wehub-resource-sync
d13100ebf3
Update draft releases / main (push) Has been cancelled
Build and push docs image / build-image (push) Has been cancelled
Build Web Application / build-web (macos-latest) (push) Has been cancelled
Build Web Application / build-web (ubuntu-latest) (push) Has been cancelled
Python Code Quality Checks / build (push) Has been cancelled
Test Python / test-python (macos-latest, 3.10) (push) Has been cancelled
Test Python / test-python (macos-latest, 3.11) (push) Has been cancelled
Test Python / test-python (ubuntu-latest, 3.10) (push) Has been cancelled
Test Python / test-python (ubuntu-latest, 3.11) (push) Has been cancelled
21 lines
702 B
Bash
Executable File
21 lines
702 B
Bash
Executable File
#!/bin/bash
|
|
|
|
default_input_lens="8,8,256,1024"
|
|
default_output_lens="256,512,1024,1024"
|
|
default_parallel_nums="1,2,4,16,32"
|
|
|
|
input_lens=${1:-$default_input_lens}
|
|
output_lens=${2:-$default_output_lens}
|
|
parallel_nums=${3:-$default_parallel_nums}
|
|
|
|
run_benchmark() {
|
|
local model_name=$1
|
|
local model_type=$2
|
|
DB_GPT_MODEL_BENCHMARK=true python pilot/utils/benchmarks/llm/llm_benchmarks.py --model_name ${model_name} --model_type ${model_type} --input_lens ${input_lens} --output_lens ${output_lens} --parallel_nums ${parallel_nums}
|
|
}
|
|
|
|
run_benchmark "vicuna-7b-v1.5" "huggingface"
|
|
run_benchmark "vicuna-7b-v1.5" "vllm"
|
|
run_benchmark "baichuan2-7b" "huggingface"
|
|
run_benchmark "baichuan2-7b" "vllm"
|