CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-05 09:36:52 +00:00
Code Issues Packages Projects Releases Wiki Activity

mtmd: add --image-min/max-tokens (#16921)

Browse Source
This commit is contained in:
Xuan-Son Nguyen
2025-11-03 11:11:18 +01:00
committed by GitHub
parent bf7b0c9725
commit 070ff4d535
8 changed files with 79 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
tools/server/server.cpp
View File

@@ -2452,11 +2452,13 @@ struct server_context {
std::string & mmproj_path = params_base.mmproj.path;
if (!mmproj_path.empty()) {
mtmd_context_params mparams = mtmd_context_params_default();
mparams.use_gpu = params_base.mmproj_use_gpu;
mparams.print_timings = false;
mparams.n_threads = params_base.cpuparams.n_threads;
mparams.verbosity = params_base.verbosity > 0 ? GGML_LOG_LEVEL_DEBUG : GGML_LOG_LEVEL_INFO;
mparams.flash_attn_type = params_base.flash_attn_type;
mparams.use_gpu = params_base.mmproj_use_gpu;
mparams.print_timings = false;
mparams.n_threads = params_base.cpuparams.n_threads;
mparams.verbosity = params_base.verbosity > 0 ? GGML_LOG_LEVEL_DEBUG : GGML_LOG_LEVEL_INFO;
mparams.flash_attn_type = params_base.flash_attn_type;
mparams.image_min_tokens = params_base.image_min_tokens;
mparams.image_max_tokens = params_base.image_max_tokens;
mctx = mtmd_init_from_file(mmproj_path.c_str(), model, mparams);
if (mctx == nullptr) {
SRV_ERR("failed to load multimodal model, '%s'\n", mmproj_path.c_str());