llama : fix Gemma3 SWA KV cache shift (#12373)

* llama : fix Gemma3 SWA KV cache shift

ggml-ci

* hparams : add comment [no ci]
This commit is contained in:
Georgi Gerganov
2025-03-13 19:08:07 +02:00
committed by GitHub
parent be7c303410
commit 84d5475541
6 changed files with 37 additions and 43 deletions

View File

@@ -168,6 +168,8 @@ private:
ggml_tensor * cur,
ggml_tensor * shift,
ggml_tensor * factors,
float freq_base,
float freq_scale,
ggml_backend_buffer * bbuf) const;
llm_graph_result_ptr build_kv_self_shift(