From e9c1cecb9d7d743d30b4a29ecd56a411437def0a Mon Sep 17 00:00:00 2001 From: xaedes Date: Tue, 7 Nov 2023 09:04:51 +0100 Subject: ggml : fix backward rope after YaRN (#3974) * fix backward process of rope rope backward process was broken after YaRN RoPE (#2268) implementation, due to missing changes in backward functions. the code for the backward process is nearly identically to the forward process: the only difference is the sign of the sin-values. to avoid future regressions remove the near-duplicate backward functions and reuse the forward code: for this a new function argument `bool forward` was added to `ggml_compute_forward_rope_f32` and `ggml_compute_forward_rope_f16`. the sin-values will be negated when forward is false. * fix finetune rope call to use correct default attn_factor of 1.0f * remove unused `ggml_rope_xpos_back` it is better to have only one `ggml_rope_back` function that accepts all rope parameters, so that `ggml_compute_backward` can propagate all parameters without having to switch between different rope_back variants. * fix comments explaining the sinus sign in ggml_forward_rope * add missing function arguments in declaration * fix function argument type in declaration --- ggml.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'ggml.h') diff --git a/ggml.h b/ggml.h index 70eb25a6..26654fc8 100644 --- a/ggml.h +++ b/ggml.h @@ -1372,8 +1372,13 @@ extern "C" { int n_dims, int mode, int n_ctx, + int n_orig_ctx, float freq_base, float freq_scale, + float ext_factor, + float attn_factor, + float beta_fast, + float beta_slow, float xpos_base, bool xpos_down); -- cgit v1.2.3