From d29f8d3d404279b081111135321b6833a3bf6441 Mon Sep 17 00:00:00 2001 From: Iwan Kawrakow Date: Thu, 6 Mar 2025 18:38:32 +0200 Subject: [PATCH] Add the --custom-q option to the help --- examples/quantize/quantize.cpp | 1 + 1 file changed, 1 insertion(+) diff --git a/examples/quantize/quantize.cpp b/examples/quantize/quantize.cpp index fac504df..89de794b 100644 --- a/examples/quantize/quantize.cpp +++ b/examples/quantize/quantize.cpp @@ -144,6 +144,7 @@ static void usage(const char * executable) { printf(" --exclude-weights tensor_name: use importance matrix for this/these tensor(s)\n"); printf(" --output-tensor-type ggml_type: use this ggml_type for the output.weight tensor.\n"); printf(" --token-embedding-type ggml_type: use this ggml_type for the token_embd.weight tensor.\n\n"); + printf(" --custom-q regex1=type1,regex2=type2...: use this to specify custom quantization type rules.\n\n"); printf("Additional specific tensor quantization types used in the custom quant scheme 'CQS (default is Q2_K):\n"); printf(" --attn-q-type ggml_type: use this ggml_type for the attn_q.weight tensor.\n"); printf(" --attn-k-type ggml_type: use this ggml_type for the attn_k.weight tensor.\n");