From 6e08281e588bbba1a5d180290a94a43f167f3a1a Mon Sep 17 00:00:00 2001 From: Kerfuffle <44031344+KerfuffleV2@users.noreply.github.com> Date: Sun, 29 Oct 2023 11:31:40 -0600 Subject: Extend llama_kv_cache_seq_rm to allow matching any sequence (#3843) * Extend llama_kv_cache_seq_rm to allow matichng any sequence * Replace llama_kv_cache_tokens_rm with llama_kv_cache_clear Use llama_kv_cache_clear for cache clearing Change calls to llama_kv_cache_tokens_rm that want to delete by position to use llama_kv_cache_seq_rm functionality --- common/common.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'common/common.cpp') diff --git a/common/common.cpp b/common/common.cpp index f81f4d35..c187128d 100644 --- a/common/common.cpp +++ b/common/common.cpp @@ -889,7 +889,7 @@ std::tuple llama_init_from_gpt_par std::vector tmp = { llama_token_bos(model), llama_token_eos(model), }; llama_decode(lctx, llama_batch_get_one(tmp.data(), std::min(tmp.size(), (size_t) params.n_batch), 0, 0)); - llama_kv_cache_tokens_rm(lctx, -1, -1); + llama_kv_cache_clear(lctx); llama_reset_timings(lctx); } -- cgit v1.2.3