sglang

ascend

[Bugfix] Fix Qwen3/DSV3/DSV3.2 model support (#11510 )

2025-10-16 15:14:09 +08:00

batch_invariant

Support DeepGEMM for deterministic inference (#12142 )

2025-10-26 22:36:17 +08:00

configs

…

cpu

[Fix] Set global args in cpu test (#12105 )

2025-10-24 21:46:17 -07:00

entrypoints/http_server

…

ep

[2/N] Added the core structure of elastic EP and the eplb algorithm with faulty rank (#10606 )

2025-10-22 01:13:31 -07:00

function_call

…

hicache

Fix flaky hicache test with mooncake backend (#11953 )

2025-10-22 21:00:47 +08:00

layers/attention

Feature/nano v2 offline modelopt fp8 and nvfp4 (#12018 )

2025-10-23 11:16:46 -07:00

lora

Feature/Add GET endpoint to query loaded LoRA adapters (#12229 )

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

models

Add support for Matryoshka embeddings (#126 ) (#11142 )

2025-10-28 02:49:36 +08:00

openai_server

Add continuous_usage_stats support for streaming responses (#12241 )

2025-10-29 10:01:23 +08:00

quant

[ci] Try fixing broken CIs (#12317 )

2025-10-29 01:13:51 -07:00

rl

[minor] sync code on python/sglang/test/test_deterministic.py and improve ci tests (#11777 )

2025-10-17 14:25:22 -07:00

rotary_embedding

Support mrope triton kernel and add unit test (#11722 )

2025-10-20 11:51:07 +08:00

xpu

Init attention backend for Intel XPU (#10656 )

2025-10-21 11:41:28 +08:00

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

[CI] Add Llama 3.1 8B FP4 to B200 CI (#12182 )

2025-10-29 14:56:41 -07:00

test_abort.py

feat: return partial generation results when aborting requests in waiting queue (#11673 )

2025-10-29 22:03:00 +08:00

test_async_dynamic_batch_tokenizer.py

…

test_bench_one_batch.py

…

test_bench_serving.py

…

test_bnb.py

vlm: enforce pybase64 for image and str encode/decode (#10700 )

2025-10-21 19:05:32 +08:00

test_build_eagle_tree.py

move eagle draft post process to cuda graph (#11434 )

2025-10-14 22:50:53 +08:00

test_chunked_prefill.py

…

test_config_integration.py

…

test_cpp_radix_cache.py

…

test_create_kvindices.py

…

test_custom_allreduce.py

…

test_cutedsl_flashinfer_8gpu.py

…

test_data_parallelism.py

…

test_deepseek_v3_basic.py

…

test_deepseek_v3_fp4_4gpu.py

Remove deprecated --enable-beta-spec argument and fix b200 test (#12167 )

2025-10-26 15:54:40 -07:00

test_deepseek_v3_mtp.py

…

test_deepseek_v32_basic.py

[CI] Add CI test for DeepSeek V3.2 MTP (#11835 )

2025-10-19 17:00:25 -07:00

test_deepseek_v32_mtp.py

[CI] Add CI test for DeepSeek V3.2 MTP (#11835 )

2025-10-19 17:00:25 -07:00

test_deepseek_v32_nsabackend.py

[Test] Add dsv3.2 nsa backend testing (#11936 )

2025-10-25 17:07:37 -07:00

test_deterministic.py

…

test_disaggregation_basic.py

…

test_disaggregation_different_tp.py

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

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

test_disaggregation_dp_attention.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_disaggregation_hybrid_attention.py

[PD] Add PD support for hybrid model (Qwen3-Next, DeepSeek V3.2 Exp) (#10912 )

2025-10-15 18:59:14 -07:00

test_disaggregation_pp.py

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

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

test_double_sparsity.py

…

test_dp_attention.py

…

test_eagle_dp_attention.py

Eagle3 DP attention for Qwen3 MoE (#12002 )

2025-10-29 20:25:17 +08:00

test_eagle_infer_a.py

…

test_eagle_infer_b.py

…

test_eagle_infer_beta.py

[misc] depdencies & enviroment flag (#12113 )

2025-10-26 14:52:35 +08:00

test_ebnf_constrained.py

…

test_eval_accuracy_large.py

…

test_eval_fp8_accuracy.py

…

test_expert_distribution.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_expert_location_updater.py

…

test_fa3.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_fim_completion.py

…

test_flash_attention_4.py

[NVIDIA] FA3/FA4 Fix (#11606 )

2025-10-19 17:10:10 -07:00

test_flashmla.py

…

test_forward_split_prefill.py

…

test_function_call_parser.py

…

test_fused_moe.py

[2/2] Deepseek deterministic: support deepseek v3 deterministic inference on 8 x H200 (#12095 )

2025-10-29 11:49:04 -07:00

test_get_weights_by_name.py

…

test_gguf.py

…

test_gpt_oss_1gpu.py

…

test_gpt_oss_4gpu.py

…

test_gpt_oss_common.py

…

test_gptqmodel_dynamic.py

…

test_harmony_parser.py

…

test_health_check.py

…

test_hidden_states.py

…

test_hybrid_attn_backend.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_input_embeddings.py

…

test_io_struct.py

…

test_jinja_template_utils.py

…

test_kv_events.py

…

test_llama31_fp4.py

[CI] Add Llama 3.1 8B FP4 to B200 CI (#12182 )

2025-10-29 14:56:41 -07:00

test_load_weights_from_remote_instance.py

…

test_local_attn.py

…

test_logprobs.py

[logprobs] Enable local deterministic logrprobs testing with strict threshold (#10994 )

2025-10-19 13:30:39 -07:00

test_mamba_unittest.py

[PD] Add PD support for hybrid model (Qwen3-Next, DeepSeek V3.2 Exp) (#10912 )

2025-10-15 18:59:14 -07:00

test_metrics_utils.py

…

test_metrics.py

…

test_mla_deepseek_v3.py

…

test_mla_flashinfer.py

…

test_mla_fp8.py

…

test_mla_int8_deepseek_v3.py

…

test_mla_tp.py

…

test_mla.py

…

test_modelopt_export.py

Enable native ModelOpt quantization support (3/3) (#10154 )

2025-10-21 21:44:29 -07:00

test_modelopt_fp8kvcache.py

…

test_modelopt_loader.py

Enable native ModelOpt quantization support (3/3) (#10154 )

2025-10-21 21:44:29 -07:00

test_modelopt.py

…

test_models_from_modelscope.py

…

test_moe_eval_accuracy_large.py

…

test_mscclpp.py

…

test_multi_instance_release_memory_occupation.py

…

test_multi_tokenizer.py

…

test_ngram_speculative_decoding.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_nightly_gsm8k_eval_amd.py

…

test_nightly_text_models_gsm8k_eval.py

…

test_nightly_text_models_perf.py

…

test_nightly_vlms_mmmu_eval.py

model: support NVILA and NVILA Lite (#10399 )

2025-10-26 09:58:09 -07:00

test_nightly_vlms_perf.py

…

test_no_chunked_prefill.py

…

test_no_overlap_scheduler.py

…

test_original_logprobs.py

Sync code and test CI; rename some env vars (#11686 )

2025-10-15 18:37:03 -07:00

test_page_size.py

…

test_patch_torch.py

…

test_penalty.py

…

test_piecewise_cuda_graph.py

support more model in piecewise cuda graph (#11745 )

2025-10-24 10:31:39 +08:00

test_pp_single_node.py

…

test_priority_scheduling.py

…

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_pytorch_sampling_backend.py

…

test_quick_allreduce.py

…

test_radix_attention.py

…

test_radix_cache_unit.py

…

test_reasoning_parser.py

…

test_regex_constrained.py

…

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

…

test_retract_decode.py

[Fix] memory leak by overlap + retract (#11981 )

2025-10-23 22:59:23 +08:00

test_rope_rocm.py

Revert "[ROCm] Remove vLLM rope dependency & use AITER impl" (#12028 )

2025-10-23 12:42:59 -07:00

test_sagemaker_server.py

…

test_schedule_policy.py

…

test_score_api.py

…

test_server_args.py

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

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

test_session_control.py

…

test_skip_tokenizer_init.py

…

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

…

test_srt_engine_with_quant_args.py

…

test_srt_engine.py

…

test_standalone_speculative_decoding.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_start_profile.py

…

test_swa_unittest.py

…

test_tokenizer_batch_encode.py

…

test_tokenizer_manager.py

…

test_torch_compile_moe.py

…

test_torch_compile.py

…

test_torch_flex_attention_backend.py

…

test_torch_native_attention_backend.py

…

test_torch_tp.py

…

test_torchao.py

…

test_tracing.py

[Feature] Sglang Tracing: Fine-Grained Tracking for Request Latency - Part 2 (#10804 )

2025-10-28 01:25:46 -07:00

test_triton_attention_backend.py

…

test_triton_attention_kernels.py

Deterministic Mode: Add 1-stage triton kernel for prefill (#11147 )

2025-10-20 01:47:36 +08:00

test_triton_attention_rocm_mla.py

…

test_triton_fused_moe.py

Support true on-policy (#12058 )

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

test_triton_moe_channel_fp8_kernel.py

Support true on-policy (#12058 )

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

test_triton_moe_wna16.py

Support true on-policy (#12058 )

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

test_triton_sliding_window.py

…

test_two_batch_overlap.py

Move deep gemm related arguments to sglang.srt.environ (#11547 )

2025-10-14 00:34:35 +08:00

test_utils_update_weights.py

…

test_vertex_endpoint.py

…

test_vision_chunked_prefill.py

vlm: enforce pybase64 for image and str encode/decode (#10700 )

2025-10-21 19:05:32 +08:00

test_vision_openai_server_a.py

model: support deepseek-ocr (#11891 )

2025-10-24 03:15:17 +08:00

test_vision_openai_server_common.py

model: support deepseek-ocr (#11891 )

2025-10-24 03:15:17 +08:00

test_vllm_dependency.py

…

test_vlm_accuracy.py

…

test_vlm_input_format.py

…

test_wave_attention_backend.py

…

test_wave_attention_kernels.py

…

test_weight_version.py

…