mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-16 12:35:01 +08:00

History

[Kernel]: Cutlass 2:4 Sparsity + FP8/Int8 Quant Support (#10995 )

Co-authored-by: Faraz Shahsavan <faraz.shahsavan@gmail.com>
Co-authored-by: ilmarkov <markovilya197@gmail.com>
Co-authored-by: Rahul Tuli <rahul@neuralmagic.com>
Co-authored-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>

2024-12-18 09:57:16 -05:00

cutlass_benchmarks

[Kernel]: Cutlass 2:4 Sparsity + FP8/Int8 Quant Support (#10995 )

2024-12-18 09:57:16 -05:00

disagg_benchmarks

[Core] Implement disagg prefill by StatelessProcessGroup (#10502 )

2024-12-01 19:01:00 -06:00

fused_kernels

[torch.compile] Dynamic fp8 + rms_norm fusion (#10906 )

2024-12-13 03:19:23 +00:00

kernels

[Misc] Kernel Benchmark for RMSNorm (#11241 )

2024-12-17 06:57:02 +00:00

overheads

[Core] Deprecating block manager v1 and make block manager v2 default (#8704 )

2024-10-17 11:38:15 -05:00

structured_schemas

[Benchmark] Benchmark structured output with datasets (#10557 )

2024-12-03 17:21:06 -07:00

backend_request_func.py

Benchmark serving structured output (#10880 )

2024-12-04 16:28:21 -05:00

benchmark_guided.py

[Benchmark] Benchmark structured output with datasets (#10557 )

2024-12-03 17:21:06 -07:00

benchmark_latency.py

[Misc] Make benchmarks use EngineArgs (#9529 )

2024-10-22 15:40:38 -07:00

benchmark_prefix_caching.py

[misc] partial prefix & random input generation benchmark (#9929 )

2024-11-18 15:39:14 -08:00

benchmark_prioritization.py

[Misc] Make benchmarks use EngineArgs (#9529 )

2024-10-22 15:40:38 -07:00

benchmark_serving_guided.py

Benchmark serving structured output (#10880 )

2024-12-04 16:28:21 -05:00

benchmark_serving.py

[Misc] Add tokenizer_mode param to benchmark_serving.py (#11174 )

2024-12-13 16:19:10 +00:00

benchmark_throughput.py

[Bugfix] Prevent benchmark_throughput.py from using duplicated random prompts (#10753 )

2024-12-03 02:26:15 +00:00

launch_tgi_server.sh

[CI/Build] Add shell script linting using shellcheck (#7925 )

2024-11-07 18:17:29 +00:00

README.md

[Feature] Update benchmark_throughput.py to support image input (#9851 )

2024-11-05 19:30:02 +00:00

sonnet.txt

feat(benchmarks): Add Prefix Caching Benchmark to Serving Benchmark (#3277 )

2024-03-27 13:39:26 -07:00

README.md

Benchmarking vLLM

Downloading the ShareGPT dataset

You can download the dataset by running:

wget https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/ShareGPT_V3_unfiltered_cleaned_split.json

Downloading the ShareGPT4V dataset

The json file refers to several image datasets (coco, llava, etc.). The benchmark scripts will ignore a datapoint if the referred image is missing.

wget https://huggingface.co/datasets/Lin-Chen/ShareGPT4V/resolve/main/sharegpt4v_instruct_gpt4-vision_cap100k.json
mkdir coco -p
wget http://images.cocodataset.org/zips/train2017.zip -O coco/train2017.zip
unzip coco/train2017.zip -d coco/