repositories/exllama/sh/test_benchmark_perf.sh

echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/_test_models/iambestfeed_open_llama_3b_4bit_128g -cs
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/llama-7b-4bit-128g -cs
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/llama-13b-4bit-128g -cs
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/llama-30b-4bit-128g
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/llama-30b-4bit-128g-act
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/llama-30b-4bit-32g-act-ts -l 1550
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/koala-13B-4bit-128g-act
echo "-------------------------------------------------------------------------------------------------------------"
python test_benchmark_inference.py -p -d /mnt/str/models/wizardlm-30b-uncensored-4bit-act-order
echo "-------------------------------------------------------------------------------------------------------------"