mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-11-02 09:12:03 +00:00
31 lines
566 B
Bash
Executable File
31 lines
566 B
Bash
Executable File
#!/usr/bin/env bash
|
|
|
|
qnt=(q8_0 q6_k q5_k q5_1 q5_0 q4_k q4_1 q4_0 q3_k q2_k)
|
|
args=""
|
|
|
|
if [ -z "$1" ]; then
|
|
echo "usage: $0 <model> [qnt] [args]"
|
|
echo "default: $0 <model> \"${qnt[@]}\" \"${args}\""
|
|
exit 1
|
|
fi
|
|
|
|
if [ ! -z "$2" ]; then
|
|
qnt=($2)
|
|
fi
|
|
|
|
if [ ! -z "$3" ]; then
|
|
args="$3"
|
|
fi
|
|
|
|
model="$1"
|
|
out="../tmp/results-${model}"
|
|
|
|
set -o pipefail
|
|
set -e
|
|
|
|
mkdir -p ${out}
|
|
|
|
for q in ${qnt[@]}; do
|
|
time ./bin/llama-quantize ../models/${model}/ggml-model-f16.gguf ../models/${model}/ggml-model-${q}.gguf ${q} 2>&1 ${args} | tee ${out}/qnt-${q}.txt
|
|
done
|