -08:00

test_triton_moe_channel_fp8_kernel.py

Support true on-policy (#12058 )

2025-10-25 10:23:42 +08:00

test_triton_sliding_window.py

Fix triton sliding window test case (#6981 )

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

test_type_based_dispatcher.py

perf: optimize TypeBasedDispatcher using dict for O(1) lookup (#12001 )

2025-11-15 22:02:42 +08:00

test_utils_update_weights.py

Fix CI (#9012 )

2025-08-09 13:33:42 -07:00

test_video_utils.py

Support nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16 (and nvidia/C-RADIOv2-H) (#12277 )

2025-11-26 16:28:52 -07:00

test_vision_chunked_prefill.py

Increase timeout for TestDeepseekV3MTP for potential DeepGEMM cold start (#15239 )

2025-12-16 00:26:10 -08:00

test_vision_openai_server_a.py

vlm: refactor engine vlm params and support processor output as input (#14091 )

2025-12-20 18:31:24 +08:00

test_vision_openai_server_common.py

Reserve more memory for DeepSeekOCR model and adjust server start timeout for DeepGEMM to reduce flakiness (#15277 )

2025-12-17 13:13:05 -08:00

test_vlm_input_format.py

vlm: refactor engine vlm params and support processor output as input (#14091 )

2025-12-20 18:31:24 +08:00

test_wave_attention_kernels.py

[AMD] Support Wave attention backend with AMD GPU optimizations (#8660 )

2025-08-12 13:49:11 -07:00