mirror of
https://github.com/csunny/DB-GPT.git
synced 2025-09-05 11:01:09 +00:00
fix(model): Fix benchmarks bugs
This commit is contained in:
@@ -11,7 +11,7 @@ parallel_nums=${3:-$default_parallel_nums}
|
||||
run_benchmark() {
|
||||
local model_name=$1
|
||||
local model_type=$2
|
||||
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}
|
||||
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"
|
||||
|
Reference in New Issue
Block a user