minke.yu/sglang

Files

T

History

Xiaoyu Zhang 39d4d65a51 [diffusion] Accelerate SANA-Video linear attention in quality=high (#35728 )

2026-08-21 18:05:43 +08:00

..

Update deepep for SBO feature (#35450 )

2026-08-19 23:37:13 -07:00

[CI] Fix nightly test failures (#34523 )

2026-08-12 15:40:20 -07:00

[AMD] Add GLM-5.2 MI35x nightly accuracy and perf benchmark (#32570 )

2026-08-19 22:49:43 -07:00

Split TRTLLM MHA decode batches by KV sequence length (#34888 )

2026-08-20 00:44:26 -07:00

[FlashInfer v0.6.16] Support FlashInfer CuTe DSL NVFP4 MoE quantization (#28354 )

2026-08-13 17:33:46 -07:00

[CI] Trim redundant nightly test registrations (#34070 )

2026-08-08 01:42:46 -07:00

chunked_prefill

[VLM] split multimodal scheduling from mm_utils (#32415 )

2026-08-05 20:24:12 +08:00

constrained_decoding

[CI] Fold duplicate-server suites and prune the retract matrix on 1-gpu-5090 (#33745 )

2026-08-05 12:41:51 -07:00

[Spec] Support logprobs with DSpark speculative decoding (#34696 )

2026-08-16 15:05:13 -07:00

Fix DSpark + DeepSeek V4 prefill CP compatibility (#33865 )

2026-08-10 23:26:29 -07:00

TP/PP Consensus checker (#34406 )

2026-08-21 01:36:03 +08:00

[CI] Key scheduled CUDA suites by runner_config instead of hand-written jobs (#34186 )

2026-08-09 16:44:53 -07:00

[DCP] Drop the prefill index-selection syncs by taking each rank's rows by stride (#35084 )

2026-08-17 19:55:51 -07:00

[CI] Key scheduled CUDA suites by runner_config instead of hand-written jobs (#34186 )

2026-08-09 16:44:53 -07:00

[P/D disagg] Decode-side radix cache for SWA hybrid models (unified radix tree) (#27770 )

2026-08-21 09:25:06 +08:00

[dLLM] Make FDFO a framework capability for all dLLM algorithms (#27551 )

2026-07-11 11:05:05 +08:00

[Fix] Resolve VLM test image placeholders from the model's own chat template (#33509 )

2026-08-04 02:01:39 -07:00

[Test] split test/registered/distributed/ into topic folders (#26240 )

2026-05-24 17:02:07 -07:00

Revert "[Feature] Add DeepEPv2 (ElasticBuffer) MoE A2A backend" (#35568 )

2026-08-19 14:14:38 -07:00

[Kernel] Phase 4 batch-3: migrate tangled JIT subsystems + new groups into kernels.ops (RFC #29630 ) (#32045 )

2026-07-22 21:15:03 +08:00

test: restore GLM-4.1V nightly latency threshold (#34811 )

2026-08-15 18:13:19 +08:00

Stop losing Kimi-K3 tool calls to reasoning, constraint conflicts, and truncation (#34881 )

2026-08-18 12:42:56 -07:00

Stabilize GB300 nightly tests (#35044 )

2026-08-17 03:30:16 -07:00

Qwen3.8-27B Model Support (#34859 )

2026-08-19 16:31:43 +08:00

[UnifiedTree] feat: support runtime attach/detach (#35269 )

2026-08-19 22:49:08 +08:00

input_embedding

[misc] Use --cuda-graph-max-bs-decode in tests, examples, and docs (#29591 )

2026-06-28 18:38:28 -07:00

[diffusion] Accelerate SANA-Video linear attention in quality=high (#35728 )

2026-08-21 18:05:43 +08:00

[UnifiedTree]: move /mem_cache/unifed_cache_component dir to /mem_cache/unified_cache (#32484 )

2026-07-28 16:28:21 +08:00

[diffusion] refactor: reuse srt qwen vision and text modules (#35006 )

2026-08-19 10:12:44 +08:00

Refocus LoRA tests on regression coverage (#34464 )

2026-08-11 17:37:52 -07:00

[CI] Share VLM engines and prune launch matrices on the per-commit H100/H200 suites (#33944 )

2026-08-07 01:22:22 -07:00

[CI] Merge tokenizer worker tests and drop redundant triton attention e2e (#33641 )

2026-08-05 11:55:11 -07:00

[MLX] Window-bounded SWA KV storage and in-graph sampling (#34166 )

2026-08-09 21:28:51 -07:00

[Feature] Optimize TP LMHead with All-to-All (#32313 )

2026-08-17 19:55:27 -07:00

[Model Loading] Overlap checkpoint staging with CUDA graph capture during startup (#32017 )

2026-08-13 12:26:25 -07:00

test: switch the Inkling-Small NVFP4 deterministic suite to DSPARK (#35293 )

2026-08-20 21:54:52 +08:00

Amd/dsv4 shared experts fusion top6 (#32340 )

2026-08-19 22:57:58 -07:00

chore: cleanup garbage code (#29770 )

2026-07-02 16:14:01 +08:00

Add three new test cases (#35502 )

2026-08-19 17:59:45 +08:00

[Refactor] New EPD (#30398 )

2026-08-21 15:22:48 +08:00

Add --http2-max-concurrent-streams server arg (#34796 )

2026-08-15 10:34:49 -07:00

config: route parallel config-leaf reads through get_parallel() (#33170 )

2026-08-01 08:58:39 -07:00

feat(mem_cache): page-major (layer-major within a page) KV/state layout (#29533 )

2026-06-29 14:49:54 -07:00

[CI] Key scheduled CUDA suites by runner_config instead of hand-written jobs (#34186 )

2026-08-09 16:44:53 -07:00

[CI] Trim the base-c 4-gpu-h100 stage from 5 shards to 4 (#35407 )

2026-08-19 00:48:07 -07:00

config: retire the multi-engine accommodation in the runtime context (#35022 )

2026-08-17 16:15:33 -07:00

[CI][AMD] Run the profiling suite without CUDA graphs on ROCm (#34452 )

2026-08-20 09:41:28 +08:00

[AMD][Quantization] Online MXFP4 quantization 4/N - NVFP4 to MXFP4 Online Requantization on AMD GPUs (#29328 )

2026-08-14 21:59:39 -07:00

[P/D disagg] Decode-side radix cache for SWA hybrid models (unified radix tree) (#27770 )

2026-08-21 09:25:06 +08:00

…

[CI] Trim the base-c 4-gpu-h100 stage from 5 shards to 4 (#35407 )

2026-08-19 00:48:07 -07:00

[rotary] Rebuild the shared RoPE cache entry when its buffers are dead (#33575 )

2026-08-04 17:24:42 -07:00

Build Rust extensions on demand in source checkouts (#34994 )

2026-08-16 14:58:06 -07:00

[Sampling] Restore finite top-k requirement for sampling masks (#35205 )

2026-08-20 16:55:25 -07:00

Fix _GenerationStreamAccumulator logprob_end off-by-one under retract (#26510 )

2026-08-20 16:03:13 -07:00

scripted_runtime

[AMD] Register 2 CPU-bound 1-GPU tests (phase_checker, scripted_runtime_core) for AMD PR CI (#30446 )

2026-07-08 14:04:23 -07:00

[CI] Share VLM engines and prune launch matrices on the per-commit H100/H200 suites (#33944 )

2026-08-07 01:22:22 -07:00

[Memory] Borrow CUDA graph pool storage for EAGLE sampling (#35375 )

2026-08-18 16:19:41 -07:00

…

[CI] Route mmlu and GB300 MMMU-Pro evals through sgl-eval (#34477 )

2026-08-12 19:33:21 -07:00

Fix overlap prebuilt row reuse race (#35748 )

2026-08-21 02:00:13 -07:00

fix: make automatic NUMA binding configurable (#30394 )

2026-08-12 17:40:28 -07:00

[perf] overlap page preprocessing, pack the vit, enable prefill CUDA graph for paddle-ocr (#35318 )

2026-08-19 08:20:55 +08:00

[XPU] upgrade sglang xpu backend to PyTorch 2.13 (#31751 )

2026-08-17 18:29:15 +08:00

README.md

Add pr tests (#31952 )

2026-08-01 15:03:21 +08:00

README.md

Registered Tests

Tests under this directory are auto-discovered by run_suite.py via CI registration decorators.

Where Should I Put My New Test?

No server / engine launch required

What you're testing	Directory	Requires
Component logic in isolation (cache, scheduler, config, parser, etc.)	`unit/<module>/`	CPU or GPU
CUDA kernel correctness	`kernels/`	GPU

Server / engine launch required (E2E)

What you're testing	Directory	Requires
Model inference correctness	`models/`, `4-gpu-models/`, `8-gpu-models/`	GPU
Feature-specific (OpenAI API, LoRA, speculative, distributed, VLM, etc.)	`openai_server/`, `lora/`, `spec/`, `distributed/`, ...	GPU
Benchmarks (performance, accuracy, stress)	`benchmark/`	GPU
Platform-specific	`amd/`, `npu/`	Vendor GPU

See unit/README.md for unit test conventions.