fix: fix issue when applying lora

This commit is contained in:
leejet 2024-08-25 22:39:39 +08:00
parent 5c561eab31
commit 8847114abf

View File

@ -82,6 +82,7 @@ struct LoraModel : public GGMLRunner {
zero_index = ggml_new_tensor_1d(compute_ctx, GGML_TYPE_I32, 1);
set_backend_tensor_data(zero_index, zero_index_vec.data());
ggml_build_forward_expand(gf, zero_index);
std::set<std::string> applied_lora_tensors;
for (auto it : model_tensors) {