Skip to content

LLM inference in C/C++

C++ ggml

model : Add label for LFM2.5-230M (#25008)

Tarek Dakhran9d5d882d8cd0
diff --git a/src/llama-model.cpp b/src/llama-model.cpp
index d041a9ce3e..6cb0ec3791 100644
--- a/src/llama-model.cpp
+++ b/src/llama-model.cpp
@@ -700,6 +700,7 @@ const char * llm_type_name(llm_type type) {
case LLM_TYPE_160M: return "160M";
case LLM_TYPE_190M: return "190M";
case LLM_TYPE_220M: return "220M";
+ case LLM_TYPE_230M: return "230M";
case LLM_TYPE_250M: return "250M";
case LLM_TYPE_256M: return "256M";
case LLM_TYPE_270M: return "270M";
diff --git a/src/llama-model.h b/src/llama-model.h
index f4718f6d58..77d8d3b625 100644
--- a/src/llama-model.h
+++ b/src/llama-model.h
@@ -36,6 +36,7 @@ enum llm_type {
LLM_TYPE_160M,
LLM_TYPE_190M,
LLM_TYPE_220M,
+ LLM_TYPE_230M,
LLM_TYPE_250M,
LLM_TYPE_256M,
LLM_TYPE_270M,
diff --git a/src/models/lfm2.cpp b/src/models/lfm2.cpp
index 07b7346ee4..70e837d6eb 100644
--- a/src/models/lfm2.cpp
+++ b/src/models/lfm2.cpp
@@ -13,6 +13,7 @@ void llama_model_lfm2::load_arch_hparams(llama_model_loader & ml) {
hparams.n_layer_dense_lead = hparams.n_layer();
switch (hparams.n_ff()) {
+ case 2560: type = LLM_TYPE_230M; break;
case 4608: type = LLM_TYPE_350M; break;
case 6912: type = LLM_TYPE_700M; break;
case 8192: type = LLM_TYPE_1_2B; break;