mikhail/llama.cpp-mtp-turboquant
1
0
Fork 0
Code Issues Pull Requests Actions 28 Packages Projects Releases Wiki Activity

model, mtmd: fix gguf conversion for audio/vision mmproj (#21309)

Browse Source 
* fix gguf conversion for audio/vision mmproj

* fix test
This commit is contained in:
Xuan-Son Nguyen
2026-04-02 17:10:32 +02:00
committed by GitHub
parent 223373742b
commit 63f8fe0ef4
27 changed files with 1462 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/models/models.h
+11
View File
@@ -266,6 +266,17 @@ struct llm_build_gemma3n_iswa : public llm_graph_context {
ggml_tensor * altup_correct(ggml_tensor * predictions, ggml_tensor * activated, int il);
};
struct llm_build_gemma4_iswa : public llm_graph_context {
const llama_model & model;
const int64_t n_embd_per_layer;
llm_build_gemma4_iswa(const llama_model & model, const llm_graph_params & params);
ggml_tensor * view_2d_slice(ggml_tensor * x, int idx);
ggml_tensor * get_per_layer_inputs();
ggml_tensor * project_per_layer_inputs(ggml_tensor * inputs_embeds, ggml_tensor * inp_per_layer);
};
struct llm_build_gemma_embedding : public llm_graph_context {
llm_build_gemma_embedding(const llama_model & model, const llm_graph_params & params);
};