ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-05-01 20:01:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix double print

Browse Source
This commit is contained in:
Saood Karim
2025-05-30 02:06:29 -05:00
parent 31c20e2c2d
commit ae3816e13d
1 changed files with 14 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files

40
src/llama.cpp
View File

@@ -20699,42 +20699,22 @@ struct llama_context * llama_new_context_with_model(
}
if (memory_size_k + memory_size_v > 0) {
if (cparams.mla_attn != 0 && !cparams.flash_attn) {
LLAMA_LOG_INFO("%s: KV self size = %7.2f MiB, c^KV (%s): %7.2f MiB, kv^T (%s): %7.2f MiB\n", __func__,
(float)(memory_size_k + memory_size_v) / (1024.0f * 1024.0f),
ggml_type_name(type_k), (float)memory_size_k / (1024.0f * 1024.0f),
ggml_type_name(type_v), (float)memory_size_v / (1024.0f * 1024.0f));
} else if (cparams.mla_attn != 0 && cparams.flash_attn) {
LLAMA_LOG_INFO("%s: KV self size = %7.2f MiB, c^KV (%s): %7.2f MiB, kv^T: not used\n", __func__,
(float)(memory_size_k + memory_size_v) / (1024.0f * 1024.0f),
ggml_type_name(type_k), (float)memory_size_k / (1024.0f * 1024.0f));
} else {
LLAMA_LOG_INFO("%s: KV self size = %7.2f MiB, K (%s): %7.2f MiB, V (%s): %7.2f MiB\n", __func__,
(float)(memory_size_k + memory_size_v) / (1024.0f * 1024.0f),
ggml_type_name(type_k), (float)memory_size_k / (1024.0f * 1024.0f),
ggml_type_name(type_v), (float)memory_size_v / (1024.0f * 1024.0f));
}
}
{
size_t memory_size_kv = 0;
size_t memory_size_kvt = 0;
ggml_type kv_type = GGML_TYPE_COUNT;
ggml_type kvt_type = GGML_TYPE_COUNT;
for (auto & kv : ctx->kv_self.k_l) {
memory_size_kv += ggml_nbytes(kv);
kv_type = kv->type;
}
for (auto & kvt : ctx->kv_self.v_l) {
memory_size_kvt += ggml_nbytes(kvt);
kvt_type = kvt->type;
}
if (memory_size_kv + memory_size_kvt > 0) {
if (cparams.mla_attn == 1 && !cparams.flash_attn) {
LLAMA_LOG_INFO("%s: KV self size = %7.2f MiB, c^KV (%s): %7.2f MiB, kv^T (%s): %7.2f MiB\n", __func__,
(float)(memory_size_kv + memory_size_kvt) / (1024.0f * 1024.0f),
ggml_type_name(kv_type), (float)memory_size_kv / (1024.0f * 1024.0f),
ggml_type_name(kvt_type), (float)memory_size_kvt / (1024.0f * 1024.0f));
} else {
LLAMA_LOG_INFO("%s: KV self size = %7.2f MiB, c^KV (%s): %7.2f MiB, kv^T: not used\n", __func__,
(float)(memory_size_kv + memory_size_kvt) / (1024.0f * 1024.0f),
ggml_type_name(kv_type), (float)memory_size_kv / (1024.0f * 1024.0f));
}
}
}
// graph outputs buffer