sglang

attention

[Auto Sync] Rename is_hybrid to is_hybrid_swa (#14252 )

2025-12-01 23:24:24 -08:00

ci

[NPU]add nightly-test-npu (#14143 )

2025-12-05 00:43:35 +08:00

kits

feat: Naive support Spec V2 + Constrained Decoding (#13425 )

2025-11-27 20:31:46 +08:00

longbench_v2

Feature/longbench v2 evaluation utils (#10949 )

2025-10-07 14:17:31 +08:00

speculative

Support FlashAttention3 page_size > 1 and topk > 1 case with paged attn and spec decode (#7725 )

2025-11-26 11:44:41 +08:00

__init__.py

…

doc_patch.py

…

few_shot_gsm8k_engine.py

[Fix] Fix uvloop get_event_loop() is not suitable for 0.22.x (#13612 )

2025-11-25 01:20:00 +08:00

few_shot_gsm8k.py

…

get_logits_ut.py

Support DeepSeek V3.2 Exp (#11061 )

2025-10-06 00:24:15 -07:00

gsm8k_mixin.py

modularize gsm8k and mmmu test classes (#13506 )

2025-11-22 16:06:53 +08:00

long_prompt.txt

…

mmmu_vlm_mixin.py

modularize gsm8k and mmmu test classes (#13506 )

2025-11-22 16:06:53 +08:00

nightly_bench_utils.py

[Auto Sync] Improve profilers and simplify bench_one_batch_server.py (#13866 )

2025-11-25 12:13:31 -08:00

run_eval.py

[Fix] Improve longbench prompt and other logics (#11474 )

2025-10-12 15:04:28 -07:00

runners.py

[NPU]add nightly-test-npu (#14143 )

2025-12-05 00:43:35 +08:00

send_one.py

[Auto Sync] Improve profilers and simplify bench_one_batch_server.py (#13866 )

2025-11-25 12:13:31 -08:00

simple_eval_common.py

Allow benchmarking tool to handle empty response (#12174 )

2025-10-29 20:26:23 -07:00

simple_eval_gpqa.py

[Lint] Add python/sglang to ruff F401 checks and remove unused imports in files (#11685 )

2025-10-17 16:49:46 -07:00

simple_eval_humaneval.py

Allow benchmarking tool to handle empty response (#12174 )

2025-10-29 20:26:23 -07:00

simple_eval_longbench_v2.py

[Fix] Improve longbench prompt and other logics (#11474 )

2025-10-12 15:04:28 -07:00

simple_eval_math.py

Allow benchmarking tool to handle empty response (#12174 )

2025-10-29 20:26:23 -07:00

simple_eval_mgsm.py

…

simple_eval_mmlu.py

Allow benchmarking tool to handle empty response (#12174 )

2025-10-29 20:26:23 -07:00

simple_eval_mmmu_vlm.py

Allow benchmarking tool to handle empty response (#12174 )

2025-10-29 20:26:23 -07:00

test_activation.py

…

test_block_fp8_deep_gemm_blackwell.py

[1/2] Refactor DeepGeem requant for FP8 Linear on Blackwell (#13601 )

2025-11-23 16:07:56 -08:00

test_block_fp8.py

[Lint] Add python/sglang to ruff F401 checks and remove unused imports in files (#11685 )

2025-10-17 16:49:46 -07:00

test_custom_ops.py

…

test_cutlass_moe.py

[sgl-kernel][5/N]Support Expert Specialization Grouped GEMM (#12666 )

2025-11-11 21:23:25 -08:00

test_cutlass_w4a8_moe.py

[Lint] Add python/sglang to ruff F401 checks and remove unused imports in files (#11685 )

2025-10-17 16:49:46 -07:00

test_cutlass_w16a16_moe.py

Add support for bf16 x bf16 cutlass fused MoE (#10275 )

2025-11-24 18:49:39 -08:00

test_deepep_utils.py

…

test_deterministic_utils.py

[Auto Sync] Update test_deterministic.py, test_deterministi... (20251024) (#12083 )

2025-10-24 22:28:01 -07:00

test_deterministic.py

[Auto Sync] Improve profilers and simplify bench_one_batch_server.py (#13866 )

2025-11-25 12:13:31 -08:00

test_disaggregation_utils.py

[ci]use H20 to run disaggregation test (#11543 )

2025-10-16 11:42:42 -07:00

test_dynamic_grad_mode.py

…

test_kvfp4_quant_dequant.py

feat: Add FP4 (E2M1) KV Cache Support with Quantization Utilities for MLA (#10078 )

2025-11-01 22:24:58 -07:00

test_layernorm.py

Enable mixed type LayerNorm kernel for NSA indexer (#12044 )

2025-11-03 16:50:41 -08:00

test_marlin_moe.py

[kimi k2 thinking] Avoid useless torch.zeros_ (#13596 )

2025-11-21 13:15:27 +08:00

test_marlin_utils.py

…

test_programs.py

[AMD] Add AITER Custom All-Reduce (#13102 )

2025-11-12 21:53:44 -08:00

test_utils.py

[NPU] chore: bump to CANN 8.3.RC1 and Pytorch 2.8.0 (#13647 )

2025-11-21 17:07:08 +08:00