vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Readd variable attn_k, attn_q, attn_o after merge

Browse source
This commit is contained in:
Nexesenex 2024-08-22 19:12:25 +02:00
parent fb2b9ea667
commit 596a4aec86
1 changed files with 3 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

3
src/llama.cpp
View file

@ -17328,6 +17328,9 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
n_attn_layer *= 3;
}
GGML_ASSERT((qs.n_attention_wv == n_attn_layer) && "n_attention_wv is unexpected");
GGML_ASSERT((qs.n_attention_wk == n_attn_layer) && "n_attention_wk is unexpected");
GGML_ASSERT((qs.n_attention_wq == n_attn_layer) && "n_attention_wq is unexpected");
GGML_ASSERT((qs.n_attention_wo == n_attn_layer) && "n_attention_wo is unexpected");
}
size_t total_size_org = 0;