Skip to content

LLM inference in C/C++

C++ ggml

Add cohere2moe to llama-vocab for TINY_AYA (#24601)

Bartowski8ed274ef46e2
diff --git a/src/llama-vocab.cpp b/src/llama-vocab.cpp
index 8543e178db..6e78a3f6c0 100644
--- a/src/llama-vocab.cpp
+++ b/src/llama-vocab.cpp
@@ -2280,7 +2280,8 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {
clean_spaces = false;
ignore_merges = true;
} else if (
- tokenizer_pre == "tiny_aya") {
+ tokenizer_pre == "tiny_aya" ||
+ tokenizer_pre == "cohere2moe") {
pre_type = LLAMA_VOCAB_PRE_TYPE_TINY_AYA;
clean_spaces = false;
} else if (