Lianmin Zheng bc1534ff32 Fix a draft model accuracy bug in eagle; support step=1; return logprob in eagle (#4134) ...

Co-authored-by: Sehoon Kim <kssteven418@gmail.com>
Co-authored-by: SangBin Cho <rkooo567@gmail.com>
Co-authored-by: Sehoon Kim <sehoon@x.ai>

2025-03-06 06:13:59 -08:00

..

configs

minor: add random flashinfer vs triton use case (#2409)

2024-12-09 04:15:21 +08:00

models

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -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

feat: support custom task runner (#2407)

2024-12-09 02:29:55 +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

run_suite.py

[QUANT] Add GPTQModel Dynamic Quantization + lm_head Quantization (#3790)

2025-03-05 01:11:00 -08:00

test_abort.py

Fix memory leak during abort (#2238)

2024-11-28 02:22:15 -08:00

test_bench_one_batch.py

Fix CI and install docs (#3821)

2025-02-24 16:17:38 -08:00

test_bench_serving.py

Fix a draft model accuracy bug in eagle; support step=1; return logprob in eagle (#4134)

2025-03-06 06:13:59 -08:00

test_block_int8.py

Feature DeepSeek V3/R1 INT8 Quantization (block-wise) (#3730)

2025-02-24 05:43:35 -08:00

test_cache_report.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_chunked_prefill.py

[CI] Split test cases in CI for better load balancing (#2180)

2024-11-25 04:58:16 -08:00

test_create_kvindices.py

Organize Attention Backends (#1547)

2024-09-30 15:54:18 -07:00

test_custom_allreduce.py

adapt custom allreduce for tensorrt llm (#2511)

2025-01-16 04:57:35 +08:00

test_data_parallelism.py

udate weights from disk (#2265)

2024-11-30 01:17:00 +00:00

test_double_sparsity.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_dp_attention.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_eagle_infer.py

Fix a draft model accuracy bug in eagle; support step=1; return logprob in eagle (#4134)

2025-03-06 06:13:59 -08:00

test_ebnf_constrained.py

Misc clean up; Remove the support of jump forward (#4032)

2025-03-03 07:02:14 -08:00

test_embedding_openai_server.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_eval_accuracy_large_chunked_prefill.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_eval_accuracy_large_mixed_chunked_prefill.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_eval_accuracy_large.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

test_eval_accuracy_mini.py

Revert "[feat] Add math eval to CI" (#2656)

2024-12-30 15:05:50 +08:00

test_fp8_kernel.py

Add activation parameters to fused_moe (#3170)

2025-01-27 00:23:37 -08:00

test_fp8_kvcache.py

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

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

test_function_calling.py

Feature/function calling update (#2700)

2025-01-26 09:57:51 -08:00

test_fused_moe.py

optimize test_fused_moe style (#3268)

2025-02-03 18:56:18 +08:00

test_get_weights_by_name.py

[Minor] Fix code style (#2311)

2024-12-02 02:27:36 -08:00

test_gguf.py

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -08:00

test_gptqmodel_dynamic.py

Fix a draft model accuracy bug in eagle; support step=1; return logprob in eagle (#4134)

2025-03-06 06:13:59 -08:00

test_health_check.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

test_hidden_states.py

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -08:00

test_input_embeddings.py

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -08:00

test_json_constrained.py

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -08:00

test_large_max_new_tokens.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_matched_stop.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_metrics.py

Split the __init__ of scheduler as smaller functions. Improve the eagle tests (#4128)

2025-03-06 00:13:20 -08:00

test_mla_flashinfer.py

Update nextn ci test (#4071)

2025-03-04 13:01:24 -08:00

test_mla_fp8.py

Use only one GPU for MLA CI tests (#2858)

2025-01-13 03:55:33 -08:00

test_mla_tp.py

Update nextn ci test (#4071)

2025-03-04 13:01:24 -08:00

test_mla.py

Update nextn ci test (#4071)

2025-03-04 13:01:24 -08:00

test_modelopt_fp8kvcache.py

Add support for nvidia modelopt fp8 kv cache (#3223)

2025-02-22 07:04:58 +08:00

test_models_from_modelscope.py

Improve process creation (#1534)

2024-09-29 02:36:12 -07:00

test_moe_ep.py

Split the __init__ of scheduler as smaller functions. Improve the eagle tests (#4128)

2025-03-06 00:13:20 -08:00

test_moe_eval_accuracy_large.py

Misc clean up; Remove the support of jump forward (#4032)

2025-03-03 07:02:14 -08:00

test_nightly_gsm8k_eval.py

Update thresholds in test_nightly_gsm8k_eval.py (#3176)

2025-01-27 03:02:09 -08:00

test_nightly_human_eval.py

Separate two entry points: Engine and HTTP server (#2996)

2025-01-19 22:09:24 -08:00

test_nightly_math_eval.py

[feat]: Add math eval to CI nightly run (#2663)

2025-01-01 15:29:35 -08:00

test_no_chunked_prefill.py

[CI] Split test cases in CI for better load balancing (#2180)

2024-11-25 04:58:16 -08:00

test_no_overlap_scheduler.py

[CI] Split test cases in CI for better load balancing (#2180)

2024-11-25 04:58:16 -08:00

test_openai_server.py

Feature/function calling update (#2700)

2025-01-26 09:57:51 -08:00

test_penalty.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

test_pytorch_sampling_backend.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_radix_attention.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_reasoning_content.py

Reasoning parser (#4000)

2025-03-03 21:16:36 -08:00

test_regex_constrained.py

Misc clean up; Remove the support of jump forward (#4032)

2025-03-03 07:02:14 -08:00

test_release_memory_occupation.py

CUDA-graph-compatible releasing and resuming KV cache and model weight memory (#2630)

2025-01-13 11:38:51 -08:00

test_request_length_validation.py

[Bugfix] Fix tokenizer_manager not getting 400 when req is too long (#3678)

2025-02-27 22:59:43 -08:00

test_retract_decode.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_sagemaker_server.py

feat: Add SageMaker support (#3740)

2025-02-21 19:31:09 +08:00

test_schedule_policy.py

Refactor SchedulePolicy to improve code organization (#2571)

2025-01-04 00:05:16 +08:00

test_server_args.py

Update vllm to 0.6.3 (#1711) (#1720)

2024-10-19 20:45:41 -07:00

test_session_control.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

test_skip_tokenizer_init.py

example: add vlm to token in & out example (#3941)

2025-03-04 22:18:26 -08:00

test_srt_endpoint.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

test_srt_engine_with_quant_args.py

[unittest] add unit test to test quant args of srt engine (#2574)

2024-12-26 06:54:43 -08:00

test_srt_engine.py

Separate two entry points: Engine and HTTP server (#2996)

2025-01-19 22:09:24 -08:00

test_torch_compile_moe.py

Improve torch compile for fused moe (#2327)

2024-12-03 01:58:25 -08:00

test_torch_compile.py

Add accuracy and latency tests of eagle into CI (#3027)

2025-01-21 02:55:14 -08:00

test_torch_native_attention_backend.py

Add a simple torch native attention backend (#2241)

2024-12-01 03:01:25 -08:00

test_torch_tp.py

Rename sglang.bench_latency to sglang.bench_one_batch (#2118)

2024-11-21 20:07:48 -08:00

test_torchao.py

Crash the server correctly during error (#2231)

2024-11-28 00:22:39 -08:00

test_triton_attention_backend.py

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -08:00

test_triton_attention_kernels.py

Support Eagle2 for Triton backend (#3466)

2025-02-10 20:00:42 +08:00

test_triton_attention_rocm_mla.py

[ROCm] Enable Fused MLA Triton kernel for DeepSeekV3 (#3237)

2025-02-24 18:14:31 -08:00

test_update_weights_from_disk.py

Add update_weights_from_disk endpoint to Engine (#4102)

2025-03-05 12:25:18 -08:00

test_update_weights_from_distributed.py

Add update_weights_from_disk endpoint to Engine (#4102)

2025-03-05 12:25:18 -08:00

test_update_weights_from_tensor.py

[Feature] SPMD for SGLang + Verl (#3852)

2025-02-28 09:53:10 -08:00

test_verl_engine.py

Support penalty in overlap mode; return logprob with chunked prefill; improve benchmark scripts (#3988)

2025-03-03 00:12:04 -08:00

test_vertex_endpoint.py

Simplify eagle tests and TP sync in grammar backend (#4066)

2025-03-04 13:40:40 -08:00

test_vision_chunked_prefill.py

Fix CI and install docs (#3821)

2025-02-24 16:17:38 -08:00

test_vision_llm.py

[Fix] Address remaining issues of supporting MiniCPMV (#2977)

2025-01-28 00:22:13 -08:00

test_vision_openai_server.py

Fix CI and install docs (#3821)

2025-02-24 16:17:38 -08:00

test_w8a8_quantization.py

Add ut for w8a8 int8 quantization (#2897)

2025-01-15 18:29:14 +08:00