vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

llama : print DeekSeek-V2-specific parameters in llm_load_print_meta()

Browse source
This commit is contained in:
Stanisław Szymczyk 2024-05-27 15:30:17 +02:00
parent 5cc7ec161c
commit d02130d549
1 changed files with 10 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

10
llama.cpp
View file

@ -4989,6 +4989,16 @@ static void llm_load_print_meta(llama_model_loader & ml, llama_model & model) {
if (vocab.special_suffix_id != -1) { LLAMA_LOG_INFO( "%s: SUF token = %d '%s'\n", __func__, vocab.special_suffix_id, vocab.id_to_token[vocab.special_suffix_id].text.c_str() ); }
if (vocab.special_middle_id != -1) { LLAMA_LOG_INFO( "%s: MID token = %d '%s'\n", __func__, vocab.special_middle_id, vocab.id_to_token[vocab.special_middle_id].text.c_str() ); }
if (vocab.special_eot_id != -1) { LLAMA_LOG_INFO( "%s: EOT token = %d '%s'\n", __func__, vocab.special_eot_id, vocab.id_to_token[vocab.special_eot_id].text.c_str() ); }
if (model.arch == LLM_ARCH_DEEPSEEK2) {
LLAMA_LOG_INFO("%s: n_layer_dense_lead = %d\n", __func__, hparams.n_layer_dense_lead);
LLAMA_LOG_INFO("%s: n_lora_q = %d\n", __func__, hparams.n_lora_q);
LLAMA_LOG_INFO("%s: n_lora_kv = %d\n", __func__, hparams.n_lora_kv);
LLAMA_LOG_INFO("%s: n_ff_exp = %d\n", __func__, hparams.n_ff_exp);
LLAMA_LOG_INFO("%s: n_expert_shared = %d\n", __func__, hparams.n_expert_shared);
LLAMA_LOG_INFO("%s: expert_weights_scale = %.1f\n", __func__, hparams.expert_weights_scale);
LLAMA_LOG_INFO("%s: rope_yarn_log_mul = %.4f\n", __func__, hparams.rope_yarn_log_mul);
}
}
// Returns false if cancelled by progress_callback