llama : fix KV shift for qwen2vl (#13870)
* llama : fix KV shift for qwen2vl * add ref to the PR
This commit is contained in:
1 parent
10961339b2
commit
763d06edb7
2 files changed
+11
-3
No files matched your search
+1
-1
@@ -455,7 +455,7 @@ llm_graph_context::llm_graph_context(const llm_graph_params & params) :
|
||||
}
|
||||
|
||||
int64_t llm_graph_context::n_pos_per_embd() const {
|
||||
return arch == LLM_ARCH_QWEN2VL ? 4 : 1;
|
||||
return hparams.rope_type == LLAMA_ROPE_TYPE_MROPE ? 4 : 1;
|
||||
}
|
||||
|
||||
void llm_graph_context::cb(ggml_tensor * cur, const char * name, int il) const {
|
||||
|
||||
Reference in new issue
Block a user