diff --git a/quant_config.json b/quant_config.json new file mode 100644 index 0000000..4002036 --- /dev/null +++ b/quant_config.json @@ -0,0 +1,6 @@ +{ + "zero_point": true, + "q_group_size": 128, + "w_bit": 4, + "version": "GEMM" +} \ No newline at end of file