File tree 2 files changed +0
-4
lines changed
2 files changed +0
-4
lines changed Original file line number Diff line number Diff line change @@ -74,7 +74,6 @@ namespace ov {
74
74
llama_model* m_llama_model_ptr = nullptr ;
75
75
llama_context* m_llama_ctx = nullptr ;
76
76
std::shared_ptr<ov::Model> m_fake_model;
77
- size_t * num_tokens_processed_ptr = nullptr ; // TODO: (vshampor) find a better place for this kind of storage
78
77
79
78
std::vector<ov::Output<const ov::Node>> m_fake_inputs;
80
79
std::vector<ov::Output<const ov::Node>> m_fake_outputs;
Original file line number Diff line number Diff line change @@ -37,14 +37,11 @@ LlamaCppModel::~LlamaCppModel() {
37
37
llama_free (m_llama_ctx);
38
38
llama_free_model (m_llama_model_ptr);
39
39
llama_backend_free ();
40
- delete num_tokens_processed_ptr;
41
40
}
42
41
43
42
LlamaCppModel::LlamaCppModel (const std::string& gguf_fname, const std::shared_ptr<const IPlugin>& plugin)
44
43
: ICompiledModel(nullptr , plugin),
45
44
m_gguf_fname (gguf_fname) {
46
- num_tokens_processed_ptr = new size_t ; // TODO (vshampor): hack, remove
47
- *num_tokens_processed_ptr = 0 ;
48
45
OPENVINO_DEBUG << " llama_cpp_plugin: loading llama model directly from GGUF... " << std::endl;
49
46
llama_model_params mparams = llama_model_default_params ();
50
47
mparams.n_gpu_layers = 99 ;
You can’t perform that action at this time.
0 commit comments