sglang/benchmark/latency_throughput

Benchmark Latency and Throughput

SGLang

Launch server

python -m sglang.launch_server --model-path meta-llama/Llama-2-7b-chat-hf --port 30000

Benchmark one batch

python3 bench_one.py
python3 bench_one.py --batch-size 64

Benchmark online serving with many requests

python3 bench_serving.py --backend srt --port 30000 --tokenizer meta-llama/Llama-2-7b-chat-hf --num-prompt 1000 --request-rate 100 --input-len 1024 --output-len 256

Benchmark online serving on the ShareGPT dataset

Download data

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

Run ShareGPT

python3 bench_throughput.py --backend srt --port 30000 --tokenizer meta-llama/Llama-2-7b-chat-hf --dataset ShareGPT_V3_unfiltered_cleaned_split.json --num-prompts 10 --request-rate 10

Other baselines

vLLM

python3 -m vllm.entrypoints.api_server --model meta-llama/Llama-2-7b-chat-hf --tensor-parallel 1 --disable-log-requests --swap-space 16 --port 21000

# run synthetic
python3 bench_throughput.py --backend vllm --port 30000 --tokenizer meta-llama/Llama-2-7b-chat-hf --num-prompt 1000 --request-rate 100 --input-len 1024 --output-len 256

# run ShareGPT
python3 bench_throughput.py --backend vllm --port 21000 --tokenizer meta-llama/Llama-2-7b-chat-hf --dataset ShareGPT_V3_unfiltered_cleaned_split.json --num-prompts 10 --request-rate 10

LightLLM

python -m lightllm.server.api_server --model_dir ~/model_weights/Llama-2-7b-chat-hf --max_total_token_num 15600 --tokenizer_mode auto --port 22000

python3 bench_throughput.py --backend lightllm --port 22000 --tokenizer meta-llama/Llama-2-7b-chat-hf --dataset ShareGPT_V3_unfiltered_cleaned_split.json --num-prompts 10 --request-rate 10