From 4f237d44f6d75afbb5cef39d4d6b0b35b2a517c7 Mon Sep 17 00:00:00 2001 From: Iwan Kawrakow Date: Tue, 30 Jul 2024 16:11:25 +0300 Subject: iq3_k: Basics Quantize/dequantize, CUDA dequantize. PPL of LLaMA-3.1-8B is better than iq3_s and iq3_m. --- examples/quantize/quantize.cpp | 1 + 1 file changed, 1 insertion(+) (limited to 'examples/quantize') diff --git a/examples/quantize/quantize.cpp b/examples/quantize/quantize.cpp index 17e87e53..0b4c3444 100644 --- a/examples/quantize/quantize.cpp +++ b/examples/quantize/quantize.cpp @@ -41,6 +41,7 @@ static const std::vector QUANT_OPTIONS = { { "IQ4_NL", LLAMA_FTYPE_MOSTLY_IQ4_NL, " 4.50 bpw non-linear quantization", }, { "IQ4_XS", LLAMA_FTYPE_MOSTLY_IQ4_XS, " 4.25 bpw non-linear quantization", }, { "IQ2_K", LLAMA_FTYPE_MOSTLY_IQ2_K, " 2.375 bpw non-linear quantization",}, + { "IQ3_K", LLAMA_FTYPE_MOSTLY_IQ3_K, " 3.44 bpw non-linear quantization", }, { "IQ4_K", LLAMA_FTYPE_MOSTLY_IQ4_K, " 4.5 bpw non-linear quantization", }, { "IQ5_K", LLAMA_FTYPE_MOSTLY_IQ5_K, " 5.5 bpw non-linear quantization", }, { "Q4_K", LLAMA_FTYPE_MOSTLY_Q4_K_M, "alias for Q4_K_M", }, -- cgit v1.2.3