From f1f4eb988fe5ee969100cd0d3782fd7460d13949 Mon Sep 17 00:00:00 2001 From: Kawrakow Date: Tue, 3 Dec 2024 14:48:26 +0100 Subject: Q6_0_R4 (#122) * Adding q6_0_r4 We get PP-512(LLaMA-3.1-8B) = 257 t/s on a Ryzen-7950X. * q6_0_r4: NEON We get PP-512(LLaMA-3.1-8B) = 95 t/s on M2-Max. In terms of ops, q6_0_r4 is identical to q5_0_r4 except for loading the high bits being vld1q_u8_x2 instead of vld1q_u8. It is strange that this can make a 5% difference in performance, especially considering that this is amortized (re-used) over 8 columns in the right matrix. Or am I running out of vector registers? * Fix AVX2 --------- 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 b638107f..a8b4a44e 100644 --- a/examples/quantize/quantize.cpp +++ b/examples/quantize/quantize.cpp @@ -43,6 +43,7 @@ static const std::vector QUANT_OPTIONS = { { "IQ4_NL_X4",LLAMA_FTYPE_MOSTLY_IQ4_NL_X4," 4.50 bpw non-linear quantization", }, { "Q4_0_R4", LLAMA_FTYPE_MOSTLY_Q4_0_R4, " 4.50 bpw quantization", }, { "Q5_0_R4", LLAMA_FTYPE_MOSTLY_Q5_0_R4, " 5.50 bpw quantization", }, + { "Q6_0_R4", LLAMA_FTYPE_MOSTLY_Q6_0_R4, " 6.50 bpw quantization", }, { "Q8_0_R4", LLAMA_FTYPE_MOSTLY_Q8_0_R4, " 8.50 bpw quantization", }, { "IQ4_XS", LLAMA_FTYPE_MOSTLY_IQ4_XS, " 4.25 bpw non-linear quantization", }, { "IQ4_KS", LLAMA_FTYPE_MOSTLY_IQ4_KS, " 4.25 bpw non-linear quantization", }, -- cgit v1.2.3