happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

clean up

This commit is contained in:
Xuan Son Nguyen 2025-12-18 23:39:00 +01:00
parent d244fe1a3e
commit 1bd8d1870f
3 changed files with 10 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
tools/mtmd/clip-model.h
View File

@ -311,11 +311,11 @@ struct clip_model {
ggml_tensor * mm_boi = nullptr;
ggml_tensor * mm_eoi = nullptr;
// lfm2
// lfm2 audio
std::array<ggml_tensor *, 7> pre_encode_conv_X_w = {nullptr};
std::array<ggml_tensor *, 7> pre_encode_conv_X_b = {nullptr};
ggml_tensor * pre_encode_out_w = nullptr;
ggml_tensor * pre_encode_out_b = nullptr;
ggml_tensor * pre_encode_out_w = nullptr;
ggml_tensor * pre_encode_out_b = nullptr;
bool audio_has_avgpool() const {
return proj_type == PROJECTOR_TYPE_QWEN2A

2
tools/mtmd/mtmd-audio.cpp
View File

@ -565,7 +565,7 @@ bool mtmd_audio_preprocessor_conformer::preprocess(
params.hop_length = hparams.audio_hop_len;
params.sample_rate = hparams.audio_sample_rate;
params.center_padding = true;
params.preemph = 0.97f; // disabled
params.preemph = 0.97f;
params.use_natural_log = true;
params.norm_per_feature = true;

8
tools/mtmd/mtmd-cli.cpp
View File

@ -323,7 +323,9 @@ int main(int argc, char ** argv) {
LOG_WRN("WARN: This is an experimental CLI for testing multimodal capability.\n");
LOG_WRN(" For normal use cases, please use the standard llama-cli\n");
eval_system_prompt_if_present();
if (eval_system_prompt_if_present()) {
return 1;
}
if (is_single_turn) {
g_is_generating = true;
@ -383,8 +385,10 @@ int main(int argc, char ** argv) {
ctx.n_past = 0;
ctx.chat_history.clear();
llama_memory_clear(llama_get_memory(ctx.lctx), true);
if (eval_system_prompt_if_present()) {
return 1;
}
LOG("Chat history cleared\n\n");
eval_system_prompt_if_present();
continue;
}
g_is_generating = true;