From e0adb8b1227dd4622a91a5b3680b4af2e36d32f4 Mon Sep 17 00:00:00 2001 From: Kawrakow Date: Wed, 11 Dec 2024 11:19:00 +0100 Subject: Q3_K_R4 (#134) * q3_k_r4: Zen4 works, but not as good as it should be 238 t/s, so sloghtly slower than q6_k_r4. * q3_k_r4: NEON We get PP-512(LLaMA-3.1-8B) = 106.9 t/s. This is 1.93X faster than q3_K_S! --------- Co-authored-by: Iwan Kawrakow --- 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 db0fc0d4..2c8b33c2 100644 --- a/examples/quantize/quantize.cpp +++ b/examples/quantize/quantize.cpp @@ -36,6 +36,7 @@ static const std::vector QUANT_OPTIONS = { { "IQ3_S", LLAMA_FTYPE_MOSTLY_IQ3_S, " 3.44 bpw quantization", }, { "IQ3_M", LLAMA_FTYPE_MOSTLY_IQ3_M, " 3.66 bpw quantization mix", }, { "Q3_K", LLAMA_FTYPE_MOSTLY_Q3_K_M, "alias for Q3_K_M" }, + { "Q3_K_R4", LLAMA_FTYPE_MOSTLY_Q3_K_R4, "Q3_K_S repacked" }, { "IQ3_XS", LLAMA_FTYPE_MOSTLY_IQ3_XS, " 3.3 bpw quantization" , }, { "Q3_K_S", LLAMA_FTYPE_MOSTLY_Q3_K_S, " 2.75G, +0.5551 ppl @ LLaMA-v1-7B", }, { "Q3_K_M", LLAMA_FTYPE_MOSTLY_Q3_K_M, " 3.07G, +0.2496 ppl @ LLaMA-v1-7B", }, -- cgit v1.2.3