From 71a742256ca032e4b26440402fa64fda0307ace5 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Stanis=C5=82aw=20Szymczyk?= Date: Sun, 19 May 2024 12:22:54 +0200 Subject: [PATCH] Temporarily hard-coded mscale value for DeepSeek-V2 (FIXME!). --- ggml.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/ggml.c b/ggml.c index 55152bce4..2618edf59 100644 --- a/ggml.c +++ b/ggml.c @@ -14073,6 +14073,8 @@ static void rope_yarn( // Get n-d magnitude scaling corrected for interpolation mscale *= 1.0f + 0.1f * logf(1.0f / freq_scale); } + // TODO ugly hack for DeepSeek-V2 until we find a solution + mscale = 1.0; *cos_theta = cosf(theta) * mscale; *sin_theta = sinf(theta) * mscale; }