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

llama : code style corrections

Browse source
This commit is contained in:
Stanisław Szymczyk 2024-05-27 12:47:53 +02:00
parent f3b5e7d436
commit abef8b2634
1 changed files with 4 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

6
llama.cpp
View file

@ -4466,8 +4466,9 @@ static void llm_load_hparams(
bool is_lite = (hparams.n_layer == 27); bool is_lite = (hparams.n_layer == 27);
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps); ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
ml.get_key(LLM_KV_LEADING_DENSE_BLOCK_COUNT, hparams.n_leading_dense_layer); ml.get_key(LLM_KV_LEADING_DENSE_BLOCK_COUNT, hparams.n_leading_dense_layer);
if (!is_lite) if (!is_lite) {
ml.get_key(LLM_KV_ATTENTION_Q_LORA_RANK, hparams.n_lora_q); ml.get_key(LLM_KV_ATTENTION_Q_LORA_RANK, hparams.n_lora_q);
}
ml.get_key(LLM_KV_ATTENTION_KV_LORA_RANK, hparams.n_lora_kv); ml.get_key(LLM_KV_ATTENTION_KV_LORA_RANK, hparams.n_lora_kv);
ml.get_key(LLM_KV_EXPERT_FEED_FORWARD_LENGTH, hparams.n_expert_ff); ml.get_key(LLM_KV_EXPERT_FEED_FORWARD_LENGTH, hparams.n_expert_ff);
ml.get_key(LLM_KV_EXPERT_SHARED_COUNT, hparams.n_expert_shared); ml.get_key(LLM_KV_EXPERT_SHARED_COUNT, hparams.n_expert_shared);
@ -6330,8 +6331,9 @@ static bool llm_load_tensors(
auto & layer = model.layers[i]; auto & layer = model.layers[i];
layer.attn_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_NORM, "weight", i), {n_embd}); layer.attn_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_NORM, "weight", i), {n_embd});
if (!is_lite) if (!is_lite) {
layer.attn_q_a_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_Q_A_NORM, "weight", i), {q_lora_rank}); layer.attn_q_a_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_Q_A_NORM, "weight", i), {q_lora_rank});
}
layer.attn_kv_a_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_KV_A_NORM, "weight", i), {kv_lora_rank}); layer.attn_kv_a_norm = ml.create_tensor(ctx_layer, tn(LLM_TENSOR_ATTN_KV_A_NORM, "weight", i), {kv_lora_rank});
if (!is_lite) { if (!is_lite) {