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

rename n_ctx to kv_size

Browse source
This commit is contained in:
Pierrick HYMBERT 2024-02-18 20:59:26 +01:00 • committed by Georgi Gerganov
parent ef96e8b1f7
commit 606873401c
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
48 changed files with 403 additions and 393 deletions
Download patch file Download diff file Expand all files Collapse all files

12
examples/batched/batched.cpp
View file

@ -78,7 +78,7 @@ int main(int argc, char ** argv) {
llama_context_params ctx_params = llama_context_default_params();
ctx_params.seed = 1234;
ctx_params.n_ctx = n_kv_req;
ctx_params.kv_size = n_kv_req;
ctx_params.n_batch = std::max(n_len, n_parallel);
ctx_params.n_threads = params.n_threads;
ctx_params.n_threads_batch = params.n_threads_batch == -1 ? params.n_threads : params.n_threads_batch;
@ -90,14 +90,14 @@ int main(int argc, char ** argv) {
return 1;
}
const int n_ctx = llama_n_ctx(ctx);
const int kv_size = llama_kv_size(ctx);
LOG_TEE("\n%s: n_len = %d, n_ctx = %d, n_batch = %u, n_parallel = %d, n_kv_req = %d\n", __func__, n_len, n_ctx, ctx_params.n_batch, n_parallel, n_kv_req);
LOG_TEE("\n%s: n_len = %d, kv_size = %d, n_batch = %u, n_parallel = %d, n_kv_req = %d\n", __func__, n_len, kv_size, ctx_params.n_batch, n_parallel, n_kv_req);
// make sure the KV cache is big enough to hold all the prompt and generated tokens
if (n_kv_req > n_ctx) {
LOG_TEE("%s: error: n_kv_req (%d) > n_ctx, the required KV cache size is not big enough\n", __func__, n_kv_req);
LOG_TEE("%s: either reduce n_parallel or increase n_ctx\n", __func__);
if (n_kv_req > kv_size) {
LOG_TEE("%s: error: n_kv_req (%d) > kv_size, the required KV cache size is not big enough\n", __func__, n_kv_req);
LOG_TEE("%s: either reduce n_parallel or increase kv_size\n", __func__);
return 1;
}