From 7f61b3068e18728e5e7e2b95546ff03dd2fd41ac Mon Sep 17 00:00:00 2001 From: Kawrakow Date: Thu, 6 Feb 2025 14:08:52 +0200 Subject: IQ1_M_R4: better 1.75 bpw quants (#187) * iq1_m_r4: basics (quantize/dequantize) * iq1_m_r4: Zen4 gemm * iq1_m_r4: neon gemm * iq1_m_r4: switch to q8_0_x4 also on AVX2/Zen4 With the deltas being per group of 8, we cannot make use of the q8 sums stored in q8_1, so we get a tiny gain by using q8_0_x4. * iq1_m_r4: rename mul_mat_iq1_m_r4_q8_1 to mul_mat_iq1_m_r4_q8_0 --------- Co-authored-by: Iwan Kawrakow --- include/llama.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/llama.h') diff --git a/include/llama.h b/include/llama.h index 0f6d15ac..3f25b296 100644 --- a/include/llama.h +++ b/include/llama.h @@ -197,6 +197,7 @@ extern "C" { LLAMA_FTYPE_MOSTLY_IQ3_S_R4 = 226, // except 1d tensors LLAMA_FTYPE_MOSTLY_IQ2_M_R4 = 229, // except 1d tensors LLAMA_FTYPE_MOSTLY_IQ4_XS_R4 = 230, // except 1d tensors + LLAMA_FTYPE_MOSTLY_IQ1_M_R4 = 231, // except 1d tensors LLAMA_FTYPE_MOSTLY_Q6_0_R4 = 335, // except 1d tensors LLAMA_FTYPE_MOSTLY_BF16_R16 = 232, // except 1d tensors LLAMA_FTYPE_MOSTLY_IQ2_BN_R4 = 337, // except 1d tensors -- cgit v1.2.3